Home

Simulation - Agilent Technologies

1. 225 use previous solution as starting point 225 234 TOES Uo a E T S E A EEE 130 Uer TOCO eai 1 TATI E O dM 170 N DG 531 yendors ppled models eth 1 viaholes VIeWetrexample aues Fini diede ecerette 175 with Z directed ports si ce eerte tese 100 Ri e E 70 aera e ere nrny ri 97 Rit ica reer errr perrere ores Drereeerenrsnerererer sre 161 VIC W CT ion loiid 122 132 161 177 178 180 Virtual Tone Channel Power 524 D E E 526 AVRO 1a T TTT m 13 VONA G tine n EEEE 161 Votice D aeaa N 531 VPOR LE xe 236 S MM NM 292 RU M Om 11 NEC IPM 524 Simulation 564 W VPOR T c ec o INNEN 236 E ON DUM MM stet te UR e ne aT PIO 1 WAV ICHIGO CA ironin eana eana 973128 Weichted Norminis 225 MUbIteupo VINO OS oescene concor LLL LEER 2 White PM usse WERE RE Rea 271 RV IRE S50 DEC E EEEE N 170 write internal data files du e t ht eee 179 o e NT 184 Y Wilkes NM ID 14 PD MN HH 11 bouis 297 A ae ne H 297 ADT 11 PT EET PH 11 yX palaneet sais etm doi ote 6 11 184 292 p dri ERRANT TTE 11 Z Z directed ports 92 100 132 134 153 175 177 178 ZIN MT H 11 LAE EE 297 A uunc 297 V NTE ce icon e ee ae ticceceecceactacesocsacseccsasssqceacae 11 Li Oe 292 TE E a re EC ERWIN 11 LUSE ai E ene SENSN SN S IM 257 LS DATADYOEGES
2. L a Log Pointsi Decade O Linear Step Sige GHz Adaptive AF5 Max Points CO Lit of Frequencies GHz The default co simulation uses EM simulation frequencies Use EM simulation frequencies box is checked but the frequencies may be defined independently by a frequency band Start and Stop frequencies and number of points In this case the EM simulation Yraw parameters will be interpolated or extrapolated for all co simulation frequencies missing in the EM analysis simulation frequencies set by Fraw If a simulated circuit has nonlinear elements its behavior is dependent on the accuracy of the DC operating point The higher the lowest EM simulation frequency the larger the error of the EM data extrapolation to the DC frequency Be sure to include 0 or close to 0 EM analysis frequency for the nonlinear circuit even if your analysis frequency band is far away from the O frequency 372 Momentum GX After the Momentum GX analysis has calculated the co simulation design may be placed on the tree to be available for any linear or nonlinear simulations To create a design on tree right click the Momentum analysis and choose Create Cosimulation Design Works nes ce on i 7e EM HE Cosim E Analyses m T HB1 Momentuml Cos E HB1 Data 7 S Lineari Amplifier 0 t paza i Lineari Data 55 Momentum Sit Moment Ant riea Momentum Rename 1 Momentum Delete
3. a 11 0 212165 j0 640765 hb _Larges vin ar ALS21 1 6195 j2 81205 hb LargeS Vin 1 VPORT 2 2 0 SEIPPORT Hamme it VPORT HAW VPORT H zZPORT HB Data RALS12 0 270294 j0 395119 hb LargeS Vin2 vPORT 2 1 0 H7 52292 15183 0 969638 hb LargeS Vin2 VPCRT 2 2 1 RAPPORT 4f Time S VPORT HW VPORT S ZPORT The voltage amplitudes of input signals Vinl and Vin2 used in the equations are calculated from the input powers of the signal ports in the global equations block 34 Getting Started Large 5 input f 271e 005 z0 50 tidy fm le ed 3 Pin1 100e 6 Pindbm 7 10 f uPin2 1 amp 3 2 Pindbm 10 Pini dbmtow Pindbm fy Vin1 0 2 Vinl 2 aqrt 2 Pin1 z0 ay Vin2 0 632 Pal z0 50 Pin Pinl 10 Gain 10dBm Vin292 sgrt 2 PinZ2 z Note It s strictly not recommended to mix up the pre processed data input design or analysis parameters with post processed data in the same equations block because after deleting the post processed data from the workspace for example after recalculating the analysis its dataset is cleared and post processed data does not exist anymore It results in an error parsing the equation block after which all the input data used by the updated design or analysis will be not accessible more As a result it stops the workspace updating process For example the above global equation block includes only pre processed data and no post proce
4. arc resolution is 0 04 degrees arc reaolutian Is 45 degrees sa D 10 T ES results are r ox identical Ta a 40 ed D o TU To B o SY 2S Frequency GHz Arc Resolution Arc Resolution 45 0 01 Process size 6 18 MB 12 791 MB User time 12 min 20 min Elapsed time 15 min 24 min 361 Simulation 362 Rectangular cells 96 80 Triangular cells 108 200 Via cells 16 36 Unknown currents 337 453 Mesh Patterns and Simulation Time The mesh generator is relatively fast even for complex meshes But be aware that the complexity of the mesh can have a significant on effect simulation time In particular the number of cells in the mesh especially triangular cells and the number of unknown currents can affect simulation time Simulation times will be faster when the mesh consists mostly of rectangular cells A simple mesh is one consisting of a few rectangles of similar size A worst case complex mesh would consist of many dissimilar triangles You should try to minimize the number of triangular cells in the mesh The second variable number of unknown currents depends on the density or total number of cells in the mesh You should try to minimize the total number of cells in the mesh This reduces the number of unknown currents and thus reduces the size of the matrix to be solved during simulation The amount of time required to solve the matrix is N N N where N is the number of unknown currents
5. i 4 iiil gt Schematic Having established a topology with a broken feedback loop we perform a two port linear analysis on the oscillator to determine if oscillation is possible given the circuit and active device gain In analyzing our circuit at the desired output frequency two conditions must be met that are required but not necessarily sufficient to ensure oscillation At the desired output frequency the open circuit gain must be greater or equal to one 1 and the phase shift is a multiple of 27Un n 360 degrees where n is an integer including zero Using the tuning option in GENESYS allows for the exact selection of circuit elements to ensure a gain margin of 71 and phase shift of zero degrees at the desired frequency output Considering the fact that our closed loop requires the connection of both potts it is helpful to view the return loss at the ports of out netwotk to determine 1f additional matching structures would be necessary to aide in the maximum transfer of power This data is already available as a result of our linear two port simulation Ideally 11 S22 In 251 Simulation the case that the port impedances are divergent closing the loop might prevent sustained oscillation Response Sele Open Loop Response ibaa TH i Bap TL zlsjBue 2 1 dB 252 HARBEC Harmonic Balance Analysis Match 5 1 1 3 5 2 2 FE Cos Uy J S Havi
6. 42 Getting Started en Chapter 2 Advanced Modeling Kit Advanced Modeling Kit O verview The GENESYS Advanced Modeling Kit AMK consists of three main parts Approximately 12 additional nonlinear models for use in HARBEC and CAYENNE These models are ready to use and do not require knowledge of the Verilog A language A built in Verilog A compiler for creating your own nonlinear models Verilog A source code for some of the non proprietary nonlinear models contained in GENESYS These files allow you to make custom changes to existing nonlinear models For example you can make a new model identical to a built in transistor but with a change to the nonlinear capacitance equations Hardware description languages were developed as a means to provide varying levels of abstraction to designers Integrated circuits are too complex for an engineer to create by specifying the individual transistors and wires HDLs allow the performance to be described at a high level and simulation synthesis programs can the take the lancuage and generate the gate level description As behavior beyond the digital performance was added a mixed signal language was created to manage the interaction between digital and analog signals A subset of this Verilog A was defined Verilog A describes analog behavior only however it has functionality to interface to some digital behavior Most other Verilog A implementations are interpreted language
7. IE 418 EEE Pope MERE 170 a0 O82 50 287 5D 4B2 50 B82 50 882 51 0072 51 2827 51 482 51 682 51 882 52 082 Frequency MHz PSPNOISE The next figure shows the results of the oscillator phase noise simulation for all carriers from 1 to 9th harmonics it shows that phase noise results around 9th harmonic compating with fundamental is seen to be increased 20 log 9 or 19 1 dB as expected Phase noise of all Oscillator Harmonics 277 Simulation 278 HBOSCT HPM 1 es EJ Oscillator sSB Phase Noise for all Harmonics from 1 to 9 30 1 1000Hz a 110 16268 45 b 104 18dB E c 100 637 dB d 38 118dB B0 T e a5 146dB E D 84 514dB ER g 33 15996 Pb h 81 83dB k Si iy 80 5506 40 105 120 135 150 165 CD 2 CX zt LL z LL z LL lt LL a uy d e d a es tL tL 100 1000 10000 100000 1 6 Frequency Hz HPM2 NPMS NPA Er MPM APM NPI MEMB NPM Below 1s the oscillator output power spectrum with noise sidebands calculated for all carriers in for noise frequency sweep from 1Hz to 1MHz Oscillator Output Power Spectrum with all noise sidebands absolute noise power spectrum HARBEC Harmonic Balance Analysis HBOSCI PSPHOISE 1 E L 25 Lu uy C Lu Ur CL 101 82dBm BIS mE EEZE 2 143 23037 27644 322 52 38858 414 66 460 74 Frequency MHz PSPNOISE Basics of Ph
8. low and upper frequencies of band where the small the signal oscillation criteria and oscillation frequency are calculated Number of Points number of points in the frequency band 2 Nonlinear oscillator solver Use Osc Solver checkbox enabling using nonlinear HB oscillator solver If it s disabled then the analysis calculates the only small signal oscillation frequency may be useful for fast VCO developing when nonlinear solver uses too much time to converge Osc Port Initial Voltage an initial amplitude of OSCPORT voltage probe If it set 0 then the solver uses internal default initial guess Absolute Current Tolerance minimal absolute value of amplitude of OSCPORT current convergence criteria The criteria is checked the only after the solver has found the OSCPORT probe voltage amplitude band where the real value of the current changes sign from negative to positive Number of Curve iteration maximal number of the tracing OSCPORT voltage probe amplitude Vprobe iterations Max step of Curve tracing maximal curve tracing step of voltage probe amplitude Vprobe Number of Adjust iterations maximal number of final iterations searching the minimal absolute value of real part of the OSCPORT current Iprobe Vprobe in band Vprobe in Vproben 1 Vproben where the function Re Iprobe Vprobe changes sion from negative to positive Vproben 1 lt Vproben
9. 357 Simulation 358 Using the wavelength of the frequency a linear function is approximated also referred to as a rooftop basis function The higher the frequency the more wavelengths fit across the structure The cells wavelength is the minimum number of cells that fit under each wavelength The mote cells the better the sinusoid is represented and the more accurate the simulation is For example if you use 30 cells per wavelength the maximum deviation between the sinusoid and the linear approximation is about 1 These parameters affect longitudinal current Higher frequencies will result in a greater number of cells increased density for a mesh Similarly increasing the minimum numbet of cells per wavelength will also increase the density In general for optimal density it is better to increase the number of cells per wavelength rather than increasing the mesh frequency The optimal value for the mesh frequency is the highest frequency that will be simulated This may avoid having to recalculate the substrate frequency band if it is not sufficient You should be aware that the value entered for the number of cells per wavelength is only a lower limit and not the exact value used by the mesh generator The following discussion explains how the actual number of cells used by the mesh generator may be greater For example suppose you enter 20 for Cells per Wavelength In general the cell size used when the geometry is infinitely
10. Analysis to Sweep He Parameter to sweep tone HE input RFpower Quitput Dataset Parameter Range Type Of Sweep Start 30 C Linger Number of Points 25 Stoo 10 C Log Points Decade 25 Linear Step Sige 5 Unit of Measure one C ust y Gear List Show Long Parameter Names v Propagate Al Variables x When Sweeping ar only 1 analysis variables Notice that after running the sweep the variables have been propagated into the sweep data SweepPowerl Data De 1 tone HB input RFpower A 1 tone HB input RFpower Lm 1 tane HB input RFpower Su Freq 7 661 7 657 7 643 7 597 fa FregiD FA PregindexIM Fe cain fai Gainv FAPI P2 Pout 1 HEIPPORT Sy Time iif VPORT i W_VPORT ig ZPORT 5 786 1 842 2511 amp 758 A a ta fa c e Jun 7 3 pF These variables can now be plotted vs the swept variable 242 HARBEC Harmonic Balance Analysis Gain P amp Compression P Properties Ca The 2nd technique The sweep of gain may be calculated directly from the spectral data of PowerSweep1_Data The frequency index of n th harmonic in the 1 tone HB analysis dataset is alyays equal to the harmonic number f Gain db10 P2 2 P1 2 or using operator lt array gt lt value gt which ret
11. 2110 2120 2130 2140 2150 2160 2170 e NodeNames is a variable placed in the path dataset which is the dependent variable for every path measurement Note that the NodeNames variable is really just an array of strings The left most column is really just a row number like a spreadsheet or the zth entry in the array For example NodeNames 3 MixIn BE System1_Data_Path1_ m EJ Variable NodeNames IP1DB ipl dbi OCP SOR In MOS mdsf CNP CNF 1 MishlatchLoss 1 immilf C NOCR E BENY NodeNames GOPTDBE opldbi OCP SOF Partha Parts e The operator returns the index in the array of the operand For example the index of NodeNames MixIn is 3 443 Simulation e nsummary then the measurment CNF NodeNames Out will return all the swept values of CNF at the node named Out e By default the names of the nodes used in the schematic are numbers so the syntax of CNF NodeNames 2 where 2 is the output node is very common e Another way to understand this is to examine the CNF measurement in the swept path dataset The dependent data is in the right most column titled CNF This variable has two independent variables 1 the swept frequency and 2 the name of the node The units of this measurement are show in the header row of the first column Or in other words dB The left most column is simply a row ot index number For example CNF 12 of the swept path dataset would be 0 517 dB which is the 12
12. Calculate port wave data The simulation will calculate time domain waveforms for all circuit ports voltage test points and current probes Save solution for all nodes Checking this box will cause the harmonic balance simulation save in output dataset solution for all circuit nodes and branches Calculate Now Dismisses the dialog box and starts the simulator Any model caches are removed 229 Simulation OK Dismisses the dialog box If automatic recalculation is on and a simulation is needed the simulator will run after the box is dismissed Cancel Dismiss the dialog box canceling any changes made Noise Tab Harmonic Balance Analysis Options General Calculate Moise Advanced Noisy Pork Input Port WS1 Partd Gutpuk Pork Port w F Noise Frequency Bandwidth Moise Frequency Sweep Sweep Type Single point Ww Noise Frequency Improve accuracy of Low Frequency Oscillator Phase Noise Polishing Dutput Calculate 556 Noise of C all carriers Carrier wth Index vector Calculate Noise Contributors An pl See Nonlinear Noise Overview Advanced Tab defines advanced parameters that control Harbec convergence methods 230 HARBEC Harmonic Balance Analysis Harmonic Balance Analysis Options General Calculate Noise Advanced Iterations Max Newton Iterations ra Max 1 D Subiterations c c Max Source A
13. E 3 Designs Properties E gt Amplifier i Export B amp Graphs peN EE i i HB1 PPOR Automatically Recalculate el HB Ww WP Cakulate Now k Momentum gt a oo Cached ies cS i un Momentum E AM Cosi T ir ry Desi ign rer Ellos import Momentum Mesh inba Layout Import Initial Pre mesh Polygons into Layout Copy Pashe This will create a new ze ih name Momentum Analysis name _cosim In this case it creates the co simulation design Momentum1_cosim This design may be used like any other Genesys design for any simulation engine For more information see the example located at Examples MomentumGX EM HB Cosim Momentum Theory of O peration Momentum is based on a numerical discretization technique called the method of moments This technique is used to solve Maxwell s electromagnetic equations for planar structures embedded in a multilayered dielectric substrate The stmulation modes available in Momentum microwave and RF are both based on this technique but use different variations of the same technology to achieve their results Momentum has two modes of operation microwave or full wave mode and the RF or quasi static mode The main difference between these two modes lies within the Green functions formulations that are used The full wave mode uses full wave Green functions these are general frequency dependent Green functions that fully characterize the substrate with
14. LogOutput fene Frequency 4 Hz Rectangular cells 134 Triangular cells 153 via cells MB Edge currents 425 Simulation Log Ready Some of the information returned includes e The number of rectangular cells in the structure e The number of triangular cells e The number of vertical rectangular cells used to model the vias e The number of unknown currents e Computer resources used and time to solve The number of cells and the number of unknown currents will affect simulation time The more cells the denser the mesh and the longer the simulation will be Each unknown current is a variable in the matrix equation that is solved by the simulator The greater the 356 Momentum GX number of unknown currents the larger the matrix and the longer the simulation time will be The memory required to solve the structure increases as well About the Mesh Generator The mesh generator provides the algorithm that divides the circuit geometry into cells rectangles and triangles When the Mesh Reduction option is selected the mesh is automatically reduced to remove slivery cells of low quality and to remove electromagnetically redundant cells resulting in a mesh with polygonal cells By discretizing the surface it is possible to calculate a linear approximation of the arbitrary varying surface current The linear approximation involves applying a rooftop function to the cells The better the linear approximation
15. O verview Overview 10 Getting Started GENESYS supports a rich set of output parameters All parameters can be used for any purpose including graphing tabular display optimization yield and post processing Linear Measurements The following table shows the available Measurements Where 7 and j are shown in the chart port numbers can be used to specify a port Some parameters such as Az use only one pott e g Al or VSWR2 Or on a tabular output the ports can be omitted ie S or Y and measurements for all ports will be given Tip All available measurements and their operators for a given circuit or sub circuit with their appropriate syntax are shown in the measurement wizard To bring up the measurement wizard select measurement wizard from the graph properties dialog box The fundamental measurements S CS Freq exist as arrays in the result datasets The short form shown here such as 12 is created by parsing a formula placed automatically into the dataset Note The section in this manual on S Parameters contains detailed information about many of these parameters Meas Sj Hj YPj ZPj ZINz YINz ZPORTZ VSWRz Nj Desctiption S Parameters H Parameters Y Parameters Z Parameters Impedance at port 7 with network terminations in place Admittance at port 7 with network terminations in place Reference Impedance at port 7 VSWR at pott 7 Voltage gain from port z to port
16. O verview This section gives a technical description of the basic EMPOWER algorithms Unlike most similar tools on the market EMPOWER is based on the method of lines MoL and comprises a set of numerical techniques designed to speed up calculations while increasing 206 EMPOWER Planar 3D EM Analysis accuracy of computations Incorporation of geometrical symmetries including rotational reduction of problem complexity using thinning out and linear re expansion procedures and multimode deembedding by the simultaneous diagonalization method are outlined here This theory section is for EMPOWER users familiar with numerical electromagnetics foundations We have added this material because MoL is less well known than the method of moments or the finite difference method MoL can be represented as a simple combination of both method of moments and finite difference method Thus we have skipped common parts and given our attention to the original parts of the algorithm More details on particular algorithm parts accuracy and convergence investigation results can be found in publications listed in the References section in the EMPOWER Engine Theory and Algorithms section Basically the theory behind the simulator can be reduced to the following An initial 3D problem in a layered medium is reduced to a 2D problem through a partial discretisation of the Maxwell s equations and its solution for a homogeneous layer in a grid spectral doma
17. Traditional Cascaded Noise Figure NFcascade F F2 1 G F5 1 G Go Fast Fn 1 G1G O 4 Note Traditional cascaded noise figure equations are not used The are very restrictive and suffer from the following conditions e Ignore effects of VSWR frequency and bandwidth e Assume noise contributions are from a single path e Ignore mixer image noise e Can ignore effects of gain compression 505 Simulation 506 The SPARCA Spectral Propagation and Root Cause Analysis technique used SPECTRASYS does not suffer from these restrictive assumptions Occasionally users are troubled if SPECTRASYS simulations give different answers than the traditional approach Using SPECTRASYS under the same assumptions as the traditional approach will always yield identical answers To use SPECTRASYS under the same assumptions only frequency independent blocks like attenuators and amplifiers must be used Even so amplifiers must be used as linear devices with infinite reverse isolation Filters cannot be used since their impedance varies with frequency Mixers cannot be used since noise from the image band will be converted into the mixer output Only 2 port devices must be used because of the single path assumption Of course the bandwidth must also be very narrow Channel Noise Power CNP This measurement is the integrated noise power in the main channel along the specified path For example if the Channel Measurement Bandwi
18. Values Complex value vetsus frequency Simulations Linear Default Format Table RECT Graph RE Smith Chart GM7 Examples Measurement Result in graph Smith chart Result on table optimization or yield YM real imaginary parts of admittance for all ports ZM1 re ZM1 ZM1 Maximum Available Gain GMAX The Maximum Available Gain measurement is a real function of frequency and is available for 2 port networks only For conditions where the stability factor IX is greater than zero i e the system is unconditionally stable then GMAX Sa Si2 IK sqrt K 1 If K 1 then GMAX is set to the maximum stable gain therefore GMAX Sai S12 Values Real value versus frequency Simulations Linear Default Format Table dB Graph dB Smith Chart none Available Gain amp Power Gain Circles GA GP An available gain input network circle is a locus of source impedances for a given gain below the optimum gain This locus which is specified via a marker is plotted on a Smith chart and is only available for 2 port networks The center of the circle is the point of maximum gain Circles are displayed for gains of 0 1 2 3 4 5 and 6 dB less than the 5 5 5 23 Simulation 24 optimal gain Similarly the power gain output network circle is a locus of load impedances for a given gain below the optimum gain If the stability factor K is less than unity then the 0 dB circle is at GMA
19. 1 Ptone Where IPn is the nth order intercept point n is the intercept order Pintermod n is the power level of the nth order intermod created by the two tones All power levels are measured in dBm This equation only applies for intermods created with two tones Intermods created with more than two tones have slightly different amplitudes and don t fit the above relationship directly However the measurement technique in the lab follows the same process except the frequency of the intermod will changed based on the order of the intermod For instance even order products are generally located at twice the frequency of the two tones ot closed to DC Intermod Path Measurement Basics Intermods are automatically created by all nonlinear behavioral models as long as intermod and harmonic calculation has been enabled 457 Simulation 458 Measuring intermods in Spectrasys is very similar to measuring intermods in the lab Cascaded intermod equations are NOT used in Spectrasys because of their serious limitations As such new measurements were created to removed these restrictions A model with either create generate intermods and conduct them from a prior stage or just conduct them from a prior stage All nonlinear behavioral models will both create intermods and conduct them from the input Linear models will only conduct intermods from the prior stage Intermod measurements can show both the generated conducted and tota
20. 20 30 1 0 2 2 2 2 datapoints plotpoints GAIN The first function contour generates the contours based on the parameters passed to the function See Built In Functions for a description of these functions and their parameters 29 Simulation 30 Contours E EEJ To create contours from a load pull data file 1 From Fz menu select the Import submenu followed by Load Pull Data File From this submenu select the type of file you are trying to import Focus or Maury format Select the file you are importing and click OK Add a Smith Chart to the workspace In the Smith Chart measurement properties dialog box click measurement wizard Select the load pull data set from the first dialog From the second measurement wizard box select Contours or Plotpoints from the first column Select the data to plot from the second column Click OK to close the measurement wizard box and click OK to close the Smith Chart properties box Getting Started You should then see contours like the ones in the Load Pull Contours Example shown above Nonlinear Port Power PPORT This power measurement array is the RMS power delivered at the port The port number is the column index into the array Values Real value in specified units Simulations Nonlinear dc analysis Default Format Table DBM Graph DBM Smith Chart none Short Form Pport Example P1 is equivalent to PPORTTI Examples Meas
21. Each frq The estimated calculation time per frequency in the current mode Estim RAM The estimated total memory required for the current simulation The fourth line displays the simulation time of the current frequency and symmetry plus the symmetty stage The fifth line displays the calculation stage The lines below the fifth line describe the calculated data for each frequency During line analysis the impedance Z and propagation constant G are displayed for each frequency In the discontinuity calculation mode the first row of the s matrix is displayed at each frequency Batch Runs Multiple workspaces can be loaded simultaneously and all EMPOWER simulations can be updated sequentially Simply open as many Workspace files as you need Select Options from the Tools menu and check Allow Multiple Open Workspaces Click the red Calculator button on the toolbar to run all of the analyses Note You should check Automatically save workspace after calc if you are running long or overnight batches so that if there is a power outage you will not lose your results EM PO WER O peration O verview An EMPOWER simulation requires a board layout description The easiest and recommended method is to use LAYOUT to create a graphical representation of the desired layout pattern The board can then be simulated by creating an EMPOWER Simulation This chapter describes how to use LAYOUT to construct a board layout an
22. Eine Methode zut Losung der Maxwellschen Gleichngen for Sechskomponentige Feleder auf Dikreter Basis Arch Electron Uebertragungstech v 31 N 3 1977 p 116 120 Computer aided design of microwave devices in Russian Edited by V V Nikol skii Moscow Radio 1 Sviaz 1982 R H Jansen The spectral domain approach for microwave integrated circuits IEEE Trans v MTT 33 1985 N 10 p 1043 1056 219 Simulation 220 S G Vesnin Eectromagnetic models for design of microstrip microwave structures in Russian Ph D Thesis MPEI Moscow 1985 E F Johnsom Technique Engineers the Cavity Resonance in Microstrip Housing Design MSN amp CT 1987 Feb p 100 102 107 109 J C Rautio R F Harrington An electromagnetic time harmonic analysis of shielded microstrip circuits IEEE Trans v MTT 35 1987 N 8 p 726 730 B V Sestrotetzkiy V Yu Kustov Electromagnetic analysis of multilevel integrated circuits on the base of RLC networks and informational multiport approach in Russian Voprosi Radioelektroniki ser OVR 1987 N 1 p 3 23 L P Dunleavy P B Katehi A generalized method for analyzing shielded thin microstrip discontinuities IEEE Trans v MTT 36 1988 N 12 p 1758 1766 T Uwaro T Itoh Spectral domain approach in Numerical techniques for microwave and millimeter wave passive structures Edited by T Itoh John Willey amp Sons 1989 R H Jansen Full wav
23. Lower Side n Channel Number any integer gt 0 For example ACFU1 if the first adjacent channel above that specified by the Channel Frequency If CF was 100 MHz and the channel bandwidth was 1 MHz then the main channel would be 99 5 to 100 5 MHz Consequently then ACFU1 would then be the channel 100 5 to 101 5 MHz and ACFL1 would be 98 5 to 99 5 MHz NOTE Only the first 2 adjacent channels on either side of the reference channel is listed in the Measurement Wizard However there is no restriction on the Adjacent Channel Numbet Channel or Path Frequency CF Since each spectrum can contain a large number of spectral components and frequencies Spectrasys must be able to determine the area of the spectrum to integrate for various measurements This integration area is defined by a Channel Frequency and a Channel Measurement Bandwidth which become the main channel for the specified path Spectrasys can automatically identify the desired Channel Frequency in an unambiguous case where only one frequency is on the from node of the designated path An error will appear if more than one frequency is available For this particular case the user must specify the intended frequency for this path in the System Simulation Dialog Box A Channel Frequency exists for each node along the specified path Consequently each node along the path will have the same Channel Frequency until a frequency translation e
24. Momentum GX there is no metal above the layers the charge density is nearly zero where there is no overlap and the change is very rapid nearly a step at the point of overlap aie A layer 1 dietectnc layer 2 cross section view of layers y area of overlap high charge density If no overlap extraction is performed the mesher can create cells that will cross the border of the overlap In the actual circuit this cell would have a partially high charge density area where the overlap is present and a partially low charge density area where there is no overlap However Momentum simulates cells with a constant charge density and approximating the strongly varying charge in the cell with a constant is not an adequate representation Thus by enabling thin layer extraction no cells will be created that cross an overlap region mesh call crosses into area of overlap top views of layer 1 with overlap extraction celis are resized there le no crossover into area of overlap Because enabling this feature for all possible overlap in a circuit would create too dense a mesh some empirical rules are implemented that switch the overlap extraction off if the layers are not close h gt 0 02 estimated cell size or if the overlap area is small To calculate the estimated cell size the mesh frequency the global number of cells per wavelength and an estimation of the effective refractive index of the substrate are
25. Text Files vs Binary Files There are two basic types of data files text sometimes called ASCII and binary Text files are human readable files They are universal and can be edited with many different programs such as NOTEPAD Among the text files used by EMPOWER are batch topology listing and S Parameter files Note Word processors can also edit text files however they will store binary formatting information in the file unless explicitly told not to Save as Text so we do not recommend their use for editing text files In contrast binary files are not human readable They contain information encoded into the numbers which make up the file which are ultimately turned into ones and zeros thus the name binary Unlike text files binary files are not universal and should only be edited by a program designed for the particular type of binary file you are using Editing a binary file in a regular word processor or text editor will undoubtedly destroy it Some 179 Simulation binary files used by EMPOWER and GENESYS are workspace line and Y Parameter files File Extensions You can normally tell the kind of file you have by looking at its extension the part of the name after the last period Some commonly used extensions include EXE executable TXT text and HLP help Each kind of file used by EMPOWER has its own unique extension These extensions are shown here Each of these types will be discussed indiv
26. The series lines and the stub dimensions were calculated using T LINE and were rounded to the nearest 5 mil increment The final line dimensions are shown below Input Line Output Line 200 mils gt 200 mils Open Stub 225 mils 25 mils Note Before beginning this example you should be sure your Workspace Window is visible Select Docking Windows NWorkspace Window from the View menu if necessary To begin select New from the GENESYS File menu Since we do not need a schematic for this circuit we will delete the schematic In the workspace window Right click on Sch1 Schematic and select Delete Next we will create a layout Click on the New Item button in the workspace window and select Designs Add Design with a Layout Enter Stub for the design name Note For all dialog boxes be sure that your entries are exactly like the ones shown in the figures Creating a Layout A board layout can be created one of two ways e By starting without a schematic e By starting from an existing schematic 109 Simulation 110 The first method starts in the GENESYS Environment by creating a layout without an associated schematic The layout is created by drawing lines and placing footprints in the LAYOUT editor The second method begins in with a schematic and creates a board layout based on the schematic objects This method is normally used when a linear simulation u
27. where Total Harm Order is the total harmonic order of the component of residual vector It increases weights of the higher order spectral components of solution spectrum and improves their accuracy Sometimes it helps also to improve convergence HARBEC Harmonic Balance Analysis Reduce internal nodes controls internal nodes amp branches reduction algorithm Default None Reduction of internal nodes amp branches deceases size of solution vector and significantly speeds up convergence Sometimes the reduction may significantly increase truncation error of Jacobian calculation which may result to convergence failure or dropping accuracy of the solution Always compare results calculated with and without nodes branches reduction to be sure that it s acceptable for the analyzed circuit None none reduction of internal nodes or branches Nodes Only reduce internal nodes only Nodes amp Branches reduce all internal nodes or branches HARBEC Popup Menu By right clicking on the HARBEC simulation icon on the workspace window the following menu appears ae Tiny Amp shee B J Designs URS fs Amp pM vel Compare E s t DCI B iij DC1 Data o HE Rename Sij fni iip Properties Em D Export f Tr i i Tre Examp Calculate Now E q25C5 Copy To ke Delete Automaticall Recalculate Copy Paste Rename Allows the name of the icon to be changed Delete Removes the icon and all of it
28. 213 Simulation 214 empower S B Worm R Pzegla 1984 K S Yee 1966 Also T Weiland 1977 B Sestroretzkiy 1977 G Kron 1944 Grid Green s Function The Grid Green s Function GGF has been mentioned quite a few times The GGF is a solution of the differential difference analogue of Maxwell s equations A 1 excited by a unit grid current Jx Jy or Jz The solution or response function is a discrete function in the xy plane and continuous inside layer along the z axis Actually to solve the formulated problem we need just a contraction of the GGF to the signal plane and to the regions with non zero z directed currents This contraction is a matrix due to the discretization EMPOWER Planar 3D EM Analysis To find the GGF matrix we used a spectral approach similar to one used in the spectral domain technique or in the method of moments Nikol skii 1982 Vesnin 1985 Jansen 1985 Rautio Harrington 1987 Dunleavy Katehi 1988 Instead of continuous TE and TM rectangular waveguide eigenwaves Samarskii Tikhonov 1948 their grid analogues are used as a basis to expand the electromagnetic field inside a layer The number of the grid TE and TM waves is finite and their system is complete This means that instead of a summation of series as in the spectral domain approach we have finite sums Moreover each basis grid eigen wave has a grid correction that provides convergence of sums to the series obtained by the continuou
29. ACF Side iChanNo None Side U or L Upper Lower iChanNo any int gt 0 CF Channel Frequency None OCF Channel Frequency OCF None Offset DCR Desired Channel DCR Desired Average resistance at Resistance CF Group Delay ICF Interferer Channel ICF None Frequency IMGF Image Channel IMGF None Frequency Spectrasys Power Measurements Name Description Syntax Spectrum Equation Type ACP Adjacent ACP Side Total Channel power at ACF Side Chan Channel Power iChanNo No AN Added Noise AN Same as AN i CNF i CNF i 1 Where Power CNF AN 0 0 dB CNR i cnr DCP Same as CNR i DCP i CNP i CNP DCP amp CNP CNDR cndr DCP Same as CNDR i DCP i NDCP i NDCP DCP amp NADP 491 Simulation CGAIN Cascaded Gain cgain X Same as X CGAIN i X i X 0 CNF Cascaded Noise enf DCP Same as CNF CNP CNP 0 Figure CNP CNP cgain DCP hy CCOMP Cascaded CCOMP None CCOMP i Summation Compression COMP i from 1 to i dB Point CNP Channel Noise CNP Noise Noise power at CF Power COMP Compression COMP None Calculated compression of each Point stage DCP Channel Power DCP Desired Desired power at CF Desired OCP Channel Power OCP Total Total power at OCF Offset GAIN Power Gain gain X Same as X GAIN i X i X i 1 Where GAIN 0 ra IP Same as IIP i OIP i ICGAIN i where OIP amp ds IIP contain all
30. B V Sestroretzkiy Yu O Shlepnev Electromagnetic analysis of planar devices with resistive films and lumped elements Proc of Europ Symp on Numerical Methods in Electromagnetics JEE 93 Toulouse France 17 19 November 1993 p 227 234 V Yu Kustov B V Sestroretzkiy Yu O Shlepnev Three dimensional electromagnetic analysis of planar devices with resistive films and lumped elements Proc of 27th Conference on Antenna Theory and Technology ATT 94 Moscow Russia 23 25 August 1994 p 352 356 K N Klimov V Yu Kustov B V Sestroretzkity Yu O Shlepnev Efficiency of the impedance network algorithms in analysis and synthesis of sophisticated microwave devices Proc of the 27th Conference on Antenna Theory and Technology ATT 94 Moscow Russia 23 25 August 1994 p 26 30 V Yu Kustov B V Sestroretzkiy Yu O Shlepnev TAMIC package for 3D electromagnetic analysis amp design of MICs Proc of the 5th Intern Symp on Recent Advances in Microwave Technology ISRAMT 95 Kiev Ukraine September 11 16 1995 p 228 233 Yu O Shlepnev B V Sestroretzkiy V Yu Kustov A new method of electromagnetic modeling of arbitrary transmission lines Proc of the 3rd Int Conference Antennas Radiocommunication Systems and Means ICARSM 97 Voronezh 1997 p 178 186 Yu O Shlepnev B V Sestroretzkiy V Yu Kustov A new approach to modeling arbitrary transmission lines Journal of Comm
31. GENESYS is the segment in the middle with frequencies and S parameter data Lines in the data file beginning with are comments and are ignored The noise data at the end of the file is used for noise figure analysis Noise is discussed in a later section AT41435 S AND NOISE PARAMETERS Vcez8V Ic 10mA LAST UPDATED 06 1 89 GHZ S MA R 50 IFREQ S11 821 812 822 0 1 80 32 24 99 157 011 82 93 12 0 5 50 110 1 2 30 108 033 52 61 28 1 0 40 152 6 73 85 049 56 51 30 1 5 38 176 4 63 71 063 59 48 32 2 0 39 166 3 54 60 080 58 46 37 2 5 41 156 2 91 53 095 61 44 40 3 0 44 145 2 47 43 115 61 43 48 3 5 46 137 2 15 33 133 58 43 58 4 0 46 127 1 91 23 153 53 45 68 4 5 47 116 1 72 13 178 50 46 75 5 0 49 104 1 58 3 201 47 48 82 6 0 59 81 1 34 17 247 36 43 101 If Fmin Gammaopt Rn Zo GHz dB MAG ANG Simulation 0 1 1 2 12 3 0 17 0 5 1 2 10 14 0 17 1 0 1 3 05 28 0 17 2 0 1 7 30 154 0 16 4 0 3 0 54 118 0 35 A sample 1 port Z parameter data file is shown below This data file could be used to specify a port impedance that varied over frequency Notice that the data is real and imaginary RI impedance Z data taken across several frequency points 13 90 to 14 45 MHz that has been normalized to 1 ohm R 1 MHZ ZRIR1 13 90 30 8 29 2 14 00 31 6 6 6 14 05 32 0 4 7 14 10 32 4 16 0 14 15 32 7 27 2 14 20 33 1 38 4 14 25 33 5 49 5 14 30 33 9 60 7 14 35 3
32. Next to Add a System Analysis Add a System Analysis After creating a schematic a system analysis must be created There a several ways to accomplish this Only one way will be shown here For additional information on adding analyses click here 416 Spectrasys System To add a system analysis 1 Right click on a folder in the workspace tree where you wan the analysis located Workspace Tree Hr S E BH Tal Default Designa A From Library Rename Delete Properties Analyses Add Harmonic Balance Oscillator Analysis Designs Evaluations Graphs Syntheses t 1Add Data S Add Equation Add Folder Add Note Add Script amp Add Substrate 2 amp dd Table ts Add DC Analysis EA odd Empower Analysis TE Add Harmonic Balance Analysis Add Harmonic Balance Oscillator Analysis Sij Add Linear Analysis Add Sonnet Analysis C Odd System Analysis 1Add TESTLINE Analysis fy amp dd Transient Analysis Export 2 Select Add System Analysis from the selected submenus as shown above 3 The following System Analysis dialog box will appear 417 Simulation System Simulation Parameters General Paths Calculate Composite Spectrum Options DRE Dora 2 5c0mi vata Morminal Impedance Ohms Measurement Bandwidth Channel MHz Retain Level of Data Calculate Mow Schematic Sour
33. Notice how the EM response is slightly down in frequency The linear simulator does not take parasitic losses and box effects into account like the EM simulator does The main reason why the EM response is shifted down in frequency is because the footprint pads for the capacitors actually add more capacitance to the filter The filter responses are shown below The red and blue response is 21 and S11 of the linear simulation and the orange response is S21 and the green response is S11 of the EM simulation 203 Simulation BE MFilter1 Response Workspace EmWalkthru DB 521 MF ilter1 EM1 DB S21 LESSIG EG Haw TELS TB Freq MHz amp DB S21 DB E11 MFiterl EM1 DB S21 MFilerl EM1 DB S11 23 Now is the the time to see the true power of Eagleware s Co Simulation Co Simulation allows your to tune your filter in real time without having to re run the EM simulation In other words you are able to tune your capacitor values without re running EMPOWER Since the response has shifted down in frequency we will need to decrease the amount of capacitance in all caps We can manually change the capacitor values by tuning them in the tune window J GENESYS File Edit View Workspace Actions D Gl ml o xl mu CAPT MFILTER CAPS MFILTER 204 EMPOWER Planar 3D EM Analysis 24 However we can use the optimizer to tune the filter for us We need to open up the optimization targets titled MFilter1 lo
34. Set the checkbox to include noise from the port into output noise 6 To calculate noise propagation parameters as Noise Figure specify the input port the same way as output from the pull down list including only the names of port load elements The checkbox must be set to activate the Input port dialog 7 Set the checkbox Noisy Port for the port to include its thermal noise into output noise 265 Simulation Harmanic Balance Analysis Options General Cacdate Moka Advanced Calculate Nonlinear Hose e put Port RFiPart Qutput Port TF Hose Frequency Bandeidth NossBand Noise Frequency Sweep sweep Type Mirim Frequency Maximum Frequency Number of Points Cub put Calculate 558 Moker of CO Bl carriers Caner with Index Vector lil e Calculate Hote Contributors 8 Define Sweep Type Single point Linear or Log Default sweep type is Log Logarithmic It s recommended to use for sideband noise analysis SSB DSB The Single point sweep type is efficient for noise figure calculation at frequency offsets not close to carrier where the parameter is independent on noise frequency offset from a carrier 9 Define noise sideband frequencies for the noise analysis in the Noise Frequency Sweep group For a swept noise analysis set the Minimum 20 cc Frequency Maximum Frequency and Number of Points per decade for Log sweep type parameters 11 Def
35. The center frequency of these noise points 1s the center frequency of each desired signal This parameter is used when the user wants greater resolution of the noise like through a narrowband Intermediate Frequency IF filter This bandwidth defaults to the channel bandwidth Calculate Phase Noise When checked behavioral phase noise is calculated System Simulation Parameters Composite Spectrum Tab This page controls how the data is displayed on a graph Tabs General Paths Calculate Composite Spectrum Options Output Spectrasys System System Simulation Parameters General Paths Calculate Composite Spectrum Options Gutput Spectrum Plot Options Show Individual Signals Show Totals Show Individual Intermods amp Harmonics Show Individual Spectrums Show Individual PhaseNoise Show Individual Noise Enable Analyzer Mode Resolution Bandwidth REWJ EE MHz v Limit Frequencies t Defaults to channel bandwidth Start MHz Filter Shape Gaussian to 118dBc 60 REW w Stop 2000 MHz Randomize Noise Step 1 MHz M Factory Default Parameter Information SPECTRUM PLOT OPTIONS This information only affects the displayed output and not internal calculations Spectrums can be displayed in groups or individually Show Totals Shows a trace representing the total power traveling for each direction of travel through a node For example if three elements were connected at a
36. Two ot mote ports excited with the same absolute potential and the same polarity The ports are simulated as a single port Use an explicit ground for a single strip internal or common mode port Implicit ground is made available through the closest infinite metal when no explicit ground port is present Port Connected to Edge of object Edge of object Edge or surface of object Edge of object Edge of object Edge of object Edge or surface of object Object on Strip ot slot layer Strip or slot layer Strip layer Strip layer Slot layer Strip layer Strip layer Additional details about each port type and how to define them are given in the following sections 311 Simulation Defining a Single Normal Port Single is the default port type It has the following properties 312 It is connected to an object that is on either a strip or slot metallization layer It can be applied only to the edge of an object The port is external and calibrated The port is excited using a calibration process that removes any undesired reactive effects of the port excitations mode mismatch at the port boundary This is performed by extending the port boundary with a half waveleneth calibration transmission line The frequency wavelength selected during the mesh or simulation process is used to calculate the length of the calibration line For more information about the calibration process ref
37. Type Binaty Can be safely edited No Average size 2 to 25Kbytes but may be larger Use Internal data file for EMPOWER EMPOWER Planar 3D EM Analysis This file contains the calculated Y parameters before deembedding If merge ME is specified the previous data stored in this file is combined with the newly calculated data and the SS S Parameter file is rewritten SS RG etc Backup Files All files with a name or an extension starting with tilde are backup files and can be safely deleted Examples of these files are OMBINE TPL and COMBINE RG EM PO WER Advanced M FILTER Example EMPOWER Advanced Example Filter Synthesis This advanced example shows how to combine M FILTER circuit simulation and electromagnetic simulation We will design a bandpass filter with a lower cutoff frequency of 2100 MHz and an upper cutoff frequency of 2200 MHz We will use the M FILTER module to design the filter then we will perform a linear and EM simulation of the filter 1 First all units in this example use mils In order to get the results in this example the default units should be changed to mils This can be done by selecting Tools and then Options from the main menu then selecting the Umits tab Make sure the Length parameter says mils as shown below 185 Simulation GENESYS Global Options General Startup Graph Schematic Directories Language Units Default units Far graphs tables new schematic
38. Vout G 0 F 2 Ke Zour F Pin where Vout k is the complex voltage amplitude of the k th output noise carrier and F k is the frequency of the k th output noise carrier Zout F k is the impedance of the output load at the carrier frequency F k Pin is the available power of the input noise carrier Entering Nonlinear Models GENESYS supports four different way to enter nonlinear models e Direct Schematic Entry e Single Part Model e Nonlinear Model Library The simple way is to enter a nonlinear model is through direct schematic entry You place a nonlinear device such as an NPN transistor from the schematic tool bar Then double click the device and type in the device parameters The advantage of this technique is that it is simple The disadvantage is that it is not as easy to reuse the device in another design 287 Simulation 288 Another way to enter a nonlinear model is to create a single part model This is similar to using a model statement in other simulators See the Designs Single Part Model section in this User s Guide for details A third way to enter nonlinear models is to choose one from the supplied library of parts To do this just enter the base nonlinear model that you would like for example a PNP then change the model to the desired part using the Model button on the element parameter dialog Chapter 7 Linear Analysis Overview Linear simulation calculates S par
39. by double clicking the layout background and then selecting the Associations tab Then proceed to click on the Change button and choose CC1608 0603 Chip Capacitot from the SM782 LIB 195 Simulation LAYOUT Properties General Associations General Layer EMPOWER Layers Fonts Hemen type Defaut Footprm Library 1 o GROUND BOTTOM emi SAMPLELIB Change a RES RC3216 1208 ChipResistr sm7a2LB_ Change a cer ccena 10603 Chip Capactor SM762LI_ change 5 soD SIGNAL COPLANAR 50mi SAMPLELIB change B rt sot3pnsot smrazuB change 7 BP SODSinnsor SMPSALB Change a TE TRANSMSSONLNE 120 28mi__ SAMPLELIB change B TRF SOWWwieinnsoc SM762LI Change aof xT Hee Quartz Crystal LEADEDLIB Change M wo SOUS3ensor SM762LI Change 12 co COXXALLANDNG iDPSOni SAMPLELIB change fa cU COAXIAL LANDING 100x50mil SAMPLELIB Change aa CL COUPLED TRL 120 25mi S 18mil_ SAMPLELIB change a o SO amp WwWeinnsOC smaze nange a mi SO amp WwieinnsoIc sm7a2LB_ Change ora rar SANPLELE change ee Te Te gG UG Current T able Default thl Save Table As Load T able Cancel Apply Help 14 The next step is center the components on the PWB This is done by selecting the MFilter1_Lay Layout window and from the Edit menu select Select All Then from the Layout menu select
40. circuits startup distribution Benefits Uses manufacturer DC biasing Automatic Automatic Lots of vendor provided information deembedding deembedding supplied models measured data Predicts box mode Predicts box mode Use frequency Requires very Lots of vendor effects e g What effects e g What dependent little memory supplied models happens if the circuit happens if the circuit equations and is placed in a box is placed in a box post processing Non linear Can use arbitrary Can use arbitrary Easily use kaa modeling of shapes does not shapes does not Use measured 4 crossover require an existing require an existing data in simulation user functions distortion etc model for them model for them No time Very slow for near Much slower slow Extremely slow l domain harmonic data than linear Very hard to model No biasing frequency domain May require lots of Requires lots of Takes a lot of information behavior e g memory memory memory and time unloaded Q Drawbacks Everything is Does not model Discretizes metal Discretizes metal Requires linear noise performance patterns as polygons patterns to fit grid nonlinear models Requires is ledge of Requires knowledge Sun SPI W l TAE pun 5 Can be difficult to Can be difficult to set transient circuit of circuit coupling oe ae cotrectly set up a up a circuit for behavior for coupling factors parasitics A circuit for si
41. complexity of the circuit and the density of the mesh The size N of the interaction matrix Momentum GX equation is equal to the number of edges in the mesh For calibrated ports the number of unknowns is increased with the edges in the feedlines added to the transmission line ports CPU Time The CPU time requirements for a Momentum simulation can be expressed as CPU time A BN CN DN gt whete N number of unknowns A B C D constants independent of N The constant term A accounts for the simulation set up time The meshing of the structure is responsible for the linear term BN The loading of the interaction matrix is responsible for the quadratic term and the solving of the matrix equation accounts for e part of the quadratic term when using the iterative solver e the cubic term when using the direct solver It is difficult to predict the value of the constants A B C and D because they depend on the problem at hand Memory Usage The memory requirement for a Momentum simulation can be expressed as Memory X YN ZN where N number of unknowns X Y Z constants independent of N Like with the CPU time expression the constants X Y and Z are difficult to predict for any given structure For medium to large size problems the quadratic term which accounts for storing of the interaction matrix always dominates the overall memory requirement For small structures memory usage can also be dependent upon the su
42. connected to the bottom wall ground plane a physical representation of ground Z directed internal ports can be used in GENESYS to connect elements just like a node in a schematic In other words components like resistors and transistors can be connected directly to these ports You simply place a z directed port in the center of the pad for the component in these cases Note GENESYS does this automatically as described later in this section 153 Simulation 154 Manually Adding Lumped Elements Note GENESYS will automatically add lumped elements to your simulation if components are on your layout This section is for background information and advanced applications The circuit shown below contains an EMPOWER circuit which was drawn completely in LAYOUT The schematic for this network was blank It has 4 ports ports 1 and 2 are external and ports 3 and 4 are internal EMPOWER will create a 4 port data file for this circuit NET 1 Note Internal ports and no deembed ports must always have higher numbers than normal external deembedded ports In the figure above the internal ports are numbered 3 and 4 while the external ports are numbered 1 and 2 The data file created by EMPOWER can then be used in GENESYS The circuit on the right above uses the resulting data in a complete network First a FOU four port data device was placed on the blank schematic The name assigned to this FOU block was the name of the in
43. enclosure causes energy to be lost to free space and resonance effects are reduced This greatly reduces coupling between metal segments of the circuit and it 1s evident in the responses given in the Box Mode example cited eatlier with the cover removed Effects of removing a top cover are illustrated in the Examples Filters EdgeCoupledOpen WSP and Components V Box Modes WSP See your Examples manual for details Cavity Absorber A similar benefit may be derived by placing absorber material on the cover or in the cavity While the poor ultimate rejection in the stopbands of filters is not recovered heavy coupling between segments is avoided This is sometimes necessary to eliminate oscillations of high gain amplifiers in oversize enclosures By far the most elegant and safest approach to minimizing box mode problems is placing circuits in small enclosures EMPOWER Planar 3D EM Analysis EM PO W ER Viewer and Antenna Patterns O verview This section describes how to launch the EMPOWER viewer program and how to use it to visualize and interpret currents or voltages generated by EMPOWER It also describes the viewer interface The EMPOWER viewer helps you visualize current distribution and densities in a board layout It processes current density magnitude and angle and plots them as two or three dimensional static or dynamic graphs These plots provide insight into circuit behavior and often suggest modifications which improve the p
44. from the Graph menu 428 Spectrasys System Workspace Tree F EJ From Library a Analyses Designs m Evaluations 2 sraphs Add Rectangular Graph Syntheses 4 4dd Data S Add Equation CJ Add Folder Add Mote Add Script amp Add Substrate 2 Click the Graph Properties tab 3 Click the Measurement Wizard button C Heare ulead dole J 4 Select the desired path data set BM Add 3D Graph iM Add Antenna Plat 9 Add Polar Chart Add Rectangular Graph dp Add Smith Chart Measurement Wizard What workspace has the Dataset you wish to use Feed Forward Amplifier Which Dataset contains the data vau wish bo use vsktemi Data Systemi Data MainPath 5 Click the Next button 6 Selectthe desired path measurement 429 Simulation Measurement Wizard System Data MainPath Which measurement would vau like ba display Adjacent 2nd Upper Channel Power Added Noise Channel Frequenc Cascaded Gain Desired Only Conducted Second Order Intermods Conducted Third order Intermads Carrier to Noise and Distortion Ratio Cascaded Noise Figure Channel Noise Power Carrier to Moise Ratio Channel Power Channel volkage Desired Channel Power Desired Channel Voltage Limit Sweep Range What display Format operation would you like to use dbi dB Magnitude Min i jo 7 Click the Finish button New Level Diagr
45. s mutual customets to have an easy to use interface between GENESYS and Sonnet Most common Sonnet features are supported directly by this interface Additionally a manual editing mode is available which allows access to all Sonnet features while maintaining connectivity with GENESYS An added advantage to Eagleware s interface is direct lumped element support just as in Eagleware s EMPOWER simulator Before using this interface Sonnet Version 9 52 or later must be installed on your computer This interface hides most of the Sonnet details and an in depth knowledge of Sonnet is not required However we still recommend that you follow through the Sonnet Tutorial manual before attempting to use the Sonnet interface Creating a Layout Without a Schematic To create a layout for Sonnet you must follow essentially the same steps as when you create a layout for EMPOWER In this tutorial we will use the same stub example used in the EMPOWER tutorial To create this layout follow these sections in the EMPOWER Operation chapter e Creating a Layout Without a Schematic e Box amp Grid Settings e General Layers e EMPOWER Layers e Drawing the Layout e Centering the Layout e Placing EMPOWER Ports If you have already created this layout and simulated it with EMPOWER you may use your existing file If you do not want to create this layout yourself you can load the file from LayoutOnly wsp in the root of the examples directory Simulati
46. since a wide line is simulated across the entire length of the box However line analysis is always symmetrical and may be symmetrical in two planes if the port lines are placed exactly in the middle of the box EMPOWER also caches the line analysis results so if the box and port lines are not changed between runs previous data will be used The data for these lines are stored internally in the Workspace WSP file using internal files namedEMPOWER R1 EMPOWER R2 etc EMPOWER also has the intelligence to detect when two or more ports have the same configuration width position etc and will only run the line analysis once See Microstrip Line for a complete example which examines deembedding M ultiM ode Ports Until now all ports which we have looked at have been single mode ports Single mode ports act just like regular nodes in GENESYS and external components can be added directly to these ports EMPOWER also supports external multimode ports where two EMPorts are close enough together that they are coupled This circuit uses multimode ports with ports 1 2 and 3 being a 3 mode port 4 being a normal single mode port and ports 5 and 6 being a 2 mode port 139 Simulation HEE Multimode ports have the following features e They much mote accurately characterize the performance of a network with two or more lines close together on one wall e They cannot be used like normal GENESYS nodes They can only be connecte
47. while Phi is being swept from 0 to 90 degrees also in 1 degree increments Measurements and Plotting Once far field radiation data is generated the following measurements can be plotted ETHETA phis thetas freqs the theta component of the total electric field Phzs Thetas and freqs can either be single values or ranges of values EPHI pAis thetas freqs the phi component of the total electric field Phzs Thetas and fregs can either be single values or ranges of values ETOTAL pAis thetas freqs the magnitude of the total electric field P zs Thetas and freqs can either be single values or ranges of values ELHCP E field Left Hand Circular Polarization ERHCP E field Right Hand Circular Polarization EAR E field Axial Ratio The measurement wizard can be used to to select these measurements and the proper syntax is automatically generated Rectangular Antenna Polar and 3D charts may be generated to display the antenna data Only one variable out of Phi Theta and Frequency may be swept when displayed on the two dimensional charts and two variables may be swept when displayed on the 3D chart Below is both a rectangular and Antenna plot polar of the ETOTAL measurement where Theta is being swept from 0 to 360 degrees Phi is held constant at 0 and the frequency is held constant This particular antenna is a very small dipole located one wavelength above a ground plane on top of a substrate EMPOWER
48. 128 130 P iue US 498 el M HR 502 Adaptive Frequency Sampling 367 ONS NINOS sacar create fe aera TT 503 adjacent channel frequency eerie atsees 498 adjacent channel powet ee eterne 502 PCN CATO a i 11 297 Advanced TAD srssacsesescsestsitiameararramaiesoosoesennssen 225 att QDOVeu oo EHE S i oss d S 02 151 ioc aleon DE e EE ENE 206 Alov ID BEP S iaa titii 225 ampBtude S EPPING ac eot eese tte titres 234 P a M X 503 PERI AS T ET A ETA 548 SE a nea eater ot iad ean enue 13 Jodi 151010 MM HMM TION 13 INO A E E 13 161 ALIA AION MEER 161 170 INO E I EI E nacio E 96 artificial intelligence techniques 234 automatic CEEMMDCACING ix se a aveva iioii 1 automatic port placement sss 154 automatic recalculation 86 225 automatica ealeulat esee edt tee tecto 229 Auto save Workspace After Calculation 225 available sati cC I6 Sq eerie tana tetro nes 11 B111 294 pac kanno A ea ER RN 86 palanced ampu NETS aoei 8 PASC ECO rd NEED enous 02 Basic HB analysis Measurements 244 Basic HB oscillator analysis Measurements sss 261 Indc Donde snos ele bor 107 Behavioral Phase Noise 2 2 otim 475 binary eS te T 179 DOUNON COV T ccce cad esa ob eee 92 100 DOM M 1 97 110 129 135 153 box modes ees 1 131 157 159 160 DONO cue c MM
49. 13 DC analysis oe telat aunties 85 86 Ea e Re NS Re Per rere eres 1 OG NEEE E ETS 507 pie TOREM 500 Ir saessnnesssasees 526 decomposition 139 142 143 145 151 deembecdded POLIS mede d eee 154 deem beddi c eun niet 136 138 184 de embedding alg OFtDIPEL dere reti deant 218 default ODePAMtoE ssp ot tu IH 19 default simulation data ess 14 delete simulation data s 233 UEFIVAU VES uec a odes AR NIS TH ROTER DEDERE UE 225 desited channel PO wel ees tptieien i aftseceice 507 D sired Chantiel Resistance aote 500 Desired Channel Resistance DCR 500 Desired Channel Voltape cuiu enitn 526 diasonaliz4to Boe pc D M c aus 206 208 Gi She Gi CS NS Un meee 92 160 dielectric CONS TAN Ey addio ree Sensations 100 dielectric oan ioci NEED RN 159 CIELO TES DOES coi coo de I HET enter Ln erroe cin add 92 110 Dimensions slo 97 directional ente eee aa eeens a seee EEN 485 discre CSa OO anann 206 CISCHEHZES Tne C osse dete t I e 1 dis soy kale Ke ea E 1 175 Edit 33 etYUs aes trit ade EU Ue de se eivai 129 effective noise input temperature s es 11 D ISEBIHIOOEe adsidua alvin 174 177 PEASE WAVE cci wed e needs IE T8 Dt M 11 IC CUNO Cie E erea 1 electromagnetic simulation 1 91 225 AI ole ueris 97 101 135 136 308 EMPOWER WIG WGK EEEE N EET I 119 122 170 valle thEOUb Dos
50. Channel Frequency was 220 MHz then the DCV is the average voltage from 219 985 to 220 015 MHz This voltage measurement will not even be affect by another 220 MHz sional traveling in the reverse direction even if it is much larger in amplitude Default Unit dBV Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used ONLY DESIRED SIGNALS Travel Direction Only in the FORWARD direction Equivalent Input Noise Voltage VNI This measurement is the stage equivalent input noise voltage Noise figure source resistance and temperature are used to determine equivalent input noise voltage The following equation is used to take the noise factor at each stage the desired channel resistance at each stage and the system analysis temperature in Celsius and converts these parameters to an equivalent input noise voltage VNI n NFtoVNI CNF n DCR n TempC nV sqrt Hz where n stage number See SVNI for additional information Spectrasys System Default Unit nV sqrt Hz Channel Used No channel is used for this measurement Types of Spectrums Used None Travel Direction N A Interferer Channel Voltage ICV This measurement is the average voltage in the interferer channel This measurement is simply a Desired Channel Voltage measurement at the Interferer Channel Frequency Channel Used Interferer Channel Frequency and Channel Measurement Bandwidth Types of Spectrums U
51. Different levels of mesh granularity on the metalization side are possible e One cell in the horizontal direction in case edge mesh is turned off and no Transmission Line Mesh specification is entered i This is the advised setting in case the conductors are specified as automatically expanded e Multiple mesh cells in the horizontal direction in case edge mesh is turned off and Transmission Line Mesh specification is entered e Edge mesh in the horizontal direction in case the global edge mesh toggle is turned on an edge mesh is generated on all sides top bottom and both sides of the automatically expanded conductor When the substrate definition of a circuit is specified from the Layout Properties dialog gt Layers tab the layer stack specifies only the vertical dimension of the circuit and not the horizontal dimension Given this definition the layers extend all the way to infinity in the horizontal direction For many circuit designs this is not relevant and does not affect 341 Simulation the simulation However there may be instances where you want to introduce horizontal boundaries For these instances you can use boxes Boxes enable you to specify substrate boundaries in a horizontal direction on four sides of the substrate Mote specifically for a box you define four perpendicular vertical planes of metal as the horizontal boundaries of the substrate These four vertical planes or walls form a rectangle if
52. EM1 The default plot seen in the main window is an animated surface electric current density distribution function reflecting the surface currents in the strip plane At the initial time t 0 it will look similar to the graph shown below empower Viewer Y6 5 OL E3 File View xy Real Solid Freq GHe 18 1a lalefl of a 4 To Front Side Oblique To get this snapshot we stopped animation by clicking the Animation camera icon adjusted the view slightly and toggled the background color to white To obtain this view EMPOWER Planar 3D EM Analysis simply press the Oblique button on the toolbar after starting the viewer All other settings are the default e Show XY current density distribution XY X Y Z button e Show Real part of the current density distribution View Menu Switches Value Mode or Value Mode button e Show Absolute values of the current density quantities View Menu Switches Absolute Value Display e Animation is off and time is set to initial View Menu Switches Animation or Animation Camera button e Scaleis on View Menu Switches Scale e Solid polygons view View Menu Switches Wireframe or Solid Wire button Note For printing Togele Background Color from the File menu was also used to change the background to white To reset the time to zero the animation was turned off and the Real Mag Angle button was clicked th
53. EMPort button Eh on the LAYOUT toolbar and click on the center left end of the series line The EM Port Properties dialog appears Set the Draw Size to 25 This controls how large the port numbers will be drawn on the LAYOUT screen but has no effect on simulation Note that the default port number is 1 Select the OK button EMPOWER Planar 3D EM Analysis Next select the EMPort button on the toolbar again Click on the center right end of the series line The EM Port Properties dialog appears Again type 25 into the Draw Size box Note that the default port number is 2 Select the OK button The screen should now look like Be Layoull workspace WorkSpace 1 For simulation EMPOWER will take S Parameters from these ports Simulating the Layout To run EMPOWER you must create a simulation Click on the Add Item button in the Workspace Window and choose Analyses Add Empower Analysis This displays the EMPOWER Options dialog This dialog is shown below For a description of the dialog options see the section on External Ports For now just set the prompts as shown below 117 Simulation EMPOWER Options Layout to simulate bo Fort impedance 50 Generalized Cancel Automatic Recalculation Recalculate Now Electromagnetic simulation frequencies Start freg MHz 8000 mm Automatically save workspace after calc Stop freg MHz 11000 Viewer data Iv Generate viewer data slower Number of point
54. First check all of the channel frequencies in question Measurements are based on these frequencies If these frequencies are NOT what the user expects then channel measurements will also not be what the user expects 3 Check channel power measurements Remember Channel Power CP and Desired Channel Power DCP are two different measurements Channel power includes forward and reverse traveling power whereas desired channel power only includes forward traveling power Most measurements depend on desired channel power i e GAIN CGAIN etc 4 Use the desired channel resistance DCR measurement to quickly determine if there are VSWR related problems 5 Click here for information on troubleshooting intermod path measurements Use Spectrum Identification on Graphs to Locate Root Spectrum Causing a Problem 1 Once the measurements have been use to identify the node where the problem occurs then add a graph at that particular node and examine the spectrum in the channel in question 2 Use the spectrum identification capability to identify the suspect spectrum NOTE You can Zoom to Spectrum by right clicking on the spectrum of interest and selecting Zoom to Fit Spectrum Trace Use Path Powers Voltages and Impedances 1 Enable path powers voltages and impedances so that these values can be added to the dataset 2 The path dataset will now contain specific information regarding values that spectrums are seeing along
55. For advanced applications you can nest Parameter sweeps creating 4 D 5 D or higher data This data can then be viewed on a table Performing a Parameter Sweep A parameter sweep gives you a set of responses for a set of parameter values You can perform a parameter sweep on any tuned variable To create a parameter sweep 1 Click the New Item button on the Workspace Tree toolbar and select Add Sweep from the Evaluations menu Getting Started 2 You will see the sweep properties box which will be similar to this Parameter Sweep Properties Sweep Mame qas Lid Calculate Mow Analysis to Sweep Linearl v ha Factory Defaults Parameter to sweep DesignsiDesignll 1 0 ww Output Dataset Description Parameter Range Type OF Sweep Start 100 pF S Liner Number of Points 10 Leg Paints Decade Linear Step Size pF List pF Unit of Measure pF Show Long Parameter Names Propagate All Variables When Sweeping Cor anky m 3 By default the sweep settings will be the same as the last time you created a sweep The default parameter to sweep is just the first in the list Here the parameter is in the Designs folder in the design named Design in the part named C1 as parameter C In the list are all tuned parameters ot equation variables Use the settings shown above then click Calculate Now to calculate
56. For example a thru line at a global seeding of 30 might produce a 50 x 50 matrix to solve Since the time required to solve the matrix is proportional to the matrix size the same thru line at a seeding of 60 might produce a 200 x 200 matrix because number of cells would be twice as dense in both X and Y dimensions This means that it will take longer to solve Adjustments to your mesh parameters can be made to control complexity of the mesh and in turn the amount of time required to simulate For complex circuits where simulation times may be long even with a simple mesh reducing the complexity of a mesh can save significant amounts of simulation time Mesh Patterns and Memory Requirements The number of unknown currents which is calculated when the mesh is precomputed is based on the complexity of the mesh The more unknowns the larger the matrix that must be solved during simulation The amount of memory required to solve is proportional to N N where N is the number of unknown currents Processing Object Overlap The parameter Thin layer overlap extraction should be used in designs that include thin layers that are both close together and overlapping It is possible that mesh cells can be generated that cross the overlap region and this is not desirable As an example consider two layers one above the other and close together They carry a large charge density where they overlap which increases proportionally to 1 distance If
57. Hew Layout x General Associations General Layer EMPOMWER Layers Fonts Height or Tand Surface Imp n eo ee O ST 255 popa 235 poo L 33 LT O n Sub Teflon 1 42 Carcel A 94 EMPOWER Planar 3D EM Analysis The EMPOWER Layer Tab consists of the following main entries Top Cover and Bottom Cover Describes the top and bottom covers ground planes of the circuit e Lossless The cover is ideal metal e Physical Desc The cover is lossy These losses are described by Rho resistivity relative to copper Thickness and Surface Roughness e Electrical Desc The cover is lossy and is described by an impedance or file See the description below under metal for more information e Semi Infinite Waveguide There is no cover and the circuit is simulated as if the box walls and uppermost substrate air layer extend up or down forever an infinite tube e Magnetic Wall The cover is an ideal magnetic wall This setting is only used in advanced applications e Substrates Choosing a substrate causes the cover to get the rho thickness and roughness parameters from that substrate definition We recommend using this setting whenever possible so that parameters do not need to be duplicated Ait Above and Air Below The presence of air at the top of the box as in microstrip or the bottom of the box as in suspended microstrip is so common that special entries have been provide
58. ICP Interferer Channel Power ICP This measurement is the total integrated power in the interferer channel This power is used for intermod measurements such as IIP3 OIP3 SFDR etc This measurement is simply a Desired Channel Power measurement at the Interferer Channel Frequency Channel Used Interferer Channel Frequency and Channel Measurement Bandwidth Spectrasys System Types of Spectrums Used Same as DCP Travel Direction Same as DCP Interferer Gain IGAIN This measurement is the gain of the interferer tone channel along the specified path The Gain is the difference between the Interferer Channel Power output of the current stage minus the Interferer Channel Power output of the prior stage as shown by IGAIN n ICP n ICP n 1 dB where IGAIN 0 0 dB n stage number Channel Used Interferer Tone Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as ICP Travel Direction Same as ICP Image Channel Noise Power IMGNP This measurement is the integrated noise power of the image channel from the path input to the first mixer After the first mixer the Mixer Image Channel Powet measurement will show the same noise power and the main channel noise power For example if we designed a 2 GHz receiver that had an IF frequency of 150 MHz using low LO side injection then the LO frequency would be 1850 MHz and image frequency for all stages from the input to t
59. Impulse Response section of the convolution tab are used a Independent impulse responses are created for each matrix entry for the element For a two node element that will generally be four entries For each entry the minimum number of points is attempted first 128 by default The frequency domain response is calculated from DC to the maximum Frequency An inverse FFT of the data is taken to get the impulse response If any entries in the last half of the impulse response are larger than the Convolution Absolute Truncation or are larger than the Convolution Relative Truncation times the maximum entry in the convolution response then the accuracy is considered insufficient the number of points is doubled up to the maximum and we go back to step C The convolution response is truncated according to the same criteria from step E For a simple transmission line this often results in a single point convolution entry with a delay CAYENNE Transient Analysis g If Save Impulse Responses is checked then the impulse responses ate saved to the dataset with the name h zodelname row column numposnts The Accuracy Testing and Most Accurate Frequency settings ate very important and unique features of CAYENNE These features allow the user to make tradeoffs between accuracy and simulation time while the default settings give generally acceptable accuracy while avoiding most of the difficulties associated with convolution base
60. In equation form the generated third order intermod power ts GIMCP n integration of the intermods generated at stage n across the channel bandwidth dBm Each column in this measurement is for a different intermod order up to the Maximum Order specified on the Calculate Tab of the System Analysis Dialog Box The column number is the same as the order starting from the left with order 0 Remember intermod bandwidth is a function of the governing intermod equation For example if the intermod equation is 2F1 F2 then the intermod bandwidth would be 2BW1 BW2 Note Bandwidths never subtract and will always add The channel bandwidth must be set wide enough to include the entire bandwidth of the intermod to achieve the expected results The Automatic Intermod Mode will set the bandwidth appropriately Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used ONLY INTERMODS and HARMONICS separated according to their order Travel Direction All directions through the node Total Intermod Channel Power All Orders TIMCP This measurement is the integrated total intermod power conducted from the prior stage plus the intermod power generated by the current stage In equation form the conducted third order intermod power is Spectrasys System TIMCPIn integration of the total intermod spectrum at stage n across the main channel Each column in this measurement is for a different i
61. M 341 btOad Dat fIDIS e assesses edt aod IU 436 calculate intermods along path 545 Calculate Jacobian Numetically 225 CalculatertiolSe m aves cimi c dn 545 Calculate NOW deese eme oo erecti dd 225 Calculate port wave data sva 225 Calculate wave data css ctii ttes 229 Calculated Stage Compression 508 Carrier to Noise and Distortion Ratio 504 Carrier tO NOISE FAO S ove ees ede va eei dewere 503 Cascaded Compresi ON Gesund tese 504 cascaded cA s escrita od SRI eret 504 Cascaded Intermod Equations 451 Cascaded Noise Analysis ueris 466 cascaded noise figure oet 504 533 Cascaded Noise Figure Equations 470 CAVE ADSOT DCR cu esed AE IIR MILIA 160 DAVID ESO NACO eeren 157 C O NN DEI due 504 Cell 617 oue veces aus 97 128 132 170 206 centermb LAV OUTS y ien 116 p E 499 S COHDUEIN aet RII PARRA tes 504 COND uu UL UU ME 2525 chamieted COFPeEs se UI RN dires 97 channel path MequenC y esser cese 424 channel frequene y ec an eiii p qus 424 499 channel MOISE DOWEE bia Diete des 506 Channel Noise WO AGC ec scencnconorassosacacsesess 525 channel OSEE RN NIMM 506 Channel WO PASC souci deos 525 SATIN Sat va tia Ue ced idu 424 Characteristic HTIDedatiCe a aedes seien a 142 CIMSP e M M H 521 Ne SE E T 2521 Simulation 558 CIECIOS Set dee Dt 11 14 292 298 299 LOB M S 504 INTE P
62. MS Windows Windows NT and MS DOS are U S registered trademarks of Microsoft Corporation Pentium is a U S registered trademark of Intel Corporation PostScript and Acrobat are trademarks of Adobe Systems Incorporated UNIXQ is a registered trademark of the Open Group Java TM is a U S trademark of Sun Microsystems Inc SystemC is a registered trademark of Open SystemC Initiative Inc in the United States and other countties and 1s used with permission Drawing Interchange file DXF is a trademark of Auto Desk Inc EMPOWER ML GENESYS SPECTRASYS HARBEC and TESTLINK are trademarks of Eagleware Elanix Corporation GDSII is a trademark of Calma Company Sonnet is a registered trademark of Sonnet Software Inc Contents Chapter 1 Chapter 2 Cetna Started ooo di den E uses it tue dioc iu eU cune 1 STU ATO DIS sans ec adibt emploi t dC T A Ac app SUED DIR etai a ud 1 When smilitor should USE os os arid ln rater tus tavkvsutost die eater vett Petite iod Pete Reda pii 1 Usne Eorsi mtie Aesi beste disco A OEN 4 Device Darkane DT Kt 5 Linea ver Nonnen Derice Models ior ar b ere ee ai eee aarti 5 Enea Data COVOPVIOM atout pt voa tutius b a epa epa pu ta sU E REM ELI DE Ege 5 Wein a Data File OENES Y Iriri a n nn a O AEA E AAA 5 Proceed Device Daties eate Duedia tate asta a aate i e esM ania atten 6 Creato New Linrcat Data File sais eoe os tede o qoot eo test boots bu nC 6 Fils Recon kep ii ida beet ih
63. Momentum by defining ground reference ports Figure 15 The S parameters obtained with Momentum are identical to the S parameters of the two port for the correct recombination scheme Figure 14 c substrate layer stack metallization Layer GaAs H 100 Hm 12 9 sheet realator 387 Simulation Figure 15 CPW step in width structure with ground references for the internal ports Limitations and Considerations This section describes some software limitations and physical considerations which need to be taken into account when using Momentum e Matching the Simulation Mode to Circuit Characteristics e Higher order Modes and High Frequency Limitation e Parallel Plate Modes e Surface Wave Modes e Slotline Structures and High Frequency Limitation e Via Structures Limitation e Via Structures and Substrate Thickness Limitation e CPU Time and Memory Requirements Matching the Simulation Mode to Circuit Characteristics The RF mode can be used to simulate RF and microwave circuits depending on your requirements However the RF mode is usually the more efficient mode when a circuit is electrically small geometrically complex and does not radiate This section describes these characteristics Radiation RF mode provides accurate electromagnetic simulation performance at RF frequencies However this upper limit depends on the size of your physical design At higher frequencies as radiation effects increase the
64. Noise tab and enable Nonlinear noise 2 To calculate oscillator phase noise specify an output port or output voltage where the noise will be calculated Set the checkbox Noisy to include noise of the port in phase noise contributors The phase noise does not need specification of an input pott it may be disabled 3 Set parameters of the frequencies sweep for example Minimum Frequency 1Hz Maximum Frequency 1MHz Number of Points per decade 3 4 Set the checkbox Improve accuracy of Low Frequency Oscillator Phase Noise Polishing to improve phase noise accuracy at low frequencies where the characteristic becomes flattened 5 Set the index vector of output carrier For an oscillator it is typically its index of 1st harmonic 1 for some applications it may be any harmonic of the oscillator spectrum 6 Set the checkbox Calculate noise contributors to perform the noise contributors analysis 271 Simulation Oscl_Closedl p N1 Osc1 OpenLoop OSCPORT OSCPORT FOSC 51 081MHz VPROBE 0 588V 272 HARBEC Harmonic Balance Analysis Oscl_OpenlLoop EU Lz11505 67443 nH i a3 E2208 nH CB 34984 pF Pon d PORT IET MUN LEM LL NC L 2122 09098 nH C258 40018 pF 72143054 pF ca u Cz kd a DPI Te The oscillator simulation noise page is shown below Parameters for oscillator noise analysis ate specified here 273 Simulation Harmanic Balance Oscil
65. Normal ports are external ports which are deembedded and may be multi mode They are shown in gray on the layout e No Deembed ports are external ports which are not deembedded and cannot be multi mode They are shown in white on the layout e Internal ports are also not deembedded and cannot be multi mode They are shown in white on the layout For more information on dembedding and multi mode lines see below Deembedding If you are actually building your circuit in the same style as an EMPort that is if your ports consist of a line which stops just short of the end wall as is often the case with a coax microstrip junction then you may not need to use deembedding because EMPOWER is simulating the circuit as you are actually going to build it However you may not have this kind of construction or you may be simulating a small segment of a larger circuit In an external port there is capacitance at the port due to coupling from the open end of the line to the wall Deembedding removes this extra reactance perfectly matching the transmission line modeling it as though the line and box extend out to infinity Deembedding also allows you to define a reference plane shift By default the reference plane shift is zero which means that the resulting data is measured at exactly the side wall If the reference plane shift is negative then the data is measured from inside the box effectively subtracting length from the circuit If the re
66. Output SOIV None oltage Intercept SOIV2 None SOIV3 None SOVIDBjStage Output 1 dB SOV1dB None Compression Point SOVSAT Stage Output SOVSAT None oltage Saturation Point SVGAIN Stage Voltage Gain SVGAIN None SVNI Equivalent Input SVNI None oise Voltage TNV otal Node Voltage Total VDC DC Voltage Total VNI Equivalent Input FtoVNI Same as CNF oise Voltage CNF DCR and DCR emperatute General Adjacent Channel Frequency ACFT U or L n SIVSAT i SOVSAT i SVGAIN Stage Entered Intercept Point Stage Entered Intercept Point Stage Entered Intercept Point Stage Entered Compression Point Stage Entered Output Saturation Voltage Stage Entered Voltage Gain Conversion from stage entered noise figure Peak average voltage at the node DC voltage VNI i NFtoVNI CNF i DCR Temperature This measurement is the frequency of the specified adjacent channel All adjacent channel frequencies are relative to the main Channel Frequency Consequently channels exist above and below the main reference channel frequency The user can specify which side of the main or reference channel that the adjacent channel is located on and also the channel number The channel number is relative to the main or reference Spectrasys System channel Therefore channel 1 would be the first adjacent channel channel 2 would be the second adjacent channel and so on U Upper Side L
67. Parameters from GENESYS Check the Generalized box in the EMPOWER properties dialog box When EMPOWER is run it outputs a file in the structured storage when run from GENESYS for each port with impedance data with extensions R1 R2 R3 etc so for a 2 port network in file EMPOWER analysis EMT using Generalized impedance is equivalent to using an impedance of WSP Simulations EM1 EMPOWER R1 WSP Simulations EM1 EMPOWER R2 See the examples manual an example of the use of generalized S Parameters EMPO WER Decomposition 142 Overview NOTE The Single and Multi Mode transmission lines required to use decomposition are not available in GENESYS In EMPOWER it is possible to break down large circuits into smaller segments which are connected by transmission line sections Decomposition can be tedious to implement but its reward is that simulations can be performed accurately in much less time and with fewer frequency points The principal benefits of decomposition are EMPOWER Planar 3D EM Analysis e Ability to tune single or coupled transmission line sections inside a circuit which was simulated by EMPOWER For example you can change the size of a meander line or adjust the tap point on an interdigital filter without rerunning the EMPOWER simulation e Most circuits require far fewer frequency points for accurate analysis This is due to the fact that quarter wave resonant lines are broken down into much smalle
68. Spec Divider cal Spectrum PE P1 2 4GHz 0 dem 0 VERE Source Divider Input P1 2 5 GHz 5U0dBrmi 14D SSB Source oO T DL E Lae m E QE EJ 22 2 3 2 4 Frequency GHZ Divider Input Spectrasys System gt Output Spec m EK Divider Output Spectrum 4 8 GHz T dBm 2nd Harmonic i Ci 4 a I _ ri 2 L Frequency GHz i Divider Output 1 William F Egan 2000 Frequency Synthesis by Phase Lock 2nd Ed John Wiley amp Sons pp 71 78s Behavioral Phase N oise The SPARCA engine supports behavioral phase noise Phase noise can be specified on certain source and oscillator models Phase noise is an independent type of spectrum and as such measurements can operate on this spectrum in the presence of others spectrums of different types This independence allows phase noise to be modified through mixers multipliers and dividers without affecting the parent spectrum This 1s illustrated in the figure below where the signal spectrum type is shown in one color and the phase noise in another NOTE Phase noise must be enabled on the Calculate tab of the system analysis AND enabled on the source model 475 Simulation LO Spectrum E ri D L o TEIL jour 1528 1529 19530 1531 TEE 15 Frequency MHz Phase Noise Specification Phase noise is specified by two lists e A list of frequency offsets e Alist of power l
69. Substrate 1 G Copy From Description Copy To p B Remove EX se RET E Decore constant 38 Tani Loss Tangent MeS Rh eesti o Tik MetalThickness pa mb S emamemes tet Height Substrate Het Stl Set To Factory Defaults Cancel Help E 5 We want to choose Bandpass as the type and Combline as the Subtype For this walkthrough we want to use the Chebyshev filter shape Your topology tab should look like what we have below NOTE Ignore locals errors created during this process since the design has not been completed 187 Simulation M FILTER Properties Topology Settings Options G Values Summary Optimize Unco andpass Shape Chebyshev x Subtype C Stepped C nterdigital C End Coupled Elliptic C Edge Coupled C Hairpin Combline User File Browse lssues Output Resistance 50 S06 Frequencies are 2099 94MHz to 2300 UBMHBHz 6 The next step is to specify all the filter parameters in the Ses zngs tab as we have below M FILTER Properties Topology Settings Options G Values Summary Optimize Unao 0432 Output Resistance 100 300 me bE A ote dim flee mn bh A a e ce bu ce n ERTE Output Resistance 50 S06 Frequencies are 2099 94MHz to 2300 06 Hz 7 Under the Options tab we should select the manufacturing process For this exampl
70. This technique is crucial for all frequency translation models such as mixers multipliers and dividers Noise Point Removal As more and more noise points are added to the simulation the simulation will become slower and slower For this reason noise points that add no value ate removed from the spectrum NOTE The Calculate Noise option must be enabled before noise figure measurements will be added to the path dataset 469 Simulation 470 Cascaded Noise Figure Equations The traditional cascade noise equation is as follows Feascade Fy Fo 1 Gi F3 1 G1G Fs 1 GiG2 Gn 1 where F is the noise factor of stage n and Gi is the linear gain of stage n Note This equation contains no information about e frequency e impedances or VSWR see Two Port Amplifier Noise e bandwidth e image frequencies e or multiple paths These limitations are very restrictive and lead to additional design spins For these reasons traditional cascaded noise figure equations are NOT used in Spectrasys A more general formulation is used which will include all the effects of frequency VSWR bandwidth images and multiple paths and will reduce to the traditional case under traditional assumptions The general formulation is Cascaded Noise Figure n Channel Noise Power n Channel Noise Power 0 Cascaded Gain n dB where n stage number Cascaded Gain n Desired Channel Power n Desired Channel Power 0
71. Total Node Voltage TNV This measurement is the peak average voltage of the entire spectrum at the node This is an extremely useful measurement in determining the total voltage present at the input of a device This measurement includes ALL SIGNALS INTERMODS HARMONICS NOISE and PHASE NOISE traveling in ALL directions through the node Default Unit dBV Channel Used No channel is used for this measurement Types of Spectrums Used All SIGNALS INTERMODS HARMONICS NOISE and PHASE NOISE Travel Direction All directions through the node Voltage DC VDC This measurement is the DC voltage along the specified path This measurement includes ALL SIGNALS INTERMODS HARMONICS NOISE and PHASE NOISE traveling in ALL directions through the node Channel Used No channel is used for this measurement Types of Spectrums Used A SIGNALS INTERMODS HARMONICS NOISE and PHASE NOISE Travel Direction All directions through the node 531 Simulation Troubleshooting General RF Architecture Troubleshooting These steps can be followed for general RF architecture troubleshooting Compare Measurements to Get an Understanding of the Problem 1 Level diagrams are not as useful as tables when debugging problems A default table can be added to the workspace by right clicking on the node at the end of a path and then choosing System1_Data_Path1 New Table of Measurements from the Add New Graph Table submenu 2
72. Travel Direction Same as CP Added Noise AN This measurement is the noise contribution of each individual stage in the main channel along the specified path as shown by AN n CNF n CNF n 1 dB where AN 0 0 dB n stage number This measurement is simply the difference in the Cascaded Noise Figure measurement between the current node and the previous node This measurement is very useful and will help the user identify the contribution to the noise figure by each stage along the path Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as CNF Travel Direction Only spectrum traveling in the forward direction are included in this measurement Carrier to Noise Ratio CNR This measurement is the ratio of the Desired Channel Power to Channel Noise Power along the specified path as shown by CNR n DCP n CNP n dB where n stage number Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as DCP and CNP Travel Direction Same as DCP and CNP 503 Simulation 504 Carrier to Noise and Distortion Ratio CNDR This measurement is the ratio of the Desired Channel Power to Channel Noise and Distortion Power along the specified path as shown by CNR n DCP n NDCP n dB where n stage number Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as DCP an
73. a Designs EN MH Part Layout be E Part Layout eff SPIRAL Schem i COMBINE Sche o C4 INDUCTOR Sel B X Simulationz D ata 0 RS EMPartl Parti S EMPart2 Part o iS Linear D to 100 2e Outputs a Substrates 5 mer al Freq MHz Error m The results from this are shown above Notice that even with only 5 analysis points across the band the interpolation is very good To illustrate this the spiral inductor was recalculated with 10 points below You can see quite good agreement between the two To test the validity of the decompositional analysis the entire spiral was analyzed and the results are given in the second figure below This full analysis took hours on a 266MHz Pentium II and if the lengths of the lines in the spiral are changed it must be rerun DB E21 DB E11 149 Simulation 4 GENESYS V7 0 Graph Workspace Combine E a Designs M Part Layout i Part Layout p SPIRAL 5chem COMBINE Sche P0 INDUCTOR Sel 4 y Simulations D ata 0S EMPartl Parti DB E21 DB E11 J Sij Linear 0 to 100 ze Outputs Freq MHz 150 EMPOWER Planar 3D EM Analysis 4 SuperStar V6 5 FULL SCH Response EMPOWER 50 m File Edit Tune Optimize Stats Utils Window Export LAYOUT empower Shell Help e x L1 MRIN1 L2 MRIN1 L3 MRIN1 W MRIN1 2 S MRIN1 2 LIND3 4 n j B B5 Bb pem ae Litt tt
74. a Sinele Normal BORUS as Gotestusto tun titus tocum aba tashed obsit cout irn tou Ead 312 AVOINE O yei Doe 312 7 NPP Vine Reterenece C EESEES seis tuastt itn Pho tien teint tos No At PIER Eu Oa CDI bI e odd 313 Pllonvine for Coupling EHC oati aa nu het tes NONA 314 Peine au Internal POR nae ea iod dedii beato ma rd ENE 316 Dehnine a Ditterentab POL basado ted ento tie hea A usta uo eu dd bac Ute unep ient 318 Define axoplanat POS don iota c iR Dodd Dave rrr ditis errr errr 321 Pennino a Common Mode POr urisi eti n E E E EE 326 ID tinine d OroUnd Ke FONC aiara a eode Dec r E N 329 Append Drawno VIAS anar aana betta T EA EE EO 332 Vids thtoue Multiple Substrate LAYE dssdo n AA 333 Thick Condactor Mode liio aa eati uet ud aque REG 337 Modeling Horizontal Side Currents for Thick Conductors esses 340 TE S AEE E I E dst cigs was et aes Soa ies aca aa ese aeons ea emcee 341 BUD Mie qe 342 MewinsLavout Laveretunps ea DOXan bateau iusta Uu tst Nu tov iuit 344 Aout DOX6ES tameii ird EO E ud temores ido o abest 345 Adding Absorbine Laver under a COVefsu destrui ostiis N 346 Doxescamdadiatton Datte FS eoim tS itunes ita ecce eae m E EE 346 Me EN Td 346 Definitio d MeSbisenee ta bcne oit eio eevee cinta tuse ltd s A 346 Defining Mesh Parameters for the Entire CECulbie dee boi ios oi 347 Calculatie Ere Vies iraniane a T A 351 Detinine Mesh Pa
75. a uu 59 Simulation Chapter 3 Chapter 4 Chapter 5 vi bableware Verlor A Extensions siguessen TUE UI pair uum d E D E 61 CAYENNE Transient Analysts io etr HS ode ab D D Hn EM M MN RI ERU 65 C AX BEININE OVON CW dient boom det ath dde ie ie aaa teed 65 CAYENNE Walt yEOUO TE ciiideoea ceti ra Rho aad A c ie OA tutor no tutte 65 DUM ATOR Tine Ste S ooo ios hte ttai stade NA istud bad inia uides eto ides 69 Numerical PESCISIOP opsein deci ioa aem E E 71 Lrequencysdependent MOGEIS cea censeri ba a E a Ronde n ios eoru e a2 COT O E Suse m 75 Inteeratorand Predietefzususte testudine Te ee Ter eMart Te oe ta tide marr yer eer eke he 76 AVP ODO INS ect oda t asec ag lee dde insu tae cip en aM dept t eue duae DD CAMA S18 assesses sade M 85 IB peter aS adl ms 85 1D Cw malysis POPETI Sarii tor un ose raheem Aceh te mas tao eso shin eee kenaisolns 86 Ic zunalvsis OUP OE teen pti cm Reve Perper ceet corer rere bm ed o Somer err ear er er 89 EMPOWER Planar 3D EM Analysis iciicscccccsccenscsecsvacdsanontecaesesesoncendescunsnuvevaenuedeness 91 PVE OVE RAB ASCs nectit T A qoasi Reb oatvenudaamsseutbuanaederis 91 cu 91 CABE S au dam uada te dete cc e eg acetate acta Vtde c dans LUIS 91 Pasie CONS Uy M rM IM 92 POE A N E CM ed car nnd CM mda aM NND 97 Viaholes andi Z Directed POTES eerte tb n a a 100 EM olur n A E AE T A IN 101 EMP ONER ODIOO estes oa cas oaa a E EN AEE 101 G oneal Ta Dna T E E RE 101 Miewer Pa
76. analyzer that has a sweeping receiver and peak detects the total power within the resolution bandwidth The user can specify the resolution bandwidth of this sweeping receiver The default resolution bandwidth is the Measurement Channel Bandwidth Filter Shape This parameter determines the shape of the resolution bandwidth filter This filter shape is analogous to the resolution bandwidth filter shape in a spectrum analyzer which uses a 5 pole Guassian filter Likewise in the system analysis this same filter is also used The user is able to select three widths for this particular filter which are based on an integer number of channel bandwidths No spectrum integration will occur outside the width of this filter This filter width is used to reduce the amount of data collected saved and processed A brickwall filter can be created theoretically and is also included Brickwall Ideal This filter is an ideal rectangular filter whose skirts are infinitely steep Gaussian to 100 dBc 30 Chan BW Data will be ignored that is farther than 30 channels away from the center frequency Attenuation 30 channels from the center will be about 100 dBc Gaussian to 117 dBc 60 Chan BW Spectrasys System Data will be ignored that is farther than 60 channels away from the center frequency Attenuation 60 channels from the center will be about 117 dBc Gaussian to 150 dBc 200 Chan BW Data will be ignored that is farther t
77. and BW 1 MHz then the resulting bandwidth would be 2 03 MHz The user needs to make sure that the Channel Measurement Bandwidth is set wide enough to integrate all of this energy 1 Jose Carlos Pedro Nuno Borges Carvalho Intermodulation Distortion in Microwave and Wireless Circuits Artech House 2003 Spectrasys System Cascaded Intermod Equations Cascaded intermod equations are NOT used by Spectrasys There are serious drawbacks using these cascaded equations Background Using basic assumptions the intercept point for a cascade can be determined Cascaded equations come in two flavors coherent and non coherent If the intermods throughout the cascade are assumed to be in phase then coherent addition should be used This will yield a worst case intercept point However if the intermods are assumed to be out of phase then non coherent addition can be used Cascaded Intercept Coherent Addition 1 ITOIcascade 1 ITOH 1 ITOI2 G1 4 1 ITOIn G1G2 Gn 1 Cascaded Intercept Non Coherent Addition 1 ITOlcascade 2 1 ITON 2 1 ITOD2 G1 2 1 ITOIn GIG 6nd Where ITOI Numeric Stage Input Third Order Intercept G Number Stage Gain See McClaning K and T Vito 2000 Radio Receiver Design Atlanta GA Noble pp 605 626 for additional information The basic assumptions ate 1 There is no concept of frequency 2 All stages have been perfectly matched 3 No cons
78. antenna above a ground plane with no substrate the Air Below layer should be set to the height the antenna is to be above the ground plane The Bottom Cover should be set to Lossless type and the Top Cover should be set to Electrical type with surface impedance set to 377 ohms To simulate a microstrip antenna the Air Below layer should not be used The substrate layer instead should be used The Bottom Cover should be set to Lossless type and the Top Cover should be set to Electrical type with surface impedance set to 377 ohms EMPOWER Options ET E X General Viewer Far Field Advanced v Generate Viewer Data slower Fart number to excite 1 Mode number ta excite f v Generate Far Field Radiation Data M Sweep Theta Start Angle fo Stop 180 Step f degrees M Sweep Phi Start Angle fo Stop 30 Shep f degrees Cancel Apply Help Specifying Sweep Parameters In order to generate far field radiation data Generate Viewer Data slower and Generate Far Field Radiation Data must be checked You may then select either Theta Phi or both to 167 Simulation 168 be swept Data is generated for all points between Start Angle and Stop Angle for both Theta and Phi with a step size specified in the Step field All angles are in degrees In the above figure data is being generated sweeping both Theta and Phi Theta is being swept from 0 to 180 degrees in 1 degree increments
79. base frequency F1 and 3 order IM component with frequency Fim3 2 F1 The Intercept Point for input power scale hb iipn SpectrPout FregIn dexIM IndexS1 IndexS2 PindBm The Intercept Point for output power scale hb_oipn SpectrPout FreqIn dexIM 248 HARBEC Harmonic Balance Analysis F2 1958 5MHz using HB2 Amp NL IndexS IndexIM Pfundd dbm P2 1959 5 Example Pintermod dbm P2 1958 5 IP3 TOI for 2 tones Pdiff Pfund Pintermod signal between spectral TOI Pin Pdiff 2 components of base frequency F1 and 3 order IM component with frequency Fim3 2 F1 1 F2 OIP3 hb oipn P2 FregI ndexIM 1 0 2 1 For mixer with 2 tone RF signal 2 RF and 1 LO frequencies signal frequencies vector F Frfl Frf2 Flo Output IP3 OIP3 hb oipn P2 FregIn dexIM 1 0 1 1 2 1 The Input IP relative to power in dBm of 1 of RF tones Prf_dBm IIP3 hb iipn P2 FregIndexIM 1 0 1 1 2 1 Prf dBm D 1 tone HB oscillator analysis Multi tones HB oscillator analysis oscillator with external signals synchronized oscillator self oscillating mixer d the HB analysis properties window flag Calculate Wave Data must be set the HB analysis properties window flag Calculate Port Wave Data must be set any of or flags must be set in order that the dataset has created the independent variable Time HB Oscillator Analysis Oscillator Design O verview Oscillator design b
80. by setting Maximum Amplitude Step to the ideal step Krylov Subspace Iterations When the Jacobian matrix gets very large it can become very slow to calculate and use Krylov subspace iterations can dramatically reduce the size of the matrix and thus speed 235 Simulation up calculations of very large circuits In general however Krylov will have more convergence issues than full Jacobian steps Also for smaller circuits Krylov may be slower than full Jacobian steps For very large problems try selecting Krylov to reduce memory requirements and speed convergence Genesys 2006 HARBEC Dataset Variables Freq An array nFout of the HB solution frequencies FreqID A string array nFout of ID s of the HB solution frequencies FregIndexIM An array nFout of index vectors nFin defining the harmonic order of each of the signal source frequencies in creating of the HB solution frequency For example for a 1 tone HB analysis nFin 1 and the elements of the array are scalars equal to the harmonic number of the HB frequency H Hel Data e UE FregindexIM no Gain 7 663 db 10 hb transgain P 1 P 2 FregIndexTM 1 11 FE Gain 2 2 14212 j1 11808 b transgain VPORT 1 VPORT 2 Freaindexim 1 1 amp amp yP1 PPORT 1 cm dx a 3 n S amp P 2 PPORT 2 ij Pout1 29 015 dbm P2 2 Hay PRORT S3 VPORT a Wt VPORT FS ZPORT Figure 1 For a 2 tone HB analysis the FreqIndexIM is an array o
81. can be placed at the connection point so even though the device is not part of the circuit Momentum GX you ate simulating the coupling effects that occur among the ports and around the device will be included in your simulation Internal ports are often used in conjunction with ground references For more information refer to Defining a Ground Reference and to Simulating with Internal Ports and Ground References An internal port type has the following properties e Itcan be applied to the interior of a circuit by applying it to the surface of an object e It can be applied to the edge of an object e It can be applied to objects that are on strip layers only e The orientation of the port is not considered if it is on the surface of an object If an internal port is placed on an edge it is oriented to be perpendicular to it e No calibration is performed on the port Because no calibration is performed on the port the results will not be as accurate as with a single port However the difference in accuracy is small Pors 1 and 2 are interna ports This design has 2 internal ports 1 and 2 To define the port mode set the Port Type to Internal 317 Simulation Copl4strip EM Port Properties Draw Size Mam Ref Plane Shift Port Number 1 Location Layer B TOP METAL Width jo Line Direction Current Direction Height Pork Type Polarity Single mode STRIP port transmission
82. capacitors and inductors in the network are complex frequency dependent and mutually coupled as all Momentum GX basis functions interact electrically and magnetically Figure 3 The ground in this equivalent network corresponds with the potential at the infinite metallization layers taken up in the layer stack In the absence of infinite metallization layers the ground corresponds with the sphere at infinity The method of moments interaction matrix equation follows from applying the Kirchoff voltage laws in the equivalent network The currents in the network follow from the solution of the matrix equation and represent the amplitudes of the basis functions Figure 2 The equivalent circuit is built by replacing each cell in the mesh with a capacitor to the ground reference and inductors to the neighboring cells Figure 3 Equivalent network representation of the discretized MoM problem The Momentum Solution Process Different steps and technologies enable the Momentum solution process e Calculation of the substrate Green s functions 377 Simulation 378 e Meshing of the planar signal layer patterns e Loading and solving of the MoM interaction matrix equation e Calibration and de embedding of the S parameters e Reduced Order Modeling by Adaptive Frequency Sampling Calculation of the Substrate G reen s Functions The substrate Green s functions are the spatial impulse responses of the substrate to Dirac type excitat
83. channel All adjacent channels are relative to the main channel identified by the Channel Frequency and Channel Measurement Bandwidth Consequently channels exist above and below the main reference channel frequency The user can specify which side of the main channel the adjacent channel is located on along with the channel number The channel number is relative to the main channel Therefore channel 1 would be the first adjacent channel channel 2 would be the second adjacent channel and so on U Upper Side L Lower Side Spectrasys System n Channel Number any integer gt 0 For example ACPL2 is the power of the second adjacent channel below that specified by the channel frequency If CF was 100 MHz and the channel bandwidth was 1 MHz then the main channel would be 99 5 to 100 5 MHz Consequently then ACPL2 would then be the integrated channel power between 97 5 and 98 5 MHz and ACPL1 would be the integrated channel power between 98 5 and 99 5 MHz Note Only the first 2 adjacent channels on either side of the reference channel are listed in the Measurement Wizard However there is no restriction on the Adjacent Channel Number other than it must be non negative and greater than or equal to 1 This measurement is simply a Channel Power measurement at the Adjacent Channel Frequency Channel Used Corresponding Adjacent Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as CP
84. condition It comprises information about 215 Simulation 216 all possible structures that could be formed by different combinations of the additional conditions The boundary condition superimposing can be represented as a set of simple manipulations with the informational multiport terminals We have added this section to clarify connections of the numerical electromagnetic solution with the circuit theory This technique is also known as the impedance interpretation of boundary condition superimposition The GGF matrix obtained in the previous section can be represented as an impedance matrix Z of a multiport shown on the left below Ft u ideal aa A EN metallization ff ff _ a ff SUTA ce Eumped efement The multiport terminals are conceptual and their positions are just a schematic representation Four conceptual ports or pairs of terminals correspond to a grid cell as shown in the figure The total number of ports oriented along the x axis is M L 1 The total number of ports oriented along the y axis is L M 1 The multiport can also have a set of z directed ports corresponding to via holes or z directed internal inputs Note that we do not need to calculate all elements of the multiport impedance matrix and its order can be reduced taking into account that some ports are no loaded or short circuited The no loaded terminals correspond to regions of the signal layer without any conductivity currents The right h
85. convention MAG S21 which is linear and DB S21 which is the decibel form With reflection parameters the linear form is often referred to as a reflection coefficient and the decibel form as return loss Si1 dB input reflection gain 20 log S11 S22 dB output reflection gain 20 log S22 S21 dB forward gain 20log 21 S12 dB reverse gain 20log 12 S21 and S12 are the forward and return gain or loss when the network is terminated with the reference impedance The gain when matching networks are inserted at the input output or both ts described later Si1 and S22 coefficients are less than 1 for passive networks with positive resistance Therefore the input and output reflection gains S11 and S22 are negative decibel numbers Throughout Eagleware material the decibel forms S11 and S22 are referred to as return losses in agreement with standard industry convention To be mathematically correct they have been left as negative numbers As such the rigorous convention would be to call them return gain Input VSWR VSWR and S11 are related by VSWRi 1 Su C1 Su The output VSWR is related to S22 by an analogous equation A circle of constant radius centered on the Smith chart is a circle of constant VSWR The complex input impedance is related to the input reflection coefficients by the expression Ii Z 14 S811 1 S11 The output impedance is similarly related to S22 Stability Because S12 of devices i
86. dB Basically the way to look at this is the cascaded noise figure is equivalent to the noise power added between the first and last stage minus the cascaded gain Channel noise power includes the amplified noise as well as the noise added by each model Therefore cascaded gain must be subtracted to get just the noise added by the system Coherency Since all signals in Spectrasys are treated on a individual basis so must coherency for each of the spectrums created during the simulation Coherent signals will add in voltage and phase whereas non coherent signals will add in power For example if two coherent voltages had the same amplitude and phase the resulting power would be 6 dB higher If they were exactly 180 degrees out of phase having the same amplitude the two signals would cancel each other If the two signals where non coherent then the power would only increase by 3 dB irrespective of the phase Spectrasys System Some of the coherency of Spectrasys can be controlled by the user The user can determine whether intermods and harmonics add coherently and whether mixer output signals consider the LO signal when determining coherency See the Calculate Tab of the System Simulation Dialog Box for more information on this setting How it Works When a new spectrum is created a coherency number is assigned to each spectrum These coherency numbers are used to group spectrums together to determine what the resulting total spectrum
87. data since the ports are a more accurately represent the physical connection of lumped elements The circuit shown in below contains an EMPOWER circuit which was drawn completely in LAYOUT The schematic for this network was blank It has 3 ports ports 1 and 2 are external and port 3 is internal with current direction Along X EMPOWER will create a 3 port data file for this circuit however you must be aware that port 3 will be a series connected port and cannot be used in the normal mannet The data file created by EMPOWER can then be used in GENESYS as described in the previous section using WSP Simulations EM1 EMPOWER SS The circuit on the right uses the resulting data in a complete network First a THR three port data device was placed on the blank schematic using the EMPOWAER SS file from the EMPOWER run An input and output were added on nodes one and two of the THR block the ground was added to the ground node and a capacitor was connected from port 3 to ground 156 EMPOWER Planar 3D EM Analysis This has the effect of putting the capacitor across port 3 in the EMPOWER simulation The rules to follow for Along X and Along Y internal ports are simple e Do not attempt to use them for transistors or other 3 terminal or more devices e Set the Current Direction of the EMPort to Along X along the x axis if the current along the component flows from left to right as on the layout on the lef
88. determining the total power present at the input of a device i e amplifier or mixer LO This measurement includes ALL SIGNALS INTERMODS HARMONICS NOISE and PHASE NOISE traveling in ALL directions through the node Channel Used No channel is used for this measurement 523 Simulation 524 Types of Spectrums Used All SIGNALS INTERMODS HARMONICS NOISE and PHASE NOISE Travel Direction All directions through the node Virtual Tone Channel Power VTCP This measurement is the virtual tone channel power The virtual tone is the power level of an un attenuated tone along that cascade of stages This tone consists of the power level of one of the tones at the input of a cascade plus the in channel cascaded gain See Intercept Measurements in the Lab and Cascaded Intermods and SPECTRASYS for additional information VTCP n ICP 0 CGAIN n dB where n stage number Channel Used Same as ICP and CGAIN Types of Spectrums Used Same as ICP and CGAIN Travel Direction Same as ICP and CGAIN Voltage Voltage Gain GAINV This measurement is the voltage gain of the main channel along the specified path The Voltage Gain is the difference between the Desired Channel Voltage output of the current stage minus the Desired Channel Voltage output of the prior stage as shown by GAINV n DCV n DCV n 1 dB20 where GAINV 0 0 dB20 n stage number See the Desired Channel Voltage measurement to
89. different lumped sources used in the calibration process to feed the circuits with calibration lines a grounded source and a floating source The grounded source works well for low frequencies however at higher frequencies when the port ground distance becomes electrically large this source provides less accurate results due to unwanted substrate coupling in the calibration process The floating source works well at higher frequencies unwanted substrate coupling is reduced however it fails at low frequencies because the capacitive internal impedance of the source blocks the flow of the low frequency currents Momentum automatically switches between these two soutces depending on the frequency range of the simulation Determining the Port Type to Use 310 There are seven port types in Momentum The purpose of potts is to inject energy into a circuit and to allow energy to flow into and out of a circuit The different port types enable you to tailor the ports in your circuit according to your type of circuit and its function in the circuit In general you should select the port type that best matches the intended application of your layout The table below gives a brief description of each port type You can use a combination of port types in your circuit although you should note that port types have limitations on where they can be applied Only the Single port type can be applied to objects that are on either strip or slot metallization
90. expressions and shall evaluate to a positive or negative integer or Zero Example real X 1 10 Tox Xj Ces net discipline Advanced Modeling Kit The net_discipline is used to declare analog nets and for declaring the domains of digital nets and regs A net is characterized by the discipline that it follows A net is declared as a type of discipline and so a discipline can be considered as a user defined type for declaring a net A discipline is a set of one or more nature definitions forming the definition of an analog sional whereas a Nature defines the characteristics of the quantities for the simulator A discipline is characterized by the domain and the attributes defined in the natures for potential and flow The discipline can bind e One nature with potential e One nature with potential and a different nature with flow e Nothing with either potential or flow an empty discipline The disciplines are typically predefined in the disciplines vams file a portion of which is shown below Electrical Current in amperes nature Current units A access I idt_nature Charge ifdef CURRENT_ABSTOL abstol CURRENT ABSTOL else abstol 1e 12 endif endnature Charge in coulombs nature Charge units coul access Q ddt_nature Current ifdef CHARGE ABSTOL abstol CHARGE ABSTOL else abstol 1e 14 endif endnature Potential in volts nature Voltage
91. fioi 149 8GHz e 2 4 n EMPOWER Planar 3D EM Analysis For example in air er 1 0006 with a 2 4 inch 0 5 inch high box b 101 6mm a 50 8mm and h 12 7mm Then amp 797 69 14 and f797 3297 MHz Higher Order Box Modes It is interesting to note that if lt a and 4 lt d then the frequency of the dominant mode is not a function of the cavity height This is not the case for certain higher order modes The mode which is next higher in frequency than the dominant mode is a function of the relative values of 7 a and b Consider for example the previous 2x4x0 5 inch box or any size box with the size ratios b 2a and 1 4 Therefore the wave numbers are k m 16r 2 a 4 J The wave numbers for the lowest frequency modes for this shape box and the resonant frequencies with 4 2 inches are listed here Freq MHz a 2 Mode Wave inches Notice that higher order modes occur frequently after dominant mode resonance It is possible to minimize perturbations in narrowband applications by operating between resonant frequencies However the above analysis assumes a pure homogeneous rectangular cavity and dielectric Partial dielectric loading and signal metal within the cavity will influence the frequency A more conservative and safer approach is to enclose the circuit in a box with the dominant resonant mode higher than the highest frequency of interest 159 Simulation 160 Partial Dielectric Loading If the
92. fits the actual current the more accurate the results Thus a dense mesh generally provides better results at microwave frequencies The mesh generator attempts to apply an optimal pattern of cells so that an accurate simulation can be achieved with a minimal number of cells In Momentum mode the minimal number of cells can be rather high resulting in a dense mesh For a given mesh density the mesh reduction technology achieves an optimal mesh with significantly fewer cells hence improving the simulation performance with reduced memory usage and simulation time Since the current in each cell is calculated in a simulation a very dense mesh can increase simulation time The mesh parameters that you specify provide the mesh maker with the information required to divide your geometry into the various cell shapes and sizes If you do not want to set mesh parameters default ones will be used instead The default mesh setting is 30 cells wavelength using edge mesh Topics in this section Adjusting Mesh Density Effect of mesh reduction on simulation accuracy About the Edge Mesh About the Transmission Line Mesh Combining the Edge and Transmission Line Meshes Using the Arc Resolution Mesh Patterns and Simulation Time Mesh Patterns and Memory Requirements Processing Object Overlap Mesh Generator Messages Adjusting Mesh Density Two mesh parameters Mesh Frequency and cells wavelength are used in combination in determining mesh density
93. for 2 tones with equal amplitude Channel Used Interferer Channel Frequency Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as ICP CGAIN and TIMCP Travel Direction Same as ICP CGAIN and TIMCP Percent Noise Figure PRNF This routine calculates the Percent Noise Figure contribution by each stage to the final Cascaded Noise Figure of the path PRNF n AN n CNF nLastStage 100 this is a ratio of dB values where PRNF 0 0 n is the current stage and nLastStage is the last stage along the designated path This measurement will help the user pinpoint all stages and their respective contribution to the total cascaded noise figure of the selected path This measurement in unit less since the measurement is a percentage There can be a few cases where the percentage sum of all the stages in the path does not equal 100 For instance if the architecture contains parallel paths then each path would contribute to the total cascaded noise figure but only a single path is considered in the measurement Another case would be where there is sufficient VSWR interactions between stages that effect the noise so it does not change linearly with the gain Reducing the architecture to Spectrasys System the spreadsheet case will always yield the expected spreadsheet answers with respect to percentages See the Cascaded Noise Figure measurement for additional information Channel Used Main Ch
94. gamma opt S CS 2 1 Complex impedance ZPORT 1 2 1 NEMIN inimum noise figure noise nfmin S CS 2 Scalat oise Circles noise_circles S CS Complex 0 25 0 5 1 1 5 2 2 5 29 6 2 D vector two entries for evety circle including zero dB All noise measurements below use port i as input and port j as output noise figure S CSy 7 Scalar oise Resistance noise_rn S CS ZPORT 7 2 Scalar oise optimum noise_gamma_opt S CS 7 2 Complex reflection coefficient zin noise gamma opt S CS 7 2 Complex ZPORT 1 303 Simulation 304 yin noise_gamma_opt S CS 7 2 Complex ZPORT 1 noise_nfmin S CS 7 2 Scalar noise_circles S CS 0 25 05 1 1 3 2 2 5 3 6 72 Simultaneous match sm_gammaz S d gamma at port 7 2 port Simultaneous match yi Complex admittance at port 7 2 port circuit only Simultaneous match e Complex impedance at port 7 2 port circuit only Unilateral gain circles at Jou_circles S 1 2 3 4 5 6 z gp circles S 1 2 3 4 5 6 GA Available gain circles ga citcles S 1 2 3 4 5 6 All circle measurements are in packed complex vector form Each circle including any zero radius circles for zero dB is represented by two complex entries The first entry is the center of the circle The second complex entry has the radius of the circle in the real part and extra information the dB level of the circle in the imaginary pa
95. greater than zero B1 1 Siu S22 2 D 7 gt 0 Note See the section on S Parameters for a detailed discussion of stability analysis Values Real value versus frequency Simulations Linear Default Format Table Linear Graph Linear Smith Chart none Commonly Used Operators None Examples Measurement Result in graph Smith chart Result on table optimization or yield K stability factor stability factor B1 stability measure stability measure Not available on Smith Chart Input Output Plane Stability Circles SB1 SB2 A output stability circle is a locus of load impedances for which the input reflection coefficient 11 1s unity This locus which is specified via a marker is plotted on a Smith chart and is only available for 2 port networks This locus is a circle with radius Rout about a point Cout where 25 Simulation 26 Rout S12821 822 D Cou S22 DSi1 aa D 9 The region inside or outside the circle may be the stable region The filled areas of the graphs are the unstable regions The input plane stability circle equations are the same as the output plane equations with 1 and 2 in the subscripts interchanged If SB1 and SB2 are placed on a table you can see the PAR value If it is zero then the region is outside the circle is stable If it is 180 then the region inside the circle is stable Note See the section on S Parameters for a detailed discussion of stability a
96. higher frequencies where the device S12 is typically larger this assumption is less valid The assumption simplifies manual and graphical design but is unnecessary in modern computer assisted design The assumption also allows factoring the above equation into terms that provide insight into the design process If 12 0 then Gta Sai 1 RAA Rc 2 1 Su R1 S2R1 where Gtu unilateral transducer power gain When both ports of the network are conjugately matched and S12 0 Gtu 3 2 1 S 3 1 S22 2 The first and third terms indicate the gain increase achievable by matching the input and output respectively If 81 or S22 approach 1 substantial gain improvement is achieved by matching Matching not only increases the network gain but reduces reflections from the network When network gain flatness across a frequency band is more desirable than minimum reflections the lossless matching networks are designed to provide a better match at frequencies where the two port gain is lower By careful design of amplifier matching networks it is possible to achieve a gain response flat within fractions of a decibel over a bandwidth of an octave or more Gain Circles When the device is complex conjugately matched the transducer gain is Gmax and if the device is terminated with the same resistance used to measure the device S parameters the transducer gain is S21 The gain with arbitrary terminations can be visualized
97. in GENESYS Interface Category 1 Sonnet features directly affecting frequencies geometry and subsectioning 401 Simulation 402 Auto Grounded ports and their reference plane shifts Polygons Descriptions e Diagonal and conformal fill types always staircase e Min Max subsection size xmin xmax ymin ymax e Conformal mesh subsection maximum length e Edge mesh setting always on Geometry Subdividers Parallel Subsections Metal Types and Features e Current ratio e Rdc Rtf type e General Metal rdc rrf xdc ls e Sense metal e NumSheets is always set to 2 for thick metal Brick polygons and their dielectric materials Magnetic loss in a dielectric Z Parts parameter for dielectric layers is always set to 2 from GENESYS All advanced amp multiple frequencies sweeps are not supported e Exponential sweep e External file for frequencies e Find maximum and minimum frequency response Response file formats eg Databank Scompact Excel CSV SPICE netlist generation Ports with negative and or duplicate numbers Push Pull etc Category 2 Features available in GENESYS other ways Port impedances in Sonnet file Viewer data defaults to 50 ohms in Sonnet Geometry parameters Optimization Parameter Sweep Sonnet Interface Category 3 Cosmetic Sonnet features which do not directly affect simulation Edge Vias use regular viaholes or Z directed polygon layers instead Medium and High importance file change dates Comments
98. in Momentum GX another dimension is added to the object in order for the shape to cut through the substrate Thus a line becomes a sheet a circle becomes a cylinder representations of vias in Layout am vias extruded through a substrate layer in Momentum Regardless of how you draw vias avoid having them extend over the sides of the objects that they connect to Vias must be on the edge or inside the object Any portion of the via outside of the object boundaries will not be taken into account during the simulation The fioure below illustrates various vias connecting two strips Momentum GX Incorrect vias overlap the object edges Efficient via sheet via with width matching wicth of strip Circular Via EE DES correct hut Correct but may not be efficient notas efficient Use an appropriate as sheet via arc ree amp agnition with matching wicth to strip Vias ate treated in the following mannet e Vias are represented as infinitely thin vertical sheets of metal If a cylinder is drawn as a via then it 1s also treated as separate sheets of metal that comptise the cylinder walls e Vias ate assumed to be opened not covered with current traveling on the sheet Vias can be drawn as covered using a cap cover of the same size as the via end which is a terminating metal strip layout layer e Avia cannot be coincident with another via if they cut through the same substrate layer and they
99. in file Polygon snap for reference planes Metal pattern ID selection Sweep with step size use number of points instead Dimensions on the geometry EM PO W ER features not supported by the Sonnet Interface The following EMPOWER features are not supported by the Sonnet Interface Surface Roughness for metal Slot mode layers Magnetic cover Multi mode ports and the multi mode transmission line circuit model X only and Y only current direction XYZ and Z directions are supported Thick metal going down Thick metal model for lossless metal or resistive films Line direction determined automatically from touching polygon Different reference plane shifts on the same sidewall Negative reference plane shifts reference planes outside the box Using specific HARBEC simulation frequencies HARBEC co simulation is otherwise fully supported Mixing standard and no deembed ports on the same layout Component pads will not use the Current Direction setting for the layer and are always output using XY currents In other words component pads on XYZ or Z layers will be output with the standard XY Current Direction Pads manually placed not part of a component still respect the direction setting Note that this behavior is different from EMPOWER where component pads also respect the current direction setting 403 Simulation Tutorial 404 Introduction The Genesys Sonnet interface has been created to allow Eagleware and Sonnet
100. in the X and Y dimensions I zs strongly recommended to turn this checkbox on whenever you are creating a layout for EMPOWER Grid Spacing X and Grid Spacing Y These control the cell size for the EMPOWER run as well as the grid snap feature in LAYOUT When using the EMPOWER Grid Style there will be LAYOUT snap points between each grid line which allow lines to be centered between two grid points if necessary They are often referred to as dx and dy and should be small with respect to a wavelength at the maximum frequency to be analyzed preferably less than wavelength 20 and always less than wavelength 10 Box Width and Box Height These are the box size for EMPOWER simulation They correspond directly to the SIZE statement in the TPL file The number of cells across the 99 Simulation 100 box equal to Width or Height divided by Grid Spacing X or Y is displayed for your convenience and can be changed to adjust the page width Note Any metal put down completely outside the box will be ignored by EMPOWER This can be used to your advantage to temporarily or permanently remove metal or components from the EMPOWER simulation Default Viahole Layers The Start Layer and End Layer combo boxes control the default layers for the viaholes These can be overridden individually for each viahole Viaholes and Z Directed Ports The grid in EMPOWER is a truly three dimensional grid rectangular lattice Z Directed currents and
101. input power level When the on channel signal generator is disabled and two tones are injected into the receiver the power level of the two tones can be adjusted until the S N ratio decreases by 3 dB At this point we know that the power level of the intermod is equivalent to the power level of the on channel signal generator Intercept Measurement Summary The process of measuring the intercept point is very different than using cascaded equations When determining the intercept point in the lab the user must physically create both tones at the frequencies of interest The spacing between the tones needs to be such that the third order intermods will appear at the desired locations The power level of the input tones needs to be sufficiently low to keep the DUT in linear operation but large enough to be seen in the dynamic range of the spectrum analyzer When measuring the intermod and tone power levels the user must select the appropriate frequencies on the spectrum analyzer to place the markers Out of band intermod measurements require an additional step to measure the in channel cascaded gain This is a very different process than using cascaded intermod equations that remain ignorant of frequencies impedances directions and power levels Intercept Points Other Than 3rd Order Thus far only 3rd order intercept points have been addressed However for a two tone soutce the general intercept equation is IPn Ptone Pintermod n n
102. intermod equations and is not restricted to their limitations Measuring intercept points in Spectrasys is akin to making measurements in the lab Spectrasys needs to know both the frequency of the intermod and the frequency of one of the tones These values are specified on a path The path frequency is the frequency of the intermod being measured and the Tone Interferer Frequency is the frequency of the tone used to determine the intercept point Unlike a marker on a spectrum analyzer Spectrasys has the ability to measure the power of a spectrum across a channel Care must be given to ensure the channel bandwidth is wide enough to cover the bandwidth of the intermod yet narrow enough to exclude the power level of any tones used to create the intermods Spectrasys System Because Spectrasys has the ability to differentiate between signals and intermods the two step out of band process can be done in a single step as shown in the following figure The out of band configuration is the general case and will also work for in band intercept measurements signal senerator Test Sig signal Generator Tone 1 spectrum Analyzer signal senerator Tone 2 i ioe Op f rx lest f fa The user can view measurements showing frequencies and power levels of the tones and intermods being used in intercept calculations The following tables show a mapping of Spectrasys measurements to the intermod test setup and results Ta
103. is approximated for other configurations e The two ports must be on the same reference plane The illustration below shows ports pairs applied to a layout of coplanar waveguide slots The coplanar port type is applied to each pair of ports in the design The arrows in this illustration indicate the polarity assigned to each port and the direction of the voltages over the slots arrows indicate the direction of the voltages ground planes normal reversed polarity polarity R Arrows indicate orientation petanty af the volteges over the slots Be careful when assigning polarity to coplanar ports An incorrect choice of polarity can change the phase of transmission type S parameters by 180 degrees T Coplanar ports can be applied to objects on slot layers only To define the Slot metal layer Physical Slot must be checked and the combo box for Momentum Slot type must be set to Slot 322 Momentum GX LAYOUT Properties General Associations Layer Ponts Show Columns Physical 12 Model Stat Height Open 7 a Sun Physica win Tend 46 1 383 10000 CB Losses c BOT METAL BOT SLE BOT MASK ar Below Boton Cover as insert j Showat S Use Al Frequency Units MHz BF Lead From Layer File Delete 5 Bown 3 Hide al H Use None Length Urits ml mm me This means that Momentum GX will use the layer as a physical slot where all metal areas are swappe
104. large corresponds with the number of cells per wavelength But when other details edges and user defined mesh settings are included cells might A 20 appear that are smaller than resulting in more cells per wavelength than the value of 20 that you entered Here are two specific cases where the actual number of cells per wavelength is greater e When the layout has details that are smaller than Az 20 the mesh follows the shape of the details Since the mesh consists of triangles and rectangles only the 4 20 cells will be less than e When you change the default settings in the Mesh Setup Controls dialog box such as the number of cells per transmission line width you directly influence the number of cells over the width of the transmission lines This might lead to cells A 20 that are smaller than Effect of mesh reduction on simulation accuracy Mesh reduction is a technology that aims at removing mesh complexity originating from the meshing of geometrically complex shapes In a normal situation the user specifies the mesh density the number of cells per wavelength that needs to be used in the simulation in order to obtain a specific accuracy However due to geometrical constraints the mesher can be forced to use more cells than strictly needed by the wavelength criterion In this case the mesh introduces redundant degrees of freedom in the solution process redundant with respect to the electromagnetic behavior M
105. line no calibration For internal ports you should specify width and height Defining a Differential Port Differential ports should be used in situations where an electric field is likely to build up between two ports odd modes propagate This can occur when e The two ports are close together e There is no ground plane in the circuit or the ground plane is relatively far away e One port behaves to a degree like a ground to the other port and polarity between the ports is developed e The ports are connected to objects that are on strip metallization layers The electric field that builds up between the two ports will have an effect on the circuit that should be taken into account during a simulation To do this use differential ports Differential ports have the following properties e They can be applied to objects on strip layers only 318 Momentum GX e They are assigned in pairs and each pair is assigned a single port number e ach of the two ports is excited with the same absolute potential but with the opposite polarity The voltages are opposite 180 degrees out of phase The currents are equal but opposite in direction when the ports are on two symmetrical lines and the current direction is approximated for other configurations e The two ports must be on the same reference plane An electric field will likely build up between these two lines Assign the ports as differential ports They will be tr
106. losses 183 Simulation 184 These files are used to specify the impedance of conductors in ohms per square These files are used in the EMPOWER layers setup dialog box or in the TPL file The files are formatted just like RX files in GENESYS SS S Parameter Files Written by EMPOWER Type Text Can be safely edited Yes Average size 5 to 50 Kbytes but may be larger Use Contains S Parameter data calculated by EMPOWER This file contains the S Parameter data written by EMPOWER It is in the industry standard S2P format and can be loaded into most RF and Microwave simulators Even though these files can be edited they will be overwritten whenever EMPOWER is rerun TPL Topology Files Written by User or GENESYS Type Text Can be safely edited Yes Average size 1 to 5Kbytes Use Desctibing circuit to EMPOWER This file contains a complete description of the circuit to be analyzed by EMPOWER GENESYS will create this file automatically whenever EMPOWER is run from the EMPOWER menu in GENESYS Even though this files can be edited it will be overwritten if EMPOWER is rerun from within GENESYS W SX Workspace Files Written by GENESYS Type Binary Can be safely edited Yes but only using GENESYS Average size 10 to 2 000 Kbytes Use Contains complete simulation graph schematic and layout information from GENESYS Contains a complete GENESYS workspace Y Y Parameter Files Written by GENESYS
107. mode that matches the application RF mode ts usually the mote efficient mode when a circuit e is electrically small e is geometrically complex e does not radiate 307 Simulation For descriptions about electrically small and geometrically complex circuits see Matching the Simulation Mode to Circuit Characteristics For infinite ground planes with a loss conductivity specification the microwave D mode incorporates the HF losses in ground planes however the RF mode will make an abstraction of these HF losses EM Ports Ports enable energy to flow into and out of a circuit Energy is applied to a circuit as part of the simulation process A circuit solved using Momentum can have at the minimum one port Ports are defined in a two step process First ports are added to a circuit when the circuit is drawn Then in Momentum you specify the type of port in order to tailor the port to your circuit This facilitates the simulation process This chapter begins with suggestions to keep in mind when adding ports to a circuit that will be simulated using Momentum The remainder of this chapter describes the various port types in Momentum and gives instructions on how to specify a port type Sections in this chapter Adding an EM Port to a circuit Port Calibration Determining the Port Type to Use Defining a Single Normal Port Defining an Internal Port Defining a Differential Port Defining a Coplanar Port Defining a Common Mode
108. n number of tracing iterations used to find the band Results of the oscillator analysis are saved in the dataset It includes all variables of a regular HB analysis plus variables specific for the oscillator analysis which are 1 The Small Signal oscillator analysis variables 260 e 7 osc HARBEC Harmonic Balance Analysis the complex array of frequency sweep values of circuit node impedance Zosc F connected to OSCPORT SweepFreq the double array of the frequencies at which the Zosc f is calculated the small signal oscillation frequency Foscss 1 Osc Freqs Osc dPhaseZdF the derivative 1 d arg Zosc P dF at frequency F Foscss e Osc Rez Re Zosc Foscss real value of the Zosc F at F Foscss 1 2 The Latge Signal nonlinear oscillator analysis variables 2 e TraceFprobe e TraceIprobe e TraceVprobe atrays of dependences Fosc Vprobe Iprobe Vprobe for tracing values Vprobe 3 1 If the circuit has more then one oscillation frequency the variable is array has the value per each oscillation frequency 2 Specific for oscillator analysis only variables 3 The last point of the arrays saves values of Fosc Iprobe and V probe corresponding to the steady state oscillator solution The table summarizes the difference in syntax for HB oscillator measurements in Genesys 2004 vs Genesys 2006 and later Table 2 Basic
109. node 1 ICP 0 CGAIN i Spectrasys System Spectrasys Voltage Measurements Name Description Syntax Spectrum Equation Type V CGAINV Cascaded Voltage cgain DCV Same as DCV CGAINV i DCV i Gain DCV 0 CNRV Cartier to Noise cnr DCV Same as DCV CNRV i DCVIJ oltage Ratio CNV and CNV CNV IJ CNV Channel Noise CNV Noise Average noise voltage in oltage the channel DCV Channel Voltage DCV Desired Desired average voltage at Desired CF GAINV Voltage Gain gain DCV Same as DCV GAINV i DCV i DCV 1 Where GAINV 0 0 dB ICV Interferer Channel ICV Desired Average interferer voltage NNV ode Noise Voltage NNV Noise Peak average noise voltage at the node OCV Offset Channel OCV Total Avetage voltage at OCF oltage PNCV Phase Noise PNCV Phase Noise verage phase noise Channel Voltage voltage at CF SIIV Stage Input Voltage siiv SOIV Same as SIIV i SOIV Intercept Point All SVGAIN SOIV and 5VGAIN I Orders SVGAIN SIIV2 Stage 224 Order sivn SIIV 2 None STV for 29 Order Input Voltage Intercept Point SIIV3 Stage 3 4 Order sivn SIIV 2 None SHV for 3 Order Input Voltage Intercept Point SIV1DB Stage Input 1 dB sividb Same as SIVIDb SOV1DB i oltage SOV1DB SOVIDB and SVGAIN Compression Point SVGAIN SVGAIN 497 Simulation 498 SIVSAT Stage Input Voltage sivsat Same as Saturation Point SOVSAT SOVSAT and SVGAIN SVGAIN SOIV stage
110. noise parameter Bandwidth The default bandwidth is 1 Hz so that the results have units of Notes Bandwidth is for spectral noise simulation 1 Hz is the recommended bandwidth for measurements of spectral noise power The noise contributor data do not scale with noise bandwidth To Calculate Noise Figure set Simulation Temperature on General page of Harbec options 16 85 C or 290 Kelvin It is the standard temperature for noise figure measurement as defined by the IEEE definition for noise figuration The option Include port noise setting checkboxes Noisy for noise simulation ports tells the simulator to include the contributions of port noise in the analysis of noise voltages and currents Once you have entered these settings you can switch the noise simulation off to speed up intermediate simulations by disabling Nonlinear noise at the bottom of the dialog Your settings will remain in the Noise tabs and become active when Nonlinear noise is enabled again Basic Concepts The results of the harmonic balance simulation HB solution and Jacobian matrix are used to determine the periodic operating point for the nonlinear noise simulation The periodic operating point is the set of steady state voltages and currents within the circuit at the fundamental harmonic and all mixing frequencies Every upper and lower noise sideband is modeled for each large signal spectral component consequently the number of noise freq
111. not directly track current error and will not warn the user if the current is not within a reaonable tolerance For more details on this subject see Ken Kundert s book by Kluwer Academic Publishers The Designer s Guide to SPICE and Spectre Frequency dependent Models Frequency dependent models can cause difficulties for all time domain simulators Traditional SPICE simulation does not allow any use of frequency dependencies In later versions and derivations of SPICE the capability to simulate s domain devices defined by rational polynomials was added While these models are good for simple structures like 72 CAYENNE Transient Analysis filters or first order frequency roll off effects they are of little use for much more complicated models typically found in RF and Microwave simulation such as dispersive and coupled transmission lines measured s parameter data and ideal elements with frequency dependent losses and skin effects Even though CAYENNE is a time domain simulator it has several different strategies for simulating frequency dependent models The two basic methods are approximate models and convolution The process CAYENNE uses to setup each element is as follows 1 Determine whether the model is nonlinear frequency dependent and or time dependent a Frequency dependent models include i ii iii iv Models which use the FREQ variable in equations which define their parameters Internal mode
112. of about 200 ohms are unstable Circles 2 and 3 are also unstable with low resistance and certain inductive source impedances At the output plane on the right at 500 MHz a wide range of inductive loads is potentially unstable Response SPARAM 50 _ i amp MAaG E1 FA a zE1 HA a zEz Bn 1500 qun Enn Bog er ion Enn 1 03644 1 2212 21 9521 S 0TSRS Ti d 4 57954 1 26165 n EzBTaL 1 03645 1 2314 31 9521 S NTSRS n n When designing an amplifier the first step 1s to examine the stability circles of the device without the matching circuit present The grounding which will be present at the emitter or source should be included in the analysis This stability data is used to 1 add stabilizing components such as shunt input and output resistors for bipolars or inductance in the source path for GaAsFETs and to 2 select an input and output matching network topology which properly terminates the device at low and high frequencies for stability In the example above matching networks with a small series capacitor adjacent to the device would insure capacitive loads at low frequencies thus enhancing stability This is probably sufficient for the input However considering that device S parameter data is approximate and since the output plane of this device is more threatening it would be prudent to stabilize this device in addition to using series capacitors 295 Simulation 296 Note Stability should be c
113. orders ICGAIN IIP2 27d Order Input iipn IIP 2 Same as IIP HP for 27 Order Intercept Point IIP3 3rd Order Input iipn IIP 3 Same as IIP HP for 5 Order Intercept Point IPIDB Input 1 dB ipldb Same ast IP1DB i DCP 1 min SDR i Compression DCP SDR DCP amp dBm SDR RX IIP Same as RX IIP i RX_OIPIi RX OIP amp CGAIN where RX_OIP and IP CG AIN contain all orders RX IIP2 Receiver 2 4 ii Same as RX IIP for 2 Order Order Input m RX IIP Intercept Point RX IIP3 Receiver 3rd iipn Same as RX IIP for 3 Order Order Input RX IIP 3 RX IIP 492 ICGAIN ICP IGAIN IMGNP IMGP IMGNR IMGR MDS NDCP OIP OIP2 OIP3 OP1DB Spectrasys System Intercept Point Interferer cgain ICP Same as ICGAIN i ICP i ICP 0 Cascaded Gain ICP Interferer ICP Desired Interferer power at ICF Channel Power IGAIN ICP i ICP i 1 Where IGAIN 0 0 dB Interferer Gain gain ICP Same as ICP Image Channel IMGNP Noise Noise power at IMGF oise Power Image Channel IMGP Total Total power at IMGF Power Image Channel oise Rejection Same as IMGNR i CNP i IMGNP i CNP amp IMGNP imgnt CNP IMGNP Same as IMGR i DCP i IMGP i DCP amp IMGP Image Channel Rejection Same as MDS i CNP 0 CNF i CNP amp CNF inimum Detectable mds CNP CNF Same as PNCP NDCP i P
114. oscillation frequency is the new additional variable of the HB Newton equations system satisfying the tracing equation Im Iprobe1 0 The tracing algorithm starts from very small amplitude of the Vprobe and increases it while Re Iprobel lt 0 After the condition fails the algorithm finds the exact oscillator solution which is located in the limited band between the previous and the last value of Vprobe of the tracing iterations satisfying the condition Re Iprobe1 0 where 1 means the 1st harmonic of the oscillation frequency The Genesys 2006 oscillator analysis method is much more robust and efficient than the one implemented in G4 It allows the Genesys 2006 user to solve many of the oscillator analysis tasks which couldn t be solved in Genesys 2004 The circuit used for the oscillator analysis must have one OSCPORT circuit element which must be connected to a node where the small signal oscillation conditions may be satisfied at a frequency Fosc 257 Simulation Re Zin Fosc lt 0 Im Zin Fosc 0 where Zin is the complex impedance of the circuit node connected to the oscillator port Typically when a tested circuit could oscillate the conditions are satisfied at a node where a resonator is connected to the input or output node of the active device or at the output port of the circuit The following steps must be performed to successfully start the oscillator analysis for a new circuit to be sure that
115. peak spectrum values that may fall in between discretization points This process guarantees that peaks will not be missed 4 The analyzer noise floor will be added 5 The discretized spectrum will be convolved with the gaussian filter 479 Simulation 480 6 Noise randomization is added 7 Analyzer spectrum will be displayed on a graph NOTE Noise floor spectrums between spectrum groups is not shown in the spectrum analyzer mode Frequency Limits Frequency limits have been added as a simulation speedup There is a lot of data and simulation time required for the analyzer mode especially as the frequency ranges become very large The users can specify a frequency band of interest to apply the spectrum analyzer mode to These settings require a start stop and step frequency The step frequency can be set within a given range Warnings will be given to the user if the step size becomes so small that maximum number of analyzer simulation points are exceeded On the other end of the range a warning will also be given if the step size becomes so large that the accuracy has been compromised In these cases the analyzer will select a good default Example The following figure shows a good example of the spectrum analyzer mode This plot is taken from an Image Rejection Mixer example where 2 signals at 70 MHz add coherently to give an increase of 6 dB and where 2 signals at 100 MHz are 180 degrees out of phase and thus canc
116. points can be determined from an in band or out of band method Both techniques will give identical results in band However if an in band method is used in a system where the interfering signals are attenuated like in the IF filter in a receiver incorrect intercept points will be reported Caution The method used to determine the intercept point is only valid for 2 tones with equal amplitude Remember Intermods travel BACKWARDS as well as forward Backward traveling intermods like those through reverse isolation paths will be included in channel measurements and must be considered when making comparisons to cascade intermod equation results Please consult the specific intermod measurements of interest for details Intermod Path Measurement Summary e Adda source that will create an intermod at the path frequency e Set the Path Frequency to the frequency of the intermod e Setthe Channel Measurement Bandwidth to the widest order of interest not too wide to include interfering tones e Setthe Path Interferer Frequency in the Edit Path dialog box to the frequency of the interfering tone e Make sure the Maximum Order is set high enough to include the intermods of interest e Add intermod measurements to a level diagram or table e Remember Intermod can and do travel backwards be careful when only expecting forward traveling results Cascaded Intermod Equations and Spectrasys Spectrasys doesn t use cascaded
117. port and choosing Details from the Edit menu A typical EM Port Properties dialog box is shown below The following sections describe the entries in this dialog box EH Port Properties Draw Size This has no effect on the simulation It controls the size that the port number appears on screen and on printouts Ref Plane Shift This parameter is only available if Port Type is set to Normal see below On most complete circuits this value can be left at zero A positive Reference Plane shift causes the deembedding to add extra line length to the circuit A negative value is more common and causes the reference planes to move inside the box See the Patch Antenna Impedance example for an example of a patch antenna simulation and the Edge Coupled Filter example which uses a reference plane shift The reference plane is shown as an arrow on the layout Additionally when the EMPort is selected Handles appear on the reference plane allowing it to be moved with the mouse Port Number When EMPOWER is tun the port numbers specified here correspond to the port numbers in the resulting data These port numbers must be sequential numbers cannot be skipped and Normal ports must always have lower numbers than non deembedded and internal ports LAYOUT assigns a new port number automatically when an EMPort is placed and the port number is displayed on the layout at the port EMPOWER Planar 3D EM Analysis W
118. ports This yields simulation results that more accurately reflect the behavior of an actual circuit The following figure helps illustrate which ports will be grouped in order for the calibration process to account for coupling among the ports In this setup only the first two ports will be grouped since the third port is an internal port type and the fourth port is on a different reference plane Note that even though the second port has a reference offset assigned to it for this process they are considered to be on the same plane and their reference offsets will be made equal If you do not want the ports to be grouped you must add metal to the edge of the object that one of the ports is connected to The ports will no longer be on the same plane and will not be considered part of the same group If you would like to exclude the coupling effect calculation between single ports with the same reference plane slightly shift the feeding line edges of the ports or use the No Deembed port mode instead of single For Momentum GX the No Deembed port mode means calibrated with 0 reference plane offset and without accounting coupling between ports For example this design has 4 single mode ports at each side left and right All the ports from each side are connected to metal edges having same X coordinates and therefore the same reference plane Momentum will calculate coupling between ports 1 4 and between potts 5 8
119. real value The left hand side specifies the source branch signal to assign the RHS It consists of a signal access function applied to a branch The form is V al n2 lt expression Branch contribution statements will implicitly define source branch relations The branch is goes from the first net of the access function to the second net If the second net is not specified in the call the global reference node ground is used as the reference net Ports Ports provide a way to connect modules to other modules and devices A port has a direction input output or inout which must be declared The ports are listed after the module declaration The port type and port direction must then be declared in the body of the module Examples module resistor p n inout p n electrical p n module modName outPort inPort output outPott input inPort electrical out in Ports can support vectors buses as well Advanced Modeling Kit Analog Functions Analog functions provide a modular way for a user defined function to accept parameters and return a value The functions are defined as analog or digital and must be defined within modules The analog function is of the form analog function real integer function name input_declaration statement _block endfunction The zuput declaration describes the input parameters to the function as well as any variables used in the statement block input passed parameters r
120. relative levels of spectral components for the small signal regime and equal amplitudes of the signal s tones is shown above Definitions of symbols P Fundamental Tone Power IP Nth Order Intercept Point H Fundamental Tone H2 2nd Harmonic H 3rd Harmonic IM Nth Order Intermods IMs Nth Order Intermods due to M tones 2nd Order Intermod Products The amplitude of the second order intermod products F2 F1 and F1 F2 are equal to the tone power level minus IP2 or in other words IM2 Ptone IP2 2nd Harmonics The amplitude of the second harmonics are calculated as follows The amplitude of the second harmonic is equal to the tone power level minus the difference between IP2 second order intercept and the tone power level of the device 3rd Order 2 Tone Products 449 Simulation 450 The amplitude of the third order products 2F1 F2 2F2 F1 2F1 F2 and 2F2 F1 are equal to 2 times the quantity of the tone power level minus IP3 or in other words IM3 2 Ptone IP3 Carrier Triple Beats 3rd Order 3 Tone Products When mote than two carriers are present in a channel 3rd order intermod products can be created by the multiplication of three carriers These intermods are called carrier triple beats Spectrasys automatically creates triple beats for all combinations of 3 or more carriers Working out the math carrier triple beats will be 6 dB higher that the 3rd order 2 tone products This calculati
121. should be taken with this recombination as the ground reference for the four individual ports does not act as the ground reference for the two recombined ports Therefore the top recombination scheme shown in Figure 10 is incorrect The correct recombination scheme is the bottom one shown in Figure 10 substrate layer stack metallizatian layer Lz3inches air 1 0 FRA H 59 mil 4 4 4 stack E air Ep 1 0 finite groundplane rat taken up in the layer i gi tour internal parte Figure 9 PCB substrate layer stack and metallization layers 384 Momentum GX Moment 4 port Mornmentum 4 port 3 l Figure 10 Simulation results The bottom illustration shows the correct recombination of ports Finite Ground Plane Ground References The process of adding two extra internal ports in the ground plane and recombining the ports in the correct way can be automated in Momentum by defining two ground reference ports Figure 11 The S parameters obtained with Momentum are identical to the S parameters of the two port after the correct recombination in the section above also shown in Figure 12 gutxatrate layer stack metallization layer L 3 inches air 4 1 0 W 2 Inches FRA H 58 mil 4 4 air Ep 1 0 finite groundplane mot taken up In the layer stack Two Internal parts with associated ground reference porta 385 Simulation Figure 11 Substrate layer stack and meta
122. simulation time and restrict the amount of collected data Filter Shape The user can use a brick wall filter or the traditional spectrum analyzer gaussian filter The variations in gaussian filters are used to determine frequency at which the convolution will begin and end These values are specified in terms of channel bandwidths Obviously the wider the filter end frequency is the longer the simulation time will be and the more data will be collected The amplitude range of the filter that is associated with the given end frequency is also given for convenience Filter Shape Gaussian Eo 118dBc 60 ChanBw Brickwall Ideal Gaussian to 100dBc 30 ChanByy Gaussian to 118dBc 60 chanBw Gaussian to 150dBc 200 chanBw The analyzer modes settings are found on the Composite Spectrum tab of the system analysis Analyzer Simulation Process The analyzer goes through the following steps before the analyzer trace is ready to display in a graph 1 The total spectrum trace for the path is created 2 This spectrum is broken down into bands to find groups of spectrums Each group will have guard bands equivalent to the stop band of the selected filter 3 Before the convolution occurs the group must be discretized Since some spectrums have bandwidths as small as 1 Hz they can be completed ignored during a standard discretization process To eliminate skipped spectrum the discretization process will keep track of
123. slot layer with coplanar ports The slot layer model is more accurate than its strip equivalent because coplanar ports are calibrated and minimum possible distance is used between sional port location and locations of its ground reference ports Defining a Common Mode Port Use common mode potts in designs where the polarity of fields is the same among two or more ports even modes propagate The associated ports are excited with the same absolute potential and are given the same port number 326 Common mode ports have the following properties They can be applied to objects on strip layers only A ground plane or other infinite metal such as a cover is required as part of the design Two ot mote ports can be associated Associated ports are excited with the same absolute potential and same polarity The ports must be on the same reference plane Befare Three ports are treated as one after ihey are associated as commvon mete parta To define a common pott E p I ede er a Place a new EM pott in layout ot select any port that you want to assign this type to Note the port number Open the port dialog window double click or clicking Enter for selected port For Port Type select Common mode Click OK Select the second port and open its dialog For Port Type select Common mode Momentum GX 7 Under Associate with port number enter the number of the port that you selected first Make sure that t
124. source file Only white space or comments can appear on the same line as the include directive A file included in the source using include can contain other include compiler directives however infinite nesting is not permitted include filename Examples include user include global_decl vams include mylIncludes txt include myFunctions va Macros String substitution can be performed with the define directive both inside and outside module definitions The macro is used in the source file by insert the character followed by the macro name The preprocessor then substitutes the text of the macro for the string text macro name All compiler directives are considered predefined macro names and so redefining a compiler directive as a macro name is not allowed A text macro can be also be defined with arguments to provide much more flexibility However the use of macros can complicate symbolic debugging so the user should be careful in their use Examples define EPSSI 1 03594e 10 define KboQ P K P Q define strobe flag xName X if debug gt flag strobe n s vog xName 1 0 X The macros are then accessed in the code as 49 Simulation 50 factorl sqrt EPSSI EPSOX tox strobe 1 Vth Vth ifdef else endif These are conditional compiler directives for optionally including lines of Verilog A source file ifdef checks for if variab
125. technique is used to include the effects of frequency impedances or VSWR bandwidth image frequencies and multiple paths in cascaded noise measurements NOTE Because traditional cascaded noise figure equations are not used in the Spectrasys a new formulation to calculate noise figure was developed This formulation relies on the cascaded gain measurement to accurately determine the cascaded noise figure The cascaded gain measurement requires a signal to be present in the channel for this measurement to be made Only noise or intermods in the channel are insufficient In general no adjustments in the noise set up is need to obtain accurate noise analysis There are a few cases where noise accuracy can be improved Signal bandwidths are as wide or wider than the narrowest filter bandwidths in the simulation Traditional cascaded noise figure equations assume NO bandwidth and that the cascaded noise figure is ONLY valid at a single frequency where noise factor and gain values ate taken In practice bandwidths must be considered Rarely do impedances remain constant across a channel especially when any type of filter is involved Consequently channel signal and noise power may not be constant across a channel When using wide channel bandwidths relative to narrow filter bandwidths channel noise power may appear to decrease through a cascaded This is due to bandwidth reduction Furthermore power levels due to channel integration can be less acc
126. the dataset This allows you to output fewer or more points than are calculated by the simulator e Output Start Time controls when data collection begins If it is not zero then no data is saved until that time point is reached Note that simulation always starts at time zero no matter what the output start time is set to e When Force Output at Exact Step is checked Cayenne will use the predictor generally third order gear to interpolate the simulated points to ensure that the output waveform is sampled uniformly at exactly the time points requested needed for performing FFT analysis on the output waveform If this is not checked then Cayenne will output exact values at time points simulated but which will vary in spacing e Output All Simulated Points outputs every point that the simulator calculated which can often output ten times more data This option can be useful when debugging your circuit to determine where the simulator is actually calculating values Hint To see these points graph a response turn on symbols on the traces then zoom in to see exactly which points the simulator chose Numerical Precision Cayenne is different from traditional SPICE simulation in one significant way Cayenne tracks and controls the error in current at each node Traditional SPICE tracks chatge which is not normally observed by the user This difference is important when a circuit has a capacitance that is large relative to the time step
127. the described MoSD application to planar structures EMPOWER References General Background J A Stratton Eectromagnetic theory McGraw Hill Co New York 1941 G Kron Equivalent circuit of the field equations of Maxwell Part I Proc of IRE 1944 May p 289 299 C G Montgomery R H Dick E M Purcell Principles of microwave Circuits McGraw Hill Co New York 1948 O Heaviside Edectromagnetic theory AMS Chelsea Publishing Co New York 1950 A A Samarsku A N Tikhonov About representation of waveguide electromagnetic fields by series of TE and TM eigenwaves in Russian GTF Journal of Theoretical Physics 1948 v 18 p 959 970 P I Kuznetsov R L Stratonovich The propagation of electromagnetic waves in multiconductor transmission lines Pergamon Press Oxford 1964 originally published in Russian 1958 K S Yee Numerical solution of initial boundary value problems involving Maxwell s equations in isotropic media IEEE Trans v AP 14 1966 p 302 307 V V Nikol sku Variational approach to internal problems of electromagnetics in Russian Moscow Nauka 1967 J Meixner The behavior of electromagnetic fields at edges IEEE Trans v AP 20 1972 N 7 p 442 446 B V Sestrotetzkiy RLC and Rt analogies of electromagnetic space in Russian in Computer aided design of microwave devices and systems Edited by V V Nikol skii Moscow MIREA 1977 p 127 128 T Weiland
128. the output node versus the input frequency or the channel power at the output node versus a sweep of the input powet To create a plot of a swept variable versus a measurement do the following e Adda rectangular graph e Set the Default Dataset or Equations to the swept path dataset of interest e Add the measurement and then selected the array entry in the measurement for the node of interest To do this use the following syntax e Measurement NodeNames Name of Node e In this example this would be CNF NodeNames Out as shown below 441 Simulation ad NF vs Freg Properties General Graph Properties Default Dataset or Equations Svskemi Data Pathl 2 5weepl d Measurement Label Optional On Right Hide Color C ha CNF NodeNamest Out CNF at Output am PT mmm Add Series Measurement Wizard 4 Equation Wizard Left Y Axis Right Y Axis s xis Add a Series then enter Title Title Title dataset measurement Auto Scale Log Auto Scale Log Auto scale Log Scale to graph a measurement or press 4 Wizard button bo guide you through Max 2170 the process of defining a measurement Units dB Units None Units None Min Min c Min 2110 Max Max i Divisions 10 Divisions Divisions e The resulting graph is shown below 442 Spectrasys System Noise Figure versus Frequency at the Output CNF at Output dB
129. the same category of signals For example a source spectrum can never be coherent with an intermod spectrum and vice versa Source spectrum can be coherent with source spectrum and intermod spectrum can be coherent with intermod spectrum etc Coherency of phase noise is determined by the coherency of the phase noise parent spectrum If the two parent spectrums that have phase noise are coherent then the phase noises themselves are coherent 3 Signals must have the same center frequency and bandwidth All coherent signals must have the same center frequency and bandwidth For example a 2nd harmonic cannot be coherent with a 3rd harmonic since both the center frequency and bandwidths are not the same However if we had a cascade of two amplifiers then the 2nd harmonic generated in 1st amplifier would be coherent with the 2nd harmonic generated in the 2nd amplifier from the same 471 Simulation signal source In this case both the center frequency and bandwidth are the same with both harmonics being created from the same signal source 4 Must have the same LO source mixets only When a new spectrum is created at the output of a mixer Spectrasys will determine the coherency of the mixer input signal as well as the LO signal A new coherency number will be assigned for unique input and LO signals If there is more than one mixer in the simulation then coherency numbers for the second mixer may come from the first mixer if the all of the
130. the sweep 4 Note that a Sweep1_Data dataset is built 5 Double click the S21 graph and change the Default Dataset or Equations to Sweepl Data so you plot S21 for the swept data Turn off symbols by clicking the Symbols button the last button on the Graph toolbar You get a range of traces that looks like this 39 Simulation 40 TM LLI T LA ur 4 5 Frequency MHz S 2 1 Here the mouse is hovering over a dot on the dark green trace and the popup identifies the trace and value To look at the range at 200 pF enter the following formula into the graph line Enter S C1_C_Swp_F 200 2 1 Note that the swept C value is in the Sweepl Data set and named C1 C Swp F C1 C swept on F The graph now is Getting Started EEK i r3 e Hl fa a oo a m TI a F 4 2 20 cu ui a a e uw i ha D e e 5 B Frequena Fal S C1 C Swp Fee 41 Simulation Parameter Sweep Properties Parameter Sweep Properties Sweep Mame Analysis to Sweep OC Mi Eactory Defauts Parameter to sweep Designs 4mpiR2 R v Description Parameter Range Type OF Sweep Start Ohm Linear Number of Points les Stop 1000 Ohm CO Lag Points Decade les Linear Step Size Ohm Unit of Measure Ohm List Ohm KA Clear List Show Long Parameter Mames Propagate All Variables When Sweeping Cor only analysis variables Loox cance
131. the total integrated power in the main channel along the specified path This measurement includes ONLY DESIRED SIGNALS on the beginning node of the path traveling in FORWARD path direction All other intermods harmonics noise and phase noise signals are ignored Note A D is placed next to the equation in the identifying flyover help in a spectrum plot to indicate desired signals For example if the Channel Measurement Bandwidth was specified to 03 MHz and the Channel Frequency was 220 MHz then the DCP is the integrated power from 219 985 to 220 015 MHz This power measurement will not even be affect by another 220 MHz signal traveling in the reverse direction even if it is much larger in amplitude Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used ONLY DESIRED SIGNALS Travel Direction Only in the FORWARD direction Offset Channel Power OCP The Offset Channel is a user defined channel relative to the main channel The Offset Channel Frequency and Offset Channel Bandwidth are specified on the Options Tab of the System Analysis Dialog Box As with the Channel Frequency measurement Spectrasys automatically deals with the frequency translations of the Offset Channel Frequency through frequency translation devices such as mixer and frequency multipliers For example if the Channel Frequency was 2140 MHz Offset Channel Frequency was 10 MHz and the Offset Channel
132. then this button has no effect J Clockwise Button Rotates the current image clockwise in the plane of the screen The center of the viewer image window is always the center of rotation This option can also be selected by pressing Page Down K Counter Clockwise Button Rotates the current image counter clockwise in the plane of the screen The center of the viewer image window is always the center of rotation This option can also be selected by pressing Page Down L Rotate Right Button Rotates the current image counter clockwise in a horizontal plane perpendicular to the screen The center of the viewer image window is always the center of rotation M Rotate Left Button Rotates the current image clockwise in a horizontal plane perpendicular to the screen The center of the viewer image window is always the center of rotation N Rotate Down Button Rotates the current image backward in a vertical plane perpendicular to the screen The center of the viewer image window is always the center of rotation O Rotate Up Button Rotates the current image forward in a vertical plane perpendicular to the screen The center of the viewer image window 1s always the center of rotation P Top Button Shows a top down view of the current image This option can also be selected by pressing the Home key Q Front Button Shows a front view of the current image This view is from the y axis at z 0 This option can als
133. this measurement Types of Spectrums Used None Travel Direction N A Stage Input Saturation Power SIPSAT This measurement is the stage input saturation power calculated by using the Stage Output Saturation Power and the Stage Gain When a stage doesn t have this parameter 100 dBm is used SIPSAT n SOPSAT n SGAIN n dBm where n stage number Channel Used No channel is used for this measurement Types of Spectrums Used None Spectrasys System Travel Direction N A Stage Output 1 dB Compression Point SOP1DB This measurement is the stage 1 dB compression point entered by the user When a stage doesn t have this parameter 100 dBm is used Channel Used No channel is used for this measurement Types of Spectrums Used None Travel Direction N A Stage Output Intercept All Orders SOIP This measurement is the stage output intercept point entered by the user When a stage doesn t have this parameter 100 dBm is used Each column in this measurement is for a different intermod order up to the Maximum Order specified on the Calculate Tab of the System Analysis Dialog Box The column number 1s the same as the order starting from the left with order 0 Channel Used No channel is used for this measurement Types of Spectrums Used None Travel Direction N A Stage Output Saturation Power SOPSAT This measurement is the stage saturation point entered by the user When a stage doesn t have this parameter 1
134. to do so may yield erroneous simulation results 381 Simulation direct excitalion point feed Figure 4 Internal port and equivalent network model The following sections illustrate the use of internal ports with ground planes and with ground references and the results Internal Ports and Ground Planes in a PCB Structure Figure 5 shows the layout for a PCB island structure with two internal ports The infinite ground plane is taken up in the substrate layer stack and provides the ground reference for the internal ports The magnitude and phase of the S21 parameter calculated with Momentum are shown in Figure 6 The simulation results are validated by comparing them with the measured data for the magnitude and the phase of S21 substrate layer stack metalltzation layer L 3 Inches W 2 inges FR4 H 59 mil E 4 4 infinite ground Hane taken up im the P Figure 5 PCB Substrate layer stack and metallization layer 382 Momentum GX J 180 10 140 10 20 uu LH 60 30 Le d pn zi rm 40 je m 370 m 50 e B 260 a0 100 70 ido Biu 180 0 4 0 4 1 9 1 6 2 4 2 0 J U 0 0 0 5 1 0 1 5 zZ D 72 5 J0 freq GHz Trea GHz Figure 6 Magnitude and phase of S21 The thicker line is Momentum results the thinner line is the measurement Finite Ground Plane no Ground Ports The same structure was resimulated with a finite groundplane Figure 7 The substrate layer stack contains no infinite m
135. total intermod channel power measurements e first interfering signal e second interfering signal 515 Simulation 516 The Channel Frequency must be set to the intermod frequency and the Interferer frequency must be set the first or second interfering frequency See the Calculate Tab on the System Analysis Dialog Box to set the Interfering Frequency Furthermore the spacing of the interfering tones needs to be such that intermods will actually fall into the main channel If these conditions are not met then no intermod power will be measured in the main channel Each column in this measurement is for a different intermod order up to the Maximum Order specified on the Calculate Tab of the System Analysis Dialog Box The column number is the same as the order starting from the left with order 0 See the Intermods Along a Path section for information on how to configure these tests Remember intermod bandwidth is a function of the governing intermod equation For example if the intermod equation is 2F1 F2 then the intermod bandwidth would be 2BW1 BW2 Note Bandwidths never subtract and will always add The channel bandwidth must be set wide enough to include the entire bandwidth of the intermod to achieve the expected results The Automatic Intermod Mode will set the bandwidth appropriately Cascaded intermod equations are not used in Spectrasys Caution This method used to determine the intercept point is only valid
136. vertices will be resolved by the mesh generator if their distance is less than 2 dbu Use the Momentum Analysis Properties dialog to change the Layout resolution for your fabrication or layout process as needed Momentum Options General Simulation Options Mesh Mesh Frequency Mesh Density cells wavelength Arc Resolution 45 max deg w Layout Mesh Resolution 1e 3 mm w Momentum GX Gap is 0 1 mi actual EXAMPLE Precision dbu gat to 0 Od mil Mash recognizes gap because precision ig within the dbu units setting EXAMPLE Precision dbu set to 1 mil Gap le collapsed in the mesh because the dbu precision value is greater than the gap Adaptive Frequency Sampling Adaptive Frequency Sampling AFS is a method of comparing sampled S parameter data points to a rational fitting model In order to accurately represent the spectral response of the circuit the AFS feature in Momentum takes a minimal number of frequency samples and then applies its algorithm to the sampled data Wherever the S parameters vary the most mote samples are taken When the fitting model and the sampled data converge the AFS algorithm is then complete and the S parameter data is written into the dataset In this way the maximum amount of information can be obtained from the minimum amount of sampling By using the AFS feature you can greatly reduce the amount of time required to simulate circuits that have resonances or othe
137. where the source is connected Description This is a summary description of the source Edit Button When clicked this button will bring up the edit parameters dialog for this part Minimum number of source data points When checked each source signal is represented by the specified number of data points When unchecked 2 points will be used This parameter is ignored for all continuous frequency sources whose amplitude is dependent on several data points This option is extremely useful when wideband signals are being simulated through filters since impedance can vary drastically across its bandwidth If only 2 data points are used then spectrum power and noise will only be represented by these 2 points Factory Defaults Sets all properties to their default values System Simulation Parameters Paths Tab Many measurements require the definition of a path For more information on paths click here Tabs General Paths Calculate Composite Spectrum Options Output Spectrasys System System Simulation Parameters General Paths Calculate Composite Speckrum Options Output Tu Add All Paths From All Sources Tu Add Path amp Delete All Paths m Parts Interferer Output Frequency FDesired MHz Add vot Pwr and Z Adj Chan 2 1 1 2 Ady Factory Defaults Parameter Information Add All Paths From All Sources Automatically adds all possible paths between sources and output ports Add Path Invokes th
138. which to run Sonnet If you have lumped elements in your simulation you can often turn down the number of frequencies here and increase the number of frequencies in the Co simulation sweep specified below Start Freq MHz Specifies the minimum frequency to analyze Stop Freq MHz Specifies the maximum frequency to analyze Number of Points Specifies the number of frequency points to analyze Points are distributed linearly between the low and high freq specified above For an ABS sweep this number of points is used on the GENESYS graphs unless a different Co simulation sweep is specified Use Adaptive Sweep ABS Turns on Sonnet s Adaptive Sweep simulation If this box is checked the Number of Points entry above is not sent to Sonnet Co Simulation Sweep Specifies the frequencies at which to run simulate the lumped elements Sonnet data combination If you have no lumped elements in your simulation you should normally check the Use EM Simulation Frequencies box For circuits with lumped elements you can often save much time by using fewer points in the electromagnetic simulation frequencies above allowing the co simulation to interpolate the Sonnet data before the lumped elements are added Automatically save workspace after calc This checkbox is handy for overnight runs to help protect against a power outage Note that checking this box will force the entire workspace to be saved after each run 399 Simulation 4
139. 00 Fart number to excite Mode number ta excite Turn off physical losses Faster Onl check errors topology and memory do nat simulate Advanced Options Lo simulation sweep Setup Layout Port Modes Use EM simulation frequencies Start freg MHz 1400 Thinning out subarid Stop freq MHz 2600 ps E w Number of points fio M Thin out electrical lossy surfaces Solid thinning aut slower accurate v Use planar ports for one port elements Add extra details to listing file Iw Show detailed progress messages Command line Click the Recalculate Now button If anything has been modified since the last EMPOWER run this launches EMPOWER to simulate the layout Note EMPOWER has been given a lot of intelligence to determine when it needs to calculate Clicking Recalculate Now will not do anything if EMPOWER believes it is up to date To force EMPOWER to recalculate from scratch right click on the electromagnetic simulation in the workspace window and select Force Complete Analysis Now Once EMPOWER calculation is completed GENESYS displays the calculated data The graphs below show GENESYS after EMPOWER simulation Double click the graph items in the workspace window to open them and select Tile Vertical from the Window menu to organize them as shown below 125 Simulation GENESYS 7_0 Ioj x File Edit View Workspace Achons Tools Synthesis Window Hel
140. 00 Sonnet Interface Options Sonnet Advanced SpeediMemor Compute Current Density peed Y E Memory Save Box Resonance Info More Accurate Faster Analysis More Memory Less Memory Multi Frequency Caching d F Symmetry Use planar ports For one port elements FinefEdge Meshing ABS Freg Resolution Per Sweep Auto 300 Fregs Target CO Manual MHz Advanced Subsectioning ABS Caching Level Max Subsection Size 20 Lambda F None Estimated Epsilon EFF Stop Restart Polygon Edge Check Levels Multi Sweep plus Stop Restart Based on fixed freq MHz The Advanced tab allows you to customize advanced Sonnet options For more details on these options please consult your Sonnet reference manuals Use planar ports for one port elements This box should almost always be checked When not checked EMPOWER uses z directed ports at each terminal for all devices When it is checked EMPOWER uses in line ports for elements like resistors and capacitors two terminal one port devices The only time this can cause a problem is when you have a line running under an element for example running a line between the two terminals on a resistor in the same metal layer as the resistor pads Note EMPOWER planar ports cannot be used for ground referenced elements such as transmission lines even though the element might only have terminals Sonnet features supported in G EN ESYS e Last modify date Cu
141. 00 dBm is used Channel Used No channel is used for this measurement Types of Spectrums Used None Travel Direction N A Conducted Intermod Channel Power All Orders CIMCP This measurement is the total intermod power in the main channel conducted from the prior stage This measurement includes all intermods that are traveling in the forward path direction In equation for the conducted third order intermod power is CIMCP n TIMCP n 1 GAIN n dBm where CIMCP 0 0 dB and n stage number 521 Simulation 522 Each column in this measurement is for a different intermod order up to the Maximum Order specified on the Calculate Tab of the System Analysis Dialog Box The column number is the same as the order starting from the left with order 0 Remember intermod bandwidth is a function of the governing intermod equation For example if the intermod equation is 2F1 F2 then the intermod bandwidth would be 2BW1 BW2 Note Bandwidths never subtract and will always add The channel bandwidth must be set wide enough to include the entire bandwidth of the intermod to achieve the expected results Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as TIMCP and GAIN Travel Direction Same as TIMCP and GAIN Generated Intermod Channel Power All Orders GIMCP This measurement is the generated intermod power in the main channel created at the output of the current stage
142. 1 650 175 6 0 0047 66 0 0 9693 3 8 0 100 0 8391 10 7 21 434 171 7 0 0092 74 1 0 9662 7 7 0 150 0 8420 16 8 21 349 167 3 0 0138 74 1 0 9584 11 6 0 200 0 8312 21 8 21 109 163 1 0 0183 76 1 0 9477 15 4 0 250 0 8150 27 0 20 679 159 1 0 0221 72 0 0 9326 19 2 0 300 0 8049 32 8 20 328 155 0 0 0267 70 8 0 9157 22 9 0 500 0 7349 52 7 18 378 140 2 0 0409 63 9 0 8256 36 4 Simulation 0 700 0 6653 71 1 16 211 127 7 0 0531 57 2 0 7212 47 9 0 900 0 5930 87 2 14 148 117 7 0 0614 52 2 0 6237 57 4 1 100 0 5403 101 2 12 427 109 4 0 0695 48 8 0 5440 65 0 1 300 0 4982 113 9 11 019 102 6 0 0747 46 1 0 4736 72 0 1 500 0 4710 125 4 9 834 96 5 0 0795 44 0 0 4168 77 5 1 700 0 4495 135 7 8 861 91 3 0 0850 41 9 0 3619 83 3 1 900 0 4312 145 4 8 013 86 4 0 0893 40 4 0 3209 88 2 2 000 0 4229 150 0 7 670 84 1 0 0917 40 1 0 3021 90 9 3 000 0 4130 172 7 5 243 64 8 0 1147 34 7 0 1662 120 6 4 000 0 4749 144 9 3 914 48 5 0 1359 27 8 0 1538 173 1 5 000 0 5311 125 9 3 037 34 4 0 1524 21 3 0 1899 151 8 6 000 0 5797 113 2 2 457 22 5 0 1699 15 8 0 2239 124 4 f Fmin Gammaopt Rn Zo where Zo 50 GHz dB MAG ANG 0 900 0 64 0 22 25 0 12 1 800 0 71 0 09 97 0 08 2 400 0 75 0 06 139 0 09 3 000 0 87 0 09 175 0 10 4 000 0 99 0 19 147 0 08 5 000 1 17 0 26 125 0 11 6 000 1 34 0 38 101 0 17 INFINEON TECHNOLOGIES Munich See reference 38 for more information on the relationship between noise figure and noise parameters M easurements
143. 2 gt 1 3 gt 2 4 gt 2 5 gt 6 6 gt 1 Incorrect group port association 1 2 gt 1 3 gt 1 4 gt 5 5 gt 6 6 gt 4 This is incorrect because the port association creates 2 independent not connected port groups 1 3 and 4 6 instead of one port group 1 6 Only ports having the same reference plane and feeding line direction may be grouped as a common mode ports Momentum GX does not take into account the Line direction parameter T of EM pott properties The direction 1s defined as the direction of the perpendicular to the closest port metal edge Defining a Ground Reference Ground references enable you to add explicit ground references to a circuit which may be necessary if no implicit grounds exist in your design Implicit ground is the potential at infinity and it is made available to the circuit through the closest infinite metal layer of the substrate Implicit grounds are used with internal ports and with single ports that are connected to objects on strip metallization layers 329 Simulation 330 There are instances where the distance between a port and its implicit ground is too large electrically or there are no infinite metal layers defined in the substrate In these cases you need to add explicit ground references to ensure accurate simulation results For more information on using ground references refer to Simulating with Internal Ports and Ground References You can apply g
144. 223 Chapter 6 HARBEC Harmonic Balance Analysis HB Harmonic Balance Analysis H armonic Balance O verview The HARBEC harmonic balance simulator simulates the steady state performance of nonlinear circuits Circuits can be stimulated with a variety of periodic signals voltage current and power such as single CW tones pulsed waves or dual tones Complex waveforms can be constructed by combining various periodic signals HARBEC makes this through the custom voltage and current sources The two assumptions that harmonic balance uses are 1 the signals in the circuit can be accurately modeled using a finite number of spectral tones and 2 the circuit has a steady state solution HARBEC works by solving Kirchoff s current law in the frequency domain It applies the stimulus sources to the designed network It then searches for a set of spectral voltages that will result in currents that sum to zero at each node and each frequency in the circuit It adjusts the voltage levels a spectrum of voltages at each node through a variety of methods until the sum of the currents is less than a user specified level see Absolute Error and Relative Error on the Harmonic Balance dialog box in the Reference Manual This process of searching is known as convergence The length of time it takes to take a search step is roughly equal to the cube of the product of the number of frequencies and the number of nonlinear nodes Thus if you double the num
145. 315 Simulation EM Port Properties Draw Size P5 Ref Plane Shift Port Mumber Location i Layer BM TOP METAL matlo PartList gt Schematic Height Line Direction Default Current Direction Default Port Type Normal el Be An ey es Hee PEOP T ert Associate with part number Polarity Normal Inverse Single mode STRIP port transmission line extended calibration For normal and no deembed types setting width and or height En zero will cause LAYOUT to aubodetect their dimensions Defining an Internal Port 316 Internal ports enable you to apply a port to the surface of an object in your design By using internal ports all of the physical connections in a circuit can be represented so your simulation can take into account all of the EM coupling effects that will occur among ports in the circuit These coupling effects caused by parasitics are included in your simulation results because internal ports are not calibrated You should avoid geometries that allow coupling between single and internal ports to prevent incorrect S parameters An example of where an internal port is useful is to simulate a bond wire on the surface on an object Another example of where an internal port is necessary is a circuit that consists of transmission lines that connect to a device such as a transistor or a chip capacitor but this device is not part of the circuit that you are simulating An internal port
146. 4 3 71 7 14 45 35 1 93 7 File Record Keeping Most device files provided with GENESYS are S parameter files in the usual device configuration typically common emitter or common source Devices you add to the library may use the ground terminal of your choice However if you always keep data in a consistent format record keeping chores are greatly minimized Exporting D ata Files Export S Parameters in the File menu writes S parameter data from any simulation or data source This output data file has exactly the same format as S parameter files used to import data This allows the user to analyze tune and optimize sub networks which are then stored as S parameter data files for use later in other circuit files The S parameter data file written by GENESYS has one line of data for each simulation frequency If there are two ot mote available simulations or designs in the circuit file GENESYS displays a dialog box to allow you to select the simulation or design to use Noise D ata in Data Files Some of the data files provided with GENESYS also include noise data used fot noise figure analysis This data includes the optimum noise figure NFopt the complex source impedance to present to the device to achieve the optimum noise figure Gopt and the effective noise resistance Rn The best noise figure in a circuit is achieved when the device is presented with an optimum source impedance The optimum input netwotk to achieve this objecti
147. 4 Save as Favorite Schematic Source Summary Hame Het Hame Description m A interferer Source CVM Pyyr poh Carriers 1 Dataset 6GSM Freq FDesiredMHz Pwr aoe Es a Osc Oscillator Power Minimum number of source data points aa Parameter Information Design to Simulate User specified name This is the source name that will appear in the spectrum identification Dataset Name of the dataset where the simulation data is stored Frequency Units These are the frequency units used for the entire system simulation dialog box Nominal Impedance Default impedance used for power measurements when no impedance information is available during the simulation Measurement Bandwidth Width of the channel used in path measurements This is analogous to the resolution bandwidth on a spectrum analyzer Automatic Recalculation When checked enables Spectrasys to automatically recalculate the simulation every time the design or system simulation is out of date 537 Simulation 538 Calculate Now Button Closes the dialog and initiates a simulation Save as Favorite Button When clicked will save all the parameters associated with the system analysis dialog as the default to be used when the next system analysis 1s created SCHEMATIC SOURCE SUMMARY This section summarizes all the sources found in the given design Name This is the name of the source or part designator Net Name This is the name of the net
148. 510 or 4 10 Amps If the default precision of 10 1 Amps is used then convergence will never be obtained In reality convergence problems actually will begin occurring if convergence tighter than about 100 times this minimum or around 10 Amps in this example Cayenne detects this condition and increases the time step but sometimes that large time step will cause accuracy errors and Cayenne will throw an error You have a few choices in this case 1 Increase the current tolerance Increasing it to a large value will prevent the error and force Cayenne to give an answer similar to traditional SPICE 2 Increase the minimum and maximum time step values Notice from the equation above that any increase in the time step value causes an corresponding increase in the precision of the current 3 Change the time step method to fixed which will always use the max step If you ate forcing very small time steps for output it may be inappropriate to cause the simulator to take even smaller steps In the example above we may have requested output every 10ps a 100 GHz sampling rate Using the defaults the simulator would then take steps as small as every 0 1ps a 10 THz sampling rate which is probably excessive 4 Reduce the value of the capacitors or inductors In the example above we are sampling at high GHz to THz rates Decreasing a 10 microfarad capacitor to 10 nanofarads is unlikely to have a significant effect Note that SPICE does
149. 60 264 270 273 275 280 286 288 294 297 300 308 312 315 320 324 325 330 336 343 350 351 352 360 364 375 378 384 385 390 392 396 400 405 416 420 429 432 440 441 446 450 455 462 468 480 486 490 495 500 504 512 520 525 528 539 540 546 550 560 567 572 576 585 588 594 600 616 624 625 630 637 640 648 650 660 672 675 686 693 700 702 704 715 720 728 729 735 750 756 768 770 780 784 792 800 810 819 825 832 840 858 864 875 880 882 891 896 900 910 924 936 945 960 972 975 980 990 1000 1001 1008 1024 1029 1040 1050 1053 1056 1078 1080 1092 1100 1120 1125 1134 1144 1152 1155 1170 1176 1188 1200 1215 1225 1232 1248 1250 1260 1274 1280 1287 1296 1300 1320 1323 1344 1350 1365 1372 1375 1386 1400 1404 1408 1430 1440 1456 1458 1470 1485 1500 1512 1536 1540 1560 1568 1575 1584 1600 1617 1620 1625 1638 1650 1664 1650 1701 1715 1716 1728 1750 1755 1760 1764 1782 1792 1800 1820 1848 1872 1875 1890 1911 1920 1925 1944 1950 1960 1980 2000 2002 2016 2025 2048 2058 2079 2080 2100 2106 2112 2145 2156 2160 2184 2187 2200 2205 2240 2250 2268 2275 2288 2304 2310 2340 2352 2376 2400 2401 2430 2450 2457 2464 2475 2496 2500 2520 2548 2560 2574 2592 2600 2625 2640 2646 2673 2688 2695 2700 2730 2744 2750 2772 2800 2808 2816 2835 2860 2880 2912 2916 2925 2940 2970 3000 3003 3024 3072 3080 3087 3120 3125 3136 3150 3159 3168 3185 3200 3234 3240 3250 3276 3300 3328 3360 3375 3402 3430 3432 3456 3465 3500 3510 3520 3528 3564 3575 3584 3600 3640 3645 3675 3696 3744 3750 3773 3780 3822 3840 3850 3861 3
150. 66 61 Estim RAM 346K eee Starting Line Analysis to De embed PORT 1 9588 MHz Zo 44 732 G i280 930 11866 MHz Zo 45 H4H G 1326 6851 ae Starting Discontinuity Analysis S666 MHz 11 1 46 lt 295 8521 5 924 2H4 9506 MHz 11 8 263 lt 1 69 521 16 24 95 8 Hum o 2 w L Viewing Results After EMPOWER simulation of the layout the data must be displayed in GENESYS This can be done first by examining the DataSet and then by creating a Data Output such as a Rectangular Graph To create a rectangular graph in this workspace using InstaGraph 1 Double click EM1_Data in the workspace tree to open the EMPOWER results 2 Right click on S and select Graph Rectangular Graph This instructs GENESYS to create a graph containing EMPOWER S parameter data Data will be displayed at 8000 9500 and 11000 MHz For a complete description of rectangular graphs see the GENESYS User s Guide The GENESYS display below shows the EMPOWER run with 3 sample points 119 Simulation 120 ORBEEEEZE st IN TT TT IAT UL a n c rm u Ci Freq Hz DB 521 lt DB 311 In this response the notch frequency appears to occur exactly at 9 5 GHz but we don t have enough points to tell for sure To add more frequency points to the EMPOWER simulation To re simulate adding more points 1 Double click EM1 under Simulations Data in the Workspace Window 2 Change the Numb
151. 888 3900 3920 3960 3969 4000 4004 4032 4050 4095 4096 4116 4125 4158 4160 4200 4212 4224 4290 4312 4320 4368 4374 4375 4400 4410 4455 4459 4480 4500 4536 4550 4576 4608 4620 4680 4704 4725 4752 4800 4802 4851 4860 4875 4900 4914 4928 4950 4992 5000 5005 5040 5096 5103 5120 5145 5148 5184 5200 5250 5265 5280 5292 5346 5376 5390 5400 5460 5488 5500 5544 5600 5616 5625 5632 5670 5720 5733 5760 5775 5824 5832 5850 5880 5940 6000 6006 6048 6075 6125 6144 6160 6174 6237 6240 6250 6272 6300 6318 6336 6370 6400 6435 6468 6480 6500 6552 6561 6600 6615 6656 6720 6750 6804 6825 6860 6864 6875 6912 6930 7000 7007 7020 7040 7056 7128 7150 7168 7200 7203 7280 7290 7350 7371 7392 7425 7488 7500 7546 7560 7644 7680 7700 7722 7776 7800 7840 7875 7920 7938 8000 8008 8019 8064 8085 8100 8125 8190 8192 8232 8250 8316 8320 8400 8424 8448 8505 8575 8580 8624 8640 8736 8748 8750 8775 8800 8820 8910 8918 8960 9000 9009 9072 133 Simulation 134 9100 9152 9216 9240 9261 9360 9375 9408 9450 9477 9504 9555 9600 9604 9625 9702 9720 9750 9800 9828 9856 9900 9984 10000 Thick M etal Using Thick Up or Thick Down metal will greatly increase the complexity of an EMPOWER run as all metal layers must be duplicated for the top and bottom of the thick metal and z directed currents must be added along the sides of all metal The detailed of defining metal layers is found in the EMPOWER layers dialog box as follows Metal Layets All metal layers from the General
152. B Gompressiofz c netten 508 Input InteFee DE oos DAS RUUN INNEN 509 Input Intercept Recevet iiias 509 Input Intercept RECely et aiiiar 509 input third order IntetCepticneee tacente 509 IOpEt VO WB erer AA 292 Intefee Dto eden PIE Hr DET E 238 EOULS KON Oa TERT tees teeta wena a RR UR RA 142 Interferer Cascaded Gain ia cesse seo ecsctii sene 510 Interferer Channel Frequency sauce 502 Intecterer Channel Posyetacicenoc ticus 510 Ia nes ie ch sos wk GUNo eu sedo t root CU ANES Enan 511 Intermod and Harmonic Basics 445 Intermod Path Measurement Basics 457 BYTES ETOCS 240 HUM UN CUR LM DU 451 intermods and harmonics 445 545 intermodulation ost Et ts 225 Simulation 560 internal ports 100 101 145 153 177 208 IP IDB iiit n Ebo et a Ende 508 loire o T A 13 Lu ptr AEE S 86 Brno iP 225 234 K 11 Gau 225 234 L L1181 layers iiit Siva iem E 134 layouts CHASE acsi eS 114 SITHIULTO CHOR nienn a cdd 117 level CLAS eE AA 427 arean e a n has Re UE 136 lineo Dedances eoa n Mme vi E 131 181 acir pap DI GUCIe Tuc esae od Reda 13 14 linear Measurements aesacsaaaandtgren tees 11 linear simili oneen 1 289 linear simulation properties 290 linear S Dat AIM Cte TS occu eene tei ao o 292 liStitie D e coe ise b e eeieie toe idude 129 181 ENMETI NW SPa se scretnrerarahosasasssasiscgeaecs 174 load pull C OBECOUE
153. B1 NCI Result in graph Smith chart dB Magnitude of 22 Loaded Q of S21 Linear Magnitude of S21 Input reactance at port 1 On a Smith chart S11 will be displayed while IM Zin1 will be used for the marker readouts show all S parameters On Smith or polar chart shows input plane stability circles On Smith or polar chart shows constant noise circles Using Non Default Simulation Data Result on table dB Magnitude plus angle of S22 default Loaded Q of 821 Lineat Magnitude of S21 Input reactance at port 1 Shows dB Magnitude alterable via complex format Displays center radius and stability parameter of input plane stability circles Displays center and radius of all noise circles 27 numbers per frequency Getting Started In all dialog boxes which allow entry of measurements there is a Default Dataset combo box Any measurement can override this default The format to override the dataset is operator dataset measurement where dataset is the name of the Dataset from the Workspace Window and operator measurement are as described in previous sections An override is most useful for putting parameters from different simulations on the same graph Some examples of overrides are Meas Meaning db Linear1 S 2 1 Show the dB magnitude of S21 from the Linear1 dataset EMI1 S11 Shows 11 from the EM1 dataset Filter ql Filter1 S21 Shows the loaded Q of the Filter design using the current simu
154. Bandwidth was 1 MHz then the OCP is the integrated power from 2149 5 to 2150 5 MHz This measurement is simply a Channel Power measurement at the Offset Channel Frequency using the Offset Channel Bandwidth Channel Used Offset Channel Frequency and Offset Channel Bandwidth 507 Simulation 508 Types of Spectrums Used Same as CP Travel Direction Same as CP Calculated Stage Compression COMP This measurement is the calculated compression point of each individual stage This value is determined during the simulation process based on the total input power of the stage This measurement includes ALL SIGNALS INTERMODS HARMONICS NOISE and PHASE NOISE traveling in ALL directions through the node that fall within the main channel Channel Used No channel is used for this measurement Types of Spectrums Used All SIGNALS INTERMODS HARMONICS NOISE and PHASE NOISE Travel Direction All directions through the node Gain GAIN This measurement is the gain of the main channel along the specified path The Gain is the difference between the Desired Channel Power output of the current stage minus the Desired Channel Power output of the prior stage as shown by GAIN n DCPIn DCP n 1 dB where GAIN 0 0 dB n stage number See the Desired Channel Power measurement to determine which types of signals are included or ignored in this measurement Channel Used Main Channel Frequency and Channel M
155. Can reduce memory requirements under some circumstances Onz Use a smaller line segment 7 times smaller for de embedding calculations Can speed up line analysis IT Output viewer data file in text format PLX Simulation Status Window E EM1 EMPOWER Log Running workspace layonly Press Escape to stop the EMPOUER run EMPOWER Planar 3D EM Simulator Version 7 66 C gt 1778 9 7 Eagylevare Corp FREQ 11HHH MHz ModeC DISC Uiewta gt Lossta gt Thinta gt 5ummt amp YZ MIRR Estim time 44 08 61 Each frg 66 44 01 Estim RAM 348K see Starting Line Analysis to De embed PORT 1 9588 MHz Zo 44 732 G i280 930 11866 MHz Zo 45 H4H G i1326 651 eee Starting Discontinuity Analysis S666 MHz 511 1 464 295 8521 5 924 2H4 9566 MHz 511 H 2634 1 69 521 16 24 95 8 The window above is shown when EMPOWER is running The objects on the second line are FREQ The current calculation frequency 106 EMPOWER Planar 3D EM Analysis Mode DISC discontinuity LINE line analysis or LN D both View Checked if viewer data is to be generated Loss Checked if physical loss is being modeled Thin Checked if thinning 1s enabled Symm Displays the type of symmetry possessed by the circuit being analyzed This option can be XZ YZ Mirror 2 way mirror or 1800 rotational The objects on the third line are Estim Time The estimated total time to complete the current calculation mode
156. Claning and Tom Vito Noble Publishing pg 174 Spectrasys has the ability to show group delay results including across frequency translations by using a wide carrier source and the linear simulator The group delay measurement function will then be applied to any path voltage gathered by Spectrasys The path voltage in Spectrasys is equivalent to the voltage gain E21 To convert E21 to 521 for group delay calculation the following equation must be used Sij Ey conj ZPortj ZPortj Sjj ZPorti sqrt real ZPortj real ZPorti where ZPott is the port impedance the circuit sees For ZPorti ZPortj R j0 Then Sij Hi Of Spectrasys System j 1 5j S21 E21 1 S11 Since Spectrasys doesn t have an S11 measurement the linear simulator must be used A frequency vector can be retrieved from the Spectrasys path dataset so that the frequencies gathered by the linear simulator are the same as those used in Spectrasys STEPS FOR GROUP DELAY MEASUREMENT 1 8 Add a wideband signal source The bandwidth should be wide enough to cover the group delay range of interest Caution must be exercised when setting the carrier power since a wideband carrier can easily drive non linear devices into saturation Note A multisource can be used here and the number of points for the wideband cartier can be specified per source Increase the number of simulation points for the carrier on the General t
157. Connect Selected Parts and then Center Selected on Page Then click the zoom maximize button crossed arrows The layout should now be centered on the PWB However the transmissions lines might not be lined up exactly on grid These can be placed on grid by placing the mouse ovet the center of one of the bottom capacitor pads and dragging the entire filter structure up to the nearest grid line all parts must be selected in order to drag the entire filter 196 IS Planar 3D EM Analys EMPOWER RUE OO PP C eC eC eC poA DAMES poen SP eet he tete ete ees eee C a bee ie be ee eo PM ho oe ee ie CoU CIC AE ae dra aS pp Ree Ct CIL CC oM pP PE seat PD UE eee EEEE CX M iced de P id Db L4 4 4 4 4 0 Sce oo se Se Se SE Se ese SES ET eel eee od eal eG 04 od od 1 dod e eu aS SS XXX EE 0604 eee et D da Kd d phe De hn he EE Sy stig Ru Mang tha Ma ie Mo Maal Le he ee ee oh a be ee eb x x X p eee queen e UR pr E pe pee bebe bee n C en ee ee ee ee ee eb E DE xir deos E Se oe oe oe E should be moved up so the drill hole is at the top of the upper capacitor pad as extend beyond the length of the resonators Furthermore the ground vias shown on the left resonator 15 We need to move the capacitor footprints so that the capacitor pads do not 197 we need to change the placement of the input and output lines We want to pull them away from the center of the resonators to the
158. Default None Automatic or manual e Specify Width enter the value of the Edge Mesh Width in layout length units in this screen units mil The relationship between wavelength and cells per wavelength is best desctibed by this example If the portion of the circuit on this layet is 2 wavelengths long and the number of cells per wavelength is 20 the length of the circuit on this layer will be divided into 40 cells Select another metal layer and repeat these steps fot setting the mesh parameters Click OK to accept the mesh parameters for the layers You also may specify a metal layer strip model and a dielectric layer via model in this table Default settings are defined in Momentum Analysis Precomputing a Mesh In order to calculate a mesh you must first pre compute the substrate and apply ports to the circuit Using the specified frequency and any user defined mesh parameters a pattern of rectangles and triangles is computed and applied to the circuit During a simulation the 354 Momentum GX surface current in each cell is calculated and this information is then used to solve S parameters for the structure To precompute the mesh without calculating the S Parameters 1 Open the Momentum Options dialog and select the Simulation Options tab f Momentum Options General Simulation Options Mesh Simulation Made S RF ffaster no radiation effects CO Microwave recalculates substrate for e
159. EM simulation Frequencies Stop freg MHz 11000 Start freq MHz Number of points 31 Number of paints Stop freg MHz Use Adaptive Sweep ABS Recalculate Now Automatic Recalculation Automatically save workspace after calc General Tab Layout to Simulate Allows you to select which layout in the current workspace to simulate Since wotkspaces can have multiple layouts and multiple simulations you can simulate many different layouts within the same workspace Port Impedance When Sonnet results are plotted on a graph it will be normalized to this impedance Different impedances can be used for each port by separating impedances with commas A 1 Port Device Data File can be used in place of any impedance file to specify frequency dependent or complex port impedances Note This impedance is not used in Sonnet so any viewer data generated defaults to 50 ohms You must change the port impedance inside the Sonnet current viewer for non 50 ohm ports 398 Sonnet Interface Use ports from schematic Check this box when co simulating with HARBEC harmonic balance nonlinear simulation This forces all sources and impedances to be considered in the simulation Note Be sure to check Use ports from schematic if you will be using this simulation as the basis for a HARBEC Simulation otherwise there will be no nonlinear sources available Electromagnetic Simulation Frequencies Specifies the frequencies at
160. ENNE with time domain simulation and a linear simulator with frequency domain simulation Actually many circuits have data of interest in both the time and frequency Simulation domains which could warrant the use of both simulators For example an oscillator has phase noise transmission and phase characteristics which are all frequency domain measurements Oscillators also have waveform magnitude starting time and startup transients which are all time domain measurements In this case both simulators can be used in the circuit design There are some guidelines for deciding between CAYENNE and linear simulation 1 Does the circuit depend on time domain characteristics If so CAYENNE must be used for this portion of the design If the circuit depends entirely on the time domain CAYENNE can be used exclusively Howevet if a frequency domain response is also of interest linear simulation may be used in addition to CAYENNE Often both CAYENNE and linear simulation are useful in a design For example in amplifier design the linear portion gain matching can be done in the linear simulator and the device biasing can be done in CAYENNE U sing Ports in the Design Note The information in this section does not apply to EMPOWER simulation In GENESYS ports serve two different purposes 1 Ina top level design specified for stimulation ports used directly for simulation The impedances of the ports define the impedances use
161. ER planar ports cannot be used for ground referenced elements such as transmission lines even though the element might only have terminals Add extra details to listing file If checked extra information which can be used to double check your setup is inserted into the listing file Show detailed progress messages Turning this option off suppresses almost all output in the EMPOWER log The listing file is not affected Turning it off can dramatically speed up vety small runs Command Line Some options ate available which are not shown on this dialog box One common example is the O option which controls the size of the box for line analysis 105 Simulation NC If this option is used EMPOWER will allow de embedded ports to be away from the wall This option is especially useful for finline and slotline configurations VM Allow virtual memory usage To solve a complex problem EMPOWER always limits usage of computer virtual memory hard disk space in a rational way It will not use it for some numerically intensive parts of the simulation The option VM tells EMPOWER to use virtual memory more freely But even with this option the program stops calculations if substantial hard disk space is involved in some parts of the simulation Check the MEMORY lines in the listing file to have an idea how much memoty your computer lacks or how to reduce the problem Sg Use an alternate method of thinning out Global thinning
162. ERE ROT ere ee 185 EN A A 177 180 ROLA ELE a 5 cR dede e nA E 206 exporting data MES ados eei AR 8 E ET EEE E E EM ERE 180 EROALE DON eraa E 135 153 179 eea CIS aS ST 181 fast intermod shape saw dida ERR tapes esl 545 fast NEWCOM iaon 225 234 Te Oe atau NTS 225 De deum NTE PEE to OUR MM 489 flicker FM noise ccecccccssscssscsseessecssscsssesssceeees 271 Foeke PM ae aire es aac 271 TINTS Biss doceo SIS DIM MUI 271 Force Ic IDRPE Tarer aR 225 S NR 5 DEDOS o Ud 236 Preq D outta iit ara Ei 236 Freglades IM o eret n iiis 256 Deque FICIOs Los HINTEN UDINE 290 FREQUEN 7 2 CCU ACY sus EE 225 PORNO Ehud os neta cut esce RM cd lE og igi FeV are leche cr rece rearerererte ner 225 234 Pull aco plate Methodi 225 I NEL A TEE 239 508 AICI e Ar 11 292 296 297 GIN Go E MM 524 Ed RTI O 13 general background references 219 General RF Architecture Troubleshooting 532 Ceneta zed ocasiona tatus 206 208 o Oneralzed SCAbEBEIID euis c NI dH 177 generalized S parametets 142 183 generate viewer data 132 161 174 177 178 180 Generated Intermod Channel Power 522 generated third order intermod power 522 Genesys 2006 HARBEC Dataset Variables eee EEE 236 CINO To 522 GIMCP sss DD e 522 Ca scnianHHHHHEUI ee 13 E E EAEE EDU 297 E A OEA E 297 ONDE ut die 11515 292 29 GNI o A nU Md ei cdit 11 Sun 86 225 1 508 Uh 6c DS oq ere Ta OE Ra
163. Flag 0 from O inf exclude 10 100 exclude 200 400 The general format is parameter real integer ast of assignments where the st of assignments is a comma separated list of parameter identifier constant va ue range where value range is of the form from value_range_specifier exclude va ue range specifier exclude constant expression Advanced Modeling Kit where the va ue_range_specifier is of the form start_paren expression exibression2 end_paren where szart paren is and ezd parem is 11 and expression is constant expression inf where exptession2 is constant expression inf and where a constant_param_arrayinit is param arrayinit element ist where param arrayimit element fist 1s made of param arrayinit element param_arrayinit_element where param_arrayinit_element is a constant_expression The type real integer is optional If it is not given it will be derived from the constant assignment value parenthesis indicates the range can go up to but not include the value whereas a squate bracket indicates the range includes the endpoint The value range specification is quite useful for range checking Some examples of this ate parameter real Temp 27 from 273 15 inf parameter R 50 from O inf and value ranges can have simple exclusions parameter R 50 from O inf exclude 10 20 exclude 100 Analog Block Expressions and statements Condition
164. Frequency Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as ICP and TIMCP Travel Direction Same as ICP and TIMCP Output Intercept Receiver All Orders RX OIP This measurement is the receiver output intercept point along the path This is an out of band type of intermod measurement RX OIP n VICP n RX_Delta n dBm where n stage number and Order order of the intermod VTCP n ICP 0 CGAIN n dBm RX_Delta n VICP n TIMCP n Order 1 dB Delta is the difference in dB between the Total Intermod Channel Power in the main channel and the interfering signal present in the Interferer Channel including the effects of the order In order to correctly calculate OIP due to out of band interferers a Virtual Tone is created whose virtual power is that of an un attenuated in band tone This power level is simply the Interferer Channel Power at the input plus the Cascaded Gain This Virtual Tone Channel Power is different than the Interferer Channel Power measurement because the Virtual Tone Channel Power is not attenuated by out of band rejection whereas the Interferer Channel Power can be For in band interferers the Virtual Tone Channel Power and the Interferer Channel Power measurement will be identical In order to make this measurement a minimum of three signals tones must be present at the input e main channel signal used for cascaded gain and
165. GENESYS using decomposition e Deembedded or non deembedded ports e Viaholes including generated fields e Any number of dielectric layers e Dielectric and metal loss e Includes box modes and package effects e Slot mode for slot and coplanar circuits e Thick metal simulation with EMPOWER ML e 32 bit code for Windows 95 98 NT XP Basic Geometry All circuits in EMPOWER exist in a rectangular box as shown below The Media substrate layers each have specific dielectric and permittivity constants and loss tangents There must be at least two media layers One above the metalization layer and one below For standard microsttip there 1s a substrate below and air above For suspended mictosttip there are three media layers two air and one substrate For buried microstrip there are also three media layers two substrate and one air EMPOWER Planar 3D EM Analysis SIDEWALLS The dialogs below show two typical EMPOWER Layer Tab setups one for microstrip and one for stripline triplate The EMPOWER Layer Tab must be carefully checked when a new problem is created as it is probably the most likely source of errors when setting up an EMPOWER run 93 Simulation Create New Layout x General Associations General Layer EMPOMWER Layers Fonts Height or Tand Surface Imp V Sub Teflon oos Sub Teflon NENNEN Suk Sub Teflan mE w ay T Sue Teton a 2 om Ll E r ooo O Create
166. GHz The following rule of thumb generally works well use 10 Delay which is 20GHz in this example e The convolution Maximum Frequency is generally not very sensitive The most important thing is to be sure for example that it is not set to 100GHz if your circuit is operating in the 10 or 20 MHz range Convergence Criteria CAYENNE has three convergence criteria all on the General tab voltage current and relative We will refer to this as VoltTol CurrentTol and RelTol CurrentTol is sometimes called AbsTol since it is also used for voltage errors At each time point the following steps are used to determine whether or not convergence has been reached 75 Simulation 1 Find the maximum contribution for each node type At standard nodes this is the largest current from a single element going into the node In branch currents this is the maximum voltage At temerature nodes as in an LDMOS model this is the maximum power 2 Next the error at each node type is calculated According to Kirchoff s laws this should be zero This error must be less than a For current error in voltage nodes CurrentTol RelTol MaxCurrent b For voltage error in branch currents CurrentTol RelTol MaxVoltage c For power error in temperature nodes 1e 5 RelTol MaxPower d For other node types defined in Verilog A Within tolerances specified in the Verilog A source If these tolerances are not met at all nodes then the circui
167. HB oscillator analysis Measurements GENESYS 2004 and GENESYS 2005 11 Oscillation Measurement earlier and later 4 i5 solution FHB T requencies 2 Steady State FHB 1 Freq 2 261 Simulation 262 Frequency FOSHB MeasContext hb getspcomp Freq Qu l0 0 0 omall Signal 3 Oscillation Frequency FOSC O0 Osc_Freqs 1 Oscillator port current loscport MeasContext Oscillator port voltage Voscport MeasContext of the complex Frequency sweep impedance at the oscillator port node ZOSC ZOSC Trace of oscillation frequency TraceFprobe Trace of oscillator port current Tracelprobe 8 port voltage Vprobe Trace of oscillator TraceVprobe 1 1 tone HB oscillator analysis 2 Multi tones HB oscillator analysis oscillator with external signals synchronized oscillator self oscillating mixer 3 Created as a basic HB dataset variable if the Save Solution For All Nodes flag is set 4 Tracing vs Vprobe Tracing conditions Im Iprobe Vprobe 20 Tracing while Re Iprobe Vprobe lt 0 5 Independent variable Vprobe for tracing the oscillator measurements Harmonic Balance O scillator O ptions HARBEC Harmonic Balance Analysis Harmonic Balance Oscill
168. Home Front Ctrl Home Shows a front view of the current image This view is from the y axis at z 0 This option can also be selected by pressing Ctrlt Home Side Ctrl End Shows a side view of the current image This view is from the x axis at z 0 This option can also be selected by pressing Ctrl End Oblique End Shows an oblique view of the current image This view 1s top down on the x y plane with a slight offset This option can also be selected by pressing End Rotate The objects in this sub menu rotate the current image Rotate Left Left Arrow Rotates the current image clockwise in a horizontal plane perpendicular to the screen The center of the viewer image window is always the center of rotation Rotate Right Right Arrow Rotates the current image counter clockwise in a horizontal plane perpendicular to the screen The center of the viewer image window is always the center of rotation Rotate Up Up Arrow Rotates the current image forward in a vertical plane perpendicular to the screen The center of the viewer image window is always the center of rotation Rotate Down Down Arrow Rotates the current image backward in a vertical plane perpendicular to the screen The center of the viewer image window is always the center of rotation Rotate Clockwise PgDn Rotates the current image clockwise in the plane of the screen The center of the viewer image window is always the cent
169. IM IndexS1 IndexS2 PindBm hb ipn SpectrPout FreqIndexIM IndexS1 IndexSZ PindBm see more calculates the input intercept point of the 2 component of the power spectrum SpectrPout HARBEC Harmonic Balance Analysis PindBm input signal power in dBm hb oipn SpectrPout FreqgIndexIM IndexS1 IndexSZ hb_ipn SpectrPout FreqindexIM IndexS1 IndexSZ PoutdBm see more calculates the output intercept point of 2 component of the power spectrum SpectrPout where PoutdBm hb getspcompdbm SpectrPout FreqIndexIM IndexS1 see more the output signal spectral component power in dBm 3 Transducer Gain hb transgain SpectrIn SpectrOut FregIndexIM IndexIn IndexOurt see more calculates the transducer gain from input spectrum SpectrIn component with IM index IndexIn to the output spectrum SpectrOut component with IM index IndexOut hb_transgaindb SpectrPin SpectrPout FreqIndex1M IndexIn IndexOut db10 hb_transgain SpectrIn SpectrOut FreqIndexIM IndexIn IndexOut see more hb_gain In SpectrOut FreqgIndexIM IndexS hb_getspcomp SpectrOut FreqIndexIM IndexS In see more calculates the complex gain where In voltage or power amplitude of the input signal SpectrOut voltage or power output spectrum IndexS IM index of the output spectral component 4 Large Signal S parameters hb_LargeS Vin Vout sameport see more calculates the Large Signal S parameter LS par
170. If an EMPOWER simulation is selected electromagnetic results will be co simulated with the circuit elements associated with the layout Note If an EM simulation is selected it is very important that the Use Ports from Schematic option be properly checked on the EMPOWER Properties dialog Frequency Table and Order Control 226 HARBEC Harmonic Balance Analysis Name The schematic designator of the source GENESYS searches the specified design for all sources and places them in the table Freq The frequency specified on the source GENESYS fills in this value by reading the frequency from the schematic Order The number of harmonics to be analyzed The larger the number of harmonics the more accurately waveforms will be represented However the length of time to find a solution increases as roughly the cube of the number of frequencies Order must always be set large enough to model the majority of the energy in each branch current Typical numbers for mildly nonlinear circuits are 4 5 For circuits deep in compression square waves present the order may need to be 8 16 to achieve the desired accuracy Maximum Mixing Order Specifies the maximum combined order of signals to be simulated In the example shown all 4th order products will be calculates For example the 1900 2 1905 1 1800 1 95MHz the mixer third order intermodulation term is a 4th order term 2 1 1 and will be calculated This term only affects the mi
171. If we have reached the next time point needed for output then save data to the dataset 5 Repeat steps two three and four until the stop time has passed The most important concept is that the time points output in the dataset are NOT the same time points calculated by the simulation To ensure that the data at the output time points is accurate CAYENNE simulates at many many more time points that are not saved CAYENNE uses the following basic rules to set the internal simulation time steps 1 The first time step is set to the Minimum Time Step Also time steps are always at least as large as the Minimum Time Step This value is specified on the Integration Time Step tab and defaults to 1 of the Maximum Time Step Note The only exception is when the circuit cannot converge due to nonlinearities In that case smaller time steps may be used if absolutely necessary 2 The simulation time step will never under any circumstance be larger than the Maximum Time Step value as specified on the General tab 3 After a successful initial step CAYENNE will change the time step as follows a If the simulator is using fixed time steps the step will be doubled until the Maximum Time Step is reached The Maximum Time Step will then be used for the remainder of the simulation b Ifthe simulator is using robust truncation error controlled time steps then each point will be checked for accuracy BEFORE it is accepted The accuracy will be che
172. In Layout Place Another OUT_OUTPUT PartList gt Schematic 421 Simulation The following level diagram will appear System Data Path1 CGAIM Systemi Data Fathi CGAIN GAIN dE Follow the same process as adding a level diagram to add a predefined table of common measurements except select System1_Data_Path1 New Table of Measurements from the Add New Graph Table submenu For additional path measurement information click here The default table will look like System Data Path1 Measurements HodeHames Parts CHP CGAIH 1 CwWEource 1 113 82B 2 RFAmp 1 90 526 Attn_1 93 504 3 4 5 Couplerl 1 3 84 288 T a Izolator 1 94 793 HINT Right click on the table data to see additional table options 422 Spectrasys System Back to Run the Simulation Fundamentals General Behavioral Model O verview Behavioral models are unique in Spectrasys Because of the unique simulation technique that is used Spectral Propagation and Root Cause Analysis SPARCA behaviors for different types of spectrums can be modeled SPARCA supports the following types of spectrums e Signal e Intermods and Harmonics e Broadband Noise e Phase Noise The SPARCA simulation technique is so flexible that additional spectrum types can be added to support future needs Each model manages its behavior with respect these spectrum types Furthermore SPARCA knows which directions sig
173. J Ve 101 normal DOU S cee risit iR AS 136 normalized noise resistance eee 11 IND oc etu en DS 271 INDORE NEU MM t dus 5 D pU RS 5 Number of Curve Iterations 262 Number of Final Iterations 262 numerical acceleration procedures 217 OCT enemas 499 CONG UU 507 een eR 527 odd orders SOY e oi n esaeen RANAS 545 OPETE NAO aa N 540 offset channel frequency ieu ees 499 Offset Channel Power essentia 507 COTE tune tao ce DE dM 514 EDS uu M Mu uu M MM 514 Hnc MA 5 Gnesdimenistonal FET eoo a iones 229 OETDP S eeu un E II iUd 513 ODGTALOES S ases nt qn Mateo en neos o ese 14 ODE Vield ge coll open occi aC a eat 233 optimal admittance osito eats st etate 11 opumal e AE EO 2 Rc RUNG d ederet 11 Opti ZatlO ansni 10 11 13 14 optimizing simulation performance 234 Osc dPhaseZAF eere 257 USC S Gl Le Sra dede e rq A 257 Osc S Co ZAR 257 OSCIBEATO oed ne 225 249 OSC ALO desi ol oiu aevo vui SOe UI co 292 Oscillator Noise Simulation 271 Oscillator Port Initial Voltage 262 oscillatotsyalkthtOupli iue eite 249 Output 1 dB Compression ur re repe 513 Output InterCept esiin 514 Output Intercept Receivef eseese 515 output third order intercept esseere 514 Overamplhne FACE eo 225 parameter SwWeeD DIODGELES ninisi 42 PALATES CEE SWEEPS qan eee anai 38 234 paramete
174. LBILLILILILBITIMISE TRE ESEE E0 ERR SR SR B1 2 NBN L Cheep EPpy Ty ty ey Tt tt C rp TETT p rp p BSE tees f 6 OM E TEE E E FSE TET E E ES EE D JA Lin SEE SE EE i GLS E CEU CEE EUU HE E T SLL SEES ERE ER Eee Be Li ee oe p a 4 ML pata ata ame la abel all 8 ranana RRR ERREREOROEROERROUROREORE RE RR ORR OR ORT 6 CeCe eee eee eee eee eee eee LITT LETT CLEP ee rer te LI TL T A E E E E CL eee eee eee eee eee eee eee eee eee ee ee BEF PRR P PR N Cae Pee EBENEN mmm mm mm mm mm m mmmn LL eee eee eee eee eee eee eee eee eee eee OO BEF cece Pe EEEE BR eee ee Pee eee eee eee a Pe Eee 217 This looks like a pseudo non equidistant grid over the regular grid that is finer near edges corners and via holes and coarser inside the solid metal regions The enlarged secondary erid cells after the thinning out consist of non divergent current borders along each side problem and leaves the currents that are shown by the thick lines in the right half above that can be substituted by two variables on the grid using linear re expansion Thinning out is a simple elimination of the grid currents in metalized regions that can be represented by a smaller number of currents without loosing accuracy As an illustrative example the left half above shows a three resonator filter mapped on the grid The grid cells with possible non zero conductivity currents metalization regions are depicted by the thick line
175. Layer Tab are also shown in the EMPOWER Layer tab These layers are used for metal and other conductive material such as resistive film The following types are available e Lossless The layer is ideal metal e Physical Desc The layer is lossy These losses are described by Rho resistivity relative to copper Thickness and Surface Roughness e Electrical Desc The layer is lossy and is described by an impedance or file This type is commonly used for resistive films and superconductors If the entry in this box is a number it specifies the impedance of the material in ohms per square If the entry in this box is a filename it specifies the name of a one port data file which contains impedance data versus frequency This data file will be interpolated extrapolated as necessary See the Device Data section for a description of one port data files e Substrates Choosing a substrate causes the layer to get the rho thickness and roughness parameters from that substrate definition We recommend using this setting whenever possible so that parameters do not need to be duplicated between substrates and layouts Caution Unless thick metal is selected thickness is only used for calculation of losses It is not otherwise used and all strips are calculated as if they are infinitely thin Metal layers have three additional settings available Slot Type Check this box to simulate the non lossless metal areas as opposed to the metal areas
176. NCP i CNP i TIMP i CNP amp TIMP Same as OIP i ICP i Delta i ICP Order 1 Delta i ICP i TIMCP Same as OIP for 24 Order OIP Same as OIP for 3 4 Order OIP Same as DCP amp Output 1 dB Compression Point OP1DB i DCP LastStage DCP SDR min SDR i dBm SDR 493 Simulation 494 RX_OIP Output rx oip Same as OIP i VICP Delta i Intercept Point VTCP VTCP Order 1 Receiver All DELTA Virtual Tone Channel Power i Orders ICP 0 CGAIN Delta i VICP i TIMCP i Same as RX OIP for 2 4 Order RX OIP Same as RX_OIP for 34 Order RX_OIP Phase Phase noise power at CF Noise Phase Noise Channel Power PRNF Percent Noise Same as AN PRNF i AN i CNF iLastStage amp CNF PRIM prim GIMCP Same as IMREF GIMCP i GIMP CGAIN iLastStage CGAIN CGAIN amp PRIM i TIMP IMREF i TIMP iLastStage this is a ratio in Watts Where PRIM 0 0 9 o ICGAIN TIMCP PRIM2 primn Same as See PRIM PRIM 2 PRIM PRIM3 primn Same as See PRIM PRIM 3 PRIM SFDR SFDR Same as SFDR i 2 3 IIP3 i MDS Dynamic Range IP3 amp MDS RX_SFDRiReceiver RX_SFDR Same as RX_SFDR i 2 3 RX_IIP3 i Spurious Free RX IIP3 amp MPS MDS SDR Stage Dynamic SDR Sameas SDR i SOPIDB i TNP i Range TNP SNF Stage Noise SNF None Stage entered value Figure Dynamic Range Spectra
177. NFET PFET DEVICE_CLASS MOS NMOsS type 1 PMOS type 1 The general format of the keyword is DEVICE_CLASS Ape option varl valuet option2 var2 value2 o type is required and should be one of DIODE BJT BJT4 BJT5 FET JEET MOS RESISTOR CAPACITOR CCCS CCVS VCCS VCVS BJT4 adds a substrate node and BJT5 adds substrate and temperature nodes MOS supports both three and four pin devices 5 option option2 ate not required If they are given GENESYS will create one model for each option Additionally if varis not given for an option the value of the option will be set to 1 For the FET example above GENESYS will make two models with NFET and PFET added to the base model name For the NFET model the value NFET will be set to one Additionally NFET and PFET will not be shown as parameters in GENESYS GENESYS will also use appropriate symbols for any recognized option The following options are recognized by GENESYS BJT BJT4 BJT5 NPN and PNP FET and JFET NFET NJF PFET and PJF MOS NMOS and PMOS If var var2 ate specified they are set to the specified value instead of the option being set to a value Additionally the parameters referenced are not shown as parameters in GENESYS Otherwise they behave identically to the case above EAGLEWARE LAYOUT keyword Advanced keyword which allows overriding of footprints or association entries EAGLEWARE OPTIONS keyword Advanced keywo
178. Nonlinear noise analysis may be performed for one output voltage between 2 circuit nodes or for one output circuit port If you need to calculate noise figure you must define a 2nd port input port because the noise parameter is meaningful only with respect to two potts If the circuit has external injected noise sources place them where noise is to be injected and edit the component as required Noise sources are found under the Basic Sources palette Note When simulating noise figure noise sources should be added at the input and the input port must be Noisy which is set via the checkbox Noisy Port for the input port Harbec Noise properties page All other circuit ports that are not defined as input and output ports of Harbec noise analysis will contribute thermal noise to the output noise 2 Open the Harbec Analysis or Harbec Oscillator Analysis Noise properties dialog page 3 Select the Calculate Nonlinear Noise checkbox and define noise analysis parameters 4 To specify the nodes at which noise will be computed choose the element Ouput port from the pull down list of circuit element names Nodes of the circuit element define output for noise analysis The list includes port loads and 2 terminal elements from the upper level hierarchy of the circuit excluding elements with branch equations voltage sources and inductances 5 If output is a port it activates the Noisy checkbox for the output port
179. Not available on Smith Chart Normalized Noise Resistance RN The Normalized Noise Resistance measurement is a real function of frequency and is available for 2 port networks only The noise resistance is normalized with respect to the input impedance of the network Zo See the definition of Nosie Figure NF for a discussion of Rn Values Real value versus frequency Simulations Linear Default Format Table Linear Graph Linear Smith Chart none Commonly Used Operators none Examples Measurement Result in graph Smith chart Result on table optimization or yield RN noise resistance noise resistance Not available on Smith Chart Reference Port Impedance ZPORTj The reference impedance measurements are complex functions of frequency The measurements are associated with the network terminations The frequency range and intervals are as specified in the Linear Simulation dialog box A port number 7 is used to identity the port ZPORT is the reference impedance for pott z Values Complex value versus frequency Examples Measurement Result in graph Smith chart Result on table optimization or yield ZPORT 2 re ZPORT 2 real imaginary of ZPORT 2 mag ZPORT 1 Lineat Magnitude of ZPORT 2 Linear Magnitude of ZPORTT I1 ZPORT the ZPORT array Shows teal imaginary parts of all ports Not available on Smith Chart Getting Started Load Pull Load Pull Contours Note See example named Load Pull Con
180. Now button This launches Sonnet to simulate the layout Note GENESYS has been given a lot of intelligence to determine when it needs to calculate Clicking Recalculate Now will not do anything if GENESYS believes it 1s up to date To force GENESYS to recalculate from scratch right click on the Sonnet simulation in the workspace window and select Delete all internal files Once EMPOWER calculation is completed you can display the calculated data Add a rectangular graph called Sonnet Simulation and change the Default Simulation to Sonnetl Layoutl Add 21 and S11 to the graph The graphs below show GENESYS after Sonnet simulation Note that the Sonnet simulation will be slightly different from the EMPOWER simulation which was previously displayed mainly because EMPOWER was run without losses in this example Sonnet Interface Z GENESYS 2004 03 File Edit View Workspace Actions Tools Schematic Layout Graph Window Help OO Ge ad im a DA ASA gu 4 au e MRA ES Gs Combined Simulat _ fox E ayout1 Worksp Mormal sl Freq MHz DB 521 m DB S11 m Sonnet Simulatio Sag m F2000 Workspac O a El B Tuned Bandpass _ Analysis EM Layout Sij Lineari 1400 _ E Sonnet Simul J Sonneti Layi Tu Pa E2000 Scher EM Layoutl Layc J Equations P Equation Notes v Freq MHz Ill gt ni DB S21 DB S11 DB S21 DB S11 Click to select an object MEE Th
181. ODE ENE Yen a NA E LU a 470 Index Table O f Contents Single Sideband to AM and PM D conposItOfunaeseeted also albidis tube qde 473 behavioral MASS NOISE coste A A O 475 opecktumm Analyzer say desee sees ete Opto a Wiesel abba acorns eheess 478 Dy Ete SUS areca blair ede aaa bebe pool ne Pues T aula ada Ru uester pat evade siue 481 Directional Enetey INode Voltabe and Power sese dt tme ng ditta reves e aere 485 JranstmitieduE Gis ct od tado nte ue asleep dotati 486 Increasinp Simulation Spee istic acs ceca olei obtuvo iu east NU deo cb dada ive tea tu e iad aat 486 REGUCHIO the Tie S Oaren tdem etd ond ad ned E and neers 489 Measure POOL S iaire obe cite h A ct usted isabel liter wen vote pA rebua elu M e uei Uer du t MM ues 490 Spectrasys Measurement Inder rnaro a RUNDE UND AD DM det DM M MEME 490 Gehe E eso obe iva ded boe Donde rR Te ee Les dtc ooa ee rtm ee Rr en terme n 498 lov 502 hioc mre DR 524 B bros hootint e 522 General RF Architecture TroublesDOOtBg iiio teca eei da vtta ete Han equ ets de 502 How come my noise figure decreases through a cascade esses 533 Why don t cet thesame answeras my spreadsheets sisisi 534 NoAttenuduon Across q PIN et eene d peo Oe etu e obe ete o a ene Seah rbd 535 Dig lOe Bos Wemeren Ce T 536 System Simulation Parameters General Tab 23 ect pr tbe edet 536 System oimulation Parameters Paths Ta Daiano a etin dimito 538 ndo die Paus Dalon DOSoiade
182. Order Intercept will stay at noncoherent levels over most of the frequency range of the system However over narrow frequency ranges the SOI and TOI will increase to coherent summation levels In a well designed system where the equivalent intercept points of all the devices are equal the difference between coherent and non coherent summation is 4 to 5 dB When designing a system it is best to calculate the numbers for both the coherent and noncoherent cases to assess the variation likely to be expected over time and frequency McClanning Kevin and Vito Tom Radio Receiver Design Noble Publishing 2000 Comparing Coherency with Harmonic Balance 472 Spectrasys System Harmonic balance is a well established nonlinear circuit simulation technique Signals used it this technique have no bandwidth and all spectrums are of the same type In harmonic balance it is assumed that two spectrums having the same frequency by definition are coherent Coherency and SPARCA One of the great advantages of SPARCA 1s that not only is the coherent total spectrum available which is the only type of spectrum available in harmonic balance to the user but so are the individual spectrums that make it up This aids the user in understanding how the design is behaving and is also a great help during the architecture debugging process Single Sideband to AM and PM Decomposition This section will help the user understand how single sideband SSB s
183. Planar 3D EM Analysis 5 x ote EIN LI IE wo eh AS TT OMM Reel ar DE E T otal tato DB ETotal iat 169 Simulation 170 Examples This section illustrates the use of the EMPOWER viewer using a number of examples The WSP files for the examples are located in the VIEWER subdirectory in the GENESYS examples You should load them as you follow along The viewer displays current distributions as two or three dimensional graphs The viewer has several modes that are used to view various components of the currents from different view perspectives The best view of most problems is often found by minor adjustments of the view orientation The following examples include a few examples of such adjustments The examples are simple problems selected because the results are predictable Nevertheless they are interesting and illustrate concepts which may be applied to more complex problems Consider the possible graphs for a simple line segment analysis The schematic file for this example is METR16 WSX It contains description of a segment of the 50 Ohm standard stripline Rautio 1994 that is also discussed in the Examples Chapter The segment is 1 4423896 mm wide by 4 996540 mm long and the box size along the z axis is 1 mm The segment length is 90 degrees at 15 GHz and 180 degrees at 30 GHz Load METR16 WSP in GENESYS Run the viewer by selecting Run EMPOWER Viewer from the right click menu of Simulation
184. Port Defining a Ground Reference Re mapping Port Numbers Adding an EM Port to a Circuit You can add a port to a circuit from the Layout window If a layout has been created from a schematic all of the schematic ports will be created normal singular calibrated EM ports in the Layout and internal ports for all lumped element terminals You may 308 Momentum GX manually change the port modes in the Port Properties dialog or add any additional ports in the Layout editor EM Port Properties Draw Size Ref Plane Shift Cancel Port Number Help Location 88 363 1 0 285 mm w Layer I TOF METAL width lo Line Direction Default Current Direction Default Port Type Normal No Deembed Internal Polarity Differential Common Ground Ref Coplanar Height Single mode STRIP port transmission line extended calibration For normal and na deembed types setting width and or height to zero will cause LAYOUT to autodetect their dimensions Considerations Keep the following points in mind when adding ports to circuits to be simulated using Momentum Momentum does not use Current Direction which is an EMPOWER parameter only e Associate with port number and Polarity are used only for Momentum to define a group of ports Differential Common Ground Reference and Coplanar mode ports only e The parameter Line direction defines the feeding T line direction that the port is conne
185. R 337 Dalle aurcm EU 134 E S L e o OSEE lu ecd 92 ANNO QUEE Re retro 0 verersrsrsrsretr rer 130 206 TER EEE 5 E T e L EE RTTE 522 Mig 522 FINTE e 236 ENE aa eet renner re 523 ENP E 523 dco AE EA 531 togole backoround CO IDE auus anii dnce 170 POLCT AICO S odo nne MEL Me Hee LED EEI 86 225 tone channel fred ge Dey iarsira 502 tonc channel DONGPGuisiceseiei ci bal ge oi iiid 510 ojos M 92 160 Total Intermod Channel Power 522 Total Intetmod PONeE e rtt tec 525 COTA MODS OO Wel uisus deii iride 523 Total Node Voltage uus n dit 531 total third order intermod power 522 PPACEE PION E enoo en aeon ii t M USE 257 TracelproDe nar eas 257 TACEN ptoD6 essene 257 transducet essere 296 297 Tansu r G Ao dec dd UC Mee 238 DER SS No errr rte er ARE ULM EUM LE 5 8 298 transmission line types i die 299 transmitted ENE OY s comi c o R 486 EE od Ta PEPATE EE e eese AET 92 Troubleshooting Intermod Path Measurements No MM M E A E E T 465 EN OEE na edet eats fien UE 5 AY Ox POTE BIOL eiii 6 292 294 296 297 two port S parameters lo deneeten tees erece 9 202 WAAC team pled ausit a o RO dde 234 andateral sons sooo RUN QU UE 11 ptlaterA eA tae di bili DRM E DU 297 Unilateral bat cire eSssecanniennn iian 11 unnormalized Y parameter data 6 EENE CO ccc ee E A 11 UPA AE a E 233 Vse Chords MoO aaaea aN 225 use Krylov subspace method
186. R S Parameter data is plotted on a graph it will be normalized to this impedance Different impedances can be used for each port by separating impedances with commas A 1 Port Device Data File can be used in place of any impedance file to specify frequency dependent or complex port impedances 101 Simulation 102 Generalized When this box is checked the impedance for each line as calculated by EMPOWER are used for their terminating impedance See your EMPOWER manual for details on Generalized S Parameters Use ports from schematic Check this box when co simulating with HARBEC harmonic balance nonlinear simulation This forces all sources and impedances to be considered in the simulation Note Be sure to check Use ports from schematic if you will be using this simulation as the basis for a HARBEC Simulation otherwise there will be no nonlinear sources available Electromagnetic Simulation Frequencies Specifies the frequencies at which to run EMPOWERR If you have lumped elements in your simulation you can often turn down the number of frequencies here and increase the number of frequencies in the Co simulation sweep specified below Start Freq MHz Specifies the minimum frequency to analyze Stop Freq MHz Specifies the maximum frequency to analyze Number of Points Specifies the number of frequency points to analyze Points are distributed linearly between the low and high freq specified above HARBEC Freqs Sel
187. R den terret toS E A 380 Smulan Mellano OS S ie mtis cpu ORG aep exe ora uadit ueno e UD oou Ups abt e Oo eU 380 Simulating with Internal Ports and Ground References esses 381 IdvottyBons mnd Con deron reote mta na mem rae tectis teu a SN Soe SEE TREE RES 388 Matching the Simulation Mode to Circuit Charactetistics eee 388 Higher order Modes and High Frequency LitnitatlOtr nette 390 Parallel Pte de Sdtucaiiadasss nati a abonos Dati os setas LA t ecd a tuae red bd 390 SUr Ce Wave WOES st Mem ipie id sided EI 392 Slotline Structures and High Frequency Lirtati B npe ae epe pains ei enibotelusunt 392 VASU So O aoran fach a aa a a a rae 392 Via Structures and Substrate Thickness sii tation uoce oho eo xe ge ade 202 CPU TIimeand Memoir Redquiremetitssiu euo ud detur ie equat ue et e e utas 392 Bol npo uud EM n MEUM M T M ME NE 393 SONCE EACE canteen cust evsare RUE EVE Et erau iu ib ream de o Dr UD cec la vue a b cuc 395 Pesen P To Wee outdoor adt E cedd cn A 395 DO dS Ee ON VS Uy Ny usstoto tutatus t bo er DE ca eria fM ouo lobes fer pu e Ortu 395 I9eleuno Intertialibtulatot 02 Pd e antes be beoe iret morta c ete d repete EE 395 Manal Mode eara A den seis aibi tabo ced Eu 396 eean PAIL UU ERR TET RUNE SET TUN 398 bonnet Options Didor Bossini bees ERR e tuin i ip pot Et bti veta na oo ie quiete 398 Sonnet featutes supported im GENES Y S aosesedastoeisbar bte obest a Beirat bM pei etu 400 Sonnet features hot support
188. RT 3 PH Udeq Hybrid Hyhrid2 l IL D 2dB SPLIT2 1 IL D 2dB Bi CPL 3 0106 ISO 300B10 Mixed CPL 3DIdB R IL 3 2d610 Canvisain Bdbl SUMO LO 10dEm Click here for path dialog box information Level Diagrams Background A level diagram can display measurements of cascaded stages along a user defined path Each horizontal division of the X axis of the graph represents a stage along the path The first division represents the input to the cascade and the last division represents the output of the cascade Each vertical division is at the interface between the two stages The value of the measurements are displayed on the vertical axis Level diagrams give the user a quick visual indication of the performance of the entire cascade Node numbers are placed on the horizontal axis to show the node sequence of the path Furthermore schematic symbols are extracted from the schematic and placed at the bottom of the level diagram Sample Level Diagram 427 Simulation System Data_MamPath_CGAlN Sime System Data MainPath CGAIN D e F J l m Z z QO T ir e 12 Adding a Level Diagram Level diagrams can be added to the workspace in many different ways Common Level Diagrams Click here for instructions on adding common level diagrams Manually Adding a Level Diagram 1 Click the New Item button on the Workspace Tree toolbar and select Add Rectangular Graph
189. RU 504 CNOI E saraa E UM 271 E E E A dE MIN 506 INI E MM 503 Ci Oates rr eee rere 524 OTN edo e ei ATE A 525 goma Pe qoe OD uou osse E 91 el ol ches an Me E 470 coherent AGI OM csicceccacsscsssesancoaverdtnsoccosconoedsdl 545 COMBINE WGP iet ettet deae 145 COND stt t ue LB o A uat 508 compensation admittance scsisistavecsecusersroetesscice 181 COMPOSITE SOC CELUI a eit DR de 431 COMPTES SION resisder 5 229 299 COMPLESSO STARE xiuitiitere neta tid 508 Conducted Intermod Channel Power 521 conducted third order intermod power 521 Console ST dO Wn deti E Red Sac dei dot 106 constant noise circles esee 11 COR SCIT See ccn blunt eer er Ip o UU 11 COM COUTS sae AEAEE AIO eut 29 Controlling Analysis Data 439 EORVEPDCHCO ares eene vie sini iind 86 206 225 234 COPIAT E en A ea 152 eje 92 Cosimulation Momentum GX 370 COUPLE roS FID 2 ccarsnccsore a eA 174 COVES ACN dauid Mm bees e etes 131 EOVEDD DE doe ceret eee 131 EDL A t EEEE AAAA S 92 135 153 DE M E 506 creating dara de S EO 6 CHEATING data Fes sous oe reddo dede di 6 CEC AIO TIL OQUACTODI cocos UE related 433 current dIrectlOn iii edet 136 CUE ME VONAT ver eA 170 current viewer data 182 ON tondeuse a Mete A ELE DUE 525 data SHAM ON Secretariat ui en emend 1 db Equation Function sc eub etes 11 13 DENAU Oee EEEa 14 BEANO E ER 11 13 PPANC IOO cara
190. SC HH oe ccnl eis ADS 29 loade A ty Seni ere 13 14 lOS Sates ros OR RR dde ntu d eR 131 loss tane CI eee AA 92 VOSS A MM E E 151 lossy metil asnan aaa 132 lumped Clements eem 101 126 155 208 M mag Equation Function sss L5 NL NO S Md EEG 11 13 MATEON A lon a d 92 YAO TAC eernconesenmneateraSoassisissassasisasastatuacsed 170 PNA MEAG EHECES S Nomen ttc td tcd iud eee eerie 5 Max 1 1 Subter oN Sennen 225 Max Chords Jacobian Iter 225 Max Diagonal Jacobian Iter 225 Max Pull Jacobian Metresini 222 Max Newton Iter HONS eite irit 225 Max Source amplitude Iterations 225 Na SEC OPC UEVE une RU Us acddtctes 262 maximumapiplitude Step ime weite 234 Maximum Analysis Frequency 225 MAXIMUM mixing Ofdet caecos 225 234 AKI OTR aua eno EER 545 njaximui stable galt aieo aite 297 IMavell S eordtio DIS 24d ub weave weed 206 MD Soorten toc CERO ICE DC MM 513 hue cuire EON E EE ET 490 Measurement wizard asinina 11 measurements 11 13 14 15 86 177 292 426 faeasutino Spade ele ts arrini 292 MESE oaeiae o SAI MD cositas 346 IVlesh Cere POESIE SUE 357 meP E enana 92 130 134 e aO AAN 170 metallization AVER ariei 92 132 method of NCS aaan n diiit 210 Min amplitude factof sssssesseseesersrererre 225 Min Rel RHS Norm GBatisge ucc 225 minimum detectable smnalo soe oett 513 Dice c MU ME 552 M
191. SINPIM SEIPDPORT Variable NFSSB SSIPPORT Real drray i 1 SEIPSPMOISE SEIRPNOISE ESPORT SSIZPORT Harbec also computes the double sideband noise figure using the following equation NP roel f 10 1 uou UP Ro k Ty G G This is the value that would be computed by a noise figure meter Harbec calculates it internally and saves it in the dataset variable NFDSB If a noise simulation of a non oscillating circuit has both input and output ports defined Harbec calculates the SSB noise power conversion gain NCGAIN from the input port and input noise carrier frequency to the output port with output noise carrier frequency frequencies defined as HARBEC Harmonic Balance Analysis NCGAIN Fnoise dim Noise power of the SSB noise of the Output noise carrier propagated from 555 noise of the input noise carrier dBm Power of the 535 noise of the input noise carrier 10 Ed dB G Fuoise where G amp is the conversion gain of the input noise power for the input noise carrier frequency to the output noise power of the k th output noise carrier Fnoise is a noise offset frequency At the zero noise offset frequency and small nonlinearity of the circuit relatively to the input signal the noise conversion gain is equal to the power conversion gain from the input port frequency to the k th spectral component of the output port which is calculated from the spectral analysis of the circuit n Voute
192. Simulation Copyright Notice Copyright 1994 2007 Agilent Technologies Inc All rights reserved Notice The information contained in this document is subject to change without notice Agilent Technologies makes no warranty of any kind with regard to this material including but not limited to the implied warranties of merchantability and fitness for a particular purpose Agilent Technologies shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing performance ot use of this material Warranty A copy of the specific warranty terms that apply to this software product is available upon request from your Agilent Technologies representative U S Government Restricted Rights Software and technical data rights granted to the federal government include only those rights customarily provided to end user customers Agilent provides this customary commercial license in Software and technical data pursuant to FAR 12 211 Technical Data and 12 212 Computer Software and for the Department of Defense DFARS 252 227 7015 Technical Data Commercial Items and DFARS 227 7202 3 Rights in Commercial Computer Software or Computer Software Documentation Agilent Technologies Inc 1994 2007 395 Page Mill Road Palo Alto CA 94304 U S A Acknowledgments Mentor Graphics is a trademark of Mentor Graphics Corporation in the U S and other countries Microsoft Windows
193. The length of a vector from the center to a given point on the Smith chart is the magnitude of the reflection coefficient The angle of that vector with respect to the real axis to the right is the phase angle of the reflection coefficient Several common definitions are used to represent the length of this vector They are referred to as radially scaled parameters because they relate to a radial distance from the center towards the outside citcle of the chart Linear Analysis O utput Linear simulation produces the following variables Description Shape S Noise correlation matrix in S Parameter form N x N x M swept complex matrix Frequencies being swept M independent vector V S S parameter normalized to port impedances N x N x M swept ZPORT complex matrix C Z the schematic vector PORT Port impedances taken from the port elements in N x M swept complex Where M is the number of frequencies and N is the number of ports All variables and shortcuts except P are swept over all frequencies The following shortcuts are available see the Measurements Linear section of the Simulation manual for more details VSWR SWR at all ports vswt diag S N real vector VSWRz SWR at port 7 vswrt S z 7 Scalar ZIN Input impedance at all zin diag S ZPORT N vector ports including all other port terminations i i 301 Simulation 302 LL impedance at ew 2 ZPORT 4d i Complex In
194. To do this select None on the Edge Mesh section of the Mesh tab If edge mesh is required for individual layers use the Layer tab to set the edge mesh for a specific layer e Only in simulations using extremely high frequencies or in cases where extreme coupling is present between neighboring conductors is it necessary to use a combination of edge mesh and horizontal side currents When the Horizontal side current option is disabled meshing reverts to the previous model for thick conductors 1 e only the vertical currents on side walls are used 11 If you want to revert to default parameters click Cancel 12 Click OK to accept the global mesh parameters Calculating Pre M esh Momentum calculates pre mesh which is the union of all metal layers on polygons You can use this as an early diagnostic tool to allow checking the initial mesh formulation before calculating Momentum mesh In some cases the initial geometry is not correctly setup to create an accurate mesh For example using the synthesis tool M Filter a filter is produced which has right schematic but the wrong layout contains overlapping metal areas 351 Simulation 352 Por il Tu Ces TU TL4 AD 403mm err L 0 595mm LR TU Wed 403men Wax WO 4 3mm fee L3 117mm L 1 Di L 9 142m 153 T3 W 0 403mm Aer L 0 505mm LR fl We 630mm f amp Lend L 8 21me Uea We9 86 1mm Win Le 606mm L271 E i TLI1 WS 851m
195. To understand the problem you must first understand how a transient simulator analyzes a circuit with a capacitor Note The equations in this section are simple first order approximations sufficient to allow easy understanding of the issues In Cayenne more sophisticated and higher order methods are generally used The current through a linear capacitor is governed by the following equation I C dV dt For a simulator using finite steps this equation can be replaced by a first order approximation I C V1 V0 T1 TO where the time step being used is T1 TO the voltage across a capacitor at time TO is VO known and the voltage being solved is V1 the voltage across the capacitor at time T1 For this example use the following values VO 2 volts Vl TO 0 1 picosecond 10 P seconds 71 Simulation T1 0 2 picoseconds C 10 microfarad 10 Farads If we then use 2 0000000000000001 2 10 16 for V1 we see that the current is I 105 2 0000000000000001 2 2 105 10 15 I 10 Amps The problem is that there are no values between 2 and 2 0000000000000001 that can be stored in a double precision number Anything closer to 2 will simply round off to exactly 2 As a result given the time step of 0 1ps all solutions will cause the current will be some multiple of 10 Amps If the exact answer given infinite precision should be 6 10 Amps then we will never be able to get closer than 105 6
196. Travel Direction N A Stage Output Intercept Voltage All Orders SOIV This measurement is the stage output intercept voltage point entered by the user When a stage doesn t have this parameter 100 kV 1s used Each column in this measurement is for a different intermod order up to the Maximum Order specified on the Calculate Tab of the System Analysis Dialog Box The column number is the same as the order starting from the left with order 0 Channel Used No channel is used for this measurement Types of Spectrums Used None Travel Direction N A Stage Output Saturation Voltage SOVSAT This measurement is the stage output voltage saturation point entered by the user When a stage doesn t have this parameter 100 kV 1s used Channel Used No channel is used for this measurement Types of Spectrums Used None Travel Direction N A Stage Voltage Gain SVGAIN Spectrasys System This measurement is the stage voltage gain entered by the user For behavioral passive models the insertion loss parameter is used When a stage doesn t have either a gain or insertion loss parameter 0 dB20 is used This measurement is not dependent on the path direction through the model For example if the path was defined through backwards through an amplifier the forward path gain would be reported not the reverse isolation of the amplifier Channel Used No channel is used for this measurement Types of Spectrums Used None Travel Direction N A
197. U nnn a E E T A 130 Walke Cover SACO nn a A TA EO Eme ed 131 Do VG AA E E A E A A A 1 131 EOS yAn TE ERE 131 Dic qv Wb ic E 132 DIOE EV De SUC CLS cotes miss euedA oss O tUe e tne teres 152 Prete tired Cell COUPE ep easet dent See ub 133 T hicie Metal eiecti acted E E E N 134 EMPOWER ExtettiaUPOBScissadigettieteeno sd eto eade a aaa rdc ED 135 INCELE NETA TTN E NE EN O ENNE OE 135 Paano Eoen PO TH 135 EMP Ot sieur T 136 MY SS CA IA E 138 Mau Mode POTS sanea E em ner err err sar reer er te er 139 Generalized 5 DAFAIetets araen a a E E E E T NA 142 ENMPOWBER uDeGOIDDOSIHODqscatscuktipdasmdcadte dentus tap sveradbnsust E dedisbedabacantowabesetaeten 142 reu 142 lp c 143 opiral Inductot Example ocasion sur dieto do undi iue decides ated dbi e Cede 145 Porro E cease ace cance wea E E oat eee ee eas ke Td Port NUImDBFPIB rto b metto he ta tae e E es Cop un Ua f cuir tti t 151 EMPOWER Lumped Elements and Internal Potts retento 153 raso a E N NA 153 Plaeine Internal DOES nicata a A EAA 153 Manually Addins bumped Elements sneipas A R 154 AOE Dort PAC CNE ol eeno Pees eee E E Rite E 154 Panie and Ysireeted POPS aaa tuas O ud 155 ROTINE a T E E A A eS O EE Ia EMPOWER DG x MOI Sra a E E E N 157 Simulation viii VCE i m pete lalanigty 157 LHormocencousdcctanpgular CAVI ty as eu
198. W I branch name current spectrum at the branch Units A W_V lt node name voltage wave at the node created when the checkbox Calculate Wave Data is set Units V W I branch name current wave at the branch created when the checkbox Calculate Wave Data 1s set Units A Harmonic Balance Analysis Functions Note To use analysis functions in sweeps of analyses they must be defined directly in the analysis dataset and the checkbox Propagate All Variables When Sweeping of the Parameter Sweep Properties dialog window must be set 1 Get spectral component from the spectrum hb getspcomp Spectr FreqIndexIM IndexS see more returns the complex amplitude of the spectral component with IM index IndexS from spectrum Spectr The HB analysis dataset variable FreqIndexIM is the table of intermodulation IM indexes hb_getspcompdbm Spectr FregIndexIM IndexS dbm hb_getspcomp Spectt FreqIndexIM IndexS see more 2 Intercept Points hb_ipn SpectrPout FreqIndexIM IndexS1 IndexSZ see more calculates the relative part of intercept point equation Pim1 Pim2 Nim1 Nim2 for 2 components with IM indexes IndexSk of the power spectrum SpectrPout where Pimk the power in dBm of the k th spectral component with the IM index IndexSk Nimk the harmonic orders of the k th spectral component equal to the absolute sum of the IndexIMk components k 1 2 hb i pn SpectrPout FreqIndex
199. WER is run It should be carefully checked whenever a new circuit is analyzed especially if that circuit was described manually from a text TPL file The following sections describe the contents of a listing file Note Some of the information described below is only output if Output additional info in listing file is checked or La is specified QCHK SECTION This section allows you check the quality of the solution Entries include Min media wavelength to mesh size ratios should be at least 20 181 Simulation 182 Thinning out thresholds Specifies the maximum number of lines in a row which can be thinned out Max box size to media wavelength ratios If the box is too large you will have box resonances If this line ends with an exclamation mark it may be too large See the Box Modes section for more details PACKAGE STRUCTURE This section is only present when the Extra Details in Listing File option is used It gives a summaty of the substrate and metal layers used as well as cell sizes MEMORY SECTIONS Several memory sections throughout the listing file give memory requirements for different parts of the simulation MAP OF TERMINALS This section shows the grid representation of the problem SDIC SECTION Symmetry detection sections specify whether the structure is symmetrical The symmetry processing additionally shows where any differences occurred and can be very useful in finding out where the str
200. X will be displayed during simulation in the Status window indicating the change For information about internal ports refer to Defining an Internal Port Applying Reference O ffsets Reference offsets enable you to reposition single port types in a layout and thereby adjust electrical lengths in a layout without changing the actual drawing S parameters are returned as if the ports were placed at the position of the reference offset Why Use Reference Offsets The need to adjust the position of ports in a layout is analogous to the need to eliminate the effect of probes when measuring hardware prototypes When hardware prototypes are measured probes are connected to the input and output leads of the Device Under Test DUT These probes feed energy to the DUT and measure the response of the circuit Unfortunately the measured response characterizes the entire setup that is the DUT plus the probes This is an unwanted effect The final measurements should reflect the characteristics of the DUT alone The characteristics of the probes are well known so measurement labs can mathematically eliminate the effects of the probes and present the correct measurements of the DUT There are significant resemblances between this hardware measurement process and the way Momentum operates In the case of Momentum the probes are replaced by ports which during simulation will feed energy to the circuit and measure its response The Momentum port fe
201. X and the inside of this circle is shaded as an unstable region The available power gain G and power gain Gp are defined as G available from network power available from soutce Gp power deliver to load power input to network Note See the section on S Parameters for a detailed discussion of Gain Circles Values Complex values versus frequency Simulations Linear Default Format Table center magnitude angle radius Linear Graph None Smith Chart Circle Commonly Used Operators None Examples Measurement Result in graph Smith chart Result on table optimization or yield GA available gain circles center magnitude angle radius Linear GP power gain circles center magnitude angle radius Linear Available on Smith Chart and Table only Unilateral Gain Circles at Port 7 GU1 GU2 A unilateral gain circle at port 1 1s a locus of source impedances for a given transducer power gain below the optimum gain This locus which is specified via a marker is plotted on a Smith chart and is only available for 2 port networks The center of the circle is the point of maximum gain Circles are displayed for gains of 0 1 2 3 4 5 and 6 dB less than the optimal gain Similarly the unilateral gain circle at port 2 1s a locus of load impedances for a given transducer power gain below the optimum gain The transducer power gain G is defined as G power deliver to load power available from source For
202. a space of the line eigenmodes are set equal to Y matrices describing independent modes propagated in continuous part of the line segments It gives the basic non linear system of equations relating eigenwave propagation constants and characteristic impedances a matrix of transformation from the grid functions space to the mode s space transformation matrix and an auxiliary matrix that helps to match propagated modes perfectly compensation matrix Solution of the system is based on simultaneous diagonalization of Y matrix blocks Each port of the MIC structure or discontinuity can be de embedded using the pre calculated line parameters and the transformation and compensation matrices The main advantages of this approach are the possibility of multimode deembedding without direct spectral analysis of the line cross section and ideal matching of line eigenmodes in the analysis of the line segment that increases the accuracy of discontinuity analysis EMPOWER Planar 3D EM Analysis Note that despite the theoretical ability to excite and to match any propagating line eigenwave using the surface current sources in the metal plane it does not always work in the discrete models Using a limited number of variables in the source regions it 1s sometimes impossible to separate different modes completely Moreover the success of the MoSD application depends on the high order modes that could substantially influence the result This is the main drawback of
203. ab of the System Analysis The default number of simulation points per carrier is 2 Note This will increase the number of simulation points used by all spectrums derived from the source spectrum and will increase the simulation time and file size Add a path to the System Analysis on the Path tab Include option Add Powers Voltages and Impedances to Path Dataset located on the Edit Add Path dialog box of the System Analysis Add a linear analysis so the S11 of the system can be determined Place the Spectrasys path frequency vector i e F2_Desired in the System1_Data_Path1 dataset in the List of Frequencies used by the linear analysis so 11 will be determined at the same frequencies as the Spectrasys path voltages S21 for the system will need to derived from the equation relating E21 to 821 where the desired Spectrasys path voltage at the output node is equivalent to E21 Copy the following equation block code to the equation block in the workspace This code assumes the output node is named 2 The path frequency vector in the linear analysis and the voltage vector in the equation block must be change if the output node is other than 2 i e if the output node was 6 then F6 Desired would be used in the linear analysis and V6_Desired would be used in the equation block Use the group delay function gd SystemS21 to plot the group delay Equation Block Code using System1 Data Path1 Convert Spectrasy
204. able only Noise Correlation Matrix Parameters The noise correlation matrix elements are complex functions of frequency The frequency range and intervals are as specified in the Linear Simulation dialog box For a n noise sources the elements are of the form CS for 1 j equal 1 2 Note See References 5 6 for a complete discussion of noise correlation matrix properties Values Complex matrix versus frequency Simulations Linear Default Format Table RECT Graph RE Smith Chart none Commonly Used Operators Simultaneous Match Gamma at Port 7 GM The Simultaneous Match Gamma is a complex function of frequency and is available for 2 port networks only Computes the reflection coefficient that must be seen by the input port 7 to achieve a simultaneous conjugate match at both the input and output Values Complex value versus frequency Simulations Linear Default Format Table RECT Graph RE Smith Chart GMz Examples Measurement Result in graph Smith chart Result on table optimization or yield Getting Started GM real imaginary parts of gamma for all ports GM1 re GM1 teal imaginary parts of GM1 Simultaneous Match Admittance Impedance at Port i ZM 7j YMj The Simultaneous Match Admittance is a complex function of frequency and is available for 2 port networks only This 1s the value of admittance which must be seen at port 7 to achieve a simultaneous match at both input and output
205. above rules are met for the input signal as well as the LO signal A good example of this is an image reject mixer A single input port is split 2 ways that drive the input to 2 mixers A single LO signal is also split 2 ways and phase shifted before being applied to the mixer LO ports The mixer outputs are combined back together to form the image reject mixer output Since both mixers have the same input source as well as LO source then all signals that have the same type frequency and bandwidth will have the same coherency number NOTE The coherency number is displayed in the spectrum identification information This will aid the user in understanding their circuit operation as well debugging any problems See the Spectrum Identification section for more information NOTE Phase noise uses the coherency of its parent signal spectrum Coherent vs Noncoherent Addition Coherent addition is more conservative than noncoherent addition i e the coherent assumption indicated a less linear system than the noncoherent equations indicated Ina worst case scenario coherent addition should be used When designing low noise receiving systems it was found that well designed cascades usually behave as though the distortion products are adding up noncoherently For the most patt these system have achieved the equivalent of noncoherent summation plus one ot two dB With wide band systems the cascaded SOI Second Order Intercept or TOI Third
206. accuracy of the RF mode models declines smoothly with increased frequency Similarly if the substrate allows the propagation of surface waves these are guided waves that propagate in the substrate layers the accuracy of the RF mode will gradually decline because surface waves are not included in the RF calculation Electrically Small Circuits RF mode works best for electrically small circuits as its accuracy smoothly decreases with increasing electrical size relative to a given frequency A circuit is considered electrically small relative to a given frequency if its physical dimension is smaller than half the wavelength of the frequency Depending on which value you know maximum circuit dimension or maximum simulation frequency you can determine a qualitative approximation of the circuit s electrical size Suppose you have a circuit with dimension D as shown in the following figure 388 Momentum GX For space wave radiation you can use one of the following two expressions strictly as a guideline to have an awareness about the circuit s electrical size relative to the maximum frequency you plan to run the simulation When you know the value of D use the first expression to approximate the maximum frequency up to which the circuit is electrically small When you know the maximum simulation frequency use the second expression to approximate the maximum allowable dimension Pee D Of 150 De F whete D the maximum
207. ace LNMIT3 WSP in this case before the next run Via Hole Viewer Example The last visualization example shows a structure with non zero X Y and Z current components A segment of microstrip line terminated by a via hole from Swanson 1992 is described in the file VIA WSP The line is 12 mil wide and is terminated by a metal square 24 by 24 mil with a 13 mil diameter circular via hole in the center The substrate height is 15 mil and the relative permitivity is 9 8 The box size is 120 by 120 mil Load this example in GENESYS and run the EMPOWER viewer The first figure below shows the time averaged plot View Menu Switches Value Mode or Value Mode button for additive XY current density distribution The view point is the oblique view with a few minor adjustments The plot shows how the dominant microstrip line mode currents spread across the square metal pad You can see the typical peaks in the current density function in the vicinity of the metal internal corners where the surface current changes flowing directions Toggling to the X and Y components of the current XY X Y Z button you can investigate how the surface currents change direction in different parts of the structure Switching to the Z current visualization mode will show a plot like the 175 Simulation 176 second figure below Note that the scale for the Z directed currents is in Amperes and not current density Each current represents a volume current density integrat
208. age at node Net 20 hb getspcomp Vnet 20 F reqIndex1M 1 0 Order index of the spectral component Not implemented with lt IM index gt findrow FreqIndexIM lt IM index gt The negative value means negative frequency of the spectral component 0 if the spectrum does not have the spectral component having the lt IM index gt Example P2 abs findrow FregInde xIM 1 1 LargeSContext OutPort InPort MeasContext OutFreq InFreq And based on it functions MixerS OutPort InPort 1 Large signal OutFreq InFreq 2 S parameters InFreq LargeS OutPort InPort LargeSdb OutPort InPort LargeSang OutPort InPort Examples MixerSang OutPort InPort OutFreq a LS parameters for the base tone frequency hb LargeS Vin Vout same port a LS parameters for frequency conversion 246 HARBEC Harmonic Balance Analysis Using PowerSweepl1 Sch1 LS21 LargeS 1 1 or LS21 LargeSContext 2 1 1 1 or LS21 LargeSContext 2 1 PowerSweep1 Sch1 1 1 hb_LargeSmix Vin Spectr Out sameport FregIndexIM IndexOut Example a LS parameters for base tone LS11 hb LargeS Vinl VPORT 2 1 1 LS21 hb LargeS Vinl VPORT 2 2 0 LS22 hb LargeS Vin2 VPORT 2 2 1 LS12 hb LargeS Vin2 VPORT 2 1 0 b LS parameters for frequency inulti pher by 3 LS21mix hb LargeSmix Vin1 VPORT 2 0 FreqIndexIM 3 where Vin1 Vin2 are the amplitudes of signal port volta
209. ain to the spatial domain To do it an auxiliary array called general sums array is introduced The dimension of the general sums atray is also about 3 L M Each element of the GGF matrix can be obtained as a sum of four elements of the general sums array The general sums array depends only on the box and media structure and the grid cell size Its elements are calculated via the discrete Fourier transforms of the GGF eigenvalue vector using the Prime Factor algorithm This stage is based on the maximal utilization of internal symmetries of the bounded equidistant grid and usually takes negligibly small CPU time Moteover it can be done only once for all structures with the same box media and grid The described technique is quite similar to the main matrix filling procedure designed for the spectral domain technique Hill Tripathi 1991 except that it has been done here in finite space and we calculate the GGF matrix elements without additional truncation or series summation errors It can also be reformulated in matrix form in accordance with Pregla Pascher 1989 The GGF matrix can be represented by a sum of Toeplitz and Hankel matrices and their rows can be obtained directly from the general sums arrays Informational Multiport The informational multiport term was introduced by B V Sestroretzkiy 1987 and in a nutshell means a model multiport that reflects electromagnetic properties of an object before superimposing an additional boundary
210. al from other desired signals at the same frequency Add Part Node to Path Button When clicked provides the user with a menu of parts or node names that can be selected in sequence to determine the path Each click on the menu will add the selected name to the path Spectrasys System Clear Path Button When clicked with clear the path list Path This contains a list of part or node names that define the path These names ate specified as an array and can be separated either with semicolons ot commas SPECTRASYS will find the shortest path containing all the names in the path list Additional names can be inserted into the path list to force a path in a given direction The path can be defined as a string array in an equation block to dynamically change the path based on an equation For example a string atray for the variable MyPath would be defined MyPath Input Part Name Optional Names Output Part Name Note This example uses part names but node names work equally as well However the correct selection of Part Names or Node Names must be selected Also commas ate NOT supported as a valid separator in a string array A semicolon must be used Channel Frequency Defines the frequency at which path measurements are made The channel measurement bandwidth will be used in conjunction with the channel frequency to completely specify the channel location in the spectrum and bandwidth of the spectrum integration I
211. al statement if else statement The conditional statement is used to determine whether a statement is executed or not The syntax is if expression rue statement or null else false statement or null If the expression evaluates to True non zero then the ue_statement will be executed or not if false If there is an else false_statement and the expression evaluates to False the false statement is executed instead Case statement A case statement is useful for multiple actions to be selected based on an expression The format is case casex casez expression 53 Simulation 54 case_item case_item endcase where case_item is expression expression y statement_or_null default statement_or_null The default statement is optional but if used can only be used once The case expression and the case_item expression can be computed at runtime neither expression is required to be a constant expression The case_item expressions are evaluated and compared in the exact order in which they are given If one of the case_item expressions matches the case expression given in parentheses then the statement associated with that case ez is executed If all comparisons fail then the default item statement is executed 1f given Otherwise none of the case_item statements ate executed Repeat and while looping statements The repeat statement executes a statement a fixed number of times Evalu
212. alance simulations DC analysis is very fast and will help to make sure that you have entered a workable design Note DC Simulation is not generally the same as the DC zero frequency level from a harmonic balance simulation In DC simulation all AC sources are turned off Nonlinear device models have many parameters that can be entered in error To make sute that the model is correct it is a good idea to look at the DC characteristic curves of the device before entering a complete circuit Workspace templates are available Select New from the File Menu then use the BJT Test wsp template that make it easy to create these curves In addition to analysis DC results can be optimized For example you can optimize bias resistor values to achieve a desired collector current and voltage for a bipolar transistor See the walkthrough DC Analysis Verifying Transistor Parameters for an example It is located in one of the following sections To add a DC analysis 1 Create a design with a schematic Include DC sources and use nonlinear device models Linear or S parameter models will generally not produce accurate results at DC 2 Click the New Item button SI on the Workspace Tree toolbar and choose Add DC Analysis from the Analyses menu 3 Define the Analysis Properties and click Calculate Now 4 The original schematic may now show DC voltage levels and a dataset is created that includes DC operating point values 85 Simulati
213. alf of the figure illustrates the correlation of other types of the boundary conditions to operations with the informational multiport terminals Operations with the z directed terminals are similar The operations in a discrete space of the informational multiport terminals are completely in accordance with the usual electromagnetic theory To connect a lumped element for example we performed both serial connections of terminals along the element that corresponds to the electric field integration along the element and parallel connections that corresponds to the surface current integration across the element see the Table EMPOWER Planar 3D EM Analysis Before filling the reduced GGF matrix we can additionally decrease the GGF matrix order and required storage space by means of thinning out with linear re expansion procedures electromagnetics Note that the examples given are not the only possible manipulations and by incorporating a geometrical symmetry into the problem above The analogies described are meant to facilitate understanding of numerical with the terminals with physical electromagnetic equivalents Numerical Acceleration Procedures ERE REE EELEE ll ll lll ERE RE j l De eee ee Eee EE ESES es I LL IM E 3 INI IN E E E BR pa a ma afa an amn 8 18 12 1 E Tana T 1L EN oss mE TES SRE 1 DB 1 nDo 9H TEC BENE B D B BE u l EN EEEE ER Bee Eee IEE 2 IERSBERIBSBBIBIBIILLLLILILBILILI
214. alue G eneral Layers The general layer settings for this example are shown below Click on the General Layer tab in the LAYOUT Properties dialog box 111 Simulation LAYOUT Properties Zj General Associations General Layer EMPOMWER Layers Fonts Lee ae nee eee gems he ee x x rrr wen hee Nee e MEC c Er Hee e NMi r ewet Ho Mir eret ECCE one mooie G e Woe M m Mjo js Load Fram Layer File Save to Layer File Insert Layer Delete Layer Only three layers had to be defined for this filter e Top Metal e Substrate e Bottom Metal These are the only layers that are needed to simulate the microstrip filter For a general layout more layers are often included for purposes only For example defining a silk screen or mask layer would not affect simulation since none of the filter metal is placed on those layers Note Since the bottom of the box will be used as a ground plane the bottom metal layer defined above is not technically necessary However since it is often necessary for manufacturing reasons it is normally defined here EM PO WER Layers Next click on the EMPOWER Layers tab The EMPOWER layers are automatically selected from the available general layers see the previous section They are chosen from the available metal and substrate layers and can be enabled or disabled for EMPOWER simulation 112 EMPOWER Planar 3D EM Analysis LAYOUT Properties E Gene
215. am Properties eT Graph Properties Default Dataset or Equations Systern1_Data_MainPath w Measurement Label Optional On Right Hide Color Add Series sg Measurement Wizard EN Equation Wizard Left Y Axis Right Axis m Axis duto Scale Log Guto Scale Log Auto Scale Log Scale Add a Series then enter dataset measurement Min Min En graph a measurement Max Max or press a wizard button to quide you through Units dB Units Mane Units the process of defining a measurement Divisions Divisions Divisions E 430 Spectrasys System Click here for additional information on graphs Getting the most out of a Level Diagrams Use the right Y axis to examine additional path measurements Change stage parameters directly from the level diagram by double clicking the part at the bottom of the level diagram Use the mouse wheel to zoom in an out when the path contains many stages The X axis range can be set manually by the user NOTE Indexes are used in this case NOT node numbers Index 0 is the first node along the path HINT Use tables not level diagrams when troubleshooting problems More parameters can be examined at the same time with tables than level diagrams Checking channel frequencies and power levels are very important during the troubleshooting process Spectrum Plots and Tables Spectrum plots in Spectrasys are unique because of the type of information di
216. ameter Vin the signal port voltage source amplitude Vout the complex amplitude of the output spectral component Vout hb_getspcomp SpectrV out FreqIndexIM IndexOut sameport 1 if input and output ports are the same and 0 otherwise hb_LargeSmix Vin Spectr Vout sameport FreqiIndexIM IndexOut hb_LargeS Vin Vout sameport see more calculates the LS parameters used with multi tone HB analysis or for frequency conversion Measuring Gain in HARBEC HB analysis 1 Tone Gain and compression 239 Simulation 240 Example The tested Circuit Amplifier has 2 ports input 1 and output 2 The 1st port is 1 tone sional source with frequency 850 MHz To analyze the circuit create HB analysis HB1 Amplifier Port 1 PORT 1 Port 2 Re500 0 4 b PORT22 F B50MHz F a Z0 500 z PAC 40 dBm F c3 Partlist Schematic The compression of gain is defined as Compression Pin Gain Pin GainSS where Gain dbmPout dbmPin and GainSS the small signal gain which is calculated in the HB analysis as Gain at very small input power when it is independent of its value To calculate the function create a Parameter Sweep for HB1 analysis PowerSweep1 whose parameter is the input power P from 30 to 10 dBm with step 5dBm For the tested circuit the gain is almost constant for input power lt 20 dBm Using it defi
217. ameters and noise parameters of a schematic design If the circuit contains nonlinear elements DC simulation is first run automatically and the nonlinear devices are linearized at the DC operating point Port Impedance Usage The design must have input output ports in sequential order Balanced Ports are fully supported All impedances are specified on the design s ports To use a complex port impedance you can either directly use a complex number in the port termination or you can use a 1 pott data file To add a linear analysis L 3 4 Create a design with a schematic Click the New Item button a on the Workspace Tree toolbar and choose Add Linear Analysis from the Analyses menu Define the Analysis Properties and click OK The analysis runs and creates a data set Note S Parameters are calculated with respect to ports you must have ports in the schematic numbered in ascending order from 1 to run a linear analysis See also Measurements later in this manual Outputs Overview User s Guide 289 Simulation Linear Analysis Properties Linear Analysis Properties zm Liencies mM o B A Mame Dataset Description Frequencies Data v Automatic Recalculation Mi DC Analvsis default s Frequency Range Start MHz Stop MHz Advanced Preferred Reduction Size Type OF Sweep Linear Number of Paints Log Points Decade CO Linear Step Size MHz Lis
218. ance such as 50 or 75 ohms However EMPOWER will give much more accurate results if you use generalized S Parameters With generalized S Parameters instead of the ports being terminated with 50 or 75 ohms the ports are terminated with the characteristic impedance of the line as calculated by EMPOWER This is a more internally consistent representation and the results are often far more accurate You should use generalized S Parameters if the following three conditions hold 1 You are using normal deembedded ports Ports marked No Deembed or Internal are not appropriate for reporting generalized S Parameters so they are normalized to 50 ohms if generalized parameters are requested 2 You have calculated the impedance of the lines at the ports using T LINE for instance and they are 50 or 75 ohms 3 You have run EMPOWER but it calculated the port impedances to be a little different for example 47 instead of 50 ohms This error is generally a result of the grid size A finer grid would result in less error in the impedance In this case you know that your port lines should be 50 ohms but EMPOWER reported 47 ohms If you then request Generalized S Parameters GENESYS will also use 47 ohms for the terminating impedance and a large part of the analysis error due to the grid will be cancelled The results will be close to the results obtained 1f you measured the circuit in a 50 ohm network analyzer To get generalized S
219. and equations to be solved simultaneously CAYENNE Walkthrough In this example we will modify a simple circuit to include a transient analysis First please load the example circuit Bridge T 1 Select New from the File menu If you have disabled the Start Page you will need to select Show Start Page from the Help menu instead 2 Press the Open Example button and select the Bridge T wsx file from the root of the examples folder The response should look like the figure below Note that this circuit has a substantial amount of loss at low frequencies D x Z BRIDGET lox PORT 2 su AE sp dnog m T Nees T D un 1 8 od s Naan SERERE PR E Group delay Es PartList PartList E Schematic 65 Simulation 66 All transient analyses must have a source The source may simply be DC allowing you to view startup and self oscillation Or the sources may be waveforms allowing the display of a transient response to a stimulus In our existing schematic there are no sources so any transient analysis would be all zero To perform a meaningful transient analysis on this example we will replace the input with a periodic pulsed waveform To do this and also add a transient analysis and display results 1 25 3 10 Delete the input port of the schematic From the part selector Filter by Pulse Find the Input Pulsed Volt model and place it on the schematic where the old inp
220. anged in 2 ways 1 adding more evenly spaced noise points across the entire noise spectrum and or 2 adding mote evenly spaced noise points around a user specified bandwidth of desired frequencies 467 Simulation 468 4 Point Noise Spectrum 20 40 4 50 50 100 120 Output Power dBm dBm 140 160 180 200 0 au 100 150 200 250 300 350 400 450 500 Frequency MHz Adding 15 noise points across a 100 MHz bandwidth around the desired signal of 250 MHz show a more accurate representation of the filter noise pedestal NOTE Improving the noise spectral shape will generally not improve the accuracy of cascaded noise measurements unless wide measurements are used in systems with narrow filters Spectrasys System 15 Points Across 100 MHz BW 20 40 50 100 120 Output Power dBm dBm 140 160 180 200 0 50 100 150 200 250 300 350 400 450 500 Frequency MHz Smart Noise Point Insertion Another great benefit of the SPARCA simulation technique is that we know which spectrums are desired and which are not Having this knowledge noise points can be added at the correct frequencies to ensure noise data is collected at the frequency of interest Adding these noise points to the simulation is called Smart Noise Point Insertion Without this technique noise simulation through filters would assuredly fail as shown in the figure using 4 noise simulation points
221. annel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as AN and CNF Travel Direction Same as AN and CNF Percent Intermods All Orders PRIM This routine calculates the Percent Intermod Contribution by each stage to the final Total Intermod Channel Power of the path IMREF Equivalent Intermod Power Referenced to the Output IMREF GIMCP n CGAIN nLastStage CGAIN n PRIM n IMREF n TIMCP iLastStage this is a ratio in Watts Where PRIM 0 0 n is the current stage and nLastStage is the last stage along the designated path This measurement will help the user pinpoint all stages and their respective contribution to the total third order intermod power of the selected path This measurement in unit less since the measurement is a percentage There can cases where the percentage sum of all the stages in the path does not equal 100 For instance if the architecture contains parallel paths then each path would contribute to the total third order intermod power but only a single path is considered in this measurement Another case would be where there are sufficient VSWR interactions between stages that effect the intermod levels Reducing the architecture to the spreadsheet case will always yield the expected spreadsheet answers with respect to percentages Sometimes this measurement can be greater than 100 if the equivalent intermod power referenced to the output is greater than the actual total int
222. arameters This measurement simply returns the Offset Channel Frequency for every node along the specified path Desired Channel Resistance DCR This measurement is the desired average resistance across the main channel along the specified path This is not the magnitude of the impedance but its real part This measurement includes ONLY DESIRED SIGNALS on the beginning node of the path traveling in FORWARD path direction All other intermods harmonics noise and phase noise signals are ignored Note A D is placed next to the equation in the identifying flyover help in a spectrum plot to indicate desired signals For example if the Channel Measurement Bandwidth was specified to 03 MHz and the Channel Frequency was 220 MHz then the DCR is the average resistance from 219 985 to 220 015 MHz This resistance measurement will not even be affect by another 220 MHz signal traveling in the reverse direction even if it is much larger in amplitude Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used ONLY DESIRED SIGNALS Travel Direction Only in the FORWARD direction Group Delay System This section will demonstrate how to measure group delay in Spectrasys BACKGROUND Group Delay measures the differential time delay caused by a filter i e it indicates if certain frequency components will be delayed more than other components and by how much Radio Receiver Design Kevin Mc
223. arameters and other data that doesn t change during the simulation 3 Save the file For example if the workspace tree was 3 4 Path Data Workspace Tree ox So a am Default E E Designs i Svstem1 Data Path1 i i Systemi Data Pathe i Schi Schematic System Schl i i System Data H eee Equation Z5 Notes Then the file size could be reduced by deleting 489 Simulation e System1_Data_Path1 e System1_Data_Path2 e System Data Keeping Node Data of Interest The node data that the simulator retains is controlled on the Output Tab of the system analysis System Simulation Parameters General Paths Calculate Composite Spectrum Options amp utput Retain Leyvelfs of Data Save Data For Splitter 3 CheckAl Mixer Filter J Uncheck All Output Interferer 5 Check Output Ports a 35M Ose Check Input Ports lil Factory Defaults Only checking the devices of interest will reduce the file size NOTE Even though some devices are not checked all data for these devices are calculated during the simulation These output options only affect the data being saved to the system analysis dataset M easurements Spectrasys M easurement Index Spectrasys measurements are broken up into 3 groups General Power and Voltage 490 Spectrasys System Spectrasys General Measurements Name Description Syntax Spectrum Equation Type ACF Adjacent Channel Frequency
224. are on different layout layers But two vias can be coincident on the same layout layer e Vias should be wholly contained within any primitives that are to be connected e Vias can be specified with loss but are considered to have zero thickness Vias through Multiple Substrate Layers Vias can be applied to multiple layers so that they cut through one or mote layers When drawing vias that cut through multiple substrate layers each via must be drawn only once on a single layout substrate layer The Momentum GX simulator expands the via object in both directions up and down up to the 1 metal layer met in each direction 333 Simulation In Layout op view vee One via drawn on one Layout layer In Momentum Side view Four substrate layers have one via process layers mapped to each one In Genesys Layout vias for Momentum may be drawn different ways 1 Any layout object line polygon circle pad etc drawn in a dielectric substrate layer expanded in both directions up to a first metal layer or metal cover in each direction creates a via This is a Momentum type via For example we use the substrate stack with 2 metal layers and the dielectric layer Default 0 between them To create a via by this method draw any layout object in the dielectric layer Default 0 for example place a pad 334 Momentum GX Pad Properties ad Diameter Of O Square Rect CO Wagon W
225. are the more data that needs to be analyzed most of which does not affect the answer in question 1 Disable all possible calculations like intermods and harmonics thermal noise and phase noise that don t affect the answer in question Eliminate multiple carriers that don t affect the answer in question Disable the spectrum analyzer mode if being used this is a display only type of tool and does not affect any path measurements Increase isolation for components having more than 2 ports This eliminates spectrums propagating multiple times around a loop In really tough cases stages can be removed from the analysis by the option on the simulation tab of the part properties dialog box Disable Part for All Simulations Short Circuit ALL terminals together This works well for 2 port devices How come my noise figure decreases through a cascade The equation for Cascaded Noise Figure measurement in Spectrasys 1s 933 Simulation 534 CNF n CNPIn CNP 0 CGAIN n dB where n stage number and CNP is the Channel Noise Power and the Cascade Gain measurement is CGAIN n DCP n DCP 0 dB where n stage number and DCP is the Desired Channel Power Cascade Gain is therefore a function of all forward traveling power in the channel which is subject to VSWR effects Verify that Gain and Cascaded Gain are as expected Another issue usually is that the Channel Measurement Bandwidth is much wider than the c
226. arts of all Y Parameters x Not available on Smith Chart Z Parametets This Z parameter or zzpedance parameter measurements are complex functions of frequency The frequency range and intervals are as specified in the Linear Simulation dialog box The Z parameters for an n port network are of the form ZP for i j equal 1 2 n For a two port network the equations relating the input voltage V1 and current I1 to the output voltage V2 and current 12 are V4 ZPii li ZP 12 D V5 ZPa li ZP22 Io Values Complex matrix versus frequency Simulations Linear Default Format Table RECT Graph RE Smith Chart none Commonly Used Operators Operator Description Result Type re ZP22 real part Real magang ZP21 Linear magnitude and angle in range of 180 to 180 Real Getting Started Examples Measurement Result in graph Smith chart Result on table optimization or yield LP 22 re ZP22 real part of ZP22 mag ZP21 Linear Magnitude of ZP21 Linear Magnitude of ZP21 ZP Shows teal imaginary parts of all Z Parameters Not available on Smith Chart Voltage Standing Wave Ratio VSWR The VSWR measurement is a real function of frequency The measurements are made looking into the network from the port with other network terminations in place The frequency range and intervals are as specified in the Linear Simulation dialog box A port number 7 is used to identity the port VSWERz is the Voltage Standing Wave Rat
227. ase Noise Simulation An oscillator phase noise analysis computes the noise sidebands of the oscillator carrier frequency as well as Normal frequency conversion Amplitude noise to frequency noise conversion Frequency translation of noise caused by component nonlinearities in the presence of large signal oscillator signals Up converted flicker noise is a commonly observed effect Bias changes due to the oscillator signals Any shift in DC bias that occurs in the presence of the oscillation waveforms is taken into account This bias shift calculation is needed for accurate calculations of nonlinear device noise The results from the phase noise analysis have NPM phase noise in dBc Hz and NAM AM noise in dBc Hz for specified output circuit nodes or port How Hatbec Simulates Phase Noise 279 Simulation 280 Phase noise in an oscillator can be analyzed from small signal mixing of noise The small sional mixing of noise comes from the nonlinear behavior of the oscillator where noise mixes with the oscillator signal and harmonics to mix to sideband frequencies on either side of the oscillator signal The phase noise computed is available to the user directly in dBc To model oscillator phase noise with a noise mixing analysis the noise at the sidebands on either side of the carrier are obtained from a small signal mixer analysis where noise sources mix with the oscillator large signals to produce noise sidebands The n
228. associated Single Mode port with same ref plane Internal part The following table provides specific information about the Port Dialog Info section Port Selection Port Info On STRIP layer Single Mode Single Mode strip port transmission line excitation extended calibration Internal Internal strip port lumped soutce excitation no calibration Differential Mode Differential Mode strip port transmission line excitation extended calibration Common Mode Common Mode strip port transmission line excitation 331 Simulation Applying 332 extended calibration Ground Reference Ground Reference for associated Single Mode port with same reference plane Internal port On SLOT layer Single Mode Single Mode slot port transmission line excitation extended calibration Coplanar Mode Common Mode strip port transmission line excitation extended calibration and Drawing Vias Momentum GX creates a via by extruding the object that is mapped as a via through the substrate layer it is applied to Vias are drawn as lines or closed polygons A line segment via is the simplest and most practical way of drawing a via Vias drawn as lines are often called sheet vias because when the line is extruded through the substrate it is treated as a horizontal metal sheet For vias of other shapes you draw a closed polygon For example for a cylinder via you would draw a circle When the shape is mapped to a via metalization layer
229. ast Newton method Sometimes the Jacobian will be a better direction sometimes it will be worse Try both approaches 3 If the convergence issue occurs during a parameter sweep sweep more points so that that each simulation is closer to the previous one often requiring less total time Ot if this is not practical or desired turn off Use Previous Solution As Starting Point This will cause the simulator to start fresh with each new parameter value 4 Increase the value of Absolute Tolerance and Relative Tolerance This should speed up the solution but will be less accurate particularly for low signal levels O ptimizing Simulation Performance A variety of methods and parameters are available to control the approach that HARBEC uses to find convergence The speed of performance can be improved by adapting these parameters to the specific circuit being analyzed To understand how these parameters work it is useful to understand a little about how the simulator searches To find a solution the simulator uses a Newton Raphson search to find the solution It starts with an initial guess and calculates an error function The derivative of the error function is used to extrapolate the next point In harmonic balance partial derivatives exist for every node and every frequency The full matrix of partial derivatives is known as a Jacobian Jacobian Calculation HARBEC Harmonic Balance Analysis The full Jacobian is usually the most ac
230. ate the path that the spectral component took to arrive at the node under investigation Example This example shows a spectrum at 4000 MHz whose power level is about 67 dBm It has a coherency ID of 15 and is a 3rd order intermod between Source 2 and Source 3 The intermod was created in RFAmpt1 and then followed the path through TL1 Attn2 RFAmp2 The output of RFAmp2 is where the spectrum is being viewed Pr 4000MHz 66 925d0Bm 15 H 50urce 3 H 2x Saurces 2 RFAmpt TL1 AEEnz RF Amp NOTE Spectrum identification information can only be displayed if Show Individual Spectrums has been enabled 435 Simulation 436 Broadband Noise The SPARCA simulation technique enables Spectrasys to simulate broadband noise very quickly The Ignore Frequency limits ate used to specify the frequency of the broadband noise along with the frequency range of the simulator The entire broadband noise spectrum is simulated with a small number of simulation points To ensure accurate noise measurements Spectrasys uses a special technique called smart noise point insertion to guarantee noise data is taken at desired spectrum frequencies This allows to simulation to run much faster and reduce the number of noise data needed to make accurate noise measurements Broadband noise flows in all directions through a node For example if the output port was being examined then on a spectral plot the user would see the noise power flowing
231. atic switching between Full Jacobian Diagonal Jacobian methods Never never use Full Jacobian Method Less robust but may be very efficient for too large circuit or too complicated solution spectrums when it converged Always always use Full Jacobian Method May be more robust but may be very time expensive for a very large circuit or complicated solution spectrums Use Chords Method if checked then Diagonal Jacobian method will reuse inverted jacobian for few subsequent iterations without updating it Parameter Slope of the method defines chord slope factor default 1 Relative Error Jacobian recalculation force update Jacobian if the relative norm of residual at the Newton iteration improved less then the value Reuse Jacobian At Most The largest number of times that a Jacobian matrix will be used before updated Calculate Jacobian Numerically if checked nonlinear part of Full Jacobian will be calculated using numerical algorithm default unchecked Absolute Numeric Derivation Step absolute step for numeric derivation used for 0 valued solution components only Relative Numeric Derivation Step relative step for numeric derivation used for all nonzero solution components Residual Norm Type formulation of residual norm function L 2 quadratic norm L Inf infinite norm Weighted Norm if it checked on than components of residual vector will be scaled by factor 1 Total_Harm_Order
232. ation of the expression determines how many times the statement is executed The while looping executes a statement until an expression becomes False If the expression is False when the loop is entered the statement is not executed at all The syntax for the repeat and while statements is shown tepeat expression statement while exptession statement For statement The for statement controls execution of its associated statement s using an index variable If the associated statement is an analog statement then the control mechanism must consist of genvar assignments and genvar_expressions operators no use of procedural assignments and expressions for procedural assignment expression procedural assignment statement where fot analog fot statement the format is for genvar assignment genvar_expression genvar assignment analog statement Signals Accessing net and branch signals Signals on nets and branches are be accessed only by the access functions of the associated discipline The name of the net or the branch is specified as the argument to the access function Examples Vin V in CurrentThruBranch I myBranch Events Advanced Modeling Kit The analog behavior of a component can be controlled using events which have the characteristics e Events have no time duration e Events can be triggered and detected in different parts of the behavioral model e Events do not block the executio
233. ator Analysis Options Calculate Oscillator Noise Advanced Frequency Calculation Minimum Frequency Number of Points Calculate Oscillation Frequency 4899 834927 Use Oscillator Solver abst Current Tolerance Curve Traing Maximum Number of Curve Iterations Maximum Step Size 0 001 Number of Final Iterations Minimum Frequency The smallest frequency to search for the frequency of oscillation Maximum Frequency The largest frequency to search for the frequency of oscillation Number of Points the number of frequencies in the above range linearly spaced to search for the frequency of oscillation Calculate Osc Frequency Calculates small signal frequency of oscillation for linearized circuit at DC operating point 263 Simulation Harmonic Balance Calculation Options Use Oscillator Solver If checked than performs nonlinear steady state Harmonic Balance Oscillator Analysis Oscillator Port Initial Voltage initial amplitude of 1st harmonic of voltage at OSCPORT node Vprobe Default 0 Absolute Current Tolerance maximal absolute value of OSCPORT current Iprobe accepted as steady state solution Oscillator solver convergence criteria Default 1e 8 A Number of Curve Iterations maximal number of curve tracing iterations finding Vprobe when Re Iprobe Vprobe lt 0 and Re Iprobe Vprobe Vstep gt 0 Default 200 Max Step of Curve maximal vo
234. ber of frequencies in the circuit you can expect the solution to take roughly 8 times longer However this is only a rough estimate The convergence process is complex and difficult to predict At a fundamental level harmonic balance solves a simultaneous set of nonlinear differential equations No mathematical approach is guaranteed to find a solution to the problem Years of work have gone into HARBEC to develop the most robust strategies available To add a harmonic balance simulation 1 Inthe Workspace Window click the New Item button and select Add Harmonic Balance Analysis from the Analysis submenu 2 Complete the HARBEC Options dialog box For details see the Reference manual HARBEC Options To edit a Harmonic Balance properties double click the Harmonic Balance Analysis or click the analysis and click the Properties button on the Workspace Window 225 Simulation General Tab Harmonic Balance Analysis Options General Calculate Moise Advanced Mame HE1 Calculate Now Design Schi Dataset HE1 Data Aw Factory Defaults ae Save as Favorite Temperature 16 85 oC Description DC Analvsis default Analysis Frequencies Frequency Units MHz Freg MHz Maximum Mixing Order al Maximum Analysis Frequency S set All Fregs as Harmonics of fies Frequency Accuracy Design Defines the schematic or EMPOWER electromagnetic simulation that will be analyzed
235. ble 1 Spectrasys Intercept Setup Measurements Channel Frequency Intermod Frequency TIMCP Total Intermod Intermod Channel Power for Channel Power each Order up to the Maximum Order TIMCPn Total Intermod Only Intermod Channel Power Channel Power for for the given Order n Order n Interferer Tone Tone Frequency used for Channel Frequency Intercept Calculations 461 Simulation ICP Interferer Tone Power Level of the Tone Used Channel Power for In Band Measurements Virtual Tone Channel Power Level of the Virtual Tone Power based on the Desired Channel Power of the Test Signal Used for Out of Band Measurements Table 2 Spectrasys Intercept Results Measurements Measurement Intermod Results Input Input Intercept Point for each Intercept Order up to the Maximum Point Order Input Only Input Intercept Point Intercept for the given Order n Point for Order n Output Output Intercept Point for Intercept each Order up to the Point Maximum Order Output Only Output Intercept Point Intercept for the given Order n Point for Order n Input Input Intercept Point for each Intercept Order up to the Maximum Point Order Input Only Input Intercept Point Intercept for the given Order n Point for Order n Output Output Intercept Point for Intercept each Order up to the Point Maximum Order Output Only Output Intercept Point Intercept for the given Order n Point for Order n 462 Spectrasys Syst
236. bstrate The substrate database must be read and interpolated which requires a certain amount of memory Algorithms to make trade off between time and memoty resources are implemented in the simulator These algorithms result in additional usage above that required to solve the matrix Total memory consumption is typically less than 1 5 times what is required to store the matrix for large matrices References 1 R F Harrington Feld Computation by Moment Methods Maxillan New York 1968 393 Chapter 9 Sonnet Interface Design Flow Sonnet Flow O verview The normal steps in the Sonnet interface flow are as follows 1 Create a layout for electromagnetic analysis 2 Create a Sonnet analysis 3 Recalculate the Sonnet analysis GENESYS does the following a Translate the Layout into a Sonnet son file storing it in the workspace b Write out the following directory structure and files in the same directory that the GENESYS workspace is stored assuming that the GENESYS file is WorkSpace wsp and the Sonnet analysis is Sonnet1 i WorkSpace_Sonnet Sonnetl Directory ii WorkSpace_Sonnet Sonnetl Genesys son Sonnet geometry file ii WorkSpace_Sonnet Sonnet1 SonData Genesys Directory c Runs Sonnet The Sonnet file contains a FILEOUT directive instructing Sonnet to write out Y Parameter data to WorkSpace_Sonnet Sonnet1 Genesys yp d Reads WorkSpace_Sonnet Sonnet1 Genesys yp and stores it in the wor
237. calculation error relatively straightforward algorithms that facilitate development of general purpose programs and a lot of possibilities to speed up calculations and to increase accuracy of solutions For these reasons and others we decided to use it for the electromagnetic simulator This section summarizes the theoretical backgrounds with emphasis on the problem formulation and acceleration techniques Historical Background Most commercial electromagnetic EM simulators designed for MIC and MMIC work are based on integral equations and the method of moments MoM EMPOWER is based on the method of lines MoL This technique has excellent error convergence properties and submits well to code optimization to minimize numeric complexity 207 Simulation 208 The root of EMPOWER is work which began in 1987 at the Novosibirsk Electrical Engineering Institute This lead to the commercial development of TAMIC in 1991 in Moscow TAMIC saw commercial use in the Soviet Union and elsewhere In late 1996 Eagleware acquired TAMIC and the principle contributor joined Eagleware to begin significant improvements The code was integrated into the GENESYS environment at release Version 6 5 in 1998 Problem Formulation This section describes a general mathematical formulation of the boundary value problem to be solved It defines all restrictions in the problem domain You can use this section to decide whether your particular problem fits the formu
238. can be also classified as MoL because of its semi discrete nature Originally the network impedance analogue method Kron 1944 Sestroretzkiy 1977 and a grid spectral representation inside homogeneous layers were used to analyze the layered three dimensional structures Sestrorezkiy Kustov Shlepnev 1988 that correspond to a combination of the 3D finite difference approach and the spectral domain technique Later only the discretisation in the metal plane was left but the method still retains some advantages of the network impedance analogue method That is why we sometimes refer to the EMPOWER numerical techniques as the impedance interpreted method of lines Here are the main solution stages of the impedance interpreted MoL e Partial discretisation of the Maxwell s equations only in the plane of metalization x y plane e Grid spectral representation of the EM fields in the homogeneous layers e Building Grid Green s Function GGF matrix in spectral domain using impedance form of the solution in a layer EMPOWER Planar 3D EM Analysis e Representation of each GGF matrix element as a sum of four elements of an auxiliary array obtained using DFFT technique e Equidistant grid transformation to a non equidistant grid using thinning out and linear re expansion procedures e Automatic detection of symmetry for symmetrical and nearly symmetrical problems reflection and 180 rotational e Solution of the main system of linea
239. cated under the Optimizations folder and change the Default Simulation Data or Equations to MFilter1 EM1 Then optimize the circuit by selecting Optimize Now and Automatic MFilter1 Fa Default SimulationzD ata or Equations b itel EMI ee Measurement Op Target Weight Min Max Ls LE 9 3 20 2x aja o Lu 3 19 29 sjs LBj 9 1 m 29 n f _L Lx L1 E 1 iil uw Measurement Wizard Equation Wizard Optimize Now Cancel Help 25 The final response of the optimized filter is as shown below The last step is to press F5 on your keyboard to update the new traces If you want you can add a bandwidth marker to display the final result 205 Simulation BE MFilter1 Response Workspace EmWalkthru z EH H 1 2075 MHz 2 838 dB DELTA L 2 887 dB AMNDVVIDTH 255 MHz 3 2330 MHz 3 396 dB DELTA R 3 394 dB oe Oo uy a m T m as m Lun uy DB S21 MFilterl EM1 DB E21 Freq MHz amp DB 221 DB 211 gt MFilter1 Eh1 DB S21 MFilter1 Eh1 DB 211 And the final capacitor values ate J GENES TS File Edit View Workspace Actions Dm E X Gs n lt CAPT MFILTER CAPS MFILTER Note that the original linear response is much higher in frequency than the electromagnetic simulation EM PO W ER Theory
240. cavity is not homogeneous but instead is partially filled with a dielectric and the remainder of the cavity 1s filled with air then the dominant mode resonant frequency is reduced and may be approximated using a filling factor Johnson 1987 Assuming the substrate is mounted on the floor of the cavity the resonant frequency of a partially filled rectangular cavity frartial 18 NEC partial hort 7 hy F where t is the thickness of the substrate and h is the height of the cavity without a substrate For example fio for the 2x4 inch box is reduced from 3299MHz to 3133MHz with t 62mils and e 4 8 This expression is approximate because the electric field lines are not parallel to the z axis and a component of these lines terminate on the side walls This mode is referred to as a quasi TEM101 mode Signal M etal Effects Relatively sparse signal metal has little effect on the resonant frequency Larger metal segments particularly when grounded significantly reduce the resonant frequency To obtain a feel for the significance of signal metal you may add extraneous metal to the substrate in Example 10 Box Modes and observe the shift in the transmission peaks Top Cover Transmission line discontinuities disturb current flow and energy is lost from the transmission structure While this lost energy is typically small the Q of the resonant cavity is high and coupling at these frequencies is significant Removing the cover of the
241. ce Summary C Automatic Recalculation eme perene 0 emm T OWsouce N Source Ca Pyr 4 Click the Factory Defaults button to initialize the dialog box to a known condition m Click Yes on Factory Defaults dialog box GENESTS SPECTRASYS Do vau really want replace system analysis parameters with Factory defaults 6 If path measurements are desired i e cascaded gain or cascaded noise figure click on the Paths tab 418 Spectrasys System System Simulation Parameters General Paths Calculate Composite Spectrum Options TRERRARERERERRRRSRRRRARSRERRRRRRSRARRRRRERREARRRRRIRRERRRRSRRRRRERRARRRRRERRERRARERERRRRRE T P Ta Ix Add Al Paths From AI Sources Add Path Channel Frequen Mame O Path from Hode thru Hode to Hode t cur TREE TUER E 17 3 Delete 1 6 Delete Add Powers Voltages and Impedances to Path Dataset di Factory Defaults 7 Click the Add All Paths From All Sources button Two paths should appear as shown above NOTE Node numbers may be different than shown above depending on the node numbers in your schematic For additional information on specifying paths click here 8 Click the dialog OK button Back to Create a System Schematic Next to Run the Simulation Run the Simulation Analysis data must be created before it can be plotted or displayed
242. ch is identical to the network shown in the previous section This result is fandamentally the same as the result from MYNET below When the capacitor below is tuned or optimized the networks MYNET and EMPOWER are both updated simultaneously Even if you create a file with a layout only no schematic you can still use automatic port placement Simply put the parts down onto a blank schematic connecting them into a dummy network The parts will now show up in the layout and can be moved as needed ignoring any rubber bands The rubber bands come from the meaningless connections in the dummy network When you display the EMPOWER simulation results it will include the components You do not need to display the results from the schematic Planar X and Y Directed Ports Note EMPOWER will create planar ports for lumped elements if the Use Planar Ports for one port elements box is checked in the EMPOWER options dialog See your reference manual for details In some situations you may want to place internal ports with X or Y directed currents These ports are much trickier to use manually since they are not referenced to ground For components in your layout EMPOWER will automatically place planar port and lumped elements so this section is for background or advanced applications 155 Simulation This figure shows the configuration of these ports These ports can be more accurate for manually connecting lumped elements to EMPOWER
243. ch name gt Freq Time lt branch name gt Voltage NENTOR pos Waveform at number gt the port lt port MOSS ORDHBIIDC IE or number time VPORT port number gt Freq Time Power dissipated at 7 the pen port P lt port number DAE number gt port number gt number gt resistance Complex lt variable gt Freq lt freq lt measutement gt lt freq value amplitude of MHz the Spectral component Examples 8 Sm HB1 Sch1 rect V1 100 1 eme NR equal to freq value MHz gt value gt Examples HB1_Data V1 Freq 100 1 VPORT Freq 100 1 1 P1 Freq 100 1 lt measurement gt lt index gt Complex amplitude of lt variable gt lt index gt 1 Examples Examples the Spectral DC i DC v component HB1 Sch1 rect V1 0 eee which order i st 2 Amplitude of 1 2 Amplitude of 1 harmonic D index is equal VIG harmonic to lt index gt P1Gl VPORT 2 1 P1 2 Complex hb_getspcomp lt variable gt amplitude of FreqIndexIM lt IM index gt the Spectral hb_getspcompdbm lt variab component le gt FreqIndexIM lt IM which multi Not implemented index gt dimensional Examples intermodulati 3 harmonic of power at on index is port 2 p equal to lt IM 245 Simulation index gt hb_getspcompdbm P2 Fr eqIndexIM 3 Power of IM which frequency Fim 1 F1 1 F2 at port 2 D hb getspcompdbm P2 Fr eqIndexIM 1 1 Amplitude of 1 harmonic of frequency F1 of volt
244. cked by comparing the actual answer at every node to the predicted values If the values ate not close enough then the time point is not accepted and the time step 1s repeated with a smaller value Howevet if the values do agree within tolerance then the time step will be doubled for the next step up to the Maximum Time Step c Ifthe simulator is using approximate truncation error controlled time steps then simulated points are always accepted if converged An accuracy check is then done using the Truncation Factor and a SPICE like algorithm which controls the step for the NEXT time point The further away the solution was from the predicted value the smaller the next step will be Again the step is limited by the minimum and maximum step sizes 4 If numerical precision has caused non convegence see the explanation below then the step size is doubled and the step 1s repeated If the step size had already been decreased then simulation 1s halted with an error Note that all of these rules apply only to simulated time points not output time points The output time points the ones saved into the dataset are determined from the entries 70 CAYENNE Transient Analysis on the Output Miscellaneous tab Using the default settings data points will be output starting at time 0 and at points separated by the maximum step To change these defaults use the following entries e Output Step Size is the smallest step output to
245. ct Show All to display all the layer parameters LAYOUT Properties General Associations Layer Fonts Show Columns Show All Momentum Slot Type Strip v On Bottom Mirrored ch Use Layer Mesh Use Layer TL Edge Mesh Name Plot Factor Use Mesh Density Density TL Mesh Mesh Edge Mesh Width Via Mode Top Cover Air amp bove TOP MASK o 1 TOP SILK F F TOP METAL 15 Bi SUBSTRATE o Default BOT METAL T oF F1 F1 MEE E Specify Width BOT MASK F F Air Below F Bottom Cover Or you can reduce visible information in this table just choosing show 353 Simulation i 8 Momentum checkbox LAYOUT Properties General Associations Layer Fonts Show Columns Metal C Substrate C General iaver Number and Color ScocceAcccccee Wecccescccecccccccscsccccesscccecscoseee C Show All EM EMPOWER Momentum Momentum Slot Type Strip v Use Layer Mesh Use Layer TL Edge Mesh Strip Name Use Mesh Density Density TLMesh Mesh Edge Mesh Width Via Model Model TOP METAL 15 3 Specify Width Default SUBSTRATE Default BOT METAL O E E Select the checkbox for Use Layer Mesh Density for the metal layer and enter the Mesh Density value cells wavelength for it To use a layer TL mesh density select the checkbox Use Layer TL Mesh for the metal layer and enter the TL Mesh value cells wavelength for it To use a layer edge mesh e Specify how it s calculated
246. cted to Depending on the value it rotates the ports to be perpendicular to the closest edge of the T line If the parameter has the default value the port connector pad will be oriented to be perpendicular to the closest box wall For Momentum it does not matter what the port orientation is the simulator will 309 Simulation orient any EM calibrated port internally to be perpendicular to the closest metal edge and snaps it to the center of the edge e The components or shapes that ports are connected to must be on layout layers that are mapped to metalization layers that are defined as strips or slots Ports cannot be directly connected to via For information on how to define strips and slots refer to the Layout Layers properties page e Make sure that the port and the object you are connecting it to are on the same layout layer For convenience you can set the entry layer in the Layout layer combo box e A port must be applied to an object If a port is applied in open space Momentum will automatically snap the port to the edge of the closest object This will not be apparent from the layout because the position of the port will not change e If the Layout resolution is changed after adding ports that are snapped to edges you must delete the ports and add them again The resolution change makes it unclear which edges the ports are snapped to causing errors in mesh calculations Port Calibration In Momentum GX there are two
247. ctual power dissipated at the port loads in tested operating regime of the circuit For example to calculate LS parameters of a 2 port DUT we first create two schematics for each of signal ports Pot 1 Poit 2 PORT 1 PORT2 R 50 0 z0 O 50 Q z F 1000 MHz f CATH i PAC 10dBm Pint Port 2 Port 1 PORT 2 Dru Fic R 50 0 20 sai F 1000 MHz f E PAC 0 dBm Pin2 In order to calculate the LS parameters we need to create 2 HB analyses HB1 and HB2 one per each design To calculate the LS parameters from the input power we create 2 Parameter Sweeps SweepPowerl and SweepPower2 for each HB analysis NEW In Genesys 2005 to calculate the LS parameters we need to define new variables LS11 LS21 in HB1 and L12 LS22 in HB2 analysis datasets which call the Harbec analysis function hb LargeS 33 Simulation hb LargeS Vin Vout samepott where Vin voltage amplitude of port source and Vout amplitude of 1st harmonic of voltage at node where the port is connected to the circuit The spectrum of the port voltages are saved in VPORT HB analysis dataset variable To get the n th harmonic of the k th port voltage spectrum we can use direct indices VPORT n 1 k The 1 offset is used because 1 the first element is DC So the Ist harmonic of voltage at the 1st port is calculated as Voutl VPORT 2 1 And for the 2nd pott it is Vout2 VPORT 2 2 HB1 Data
248. cuit which has a highpass response Next we will examine the response of this circuit to a higher frequency pulse 1 Change the parameters of the input pulse model on the schematic Set the pulse width to 5 ns and the Frequency to 100 MHz Be sure Use Default is not checked for the Frequency variable 2 Double click the Transient analysis on the workspace tree 3 Change the Stop Time to 50 ns and the Maximum step size to 1 ns Note that we are only using 10 points per cycle 4 Click Calculate Now The VPORT graph has changed to the following 67 Simulation 68 Transient VPORT Pot les des eL eA LL LU Ld LA T CALL KL B LN LN d A lU A ATATA A AAAH NEA WY AAAA AAA mm d bh mm bh gt xD Li oO a7 LL ka A Time n w PORTE PORTE Note the ragged edges on the square wave On first glance it appears that the simulation might not be very accurate However these problems are actually caused by the data reduction in the transient simulator We can actually look at the exact points being simulated instead of simply interpolated results To do this 1 Double click the Transient1 analysis on the workspace tree 2 Click on the Output Miscellaneous Tab Note that Output Step Size is set to Use Maximum Step Size This means that no matter how many points are simulated we will get evenly spaced points every nanosecond While this is extremely convenient fo
249. culate Tab of the System Analysis Dialog Box The column number is the same as the order starting from the left with order 0 See the Intermods Along a Path section for information on how to configure these tests Remember intermod bandwidth is a function of the governing intermod equation For example if the intermod equation is 2F1 F2 then the intermod bandwidth would be 2BW1 BW2 Note Bandwidths never subtract and will always add The channel bandwidth must be set wide enough to include the entire bandwidth of the intermod to achieve the expected results The Automatic Intermod Mode will set the bandwidth appropriately Caution This method used to determine the intercept point is only valid for 2 tones with equal amplitude Channel Used Main Channel Frequency Interferer Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as RX_OIP and CGAIN Travel Direction Same as RX_OIP and CGAIN Interferer Cascaded Gain ICGAIN This measurement is the interferer cascaded gain of the main channel along the specified path The Interferer Cascaded Gain is the difference between the Interferer Channel Power measurement at the nth stage minus the Interferer Channel Powet measurement at the input as shown by ICGAIN n ICP n ICP 0 dB where n stage number Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as ICP Travel Direction Same as
250. curate way to determine the next point However the matrix can be very large requiring a lot of time to calculate and invert To make the simulator faster HARBEC generally tries Fast Newton steps first A Fast Newton step calculates only a portion of the Jacobian and uses it to calculate the next point For many circuits the entire solution can be found quickly using only Fast Newton steps The default setting for HARBEC is to automatically switch between using Fast Newton and full Jacobian steps Artificial intelligence techniques are used to determine which technique to use and when Usually the automatic switching will find the solution quickly However for certain circuits it will be better to always use the Jacobian or never use the Jacobian On the HARBEC Options dialog box you can specify either Automatic Always or Never use of the Full Jacobian Experimenting with different values may improve convergence speed Order vs Accuracy and Time The easiest way to affect stimulation performance is to change the order of the frequencies used in simulation Harmonic balance models signals in the circuit by using a finite number of harmonics of the fundamental signals and a finite number of mixing terms The larger the number of harmonics and mixing terms the better the approximation of the actual signals However the larger the number of frequencies the longer the simulator takes to work The length of time to take a search step is
251. d NDCP Travel Direction Same as DCP and NDCP Cascaded Compression CCOMP This measurement is the cascaded compression of each stage along the path CCOMP n Summation COMP 0 to n dB where n stage number For each stage n a summation is performed on the compression point of all previous stages Channel Used Same as COMP Types of Spectrums Used Same as COMP Travel Direction Same as COMP Cascaded Gain CGAIN This measurement is the cascaded gain of the main channel along the specified path The Cascaded Gain is the difference between the Desired Channel Power measurement at the nth stage minus the Desired Channel Power measurement at the input as shown by CGAINI n DCP n DCP 0 dB where n stage number Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as DCP Travel Direction Same as DCP NOTE Under matched conditions CGAIN and 21 from a linear analysis are the same As shown in the above equation the cascaded gain at the first node by definition is 0 dB This may not be true if there is an impedance mismatch between the source and the first model in the path The cascaded gain measurement does not take into account this initial mismatch because cascaded gain is always assumed to be 0 dB at the first stage This mismatch can be accounted for by taking the difference between the power level specified in the source with the Channel Power CP at the firs
252. d for these cases Checking the box to turn these layers on is the equivalent of adding a substrate layer with Er 1 Ur 1 and Height in units specified in the Dimensions tab as specified Caution When setting up a new circuit be sure to check the height of the air above as it is often the only parameter on this tab which must be changed and is therefore easily forgotten Metal Layers In LAYOUT multiple METAL layers e g copper and resistive film are automatically converted to one EMPOWER signal layer if no media layer is in between the metal layers All metal layers from the General Layer Tab are also shown in the EMPOWER Layer tab These layers are used for metal and other conductive material such as resistive film The following types are available e Lossless The layer is ideal metal e Physical Desc The layer is lossy These losses are described by Rho resistivity relative to copper Thickness and Surface Roughness e Electrical Desc The layer is lossy and is described by an impedance or file This type is commonly used for resistive films and superconductors If the entry in this box is a number it specifies the impedance of the material in ohms per square If the entry in this box 1s a filename it specifies the name of a one port data file which contains impedance data versus frequency This data file will be 95 Simulation 96 interpolated extrapolated as necessary See the Reference manual for a descriptio
253. d in a system with source and loads equal to the reference impedance S21 is the network transducer power gain in decibels Because S11 and S22 of a network are not in general zero a portion of the available source power is reflected from the network input and is dissipated in the source The insertion of a lossless matching network at the input and or output of the network could increase the gain of the overall system if reflections toward the source were reduced Shown below is a two port network with lossless matching networks inserted between the network and the source and load Linear Analysis GMAX and MSG When the input and output networks are simultaneously designed for maximum gain there is no reflection at the source or load The maximum transducer power gain Gmax 1s given by Gmax Sai Sz K sqzt K2 1 The maximum stable gain MSG is defined as Gmax with K 1 Therefore MSG Sai S12 A GENESYS plot of GMAX shows Gmax when K gt 1 and MSG when K 1 Again achieving this maximum gain requires that the input network is designed such that Rs is the complex conjugate of S11 and Ry is the complex conjugate of S22 GENESYS returns the required reflection coefficients impedance and admittance for the input and output networks as GM1 GM2 ZM1 ZM2 YM1 and YM2 respectively The Unilateral Case Historically to simplify the complex equation for Gt in the previous section on matching Si2 was set to zero At
254. d in the simulator Inputs with sources such as INP_PAC input port with power source and INP_VDC input port with DC voltage are in full effect when placed on the top level 2 In lower level designs single terminal ports are only used to provide connection information for subcircuits Impedances and sources such as INP_PAC and INP_VAC are ignored To place sources in subcircuits you must use sources without ports such as PAC and VAC Note Balanced two terminal ports are retained from all levels in the design essentially becoming top level ports With this configuration you can test a subcircuit in linear and nonlinear modes by putting the sources for test within the ports When the subcircuit is then reused inside the larger circuits those circuits and ports will be disabled If you actually want to create a true port within a subcircuit you can place a balanced port inside the subnetwork If you only want a single ended port simply ground the negative terminal of the differential port New in GENESYS 2005 You can use equation variables to specify port numbers plus you can create user models which contain ports For an examples load the system intermod source into a workspace Use Add From library in the create new item button in the workspace tree Select the system intermod soutce then look at the design and equations New in GENESYS 2005 Balanced differential ports are supported in all simulators To place a diffe
255. d obtain an EMPOWER simulation GENESYS is then used to display and compare the linear simulation with the EMPOWER data 107 Simulation 108 Examples The examples are completely contained in the EXAMPLES manual Examples which illustrate EMPOWER include e Microstrip Line WSP e Sttipline Standard WSP e Spiral Inductor 2 WSP e Box Modes WSP e Film Atten WSP e Edge Coupler WSP e Dual Mode WSP e 8 Way WSP e Edge Coupled WSP e Coupled Stepped Z WSP e Tuned Bandpass WSP e Patch Antenna Impedance WSP The required RAM specified in the Examples manual is the value estimated by EMPOWERR They are approximate and are determined by algorithm rather than a test of memory used The execution times are for a 266 MHz Pentium II with 256Mbytes of RAM operating under Windows 98 In most cases execution time is for the discontinuity mode Creating a Layout Without a Schematic The completed file for this walkthrough is the GENESYS Examples folder EMPOWER Layout Only wsx This example demonstrates the following topics e Creating a layout without a schematic e Choosing grid spacings e Choosing the box size A microstrip stub notch filter with a transmission zero at 9 5 GHz is to be simulated The filter has the following specifications e 15 mil RT Duroid substrate er 2 2 tan d 0 0009 e Copper metalization e 50 Ohm terminations EMPOWER Planar 3D EM Analysis e The stub line should be 70 W and 90 at 9 5 GHz
256. d path as shown by IMGNR n CNP n IMGNP n dB where n stage number This measurement is very useful in determining the amount of image noise rejection that the selected path provides For this particular measurement basically two channels exist both with the same Channel Measurement Bandwidth 1 main channel and 2 1st mixer image channel Channel Used Main Channel Frequency Image Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as CNP and IMGNP Travel Direction Same as CNP and IMGNP Image Rejection Ratio IMGR This measurement is the ratio of the Channel Power to Image Channel Power along the specified path as shown by IMGR n DCP n IMGP n dB where n stage number For this particular measurement basically two channels exist both with the same Channel Measurement Bandwidth 1 main channel and 2 1st mixer image channel The only difference is between these two channels ate their frequencies one is at the Channel Frequency and the other is at the Mixer Image Frequency Channel Used Main Channel Frequency Image Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as DCP and IMGP Travel Direction Same as DCP and IMGP Spectrasys System Minimum Detectable Signal MDS This measurement is the minimum detectable discernable signal referred to the input and is equivalent to the input channel noise power plus the cascaded
257. d simulators Generally you should do the following e Set the Most Accurate Frequency to the frequency of greatest interest in your circuit such as an expected oscillation frequency Remember the modeling error will be limited to 1 by default at other frequencies e Turn on Always use Constant Loss Not Frequency Dependent Losses This setting will use approximate models for INDQ and CAPQ which are otherwise vety problematic elements for a convolution simulator since they are non causal For these elements the loss resistance or admittance as calculated at the Most Accurate Frequency will be used for the time domain simulation For example if you ate designing a filter you should set the Most Accurate Frequency to the center frequency for bandpass or cutoff frequency for lowpass or highpass The response will be almost unchanged due to the approximation e Setthe Maximum Frequency for convolution carefully It must be low enough to capture the delay of your longest transmission line and must be high enough to avoid time step errors e 1 Delay lt MaxFreq lt MinNumOfPoints Delay e Ifthe response has unexpected oscillations or noise set MaxF req gt 0 1 MaxTimeStep e For example if you have a transmission line with a 0 5 ns delay plus you need 10 ps resolution time step you should set max frequency within 1 0 5ns lt MaxFreq lt 128 0 1ns between 2 GHz and 256 GHz and greater than 0 1 10ps greater than 10
258. d to other multimode ports including multi mode lines and multimode EMPOWER data Further any multi mode elements connected together must have the same number of modes for each pott Caution Do not connect standard lumped elements to a multimode port The results will not be correct If you will be connecting directly to components you should use single mode ports Use multi mode ports only for connection only with other multi mode ports and multi mode lines e They can be used with decomposition to accurately analyze much larger structures than would be possible in a single EMPOWER circuit See the Decomposition section for more details To create a multi mode port click on the Mode Setup Button from the EMPOWER setup dialog box when you start an EMPOWER run You will see a box similar to the one at the end of this section To make ports multi mode check the boxes between them EMpPorts 1 2 and 3 form one multi mode port and EMPorts 5 and 6 form another multi mode port EMPort 4 is a single mode port To make a multi mode port you must follow these rules 140 EMPOWER Planar 3D EM Analysis e All EMPorts for a multi mode port must be on the same wall e All EMPorts must have the same length line direction current direction and reference plane shift The EMPorts may and often do have different widths as above e All EMPorts must be Normal not No Deembed or Internal e Port numbers must be seq
259. d to represent the data the larger the data file will be and most likely the simulation time will increase Analyzer Troubleshooting What does it mean when the signal doesn t seem to be ined up with the integrated spectrum All this means is the frequency resolution isn t small enough to accurately represent the signal of interest If this is the case there are a few things that can be done to increase this resolution First the resolution bandwidth can be reduced If this is inadequate the Limit Frequencies feature should be enabled and the user can specify the Start Stop and Step frequencies used for the analyzer System Simulation Parameters O ptions Tab This page contains miscellaneous Spectrasys options Tabs General Paths Calculate Composite Spectrum Options Output Spectrasys System System Simulation Parameters General Paths Calculate Composite Spectrum Options Output Ignore Spectrum Maximum Number of Spectrums To Generate Level Below dem Max Spectrums EE Frequency Below MHz Mixer LO C Strongest Signal Only Frequency Above All Signals within dBc of Strongest Frequency Above and Below are optional Range Warning For Mixer Multiplier etc The default Frequency Below is 0 Frequency Above defaults to Tolerance Range dB Max Order 1 Max Source Frequency Parameter Information IGNORE SPECTRUM This group is used to limit or restrict the numb
260. d with non metal dielectric areas The combo box for Momentum Slot type only works for Momentum GX and is ignored by other Genesys EM simulators Empower Sonnet Other EM simulators use the Physical slot checkbox setting for internal modeling of the slot layer without swapping the layer metal areas in Layout To define coplanar ports 1 Place a new EM port in layout or select any port that you want to assign this type to Note the port number 2 Open the port dialog double click or select it and click Enter 3 For the Port Type select Coplanar 4 Under Polarity make sure that Normal is selected 5 Click OK 6 Place another new EM port in layout or select the second port Note the port number 7 For the Port Type select Coplanar 8 Under Polarity select Reversed 9 Under Associate with port number enter the number of the previously selected port 323 Simulation 10 Click OK For example this layout has defined 2 pairs of coplanar ports 1 2 and 3 4 You do not need to set the Associate with port number value for both ports of a coplanar pair In this example only port 2 of the coplanar ports pair 1 2 has value of the port association 1 The coplanar port consists of pair of coplanar mode ports with different polarity The polarity must match to the polarity of signal sources connected to the ports otherwise corresponding to the incorrectly defined polarity coplanar port S parameters get n
261. daries are not aligned correctly the simulation data may be less accurate For mote information refer to Processing Object Overlap Meshing Thin Lines When the geometry has narrow lines like thin transmission paths in a spiral it may be difficult to have a mesh that is more than one cell across the width if the default global mesh is used If needed use Edge Mesh or Transmission Line Mesh to capture the current distribution across the line Meshing Slots Slots should be meshed exactly the same as strips there is no difference For example the edge mesh can be used for slots because the current distribution is basically the same that is it is concentrated on the edges of a slot Adjusting the Mesh Density of Curved Objects Vias or other curved objects when drawn in a layout have a default value for the number of facets used to draw the object If the value is small a relatively large number of facets are used to draw the circle and this can result in more triangular cells created the curved areas during the mesh process Be aware that changing the facets of an object does alter the geometry of the circuit For information on how the mesh parameter Arc Resolution affects a mesh refer to Processing Object Overlap Discontinuity Modeling A microstrip transmission line with a bend using the default mesh may have a cell size equal to the width of the line one cell per line width If it is long and the bend is not sever
262. de if this box is checked A constant resistance admittance as computed at the Most Accurate Frequency is used instead Most GENESYS built in models with loss such as microstrip viaholes and inductors capacitors with Q use the RLOSS models interanally and take advantage of this feature 81 Simulation 82 Transient Analysis Properties General Integration Time Step Convolution Output Miscellaneous QutEpuE Output Step Size Use v Maximum Step Size Force Gukput at Exact Step Output All Simulated Points Save Pork Voltages Save Node Voltages and Currents Levels of Data to Retain in Hierarchy ee Iterations Max Tries per Time Point 100 Max Newton Steps per Try 1000 Min Relative Progress Max Tries per Newton Step Miscellaneous Maximum Simulation Points Use w Infinity In Minimum Conductance trarin ed2 Output Output Step Size is the smallest step output to the dataset If Output Start Time is not zeto then no data is saved until that time point Simulation always starts at time zero Force Output at Exact Step will use the predictor generally third order gear to interpolate the simulated points to ensure that the output waveform is sampled uniformly in time needed for performing FFT analysis on the output waveform Output all simulated points outputs every point that the simulator calculated which can often output ten times more data See the Simu
263. dent wave is a harmonic function of time Its magnitude is unity and it corresponds both to one Watt instantaneous power and 1 2 Watt time averaged power The initial phase of the incident wave is zero Other eigenmodes of the structure are terminated by their characteristic impedances and are perfectly matched It numerically represents a row of the generalized scattering matrix The internal ports are often locations where lumped elements will be included by GENESYS Parameters of the lumped elements are not required for the EMPOWER simulation Thus internal ports default to 1 ohm normalization In this case the viewer data may not be as useful since the lumped elements are not taken into account by the viewer It is also possible to use an internal port as a source of energy to excite a structure The termination impedance can be specified using the option NI lt n gt In this case the internal inputs are terminated by virtual transmission lines with the specified characteristic impedance The unit incident wave is excited at the specified input Note that if option NI lt n gt is used then the external inputs are also terminated by transmission lines or loads with this impedance after de embedding and transformation into the mode space if necessary If the excitation conditions are defined EMPOWER calculates the scattering matrix S with default or defined normalization first Then it creates an excitation vector A 0 1 0 that co
264. dent power wave at the network input bi 2 reflected power wave at the network input a2 incident power wave at the network output b2 reflected power wave at the network output These new variables and the network S parameters are related by the expressions bi a1Si1 a2812 b2 a1Sa1 a2S22 11 bi 21 a2 0 12 bi az a1 0 S21 be ai a2 0 S22 bz a a1 0 Terminating the network with a load equal to the reference impedance forces a2 0 Under these conditions Siu bi ai S21 b2 a1 511 is then the network input reflection coefficient and S21 is the gain or loss of the network Terminating the network at the input with a load equal to the reference impedance and driving the network from the output port forces a 0 Under these conditions S22 b2 az S12 bi az S22 is then the network output reflection coefficient and S12 is the reverse gain or loss of the network Linear S parameters are unitless Since they are based on voltage waves they are converted to decibel format by multiplying the log of the linear ratio by 20 It is not always obvious whether an author is referring to linear or decibel parameters To avoid this confusion the 293 Simulation 294 book Oscillator Design and Computer Simulation and Versions 5 4 and earlier of SUPERSTAR use C for linear S parameters and S for the decibel form This is somewhat unconventional Version 6 0 and later of GENESYS also supports the
265. depends on how well the field generated by the current source matches with the field distribution of the parallel plate mode A distinction can be made between the fundamental mode and the higher order modes The fundamental mode which has no cut off frequency propagates at any frequency The higher order modes do have a cut off frequency Below this frequency they decay exponentially and only influence the local reactive field around the source Above this frequency they propagate as well and may take real energy away from the soutce Momentum takes the effect of the fundamental mode into account The higher order modes are taken into account as long as they are well below cut off If this is not the case a warning is issued saying that higher order parallel plate modes were detected close to their cut off frequency Simulation results will start to degrade from then on The Effect of Parallel Plate Modes Let s concentrate on the fundamental mode since this one is always present The fundamental mode has its electric field predominantly aligned along the Z axis from the top plate to the bottom plate This means that this mode creates a difference in the potential between both plates This mode can be excited by any current in your circuit electric currents on a strip or via magnetic currents on a slot It behaves as a cylindrical wave that propagates to infinity where it feels a short circuit since both plates are connected there Howev
266. determine which types of signals are included or ignored in this measurement Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as DCV Travel Direction Same as DCV Carrier to Noise Voltage Ratio CNRV This measurement is the ratio of the Desired Channel Voltage to Channel Noise Voltage along the specified path as shown by Spectrasys System CNRV n DCVIn CNV n dB where n stage number Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as DCV and CNV Travel Direction Same as DCV and CNV Cascaded Voltage Gain CGAINV This measurement is the cascaded voltage gain of the main channel along the specified path The Cascaded Voltage Gain is the difference between the Desired Channel Voltage measurement at the nth stage minus the Desired Channel Voltage measurement at the input as shown by CGAINV n DCV n DCV 0 dB20 where n stage number Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as DCV Travel Direction Same as DCV Channel Voltage CV This measurement is the average voltage across the main channel along the specified path This measurement includes ALL SIGNALS INTERMODS HARMONICS NOISE and PHASE NOISE traveling in ALL directions through the node that fall within the main channel For example if the Channel Measurement Bandw
267. domize Noise feature increase the Resolution Bandwidth and or limit the frequency tange over which a spectrum analyzer trace will be created See the Spectrum Analyzer Display section for additional information Paths The more paths contained in the simulation the longer it will take to simulate and the more data that will be collected Delete all unnecessary paths Noise Spectrasys System The more noise points that are simulated generally the longer it takes the simulation to run See Broadband Noise and Cascaded Noise Analysis sections for additional information on controlling noise Mixers The mote LO Signals used or the higher the Maximum Order to create new mixed spectrum the longer simulation time and more data that will be collected The number of LOs used in the simulation can be reduced to increase the simulation speed The intermod Maximum Order can be decreased to reduce the number of spectrums created Reducing the File Size Most of the size of a file is due to simulated data The file size can be reduce in one of two ways 1 completely removing the datasets and 2 only keeping the node data of interest Completely Removing Datasets 1 Closing all graphs and tables this will keep these items from complaining when they have no data to show 2 Deleting all system analysis and path datasets Be sure to include all those associated with sweeps NOTE Be careful not to delete static data like S p
268. dth was specified to 100 kHz and the Channel Frequency was 2000 MHz then the CNP is the integrated noise power from 1999 95 to 2000 05 MHz See comments in the Cascaded Noise Figure measurement or Broadband Noise section for additional insights Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used ONLY NOISE Travel Direction Only spectrums traveling in the FORWARD path direction Channel Bandwidth Caution When the Channel Frequency is less than 1 2 the Channel Bandwidth the lowest integration frequency used for measurements will be 0 Hz This will result is Channel Noise Power measurements being different than when the full bandwidth is used Channel Power CP This measurement is the total integrated power in the main channel along the specified path This measurement includes ALL SIGNALS INTERMODS HARMONICS NOISE and PHASE NOISE traveling in ALL directions through the node that fall within the main channel For example if the Channel Measurement Bandwidth was specified to 03 MHz and the Channel Frequency was 220 MHz then the CP is the integrated power from 219 985 to 220 015 MHz Spectrasys System Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used 4X SIGNALS INTERMODS HARMONICS NOISE and PHASE NOISE Travel Direction All directions through the node Desired Channel Power DCP This measurement is
269. e Ref Plane Shift Port Number Location Layer I TOP METAL width Line Direction Default Height ion Current Direction DeFaulE Port Type Differential Associate with port number 1 Polarity C2 Normal Inverse Differential made STRIP part transmission line extended calibration Defining a Coplanar Port This type of port is used specifically for coplanar waveguide CPW circuits It is similar to a differential port but coplanar ports are applied to objects on slot layers that is where slots are used in the design Coplanar ports should be used in situations where an electric field is likely to build up between two ports This can occur when e The two ports are close together e Polarity between the ports develops e The ports are connected to objects that are on slot metallization layers The electric field that builds up between the two ports will have an effect on the circuit that should be taken into account during a simulation To do this use coplanar ports 321 Simulation Coplanar ports have the following properties e They can be applied to objects on slot layers only e They are assigned in pairs e ach of the two ports is excited with the same absolute potential but with the opposite polarity The voltages are opposite 180 degrees out of phase The currents are equal but opposite in direction when the ports are on two symmetrical lines and the current direction
270. e Wiles airi pov edt iie nderit RERO eS p esi 183 ROP Hne Iaa E MT 183 RX Ere quenty ys Impedanee 3882 sedet iai E E E 183 OO S PA E PIES ou aaa a E e taa ede d d b Ma 184 PIP Lex CEO POlO Gy BUCS rd A 184 ANOA dedii 184 ba pd iciciatdr d ED E A E EE ET N E E oases E 184 555 CRGO COC CDaekuD E OE marnan A A A 185 EMPOWER Advanced M FILTER Example rte o ned tdeds 185 EMPOWER Advanced Example Filter Synt 88165 eee t Er p te t orien 185 EMPOWER TACO nonada na r O epe e eet tine O OEO 206 OVEEVIOU ains aio ed anii A E tnit T A E T E TE 206 Historical Baer COUN messere ia ANa EA Nodes Aion ooa ae 207 Problem FOr Oi E N er dud 208 Merodon aae TO S E E 210 Map pine on tie Gaderne T A E E O ZH Cod Geens PUDCHOD ao eon coit e So M toD toD eara DI coa NES 214 Infosmational NIoIBDOEE E Vastu tuni tum ip additum tQ DU Nd tcc 215 Chapter 6 Chapter 7 Table O f Contents Nun nicil Acceleration P EO UE S ode ebe d RIP Iu PROP ada bI ER RU Ue MR quu i cael 217 De Embed ine oori iiias a OO 218 EMPONCERB RGICIED eS ouest Nt a been peeps NN acted Cape 219 General Daekorotmt dade atest edu tn edt bi a Ge etes ed da ounce E Du Dt m ME 219 The dMethod ico an fneS eiua qat enda tute ern ME n en Meum tatu a 220 Richardson s ESA poAo Nean a usine bei ine estu uelut 221 SVMIMCEYy Processio manai iasan A A 221 EMPOWER Engine Theory atid Algorit BEDS icc viesinasiierisiacrisirestaannaneanteenisiabnoanieueus 221 Testbsamples and COMPAN ONS da riade
271. e then the default mesh may be adequate because the discontinuity is proportionally small compared to the line length However if the reference planes are moved inward or if the bend is more severe the discontinuity and resulting parasitics are in greater proportion to the rest of the line In that case the default mesh may result in simulation data with an error To correct this inaccuracy the mesh should be increased and edge mesh used When the area near the discontinuity is meshed so that the cell size is equal to a third of the line width three cells per line width the resulting error is reduced The denser mesh allows for current crowding parasitic series inductance at the interior corner of the bend and charge build up parasitic shunt capacitance at the outer edge of the bend Solving Tightly Coupled Lines An example of a tightly coupled line would be a microstrip coupled line filter on a 25 mil Alumina substrate Er 9 8 where the line width is 25 mils and the separation between the lines is 2 5 mils The filter response is sensitive to coupling between the lines and the 365 Simulation 366 default mesh is 1 cells per line width In this case the reflection coefficient data does not match measurement data To solve this design properly edge mesh must be used Mesh Precision and Gap Resolution Mesh resolution is directly related to the Layout Precision dbu data base units value All gaps or other unresolved layout
272. e Add Edit Path dialog box Delete All Paths Deletes all paths in the system analysis PATH SUMMARY TABLE This table summarizes the aspects of the path Name User defined name of the path Description User defined description of the path Enable When checked will create and add a path dataset to the workspace tree When unchecked will remove the existing path dataset from the workspace tree 539 Simulation 540 Add Edit Button When clicked will invoke the Add Edit Path dialog box Delete Button Will delete the path from the system analysis Add Edit Path Dialog Box This dialog box is used to specify the characteristics of a path A basic path consists of a beginning and ending location in a schematic and a channel frequency Path measurements can be grouped into the following groups fundamental intermod adjacent channel receiver image channel offset channel and primitive path components of powers voltages and impedances Through this dialog box the user will determine the locations in the schematic that the path should pass through along with the measurements to be calculated and saved to a dataset Spectrasys System Edit Path Description Define Path Define Using Force path through Switch state PartNames Allow path to begin on internal node Node Names Add Part Node To Path X Clear Path Channel Frequency DEN MHz Defaults to single frequency at beginning of path Qut
273. e above have a bigger effect on accuracy Symmetry Making a problem exactly symmetrical is an easy way to make a problem require less memory and time without sacrificing any accuracy There are four types of symmetry recognized by EMPOWER YZ mirror symmetry XZ mirror symmetry two mirror symmetry and 180 rotational symmetry These types are illustrated below 129 Simulation 130 YZ Mirror X Mirror Two Mirror Rotational When EMPOWER 1s running you should look at the information area at the top of the screen to see if symmetry is active If it is not recheck your problem to see if it is exactly centered on the box and to see if it is in fact symmetrical Two tools can help with this 1 Using Center Selected on Page from the Edit menu in LAYOUT This command makes it easy to make sure that your circuit is exactly centered on the page 2 Showing the listing file by selecting Show Listing File from the EMPOWER right click menu This file shows exactly how the problem was put on the grid and lack of symmetry is often obvious Making an unsymmetrical problem symmetrical will make it run 4 times faster in most cases and will make it 16 times faster if your problem can use two mirror symmetry See EMPOWER Basics section for more information on cells and the problem geometry See the Files section for more information on the listing file Thinning O ut For most examples the default thinning out should b
274. e after any initial comments is a format specifier in the form units type format R impedance where units is either Hz kHz MHz or GHz type is the type of the data file either S Y G H or Z formatis DB for dB angle data MA for linear magnitude angle data or RI for Getting Started real imaginary data impedance is the reference impedance in ohms commonly 50 or 75 One of the most common format specifiers is MHZ S MA R 50 This indicates that the data is in S parameter form normalized to 50 ohms The data is given in linear polar format magnitude amp angle The frequencies are in megahertz The data follows after the format specifier A typical line for this two port file 1s 500 64 23 12 5 98 03 70 8 37 In this case 500 is the frequency in megahertz The magnitudes of 11 21 12 and 22 are 64 12 5 03 and 8 respectively The phases are 23 98 70 and 37 degrees respectively Alternatively Y parameter data may be used The format specifier could be GHZ Y RIR1 This would indicate rectangular unnormalized Y parameter data with frequencies in GHz A typical line is 30 0 3E 4 9E 3 8E 3 2E 5 0 1E 4 1E 3 In this case the frequency in gigahertz 1s 30 The real values of Y11 Y21 Y12 and Y22 are 0 9E 3 2E 5 and 1E 4 mhos respectively The imaginary values are 3E 4 8E 3 0 and 1E 3 mhos respectively A sample S parameter data file is shown below The only portion of the file required for
275. e analysis and modeling for CAD of mm wave MMICs Alta Frequenza v LVIII 1989 N 5 6 p 115 122 A Hill V K Tripathi An efficient algorithm for the three dimensional analysis of passive microstrip components and discontinuities for microwave and millimiter wave integrated circiuts IEEE Trans v MTT 39 1991 N 1 p 83 91 The Method of Lines M G Slobodianskii A new method of approximate solution of partial differential equations and its application to the theory of elasticity in Russian Prikladnaia Matematika 1 Mekhanika Applied Mathematics and Mechanics v 3 1939 N 1 p 75 82 O A Liskovets The method of lines Review in Russian Differenzial nie Uravnentya v 1 1965 N 12 p 1662 1668 B L Lennartson A network analogue method for computing the TEM characteristics o planar transmission lines IEEE Trans v MTT 20 1972 N 9 p 586 590 U Schulz On the edge condition with the method of lines in planar waveguides Arch Electron Uebertragungstech v 34 1980 p 176 178 U Schulz R Pregla A new technique for the analysis of the dispersion characteristics of planar waveguides and its application to microstrips with tuning septums Radio Science v 16 1981 Nov Dec p 1173 1178 S B Worm R Pregla Hybrid mode analysis of arbitrarily shaped planar microwave structures by the method of lines IEEE Trans v MIT 32 1984 N 2 p 191 196 R Pregla W Pasch
276. e click select the Measurement Wizard to help place the spectrum data on the graph Alternatively double click the data generated by the analysis to view the measurements For Harmonic Balance simulations we are able to select from a range of node voltages branch currents and ports Node or port waveforms are also available via the Measurement Wizard and data set This may aide in viewing the distortion and voltage levels at various nodes in the circuit There are several examples of Oscillators analyzed with Linear Harmonic Balance and Transient simulation in the C Program Files Genesys200X YY Examples Oscillators directory Usage and Theory of HB Oscillator Analysis Genesys 2006 04 and later All measurements from the basic HB analysis may be used in the HB Oscillator analysis Genesys 2004 and Genesys 2006 have a different implementation of the oscillator analysis In Genesys 2004 the analysis is implemented as an additional feature of HB analysis while Genesys 2006 has a fully independent oscillator analysis Genesys 2004 used a double nested Newton algorithm an internal part of which is a basic HB analysis which calculates complex amplitude of current through the oscillator port which externally controls the amplitude of the port voltage and its frequency The Genesys 2006 oscillator analysis algorithm is based on the tracing function whose parameter is the voltage amplitude of the oscillator port Vprobe when the
277. e conserving computer resources Typical Momentum GX RF applications include RF components and circuits on chips modules and boards as well as digital and analog RF interconnects and packages The simulation mode is defined in the Momentum GX analysis properties window Simulation options tab Momentum Options General Simulation Options Mesh Simulation Mode RF Faster no radiation effects Microwave frecalculates substrate for every frequency Deciding which mode to use depends on your application Each mode has its advantages In addition to specifically RF applications RF mode can simulate microwave circuits The following graph identifies which mode is best suited for various applications As you can see some applications can benefit from using either mode depending on your requirements As your requirements change you can quickly switch modes to simulate the same physical design As an example you may want to begin simulating microwave applications using RF mode for quick initial design and optimization iterations then switch to microwave mode to include radiation effects for final design and optimization High Speed Digital SI BGA RF Board FR4 Duroid RF Package Plastic Planar Antennas RF Module MCM LTCC Microwave hybrid Aurina Microwave IC GaAs Initia design amp Final design amp optimization e optimization Choose the
278. e exact at this frequency and approximate at all others The Accuracy Testing section on the Convolution tab sets the accuracy requirements Calculate Now Click this to force CAYENNE to recalculate and decache models CAYENNE Transient Analysis Transient Analysis Properties General IntegrationjTime Step Convolution Gutput Miscellanenus Integration Options Integration Method Trapezoidal Max Order For Integration Degree of Mixing Coefficient mu 0 5 Step Control and Predictor Options Method Gear Max Order 3 Use Predictor Generally Faster Time Step Method C3 Fixed Controlled by Truncation Error Robust Controlled by Truncation Error Approximate Overestimate Factor 20 Minimum Step Size Use 1 of Maximum Step oF e Integration Options Integration method tells the simulator what method to use for numerical integration The options available are Trapezoidal and Gear Max Order for Integration tells the simulator what order to use for numerical integration for Gear integration Orders two through six are supported Step Control is used to calculate predicted values to determine whether the time step needs to be decreased If Use Predictor is selected then these predicted values are also used as the initial guess for the next simulation time point If Use Predictor is not selected then values from the previous time
279. e how the timer function can be used module bitStreamGen out output out electrical out parameter period 1 0 integer x analog begin timer 0 period x random 0 5 V out lt transition x 0 0 period 100 0 end endmodule Operators Analog operators operate on an expression and return a value Furthermore they can operate on more than just the current value of their arguments as they maintain their internal state and so their output is a function of both the input and the internal state Because they maintain their internal state analog operators are subject to several important restrictions These are e Analog operators can not be used inside conditional if and case or looping for statements unless the conditional expression is a genvar expression which can not change their value during the course of an analysis e Analog operators are not allowed in repeat and while looping statements e Analog operators can only be used inside an analog block they can not be used inside an initial or always block or inside a user defined analog function Advanced Modeling Kit Under most cases you can not specify a null argument in the argument list of an analog operator Operator Function Time The ddt operator computes the time derivative of its argument The derivative form is ddt expr Time The idt operator computes the time integral of its argument The integral general form is idt exp7 Linear time absd
280. e information in the EMPOWER Operation chapter on Lumped Elements and Real Time Tuning also apply to Sonnet Simulations so you should review those sections now Viewing the Geometry in the Sonnet Native Editor Note This command is only available when Manual Mode is off When using a Sonnet simulation you may want to preview memoty usage and subsectioning before running a simulation To do this right click the Sonnet simulation in the Workspace Window and select View in Sonnet Native Editor This will bring up the native Sonnet geometry editor To estimate memory select Estimate Memory from the Analysis menu To view subsectioning press View Subsections after selecting Estimate Memoty 411 Simulation Note Any changes you make will not be saved into the GENESYS workspace and may be overwritten at any time If you need to make changes to the Sonnet geometry file you should first switch to Manual Mode 412 Chapter 10 Spectrasys System Spectral Propagation and Root Cause Analysis SPARCA A new simulation technique has been created to simulate RF architecture This technique is called Spectral Propagation and Root Cause Analysis Every source spectrum at every node propagates both forward and backward to every node in the schematic Along the way they create noise intermods harmonics and phase noise and these spectrums propagate to every node in the schematic These spectrums contain spectral dens
281. e next step is create an EM simulation of the layout We do this by right clicking on Simulations Data folder and selecting Add Planar 3D EM Analysis Then you should set up your simulation by changing the number of points Your EMPOWER options should be setup just like ours below Pressing Recalculate Now will start the EM simulator This simulation takes a couple minutes on a Pentium III 500MHz CPU with 192MB of RAM Note We specified a wider simulation frequencies to get a bigger picture of the response 200 EMPOWER Planar 3D EM Analysis EMPOWER Options General Viewer Far Field Advanced Layout to simulate MFilterl Lay Port impedance 50 Generalized Setup Layout Port Modes Use ports from schematic Necessary for HARBEC co simulation Electromagnetic simulation frequencies Start freq MHz 38 Stop freq MHz 2450 Start freq MHz QE Number of points ho Stop freq MHz E F I Harles re Number of points Li Max critical freq 2200 7 Iv Turn off physical losses Faster Recalculate Now Automatic Recalculation Co simulation sweep v Use EM simulation frequencies Automatically save workspace after calc Cancel ITI Help 20 After the simulation is done you should first take a look at the Empower Listing file Right click the Empower simulation titled EM1 in the workspace tree and select Show Listing File Inspect this file to verify the s
282. e of this documentation to elaborate on noise correlation matrices and other noise simulation techniques For additional information on noise correlation matrices see Computer Aided Noise Analysis of Linear Multiport Networks of Arbitrary Topology Vittorio Rizzoli and Alessandro Lipparini IEEE Transactions on Microwave Theory and Techniques Vol MTT 33 No 12 December 1985 Two Port Noise Amplifier Analysis The noise figure of a two port amplifier is given by 437 Simulation 438 F Fmin rn gs Ys Yo 2 where rn is the equivalent normalized noise resistance of the two port i e rn Rn Zo Ys gs j bs represents the source admittance and Yo go j bo represents that source admittance which results in the minimum noise figure Fmin Ys and Yo can be expressed in terms of the reflection coefficients Gs and Go as Ys 1 Gs 1 Gs and Yo 1 Go 1 Go See G Gonzalez Microwave Transistor Amplifiers pgs 142 142 Prentice Hall Inc New Jersey 1984 What this means in layman s terms is that the noise figure of an active device is very much a function of the source admittance or impedance NOTE Cascaded noise figure equations make the assumption that noise figure of an active device is independent of the source impedance which can clearly lead to erroneous results Propagation Basics The basic operation of Spectrasys involves the propagation of individual source spectra and all of their derived pr
283. e results MIXER LO This section controls the effects of all mixer LOs in a simulation Strongest Signal Only When selected the frequency of strongest LO signal is used to determine the output frequency of all mixed signals regardless of the number of other sionals that may be present on the LO All Signals Within X dBc of Strongest When selected all frequencies of LO signals falling within the specified range of the peak LO signal will be used to create new mixed output spectrum RANGE WARNING FOR MIXER MULTIPLIER ETC This group is used to control range warnings used by some elements 554 Spectrasys System Tolerance Range This threshold range is used by some elements to warn the user when a given power level falls outside the specified range This range applies to each element on a case by case basis For example the total LO power for the given mixer will be determined by integrating the LO spectrum and then comparing this power level to the LO Drive Level for the given mixer If this power level is outside the Tolerance Range window then a warning will be issued for this mixer either indicating that the mixer is being starved or over driven System Simulation Parameters O utput Tab This page contains miscellaneous Spectrasys options Tabs General Paths Calculate Composite Spectrum Options Output System Simulation Parameters General Paths Calculate Composite Spectrum Options Output Retai
284. e set the Grid Spacing X 10 and the Grid Spacing Y 10 and the Box Width X to 640 and the Box Width Y to 800 The other properties should be set as follows We have chosen the grid spacing to be 10 because the widths and lengths of the synthesized filter are very close to multiples of 10 mils 191 Simulation LAYOUT Properties x General Associations General Layer EMPOWER Layers Fonts Units Designs to include C Milimeters Design E Box Settings The UNITS box at left show units Grid Spacing x i D IY Show Box Mils MFilter1 MFilter1 Lay p i i v Show Grid Dot Ci Cuon ERES arid Spacing r 10 M Show Grid Dots M Show EMPOWER Grid Box Width s E40 54 Cells Bax Height T gon 80 Cells Origin E jo Object Dimensions Line Width 25 Pad width 50 Drill Diameter 40 Drawing Options Port Size 50 Rot Snap Angle 30 Multi Place Parts Remove Add Mew Drawing Style C Solid Opaque S Alay mode C Hollow Default Yiahole Layers Top Layer TOP METAL Bottom Layer m Bottom Cover Cancel Apply Help 10 Now we need to slightly edit our layout The ultimate goal is for the resonators dimensions to be an exact multiple of the grid dimensions In this case the spacing between resonators is very important therefore we will not change them to much First we need to change some of the dimensions in the equati
285. e structure onto the borders of the cell not onto the space inside the cell A slightly mote complex example which does not exactly fit the grid 1s shown below There are three important things to notice in this figure 1 The stub line going up is about 2 1 2 cells wide but is approximated by EMPOWER as being 2 cells wide 2 The 97 Simulation chamfered corner is approximated by a stairstep 3 The viahole near the end of the stub is represented by an asterisk in the listing 0lz34567839501z3a4s5 B4 t l4 t 4 4 l E l3 4 l E l 0 lll o 4 4 l E g lo 4 B ltf t t t 4 4 4 Id tt t dod S 1l 1 tt tt tol 7 1l1 1 tt dE P gd td E 1 l dI E t t t t dI 4 t tet t 4 dI 3 tt t tFtt t H H Hl l dI ZI tet t 4 l dI l t t t 4 4 4 l ZZ EE Ot nn er er er eer er ee A close up is shown below where you can see how metal and ports are mapped onto the borders of the cells The presence of metal or conductors along the grid causes EMPOWER to close the connections along the grid The presence of an EMPort causes the line to be opened creating an open circuit which turn
286. e the absolute time by the value of the argument 1 e scale to the value specified in the argument The argument for realtime follows the semantics of the time unit that is it shall consist of an integer followed by a scale factor Valid integers are 1 10 and 100 valid scale factors are s seconds ms milliseconds us microseconds ns nanoseconds ps picoseconds and fs femtoseconds vt can optionally have Temperature in Kelvin units as an input argument and returns the thermal voltage KT q at the given temperature vt without the optional input temperature argument returns the thermal voltage using temperature Input output operations These functions provide access to display and file operations Pfopen fzie nage Sfclose fzie 2d strobe args monitor a7gs fopen opens the file specified as an argument and returns a 32 bit multichannel descriptor which is uniquely associated with the file It returns O if the file could not be opened for writing fclose closes the channels specified in the multichannel descriptor and does not allow any further output to the closed channels fopen reuses channels which have been closed strobe provides the ability to display simulation data when the simulator has converged on a solution for all nodes using a printf style format monitor provides same capabilities as strobe but outputs only when a parameter changes Advanced Modeling Kit Eaglewa
287. e used As a general rule you will get better accuracy for a given amount of time and memory when you use thinning Thinning out helps by removing currents which have little or no effect This reduces the number of vatiables in the problem considerably with little effect on the accuracy of the solution There are a few cases where thinning out should not be used and they generally involve very large sections of metal which are affected too much by thinning out The Dual Mode Power Divider example is one of these cases EMPOWER Planar 3D EM Analysis Wall amp Cover Spacing Generally the wall and cover spacing should match the problem which you are trying to model This will give an accurate assessment not only of circuit performance but also of box resonances However this will not be possible in a few situations 1 The structure will not be in a box 2 You are analyzing part of a larger circuit and the box walls would be prohibitively large to model 3 You are designing a component such as a spiral inductor which will be reused in many different circuits so the cover height is not known In these cases you must use an approximation Set the box size so that the walls are separated from the circuit by at least 3 times the substrate thickness preferably 6 times For microstrip set the cover spacing air above to 5 to 10 times the substrate height See for more info see Box Modes See Microstrip Line for an example of the eff
288. e we will use microstrip standard as the Process The conversion 188 EMPOWER Planar 3D EM Analysis process should look similar to what we have below Press OK and you should have an schematic that looks like ours below Lonvert Using Advanced TLINE E Defaut Untied 1 m OK Unused nl Cancel Hep pn Schematic of the synthesized unoptimized filter 189 Simulation ttt C 22 53 pF CAP 1 Ces pr ice We o3 909 mil L 200 9 76 mil IL 2 Wi 84 956 mil 2133 438 mil Lead W B4 8358 mil 2133435 mil Lead Lx s EE W 83 909 mil L 78 808 mil IL1 Notice how M FILTER automatically inserts the discontinuities to model their effect 8 The next we must optimize this filter by pressing the Optimize button at the top of the MFilter dialog box This is very important in order to obtain the expected filter performance in this example After the optimization the schematic should now look as shown below Note You should stop the optimizer by hitting the Esc key once the error value is not improving much 190 EMPOWER Planar 3D EM Analysis F Cz2 328 pF CAP 1 Cz22 326 pF CAP1 W B3 8D8 mil L 200 5 78 mil IL 2 TL We 84 958 mil Ep PME 5 o i3 c 133439 mil Lead ces 909 mil L 101 484 mil IL1 W 84 958 mil 2133 435 mil Lead 9 Now we need to set the board dimensions and the EMPOWER grid spacing For this walkthrough w
289. eal parameter ist The statement_block and analog function e Can use any statements available for conditional execution e Can not use access functions e Can not use contribution statements or event control statements e Must have at least one input declared the block item declaration declares the type of the inputs as well as local variables used e Can not use named blocks e Can only reference locally defined variables or passed variable arguments The analog function implicitly declares a variable of the same name as the function function name This variable must be assigned in the statement block its last assigned value is passed back Example analog function real B of T input B T T NOM XTB real B T T NOM XTB begin B of T B pow T T NOM XTB end endfunction The function is called by the line BF T B of T BF T T NOM XTB System tasks and functions System functions provided access to system level tasks as well as access to simulator information 59 Simulation 60 Environment parameter functions These functions return simulator environment information temperature Circuit ambient temperature in Kelvin Absolute time in seconds realtime scale vt 1 emperature realtime can have an optional argument which scales the time If no argument is given realtime s return value is scaled to the time unit of the module which invoked it If an argument is given realtime shall divid
290. easurement Bandwidth Types of Spectrums Used Same as DCP Travel Direction Same as DCP Input 1 dB Compression IP1DB This measurement is the system input 1 dB compression point referenced to the path input The value at the output of each stage represents that equivalent input compression from input to that stage output The last entry will always be the performance of the entire chain referenced to the input IP1DB n DCP 0 min SDR n dBm where n stage number Spectrasys System This measurement is made by determining the desired channel input power for the first stage and then adding to it the minimum 1 dB headroom or stage dynamic range along the path Channel Used Same as DCP and SDR Types of Spectrums Used Same as DCP and SDR Travel Direction Same as DCP and SDR Input Intercept All Orders IIP This measurement is the intercept point referenced to the path input This is an in band type of intermod measurement IIP n OIP n CGAIN n dBm where n stage number This measurement simple takes the computed Output Intercept and references it to the input by subtracting the cascaded gain The last IIP value for a cascaded chain will always be the actual input intercept for the entire chain Each column in this measurement is for a different intermod order up to the Maximum Order specified on the Calculate Tab of the System Analysis Dialog Box The column number is the same as the order star
291. eated ss one port and assigned the game port number Assign as differential ports They will be treated as one port and assigned the same port number To define a differential port 1 Placea new EM pott in layout or select any port that you want to assign this type to Note the port number Open the port dialog double click or select and click Enter For the Port Type select Differential Under Polarity make sure that Normal is selected Click OK oY GEM ae ee dp Place another new EM port in layout or select the second port Note the port number For the Port Type select Differential 8 Under Polarity select Inverse 9 Under Associate with port number enter the number of the previously selected port 319 Simulation 10 Click OK For example this design has 2 groups of differential ports 1 2 and 3 4 Each differential port consists of 2 EM ports with different polarity associated each other To define the differential port 1 2 for port 1 we use polarity Normal positive Draw Size Ref Plane Shift Port Number Location j PartL ist Layer E TOP METAL Width Line Direction Height jo Current Direction Port Type Associate with port number Polarity Mormal CO Inverse Differential mode STRIP port transmission line extended calibration For port 2 we define associated port number 1 and inverse polarity 320 Momentum GX Draw Siz
292. ect of wall spacing on line impedance Cover Type Choosing the correct cover type is absolutely critical to getting an analysis which matches measured results The choice is usually between whether to use an open cover or a closed cover Choosing the correct cover type usually has no effect on analysis time so there is no reason not to set this to the proper type With an open cover there will be radiation and this can have a huge impact on circuit performance You can choose the correct cover types in the Layers Tab when starting an electromagnetic simulation See the EMPOWER Basics and Box Modes sections for more information on covers See the Edge Coupled Filter example for an example of the impact that removing a cover has on circuit performance Lossy Analysis If you do not need information about circuit loss you can check the box labeled Don t use physical loss Faster when starting an EMPOWER run Turning off losses will generally make a problem require 1 2 the memory and 1 4 the time as a lossy problem We tecommend that you define all layers with their proper characteristics including losses You can then quickly change between lossy and lossless modes as described above A common technique is to analyze a circuit first without losses then turn on losses and run an analysis with a few points in it This allows you to determine the amount of loss and confirm that it has no other major effect on performance while not having
293. ect this box to cause EMPOWER to simulate the layout at each frequency calculated by the harmonic balance simulator Checking this box makes sure that EM results are available at all frequencies so that the data will not need to be interpolated or extrapolated for harmonic balance analysis Max Critical Freq MHZ Specifies the highest important frequency that will be analyzed on any run of this circuit MAXFRQ is specified in the units defined in the DIM block The default units are MHz Parameters of the solution quality thinning out thresholds and lengths of lines for de embedding are based on the maximum critical frequency value In other words this value influences both accuracy of simulation and calculation time Decreasing the value accelerates simulation but may increase model error especially at frequencies above the value On the other hand an unnecessarily high value may slow down the solution without visible improvements in accuracy Note An important reason to specify MAXFRQ By default this value is set equal to the highest sweep frequency specified in EMFRQ Even a small change of its value may cause the erid to change forcing recalculation of de embedding parameters and unnecessarily increasing simulation time as a consequence This change will also change the answer slightly with disastrous results if you are merging data This will not happen if you use MAXFRQ It is also important to remember to update it if you change the
294. ed across the erid cell They are shown as lines connecting the corresponding geometrical point in the grid plane and the point corresponding to the actual current values If a via hole surface shape is known using the current in Amperes it is possible to estimate a current density on the via hole surface It is obvious from the picture that the current density is higher on the via hole side that 1s closer to the microstrip line segment empower Viewer Y6 5 Mil x File View xY a Mag Sold Freq 6H L eoelolelvel lna s To Fon Side Oblique PSI x empower Viewer Y6 5 File View z Mag Solid Freq GHz 1alalel of a 4 Top Fon Side Oblique EMPOWER Planar 3D EM Analysis Viewer Theory The EMPOWER viewer is a program designed to read to process and to visualize the current distribution data created by EMPOWER To obtain a current distribution inside a structure the excitation condition must be defined This mirrors a real measurement where there are incident and reflected waves The viewer depicts the case wit
295. ed in GENESYS Interface uesoeene tie esent 401 EMPOWER features not supported by the Sonnet Interface esses 403 TEENA escent N T E TP TE E A PE A E E E EE EE E 404 IEEE 6 16 cic 0 EPE E IM E M i EAT 404 xi Simulation Chapter 10 xii Creauno a Layout Without a Schemat Cisean a i Elo ba pea Ue pq 404 PETA EO EDO T VOU edt n be stai us qitot E qusip le Nob tuit DU ea utu Emus OE m tape iA De naL USE 404 Miesvipno RESU S eso tico Mises eoo tne MTs emir rere ttre ye er meray Terre ur E Dese 406 ising the Current amd Far Field VIeWweESucpide e ente dibetenta eddie navales 407 Creatine a Layout From an Exis ng SCODeWmalte aee eve d a atu indt urbe vt UE 407 Simulating the Layout with Lumped BIeIments siio dirt eo elena ien 409 Viewing the Geometry in the sonnet Native BIalitor a s heces tide tasti toe depends 411 Spectrasys Systemi irents 413 Spectral Propagation and Root Cause Analysis SPARCA Lie paren teta ipte vas 413 Gertner otl a E aad eddie 414 opecttasys W alkthroueh COVBEUIGNV su medesime veniae tts dovete aO 414 CENE a SYST S6 DEPO oe tuti dade a tu pictum A 414 INGGL Ar SY SUCH TIAIVSIS aaa isis iii taton teta e tdtel AEA on e teda RP ua DOMUS 416 Ramine SATU AO stylist dete tont sd ped rant Sih ass toot A 419 Adda Grapbor Tibisco d then a reo pq eed thao I supe iiu 420 lutdanieBtals aaue itta art E N n Alte tate tutes Ud 423 General behavioral Model OVeEVIewisseid Dn cese on e V nitent bd n
296. ed in the Decomposition section of your EMPOWER manual If this button has exclamation points on it then multi mode lines are active Thinning out slider Control the amount of thinning The default thinning out amount is 5 Setting the slider to zero turns off thinning See your EMPOWER manual for details on thinning Thin out electrical lossy surfaces If checked lossy metal described using electrical parameters will also be thinned Since the thinning out model assumes that most current flows on the edges of the lines this option will be somewhat less accurate for resistive films where current flows more evenly throughout the material In these cases you should probably also check the Solid thinning option shown below Solid Thinning out slower If checked slower solid thinning out model is used This model restores capacitance lost due to thinning out and can be most useful for when large sections of metal have been thinned out Use planar ports for one port elements This box should almost always be checked When not checked EMPOWER uses z directed ports at each terminal for all devices When it is checked EMPOWER uses in line ports for elements like resistors and capacitors two terminal one port devices The only time this can cause a problem is when you have a line running under an element for example running a line between the two terminals on a resistor in the same metal layer as the resistor pads Note EMPOW
297. edge to match the 5 synthesized schematic by using the arrow keys Your layout should look like what we have below Next 16 Simulation Tip You can move the designator text if desired by grabbing the handle in the middle of the text block 17 Now it is time to add an input and output port EM ports are found in the GENESYS toolbar below The EM ports should line up exactly on the pwb edge and the gray shaded bar underneath the EM port will appear showing that they have been snapped correctly to the top metal layer z Aoeo Draw EMPort layout editor or Footprint Port Footprint editor Variable Filter Layout of the schematic 198 EMPOWER Planar 3D EM Analysis 18 Next we need to set the EM layout properties as shown below This can be done by bringing up the Layout Properties by double clicking the layout background and then selecting the EMPOWER Layets tab 199 Simulation LAYOUT Properties General Associations General Layer EMPOWER Layers Fonls m EL eme ERR eum tvpe Thickness Sigma Value or File Direction Height Slow Z ports Topcover guxpeeu s foo O o y y o M po fo TOP METALS Sub Detta joss Norma m Down SUBSTRATE I fv supeeus psspowp 1 r See BY E MM Argw CO T T EI ae oes _ Cancel Apply Help 19 Th
298. eding scheme also has its own unwanted effect low order mode mismatch at the port s boundary although this is eliminated by the calibration process However in order for this calibration process to work well it is necessary that the fundamental mode is characterized accurately This can only be accomplished when the distance between the port boundary and the first discontinuity is sufficiently large that is there exists a feedline that is long enough to provide this distance As a basic example consider a linewidth that varies abruptly in some part of your circuit as shown in the example below p m o Q o Strictly speaking all you need is to characterize is the variation of the step in width itself as shown below As mentioned previously it takes a little distance for the fundamental mode to settle which means that this short structure might not yield the accuracy that you expect from an Momentum simulation In this case allow for some feed line length 313 Simulation 314 Now the simulation will yield accurate results but the results will also contain the extra line lengths To remedy this use reference offsets Although the circuit has been calculated with the long lines reference offset shifting allows you to produce the S parameters as if the short structure had been simulated instead pi ji apply reference offsets at these locations The effect of the extra feed lines is mathematically elimina
299. efault name Graph2 2 Very Important Select Sonnet Stub for Default Simulation Data or Equations 3 Enter S21 for the first measurement and S11 for the second measurement Note You can also use the Measurement Wizard to select simulation results This instructs GENESYS to display a window with 21 and S11 from the Sonnet simulation as shown below For a complete description of rectangular graphs see the GENESYS User s Guide 406 Sonnet Interface E Graph Workspace Layo iL fol A in I H2 c 30 o tn Ca 25 C Ln t Ca I um CT I T I RR em I CT c 4500 Freq MHz e DB 521 DB S11 U sing the Current and Far Field Viewers Once the Sonnet run is completed the current viewer can be used if Generate Viewer Data was selected in the Sonnet options dialog Generating this data slows the simulation so it s usually only checked during last run simulations Note The viewer will NOT include the effects of lumped elements Right Click the Sonnet simulation in the Workspace Window and select Run Sonnet Current Viewer or Run Sonnet Far Field Viewer For more details on using these viewers see your Sonnet manuals Creating a Layout From an Existing Schematic The file used in this example is FiltersV Tuned Bandpass wsp This circuit is a tunable bandpass filter This example demonstrates the following topics 407 Simulation e Creating a layout fro
300. egative values n phase shift Any objects drawn on slot metal layers are slots holes in metal where T open areas define metal In Genesys layout slots are displayed identically to metal in strip layers EM Port Properties Draw Size Ref Plane Shift Port Number Location 650 Layer IB TOP METAL Current Direction Width jo Line Direction Height Pork Type 4 PartList gt Layo Associate with part number Polarity Normal O Inverse Coplanar Mode SLOT port transmission line extended calibration 324 Momentum GX Bat hr RI ata 5 cm FOr L IU DET eS Draw Size Ref Plane Shift Cancel Port Number Help Location 0 r Layer BN TOP METAL Line Direction Default Current Direction Default Port Type Coplanar Exe Layout Associate with port number 1 C2 Normal Inverse Coplanar Mode SLOT port transmission line extended calibration The slot layer with coplanar ports may be equivalently represented by a strip layer with non calibrated internal ports with ground reference ports located as close to signal internal ports as possible 325 Simulation In this case the internal port 1 has two ground reference ports 3 and 4 and the internal port 2 has ground reference ports 5 and 6 The closer the ground reference ports to their internal ports the closer the results of the strip equivalent to the
301. egins with three basic elements Amplification a frequency determining circuit or device and feedback to overcome network losses and provide power to the load 249 Simulation We start by selecting an amplifying device and topology that will provide gain at the desired frequency band of frequencies for tunable oscillators Next some form of a frequency selective network is added e g crystal L C circuit cavity or dielectric resonator And finally a feedback path that provides power flow from the amplifiers output back to the frequency selective network There are generally many topologies available to provide positive feedback however the path should be chosen such that opening the path would result in termination of oscillation A path that provides positive power flow from input to output S21 gt 1 in a broken feedback loop is an excellent starting point The easiest way to begin an oscillator analysis in GENESYS is to create a new workspace from the built in Oscillator Template in the Getting Started with GENESYS screen the first thing that comes up when you choose File gt New Create a new workspace from a template a um E Linear Simulation Select a template from the list on the right Nonlinear Simulation Click OK to continue Oscillator Template The template s schematic is shown below 250 HARBEC Harmonic Balance Analysis gt Sch_Open fax amp C 100pF E Port 1
302. egions After connection of the lumped elements the immitance matrix can be transformed into a generalized Y or S matrix using the simultaneous diagonalization method see the de embedding section Thus we have a problem formulation that is appropriate for a wide range of microwave and mm wave devices such as planar filters dividers combiners matching circuits phase shifters attenuators diplexers amplifiers as well as their components Method of Lines The method of partial discretization later called the method of lines MoL is as old as partial differential equations and the finite difference approach to their solution Traces of it can be found in the 18th century wotks of J L Lagrange Its first conscious usage fot the numerical solution of elliptical problems could be attributed to M G Slobodianskii 1939 An almost complete reference on the MoL development and applications in the petiod from the beginning up to sixties are given in Liskovets paper 1965 The network analogue method of B L Lennartson 1972 is probably the first technical application of the MoL to the static numerical analysis of planar multiconductor lines It was not quite straightforward when it was published and the actual exploration of the method for microwave integrated circuit structures began in the early eighties in works of German scientists H Diestel R Pregla U Schulz S B Worm and others Pregla Pascher 1989 The EMPOWER algorithms
303. ei be pide 423 vil PT C 423 Ban Dels o adeceened ted ead Desde dicus buat tsa baad ase aote loud aes ebat ein taii ea ao ere aE ten tiie 424 Opecin Me PAINS eodeni ipid itemm use iuit bout AE 426 IS UTA AIS m c 427 opectturmPIotsdrid Fables ossia qepidbn i bonor a d Re rasis cia ene a 431 Idenutyine Special Oo Maninoa nn an Eo ERE ECCO R It E Eh gate EN pa dEuE 433 DtoddBan DNDISE o eds te ebat pu eU tb A Cree Crrre meer ete eon 436 Two Pore Noise AMPUH ATIS S o ucc etae pee tust t esl eund vest ni i tei 437 PEO PACA tl MIDAS ICS CMM EM aS 438 Controlling Analysis Dati sete tv este tread O due e eim snore 439 DUECDS OE DAD act etta pa at bean atu Nb octauo siat it an ra E A OMM Uca un p ES 440 1 Bal sos EES is NR TM TUR IE 444 Nonlinear Model BeliaviO ss eb OI ERU reb ee etd iene 444 Interoeod and Earmonic Dasic Ssmi Doa Itane b mU tta hut 445 cascaded Enterpmod GUA ONS usi i rr o amid inane net tmn ions 451 Intercept Measurements tir the Lusso artritis dte d tb bare etse 452 Int rmod Path Measurement Basic Sinica i Medo tcv Sor t o e a o e to eate 457 C ascaded Intermod Equations and Spec tas y Sesion ditte iita aisi omis 460 Troubleshooting Int rmod Path Measurements ssssesecossti tete e pe petri dern putet 465 JAVep Neo OR senda A E T E T 466 Cascaded Nobbe Analysis eneren pat tp A E A OE A 466 Gascaded None T eure Guan OMS sninn E rt od adidas 470 C
304. el Using the spectrum analyzer mode we quickly see the peak at 70 MHz and even though we see the individual spectrums at 100 MHz we know their total must be equivalent to or lower than the noise floor Random noise is also shown in the example Spectrasys System Output Power dBm Frequency MHz Synthesis Some behavioral models can directly synthesized from Spectrasys Right clicking on the behavioral model will bring a context sensitive menu This menu will list the synthesis modules available for the given model 481 Simulation 482 Filter3 IL 1dB10 N 3 aP nnn ee nau as Ackive Filter as Microwave Filter Format VIEN b aiii Find Part In Layout C1 2 C 1000pF w Keep Connected B1 0 w Show Park Text oplitter IL 6 0206dB 10 Properties Schematic Properties The selected synthesis module will be invoked and the parameters of the behavioral model will be passed to this synthesis module as shown below Spectrasys System Filter Properties Attenuation at Cutoff dB 3 011053 Issues Estimate Order Output Resistance 50 S06 Frequency is 450MH2 Once the model has been synthesized the synthesized circuit is substituted back into the behavioral model 483 Simulation Filter3 Subnet Filter3 Schematic RFAmp2 G 20 5dB10 plitter NF 4dB10 IL 6 0206dB10 At this point the parameters for the behavioral model will be di
305. elay implements the absolute transport delay for continuous delay waveforms The general form is absdelay zeput td maxdelay Discrete transition expr td rise fume fall_time fume tol waveform filters transition slew The slew analog operator bounds the rate of change slope of the waveform The general form is slew expr max_pos_slew_rate max neg slew rate The last_crossing function returns a real value representing the simulation time when a signal expression last crossed zero The format is last crossing expr direction Laplace laplace zp implements the zero pole form of the Laplace transform transform filter The general form for each is laplace_zp expr s r e filters laplace_zd implements the zero denominator form of the Laplace transform filter The laplace np implements the numerator pole form of the Laplace transform filter laplace nd implements the numerator denominator form of the Laplace transform filter Z transform The Z transform filters implement linear discrete time filters Each filter filters uses a parameter T which specifies the filter s sampling period The zeros argument may be represented as a null argument The null argument is produced by two adjacent commas in the argument list All Z transform filters share three common arguments T t and tO T specifies the period of the filter is mandatory and must be positive t specifies the tra
306. eld in the slot is considered This electric field is modeled as an equivalent magnetic current that flows in the slot Momentum does not model finite ground plane metallization thickness Ground planes and their losses are part of the substrate definition By using slot metallization definitions entire structures such as slot lines and coplanar waveguide circuits can be built Slots in ground planes can also be used to simulate aperture coupling through ground planes for multi level circuits Structure components on opposite sides of a ground plane are isolated from each other except for intermediate coupling that occurs through the slots This treatment of slots allows Momentum to simulate slot based circuits and aperture coupling very efficiently Simulating M etallization Loss When using Momentum losses in the metallization patterns can be included in the simulation Momentum can either treat the conductors as having zero thickness or include the effects of finite thickness in the simulation In the substrate definition the expansion of conductors to a finite thickness can be turned on off for every layer For more information refer to Via Structures Limitation Momentum uses a complex surface impedance model for all metals that is a function of conductor thickness conductivity and frequency At low frequencies current flow will be approximately uniformly distributed across the thickness of the metal Momentum uses this minimum resi
307. element models Also the model and equation features provide for user creation of models However it is often necessary or desirable to characterize a device used in GENESYS by measured or externally computed data This function is provided for by the use of the ONE TWO THR FOU NPO and NPOD elements which read S Y G H or Z parameter data De embedding is done with the NEGxx and NEGDxx elements Note The information provided in this section applies to linear devices When using linear simulator circuits are assumed time invariant element values are not a function of time and thus sub components are uniquely defined by a set of port parameter sets such as two port S parameter data Although ONE TWO THR FOU and NPO ate typically used for active devices they may be used for any devices for which you can compute or measure data For example they could be used to characterize an antenna a circuit with specified group delay data or measured data for a broadband transformer or a pad Using a Data File in GENESYS Data files can be used in GENESYS in two different ways Simulation e By overriding the simulation properties ofa patt This allows measurements to refer directly to the data file without the need to create a design e By using ONE TWO THR FOU or NPO elements in a schematic Use NEG in a schematic to deembed S Data In both cases you must know in advance how many ports the device data represents For tran
308. elements and substrates Pam ums Deseription pF I medaxe lt lt S El Mote Netlists use the default GENESYS units and the units specified in the substrate Changing these parameters will only affect new objects existing schematics will mot be modified Restore Defaults OK Cancel 2 Next lets open the Microwave Filter module from the GENESYS tree to start the design process by selecting the New Item button and picking Synthesis Add Microwave Filter In the Create a new Microwave Filter dialog box change the Initialize using to Factory Default Values then select OK Create a new MFilter Name MFiter Initialize using Cancel Last Saved Values 3 Now the user will be prompted for the printed wiring board layer settings in the Select Layout Setting File dialog box Select Standard ly and then OK 186 EMPOWER Planar 3D EM Analysis Select Layout Settings File Tour new layout will be based an DK LastlsedSetlings standard we Cancel Help i Directory c program Filezsgenesus 6 7047 emplate Note You can create new layout settings by using the Layout Save Layout Settings menu item 4 Itwill ask you to specify a substrate For this example just choose all default values set Er 2 55 Height 31mil and press OK Substrate Properties x Mame
309. em NOTE In Band and Out of Band intercept measurements will yield the same results if the tones are attenuated At 4 b 9h freg ICF In Band Spectrasys Measurements 463 Simulation 464 b alt F aus tone dBm DUT Gain dB TIMCP F one in dam Pins dem Out of Band Spectrasys Measurements From the intermod power level and the measured tone power level intercept points for all orders up to and including the maximum order are calculated However only one of them will be valid since the intermod frequency will be different for each order In Band Intercept Simulation 1 Create a two tone source and connect it to the DUT 2 Create a path and set its frequency to the intermod frequency 3 Set the Tone interferer Frequency of the path to the tones that will be used to calculate the intercept point 4 Set the Channel Bandwidth to a value wider than the intermod but narrow enough to exclude any power from the tone 5 Add the IIPn or OIPn measurement to a graph or table when n is the intercept order Out of Band Intercept Simulation 1 Create a three tone source and connect it to the DUT 2 Set the frequency of the third test signal to the frequency of the intermod 3 Create a path and set its frequency to the intermod frequency Spectrasys System 4 Set the Tone Interferer Frequency of the path to the tones that will be used to calculate the intercept point 5 Se
310. enDeel ad ac debentibus dessus ia Dna dtu ipd 8 IExpostinp ata Iles oce A tubi oec i Dusike e eb oec na ud 8 Nore Datan Data Files eoo or er tad bens rata iecit bos ern ote tectus 8 Neasa vo eg Uc ak come ae nee Nr nT cnn eee rN oT eo gee ae OP een er ner re ora 10 CV IVS RM us 10 Dele al fe mem H 15 Eo Te RU UU UU ENA AEA E ATS 29 INOBITICdE erase Eia dus cu ete avaN oR Sine RESIN p i ue ME EE M EWES DM IM LE eR MEE 31 Parameter WEE DSuau oippisul apice mndE ir i uic ntu adt oum O usua DER UH DE 38 Parameter owe aen tactic DM E E tita IM EN ME 38 Petiotmine a Parameter Sweep etait bed pel entuptae A 38 Patameter Sweep PLO Peres aea deett ient odeurs elei cue uel dii bebat 42 Advanced Modeling Kit iiiio e cbe eee radere a eere rao os Fonte eo N 45 Advanced Modelo it 4 vetviesisesamaotanppd n RU ER Leste b rene eoru bs 45 Usine phe additional AMK MOdelSus sese toertad tete ata ped nb EMI A NER RUM ERU rbdU els 45 Greapne ING or ventos MOSS ae edt timete ei ed rmt adis oadi teu hte eth ipia tis 46 Customizing Built in Nonlinear Models edet pieta auiem t ea EIU 47 Veroa Lutosbal ie etenim DEMNM eC 47 Memo Os FOR CLC CN Ce T 49 Meroe Reteren ee Overview onair e AAO E EOE 49 lg d e c 49 PD Ata Iypesand Parameters ai een Eto dte T ON 50 AOT DIOC ete mob codi aeu t tuc Mc eM DID Sia aed Ut 53 ADOS PUtIe HOS ea ent euo Deb O De O ON OTN 59 Syse aS kS and CUON eda ied E bmi tud
311. enerally you should choose the electrically shortest path for this direction Substrate Media Layers All substrate layers from the General Layer Tab are also shown in the EMPOWER Layer tab These layers are used for substrate and other continuous materials such as absorbers inside the top cover An unlimited number of substrate media layers can be used The following types are available e Physical Desc The layer is lossy These losses are described by Height in units specified in the Dimensions tab Er relative dielectric constant Ur relative permittivity constant normally 1 and Tand Loss Tangent e Substrates Choosing a substrate causes the cover to get the height Er Ur and Tand parameters from that substrate definition We recommend using this setting whenever possible so that parameters do not need to be duplicated Caution For true stripline triplate be sure to check the Use 1 2 Height checkbox if you are using a substrate This forces EMPOWER to use 1 2 of the substrate height fot each substrate above and below so that the total height for both media layers is correct EMPOWER Planar 3D EM Analysis In addition to the metalization and substrate layers viaholes and other z directed currents can be used These currents can go from the metalization layer through one media air layer to either the top or bottom walls Besides conductive materials ports are placed on the metal layers and in z directed positio
312. enerated in the signal generators before appearing at the DUT input A typical setup is as shown below using two tones Signal Generator Spectrum Analyzer Signal Generator 2 hod 2 5 f f 254 Figure 1 In Band Intermod Measurement Setup The intercept point is determined from the measured power level of the two tones and the powet level of the intermods themselves on a spectrum analyzer as shown in the following figure Spectrasys System Fins dBm 2b a G IA freq Figure 2 In Band DUT Output Spectrum From this information the output third order intercept is determined as follows OTOIdBm Ptone out dBm Ap 2 The input third order intercept is ITOIdBm OTOIdBm GainDUT Out of Band Intercept Measurements Out of band measurements are more complicated since the tones have been attenuated at the IF output This is illustrated in the following two figures Out of Band Interferers Front End Filter Final IF Filter 453 Simulation Figure 3 Third Order Distortion Inside the Receiver Input Filter Signal Generator opectrum Analyzer Signal Generator 2 ES SET or 3 OL f rx Figure 4 Out of Band Intermod Measurement Setup Won t Work Without knowing what the un attenuated tone power level is the intercept point cannot be determined as shown in the following diagram ut Cant Measure Tones H h f ef freq Of f rx Figure 5 Out of Band DUT Output Spectru
313. entry of the CNF array The operator must be used to extract only the data at the node of interest E System Data Path1 2 Sweep1 E iz E hy arlable dE Equation FRF Swp NodeNames AN 2110 CCOMP 2110 CF I 2110 CAIN 2110 CEAN 2110 CNDR 2120 CMF 2120 CNP 2120 CNR 2120 CNRY 2120 CNY 2130 COMP 2130 CP 2130 CN 2130 DCE 2130 DCR 2140 DEW 2140 Design 2140 THEME FRF Sv 2140 T gt T 4 4n Intermods Non Linear Model Behavior In the real world components and stages exhibit non linear distortion such as gain compression and power output saturation To characterize non linear behavior compression points saturation intercept points and spurious free dynamic ranges are defined according to the following diagram 444 Spectrasys System Third Order Intercept Point Saturation Region Saturated Output Power Level 5 Output Power at 1 dB compression point Burnout Input Power Level Causing Burnout N Third Order Intermod Product Slope 3 1 Output Power dBm Minimum Detectable Signal Spurious Free Dynamic Range Total Noise Level The above figure illustrates 3rd order intermod and noise performance boundaries As shown a higher intercept point yields larger dynamic range Consequently intercept points are commonly used as a performance characteristic of RF systems In general the higher the intercept point the more tolerant the
314. equal to the number of basis functions The right hand side vector V represents the discretized contribution of the excitations applied at the ports of the circuit The surface currents contribute to the electromagnetic field in the circuit by means of the Green s dyadic of the layer stack In the MPIE formulation this Green s dyadic is decomposed into a contribution from the vector potential A r and a contribution from the scalar potential V r Gir r joo r pyi Y Gir ry Equation 6 The scalar potential originates from the dynamic surface charge distribution derived from the surface currents and is related to the vector potential through the Lorentz gauge By substituting the expression 6 for the Green s dyadic in the expression 4 for the interaction matrix elements yields the following form Z fm t ee Jj Joe Equation 7 with L aSByr fasc o ry Bir P asv Br f asc rV Bir C j S Equation 9 This allows the interaction matrix equation to be given a physical interpretation by constructing an equivalent network model Figure 2 In this network the nodes correspond to the cells in the mesh and hold the cell charges Each cell corresponds to a capacitor to the ground All nodes are connected with branches which carry the current flowing through the edges of the cells Each branch has in inductor representing the magnetic self coupling of the associated current basis function All
315. er The method of lines in Numerical techniques for microwave and millimeter wave passive structures Edited by T Itoh John Willey amp Sons 1989 S B Worm Full wave analysis of discontinuities in planar waveguides by the method of lines using a source approach IEEE Trans v MTT 38 1990 N 10 p 1510 1514 EMPOWER Planar 3D EM Analysis Richardson s Extrapolation L F Richardson The differed approach to the limit 1 Single lattice Philos Trans of Royal Society London ser A 226 19277 p 299 349 A Premoli A new fast and accurate algorithm for the computation of microstrip capacitances IEEE Trans v MT T 23 1975 N 8 p 642 647 G I Marchuk V V Shaidurov Dzfference methods and their extrapolations Spt Verlag 1983 originally published in Russian 1979 A G Vikhorev Yu O Shlepnev Analysis of multiple conductor microstrip lines by the method of straight lines Journal of Communications Technology and Electronics 1991 N 12 p 127 129 originally published in Radiotekhnika 1 Elektronika v 36 1991 N 4 p 820 823 Symmetry Processing M Hammermesh Group theory and its application to physical problems Pergamon Press Oxford 1962 I J Good The inverse of a centrosymmetric matrix Technometrics Journal of Statictics for Physical Chemical and Engineering Science v 12 1970 p 925 928 P R McIsaac Symmetry induced modal characteristics of uniform waveguides Pa
316. er since infinity 1s very far away no reflected wave ever comes back In a symmetric strip line structure the current on the strip won t excite the fundamental mode due to this symmetry However the presence of a feature such as a slot in one of the plates creates an asymmetry The slot will excite the fundamental mode The accuracy of the Momentum results for calibrated ports degrades when the parallel plate modes become important For calibrated ports is assumed that they excite the circuit via the fundamental transmission line mode However this excitation is not pure The source will excite the circuit via the parallel plate modes too This effect cannot be calibrated out since those contributions are not orthogonal The consequence is that the excitation doesn t correspond with a pure fundamental transmission line mode excitation Avoiding Parallel Plate Modes Since we know the field orientation of the parallel plate modes it is easy to understand that you can short circuit them by adding vias to your structure at those places where parallel plate modes will be excited Similar problems will exist with the measurements Both plates must be connected explicitly with each other This won t take place at infinity but somewhere at the side of the board If no vias are used in the real structure the fundamental mode may be excited This mode will propagate to the borders and reflect Thus the real structure differs in that from the s
317. er of Points in the Electromagnetic Simulation Frequencies to 11 3 Click the Recalculate Now button This will add to the previous EMPOWER simulation so that we have 11 instead of 3 data points EMPOWER will intelligently recalculate only the additional points The figure below shows the simulation with 11 EMPOWER data points The notch frequency now appears to be at 9 2 GHz Let s add increase the simulation 31 points to ensure that we get the actual notch frequency Repeat the previous steps to change the number of EMPOWER points to 31 and recalculate EMPOWER Planar 3D EM Analysis T wi m m mJ o1 fi Ca Freg Hz DB 224 DB E11 The display below is after the EMPOWER run with 31 points The response has not changed noticeably since the 11 point simulation so we must have found the correct notch frequency BP Graphi Workspace layonly Iof x DB S21 DB S11 Freg Hz DB 224 DB E11 121 Simulation 122 U sing the Viewer Once the EMPOWER run is completed the viewer can be loaded if Generate Viewer Data was selected in the EMPOWER options dialog Generating this data slows the EMPOWER simulation so it s usually only checked during last run simulations empower Viewer Y6 5 File View XYZ 3 Mag Seiid Freq GH2 3 2 lel 2 a ar Top Front side Oblique 0 086 0 000 Right Click the EMPOWER simulation EM1 in
318. er of rotation This option can also be selected by pressing Page Down Rotate Counter Clockwise PgUp Rotates the current image counter clockwise in the plane of the screen The center of the viewer image window 1s always the center of rotation This option can also be selected by pressing Page Down Pan The objects in this sub menu shift the apparent location of the viewer window relative to the current image Pan Left Ctrl Left Moves the viewer location to the left relative to the current image This moves the image to the right in the viewer window EMPOWER Planar 3D EM Analysis Pan Right Ctrl Right Moves the viewer location to the right relative to the current image This moves the image to the left in the viewer window Pan Up Ctrl Up Moves the viewer location up relative to the current image This moves the image down in the viewer window Pan Down Ctrl Down Moves the viewer location down relative to the current image This moves the image up in the viewer window Pan Zoom In Ctrl PeUp Moves the viewer location closer to the current image This increases the size of the image in the viewer window Pan Zoom Out Ctrl PgDn Moves the viewer location away from the current image This decreases the size of the image in the viewer window Toggle The objects in this sub menu toggle the available options listed below Toggle Absolute Value Display When selected the vie
319. er of spectrums created by Spectrasys These thresholds apply at every calculated node Consequently if a signal is heavily attenuated or outside the given frequency range during a portion of the path and are then amplified or frequency translated back into the given frequency range then these thresholds must be set so that the spectrums will not be ignored along the calculation path Once an individual spectrum is ignored it will not continue to propagate However all spectrums previously calculated will still be available at the nodes where there were within the specified limits For example If we had a 2 GHz transmitter that had an IF frequency of 150 MHz and we set the Ienore Frequency Below limit to 200 MHz then the entire IF signal would not be present and consequently neither would the 2 GHz RF signal Level Below default 200 dBm All spectrums that are below this threshold will not be created or propagated This threshold should be set to the highest acceptable level if faster simulations are important Spectrums are not actually ignored unless they are more than about 20 dB below this threshold since several spectrums can be added together to give a total result that would be greater than this threshold 553 Simulation Simulation Speed Up As with any other type of simulation the more spectral components that need to be processed the longer the simulator time Setting these limits to only calculate the frequencies and amplit
320. er response can be quite complex and require many more points to characterize 126 EMPOWER Planar 3D EM Analysis Real Time Tuning As stated before GENESYS creates ports internal to a layout structure containing lumped elements before invoking EMPOWER During calculation EMPOWER creates S parameter data with port data for all ports whether internal or external This allows GENESYS to tune the lumped elements while still using the original EMPOWER data To see an example of tuning 1 Click inside the Equation C2000 prompt in the Tune Window 2 Type a new value for the capacitor or tune using Page Up Page Down keys or the spin buttons The GENESYS screen below is shown after tuning the capacitors from 0 55 pF to 1 2 pF The response shown on the left in this figure is the linear simulation response The EMPOWER data is combined with the lumped elements in the rightmost response Wi GENESYS 7 0 2 Bl x File Edit View Workspace Achons Tools Synthesis Window Help 0 sa RS m ce RR A RI A az BP Circuit Simulation works Els E3 BB Combined Simulation uu B x Ely Designs o ped F2000 Schema Lavoutl Layout By Simulatianz D ata Si EM1 Layout o 88 Linearl 1400 tc Fy Outputs FRA Circuit Simulation o B Combined Simul zu oe Equations ay Substrates r pn Er Default 30 30 1400 2000 2600 1400 2000 2600 Freq MHz Freq MHz DB S11 amp DB S21 DB S11 127 Simu
321. er than the peak spectrum is treated as a SSB component There may be multiple SSB components along with the peak input spectrum driving one of these 473 Simulation 474 devices Each SSB component is decomposed into its AM and PM counterparts Consequently each harmonic output signal will contain all of the decomposed PM components Obviously without filtering between multiplier and dividers stages the number of spectrums grows rapidly due to the multiplication and SSB to AM and PM decomposition If input SSB spectrum are within 10 dB of the peak input spectrum a warning is given to the user indicating that the decomposition may not be accurate The PM component will change in amplitude according to 20 Log N where N is the multiplication ratio For a divider N M D where M is the harmonic of the divider output and D is the division ratio For example in the multiplier output PM spectrum was 56 dBm as shown above before multiplication the 2nd harmonic PM spectrum would be 6 dB higher 56 dBm 20Log 2 or 50 dBm Likewise the 4th harmonic PM spectrum would be 44 dBm 56 dBm 20Log 4 For a divide by 2 device the divide by 2 PM spectrum would be 62 dBm 56 dBm 20Log 1 2 and the 4th harmonic to the divide by 2 device would have PM spectrum at 50 dBm 56 dBm 20Log 4 2 The following figures and input and output spectrum from a digital divider whose division ratio is 2 and the output power is 5 dBm Input
322. er to Calibration and De embedding of the S parameters The port boundary can be moved into or away from the geometry by specifying a reference offset S parameters will be calculated as if the port were at this position For more information refer to Applying Reference Offsets When two or more single ports are on the same reference plane coupling effects caused by parasitics affects the S parameters The calibration process groups the ports so that any coupling in the calibration arms is included in the S parameter solution For more information refer to Allowing for Coupling Effects If the port is connected to an object on a strip layer the substrate definition must include at least one infinite metal layer a top cover ground plane or a slot layer ot a ground reference must be used in addition to the port For more information on ground references refer to Defining a Ground Reference If the port is connected to an object that is on a slot layer the port has polarity Physical port boundary Calibration arm remeocwves miamatzh To define a single port type open the property dialog and select the type from the combo box Avoiding O verlap Be aware that when using single ports the calibration arm applied to a port may be long enough to ovetlap another element in the circuit In this case the port will be changed to an internal port type and no calibration will be performed on it If this occurs a message Momentum G
323. ered correctly so that signals don t grow in amplitude as they traverse around a loop Once loop spectrums fall below the Ignore Spectrum Level Below threshold the spectrum will stop propagating around the loop Frequency Ranges Spectrasys System Since SPARCA is a continuous frequency simulation technique there is no upper frequency limit As such unnecessary simulation time and data may be taken on spectrums adding no value to the solution of interest Two parameters are used to control which frequencies will be propagated through the simulation engine These are Ignore Spectrum Frequency Below and Ignore Spectrum Frequency Above By default the lowest frequency limit is set to 0 Hz and the upper frequency limit is set to 5 times the highest soutce frequency Controlling Analysis D ata Spectrasys saves data in 1 or more datasets There is a main dataset associated with the system analysis that stores all the node spectral data such as frequencies voltages powets and voltages When paths are defined then a dataset is created for each path The path dataset contains all measurements for the given path Powers voltages and impedances for the path can also be saved to the dataset Spectrasys is a continuous frequency simulator and as such no frequency is outside the bounds of the simulator However since users work in frequency bands of interest the simulator can be speeded up by ignoring frequency bands outside a given window Fur
324. erformance Most electromagnetic simulators include visualization tools The EMPOWER viewer has distinct advantages such as three dimensional graphs true animation capabilities and precise information about current phase The full potential of the EMPOWER viewer is realized with practice so we encourage you to investigate your circuits with the viewer and reflect on the results you observe The viewer is started by selecting Run Viewer from the right click menu of an EMPOWER simulation Workspace Window in GENESYS Interface This section describes the viewer menu items and buttons It can be used to become acquainted with the interface in general as well as as a reference section A sample viewer screen is shown below The objects in this figure are described below A B G hHtlJKLMNOGP amp k 5 lower Viewer V6 5 A File Menu Open Opens a new viewer data file 161 Simulation 162 Exit Exits the viewer Toggle Background Color Toggles the background from black to white or white to black A white background is normally selected before a screen or window print Print Screen Sends a copy of the entire screen to a bitmap file or to a printer Print Window Sends a copy of the viewer window to a bitmap file or to a printer B View Menu The objects in this menu affect how the current image is displayed Top Home Shows a top down view of the current image This option can also be selected by pressing
325. ermod channel power A good example of this would be an amplifier where intermods are cancelled at the amplifier output In this case the generated intermod power alone may be much higher that the total intermod output power Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as GIMCP GAIN and TIMCP Travel Direction Same as GIMCP GAIN and TIMCP Phase Noise Channel Power PNCP This measurement is the integrated phase noise power in the main channel along the specified path Phase noise is displayed on the graphs in dBm Hz and the channel bandwidth is ignored while displaying phase noise However for channel measurements like this one the phase noise is scaled by the channel bandwidth before being integrated Channel Used Main Channel Frequency and Channel Measurement Bandwidth 517 Simulation 518 Types of Spectrums Used ONLY PHASE NOISE Travel Direction Only spectrums traveling in the FORWARD path direction Spurious Free Dynamic Range SFDR This measurement is the spurious free dynamic range along the specified path as shown by SFDR n 2 3 IIP3 n MDS n dB where n stage number The Spurious Free Dyanmic Range is the range between the Minimum Detectable Discernable Signal MDS and the input power which would cause the third order intermods to be equal to the MDS The MDS is the smallest signal that can be detected and will be equivalent to the receiver n
326. eseseacaataes 287 SIP TID ass cis cn ERU oA M AIME 520 SIS T sdaccedit metet M paa 520 SIV D pm M 529 DP A o os INN ME 529 SOM Perici 131 132 SM CEP d cte teintes 145 181 183 SN o S 519 OLD hearer au M NM rt nya 521 BODAS nU M URNA amarante 521 BO eee 530 SOT NEC assieme teli eei debilis senes Ense o REIS 161 170 Solving Convergence 55068 sette 234 SCOPO o PE tet p 521 v9Im d X 521 aee S uL CU DU LERNER Rr er 423 SOV DD einen pna Rd dac river MANU 530 SOV SI edicere a 530 SPARC cni RIED NUI UM 413 Sy Osc EEI E NET 225 SPECIA P AE A 426 Spacal Ora eese isa EEEa aE desose voe sers dM 433 SPE CUR AS YS Ranen a 414 SPECTRASYS Measurement Index 490 SPECTRASYS broadband noise 436 SPECTRASYS channel frequency 424 SPBCTRAS5YS cohbeteticyasn ore 470 SPECTRASYS composite spectrum 431 SPECTRASYS intermods and harmonics 445 SPECTRASYS level dllagtarmsauiee recen 427 SPECIRASYS SOUSEES I oun t tree 423 SPECTRAS YS spectral Otloini secas 433 SPECTRASYS tone channel frequency 502 DCC HUM ATA ZEE ca oeste tine deu 478 Spectrum Analyzer Display uerbis 478 SPICE determining simulator to use 1 Spittal tUe Oto ro epoca 131 143 145 151 spredde ees ciiin diia UM HU EM rRRMEAOI 534 spurious free dynamic range sss 518 Spurious Free Dynamic Range Receiver 518 Spu
327. esh reduction is a technology that automatically removes these redundant degrees of freedom prior to the solution of the problem Hence it should have a negligible impact on the accuracy of the results Momentum GX About the Edge Mesh The edge mesh feature automatically creates a relatively dense mesh pattern of small cells along the edges of metal or slots and a less dense mesh pattern of a few large cells in all other areas of the geometry Because most of the current flow occurs along the edges of slots or metals the edge mesh provides an efficient solution with greater accuracy Use the edge mesh to improve simulation accuracy when solving circuits where the modeling of current flow in any edge area is a critical part of the solution This includes circuits where the characteristic impedance or the propagation constant are critical for determining the electrical model circuits in which close proximity coupling occurs or circuits where edge currents dominate the circuit behavior Applications for using the edge mesh include e Tightly coupled lines e Patch antennas e Resonant circuits e Delay lines e Hairpin filters Edge mesh is available for all levels of meshing global layout layer and object meshing Setting the Width of the Edge Mesh When using an edge mesh you can either specify the edge mesh width or leave it blank If you leave it blank an appropriate width will be determined and used to create the mesh About the T
328. etallization layers No ground reference was specified for the ports so the sphere at infinity acts as the ground reference for the internal ports The S parameters obtained from a simulation for these ports no longer have a physical meaning Using such ports in the simulation yields incorrect simulation results as shown in Figute 8 qubstrate layer stack metallization layer Ls 3 inches air p 1 0 W z2inches FR H 58 mil Er 4 4 air Ep 1 0 finite ground plane mot taken up In the Byer stack Figure 7 PCB substrate layer stack and metallization layers 383 Simulation n 150 i0 140 100 29 P BD 10 g r4 wa O a9 pen b IF m Jl p9 E e 60 Rh i i ii i dg ag 189 7 0 0 5 1 9 1 5 7 0 2 5 4 d e g s 1 4 1 6 zu zl eu reg GHz Treg GHr Figure 8 Magnitude and phase of S21 The thicker line is Momentum results the thinner line is the measurement Finite Ground Plane Internal Ports in the Ground Plane One way to define the proper grounding for the internal ports is to add two extra internal ports in the ground plane Figure 9 The resulting four port structure is simulated with Momentum Note that because no ground reference is specified for the internal ports the resulting S parameters for the four port structure have no physical meaning Howevet by properly recombining the four ports a two port structure is obtained with the correct ground references for each of the ports Care
329. ev a tis a padded tuis axis ee totos tutu DUE soup oM 158 Hisher Order Box Modes etico te ete t ded aie d e doen d Spe tete bee teen 159 Partial Diclecric todito e E teil ss iistavna vanes Ueavia T ct leeuiadeatin 160 Biota ictal E TS6ES aua i eet P eti M ate me UR dea Htul E db ratu PA eR e Dt IUufe 160 JOD OV CE qn qu bei on odit ltissim ei esl vb edd eus 160 CAVA DSOL DET uc esiti iaa Sues ada aU s E Mira aUe riga dust aucune 160 EMPOWER Viewer and Antenna Patterns e epo aaa btesii etaient 161 EEC T 161 MACS TAC ER 161 Pat Field Riddo P aneii NIEN Cx ouest aede mutas ro ert i tet ern avenir eds 166 DExatmplesstedestossettuust EET Linee EEA 170 Moh Mode V es ox D AU seco de abd en ade taut ide da dco UA DU Des ne ee AE 174 Via Hole Vie radici 175 hu Ni Mt 177 SETI RT T M 178 EMPOWER Bile Deseo sists dixcon tps tous ep enn ta ette eus isse tex O ea pud 179 orca ll EE 1 79 Text T les Eu Marie dE m lsts cat saces i rast N mate nadeouatict 179 de EXCISION e tate nmin ren rane tray Tener aye etree Trem ercere errr te 180 EM V EMPOWER Viewer ILS adco bediedteae denied betae 180 Ed 2 s We DAV Data TIE S 2h ys cco ander na pen esaet NR 181 Jud Bits ato md PHeS eoo dbi to Bb epa don n da dena nU Utd Dad o Ute 181 PLEX Gurren Viewer Data Files cipue dicet M Menon i aD Nie M S Eq o sU d ME 182 Axle Rea En Post Inpedane
330. evels in dBc Hz Phase noise can also be specified single or double sided If double sided negative offset frequencies must be used NOTE There must be the same number of entties in both frequency and amplitudes lists If not a warning will be issued and the lists will be truncated to the smaller of the two lists The following example shows the single sided specification of an oscillator In this example phase noise is specified for 70 90 100 105 and 110 dBc Hz for the respective offset frequencies of 1 10 100 1000 10000 kHz The carrier center frequency is 1530 MHz 476 Spectrasys System Osc F 1530MHz Foff21 10 100 1000 10000KHz PhaseN 0 90 100 105 110dB10 The list data does not need to occur in ascending frequency order though this is a more likely readable format However the first frequency entry will be associated with the first phase noise power level entry the 2nd frequency entry with the 2nd phase noise power level entry etc etc Enabling Phase Noise Phase noise must be enable in two places before phase noise spectrums will propagate through the analysis e Inthe source itself see the phase noise enable parameter e Inthe system analysis Phase Noise Simulation Each phase noise spectrum is associated with a parent signal spectrum The phase noise goes through the same transfer function as its parent Through all linear models the phase noise will be transformed as is the signal spectr
331. f this field is empty the system analysis can determine the path frequency if there is a single signal source on the beginning node of the path For multiple frequencies the system analysis doesn t know the intent of the user and will display an error For intermod measurements this is the frequency of the intermod to be measuted The channel frequency will automatically change along the path through frequency translation devices such as mixer multipliers dividers etc If the user would like to force the channel frequency to a particular value at a given node along the path this can be done by adding an additional frequency to the channel frequency list by separating each frequency by commas The channel frequency and path list are paired beginning at the start of the path So the second channel frequency in the list will correspond to the second node or part in the path list For example if the user wanted to track the 2nd harmonic created at the output of the LNA and the path is specified as 1 5 2 where node 1 is the input 5 the LNA output and 2 the system output If the input frequency was 1000 MHz then the channel frequency would be specified as 1000 2000 What this means is that at node 1 the 1000 MHz frequency will be used When the path calculation reaches the next name in the path list node 5 the next channel frequency will be used which is 2000 MHz The channel from this 543 Simulation 544 point to the end of the path a
332. f vectors of size nFin 2 236 HARBEC Harmonic Balance Analysis B x FregindexiM 2 2 c Data ame 0 ay Convloss z3 997 PrfdBm Pif La Say Freq Ta FreqiD 1 FL SP1 PPORT 1II sSP22 PPORT Z S amp P 32 PPORT 3 ify Pifa 23 997 hb getspcompdbm P3 FregindexIM 1 1 HP PPORT A VPORT EAW VPORT EX ZPORT TRURTRURURU RUE AAN lellelelalelelelole Bislsls ls e ell e o Jl ke ale zl pa ale me oo ne S oe eo co na S fs ols 7 l i Figure 2 PPORT array nPorts of spectrums of the port impedance dissipated powers Units dBm VPORT array nPorts of spectrums of port voltages Units V ZPORT array nFout of port complex impedances at HB solution frequencies Units Ohm The next variables are created in the dataset only after setting one of the checkboxes defining calculation waveforms of the HB solution see Calculate tab of HB analysis properties window Time independent variable array n TimePoints of time points for creating waveforms for the HB solution Units s W VPORT port voltage waves solutions transformed to the time domain Units V 237 Simulation 238 The Next variables are created in the dataset only after setting the checkbox Save solution for all nodes in the Calculate tab of the HB analysis properties window V node name voltage spectrum at the node Units V
333. fect on memory and time requirements Be sure that your metal layers are set to lossless if you check the slot type structure box EMPOWER Planar 3D EM Analysis Preferred Cell Count The first part of an EMPOWER run involves taking Fourier Transforms of the grid These transforms will run much faster if the number of cells along the each side of the box is of the form 22355741 1 13f where e and f are either zero ot one and a b c and d are arbitrary integers In other words a circuit with a box 512 cells by 512 cells 28 by 28 will analyze much faster than a circuit with a box 509 cells by 509 cells 509 is prime Making one side a preferred number will help so a box 509 x 512 cells is better than one 509 x 509 cells Note that only the time while EMPOWER is working on the Fourier Transform is affected and this is normally only substantial with boxes 100x100 or larger If you see a status with FFT in the message for a long time check to see that the box width and height are a preferred number of cells across Preferred numbers which fit the form given above 10000 and below are 1234567891011 12 13 14 15 16 18 20 21 22 24 25 26 27 28 30 32 33 35 36 39 40 42 44 45 48 49 50 52 54 55 56 60 63 64 65 66 70 72 75 77 78 80 81 84 88 90 91 96 98 99 100 104 105 108 110 112 117 120 125 126 128 130 132 135 140 143 144 147 150 154 156 160 162 165 168 175 176 180 182 189 192 195 196 198 200 208 210 216 220 224 225 231 234 240 243 245 250 252 256 2
334. ference plane shift is positive then the data is measured from outside the box effectively adding length to the circuit RefShift RefShift CIRCUIT X This is the equivalent network used when deembedding is active The center of the figure labeled CIRCUIT contains the raw results from the EMPOWER simulation Reactance X shown as inductors above cancels the capacitance caused by the end wall as well as correcting other reactances The value of X may be negative and it is frequency dependent The RefShift lines at the outside move the reference planes to the correct location Since the RefShift lines also help to correct for the discontinuity at the box wall their lengths are normally not zero even if the reference shift specified for the port is zero The impedance of the RefShift lines is equal to the port line impedance so only the EMPOWER Planar 3D EM Analysis phase is shifted by the addition of these lines The magnitude of the reflection coefficients is not affected The parameters for deembedding are calculated prior to the analysis of the circuit EMPOWER does this automatically by analyzing two different length lines at each frequency for each port used solving for the reactance and the base RefShift value Note Deembedding requires an additional line analysis mode at the start of the run so tuns using deembedding can take substantially longer This is especially true if the lines at the ports are wide
335. ferent authors The resultant system of differential difference equations approximates the initial EMPOWER Planar 3D EM Analysis system with the second order locally inside a layer The initial boundary value problem can contain infinitesimally thin metal regions with consequent singularities of the field and conductivity currents at the metal edges Meixner 1972 That is why a global approximation order of the problem is usually lower and the largest calculation error part for integral parameters of a structure Y S matrix elements characteristic impedance decreases usually proportionally to the grid cell size That is the monotonic convergence was observed for almost all problems solved on the initial equidistant grid This makes it possible to use such powerful convergence acceleration techniques as Richardson s extrapolation Richardson 1927 Marchuk Shaidurov 1979 Note that this is an observation and it cannot be proven to work for all problems The technique used here for the descriptor matrix evaluation using current sources in the metal plane is empirical The evaluation accuracy depends on parasitic high order modes that could be excited by current sources and if they are close to their cutoffs or even are propagating the estimated descriptor matrix could be far away from the correct one This can be expected however since real circuits which have unexpected high order modes near the cutoff usually do not work properly either
336. frequency in the Mesh Frequency field and select the units The wavelength of this frequency will be used to determine the density of the mesh In general set the value of the mesh frequency to the highest frequency that will be simulated For more information refer to Adjusting Mesh Density 3 Enter the Number of Cells per Wavelength This value will also be used to determine the density of the mesh The relationship between wavelength and cells per wavelength can be described by the following example If the circuit is 3 wavelengths long and the number of cells per wavelength is 20 348 Momentum GX the length of the circuit will be divided into 60 cells For more information refer to Adjusting Mesh Density 4 Enter the arc resolution Any curved areas in the circuit will be meshed using facets In the Arc Facet Angle field enter the number of degrees that will be included in a single facet The maximum is 45 degrees per facet The lower the value the better the resolution and the denser the mesh will be The minimum value is equal to the Arc Circle Resolution which is set when the object is drawn For more information refer to Processing Object Overlap Arc Facet Angle la 30 degrees Arc Facet Angle is 18 degrees 5 Enter Layout Mesh resolution Any 2 points that differ less than this value will be considered identical 6 Enable Transmission Line Mesh to specify the number of cells along the width of a geometry It is most use
337. frequency range substantially Co Simulation Sweep Specifies the frequencies at which to run simulate the lumped elements EMPOWER data combination If you have no lumped elements in your simulation you should normally check the Use EM Simulation Frequencies box For circuits with lumped elements you can often save much time by using fewer points in the EMPOWER Planar 3D EM Analysis electromagnetic simulation frequencies above allowing the co simulation to interpolate the EMPOWER data before the lumped elements are added Turn off physical losses Faster If checked EMPOWER will ignore any losses specified in the EMPOWER Layer tab This option is very useful to speed up any preliminary runs Automatically save workspace after calc This checkbox is handy for overnight runs to help protect against a power outage Note that checking this box will force the entre workspace to be saved after each run DOOR TEN EMPOWER Options o M ES General Wiewer Far Field Advanced Generate Viewer Data lower Fart number to excite Mode number ta excite Generate Far Field Radiation Data Sweep Theta Start Angle fo Stop 2o Step degrees Sweep Phi Start Angle fo Stop 2o Step degrees UF Cancel Apply Help Viewer Far Field Tab Generate Viewer Data Slower Checking this box causes EMPOWER to generate a EMV file that can be loaded in the EMPOWER c
338. from the device driving the output port This noise is obviously flowing toward the output port The output port itself will also generate noise which will flow from the output port back towards the input Impedances that these spectrums see may be different for every direction through the node For this reason each of these total noise spectrum are displayed on a spectral plot Spectrasys uses complicated noise correlations matrices along with other special noise simulation techniques to be able to be able to propagate noise spectrums especially through multiport devices The individual noise spectrums by default are not shown on spectral plots However user can view this information if so desired There are times when debugging noise problems in an RF architecture that this information is extremely helpful The following figure illustrates multidirectional noise between an amplifier and output port The amplifier has a gain of 20 dB and a noise figure of 3 dB The measurement bandwidth is 1 Hz Spectrasys System Amplifier Output Spectrum 20 4D 60 3 E 80 5 100 5 120 140 160 Frequency GHz As can be seen from the figure the noise from the output port is thermal noise whereas the noise power from the amplifier output is thermal noise plus the amplifier gain of 20 dB plus an additional 3 dB for the amplifier noise figure Click here for additional information on noise analysis It is not the purpos
339. ful for circuits with straight line geometry Enter the number of cells that the width will be divided into in the Number of Cells Wide field This will be the total number of cells along the width of the circuit For more information about the transmission mesh refer to About the Transmission Line Mesh 7 Enable Edge Mesh to add a relatively dense mesh along the edges of objects Since most current flows along the edges of objects the edge mesh can improve the accuracy and speed of a simulation For more information about the edge mesh refer to About the Edge Mesh An edge mesh that is specified with a width larger than the cell size set by the wavelength number_of_cells_wavelength will be ignored This i is because such edge meshes would be very inefficient However if these edge mesh values must be used you can decrease the number_of_cells_wavelength which is specified in the Number of Cells per Wavelength field 8 Enable Thin layer overlap extraction in order to extract objects for the following situations e Two objects on different layers overlap e The objects are separated with a thin substrate layer If this is enabled the geometry will be altered to produce a more 349 Simulation accurate model for the overlap region For more information refer to Processing Object Overlap Si a i This should always be enabled when modeling thin layer capacitors 9 Enable Mesh reduction in order to obtain an opt
340. g multidimensional FFTs If the frequencies are evenly spaced have a large common factor it may be faster to use a one dimensional FFT On some occasions convergence can also be affected Automatic Recalculation Checking this box will cause the harmonic balance simulation to be run any time there is a change in the design If the box is not checked the simulation must be run manually either by right clicking on the simulation icon and selecting Recalculate Now or by clicking the recalculation calculator button on the main tool bat Auto save Workspace After Calculation Checking this box will cause GENESYS to save the current workspace after the simulation 1s complete This is particularly useful with long simulations or simulations that run overnight If this box 1s checked when optimizing the file will be saved after each optimization step Use Previous Solution As Starting Point Usually checked this option will start the convergence process using the previous set of node voltages If the parameters changed or swept are relatively small starting with the previous solution can dramatically speed convergence If the parameters changed are large is sometimes better to start from scratch Certain circuits will always converge faster from scratch than previous solutions Calculate wave data The simulation calculate time domain waveforms for all circuit nodes and branches if checkbox Save solution for all nodes is checked on
341. ge measurements of the first mixer NOTE For measurements that require a channel only spectrum falling within the channel will be integrated If the path center frequency is set incorrectly or the bandwidth is set too large or small measurement values may be different than expected Mathematical integration is used and is precise If a spectrum is split by the channel bandwidth then only that portion of the spectrum that falls within the channel will be measured Of course spectrum plots will show all spectrum regardless of whether they are in the channel or not Channel Bandwidth Caution When the Channel Frequency is less than 1 2 the Channel Bandwidth the lowest integration frequency used for measurements will be 0 Hz This will result is Channel Noise Power measurements being different than when the full bandwidth is used Specifying Paths Spectrasys supports multiple paths through arbitrary architectures Paths are not restricted to traditional 2 port cascaded lineups A path consists of 1 Name 2 Beginning Node 3 Ending Node 4 Frequency Spectrasys will find the shortest path between the specified nodes If the user would like to select an alternate path then a thru node can be added to the path to uniquely identify the path Through nodes can added until the path is uniquely identified For example Spectrasys System LO ConvGain bdbl PORT 2 F 1000IMHz as Part_3 PAC T3dBm LO 10dBirn PO
342. ge sources Transducer Not implemented as a function It s calculated using equations Example Transducer Gain from port 1 to port 2 from frequency of the 1 harmonic to 3 harmonic GainP dbm P2 3 dbm P1 1 Gain hb_transgain SpectrIn Spect rOut FregIndexIM IndexIn Index Out hb transgaindb SpectrPin SpectrPout FreqIndexIM IndexIn Inde xOut hb_gain In SpectrOut Freq IndexIM IndexS Example a Transducer Gain from port 1 to port 2 from frequency of the 1 harmonic to 3 harmonic GainP 247 Simulation hb_transgaindb P1 P2 Fr eqIndexIM 1 3 b Conversion gain from input port voltage base tone frequency to output port voltage IF frequency Fif 1 F1 1 F2 ConvGainV hb transgai n VPORTY 1 VPORT 2 FreqIndexIM 1 0 1 1 c Conversion gain from input port voltage base tone frequency to current of current probe CP1 for IF frequency Fif 1 F1 1 F2 ConvGainV hb transgai n VPORT 1 ICP1 Iprob e FregIndexIM 1 0 1 1 c Conversion gain from available input port power Pin in Watts to 3 harmonic of output port power D ConvGain3av hb_gain Pin P2 FreqInde xIM 3 Intercept point of 2 14 spectral components IPn Not implemented as a function It s calculated using equations Example Output IP3 TOD for 2 tones signal F1 1959 5M Hz F2 1960 5MHz HB analysis between spectral components of
343. ged with the fitting model This can be checked by the messages in the status window If converged the message will be that AFS is complete If not converged a warning will appear Examples of such warning messages are Maximum number of adaptive frequency samples reached Fitting models not completely converged Consider increasing maximum number of sample points activating the reuse option If you encounter errors try decreasing the frequency range of the frequency plan and increasing the number of sample points Theoretically some circuits may require several dozen points to converge if they are electrically long This is because the AFS program samples the data at about each 60 degree rotation in the Smith chart of the S parameter with the most phase variation Momentum GX If you simulation does not converge to help you assess the quality of the simulation you can view convergence results in the Simulation Log window Eronen E Extracting layout Adaptive Frequency sweep started T RF mode amp amp PORTZ 50 Simulation Frequency 1 1 625 GHz loading Green Functions _ loading quasi static matrix rmn FER GGG k Li p IMMENSE LI IMMENSE i FER GRO GG J solving interaction matrix Solver selection configuration setting For direct solver calibrating ports Setting Sample Points If you are using the Adaptive sweep type in a simulation you may want to select specific sample points to be u
344. h of a vector from the center of the chart with 0 length being a perfect match to the reference impedance and 1 being total reflection at the circumference of the chart The underlying grids of the Smith chart are circles of a given resistance and arcs of impedance The reflection coefficient radius of the standard Smith chart is unity Compressed Smith charts with a radius greater than 1 and expanded charts with a radius less than 1 are available High impedances are located on the right portion of the chart low impedances on the left portion inductive reactance in the upper half and capacitive reactance in the lower half Real impedances are on a line from the left to right and purely reactive impedances are on the circumference The angle of the reflection coefficient is measured with respect to the Linear Analysis real axis with zero degrees to the right of the center 90 straight up and 90 straight down The impedance of a load as viewed through an increasing length of lossless transmission line or through a fixed length with increasing frequency rotates in a clockwise direction with constant radius when the line impedance equals the reference impedance If the line and reference impedances are not equal the center of rotation is not about the center of the chart One complete rotation occurs when the electrical length of the line increases by 180 Transmission line loss causes the reflection coefficient to spiral inward
345. h one incident wave at one input The excitation conditions are passed to EMPOWER in the command line when running EMPOWER text files When EMPOWER is launched from GENESYS the excitation conditions are automatically defined from the EMPOWER Setup dialog box when the Generate Viewer Data check box is active If Generate Viewer Data is selected the default incident wave is the first eigenwave of the first input The input number can be changed in the Port number to excite box of the EMPOWER Setup dialog and the input mode number can be changed in the Port mode to excite box The control information about what input and what mode are actually excited in the structure is printed out in the listing file see PPLT Input _ mode will be incident in the listing file An output binary file with the extension EMV is created by EMPOWER to pass data to the viewer program In a GENESYS Workspace the internal name of this file is EMPOWER EMV An optional self documented ASCII data file with extension PLX can also be written for import into other programs To understand the viewer a review of EMPOWER input and mode representations is helpful A circuit can have external and or internal inputs External inputs are transformed to eigenmode space de embedded and normalized to characteristic impedances of eigenmodes They could be one mode or multimode modally coupled and the incident wave for these inputs can be one of the input eigenmodes The inci
346. han 200 channels away from the center frequency Attenuation 200 channels from the center will be about 150 dBc Randomize Noise When enabled random noise will be added around the resulting analyzer sweep The output trace will be more representative of a typical spectrum analyzer at the expense of additional computation time Add Analyzer Noise All spectrum analyzers have a limited dynamic range They are typically limited on the upper end by intermods and spurious performance at an internal mixer output On the lower end they are limited by noise of the analyzer This noise is a function of the internal architecture of the specific spectrum analyzer and internal RF attenuator Analyzer Noise Floor Specifies the noise floor in dBm Hz of the spectrum analyzer mode Limit Frequencies When checked the frequency range of the analyzer mode is limited By default the entire spectrum from the Ignore Spectrum Frequency Below lower frequency limit to the highest frequency limit of Ignore Spectrum Frequency Above will be processed by the analyzer for every node in the system In some cases this may be very time consuming In order to improve the simulation speed and just process the area of interest frequency limits can be enabled to restrict the computation range of the analyzer Start Beginning frequency of the analyzer Stop Ending frequency of the analyzer Step This is the frequency step size between analyzer data point
347. hannel signals This is ok but extra noise points may need to be added to improve the accuracy of the Channel Noise Power measurement Spectrasys interpolates between all noise and signal data points If there is a lot of amplitude ripple in the circuit sufficient noise points must be added for each signal to properly account for these variations If the noise spectrum looks very stick figure ish then extra noise points may need to be added If cascaded noise figure is being examined through a hybrid combining network the cascaded noise figure will appear to artificially peak at the internal nodes due to the hybrid network This occurs because the cascaded gain used is only for the current path and not all parallel paths used in the hybrid network See Broadband Noise Cascaded Noise Analysis and Cascaded Noise Figure Equations for additional information W hy don t I get the same answer as my spreadsheets Spectrasys accounts for VSWR between stages sneak paths reverse isolation frequency response channel bandwidth gain compression broadband noise and image noise Spreadsheets do not Cascaded noise figure equations assume no image noise and perfect matching between stages Cascaded intermod equations assume no frequency rolloff for the interfering tones This can be a bad assumption especially for a receiver blocking test In order to correlate Spectrasys data with a spreadsheet or other math packages or programs the schematic mu
348. has the typical flat white noise floor Mixer Noise Simulation This section focuses on setting up and using noise simulations for mixers It describes how to calculate Noise analysis frequency translation of the noise HARBEC Harmonic Balance Analysis Nonlinear spot noise Nonlinear swept noise Determining Mixer Noise Here we illustrate an example fot determining mixer noise The noise figure computed by a harmonic balance noise simulation is a single sideband noise figure Note Port sources and an output termination port component are necessary only if a noise figure simulation is being performed The model used in the current design is a 2 port model and allows controlling noise the input and output ports Noise of other ports of the circuit for example noise of LO port are taken into account automatically as noise of other circuit components To determine mixer noise set the analysis temperature and Noise tab parameters On General tab of Harbec options set simulation Temperature Note In nonlinear noise analyses we recommend setting simulation temperature equal to 16 85 C or 290 K which is the standard temperature for noise figure measurement as defined by the IEEE definition fot noise figuration On the Noise tab of Harbec options define input and output ports and set the Noisy checkbox for Input and clear it for Output ports Define noise frequency sweep parameters For most applications
349. he Optimal Admittance for Noise is a complex function of frequency and is available for 2 port networks only The optimal admittance is the value of the input admittance which minimized the noise fioure of the network The optimal admittance is defined in terms of the source admittance Ys and the noise resistance Rn and the noise figures NF NFMIN as NF NFMIN Rn Re Ys Ys Yorr The optimal impedance is the inverse of the optimal admittance i e Zopr 1 Yorr Values Complex value versus frequency Simulations Linear Default Format Table RECT Graph RE Smith Chart GOPT Examples Measurement Result in graph Smith chart Result on table optimization or yield YOPT real part of optimal admittance real imaginary parts of admittance Effective Noise Input Temperature NFT The Effective Noise Input Temperature is a real function of frequency and is available for 2 port networks only The effective noise temperature is defined in terms of the noise figure NF and a standard temperature To in degrees Kelvin as NFT To NF 1 where To 300 degrees Kelvin Values Real value versus frequency Simulations Linear Default Format Table Linear Graph Linear Smith Chart none Commonly Used Operators none Examples 27 Simulation 28 Measurement Result in graph Smith chart Result on table optimization or yield NFT noise temperature in degrees Kelvin noise temperature in degrees Kelvin
350. he center of the pad footprint and EMPOWER writes data for each port created whether internal or external e The 1 and 2 ports pictured in the figure above are examples of external potts This is a powerful technique since real time tuning can be employed in GENESYS once the EMPOWER data for has been calculated Simulating the Layout with Lumped Elements This workspace has an EMPOWER simulation but does not contain a Sonnet Simulation so we will add one now Click on the Create New icon in the Workspace Window and select Analysis Add Sonnet Simulation Accept Sonnet1 as the analysis name The Sonnet Options dialog will be displayed as shown below For a description of the dialog options see the details in the Sonnet Translation chapter For now just set the prompts as shown below 409 Simulation 410 sonnet Interface Options General Sonnet Advanced Layout to simulate Layouti Pork Impedance 50 Electromagnetic simulation frequencies Start freq MHz 1400 Stop Freq MHz 2600 Humber of points 4 _ Use Adaptive Sweep ABS Automatically save workspace after calculate Use ports From schematic necessary For Harbec ca simulation Co simulation sweep Use EM simulation Frequencies Start freg MHz 1400 Stop freq MHz 2600 Mumber of points 101 Recalculate Mow Click the Recalculate
351. he first mixer would be 1700 MHz If the receiver bandwidth was 5 MHz then the image channel would be from 1697 5 to 1702 5 MHz This measurement is simply a Channel Noise Powet measurement at the Image Frequency Channel Used Image Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used ONLY NOISE Travel Direction Only spectrums traveling in the FORWARD path direction Image Channel Power IMGP This measurement is the image channel power from the path input to the fitst mixer After the first mixer the this measurement will show the same power and the main channel powet For example if we designed a 2 GHz receiver that had an IF frequency of 150 MHz using low LO side injection then the LO frequency would be 1850 MHz and image frequency for all stages from the input to the first mixer would be 1700 MHz If the receiver bandwidth was 5 MHz then the image channel would be from 1697 5 to 1702 5 MHz All 511 Simulation 512 noise and interference must be rejected in this channel to maintain the sensitivity and performance of the receiver This measurement is simply a Channel Power measurement at the Image Frequency Channel Used Image Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as CP Travel Direction Same as CP Image Noise Rejection Ratio IMGNR This measurement is the ratio of the Channel Noise Power to Image Channel Noise Power along the specifie
352. he insertion loss parameter is used When a stage doesn t have either a noise figure or insertion loss parameter 0 dB is used This measurement is not dependent on the path direction through the model For example if the path was defined through the coupled port of a coupler the insertion loss of the coupler would be reported and not the coupled loss Channel Used No channel is used for this measurement 519 Simulation 520 Types of Spectrums Used None Travel Direction N A Stage Input Intercept All Orders SIIP This measurement is the stage input intercept point calculated by using the Stage Output Intercept Point and the Stage Gain When a stage doesn t have this parameter 100 dBm is used SIIP n SOIP n SGAIN n dBm where n stage number Each column in this measurement is for a different intermod order up to the Maximum Order specified on the Calculate Tab of the System Analysis Dialog Box The column number is the same as the order starting from the left with order 0 Channel Used No channel is used for this measurement Types of Spectrums Used None Travel Direction N A Stage Input 1 dB Compression Point SIPTIDB This measurement is the stage 1 dB compression point calculated by using the Stage Output 1 dB Compression Point and the Stage Gain When a stage doesn t have this parameter 100 dBm is used SIPIDB n SOP1DB n SGAIN n dBm where n stage number Channel Used No channel is used for
353. he value in the Associate with port number field is the same for additional ports For example if you were associating three ports and the first port was assigned as port 1 for the second and third port the value entered into the Associate with port number field would be 1 For the first port you choose no value is entered in this field 8 Click OK 9 Repeat these steps for other common mode ports in the circuit For example we created 2 groups of common mode ports 1 6 and 7 12 in this layout 327 Simulation BEES EM Port Properties C 7 Draw Size Ref Plane Shift Pork Number Ea PartList Ez Sc Layer IB TOF METAL v Current Direction Default Width oo Line Direction oT Height Pork Type Common Associate with port number Polarity Mormal Common mode STRIP port transmission line extended calibration It s not necessary to define the association with other ports for the first defined port in a group Other ports of a port group must have association to the first port of its group or be associated to other ports of the group It does not matter how grouped ports are associated just that they are associated to other ports of the group For example ports 1 6 may be associated in a group by defining Associate with port number 1 2s 328 Momentum GX 3 gt 1 4 gt 1 5 gt 1 6 gt 1 or using any other possible combination without closed loops 1
354. hecked not only at the amplifier operating frequencies but also over the entire frequency range for which S Parameter data is available Matching One definition of network gain is the transducer power gain Gt Transducer power gain is the power delivered to the load divided by the power available from the soutce Gt P delivered to load P available from source Other gain definitions include the power gain Gp and the available power gain Ga Gp P delivered to load P input to network Ga P available from network P available from source The S parameter data for the network is measured with a source and load equal to the reference impedance If the network is not terminated in the reference impedance Gt can be computed from the reflection coefficients of the terminations on the network and the S parameters of the network At this point we have multiple sets of reflection coefficients those of the terminations and S11 and S22 of the network To avoid confusion the termination reflection coefficients are given a different symbol G The transducer power gain with the network inserted in a system with arbitrary source and load reflection coefficients 1s 4 Gt 2 1 R 3 1 R1 3 1 SuRs 1 S2 Rz SziSizRzRs where Rs reflection coefficient of the source Ri reflection coefficient of the load If and are both zero then Gt S21 or Gt dB 20log S21 S21 dB Therefore when a network is installe
355. heel a Partlist User Ground Layer Default 0 v Don t Create Mask 2 Any layout object line polygon circle pad etc drawn in a metal layer with Z directed current expanded from the 1st metal layer in the direction of the current to the next metal layer or metal cover creates a via This is an EMPOWER type via For example to create a via between the 2 metal layers Metall and Metal2 we add a new metal layer Via under metal layer Metall and define its current direction as XYZ down down to the substrate stack Any layout object drawn on the Via layer creates a via between the Metall and Metal2 layers The via model may be defined in the substrate layer located between the 2 metal layers A Defau t via model means use the model defined by the Momentum Analysis 335 Simulation LAYOUT Properties General associations Layer Fans Show Columns v Metal Cl Substrate Cl General v Layer Humber and Color Momentum Sot Type Strip Current Thick Me Direction bos Lossless EYI Dene w Thin Lossess Phiesical with Tand Lixesless Physical with Tand KYI Dose gt Frequency Units MHz Load From Layer File Length Units mil id Soweto Layer File Lo J eme For example in this layout we created the via by drawing a polygon on the Via layer Polygon Properties 3 Any holes drilled create a round via between the metal
356. ic of the layered medium acts as the integral kernel The unknown surface currents are discretized by meshing the planar metallization patterns and applying an expansion in a finite number of subsectional basis functions Bi t BN t iv Jr gt I B An Equation 2 The standard basis functions used in planar EM simulators are the subsectional rooftop functions defined over the rectangular triangular and polygonal cells in the mesh Each rooftop is associated with one edge of the mesh and represents a current with constant density flowing through that edge The unknown amplitudes Ij j 1 N of the basis function expansion determine the currents flowing through all edges of the mesh Figure 1 Discretization of the surface currents using tooftop basis functions The integral equation 1 is discretized by inserting the rooftop expansion 2 of the currents By applying the Galerkin testing procedure that is by testing the integral equation using test functions identical to the basis functions the continuous integral equation 1 is transformed into a discrete matrix equation for i 1 N Equation 3 with i Z 4SBin fas Gir rj Bir E E E Equation 4 375 Simulation 376 v asBjn En S Equation 5 The left hand side matrix Z 1s called the interaction matrix as each element in this matrix describes the electromagnetic interaction between two rooftop basis functions The dimension N of Z is
357. ideration for filtered tones that generate the intermods 4 Multiple paths are ignored only valid for two pott lineups 5 Gain is assumed to be independent of power level 6 Intermods never travel backwards reverse isolation is assumed to be infinite The calculation of cascaded intermods is generally in a spreadsheet Note there is no relationship between these calculations and the physical measurements of the intercept points in the lab There is no mention of frequencies and power levels of tones that are need to make measurements in the lab Because of these serious restrictions new intermod measurements were created to eliminate these issues See Intermod Path Measurement Basics for additional information 451 Simulation 452 Intercept Measurements in the Lab Intercept measurements in the lab are broken down into two groups in band and out of band In band measurements are used when tones are not attenuated by filtering through the cascade For example intercept point for a power amplifier is generally done with 2 tones that exhibit the same power throughout the system Out of band measurements are used when they are attenuated like filtering in an Intermediate Frequency IF In Band Intercept Measurements For in band measurements two tones f1 and f2 are created by two signal generators and combined before entering the Device Under Test DUT Care needs to be taken in the setup to ensure reverse intermods will not be g
358. idth amp Length When placing an external port on the end of a strip type transmission line you should normally leave these at zero so that LAYOUT sizes the port automatically If you want to override the size or for slot type or internal ports you can specify width and length here Note Width and length are measured relative to the line direction so these parameters can appear to be reversed Length is the length in the direction of propagation along the line and width 1s the width of the strip Layer Specifies the metal layer on which the port is placed Location specifies the edge of the port for external ports and the center of the port for internal ports Line Direction Gives the direction of the line at the port In the default mode the nearest wall determines the direction of the line This value rarely needs to be overridden Current Dir Specifies the direction of current flow within the port The first figure below shows the default current direction for external ports on strip type structures such as microstrip and stripline The second figure shows the default current direction for external ports on slot type structures such as coplanar waveguide For internal ports the default current direction is Along Z This value also rarely needs to overridden stip conductors A Along Y B Along X 137 Simulation 138 Port Type Specifies the basic type of port Normal No Deembed and Internal e
359. idth was specified to 03 MHz and the Channel Frequency was 220 MHz then the CV is the average voltage from 219 985 to 220 015 MHz Default Unit dBV Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used A SIGNALS INTERMODS HARMONICS NOISE and PHASE NOISE Travel Direction All directions through the node Channel Noise Voltage CNV This measurement is the average noise voltage in the main channel along the specified path For example if the Channel Measurement Bandwidth was specified to 100 kHz and the Channel Frequency was 2000 MHz then the CNV is the average integrated noise voltage from 1999 95 to 2000 05 MHz 525 Simulation 526 Default Unit dBV Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used ONLY NOISE Travel Direction Only spectrums traveling in the FORWARD path direction Desired Channel Voltage DCV This measurement is the desired average voltage across the main channel along the specified path This measurement includes ONLY DESIRED SIGNALS on the beginning node of the path traveling in FORWARD path direction All other intermods harmonics noise and phase noise signals are ignored Note A D is placed next to the equation in the identifying flyover help in a spectrum plot to indicate desired signals For example if the Channel Measurement Bandwidth was specified to 03 MHz and the
360. idually in the following sections Note Unfortunately Windows can be setup to hide files extensions as well as actual files from the user We would recommend that you turn off this feature Double click on My Computer Select Options from the View menu Click the Viewer tab Click Show all files Deselect Hide MS DOS file extensions for file types that are registered and click OK Different versions of Windows may have slightly different procedures L1 L2 etc Port deembedding and line data for port 1 2 etc Listing file summarizing all EMPOWER data PLX Text Text listing of viewer currents Port deembedding and line data for a port with a user specified deembedding file name a rears sa oENES Wem Kico ince S Parameter results Netlist for EMPOWER Binary Y Parameter results Backup All files with either a name or an extension starting with tilde are backup files and can be safely deleted EMV EMPOWER Viewer Files Written by EMPOWER Type Binary 180 EMPOWER Planar 3D EM Analysis Can be safely edited No Average size 10 to 100Kbytes but may be larger Use Data for viewing currents or voltages EMV files EMPOWER Viewer files are completely self contained files containing all information needed by the viewer to display currents and voltages for a circuit These files contain information about the box and the grid mapping of the circuit as well as actual complex curren
361. ignals are decomposed into AM and PM components Non linear devices like digital dividers frequency multipliers and frequency dividers will treat all input spectrum other than the peak as single sideband input spectrum which can be decomposed into its AM and PM counterparts These AM and PM counterparts are then processed as well as the peak spectrum by the non linear element Please refer to chapter 3 Modulation Sidebands and Noise Spectrums in William F Egan s book Frequency Synthesis by Phase Lock 2nd Eid for more information SSB spectrum AM Spectrum PM Spectrum 0 dam 0 dBm 0 dem Decomposes CEER 56 dBm 56 dBm 56 dBm 56 dBm 50 dBm 1 1 2 MHz 0 8 1 1 2 MHz 1 1 2 MHZ 0 8 The above illustration shows how the 50 dBm SSB signal located at 1 2 MHz is decomposed into its AM and PM counterparts The AM sidebands will drop in power by 6 dB and both have the same phase Whereas the PM sidebands will also drop in power by 6 dB but the lower sideband will be 180 degrees out of phase with the lower sideband in the AM spectrum When the AM and PM spectrums are added together the lower sidebands cancel out and the 56 dBm upper sidebands add coherently to yield the single upper sideband of 50 dBm Digital Dividers Frequency Multipliers and Dividers Since digital dividers frequency multipliers and dividers operate in hard limiting the decomposed AM spectrum is removed leaving only the PM spectrum Every input spectrum oth
362. ignals are included or ignored in this measurement Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as CNP TIMP and PNCP Travel Direction Same as CNP TIMP and PNCP Output 1 dB Compression OP1DB 513 Simulation 514 This measurement is the output 1 dB compression point along the path The value at the output of each stage represents that output compression from input to that stage output The last entry will always be the performance of the entire chain referenced to the ouput OP1DBJ n DCP LastStage min SDR n dBm where n stage number This measurement is made by determining the desired channel input power for the last stage and then adding to it the minimum 1 dB headroom or stage dynamic range along the path Channel Used Same as DCP and SDR Types of Spectrums Used Same as DCP and SDR Travel Direction Same as DCP and SDR Output Intercept All Orders All Orders OIP This measurement is the output intercept point along the path This is an in band type of intermod measurement OIP n ICP n Delta n dBm where n stage number and Order order of the intermod Delta n ICP n TIMCP n Order 1 dB Delta is the difference in dB between the Total Intermod Channel Power in the main channel and the interfering signal present in the Interferer Channel including the effects of the order In order to make this measureme
363. ilable if Generate Far Field Data above is checked It generates data for varying phi in the spherical coordinate system Phi is the angle formed from the positive x axis to a point projected on the xy plane in 3 space If Sweep Phi is unchecked a fixed angle will be specified and far field data will be produced only at this phi angle EMPOWER Options x General Viewer Far Field Advanced Thinring out subgrid 5 Thin out electrical lossy surfaces F Solid thinning aut slower accurate Add extra details to listing file v Use planar ports for one port elements vw Show detailed progress messages Only check errors and memory da not simulate Command line UF Cancel Apply Help 104 EMPOWER Planar 3D EM Analysis Advanced Tab Only check errors topology and memory do not simulate Useful to make sure you have the simulation and layout setup properly before a long EMPOWER tun This option provides a very important means both for checking the grid mapping and required memory EMPOWER just maps the problem onto the grid and calculates the required number of the grid variables for each frequency Check the map of terminals in the listing file to see the grid model of the problem and check the MEMORY lines in the listing file to get some idea about problem complexity and probable simulation time Setup Layout Port Modes Brings up the multi mode setup dialog box as describ
364. imal mesh with fewer small cells and an improved memory usage and simulation time For more information on Mesh Reduction refer to Effect of Mesh Reduction on Simulation Accuracy 10 On the Simulation Options tab Enable Horizontal side currents thick conductors to activate horizontal side currents on thick conductors Momentum Options General Simulation Options Mesh Simulation Mode CO Microwave recalculates substrate for every frequency Calculate O Substrate O Mesh and substrate All substrate mesh and 5 Parameters 3D Metal Expansion Mane thin metal Oup P Down via Model Lumped CO1D wire 2D planar no horizontal currents default CO 3D spatial include horizontal currents Use horizontal side currents thick conductors Reuse results of last simulation 350 Momentum GX When enabled this feature provides improved modeling of thick conductors by using the horizontal currents on side walls This feature e Automatically uses edge mesh on side walls when the global mesh option is enabled e Provides higher accuracy for thick conductors with a width thickness aspect ratio up to and below 1 0 e Makes substrate database recalculation unnecessary by taking advantage of new reconstruction technology for the Green functions For optimal simulation performance we recommend that you disable the edge mesh for expanded layers
365. imulation geometry As you can see this listing matches the desired layout 201 Simulation WE EMI Listing File Workspace mfilter_temp ce E EE i 5a 54 ES EZ El 5a 3 48 47 46 21 Now we would like to see the EM response on the same graph as the linear response to compare the two We can do this by opening the linear graph s properties MFilter1 Response and typing in MFilter1 EM1 DB S21 and MFilter1 EM1 DB S11 on line three and four of the measurements just like it s illustrated below 202 EMPOWER Planar 3D EM Analysis Graph Properties X Default Simulation Data ar Equations MFilter MAF ilterd Measurement fon Right Hide Color paean o a ae r mmmwr 2 jas EL F Fr mmm Mieimnsu O Ej Lr FEE Eee Ed E E Ems Right r Axis W Auto Scale Min 50 W Auto Scale W Auto Scale Log Scale um Help Cancel Min n w Measurement Wizard Max I ax fe Maw 2450 nt a Equation Wizard Divisions i H Divizions 0 H Divisions 10 Advanced Properties E nter the name of a parameter ta graph or press a wizard button to guide you through the process of creating a measurement Note You can also use the Measurement Wizard instead of manually typing in the measurement 22 Now it s time to analyze what each simulation is showing us Below is the graph which shows us both linear and EM simulations
366. imulations where no reflection will be seen since a reflection only happens at infinity One way to counter the difference between measurements in reality and simulation is to take into account the package box around the structure Another possibility is adding 391 Simulation 392 vias to your structure both in real life and in the simulations Try to make the physical representation of your circuit as close as possible to reality However the actual shape of the via should not be that important Small sheets can be used to represent the small circular vias that are realized in the real structure Surface W ave Modes If the substrate is not closed open or half open and not homogeneous surface wave modes can exist The parallel plate mode can be seen as a special case of a surface wave mode Their behavior is identical Both are cylindrical waves that propagate radially away from the source They are guided by the substrate Both fundamental and higher order surface wave modes exist Similar conclusions can be drawn with respect to limitations such as the effects of the modes Slotline Structures and High Frequency Limitation The surface wave accuracy deterioration of the calibration process is somewhat more prominent for slotline transmission lines than for microstrip transmission lines Therefore slotline structures simulated with Momentum will exhibit a somewhat higher noise floor 10 20 dB higher than microstrip structu
367. imum Analysis Frequency set Al Frags a5 Harmonics of Frequency Accuracy We calculate SSB noise only for the output IF spectral component whose frequency is defined as Fif 1 Frf 1 Flo then the index vector of the carrier is 11 or 1 1 the sign of the vector does not matter In the general case to calculate noise of an intermodulation carrier Fp q p Frf q Flo the index vector of the carrier will be pd ot p q where p 20 1 Nrf q 0 1 Nlo where Nrf and Nlo are harmonic orders of the RF and LO frequencies taken into account in the HB analysis set in the Frequencies table General page of Harbec properties 12 To perform noise contributors analysis set the checkbox Calculate Noise contributors This creates noise contributors data in HB analysis output dataset Note The noise contributors analysis is activated only for the one carrier noise analysis mode Noise tab Output Calculate SSB noise of Carrier with HARBEC Harmonic Balance Analysis NCValue Relative power of Noise contributors dBc NCName Noise contributors names NCIndex Noise contributors indexes Index vector The noise contributors in the tables NCName NCValue NCIndex are sorted by weight NCValue of noise contributor in output SSB noise of the cartier from most significant to less significant contributors The table always consists of the same noise contributors but thei
368. in The resultant matrix relating local grid currents and voltages is reduced to an immitance matrix relating integral currents and voltages in ports To extract a generalized scattering matrix of the problem from the immitance matrix the method of simultaneous diagonalizations is used After this introduction we are ready to formulate the reasons for using MoL as a basis fot an electromagnetic simulator The 3D problem is discretized only in two directions and reduced to a 2D one that corresponds naturally to the planar MIC structures In contrast with the method of moments the MoL gives a self regularized solution with only one variable grid cell size defining all parameters of the numerical model That eventually leads to monotonic convergence of calculated data and predictable errors of calculations The high grade of internal symmetries of the MoL based algorithms makes it possible to substantially reduce the numerical complexity of the main matrix computation stage The main restriction of using a regular grid related with its potentially excessive number of variables has been overcome by introducing thinning out and re expansion procedures Basically the discrete analogue of a problem is processed in a way similar to the method of moments but in discrete space like the finite difference approach which facilitates different aspects of the solution and programming Thus the main advantages of the MoL are reliable solution with the predictable
369. in EMPOWER Use this for ground planes and other layers which are primarily metal Do not use this for lossy layers See your EMPOWER manual for details Current Direction Specifies which direction the current flows in this layer The default is along X and Y X Only and Y Only can be used to save times on long stretches of uniform lines Z Up Z Down XYZ Up and XYZ Down allow the creation of thick metal going up down to the next level ot cover EMPOWER Planar 3D EM Analysis Thick Metal Checking this box forces EMPOWER to model the metal including thickness EMPOWER does this by putting two metal layers close together duplicating the traces on each and connecting them with z directed currents If thick metal is used then Current Direction is ignored Element Z Ports This setting specifies the default direction for automatically created element ports either to the level above or to the level below Generally you should choose the electrically shortest path for this direction EMPOWER External Ports Overview Every EMPOWER circuit must contain at least one port These ports are divided into two major categories external ports which are at a sidewall and internal ports which are inside the box This section will cover only external ports internal ports are discussed in a later section of this manual Placing External Ports By now you should be familiar with the placement of external ports EMPorts If not yo
370. in metal enclosures By adding a box to your design the metal sidewalls that are present in the real structure may be included in the simulation This is useful is you suspect that the presence of these sidewalls will have an immediate effect on the behavior of the circuit For example broad coupled filters are placed in metal enclosures a box and the sidewalls can have an significant influence on the filter characteristics e Nearby metal sidewalls may affect circuit performance You may want to use a box because metal sidewalls are present in the real structure and there may be an effect from these sidewalls on the characteristics of the circuit This can be a parasitic unwanted effect If the effects of the sidewalls were not taken into account while designing the circuit you can verify any effect that the sidewalls may have on your circuit In most cases when the sidewalls are not too close to the actual circuit the effect of the sidewalls on the simulation results will be marginal There is however a specific significant condition which is unique for structures with sidewalls e The box may resonate In the case of a box this is the occurrence of one or more box resonances A box resonance is a physical effect where under the condition of certain frequency and box size combinations the box actually starts resonating at a certain frequency Because a box resonance has a significant effect on S parameters in a small band cen
371. in tables The analysis can be enabled to Automatically Recalculate or may need to be manually calculated If the analysis has been set to Automatically Recalculate datasets will appear in the workspace tree after the analysis If manual calculation is needed the calculate button 9 will appear red and so will other items on the workspace tree Click the calculate button to update the system analysis and create the necessary datasests After calculation the workspace tree should look like 419 Simulation Workspace Tree ox gel Default 4 Designs B3 Path Data 0 Ed Systemi_Data Pathi i Systemi Data Path m Schi Schematic 2 T System Schi Systemi Data p Su Equation i Motes For more information on datasets click here Back to Add a System Analysis Next to Add a Graph or Table Add a Graph or Table There are several ways to display data in GENESYS Only one way will be demonstrated here For additional information on graphs click here The easiest way to add a spectral power phase or voltage plot in Spectrasys is by right clicking the node to be viewed and selecting System1_Data New Power Plot at Node x from the submenu Add New Graph Table The output of the attenuator was selected in the following figure Coupler1_1 IL 0 5dB CPL 20 0dB Synthesize Subcircuit gt 7 Add New Graph Table System1 Data New Power Plot at Node 4 System1 Data New Voltage Plo
372. indexIM 1 0 IM3 117 694 hb_getspcompdbm P2 FreqindexIM 2 1 IM5 185 7472 hb getspcompdbmiP 2 FreqindexIM 3 2 i OIF 3 18 309 2 hb oipn P 2 FregIndexTM 1 0 1 2 3 OIPS 12 655 hb oipn P2 FregIndexIM 1 0 2 3 RAPZ PPORT 2 dis PORT im mo oF nz i Haw VPORT diay ZPORT The variables are propagated in the sweep and may be plotted 36 Getting Started IMn Pin and OIPn Properties For an example of the above in an Amplifier design see the example XExamplesVAmplifiersNAmplifier IPn Calculation wsx 37 Simulation Parameter Sweeps 38 Parameter Sweep 3D graphs in GENESYS require parameter sweeps to generate a third dimension for plotting Parameter Sweeps give you this third dimension by adjusting a tuned variable repeating another simulation for each adjustment For example to see how the response of a circuit changes when a capacitor is adjusted you can add a Parameter sweep which sweeps the linear or electromagnetic simulation while adjusting the capacitor value You can then view the results on a 3 D graph To add a Parameter Sweep Evaluation 1 Create a design with a schematic 2 Define your tunable parameters 3 Click the New Item button a on the Workspace Tree toolbar and choose Add Sweep from the Evaluation menu 4 Define the Parameter Sweep Properties and click OK The analysis runs and creates a data set
373. individually the user can find the origin and path of each spectrum by placing a marker on the graph or placing the mouse cursor over the spectrum of interest When a graph marker is added to a plot the marker will attach itself to the closest spectral data point The mouse flyover text ONLY appears when the mouse is over the spectrum data point or the marker text on the right side of the graph These spectrum data points can be enabled or disabled NOTE If you are having a difficult time getting the mouse flyover text the try enabling the spectrum data points to see the exact data locations for the mouse Placing the mouse over the data point on a spectrum yields the following 433 Simulation 1 Ss gS p e ji eem n eee omni 5 m tl P7 20 CAA E Attn1 C1 RFAmp1 TL1 AtEnZ RFAmpz 20 90 60 zi aD 100 S700 3800 3900 4000 4100 4200 4300 4400 4500 4600 4 00 Frequency MHz The format of the spectral identification is as follows GENERAL FORMAT Line 1 Measurement Name Line 2 Marker Frequency Marker Power Voltage Level Line 3 Coherency Number Signal Type Frequency Equation Origin Element Next Element Current Element Coherency Number All signals in Spectrasys are grouped according to a coherency number All signals with the same coherency number are coherent with each other Click here for additional information on coherency Signal Type D Desired Signal All spectrums are e
374. ine output data of the noise analysis Set the radio button Calculate SSB Noise of to All Carriers to calculate SSB noise for all noise carriers all spectral components of Harmonic Balance solution to Carrier with Index Vector to define only one spectral component carrier at which SSB noise will be calculated Set integer index vector of the carrier in the format k1 k2 kn where ki is the harmonic index of i th input signal frequency from Harbec General page For example the 266 HARBEC Harmonic Balance Analysis lower sideband mixing term in a mixer would be entered as 1 and 1 The indices are listed in sequential order by carrier Mixt1 Schematic OFF FORT FE2DODeeiHz PADa bbm PRF fe PFC Fa pith C1 1 1pF LG PORT F Toba PAG 10d bet PLDI fet PerlList gt Sehmmatit For example observe the simple mixer circuit above It has 2 input ports LO Flo 1750MHz and RF Frf 2000 MHz The Frequencies table created for this circuit has n 2 signal sources the 1st is RF and the 2nd is LO source frequency 267 Simulation 268 Harmonic Balance Analysis Options General calculate Moise Advanced Mame LEN Design Mic Schematic d Calculate Now Dataset Mie Diets dd Factory Def aus Description Temperature 16 85 C DC Anas defaut Analysis Frequencies Frequency Units MHz Maximum Ming Order 5 Max
375. ine the thermal noise power level The actual thermal noise is shown for convenience Noise Points for Entire Bandwidth This is the number of points used to represent the entire noise spectrum Noise will automatically be created beginning at the frequency 547 Simulation 548 specified by Ignore Frequency Below and ending at the frequency specified by Ignore Frequency Above These noise points will be uniformly distributed across this bandwidth Note The more noise points used in the simulation the longer the simulation time generally takes Since each component generates noise the more components in a schematic will also increase the simulation time Better speed performance can be achieved for a large number of components by disabling noise calculations or reducing the number of simulation points Add Extra Noise Points This 1s the numbet of extra noise points that will be inserted across the Extra Points Bandwidth parameter These additional noise points will be uniformly distributed across this bandwidth The center frequency of these noise points is the center frequency of the desired signal frequency These noise points will be added to evety created desired spectrum However unused noise points will be removed to improve simulation time See Broadband Noise and Cascaded Noise Analysis sections for additional information Extra Points Bandwidth This is the bandwidth where additional noise points can be inserted
376. ing algorithm will maximize the number if uniform rectangular cells created in the mesh The meshing algorithm is very flexible as different parameters can be set by the user number of cells wavelength number of cells width edge meshing and mesh seeding resulting in a mesh with different density It is clear that the mesh density has a high impact on both the efficiency and accuracy of the simulation results Default mesh parameters are provided which give the best accuracy efficiency trade off Both the RF and microwave modes use mesh reduction technology to combine rectangular and triangular cells to produce a mesh of polygonal cells thus reducing demand for computer resources Mesh reduction eliminates small rectangles and triangles which from an electrical modeling point of view only complicate the simulation process without adding accuracy As mentioned previously this feature may be turned on or off with a toggle switch on the Momentum Analysis properties window Loading and Solving of the MoM Interaction Matrix Equation The loading step of the solution process consists of the computation of all the electromagnetic interactions between the basis functions and the filling of the interaction Momentum GX matrix and the excitation vector The interaction matrix as defined in the rooftop basis is a dense matrix that is each rooftop function interacts with every other rooftop function This electromagnetic interaction between two basis func
377. ints 3 If your problem is very large you may want to increase the cell size or make other tradeoffs to reduce the time required for calculation If you use this technique save the file with a new name before you generate viewer data so that you do not corrupt your existing S Parameter data See the EMPOWER Viewer section for more information Slot Type Structure In the normal mode EMPOWER solves for the currents in the metal There is an additional mode where EMPOWER solves for the voltages in the gaps and in lossy metals This mode must be turned on manually by checking Slot type structure when starting an EMPOWER run from GENESYS or by using the VOLTAGE keyword when describing a LAYER in a TPL file In general you should check Slot type structure whenever the metalization layer has more lossless metal than open space This is often the case in a slot type structure such as coplanar waveguide The answer will always be identical but you will save orders of magnitude of memory and simulation time by ensuring that this checkbox is set to the tight value Note This setting has no effect on z directed metal viaholes etc which is always calculated as currents There is a caveat when describing lossy problems with this option All non ideal metal must be analyzed so if the metal in your problem is lossy turning on Slot type structure will result in both the air and the metal being analyzed which will have a disastrous ef
378. io looking in from port 7 The VSWR is a measure of the energy reflected back to the port The VSWR is related to the s parameter S11 by VSWRi 1 81 1 Su J Therefore as the reflected energy goes to zero S11 goes to zero and the VSWR approaches unity As the reflected energy increases S11 approaches unity and VSWR goes to infinity Values Real value versus frequency Simulations Linear Default Format Table RE Real Graph RE Real Smith Chart S plots s parameters Commonly Used Operators None Examples Measurement Result in graph Smith chart Result on table optimization or yield VSWR1 VSWR1 VSWR VSWR Show VSWR for all ports Not available on Smith Chart plots s parameters Note a vswr function is available in Equations Input Impedance Admittance ZIN Z YIN 19 Simulation 20 The port impedance and admittance measurements are complex functions of frequency The measurements are made looking into the network from the port with other network terminations in place The frequency range and intervals are as specified in the Linear Simulation dialog box A port number 7 is used to identify the port ZINz is the input impedance looking in from port 7 YIN is the input admittance looking in from port 7 Values Complex value versus frequency Simulations Linear Default Format Table RECT Graph RE Smith Chart Si plots s paramters Voltage Gain This voltage gain
379. ions They ate calculated for each pair of signal strip slot and or via layers mapped to a substrate level Although it is necessary to know which signal layers are mapped to a substrate level since only impulse responses are being calculated it is not necessary to know the patterns on these signal layers This implies that the Green s functions can be pre calculated and stored in a substrate database This allows the substrate Green s functions to be reused for other circuits defined on the same substrate The high frequency electromagnetic Green s functions depend upon the radial distance and the frequency The computations are performed up to very large radial distances over the entire frequency band specified by the user The frequency points are selected adaptively to ensure an accurate interpolation with respect to frequency Computations performed over very wide frequency ranges can consume more CPU time and disk space to store the results To increase speed the RF mode uses quasi static electromagnetic Green s functions based on low frequency approximation and scales the quasi static Green functions at higher frequencies Meshing of the Planar Signal Layer Patterns The planar metallization strip via and aperture slot patterns defined on the signal layers are meshed with rectangular and triangular cells in the microwave simulation mode As translational invariance can be used to speed up the interaction matrix load process the mesh
380. ip Capacitor footprint has been used for each of the lumped capacitors Whenever a lumped element is used for an EMPOWER run GENESYS creates an internal ports for the element These ports are placed e If Use Planar Ports for two port elements is checked in the EMPOWER properties box one port is created for 2 terminal elements like resistors or capacitors which are aligned horizontally or vertically e Inal other cases an internal port is used for each terminal of the element This port is placed at the center of the pad footprint and EMPOWER writes data for each port created whether internal or external e The 1 and 2 ports pictured in the figure above are examples of external ports Potts are described in the External Ports and the Internal Ports sections Automatic port placement is a powerful technique and real time tuning can be employed in GENESYS once the EMPOWER data for has been calculated Simulating the Layout Double click EM1 in the Workspace Window This displays the EMPOWER Options dialog shown below EMPOWER Planar 3D EM Analysis EMPOWER Options Layout to simulate EE aK Port impedance pO Generalized Cancel Automatic Recalculation i Recalculate Now Electromagnetic simulation frequencies Start freg MHz 4n0 E Automatically save workspace alter calc Stop freq MHz 2600 E Viewer data Generate viewer data slower Number af points E Mas critical freg 26
381. ircuit must contain at least one port This section will cover lumped elements and internal ports ports inside the box External ports along a sidewall were also covered in that section Placing Internal Ports The process of placing an internal port is similar to the process of placing an external port To summarize An internal port is placed in LAYOUT by selecting EMPort from the toolbar Internal ports can be placed anywhere in the box When the EMPort Properties dialog box appears first select Internal in the Port Type combo box Next fill in the width and length of the pad Press OK to complete the placement Note The rest of the options in the EMPort Properties dialog box were covered in the section entitled Port Options You may want to review these options now The figure below shows a comparison between ports in circuit theory and internal ports in EMPOWER In the circuit theory schematic on the left there are two ports Each port has two terminals with the bottom terminal generally being ground In the EMPOWER illustration there are two z directed ports one at each end of the line These z directed ports are mapped onto the grid along Z much in the same way as a viahole would be mapped See the Basics section for more information on mapping to the grid As in the circuit theory schematic there are two ports and each port has two terminals The bottom terminals which are true ground in the circuit schematic are
382. is after a coherent addition Coherent additions are especially important at the input to non linear devices since the total spectrum from coherent signals will yield a different power than individual spectrums This total power is needed to correctly determine the operating point of the non linear devices The coherency number can be viewed by the user when examining the spectrum identification The coherency number of a new spectrum will use an existing coherency number if the two spectrums are coherent Several rules are followed to determine if a newly created spectrum is coherent with an existing spectrum All of these rules must be followed before any two signals can be considered coherent NOTE If the Coherent Addition option is unchecked then all intermods and harmonics will always be non coherent and well as any mixed products out of a mixer regardless of the following comments 1 Each source is only coherent with itself OR if the sources have the same reference clock When no reference clock is specified for the source only signals created from this soutce will be coherent assuming the other coherency constraints are met When multiple sources have the same reference clock name and the other coherency constraints are met then the signal is considered to be coherent 2 Signals must be of the same type Signals generally have the following categories Source Intermod Harmonic and Thermal Noise Coherent signals only apply to
383. is the incident wave frequency the current distribution Ig t at time t is given by expression Ig t Ig U exp Z pi f t The same formula is valid for the voltage distributions Advancing time displays snapshots of the current or voltage distribution thus animating the display As we mentioned above the viewer reads the grid currents or voltages with their cootdinates and prepares them for plotting The preparation stage includes a transformation of the grid variables to more general current density functions surface electric current density function for strip like problems or surface magnetic current density function for slot like problems The units for the electric current density magnitudes are Amperes per millimeter A mm The units for the magnetic current densities are Volts per millimeters V mm We choose millimeters to scale graphs to more readable values The current density functions are created only for the currents in the signal or metal layer Viaholes and z directed ports are always represented as z directed currents in Amperes Summary To summarize viewer behavior e If Generate Viewer Data is selected the default incident wave is the first eigenwave of the first input e Define the input number and mode number in the EMPOWER properties dialog e An incident wave is a time harmonic function with unit magnitude and zero initial phase EMPOWER Planar 3D EM Analysis e The external ports are terminated by c
384. ismatch loss it typically several times greater than the insertion loss Signals arriving at a filter input will either be reflected or transmitted All transmitted signals through the filter will undergo the loss due to dissipation Dissipative loss is a function of the Q of the components that make up the filter Ideal components have infinite Q and will have no insertion loss When looking at out of band channel power or voltage along a cascade of stages it will appear that the stage prior to a filter has more loss than expected and the filter attenuation is much less than expected This seems counter intuitive at first but when we look at mismatch loss and transmitted and reflected power it all makes good sense The out of band impedance looking into a filter is either really high or really low There is an out of band impedance mismatch between the filter and the stage prior to it The actual out of band voltage appearing at the filter input will be affected by mismatch impedance which automatically accounts for the out of band attenuation caused by the filter In the lab the only way to accurately measure the out of band input voltage to a filter is to break the input node connection and measure the forward voltage transmitted voltage and reflected voltage Vforward Vtransmitted Vreflected In order to simplify the simulation process all SPECTRASYS measurements and spectrums only use transmitted signals not the forward signals assumed b
385. ists between the desired signal metalization and the cavity Because this coupling is reciprocal coupling occurs between segments of the signal metalization This is nearly certain to perturb the circuit responses as the operating frequency approaches or exceeds the first resonant frequency of the cavity While EMPOWER inherently predicts these effects they may have a significant destructive effect on the performance of your designs Box modes are clearly illustrated in this example 157 Simulation 158 Homogeneous Rectangular Cavity In the formulation which follows we use definitions from the section on Geometry The height of the box in the z direction is h the length of the box in the x direction is a and the width of the box in the y direction is b The resonant wave number for a rectangular cavity 1s Lf F r h 2 i 2 i k Inm HT T k uso ur S E a hj b MKS units and the resonant frequency when homogeneously filled with material with a relative dielectric constant of e is kc mnp Phir JE map where c is the velocity of light in a vacuum 2 997925x108m sec The frequency of the dominant mode is fro lowest resonant frequency and in a vacuum we have c 11 l ha RAF In air with linear dimensions in inches and the frequency in megahertz E UE 1 31 fia 2 5900 MHz inches 101 TE 2 Va b With linear dimensions in millimeters and the frequency in gigahertz 1 1
386. it can oscillate 1 Create closed loop oscillator circuit with OSCPORT Probe properly located on oscillator at a node where small signal oscillator condition may be satisfied 2 Create Harmonic balance oscillator analysis HBOSC for the circuit 3 Before using the HBOSC analysis in the Oscillator Tab run Calculate Osc Frequency and be sure to put in an appropriate range for the frequency search 4 Once Calculate Osc Frequency has been run the OSCPORT probe on the circuit is automatically calculated with Insertion Voltage and Oscillation frequency 5 Now you can run the full blown oscillator analysis The oscillator analysis dialog includes all pages of regular HB analysis and additional tab Oscillator defining specific settings for the analysis In Genesys 2006 the tab includes the same parameters as in Genesys 2004 and few new ones 258 HARBEC Harmonic Balance Analysis Harmonic Balance Oscillator Analysis Options General Calculate Oscillator FFT Oscillations Frequency Calculation Humber of Points 101 Calculate Osc Frequency 2 zezieg Use Osc Solver Osc Port Initial Voltage Absolute Current Tolerance Number of Curve iterations Max step of Curve tracing Number of Adjust iterations All parameters of the dialog page are grouped into 3 groups 1 Small signal oscillator frequency calculation e Minimal Frequency Maximal Frequency 259 Simulation
387. ither marked desired or undesired Desired spectrums are generally those of main interest in the simulation Desired spectrums 434 Spectrasys System consist of signal sources selected multiplication or division values through frequency multipliers and dividers and sum or difference products and determined by the user None Undesired Signal If there is no D displayed then the signal is an undesired signal Frequency Equation From the frequency equation the user can identify which source frequencies created the spectrum This equation is written like a typical mathematical equation The equation will contain the name and combinations of all the sources that created the spectrum Analog to Digital Equations Additional spectrum identification following an analog to digital converter is provided in the frequency equation The originating Nyquist zone of the signal will we specified along with an indication of whether the spectrum was inverted or not in the downconversion aliasing process Path The path of the spectral component can be determined by examining the comma delimited sequence of reference designators which identify the element where the spectrum was created and the element sequence that the signal took to arrive at the destination node The first reference designator after the closing frequency equation bracket shows the reference designator where the spectrum was created The subsequent reference designators indic
388. ity information so the effects of bandwidth are automatically accounted for As spectrums propagate through the system spectral genealogy is maintained providing users with the ability to identify the propagation path of every spectrum Furthermore this parentage information also includes coherency identification desired or undesired status and the frequency equation associated with the given spectrum Click here for more information on spectral identification This simulation technique is extremely fast compared to other traditional non linear simulation techniques such as harmonic balance that requires convergences criteria and mathematical inversions of large matrices to achieve simulation solutions Users specify arbitrary paths through a single block diagram to gather cascaded information along a given path Each path contains several types of paths such as desired total noise phase noise etc Each spectrum along the designated path will be placed in the appropriate path category Measurements operate on specific path types to create desired affects For example the channel noise power measurement excludes all signal intermods and harmonics and phase noise spectrums from its path spectrums giving the user only noise within the channel regardless of whether or not a much stronger signal is located at the same frequency This is a huge advantage allowing the user to see and measure true in channel signal to noise ratio SPARCA Simulation ad
389. ixer Noise Simultan 282 D hogy ape 181 183 jnode setup DOX c t necp aiit 145 MOG H 2977 multidimensional BE TS eee hot 225 multimode 139 143 145 151 174 177 206 PTE OCS dies debet eni e ettet 151 TIN A A EEE E E 27 ded rol er E TT TT TEE 271 hg esc t 282 hg oh 14 INC CIE Suec HOMI ISMEE esset 11 INC Dale EA 271 INKOINLATETG on RR TRA TRA V rol RU dr netiis 271 NCValue eese nnne nnne nenne senno ns 271 NDECP o MR i A 515 new A 5 eid OU 6 Newton RAPS O Sd eie ela 234 hj EA 11 13 HELEN DT o A AA A PEPPER 282 A e Eid ils ce aes 8 NESS occum UM ccc USD 282 c P m 11 INE NS aa 11 13 Ds A MCONE ONCE RUERE 527 No Attenuation Across a Filter 535 no deembetdding e oia 101 Node Nore Volane eair e 52 DOE Gesn tesis lasso sse aded ia 11 436 Noise and Distortion Channel Powet 513 MOSS CHECA HT etes ir eerta 11 14 298 noise correlation essen 8 11 Dole AEA esse p UM E RUE 8 MOISE OTTE iii odo eda dai 533 Hore Noure CE CIEASE Atenas N 299 HOUSE dE osccuiuntu edic UNI 1 85 nonlinear device models eet 5 nonlinear measurements 1 sees eee ee dene tectus 13 nonlinear thodelss usce RR RSEN 287 Nonlinear Noise Simulation 265 WORE STAN GUG Fetal amp ca secre rt de iS 1 normal deembedded 2 252y
390. j with network terminations in place Noise correlation matrix parameters Default Operator db re re re re re Linear real db re Shown on Smith Chart Sj SZ SZ 11 Simulation GMAX Maximum available gain db NF Noise figure ddb NMEAS Noise measure db NFT Effective noise input temperature Linear real GOPT Optimal gamma for noise db GOPT YOPT Optimal admittance for noise re GOPT ZOPT Optimal impedance for noise re GOPT RN Normalized noise resistance Linear real NFMIN Minimum noise figure db ZMz Simultaneous match impedance at port 7 re GMz YMz Simultaneous match admittance at port 7 re GMz GMz Simultaneous match gamma at port 7 db GMz K Stability factor Linear real Bl Stability measure Linear teal SB1 Input plane stability circle None Circle SB1 Circles Note Filled areas are unstable regions SB2 Output plane stability circle None Circle SB2 Circles Note Filled areas are unstable regions NCI Constant noise circles shown at 25 5 1 1 5 2 None Circle NCI Circles 2 5 3 and 6 dB less than optimal noise figure GA Available gain circles None Circle GA Circles GP Power gain circles None Circle GP Circles GUI Unilateral gain circles at port 1 None Circle GUI Circles GU2 Unilateral gain circles at port 2 None Circle GU2 Circles Can only be used on 2 port networks Gain circles are only available for 2 port networks Ci
391. ke a very long simulation you might want to increase the order of the integrator after first trying a small run to be sure that the simulation is not unstable When solving for each time step CAYENNE must use an initial guess for the solution for that time step If the predictor 1s off then the initial guess 1s simply the solution at the previous time step If the predictor is on then an Nth order polynomial approximation is fit to the previous points generally using a Gear integrator and the predicted values at the new time point is used as the initial solution Generally using the predictor results in CAYENNE Transient Analysis faster simulation However in some highly nonlinear circuits the predictor can actually hinder convergence CAYENNE may automatically turn off the predictor if it detects this situation In summaty e Only change integrator settings as a last resort in an attempt to fix simulation problems One of the problems sometimes fixed by switching to second order gear integration is high frequency ringing e If you have a very long simulation to run try running a piece of it with higher order integration up to sixth order gear with the predictor set to Auto and with the predictor both on and off Then use the best settings to speed simulation while maintaining accuracy Cayenne O ptions Transient Analysis Properties General Integration Time Step Convolution Output Miscellaneaus Mame Transient
392. ke simplifying assumptions about the layout of the structure It is therefore necessary to take these assumptions into account and follow them to get accurate solutions e The walls of the box are assumed to be infinitely far away from the structure e Ifa substrate is used it is also assumed to extend infinitely in the lateral dimensions e Fields generated from z directed currents are not taken into account therefore it is not recommended that you include vias in the layout Setting Up the EMPOWER Box To get good results for the far field radiation patterns the following rules must be observed e The structure should be centered in the box e The walls of the box should be far away from the structure e Only one layer of metal must be used e Exactly one substrate or an Air Below layer must be under the metal layer not both There are 3 different antenna types for which far field radiation patterns can be generated e Antenna in free space EMPOWER Planar 3D EM Analysis e Antenna above a ground plane e Microstrip antenna above a substrate and ground plane To simulate an antenna in free space no substrate should be used and the only layer below the metal layer should be Air Below The height of the Air Below layer in this case is irrelevant Both the Top Cover and Bottom Cover should be set to Electrical type with surface impedance set to 377 ohms 377 ohms is the intrinsic impedance of free space To simulate an
393. kspace e Creates results containing WorkSpace_Sonnet Sonnetl Genesys yp plus any lumped elements on the layout 4 Add the Sonnet results to a graph 5 If any changes to the layout or Sonnet setup are made the recalculation step 3 is done again However before the recalculation step the user has the option to delete the existing simulation data See the Deleting Internal Simulation Data section for details D eleting Internal Simulation Data When GENESYS detects a change to the layout or to the Sonnet simulation options the user is offered a choice to delete the existing simulation data Additionally the user can 395 Simulation 396 right click the Sonnet simulation on the Workspace Window and select Delete Internal Simulation Data This does the following 1 Deletes all data stored inside the workspace including the Genesys son Sonnet file and Genesys yp simulation results files 2 The next time Sonnet is run the WorkSpace_Sonnet Sonnet1 SonData Genesys directory will be emptied This clears all of Sonnet s previously stored simulation results GENESYS cooperates with the Sonnet caching mechanisms giving several benefits e Sonnet will check the simulation data for consistency before running e If more frequency points are added the previously calculated points do not need to be recalculated As a side effect of this cooperation it is possible that Sonnet will see a change in the geometry and refu
394. l Calculate Now Design A Factory Defaults Dataset d Save as Favorite Description Automatic Recalculation Time Setup Starting Conditions Stop Time 2000 O Use Zero voltages Skip Bias Aemini an E L Help Oscillators Start Tolerance voltage Truncation Factor Relative 18 3 Most Accurate Frequency MHz 77 Simulation 78 Time Setup Simulation always starts at 0 seconds The stop time tells the much time to simulate The maximum step size tells the simulator the minimum sampling rate at which to compute the output waveform The simulator may calculate samples at a higher sampling rate depending upon the settings in the Integration Time Step tab Starting Conditions Use Zero Voltages Skip Bias assumes an initial condition of zero volts at every node in the circuit which can be useful to simulate turn on Use DC Bias Point runs a DC analysis at time 0 to determine the initial voltages for the circuit The option to help oscillators start slightly randomizes the initial voltages and is often necessary when analyzing an oscillator since oscillators rely on start up transients to oscillate Tolerance The Current and Voltage tolerance tell the simulator at what precision the iterative algorithms used to solve the circuit equations can terminate Most Accurate Frequency If CAYENNE determines that the accuracy of a frequency independent model is acceptable then all models will b
395. l intermods channel powers along a path Furthermore these measurements can also segregate the data based on intermod order These channel based measurements should not be confused with a measurement called Total Intermod Power which contains the total intermod power of the entire spectrum at the given node and cannot segregate its data based on order Here is a simple diagram showing how to setup Spectrasys to make intermod path measurements Power dBm SL Channel bandwidth must be greater than the Intermod bandwidth but small enough to exclude other signals 160 180 200 Frequency MHz Intermads Determining the Intercept Point Spectrasys System Intercept point measurements are assumed to be from two interfering tones The calculations are based on what would be done in a laboratory as shown in the following fioure As can be seen two measurements are needed to determine the power of the intercept point These measurements are the power level of the intermod and that of the one of the two tones If the interfering tones are attenuated through the system like in a receiver IF a virtual interfering tone must be created by the simulator to correctly determine the intercept point This is done by injecting a small test signal at the intermod frequency to measure the in channel gain Knowing the power level of the two interfering tones to the system plus the in channel cascaded gain a virtua
396. l is used for this measurement Types of Spectrums Used None Travel Direction N A Stage Input Intercept Voltage All Orders SIIV This measurement is the stage input intercept voltage point calculated by using the Stage Output Intercept Voltage and the Stage Voltage Gain When a stage doesn t have this parameter 100 kV is used SITV n SOIVIn SVGAIN n dBV where n stage number Each column in this measurement is for a different intermod order up to the Maximum Order specified on the Calculate Tab of the System Analysis Dialog Box The column number is the same as the order starting from the left with order 0 Channel Used No channel is used for this measurement Types of Spectrums Used None Travel Direction N A Stage Input Saturation Voltage SIVSAT This measurement is the stage input saturation voltage calculated by using the Stage Output Saturation Voltage and the Stage Voltage Gain When a stage doesn t have this parameter 100 kV 1s used SIVSAT n SOVSAT n SVGAIN n dBV where n stage number 529 Simulation 530 Channel Used No channel is used for this measurement Types of Spectrums Used None Travel Direction N A Stage Output 1 dB Voltage Compression Point SOVIDB This measurement is the stage 1 dB voltage compression point entered by the user When a stage doesn t have this parameter 100 kV is used Channel Used No channel is used for this measurement Types of Spectrums Used None
397. l tone power can be determined at the output of the system which will be used to find the intercept point In the laboratory this would be done in a two step process since a spectrum analyzer is unable to separate out the cascaded gain test signal and the intermod However in Spectrasys this presents no problems and both signal can co exist at the same time Nth Order Intercept En nth Order Intercept dBm I P Interferer Power dBm Remember Intermod bandwidth is a function of the governing intermod equation For example if the intermod equation is 2F1 F2 then the intermod bandwidth would be 2BW1 BW2 Note Bandwidths never subtract and will always add The channel bandwidth must be set wide enough to include the entire bandwidth of the intermod to achieve the expected results For example CW signals have a 1 Hz bandwidth Therefore a third order intermod generated from CW signals will have a 3 Hz bandwidth If the channel bandwidth is set smaller than 3 Hz not all of the third order intermod energy will appear in the intermod measurements 459 Simulation 460 Intercept points can only be determined by measuring and interferer signal in an interferer channel The user must set the main channel frequency in Spectrasys to the frequency where the intermods are to be measured and the interferer channel must be set to the frequency of the interferer to get the correct interferer channel power Intercept
398. lation EMPOWER Tips Overview Often electromagnetic simulation involves tradeoffs and compromises to keep simulation times and memory requirements as small as possible while making accuracy as high as possible This section looks at several choices and clarifies the tradeoffs Table 3 1 lists various features and gives their impact on simulation times accuracy and memory requirements Each of these choices are looked at in detail below The values are approximate and may vary Reducing Cell Size by 2 Raising Max Critical Freq E AM Symmetry x1 4 to x1 16 LEE LLL x1 LL NME to x1 16 Turning asas Thinning Out Increasing Wall amp Cover Spacing Choosing Correct Covet iens HH Viewer 2 em x10 Data Corecting Slot x1 64 x1 256 Type Structure Using Preferred Box Cell Count Cell Size Cells should be small enough so that the result is accurate at least 10 cells per wavelength at the maximum critical frequency see below Additionally the cells should be small enough that there is at least one and preferably more cell across every line and gap Decreasing the cell size makes all stages of the solution take longer so decreasing cell size can be an expensive way to get more accuracy Conversely increasing cell size is a great 128 EMPOWER Planar 3D EM Analysis way to do an initial run of your problem to make sure that the result is close before you start a simulation that will take hours See the EMPOWER Ba
399. lation Linearl db S21 wrong ILLEGAL The operator must go around the entire measurement X Shows the global equation vatiable X which must contain post processed results Using Equation Results post processing Note An advanced post processing example called STFT of Oscillator Response wsx which produces a graph of the time dependent Fourier Transform of an oscillator s transient response can be found in the Equations subdirectory of your Examples directory Anywhere that a measurement is used post processed equation variables can be used The format is variableN ame where variableName is a variable from the equations for the workspace Inline equations can also be used anywhere a measurement can be used Start the measurement with to indicate an inline equation For example mag V1 mag V2 will use the difference of V1 and V2 Note For more information on Equation syntax and available functions see the Uszng Equations chapter of the User s Guide Linear S Parameters 15 Simulation 16 This S parameter or scattering parameter measurements are complex functions of frequency The frequency range and intervals are as specified in the Linear Simulation dialog box The s parameters assume a 50 ohm reference impedance unless otherwise specified The s parameters for an n port network are of the form Sj for i j equal 1 2 n Details on the S parameters and their application are found in Section x x of
400. lation Time Steps section above for more details If Save Port Voltages is selected then a VPORT array is created with voltages for each port This feature works with standard and differential ports If Save Node Voltages and Currents is selected then all node voltages and currents are saved subject to these limitations If Save Internal Voltages and Currents is not selected then internal nodes for example inside a transistor are not saved Levels of Data specifies how far down in the hierarchy counting the master level to save values CAYENNE Transient Analysis Iterations sets up details characteristics of the solver and should generally not be modified Miscellaneous Maximum Simulation Points is useful during automated runs or optimizations to ensure that bad value does not cause the simulator to shrink the step size too far and take too long to run Notice that this is a limitation on simulation points not output points The Mimimum Conductance is added from every node to ground to help convergence and matrix stability within the simulator In the example above a 1 teraohm resistor is added from each node to ground 83 Chapter 4 DC Analysis DC Analysis O verview DC simulation analyzes the static operating points DC voltages and currents at each nonlinear node and port in the circuit When designing circuits using non linear models you should always check the DC operating point before doing linear or harmonic b
401. lation or not For analysis a passive MIC structure is confined inside a three dimensional rectangular volume bounded by electric or magnetic walls The volume is filled by a layered medium that may consist of an arbitrary number of isotropic homogeneous dielectric or magnetic layers as shown below The electric E and magnetic H field vectors are related by Maxwell s system of equations rot H ioe E Jz rot E ion H div E 0 div H 0 x y z EQ p A 1 EMPOWER Planar 3D EM Analysis Here Jz is the volume density vector of z directed currents inside a media layer ep and muy ate permittivity and permeability of the media layer ep is a complex value for a lossy media The z directed currents are constant values inside a layer but they can change from layer to layer which gives a possibility to discretize the problem along the z axis Thus we have all six components of the electric and magnetic fields inside a layer with constant current across it X and y current components can exist only in a signal layer z dj parallel to medium layer interfaces Generalized boundary conditions for the signal layer are 1 H dj H d 2n 1 E dj E aj 0 The signal layer plane can contain arbitrarily shaped regions of perfect metalization regions with complex surface impedances lossy metal resistive films and regions modeling lumped element connections All regions have zero thickness The top and bo
402. lative path if you place your Verilog A files in a subdirectory under your User Model path 2 Create a text file containing the Verilog A source code in a text editor such as Windows Notepad and save it into the directory chosen in step 1 Be sure to use the extension va on the file See the Verilog A Tutorial for information about creating a Verilog A file 3 To use your model you must change the model on a part to module filename For example if your module is called limiter and it is in the file MyLimiter va you would set the model to limiter Q MyLimiter va If you do not put the file in the directory from step 1 then you should specify a full path drive and directory Note GENESYS also searches the directory which you saved the workspace If your Verilog A module is only used for a few workspaces you could simply place the model in the same directory as the workspaces 4 GENESYS will compile your Verilog A file If there are errors shown in your Verilog A code fix them and repeat step 3 5 When the model successfully compiles a Compiled subdirectory is created in the same directory as the Verilog A source file GENESYS creates compiled model library cml file and an XMLmodel xml file for each Verilog A file 46 Advanced Modeling Kit Customizing Built In Nonlinear Models GENESYS supplies Verilog A source code for most of the built in nonlinear models This allows you to create models identical to
403. lator Analysis Options Geners Calodete Censor Hole Advanced Calculate Nonlinear Noise E Input Port Cuitput Port Noe Frequency Bancdesdth ois Frequency Sweep Sweep Type Minimum Frequency Ue Pii Maximum Frequency PED cw Number of Points per Decade Improve accuracy of Low Frequency Oscillator Phase Noise Polishing Ciuha 9 All carriers CO Carrier with Index Vector Calculate Noise Contributors After the oscillator simulation Harbec creates a set of dataset variables for the oscillator noise analysis Name Desctiptio Dependence Data Size Units Data n Type FSPNOISE Frequencies lt indep gt 2 Nearriers Npoints Freq real of Noise MHz spectrum PSPNOISE Noise FSPNOISE 2 Ncarriers Npoints dBm real sidebands power spectrum FNOISE Noise lt indep gt Npoints Freq Hz real frequencies CNOISE Noise FNOISE Npoints 4 Nearriers Volts 2 Hz complex sidebands 274 RPNOISE NAPM NCValue NCIndex Neatriers Nearriers nHBFreqs if noise calculated for all carriers where nHBF reqs is the number of frequencies of the HB solution including DC component Npoints number of noise frequency points used to calculate noise of SSB Neontrib number of noise contributors into output noise all nonzero valued primary correlation matrix Relative FNOISE power of SSB noise Relative FNOISE power of AM SSB noise Relative FNOJISE power of PM SSB noi
404. layers Only the Internal and Ground Reference types can be applied to ports that are on the surface of an object the remaining types can be applied to ports that are connected to an edge of an object Momentum GX If you elect not to assign port types any port in the circuit will be treated as a Single Normal port type during simulation T Port Type Normal Single default No Deembed Internal Differential Coplanar Common mode Ground reference Strips and slots refer to metallization layers For more information refer to Layout Layers All port types can be defined in both of Momentum s simulation modes microwave and RF However de embedding reference offset is not available in RF mode Description The port is calibrated to remove any mode mismatch at the port boundary Single ports on slot layers have polarity Unless a port is given another type it will be treated as a single port Empower Noncalibrated single port Momentum GX Calibrated non grouped single port with 0 reference offset Non grouping is used to exclude from Momentum analysis grouping single port coupling effects The port is not calibrated It is useful for making a connection with lumped elements or for representing other connections in the circuit Two ports with opposite polarity The port pair is simulated as a single port Two ports with opposite polarity The port pair is simulated as a single port
405. layers defined in the hole properties For example we used the Layout tool Draw viahole including pads to define the via 336 Momentum GX e aag A Viahole Properties locaton 318 s o mi 4 Pad Shape PartList Sy Layout o Round Pad Diameter a Square Rect Wagon Wheel Use Default Layers Start Layer IB metal 1 End Layer E metal 2 To create a via for Momentum analysis you may use any of the methods described above Keep in mind e Methods 1 and 3 do not need require creating an additional Via layer with a Z directed current e Methods 1 and 2 may be used to create a via with an arbitrary shape e Method 1 drawing in the dielectric layer is not compatible with Empower and Sonnet simulators Thick Conductor Modeling Simulations with automatic expansion require more simulation time and memory but result in more accurate simulation results In Momentum GX simulations when the height thickness aspect ratio is smaller than a factor of 5 the effect of accounting for the finite thickness of the conductors will need to be allowed for The following figure illustrates the internal substrate model when using an up and down expansion for a conductor In both cases an extra dielectric layer is inserted indicated with new in the figure which in the case of an up expansion has the 337 Simulation dielectric properties of the layer ab
406. le name is defined If it is defined the lines following ifdef are included up to the endif directive If the variable name is not defined but an else directive exists this source is compiled The ifdef else and endif directives can appear anywhere in the Verilog A soutce file Examples ifdef Thermal module bjt c b e dt else module bjt c b e endif Note GENESYS does not support predefining a macro as is often done from a command line build You must define any necessary switches within the Verilog A soutce A useful method is to create a Verilog A file that does nothing but define macros and then includes the real Verilog A source In the example shown above that file could define Thermal before including the bjt module D ata Types and Parameters Integer An integer declaration declares one or more variables of type integer holding values ranging from 231 to 231 1 Arrays of integers can be declared using a range which defines the upper and lower indices of the array where the indices are constant expressions and shall evaluate to a positive or negative integer or zero Example integer flag MyCount I 0 63 Real A real declaration declares one or more variables of type real using IEEE STD 754 1985 the IEEE standard for double precision floating point numbers Arrays of reals can be declared using a range which defines the upper and lower indices of the array where the indices are constant
407. lement such as a mixer or frequency multiplier is encountered Spectrasys automatically deals with frequency translation through these elements The individual mixer parameters of Desired Output Sum or Difference and LO Injection High of Low are used to determine the desired frequency at the output of the mixer The Channel Frequency is a critical parameter for Spectrasys since most of the measurements are based on this parameter If this frequency is incorrectly specified then the user may get unexpected results since many measurements are based on this frequency The easiest way to verify the Channel Frequency that Spectrasys is using is to look at this measurement in a table or the dataset Offset Channel Frequency OCF The Offset Channel Frequency and Offset Channel Power are very useful measurements in Spectrasys These measurements give the user the ability to create a user defined channel relative the main channel The user specifies both the Offset Frequency relative to the main Channel Frequency and the Offset Channel Bandwidth As with the 499 Simulation 500 Channel Frequency measurement Spectrasys automatically deals with the frequency translations of the Offset Channel Frequency through frequency translations elements such as mixers and frequency multipliers Both the Offset Frequency and the Offset Channel Bandwidth can be tuned by creating a variable for each of these p
408. length in mm diagonally across the circuit F the maximum frequency in GHz The following expression provides a guideline up to which frequency the substrate is electrically small for surface wave radiation Ihickness lt Wavelength aE 1 20 with groundplane 10 without groundplane Il Freg cud lt 300 afe l Tm a During a simulation RF mode calculates the maximum frequency up to which the circuit is considered electrically small and displays that value in the status display This is similar to using the expressions above since the dimension and thickness of a layout is typically fixed and it is the simulation frequency that is swept 389 Simulation 390 W Simulation Status 2unning Momentum press the Stop button to end calculations Substrate valid Mesh generation at 2 875GHz Mesh generation Finished S parameter simulation 5 parameter simulation started Fast View Rate Layout is electrically small below 7 6 GHz space wave radiation D Redraw after Substrate is electrically small below 5 34 GHz surface wave radiation 3 each sweep Geometrically Complex Circuits The mesh generated for a simulation establishes a geometric complexity for a circuit A circuit is considered geometrically complex if its shape does not fit into a uniform rectangular mesh and the mesh generation produces a lot of triangles Layouts containing shapes such as circles arcs and non rectangular po
409. llization layers 120 Cx in zo C4 a en E bod n 86 20 E n E0 Tn 140 i180 Qf 0 45 1 0 1 5 al X 7 5 J i freq GHz Treg bHz Figure 12 Magnitude and phase of S21 The thicker line is Momentum results the thinner line is the measurement Internal Ports with a CPW Structure Figure 13 and Figure 14 show the layout for a CPW step in width structure with an integrated sheet resistor 45 m x 50 m 50 Six internal ports are added to the metallization layers The six port S parameter model is recombined into a two port model to describe the CPW mode S parameters Different recombination schemes are possible yielding different simulation results Scheme c is the correct one as the ground reference for the internal ports is the sphere at infinity and this cannot act as the ground reference for the CPW ports subatrate layer stack metallization layer Tee Sheet resistor Figure 13 CPW step in width with integrated sheet resistor 386 Momentum GX i 4 i irs Lio vn a ii i Fa EG i foe k m EEE cock ono a D CRI Dp mr EEE J es i EE Tr P a GT LA i n iI p pc i Le A CCE cou i dx 4 eel p J 5 10 1 20 725 30 35 ao Treg CHF Figure 14 CPW Step in width structure with integrated sheet resistor port configurations and results The process of adding extra internal ports to the ground metallization patterns and recombining the ports in the correct way can be automated in
410. lly added by the compiler as another variable to be solved and the matrix entries to support this Modified Nodal Analysis relationship are also added automatically Line 10 end Ends the analog equations started at line 8 Line 11 endmodule Ends the resistor module started at line 3 Other commonly used features not shown in this simple example include local variables equations and if then statements See the Verilog A examples or the Verilog A reference section of this manual Advanced Modeling Kit Verilog A Reference Verilog A Reference O verview This manual does not give a complete technical reference to Verilog A Rather the objective is to give a model developer enough details to implement complex models without being weighted down with syntax charts and excessive details To purchase a complete reference to Verilog A contact Accellera at www accellera otg Preprocessor The preprocessor supports certain directives in order to simplify code development These directives atre vety similar to their C counterparts Include The include directive is used to insert the entire contents of a source file during compilation The include can be used to simplify code by including global definitions or without repeating code within module boundaries The compiler directive include can be specified anywhere within the Verilog A file The filename is the name of the file with either the full or relative path to be included in the
411. ls which use frequency in their definitions such as transmission lines and INDQ CAPQ inductor and capacitor with frequency dependent Q Aggregate models or user models which include either of the above Frequency domain sources such as LAC and PAC are not considered frequency dependent models These sources all have direct straightforward time domain equivalents like v sin wt Charge dependent elements such as ideal inductors and capacitors are not considered frequency dependent CAYENNE can directly simulate models with an impedance ot admittance of the form R jwX where R and X are constant an w is the frequency b Time dependent models include i ii iii iv Models which use the TIME variable in equations which define their parameters Internal models which use time or delay in their definitions Currently the only models in this category are nonlinear transistor or Verilog A models which contain delay Ageregate models or user models which include either of the above Time domain sources such as IPWL and VPULSE are not considered time dependent models These sources all have direct straightforward frequency domain equivalents using Fourier transforms c Nonlinear models include elements like diodes and nonlinear transistors gummel poon BSIM etc 73 Simulation 74 2 Ifa modelis frequency dependent but is also either nonlinear or time dependent CAYENNE gives a warning a
412. ltage amplitude Vprobe step of curve tracing iterations Default 0 05 V Number of Final Iterations maximal number of adjusting voltage amplitude Vprobe terations finding Vprobe when Re Iprobe Vprobe lt AbsCurTol Default 5 Note The adjusting Iterations are performed after the tracing algorithm found Vprobe amplitude at which the Re Iprobe Vprobe function changes its sign at the next curve tracing iteration step Next See HB Oscillator Usage and Theory Harbec Nonlinear Noise 264 Nonlinear Noise O verview The nonlinear noise options in the Harmonic Balance simulator Harbec enable you to calculate Nonlinear spot noise e Swept noise If you ate not familiar with the harmonic balance simulator refer to Harmonic Balance Basics before continuing with this chapter Refer to the following topics for details on harmonic balance for nonlinear noise simulation HARBEC Harmonic Balance Analysis Performing a Nonlinear Noise Simulation describes the minimum setup for calculating noise Dasic Concepts is an overview on how nonlinear noise is calculated in a simulation Oscillator Noise Simulation focuses on setting up noise simulations fot oscillatots Mixer Noise Simulation focuses on noise simulations for mixers Performing a Nonlinear Noise Simulation Use this type of nonlinear noise simulation for spectral noise simulations of circuits For a successful analysis observe the following 1
413. lygons usually result in meshes with many triangles A measure of increasing geometric complexity is when the ratio of triangular cells to rectangular cells grows larget Complexity Triangles Rectangles The mesh reduction technology offered by RF mode and Microwave eliminates electromagnetically redundant rectangular and triangular cells This reduces the time required to complete the simulation thus increasing efficiency This mesh reduction algorithm can be turned on or off using a toggle switch Higher order Modes and High Frequency Limitation Since Momentum does not account for higher order modes in the calibration and de embedding process the highest frequency for which the calibrated and de embedded S parameters are valid 1s determined by the cutoff frequency of the port transmission line higher order modes As a rule of thumb for microstrip transmission lines the cutoff frequency in GHz for the first higher order mode is approximately calculated by Cutoff frequency fc 0 4 Zo height where Z0 is the characteristic impedance of the transmission line For a 10 mil alumina substrate with 50 ohm microstrip transmission line we obtain a high frequency limit of approximately f 80 GHz Parallel Plate M odes In the region between two infinite parallel plates or ground planes parallel plate modes exist Any current flowing in the circuit will excite all of these modes How strong a mode Momentum GX will be excited
414. m AM 123 205mm U23 2 19 T WD Otm d L 0 595mm LR 1 Xx Port TLIZ Q0 We0 403mm fatx D L 9 142mm 2 Tus WhO 03mm tax L 0 595mm LR TLIO WO 861mm At L 3 205mm L23 Ve 4mm fatta L 0 505mm LR 2 TLI Dis17 TLS Wet 403mm Max W2 639mm WiLesd Didi L 3 117mm L51 L 8 31mm Lead Tu O rren MaMa L 0 506men LR WO 351mm Min L 1 606mm L71 Momentum needs all pieces of layer metal to have closing contours which is not satisfied for this layout As a result Momentum can not calculate the mesh for it and fails But as Momentum creates the pre mesh With all hidden layers it looks like this Momentum GX which clearly shows that the pre mesh does not have the right contours The topology may be fixed by changing the layout drawing or the substrate parameters In this example it is fixed by increasing the dielectric constant Er by 2 times and decreasing dielectric height H by 2 times D efining Mesh Parameters for a Layout Layer You can define mesh parameters that affect the objects on a single layout layer If you have global parameters set in an analysis they will be overridden by the mesh parameters defined for a layout layer To set up layout layer mesh parameters 1 Open the Layout Properties dialog double click the layout window or its use the right click pull down menu command for Layout Properties 2 Select the Layer tab 3 Sele
415. m Won t Work To remedy this additional steps are needed to determine an out of band intercept point 454 Spectrasys System To solve this problem a virtual tone can be determined which is the power of the tone at the DUT input plus the in channel cascaded gain Once the virtual tone power and the in channel gain are determined then both input and output intercept points can be found NT A Fais tone dBm DUT Gain dB Fane in dam Fina dBm Figure 6 Out of Band DUT Output Spectrum Virtual Tone In the lab measuring the in channel cascaded gain and intermod power is a two step process First the in channel gain is measured using a single signal generator as shown below 455 Simulation 456 signal senerator Test Sig Spectrum Analyzer ANAN a1 f rx lest fi fa Figure 7 Out of Band Intermod Measurement Setup Step 1 Next the in channel signal generator is disabled and the two tone generators are enabled and the in channel intermod power is measured as shown below Spectrum Generator Analyzer Tone 1 Generator Tone 2 o f f f 25 OF f rx fest f fa Figure 8 Out of Band Intermod Measurement Setup Step 2 Spectrasys System Be aware that it is very difficult in practice to measure intermod power levels near the noise floor of the receiver One common technique is to measure the S N ratio at the receiver output given a known on channel signal generator
416. m an existing schematic e Tuning with Sonnet data e Using lumped elements with Sonnet Note This is the same circuit as was used in the EMPOWER tutorial In GENESYS select Open Example from the File menu Then select Tuned Bandpass wsp from the Filters directory Double Click F2000 under Designs in the Workspace Window to display the schematic for this filter shown below C2000 pF i E C2000 pF This is the schematic of a 2nd order microstrip combline bandpass filter with 50 Ohm terminations and transformer coupling on the input and output The lumped capacitors are gang tuned to adjust the resonant frequency of the two center lines Tuning in this manner affects only the center frequency and keeps the passband bandwidth constant Double Click Layout under Designs in the Workspace Window to display the layout for this schematic The layout for this example is shown below 408 Sonnet Interface A 0402 Chip Capacitor footprint has been used for each of the lumped capacitors Whenever a lumped element is used for an Sonnet run GENESYS creates an internal ports for the element These ports are placed e If Use Planar Ports for two port elements is checked in the EMPOWER properties box one port is created for 2 terminal elements like resistors or capacitors which are aligned horizontally or vertically e Inal other cases an internal via port is used for each terminal of the element This port is placed at t
417. mal distribution of samples over the frequency range and this can produce poor simulation results 369 Simulation To remedy this situation either run the Adaptive sweep plan first or make sure that the frequency ranges of the two plans do not overlap Viewing AFS S parameters When viewing the S parameters in an AFS dataset if you zoom in you may see small unexpected spurious ripples or oscillations with an amplitude less than 0 0002 The amplitude of these oscillations is always less than the AFS accuracy level which is approximately 60 dB Generally these ripples will only appear if the dynamic behavior of the S parameters is limited for example an S parameter that is nearly 1 over the frequency range simulated This is likely due to the rational fitting model have too many degrees of freedom and being too complex for this situation Momentum Cosimulation Any Momentum GX simulation performs these two internal steps 1 EM analysis of the design board producing the raw parameters of the board Sraw Nports a Nports a Nfteqsraw S parameters of the board ZPORTrtraw Npotts aw Nfreqs as reference impedances of the board ports Fraw Nfreqsraw array of the Momentum analysis frequencies Yraw Npotts aw Nports raw Nfreqsraw Y parameters of the board converted from Sraw and ZPOR raw Nfreqs a number of Momentum analysis frequencies Nports aw number of all board ports Npotts a Nportsexrt Npo
418. me as RX_IIP and MDS Stage Dynamic Range SDR This measurement along the specified path as shown by SDR n SOPIDB n TNP n dB where n stage number This simple measurement shows the difference between the 1 dB compression point of the stage entered by the user and the Total Node Power at the stage output This measurement is extremely useful when trying to optimize each stage dynamic range and determine which stage that will go into compression first See the Stage Output 1 dB Compression Point and Total Node Power measurements to determine which types of signals are included or ignored in this measurement Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as TNP Travel Direction Same as TNP Stage Gain SGAIN This measurement is the stage gain entered by the user For behavioral passive models the insertion loss parameter is used When a stage doesn t have either a gain or insertion loss parameter 0 dB is used This measurement is not dependent on the path direction through the model For example if the path was defined through backwards through an amplifier the forward path gain would be reported not the reverse isolation of the amplifier Channel Used No channel is used for this measurement Types of Spectrums Used None Travel Direction N A Stage Noise Figure SNF This measurement is the stage noise figure entered by the user For behavioral passive models t
419. measurements are complex functions of frequency The frequency range and intervals are as specified in the Linear Simulation dialog box The voltage gain Ei is the ratio of the output voltage Vj to the input voltage Vi Ej Vi Vi Note that due to reflections the gain Ej may not be unity Values Complex matrix versus frequency Simulations Linear Default Format Table dB angle Graph dB Smith Chart none Noise Measure NMEAS The Noise Measure measurement is a real function of frequency and is available for 2 port networks only The noise measure is defined in terms of the noise figure NF and maximum available gain GMAX as NMEAS NF 1 1 1 GMAX The noise measure represents the noise figure for an infinite number of networks in cascade Values Real value versus frequency Simulations Linear Default Format Table mag Graph mag Smith Chart none Commonly Used Operators Operator Description Result Type Getting Started db NMEAS noise measure in dB Real mag NMEAS magnitude of the noise measure Real Examples Measurement Result in graph Smith chart Result on table optimization or yield NMEAS mag NF mag NF db NMEAS magnitude of the minimum noise measure magnitude of the minimum noise measure Not available on Smith Chart Noise Figure NF Minimum Noise Figure NFMIN The Noise Figure measurements are real functions of frequency and are available for 2 port ne
420. minant stripline wave propagates in the structure The animation is a simple set of snapshots for the subsequent time moments The time will vary between zero and the period of the incident wave 1 f seconds The previous example illustrates the propagation of the wave For simple evaluation of the high and low current density region the time average values of the current density is more practical To obtain this plot switch to Magnitude mode by clicking on the Real button The viewer in this mode is shown below The results are as expected for a transmission line segment The current density is highest at the edges and lowest in the middle Note that the absolute values of the current density at the edges are greatly affected by the grid 171 Simulation 172 cell size used A smaller grid cell size increases the edge current density However integrated values of current density are nearly invariable as they should be Mexiner 1972 If the exact current density values are required we recommend choosing a grid cell size equal to the metalization thickness empower Viewer Y6 5 OL x File View xY a Mag Sold Freq GHe 15 1al alel ofa Top Front Side Oblique To investigate the various current components you may switch from the XY mode to the X mode XY X Y Z button You see only a small change in the graph because the current flows prima
421. mplitude Iterations Min Amplitude Factor ra Max Diagonal Jacobian Iter m Max Full Jacobian Iker Max Chords Jacobian Iter Jacobian Full Jacobian calculation method Use Chords Method Slope Relative error Jacobian recalculation Reuse Jacobian at most Calculate Jacobian Numerically 1e 6 le 3 Residual Norm Type _ weighted Norm Reduce Internal Nodes Mane Refer to Optimizing Simulation Performance in the User Manual for details on the convergence process and the use of the parameters described below Max Newton Iterations maximuml number of Newton solver iterations to achieve convergence Max 1 D Subiterations maximum number of 1 dimensional improving solution sub iterations at each Newton iteration Max Source amplitude Iterations maximum number iterations of source amplitude factor continuation iterations Min amplitude factor minimum value of source amplitude factor 231 Simulation 232 Max Diagonal Jacobian Iter maximum number of newton iterations with diagonal Jacobian method before changing it to Full Jacobian Method Max Full Jacobian Iter maximum number of newton iterations with Full Jacobian method before changing it to Diagonal Jacobian Method Max Chords Jacobian Iter maximum number of newton iterations with chords method without updating Jacobian Full Jacobian Method Auto defines autom
422. ms This line includes the definitions for electrical nodes among other things and should be the first line of most Verilog A files Note the use of the symbol Iz zs not a normal 47 Simulation 48 apostrophe On most keyboards it is located on the upper left key the same key as the tilde Line 3 module resistor p n Declares the start of a module named resistor with two external terminals p and n These terminals are used in order by GENESYS so p becomes pin 1 and n becomes pin 2 in the symbol Line4 inout p n Declares that these ports are input output potts Line 5 electrical p n Declares that these nodes are electrical If internal nodes are needed they should be added to this line Line 6 parameter real r 50 from 0 inf exclude 7 Declares model parameter r with a default value of 50 This value can range from greater than zero using opening parenthesis to indicate zero is not allowed to infinity Infinity is a legal value since square bracket was used The value 7 is specifically excluded Line 7 analog Header for the analog equations Required in all files Line 8 begin Starts the actual analog equations Often this is combined with analog on one line analog begin Line 9 V p n lt r I p n Adds a voltage due to the resistor V IR V p n is the voltage from node p to node n I p n is a branch current flowing from node p to node n Note This branch current is automatica
423. ms associated with transmission systems 36 Although the characteristics of transmission systems are defined by simple equations prior to the advent of scientific calculators and computers evaluation of these equations was best accomplished using graphical techniques The Smith chart gained wide acceptance during the development of the microwave industry It has been applied to the solution of a wide variety of transmission system problems many of which are described in a book by Philip Smith 37 The Smith chart as displayed by GENESYS is shown in below Labels for normalized real and reactive components are displayed when the level of detail permits it 299 Simulation 300 The design of broadband transmission systems using the Smith chart involves graphic constructions on the chart repeated for selected frequencies throughout the range of interest Although the process was a vast improvement over the use of a slide rule it is tedious Modern interactive computer programs with high speed tuning and optimization procedures are much mote efficient However the Smith chart remains an important tool for instructional use and as a display overlay for computer generated data The Smith chart provides remarkable insight into transmission system behavior The standard unity radius impedance Smith chart maps all positive resistances with any reactance from to onto a circular chart The magnitude of the linear form of S11 or S22 is the lengt
424. mulation 340 Momentum Options General Simulation Options pesh Simulation Mode BF Faster no radiation effects Calculate Substrate Q Mesh and substrate C5 A substrate mash and S Parametars 3D Metal Expansion 9 None thin metal Oup CO Down Vis Mite C Lumped E 1D wire CO 2D planar no horizontal currents def aui CO 3D spatial include horizontal currents Cl Use Box C include TL calculation Use horizontal side currents thick conductors Reuse results of last simulation Modeling Horizontal Side Currents for Thick Conductors Horizontal side currents for thick conductors can be modeled automatically by enabling Use horizontal side currents in Momentum Analysis properties Simulation Options tab When the Horizontal side current toggle is set the horizontal current components will be added on all conductors which have been specified as automatically expanded This results in a modeling of the current as illustrated on the following figure Boxes Momentum GX horizontal top currents t vertical side currents horizontal side currents A eJ ae w a Eon eel Sen Mons oe ta EE H k o d ae FF I Horizontal current components are not added on regular vias or when users manually expand finite thickness conductors For backward compatibility reasons by default the new horizontal side currents toggle is not set
425. mulation simulation example factots etc parasitics etc oscillator startup In determining which simulation type to use several points should be considered Getting Started Linear or Electromagnetic 1 Should I use both circuit theory and EM simulation Circuit theoty simulation in GENESYS 1s amazingly fast and interactive No other program at any price approaches the speed of GENESYS EM simulations are more accurate and do not require the use of specific geometric objects for which circuit models have been developed EM simulation complements rather than replaces circuit theory simulation 2 Whatis the highest frequency used in the circuit If below about 1 GHz lumped elements are often used in place of distributed elements In this case the final board layout usually won t add any significant parasitics or coupling concerns Often however customers use EMPOWER or Momentum to simulate the final board layout to make sure that it doesn t differ from the linear simulation 3 How big is the circuit If the circuit itself is very small compared to a wavelength at the highest frequency of concern electromagnetic simulation may not be needed This is because resonances occur at quarter wavelengths and circuits much smaller than this usually behave as predicted by a complete linear simulation 4 Does the circuit have non standard metal shapes patterns or geometries If so electromagnetic simulation may be the only optio
426. mulation speed of Spectrasys Spectrasys System NOTE For additional understanding of why these parameters affect the speed of the simulation see Propagation Basics Loops The larger the gain around a closed loop the longer it will take for spectrums to fall below the Ienore Spectrum Level Below threshold Furthermore more data is collected each time spectrums are propagated around a loop A quick test to verify if this is the problem with the simulation is to increase the isolation of one of the main elements in the loop to a very high value like 200 or 300 dB This will force loop spectrum to fall below the Tgnore Spectrum Level Below threshold Ignore Spectrum There are 3 Ignore spectrum parameters that affect the simulation speed and the amount of data collected They are 1 Ignore Spectrum Level Below 2 Ignore Spectrum Frequency Below and 3 Ienore Spectrum Frequency Above These parameters can be changed to reduce the frequency range and amplitude dynamic range for which the simulator is collecting and analyzing the data The biggest speed improvement usually comes from raising the Ienore Spectrum Level Below threshold Tight Loops The more nodes and elements in a design the more spectrums that will be created and propagated Linear elements formed in tight loops should be moved to a subcircuit and called from the parent design If only linear elements exist in the a subcircuit a linear analysis is used fo
427. n EM can simulate any arbitrary shape such as ground plane pours A linear simulator requires a netlist ot schematic to describe the circuit so models would have to exist for the pattern that you plan to simulate 5 Do any of the models in the circuit exceed or come close to exceeding the published parameter ranges If so you may want to verify the simulation with EMPOWER or Momentum or use EM exclusively Most of the models in GENESYS were detived from measured data which was only taken for particular parameter variations Linear or Harmonic Balance This question is the easiest to answer for active circuits you will usually use both For passive circuits filters couplers power dividers etc you will only use linear Passive circuits are linear harmonic balance will not give you extra information that you could not get from linear simulation Active circuits are inherently nonlinear Harmonic balance will help you analyze DC operating points and nonlinear performance For both active and passive circuits linear simulation is the workhorse of RF design Matching noise and stability studies are all completed quickly using linear simulation Harmonic balance is used to complete the analysis of most circuits Examine mixer conversion gain amplifier compression and detector efficiency using harmonic balance Linear or CAYENNE Transient Often this question does not have a quick answer For example many engineers associate CAY
428. n Levels of Data Save Data For Splitter 3 CheckAl Mixer J Filter J Uncheck All Output E T Check Gutpuk Ports GSM Ose 4 Check Input Ports Ay Factory Defaults Parameter Information Retain X Levels of Data Specifies the number of data levels that will be saved to the dataset For example 1 level of data is the data for the top level design only This parameters only refers to the data retained in the dataset and not the whether a subcircuit will be analyzed ot not All subcircuits will always be analyzed SAVE DATA FOR 955 Simulation 556 This section controls the data that will be saved to the system analysis dataset Each part in the schematic is listed and can be checked or unchecked When checked the data for all nodes of the specified part will be saved Several buttons have been added for convenience is selecting or unselecting parts Check All When clicked with select all components Uncheck All When clicked will unselect all components Check Output Ports When clicked will select only the output ports Check Input Ports When clicked will select only the input ports Index AP GD Joep errr Ca a Merc Ice TT 292 Apsolote Current bh Ole rane i eec t 229 absolute CERE Co no E Le sede 225 Absolute Numeric Derivation Step 225 absolute tolerance 225 234 Absolute Voltage Tolerance 225 PWS TO E AA 86 ACCURACY Eod A 97
429. n of an analog block e Events can be detected using the operator e Events do not hold any data e There can be both digital and analog events There are two types of analog events global events and monitored events Null arguments are not allowed in analog events cross function The cross function is used for generating a monitored analog event It is used to detect threshold crossings in analog signals when the expression crosses zero in the direction specified cross can control the timestep to accurately resolve the crossing The format is ctoss expr dir time_tol expr tol where expr is required and dir time_tol and expr_to ate optional arguments The dir argument is an integer expression the other arguments are real If the tolerances are not defined they are set by the simulator If either or both tolerances are defined then the direction of the crossing must also be defined The direction can only evaluate to 1 1 ot 0 If it is set to O or is not specified the event and timestep control will occur on both positive and negative signal crossings If dir is 1 or 1 then the event and timestep control occur on rising edge falling edge transitions of the signal only For other transitions of the signal the cross function will not generate an event expr_ o and lue fol represent the maximum allowable error between the estimated crossing point and the actual crossing point Examples The followi
430. n of one port data files e Substrates Choosing a substrate causes the layer to get the rho thickness and roughness parameters from that substrate definition We recommend using this setting whenever possible so that parameters do not need to be duplicated Caution Thickness is only used for calculation of losses It is not otherwise used and all strips are calculated as if they are infinitely thin Metal layers have three additional settings available Slot Type Check this box to simulate the non lossless metal areas as opposed to the metal areas in EMPOWER Use this for ground planes and other layers which are primarily metal Do not use this for lossy layers See your EMPOWER manual for details Current Direction Specifies which direction the current flows in this layer The default is along X and Y X Only and Y Only can be used to save times on long stretches of uniform lines Z Up Z Down XYZ Up and XYZ Down allow the creation of thick metal going up down to the next level or cover Thick Metal Checking this box forces EMPOWER to model the metal including thickness EMPOWER does this by putting two metal layers close together duplicating the traces on each and connecting them with z directed currents If thick metal is used then Current Direction is ignored Element Z Ports This setting specifies the default direction for automatically created element ports either to the level above or to the level below G
431. nals are flowing and which signals are desired or not Every pin on a behavioral models serves both as an input and output pin Each pin treats its input spectrum with the behavior appropriate to that pin and spectrum type Linear models resistors capacitors transmission lines etc use a Y matrix to determine the input to output transfer function for each spectrum For linear models the same input to output transfer function is applied to all spectrum types Non linear models amplifier multipliers mixers etc use a Y matrix that is dependent on non linear parameters such as P1dB PSAT IP3 and IP2 Non linear parameters are used to create the behavior associated with the given model input pin output pin and spectrum type Since a Y matrix is used for calculations VSWR effects along a path and at the schematic nodes is automatically accounted for Sources Sources must be placed in the schematic and connected to the device under test before a system analysis can produce any useful data NOTE Thermal noise is automatically added to the system analysis For specific source information of Element Catalog click the following links MultiSource CW Source 423 Simulation CW Source with Phase Noise Wideband Source Multicarrier Source Intermod Source Intermod Source Receiver Continuous Frequency Source Noise Source Channels All measurements in Spectrasys are based on a channel Channels consist of 1 Center Freque
432. nalysis See the Stability Example for usage Note To see stability circles on a smith chard a marker must be placed on the SB1 or SB2 traces The trace is generally located outside the smith chart area Values Complex values versus frequency Simulations Linear Default Format Table center magnitude angle radius Linear Graph none Smith Chart Circle Commonly Used Operators None Examples Measurement Result in graph Smith chart Result on table optimization or yield SB1 input stability circle center magnitude angle radius Linear par SB2 output stability circle center magnitude angle radius Linear par Available on Smith Chart and Table only Parameter indicating the unstable region Optimal Gamma for Noise GOPT The Optimal Gamma for Noise is a real function of frequency and is available for 2 port networks only The optimal gamma is defined in terms of the reference admittance Yo and the optimal value of admittance Yopt as GOPT Yo Yorr Yo Yorr Notice that gamma goes to zero if the reference admittance 1s optimal Values Real value versus frequency Simulations Linear Default Format Table Linear Graph Linear Smith Chart GOPT Commonly Used Operators none Getting Started Examples Measurement Result in graph Smith chart Result on table optimization or yield GOPT gamma coefficient gamma coefficient Optimal Admittance Impedance for Noise YOPT ZOPT T
433. nature Voltage units V access V idt_nature Flux 51 Simulation 52 ifdef VOLTAGE ABSTOL abstol VOLTAGE ABSTOL else abstol 1e 6 endif endnature genvar Genvars are integer valued variables which compose static expressions They are used for instantiating structure behaviorally e g accessing analog signals within behavioral looping constructs genvar ist_of_genvar_identifiers where Zs of genvar identifters is a comma separated list of genvar identifiers Example genvar I j Parameters Parameters provide the method to bring information from the circuit to the model Parameter assignments are a comma separated list of assignments The right hand side of the assignment is a constant expression including previously defined parameters For parameter arrays the initializer is a list of constant expressions containing only constant numbers and previously defined parameters within and bracket delimiters Parameters represent constants their values can not be modified at runtime Parameters can be modified from the declaration assignment at compilation time The purpose is to allow customization of module instances A parameter however can be modified with the defparam statement or the module_instance statement It is not legal to use hierarchical name referencing from within the analog block to access external analog variable or parameter values An example is parameter real Test
434. ncy 2 Bandwidth Channel Example The channel center frequency is 1000 MHz with a bandwidth of 1 6 MHz 999 2 to 1000 8 MHz Only the spectrums located in the yellow region will be integrated by the channel measurements 424 Spectrasys System 10 Out of Channel An In Channel 100 995 998 4 996 5 999 2 9996 1000 1000 4 1000 6 1001 2 1001 6 1002 Frequency MHz There are several different channels used in Spectrasys 1 Main Channel of the Path 2 Offset Channel 3 Interferer Channel 4 Adjacent Channel 5 Ist Mixer Image Channel NOTE The bandwidth for ALL channels except the Offset Channel is the Channel Measurement Bandwidth Main Channel of the Path Most of the measurements are based on this channel Each new path can have a new center frequency however the bandwidth for all paths will be identical Offset Channel This is a user defined channel This channel is specified as an offset from the main channel of the path Click here for additional information on specifying this channel 425 Simulation 426 Interferer Channel This channel is used to determine input and output intercept points of a path This channel frequency is specified as an absolute value Click here for additional information on specifying this channel Adjacent Channel This channel is adjacent to the main channel of the path It is provided as a convenience to the user 1st Mixer Image Channel This channel is used to make ima
435. nd ignores the frequency dependency 3 Otherwise If a model is frequency dependent CAYENNE checks the response as set in the Convolution tab in the Accuracy Testing section By default this does two things a If the only frequency dependency is due to loss and the Always Use Constant Loss box is checked then CAYENNE uses the impedance or admittance at the Most Accurate Frequency specified on the general tab to calculate equivalents of the form R jwX for all matrix entries for that element This allows elements like INDQ and CAPQ plus all elements like WIRE which are based on RLOSS or GLOSS models to avoid convolution Otherwise the response for the element is calculated over a range of frequencies as specifiedin the Accuracy Testing section 11 points from DC to twice the Most Accurate Frequency by default If the R and X values for the impedance or admittance in the equivalent of the form R jwX for any matrix entry varies by more than the tolerance 0 01 1 by default then convolution is used Otherwise the approimxation accuracy is acceptable and the R jwX equivalent at the Most Accurate Frequency is used Note that this approximation can even be used for measured s parameter data if there is not much variation with frequency Also note that transmission line responses are never in this form they are periodic and must always use convolution 4 Ifan element needs to use convolution then the settings in the
436. ne GainSS as gain at lowest power of the sweep GainSS Gain 30 dBm In Genesys 2005 and later the analyses functions have been changed To use analysis functions in a sweep The variable defined by a function must be defined in the analysis dataset right click in the dataset and choose add new variable The checkbox Propagate All Variables When Sweeping of the sweep properties dialog window must be set There are two methods to calculate the gain compression in G5 1 By defining Gain in the swept analysis dataset directly or using HB functions and propagating it to the sweep dataset 2 Directly by calculating it from parameter sweep data HARBEC Harmonic Balance Analysis The ist technique Add a new variable Gain to the HB1 analysis dataset HB1_Data Gain db10 hb_transgain P1 P2 FreqIndexIM 1 1 Note Also we can add the complex voltage gain variable if we need to know it to know information about phase GainV hb transgain VPORT 1 VPORT 2 FregIndexIM 1 1 Note the added variables in the data set below HB1 Data SayFreq Z3 FreqindexIM fay Gain 663 db 10 hb_transgain P 1 P2 FregIndextM 1 1 Fy Gainy 2 14212 4 j1 11808 hb transgain VPORT 1 VPORT 2 FregIndexIM 1 1 FAP 1 PPOAT I P2 PPORT Z FSYPPORT Now check the box to propagate all variables 1n the sweep 241 Simulation Parameter Sweep Properties
437. ne is again even pattern To excite the odd mode as an example select Generate viewer data and enter 2 in the Mode Number to Excite box of the EMPOWER properties dialog Run the viewer A snapshot of the calculated current density function is shown here You must turn Absolute Value Display off View Menu Switches All settings except two are the same as in the previous example The initial view was set to the side view View Menu Side or Side button and the polygon view was set to witeframe View Menu Switches Wireframe or Solid Wire button 174 EMPOWER Planar 3D EM Analysis empower Viewer V5 5 L GI x File View xy e Real Wire Freq GHe t 1a late ts ot To Front Side Oblique The plot confirms that this is an odd mode and shows the typical current density distribution If currents on the left strip flows in the forward direction the currents on the right strip flow in the backward direction and the center strip currents flow in opposite directions at the opposite strip sides For a dynamic view turn on the animation and rotate the plot for a better view of the propagating wave To calculate the viewer data for the other eigenwaves run EMPOWER and the viewer twice more with Mode Number to excite set to 1 and 3 Note that newly calculated data will overwrite the previous ones To avoid this and to keep viewer data for all excitation experiments you need to save a copy of the existing worksp
438. ne to edit them by hand These files should be very useful in other applications as the engineers at Eagleware used third party applications to graph currents before our EMPOWER viewer was completed R1 R2 Rn Port Impedance Files Written by EMPOWER Type Text Can be safely edited Yes Average size 1Kbyte Use Read by GENESYS when Generalized S Parameters are requested These files contain each port s impedance versus frequency These ports are read by GENESYS if the keyword GEN is used in place of a termination impedance The files are formatted just like RX files in GENESYS GENESYS always requests these files when EMPOWER is run from GENESYS Notes These files are numbered differently than Ln files When these files are numbered each port in a related group of ports is counted individually RGF Line D ata Files Written by EMPOWER Type Binaty Can be safely edited No Average size 1 to 5Kbytes but may be larger Use Internal file for EMPOWER but can also be used in the SMTLP and MMTLP models in GENESYS These files are used in place of Ln files if a filename was given on the PORT line in the TPL file When run from GENESYS this file type is not available use the Ln files instead Otherwise they are completely identical to the Ln files described earlier RX Frequency vs Impedance Files Written by User Type Text Can be safely edited Yes Average size 1 Kbyte Use Specifying electrical
439. necessary to set up mesh parameters default parameters can be used instead The more cells the more accurate the simulation will be but too many cells will slow down a simulation and provide little improvement in accuracy You can pre compute the mesh before simulating in order to view the mesh otherwise the mesh will be computed as part of the simulation process A mesh is required in order to simulate If you choose to define mesh parameters you can set them for e The entire circuit e The objects on a layout layer It is not necessary to specify parameters for all levels You can for example specify mesh parameters for a single layer and use default values for the rest of the circuit For more information about how a mesh is generated refer to About the Mesh Generator For suggestions to consider when setting mesh parameters refer to Guidelines for Meshing D efining Mesh Parameters for the Entire Circuit The global mesh parameters affect the entire circuit They ate specific to the current Momentum analysis To set up global parameters open the Momentum Options dialog gt Mesh 347 Simulation Momentum Options General Simulation Options Mesh Mesh Frequency Mesh Density cells wavelength Transmission Line Mesh Oo Mone O Automatic Edge Mesh Q Mone Automatic Include thin layer overlap Reduce Mesh Refacet Arcs 1 The global parameters are displayed 2 Enter the mesh
440. nes is computed by the simulation of a calibration standard and subsequently removed from the S parameter data A built in cross section solver calculates the characteristic impedance and propagation constant of the transmission lines This allows to shift the phase reference planes of the S parameters a process called de embedding Results of the calibration process includes the elimination of low order mode mismatches at the port boundary elimination of high order modes and removal of all port excitation parasitics Besides transmission line ports Momentum offers the user the ability to define direct excitation or internal ports These ports can be specified at any location on the planar metallization patterns as either a point or a line feed They allow to connect both passive and active lumped components to the distributed model of the planar circuits The S parameters associated with these ports are calculated from the excitation consisting of a lumped source connected to the equivalent network model at the locations of the internal ports The parasitic effects of these lumped sources are not calibrated out of the S parameters results Reduced O rder Modeling by Adaptive Frequency Sampling A key element to providing fast highly accurate solutions using a minimum of computer resources is the Adaptive Frequency Sampling AFS technology When simulating over a large frequency range oversampling and straight line interpolation can be used to
441. ng description of a sample and hold illustrates how the cross function can be used module sample and hold in out sample output out input in sample electrical in out sample real state analog begin Q cross V sample 2 0 1 0 state V in V out lt transition state 0 10n end endmodule 59 Simulation 56 The cross function maintains its internal state It has the same restrictions as other analog operators in that it can not be used inside an if case casex or casez statement unless the conditional expression is a genvar expression Also cross is not allowed in the repeat and while iteration statements but is allowed in the analog for statements timer function The timer function is used to generate analog events It is used to detect specific points in time The general form is timer start_time period time tol where start_time is a required argument but period and time tol are optional The timer function schedules an event to occut at an absolute time start time The analog simulator then inserts a time point within timetol of an event At that time point the event evaluates to True If time tol is not specified the default time point is at or just beyond the time of the event If the period is specified as greater than zero the timer function schedules subsequent events at multiples of period Examples A pseudo random bit stream generator is an exampl
442. ng met the conditions for gain phase and match the next step is to verify oscillator performance and accurately determine the frequency of operation power delivered to a load and the harmonic content To accomplish this we use the OSCPORT element along with HARBEC Oscillator analysis We begin by connecting the two ports together to close the loop To initiate or provide start up impetus to our circuit we will insert an OSCPORT component into the circuit The placement is not critical any node is useable however best results are obtained if we do not place it on the output port or node We access the OSCPORT element from the source selection icon on the GENESYS toolbar 253 Simulation 254 EX dsl Data Linear OSCPORT OSCPORT 1 FOSC 1DOMHz VPROBE I 445 Output 4 A TIT PariList f gt Schematic wr From the Simulations Data folder in the GENESYS Workspace window add a DC Analysis simulation to determine the operating point of our device This is recommended prior to any Harmonic Balance simulation From the Simulations Data folder in the GENESYS Workspace Window add a Harmonic Balance Oscillator analysis Accept the default name or choose another Note that the default analysis is for the current schematic The oscillator frequency and note voltage will be filled in by the simulator after a successful run The Harmonic balance dialog establishes a default value for the number of harmonics Increased harm
443. ng the Layout To run Sonnet you must create a simulation Click on the Create New icon in the Workspace Window and select Analysis Add Sonnet Simulation Accept Sonnet1 as the analysis name The Sonnet Options dialog will be displayed as shown below Fora description of the dialog options see the details in the Sonnet Translation chapter For now just set the prompts as shown below Sonnet Interface Sonnet Interface Options General Sonnet Advanced Port impedance Use ports From schematic Necessary For HARBEC co simulation Electromagnetic simulation Frequencies Co simulation sweep an Start Freq MHz Use EM simulation Frequencies Stop freq MHz 11000 Start freq MHz Number of points Stop freq MHz Number of points 31 Use Adaptive Sweep ABS Recalculate Now Automatic Recalculation Automatically save workspace after calc We are starting with 31 sample points in the range 8 11 GHz Click the Recalculate Now button This launches Sonnet to simulate the layout While Sonnet is calculating a window similar to the one in below will be shown This window shows the current status throughout the calculation mode For more details on this window see your Sonnet manuals 405 Simulation genesys son 9000 MHz Running File Edit View Run Project Help Genesys son Freqs 10 Complete Finish Time 8 44 AM 4 sec Subs 121 Memory 1 MB Analyzing 9000 MHz D Red
444. nn naa NA e Mb a Na o acie dd 222 HARBEC Harmonic Balance Analysis eee iee etta reete eee eene aor ne oae aeo ara 225 Hb bfatmonic Balance JATALDSIS eia estet ot tbi E N a A 225 Harmonic balance OVENI C Werer i mi motu mnisodbtuttatea mde n Dm UM eU 225 FARBECIOPUONS MH 225 EDARBEG TOPUR MoM oe perire rere tre mre EID TENT a trito Coe IP tob t hU ero 233 DOW 100 CON VCLOCIICe ISSUCS rosse titudo ud Sepatu stan onion Mir ONL 234 Optimizing Simulation Perl Ormance oue ei obest ton nite Fobu UN e OAM tob NU das Dies 234 Harmonic Balance Analysis P tictlOnso eei oa o ec tdio ieee tice ndt 238 Table 1 Basie HD atalysis Measure ments 4o tei patr e mtt i bt ires 244 HD OScillatot ATAN S S s tare Reais iva bean oa toot babes ihe a Deb iust auiuop fte tba de 249 Oscillator Desiofu Ve ye Wi eoe qeteesteo esie itin EAEN di pat potus 249 IVES AS UTE TI Db AEE T E A A ato T nests ow AE EE EA E S 261 Harmonic Dalance Oscillator CODEOFS pp Va egeo A E E TOA 262 LHatbecNoOnBneat NODE cete eso petant tex do eti ido RON Vtde ene 264 INODOIGOAT NoE OVON IEW ssi nudo sessuali tacet Ent eI dist rcu LL Od 264 Peromnine a Nonlinear NOISE Siti a EOD o cerle duis rt ed tumet Hia Uu 265 lorrirod beisiuooi MERO C 270 Oscillator Nore Slat OM fonder e ttftt etg ca oditaut que e gau dul uir tied 271 Miker Nore SITU IAG EE eost ius o as e Ne Ie Uns RU att MOS Daum tp tat INDE t A 282 Fieri INOonlInear VIO GS Spiraea tobape sevice aniaslscGt a
445. noise figure of the specified chain as shown by MDS n CNP 0 CNF n dBm where n stage number The MDS value at stage n represents the MDS of the entire system up to and including stage n Consequently the MDS of the entire system is the value indicated at the last stage in the path or chain The minimum detectable signal is the equivalent noise power present on the input to a receiver that sets the limit on the smallest signal the receiver can detect For example if the thermal noise power input to a receiver is 174 dBm Hz and the channel bandwidth is 1 MHz 10 Log 1 MHz 60 dB then the input channel power would be 114 dBm For a cascaded noise figure of 5 dB the minimum detectable signal would be 109 dBm See the Channel Noise Power measurement to determine which types of signals are included or ignored in this measurement Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as CNP and CNF Travel Direction Same as CNP and CNF Noise and Distortion Channel Power NDCP This measurement is the integrated noise and distortion channel power in the main channel along the specified path The Noise and Distortion Channel Power is the sum of the Channel Noise Power plus the Total Intermod Channel Power plus the Phase Noise Channel Power as shown by NDCP CNP n TIMP n PNCP n dB where n stage number See the above measurements to determine which types of s
446. ns The Grid All conductive surfaces and ports must be on a grid This grid is composed of regular rectangular cells An example of mapping a microstrip bend to the grid is shown below The left half of the figure shows the circuit as it appears in LAYOUT The right half of the circuit shows a part of the EMPOWER listing file Each of the plus signs in the listing file represents an intersection of two grid lines as shown on the layout Lines connecting plus signs represent metal Numbers represent port locations Notice that the ports map onto the grid in place of metal so the ports go between the end of the line and ground the wall so each port has a ground reference as would be expected lszs4557858898 01253485 lBt t 1 4 4 RF 4 1 lz 4 4 4 4 ll 4 4 GB 4 lll T t Et 4G GGG gd 6B GG G B c4 G6 l l T Tt t4 4 t4 3 L I tt to tt td amp 1 I tt gd P 0g tod 7 L I gd dL gd P 0g td 6 1 dd 5 4 tod 4 E tod 3 4 tod e dd l l Ex x Ofer rr er er er er ee EMPOWER will move all surfaces to the nearest grid cell before analyzing a circuit EMPOWER maps th
447. ns In Band Intermod Path Measurements Out of Band Intermod Path Measurements and Troubleshooting Intermod Path Measurements for additional information Powers Voltages and Impedances When checked will add path powers voltages and impedances to the path These spectrums are grouped spectrum into various categories which are then integrated by corresponding measurements Adjacent Channels When checked will calculate the specified adjacent channels and place the results in the path dataset The bandwidth of the channel is the channel measurement bandwidth specified on the General Tab of the system analysis Channels These are the adjacent channels which reside on either the lower or upper side of the main channel These values are specified as an integer array where values are separated by semicolons Negative numbers represent the Spectrasys System channels lower than the main path frequency and positive values represent channels higher than the main path frequency For example 2 1 1 2 means that first and second lower and first and second upper adjacent channels will be measured Receiver Image Channel When checked all measurements associated with the receiver image channel will be calculated and saved to the path dataset The receiver image channel is defined as the channel from the input to the first mixer Offset Channel When checked will calculate the offset channel power and frequency measurements and add them to
448. nsition time is optional and must be nonnegative zi zp implements the zero pole form of the Z transform filter The general form is zi zp expr z r T 7 70 zi zd implements the zero denominator form of the Z transtorm filter zi_np implements the numerator pole form of the Z transform filter zz nd implements the numerator denominator form of the Z transform filter Contribution assignment statements Sequential block A sequential block is a grouping of two or mote statements into one single statement The format is 57 Simulation 58 begin block_identifier 4 block_item_declaration Y statement end where block_item_declaration is parameter declaration integer declaration real declaration Indirect branch assignment An indirect branch assignment is useful when it is difficult to solve an equation It has this format V n Vp 0 Which can be read as find V n such that V p is equal to zero This example says that node n should be driven with a voltage source and the voltage should be such that the given equation is satisfied V p is probed and not driven Indirect branch assignments are allowed only within the analog block Branch contribution statement A branch contribution statement typically consists of a left hand side and a right hand side separated by a branch contribution operator The right hand side can be any expression which evaluates to or can be promoted to a
449. nt a minimum of two signals tones must be present at the input e first interfering signal e second interfering signal The Channel Frequency must be set to the intermod frequency and the Interferer frequency must be set the first or second interfering frequency See the Calculate Tab on the System Analysis Dialog Box to set the Interfering Frequency Furthermore the spacing of the interfering tones needs to be such that intermods will actually fall into the main channel If these conditions are not met then no intermod power will be measured in the main channel Each column in this measurement is for a different intermod order up to the Maximum Order specified on the Calculate Tab of the System Analysis Dialog Box The column number is the same as the order starting from the left with order 0 Remember intermod bandwidth is a function of the governing intermod equation For example if the intermod equation is 2F1 F2 then the intermod bandwidth would be 2BW1 BW2 Note Bandwidths never subtract and will always add The channel Spectrasys System bandwidth must be set wide enough to include the entire bandwidth of the intermod to achieve the expected results The Automatic Intermod Mode will set the bandwidth appropriately Note Cascaded intermod equations are not used in Spectrasys Caution This method used to determine the intercept point is only valid for 2 tones with equal amplitude Channel Used Interferer Channel
450. ntains only one unit element corresponding to the specified input 177 Simulation 178 The other elements of the vector are zeros Reflected waves vector B are calculated from the equation B S A Then the simulator defines normalized voltages and currents in mode space denormalizes them and restores the grid currents and voltages inside regions corresponding to all input surface current regions Finally using the input region variables the program calculates non zero grid currents Ig for strip like structures or voltages Vg for slot like structures The grid currents and voltages are locally defined model currents and voltages see the Theory section and their units are Amperes and Volts accordingly The grid currents and voltages together with their coordinates on the grid are stored in the EMV file The same data can be written in the self documented text file with the extension PLX The viewer reads the EMV file and to displays data Note that the initial current voltage distribution is a model representation and is treated using complex number conventions The currents voltages are complex quantities and harmonic functions of time So their magnitudes are maximal values for the excited wave petiod The real component corresponds to instantaneous values of currents and their phases reflect the phase delays of currents at the initial time t 0 Using these initial data the current distribution is calculated versus time If f
451. ntermod order up to the Maximum Order specified on the Calculate Tab of the System Analysis Dialog Box The column number is the same as the order starting from the left with order 0 Remember intermod bandwidth is a function of the governing intermod equation For example if the intermod equation is 2F1 F2 then the intermod bandwidth would be 2BW1 BW2 Note Bandwidths never subtract and will always add The channel bandwidth must be set wide enough to include the entire bandwidth of the intermod to achieve the expected results The Automatic Intermod Mode will set the bandwidth appropriately Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used ONLY INTERMODS and HARMONICS separated according to their order Travel Direction All directions through the node Total Intermod Power TIMP This measurement is the total integrated power of all intermod orders in the main channel along the path This measurement differs from the Total Intermod Channel Power in that it is a sum of all the orders of intermods whereas the Total Intermod Channel Power is separated by order Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used ONLY INTERMODS and HARMONICS Travel Direction All directions through the node Total Node Power TNP This measurement is the integrated power of the entire spectrum at the node This is an extremely useful measurement in
452. nys Rea dunt nba ANA 287 Linear Analysis naeioun a a E 289 COVOEVIe Waea T E N eae S 289 Pines Dialysis PODE GS ea n E A E esestoeb antenna 290 SAE mee Eno wise E ees ATE eae ee 292 OPEC arse gece ec ere eg A ent ly 202 S Parameter Dasicooinu ise eio pet diaeta A A 2902 CICLO ee NT 294 IN ce avo AEE cee T Ore eer eee oor Lom A Crea rT A rere rene tren A 296 EAS qo INES a esos atelier aclacln Se a eet nee tor rete 207 The Ula etal E c 297 Simulation Chapter 8 Gat CTT SS aque UMP o NM ets ettet MN ANENE EEA opens 297 PO EE AE On NC et SEP ATE Re 298 aU Na ale wea Sees haod E ec eee asd eulsn cea Li Secon A ese ci uc aei 299 Linear Anabsis COUCDUE deii Pre bate pedale ed ae E edt fecto utque 301 Whore titi OLX spese arinn FEIER ERIPE RE RA NVRETANEE UR TREN TN TREES NIS CREAN DEEST EQ PESE ENTE 305 Momentu OG X ID ASICS aeiiae esiti bibe te Snnt di ipsa eketdsai aite ttp toria isa bm ta iidmuhlus 305 Typical SSP LO CESS courier p boxed betta iau oue Go OO 305 Selecting ME CONC Mod Genen vt aep etta A A T OEA OAE 306 NP DO ec I M II E uM MM MERE RC n CLADE 308 ING Cine an EM Porto CIF UIT etit epu N ulii ute wise tases rs 308 COn TIO EITIODE Soo bts ssutius ei dct pTsD MEM eT USC M rS ED L MED EUIS M DU UE EE 309 Pot C2 DEAHOTU ses ec ines aad cases a eases tercie Ada cot ert aeons eee ies 310 eterno toe Port Type to U SG eocsusacipet telo eta S pid Na RUMP aO SE QS Ped tpe NINE a RM 310 Denning
453. o RO 86 GOP ee ee 8 11 13 Ce N 296 297 GP Eee Mm 11 Green S TUDCEOTL o Maren UR MA 208 AIG RM NUNT EAMUS T 129 159 201 Gid Green s function uuu t ttettesen 214 L E S tos Motos pedes oett 181 GEOUDIO ds LoT I aper Sox cadens 500 Gt296 S et 297 EBE cl e MM 11 SD eae eae 297 GO E e i E 11 Index Fel parabuCte ES seestevae uin IR WEISEN CN repe En ORE 11 FARBE O neat CERES 225 233 234 ELAR BEG ODUODSA 2e eet ttis iei icona 225 HARBEC convergence issues 234 HARBEC optimization ede nete 234 harmonic balance seccina 225 234 Harmonic Balance Analysis Functions 238 Highest accurate reque Cy iones nnan 100 hornos ee OUS ud aA 158 159 206 m A 502 IC AI e e RU aequ ds 510 DI AAE EA A AAN AA III 510 ui M 527 ident hca OTI 20020220 0133 yhoo dedu vseves ee EUER NUS 433 identifying spectral origin uet tetsse 433 TORIS smt RA D MELLE 511 IPAO SDECDEUDT a inii ienai 552 I eTT ER TOTP ODE OT 509 LN Coo CRM HMM A 509 iie CEU n PUNCO ecce nba A a 13 image channel noise powet sss 511 image channel pOWEEco se dedere 511 IMAGE SECQUG IO V eese PUR HEREIN IO Ie Dd 502 image noise rejection fatio 512 IMACS PEfeC HOM AO sy d Aaa SS 512 IM GE ccccccssccsssccssscessscssssessssssssssseessssseess 502 IMGN Diese ee 511 I Pea m EUR 512 IIO o 3 511 Uh 51 EMT EEEE EEEE 512 poned dd aS Bb onc i HR 5 informational multiport xit dececee 215 Input 1 d
454. o allow data to be taken from the EMPOWER simulation The number of ports is equal to the number of ports in the EMPOWER network to be analyzed They are placed in the layout using the EMPort button and can be Normal deembedded external ports gray external ports with No Deembedding white or internal ports white External Ports and Lumped Elements and Internal Ports are discussed in their respective sections EMPOWER Options To open double click or create a Planar 3D Electromagnetic Simuation EMPOWER Options X DOES D HAY General Viewer Far Field Advanced Layout to simulate Port impedance 50 Generalized Setup Layout Port Modes Use ports from schematic Necessary for HARBEC ca simulatian Electromagnetic simulation frequencies Lo simulation sweep n 1000 v Use EM simulation frequencies Stop freq MHz 5000 Start freq MHz 0 Number of points 5 Stop freg MHz s000 Harbec Frege z Number of points Max critical freg 3000 vw Turn off physical losses Faster Recalculate How Automatic Recalculation Automatically save workspace after calc LE Cancel Apply Help General Tab Layout to Simulate Allows you to select which layout in the current workspace to simulate Since workspaces can have multiple layouts and multiple EMPOWER simulations you can simulate many different layouts within the same workspace Port Impedance When EMPOWE
455. o be selected by pressing Ctrlt Home R Side Button Shows a side view of the current image This view is from the x axis at z 0 This option can also be selected by pressing Ctrl End S Oblique Button 165 Simulation 166 Shows an oblique view of the current image This view is top down on the x y plane with a slight offset This option can also be selected by pressing End T Current Plot Shows the color coded current patterns for the loaded EMPOWER generated viewer data file The menus and toolbar buttons control how this image is displayed U Color Scale For Current Plot This scale shows the relative current and current density magnitudes based on the color used to draw the plot patterns Far Field Radiation Pattern Viewer The EMPOWER far field radiation data describes the electric field patterns in the far zone region radiated from a structure The far zone is defined as the region where 21R gt gt 1 where R is the distance from the structure and lambda is the wavelength of the signal exciting the structure Far field radiation patterns are described in the spherical coordinate system where phi is the angle on the xy plane from the positive x axis and theta is the angle from the positive z axis The distance is not specified since it is assumed to be in the far zone Assumptions Made when Generating Far Field Radiation Data Data for radiation in the far field is generated using equations that ma
456. o ctu MARS EIER 11 DTA E E 11 ZPOR Eis de Nd 236
457. obtain smooth curves for the S parameters Oversampling however implies a huge amount of wasted resources Momentum allows the user to benefit from a smart interpolation scheme based on reduced order modeling techniques to generate a rational pole zero 379 Simulation model for the S parameter data The Adaptive Frequency Sampling algorithm selects the frequency samples automatically and interpolates the data using the adaptively constructed rational pole zero model This feature allows important details to be modeled by sampling the response of the structure more densely where the S parameters are changing sionificantly It minimizes the total number of samples needed and maximizes the information provided by each new sample In fact all kinds of structure can take advantage of the AFS module The Adaptive Frequency Sampling technology reduces the computation time needed for simulating large frequency ranges with Momentum significantly Special Simulation Topics 380 Some special simulation topics are discussed in this section e Simulating Slots in Ground Planes e Simulating Metalization Loss e Simulating Internal Ports and Ground References Simulating Slots in Ground Planes Slots in ground planes are treated in a special manner by Momentum An electromagnetic theorem called the equivalence principle is applied Instead of attempting to simulate the flow of electric current in the wide extent of the ground plane only the electric fi
458. od needs to create N designs differing by the port where the 1 tone exciting signal enters the circuit The design with signal port number j j 1 2 n is used to calculate LS 1j 121 2 LS parameters are calculated from set of N HB analyses with one created for each of the designs 2 Exciting all ports of the DUT simultaneously The full set of LS parameters are calculated from one design where the 1 tone exciting signals are entered at each of ports and and are analyzed by one multi tone HB analysis Frequencies of the signals slightly differ for each of the ports to extract spectral components from the analysis solution corresponding to each of the ports For example the frequencies may be defined by formula F i FO i deltaF i 1 2 n where FO frequency of the LS analysis deltaF small frequency offset deltaF lt lt FO The choice of the method should be made on the actual setup of the physical DUT If the real circuit is to be excited from one port only then the first method is preferable conversely if the circuit is to be excited from multiple ports method 2 should be employed For a linear circuit both methods calculate identical results while for a nonlinear DUT LS parameters that are calculated by the two methods may be different Getting Started and the difference will increase with rising power of tested signals Note The signal power for each of the signal ports are to be equal to the a
459. ode names SPECTRASYS will always find the shortest path that contain the list of parts or node names specified by the path Many times the input and output parts are sufficient However in the case where loops or parallel paths exits it may be necessary to specify intermediate part or node names to force the path through a direction of interest Part Names When selected informs the system analysis that part names are being used for the simulation of the path Node Names When selected informs the system analysis that node names are being used for the simulation of the path Force path through Switch state When checked will force the path to follow the state of the any switches along the path For example this allows the path to track the state of a switch bank When unchecked will allow the user the path to take through the switches Allow path to being on internal Node When checked any desired signals within the channel will be used to determine the desired channel power gain and cascaded gain When unchecked the desired signal at a soutce must be located or the desired channel power gain and cascaded gain cannot be determined NOTE When a path begins on a internal node the total power of ALL desired signals at the channel frequency will be used to determine the desired channel power gain and cascaded gain When unchecked a single signal at the source is used for these measurements and SPECTRASYS can distinguish this desired sign
460. oducts intermods harmonics etc to every node in the system These spectrums will keep propagating until no additional spectrums are created For instance any new inputs arriving at the input of an amplifier will cause intermods and harmonics to be created at the amplifier output at that particular time If additional signals arrive at the amplifier input at a future time then new intermods harmonics and other spurious products will be created at the amplifier output This process continues until no additional spectrums are created If loops exist in the system then the output from one element will feed the input of the next element and spectrum propagation could continue forever unless special features are placed within the software to limit spectral creation in this infinite loop Spectrasys has special features to control loops and limit the total number of created spectrums Loops Elements in parallel parallel amplifiers connected via a 2 way splitter at the input and combined back together with a 2 way combiner at the output can cause spectrums to be created that will propagate around this parallel path or loop If the gain of the amplifier is greater than its reverse isolation the spectrums will keep on growing as they travel around the path and will never die out we would have an oscillator The key point here is that if there are loops in the system schematic then it is very important to make sure that the element parameters are ent
461. of mixer noise simulation it may be only a single noise frequency point noise analysis corresponding to equal values of Minimum and Maximum noise sweep frequencies e Set value of Output Carrier Index Vector for example for Fif Flo Frf the value must be 1 1 or 1 1 e Simulate See Performing a Nonlinear Noise Simulation for sample output data and graphs Harbec computes single sideband noise NFSSB using the IEEE standard definition of single sideband noise figure usu Rok T G G GG sse s ie IE Gr 1 1 where amp is Boltzmann s constant 1 389658x10 23 TO is the IEEE standard temperature for noise figure 290 K 283 Simulation 284 G1 is the conversion gain of the mixer G2 is the image conversion gain of the mixer G3 Gm are conversion gains of higher order mixing products R is the resistance of the output termination Unoise is the noise voltage at the output port at the output frequency where the input and output terminations do not contribute any noise The NF SSB may be also calculated using post processing equations SN ye our TUE SSE SN 2 where on four T PNowr Fr is the output port signal to noise ratio SN Fuge k T is the input port signal to noise ratio PNoyr Fre Unicel Pye R tk Ly 7 G G t t Gp is the total output port noise power density in 1 Hz bandwidth including noise contributors from all circuit ports excl
462. of phase corresponds to a one wavelength delay period The difference of the current phases at the input and output again confirms a 90 degrees line segment 173 Simulation empower Viewer Y6 5 Mil x File View x Ang wie Freq GHe 15 1a lalel of a To Front Side Oblique The line segment example was prepared at two frequency points All graphs and explanations given here used the first frequency point 15 GHz The second point is 30 GHz and the corresponding segment length is a half of the wavelength You may display results at 30 GHz by clicking the button and then choosing the views of your choice M ultiM ode Viewer D ata This example illustrates the eigenwave multi mode excitation capabilities of EMPOWER A three conductor coupled microstrip line segment from Farr Chan Mittra 1986 is described in the schematic file LNMIT3 WSP Three microstrips are 1 mm wide and 0 2 mm apart They are on a 1 mm substrate with relative permitivity of 10 The segment is 8 mm long The structure has three modally coupled inputs at opposite segment sides We expect at least three propagating modes Load the example in GENESYS The listing file Right Click on the EMPOWER simulation in the Workspace Window and select Show Listing File gives information about the propagating waves The first eigenmode is an even mode with integral current distribution pattern the second eigenmode is odd pattern 0 and the third o
463. of the current patterns This option has a checkmark beside it when selected Load From User View 1 10 Loads the previously saved viewer settings for the selected view Saved settings can also be restored by pressing the number key corresponding to the desired setting Save To User View 1 10 Saves the current viewer settings into the selected view The settings can be restored later by selecting the desired from the load sub menu described above The options in this menu can also be selected by pressing Shift the number key corresponding to the desired save 163 Simulation 164 Tip The save and load functions are extremely useful If you rotate and pan to a view that you like press Shift plus a number not an arrow to save that view Simply press the number by itself to return to that view These views are remembered even if you exit the viewer so you can easily store your favorite views C X Y Z XY Button Pressing this button toggles between the four possible modes X Displays the x directed current density distribution Y Displays the y directed currents density distribution Z Displays the z directed currents XY Displays additive surface current density distribution function D Animate Button This button toggles viewer animation on the current image When this option is selected the button appears pressed The viewer animation is accomplished by multiplying the individual curren
464. oise at these two sideband frequencies and their correlation is then used to compute the phase noise The mixing analysis additionally computes the AM noise as well as phase noise and correlation between AM and PM noise Basic Phase Noise Theory Phase noise occurs naturally in electronic circuits It can be observed in the time domain as phase jitter of the signal on an oscilloscope display or as time fluctuations of the zero crossings see the following figure The below figure shows Phase Noise in the Time domain Act AP ds madri 0 5 Vie gini s D Vs gin ca aai t t Frequency and phase are related by equation 1 de JO YyT dt Phase noise and frequency fluctuations are the same physical phenomenon Noise in angular frequency can be obtained from the derivative of phase with respect to time The modulation of the signal phase manifests itself in the sidebands of the oscillator carrier as offsets from the carrier frequency these offsets are related to the multiples of angle modulation frequency For small angle modulation only the first term is important and HARBEC Harmonic Balance Analysis the relationship between the phase deviation and the sideband level is approximated as follows A common but indirect representation of phase noise is denoted L f see the following figure This is the ratio of the single sideband noise power in a 1 Hz bandwidth at an offset frequency from the carrier to the total ca
465. oise floor with a signal to noise ratio of 0 dB In other words the MDS 174 dBm Hz System Noise Figure 10 Log Channel Bandwidth See the Input Intercept and Channel Noise Powet measurements to determine which types of signals are included or ignored in this measurement Channel Used Main Channel Frequency Interferer Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as IIP and MDS Travel Direction Same as IIP and MDS Spurious Free Dynamic Range Receiver RX_SFDR This measurement is the spurious free dynamic range along the specified path as shown by RX SFDR n 2 3 RX_ITP3 n MDS n dB where n stage number The Spurious Free Dyanmic Range is the range between the Minimum Detectable Discernable Signal MDS and the input power which would cause the third order intermods to be equal to the MDS The MDS is the smallest signal that can be detected and will be equivalent to the receiver noise floor with a signal to noise ratio of 0 dB In other words the MDS 174 dBm Hz System Noise Figure 10 Log Channel Bandwidth See the Input Intercept Receiver and Channel Noise Powet measurements to determine which types of signals are included or ignored in this measurement Channel Used Main Channel Frequency Interferer Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as RX IIP and MDS Spectrasys System Travel Direction Sa
466. on DC Analysis Properties DC Analysis Properties Description This is a sample DC Analysis Calculate Nowy F lj Factory Defaults save as Favorite Temperature 27 0 Automatic Recalculation 86 DC Analysis 87 Simulation 88 The following table shows optional simulation parameters that can be set on the Advanced Tab note more than one simulation can be added by placing a semicolon between each parameter Ex emin 1e 6 reltol 1e 4 DC Analysis O utput DC simulation produces the following variables DC Analysis 89 Simulation 90 Chapter 5 EMPOWER Planar 3D EM Analysis EMPOWER Basics Overview A major part of any electromagnetic simulation is to break the problem down into manageable size pieces that allow an approximation of Maxwell s equations to be solved Electromagnetic simulators traditionally fall into three major categories 2 D 3 D and 2 172 1 2 D SIMULATORS 2 D simulators can only analyze problems that are infinitely continuous in one direction Ideal transmission lines and some waveguide problems are practical problems which fall into this category A 2 D simulator will analyze a slice of the line s and determine propagation impedance and coupling values 2 D simulators are the fastest but most limited type of simulator available 3 D SIMULATORS 3 D simulators can analyze virtually any type of problem and a
467. on of the triple beat level assumes that the amplitude of all input signals is the same The frequency combinations of the carrier triple beats are as follows Pl F2 53 F1 F2 F3 Bier P23 PIEP F 3rd Harmonics The amplitude of the third harmonics are 9 542 dB below the 3rd order 2 tone products Higher Orders The Maximum Order parameter on the Calculate tab of the system analysis determines the maximum intermod and harmonic order used in the simulation The intermod levels and frequencies are calculated based on a complicated mathematical process This process description is beyond the scope of this text Please see other resources for additional information Tone Dissimilar Amplitude Spectrasys automatically accounts for the amplitude of all input signals that create a given intermod This yields accurate intermod results since cascaded intermod equations ignore the effects of unequal amplitudes Channel Bandwidth and Intermods The bandwidth of third order products is greater than the individual bandwidth of the sources that created them For example if two 1 Hz tones were used to create intermods the resulting bandwidth would be 3 Hz The bandwidth follows the intermod equation that determines the frequency except for the fact that bandwidth cannot be subtracted For example if the third order intermod equation is Fim3 F1 2 F2 then the equation for the resulting bandwidth would be BWim3 BW1 2 BW2 If BW1 30 kHz
468. on the Smith chart using gain circles 297 Simulation GENESYS plots three forms of gain circles transducer gain unilateral circles GU1 for the input network and GU2 for the output network power gain output network circles GP and available gain input network circles GA Shown below are the input and output unilateral transducer gain circles GU1 and GU2 of the Avantek AT10135 GaAsFET transistor The measurement circles are plotted at the frequency of the first marker in this case 2500 MHz Marker 1 1s plotted at the center of the smallest circle the point of maximum gain The gain at the circumference of each circle of increasing radius is 1 dB lower than the previous inside circle Response SPARAM 5D DEL GU DEGLI DB GU DE GUt DIE GUI DE GU 2500 3500 S500 2500 3500 S500 12000 2 91388 4 58296 3 6764 10 5031 12 2167 8 40433 5 35212 2 31388 4 58296 3 6764 0 0 0 The arc which is orthogonal to the gain circles is the locus of smallest circle center points from the lowest to highest sweep frequency Tuning the first marker frequency moves the center of the circles along this arc Notice that a complex conjugate match at the input improves the gain by over 3 dB in relation to an unmatched 50 ohm source impedance However matching the output provides less than 1 dB gain improvement An examination of the device S parameter data at 2500 MHz reveals that the output is originally closer matched to 50 ohm
469. ones in order to view the full intermod power of 3rd order products This bandwidth must be increased accordingly for higher order products Here is an example of the intermod spectrum due to 5 carriers through the same amplifier Spectrum groups ate being displayed instead of individual spectrums Signals are shown in one color and all intermods are grouped together and shown in another color Spectrasys System 5 Carrier Output Spectrum i 10 E 20 E 30 2 p 40 F C 2 a a 50 a C 60 F jf jf jf 90 200 250 300 350 400 450 500 Frequency MHZ Notice the peaking effects of the intermods around 300 and 25 MHz as well as the in channel effects associated with carrier triple beats and 2nd order products Reverse Isolation Intermods can and do appear at the input to a nonlinear stage due to the reverse isolation of the device as shown in the following figure 447 Simulation spectrum due to Reverse Isolation Input Power dem i EH J 90 100 150 200 250 300 350 400 450 S00 Frequency MHz The 5 carriers are displayed in one color and the reverse intermods in another Calculated Products The following figure shows the nonlinear second and third order products created for two input signals F1 and F2 where F2 is greater in frequency than F1 448 Spectrasys System P dEm c XIR 3 P a IPs 5 P 6dB 6dB n D IB P B 542ZdE Fh IM H INE 3 FE IMs s Mon t The
470. onic order results in higher accuracy of the solution at the expense of slower simulation Having established a working oscillator this value may be changed to improve accuracy Generally the default values will yield sufficient accuracy HARBEC Harmonic Balance Analysis Spectrum E x ay ta Ti S of D DL DL Frequency MHz Selecting the Oscillator tab allows to enter the search range for analysis This gives the simulator a range of frequencies to search over to find the exact oscillation frequency For resonator elements such as crystals addition points may be required to find the exact frequency considering the higher C 255 Simulation Harmonic Balance Oscillator Analysis Options General Calculate Oscillator Advanced FFT Oscillations Frequency Calculation Number of Points 101 Calculate Osc Frequency Use Osc Solver Osc Port Initial Voltage Absolute Current Tolerance Number of Curve iterations Max step of Curve tracing Number of Adjust iterations Having set the range for search we are ready to perform an analysis Select the Update Icon from the GENESYS toolbar to perform the Harmonic Balance simulation To view the results of our simulation add a rectangular graph from the 256 HARBEC Harmonic Balance Analysis Outputs file in the GENESYS Workspace Window By double clicking on the graph or selecting properties from a right mous
471. ons block Change the Lead MFilter1 IL2 MrFilter1 IL1 MFilter1 and S1_MFilter1 to the values as show below Also be sure to remove the question matk since we will not reoptimize the layout dimensions Note The equation order may appear differently in your example 192 EMPOWER Planar 3D EM Analysis BP WorkSpace 1 Global Equations Workspace WorkSpace 1 MFilter MFilterl Equations a LLL LaLa Il M MFilter1 20U ILI M HFilter1 100 S1 MFilter 50 Lead MFilterl 15ll CAP1 MFilterl1 2 53025 CAP MFilter1 2 5075 Z5 MFilterl 5l I i ie aa End MFilter MFilter Equations I i LLL LLL LLa 11 Next we need to change the resonator widths for all four transmission lines TL1 etc by double clicking on each of these schematic elements and changing the Width or Width of all strips to 80 mils 193 Simulation Part Properties For MENG E ioe ese fee si te rs m DIM O O O O Lo qm DIM pm Default Untitled 1 bd 12 The optimized readjusted filter schematic should now appear as follows 194 EMPOWER Planar 3D EM Analysis fff C 2 996 pF CAP 1 C22 326 pF CAP1 TL WSO mil L 200 mil IL2 3 TLI N 8BD mil Tt WW 80 mil Lefetmil Lead cen mil eee 2 L 100 mil IL1 13 You also may need to change the capacitor footprints to 0603 depending on yout default footprints This can be done by bringing up the Layout Properties
472. oolbar 114 EMPOWER Planar 3D EM Analysis 2 Click on the left edge of the page border and drag toward the right and down until the status bar bottom of the screen shows DX 425 and DY 50 3 Release the mouse button This is the series transmission line The screen should now look as below Don t worry if the line isn t at the exact same position on the page the layout will be centered later EB Layout workspace WorkSpace 1 To draw the open stub 1 Select the Rectangle button from the toolbar 2 Click at the bottom edge of the line just drawn one grid cell left of the series line s centet 3 Drag to the right and down until the status bar shows DX 25 and DY 225 4 Release the mouse button The screen should now look like the following If the stub line isn t centered horizontally on the screen select the stub by clicking on it and drag it to the proper position 115 Simulation 116 EB Layout Workspace WorkSpace 1 Centering the Layout As a general rule EMPOWER simulation time is greatly reduced if the circuit to be simulated exhibits symmetry in any of several planes Many circuits will exhibit some form of symmetry if they are centered in the page area To center the example filter 1 Choose Select All from the Edit menu 2 Choose Center Selected On Page from the Layout menu Placing EMPO WER Ports Before running EMPOWER the filter s ports must be designated Select the
473. or is one of the operators listed in the table below or any other function described in equations and measurement is one of the measurements listed in the table in the previous section All measurements have default operators For instance on a table using 21 will display in dB angle form and Z32 will display in rectangular real amp complex form Likewise on a graph S21 graphs in dB while Z32 graphs the real part of Z32 Note For a description of mathematical operators such as and see Operators Operator Desctiption Measurement Result Is Should Be mag Linear magnitude Real Complex Real 13 Simulation 14 ang ang360 re im db gd ql time Angle in range 180 to 180 Complex Real Angle in range 0 to 360 Complex Real Real part of complex measurement Complex Real Imaginary part of complex measurement Complex Real dB Magnitude Real Complex Real Group delay Complex Real Loaded Q QL Qpif GD 2 Complex Real Converts Frequency domain to Time domain Complex Real via invetse Fourier Transform Intended for use with Voltage Current to get time waveforms Note All available measurements and their operators for a given circuit or sub circuit with their appropriate syntax are shown in the measurement wizard To bring up the measurement wizard select measurement wizard from the graph properties dialog box Sample Measurements Measurement S 2 2 ql s 2 1 mag S21 im ZIN1 S
474. orresponding mode characteristic impedances while the internal ports are terminated by 1 Ohm if another termination is not defined by the option NI lt n gt e The instantaneous power of the incident wave is 1 Watt and time average power is 1 2 Watt e Surface current density functions are used for the signal or metal layer and integral currents are used for viaholes and z directed inputs EM POWER File Descriptions O verview In performing its tasks EMPOWER creates many different types of files An understanding these different files is very helpful in understanding the operation of EMPOWERR These files contain the topology of the circuit external port line data generalized S Parameter normalizing impedances output information S Parameter data batch commands Y Parameter data viewer data and backup data Where are these files Starting with Version 2005 GENESYS uses XML storage for its workspace files The files described in this section are normally stored internally in the workspace If you need to access these internal files right click on the EMPOWER simulation on the tree and select Write Internal Files This automatically creates a directory with the same name as the simulation and places copies of the files there Note Previous versions of GENESYS used actual disk files for all internal EMPOWER files and separate subdirectories were recommended for each circuit This is no longer necessary for typical usage
475. ou want to calculate radiation patterns for it after the simulation you need to set the top and bottom planes to values between 376 and 378 LAYOUT Properties General Associations Layer Fonts Show Columns Metal Substrate General Layer Mumber and Calor show All EM EMPOWER Momentum Momentum Slot Type Strip v T Etch u 1 2 T Heiaht E Tana Surface Imp Current Thick Meti EE Factor 7S Height ype eu 7 Sigma Value or File Direction Slow Top Cover Electrical afr Air Above Air 500 1 1 TOP MASK Tnm caa D efining a Mesh A mesh is a grid like pattern of triangles and rectangles and each triangle or rectangle is a cell This pattern of cells is based on the geometry of a circuit and optionally user defined parameters so each circuit will have a unique mesh calculated for it The mesh is then applied to the circuit in order to compute the current within each cell and identity any Momentum GX coupling effects in the circuit during simulation From these calculations S parameters are then calculated for the circuit The mesh that was calculated for the double patch example is shown below Rectangular cals are applied bo mest of the geometry Lr I L t Far Dut for curved or no n rectangular abjects triangles are used to complete Ine mesh Creating a mesh consists of two parts e Defining mesh parameters e Pre computing the mesh It is not
476. out making any simplification to the Maxwell equations This results in L and C elements that are complex and frequency dependent The quasi static mode uses frequency independent Green functions resulting in L and C elements that are real and frequency independent Because of the approximation made in the quasi static mode the 373 Simulation RF simulations run a lot faster since the matrix L and C elements only have to be calculated for the first frequency simulation point The approximation also implies that the quasi static mode typically should be used for structures that are smaller than half the wavelength Both engine modes are using the so called star loop basis function ensuring a stable solution at all frequencies Both engines also make use of a mesh reduction algorithm which reduces the number of unknowns in the simulation by generating a polygonal mesh This mesh reduction algorithm can be turned on or off with a toggle switch The sources applied at the ports of the circuit yield the excitations in the equivalent network model The currents in the equivalent network are the unknown amplitudes of the rooftop expansion functions Solving the equivalent network for a number of independent excitation states yields the unknown current amplitudes A port calibration process is used to calculate the S parameter data of the circuit from the current solution when calibration is requested The following sections in this chapter contain mo
477. ove the metal layer In the case of a down expansion the new layer has the material properties of the layer below the metal layer up expansion down expansion diel diel T Extra internal metalization layers are automatically added in Momentum GX to model the currents on all four sides of the finite thickness conductor To simulate metal thickness in Momentum GX open the Layout dialog properties Layer Tab and set the metal layer type to Physical In the Thick Metal column of the table set the metal expansion to Default Thin Thick Up or Thick Down Default will use Momentum GX analysis settings Thin does not use metal thickness sheet metal layer 338 Momentum GX LAYOUT Properties General Associations Show Columns v Metal substrate C General Layer Number and Color C Show Al Clem J EMPOWER Momentum Momentum Slat Type Edge Mesh Via Model Strip T Metal Current Thick Metal Width la mode Model ype Thickness Direction Slow rn Nd NE pem Physi tar Mme ms v Physica with Tart faut eme oo DWa Physical wih ed m ii Thick Down s E Up B showal 8 seal Frequency Units MHz Load From Layer File A Delete w Down E Hide All H Use None Length Units mil Save En Layer File em apply The default metal 3D expansion parameters are defined in the Momentum analysis properties Simulation Options tab 339 Si
478. p 0 EE X E 5 cx amp 9 pA AAR H E Designs o 4 F2000 Schema Layout Layout B xg Simulations Data Si EM1 Layout Sij Lineari 1400 tc E a Outputs Bg Circuit Simulation 0 FRA Combined Simul 4 de Equations 1400 2000 2600 1400 2000 2600 Freq MHz Freq MHz E DE S11 amp DB E21 DB S11 Lumped Elements The first example in this section required several data points to find the exact notch frequency This second example only used 4 data points and produces data very close to the GENESYS simulation This is because the capacitors which load the coupled lines causing resonances at the center frequency were removed during the EMPOWER simulation This effectively removes the resonances from the simulation range producing a flat response from the open coupled lines Since a flat response is well suited for linear interpolation few data points are required in the EMPOWER simulation In the EMPOWER options dialog the Co Simulation Sweep box is used to set up a simulation with more points after lumped elements are added When GENESYS uses the EMPOWER results it replaces the lumped capacitances resulting in the bandpass response shown in the previous section This technique is similar to using approximations for inductors and capacitors when creating a lumped filter Even though the response of each element is simple and can be interpolated from just a few widely spaced points the complete filt
479. parallel with the C1 to port 2 network In a similar manner the value reported at node 5 for the Path3 2 would be the value of the measurement leaving terminal 2 of inductor L1 entering node 5 The impedance for the node looking from terminal 2 of inductor L1 is most likely to be completely different from the impedance seen by R1 or even C1 because from the inductors perspective the R1 to port 1 network is in parallel with the C1 to port 2 network 485 Simulation 486 SPECTRASYS knows about the direction of all of the paths and will determine the correct impedance looking along that path As a result all measurements contain the cotrect values as seen looking along the path of interest Remember absolute node impedance and resulting measurements based on that impedance don t make any sense since they are totally dependent on the which direction is taken through the node Path 1 2 Path 3 2 Transmitted Energy When an incident propagating wave strikes a boundary of changing impedances a transmitted and reflected wave is created Obviously the transmitted wave is only the enetgy of the wave flowing in the forward direction The SPARCA engine calculates the transmitted wave ONLY from the incident wave This transmitted energy is what is used for all path measurements See No Attenuation Across a Filter for an illustration of this principle Increasing Simulation Speed There are several options available to increase the si
480. particular node then power would be flowing in three different directions A unique color would represent each trace Show Individual Spectrums When checked individual spectrums will be displayed otherwise groups will be shown Noze Individual spectrum identification information cannot be shown for a group When grouping spectrum all spectrums in the same group are represented by a single trace Show Individual Signals Show Signal Group Shows a trace for each fundamental signal spectrum or signal group Show Individual Intermods amp Harmonics Show Intermods amp Harmonics Group 549 Simulation 550 Shows a trace for each intermod and harmonic spectrum or intermod and harmonic group Show Individual PhaseNoise Show PhaseNoise Group Shows a trace fot each phase noise spectrum or phase noise spectrum group Show Individual Noise Show Noise Group Shows a trace fot each noise spectrum or noise spectrum group ENABLE ANALYZER MODE This checkbox enables the analyzer mode and its settings This mode can help the engineer visualize what the simulated spectrum would look like on a common spectrum analyzer The analyzer mode has been added to allow the user to correlate the simulation data with spectrum analyzer data measured in the lab Noz This mode affects only the display and in nowise will affect the integrated measurements Resolution Bandwidth RBW The analyzer mode can be thought of just like a spectrum
481. patch antenna or the traces on a multilayer printed circuit board There are three ways to enter a design into GENESYS e Convert a schematic into a physical layout 305 Simulation e Draw the design in a Layout e Import a layout from another simulator or design system GENESYS can import files in a variety of formats For information on converting schematics drawing in Layout or impotting designs refer to the User s Guide 2 Assign port properties Ports enable you to inject energy into a circuit which is necessary in order to analyze the behavior of your circuit When you create the circuit you apply ports and then assign port properties in Layout There are several different types of ports that you can use in your circuit depending on your application For more information refer to Ports 3 Define the layer stack characteristics This is done on the Layers page of the layout properties dialog A substrate is the media upon which the circuit resides For example a multilayer PC board consists of various layers of metal insulating ot dielectric material and ground planes Other designs may include covers or they may be open and radiate into air A complete layer stack definition is required in order to simulate a design The definition includes the number of layers in the substrate and the composition of each layer For more information refer to Adding and Changing Substrates 4 Simulate the circuit You set up a sim
482. pecification the D microwave mode of Momentum incorporates the HF losses in ground planes however the RF mode will make an abstraction of these HF losses Simulating with Internal Ports and Ground References Momentum offers the ability to use internal ports within a structure Internal ports can be specified at any location on the planar metallization patterns and they make possible a connection for both passive and active lumped components to the distributed model of the planar circuits Refer to Figure 4 The S parameters associated with these ports are calculated from the excitation consisting of a lumped voltage source connected to the equivalent network model as shown in Figure 3 The ground reference for these ports in the resulting S parameter model is the ground of the equivalent network and this ground corresponds physically to the infinite metallization layers taken up in the layer stack In the absence of infinite metallization layers the ground no longer has a physical meaning and corresponds mathematically with the sphere at infinity It is important to mention that in this case the associated S parameters also lose their physical meaning as the applied voltage source is assumed to be lumped that is electrically small since it sustains a current flow from the ground to the circuit without phase delay To overcome this problem a ground reference must be specified at a distant electrically small from the internal port Failure
483. ports are mapped from the intersection points to the top or bottom cover There are two caveats metal and ports in the z direction are modeled as one continuous current so the viaholes should be small in compatison with a wavelength Also you cannot have both a port and metal along the same grid line so you should be extremely careful when placing a viahole directly underneath an internal port You should check the listing file select Show Listing File from the EMPOWER simulation right click menu carefully to see that both the port and the viahole are represented on the grid The physical length of a viahole in a substrate should be kept shorter than about 1 10 to 1 20 wavelength within the analysis range Longer lengths can suffer calculation inaccuracies in EMPOWER For example suppose a microstrip circuit with a 10 mil substrate and a dielectric constant of 2 4 1s to be used What 1s the highest accurate frequency for this setup Note If the substrate layer is broken down into two substrate layers by adding an additional layer each 1 2 the height of the original then the viaholes will be accurate at twice the original frequency This procedure can be repeated as necessaty a amp 8542x1077 7 u 212566x10 E ie 48 85Hg 42 A gH Z 10 mils 2 5410 m A 381x10 m c 1935x10 ds NE dace NT GHZ Ao 381x10 m EMPOWER Planar 3D EM Analysis EM Ports All circuits must contain at least one EMPort t
484. put Intermods Along Path Tone Interferer Frequency MHz Powers Voltages and Impedances iet chomet games 22 Receiver Image Channel Erequency Offset From Channel MHz Parameter Information All fields can be numeric or a variable When using a variable the name of the variable is entered into the field and the variable must be defined in an equations block For example if we wanted to use a variable for the offset channel frequency called MyOffset then we would type the string MyOffset into the Frequency Offset From Channel field then define MyOffset 10 in an equation block which is set to use display units or in other words the units used in the dialog box If the equation units are changed to MKS then MyOffset 10e6 would represent 10 MHz To make a variable tunable a is placed in front of the number in the equation block i e MyOffset 10 Offset Channel Name This is the name used to identify the path 541 Simulation 542 Description Description of the path Used for identifying the path to the user When Use auto description is checked the description will automatically be filled based on parameters set in the remaining fields of the dialog box Use auto descprition Description will automatically be created from specified path parameters DEFINE PATH This section will define the path to be calculated by the system analysis A path can be specified with either part names or n
485. put impedance at all yin diag S ZPORT vector ports including all other port terminations LL admittance at un 7 ZPORT dq Complex Specific S parameter Complex For 7 or j greater than 9 use the S form Parameters of circuit stoy S ZPORT N x N complex matrix Complex complex matrix Complex 2 x 2 n matrix Specific Y parameter For 7 ot j greater than 9 use the YPL form 7 Parameters of circuit stoz S ZPORT Specific Z parameter For 7 ot j greater than 9 use the Z j form 5 H parameters two port stoh S ZPORT Hy Specific H parameter H zj Complex in S a Complex terminals with all termination impedances Stability factor only E Scalar available for two port Bl Stability measure only stab_meas S Scalar available for two port Linear Analysis circuits Stability circles at port 7 stab_circle S z 2 Complex only available for two vector port circuits contains center radius and sien All noise measurements below which do not specity ports use port 1 as input and pott 2 as output NF Noise figure from port 1 noise figure S CS 2 1 Scalat oise Resistance noise rn S CS ZPORT 2 1 Scalar oise optimum noise gamma opt S CS 2 1 Complex reflection coefficient gamma ZOPT oise optimum input zin noise gamma opt S CS 2 1 Complex impedance ZPORT 1 YOPT oise optimum input yin noise
486. quencies fall within the channel and the Channel Power measurement will be very high IF the Total Intermod Channel Power seems to be too high the intermods may be traveling backwards from a subsequent stage The reverse isolation of this stage can be increased to verify this effect If the Input Intercept Point measurements IIP OIP RX_HP RX_OIP don t seem to be correct then first verify that the Interferer Tone Channel Frequency ICF is set to the interfering frequency If the Interferer Channel Frequency is set correctly then look at the Interferer Channel Power ICP measurement for in band intermod measurements or Virtual Tone Channel Power VI CP measurement for 465 Simulation out of band intermod measurements to verify an expected level of interferer channel power e Ifthe Output Intercept Point looks correct but the Input Intercept Point doesn t then verify that the Interferer Cascaded Gain ICGAIN measurement is cotrect for the in band intermod measurement case or the Cascaded Gain CGAIN measurement is correct for the out of band intermod measurement case See the Intermod Path Measurement Basics section for illustrations of intermod path measurements and additional information Advanced 466 Cascaded Noise Analysis Cascaded noise figure is an important figure of merit especially for receiver design Traditional cascaded noise figure equations are not used in Spectrasys A more elaborate
487. quired then GENESYS will give an error 1f the parameter is not specified If the value NOT GIVEN is used as the default then GENESYS will not show a default but will just show optional for the default value Eagleware extension keywords All Eagleware extension keywords are placed inside comments in the Verilog A and are always placed between two pairs of percent signs like D YKEYWORDY VVoKEYWORD value oo Since these keywotds are placed in comments they will be ignored by other simulators The keywords znust be placed inside the module that they are to affect after the module statement Keywords only affect one module If you have multiple modules in your VA file you will need to duplicate any keywords which are to affect multiple modules DEVICE CLASS keyword The DEVICE CLASS keywotd tells GENESYS what type of device the module represent This allows GENESYS to e Select a appropriate symbols e Create multiple models for N or P class devices such as NFET or PFET from one Verilog A source e Automatically reverse pins 1 and 2 for transistor class devices This reversal is necessary since in SPICE and in most Verilog A source the input is pin 2 and 61 Simulation 62 the output is pin 1 In GENESYS and most RF Microwave simulation the convention is for the input to be pin 1 and the output to be pin 2 Some examples of device class statements are DEVICE_CLASS DIODE DEVICE_CLASS FET
488. r lines that do not resonate and interpolation is possible For example a 7th order interdigital filter can often be simulated with just 5 frequency points in the EMPOWER run while 100 points are displayed in the output sweep e Ability to simulate problems too large to otherwise run The main disadvantages of decomposition are e Tedious to setup circuit The simulation requires multiple EMPOWER runs combined with a schematic e Box modes and other phenomena related to the entire problem are not modelled However since EMPOWER uses mode space to model coupled line connections this is less of a problem that it would be with other simulators e Losses in the connecting lines are not modelled Basics NOTE The Single and Multi Mode transmission lines required to use decomposition are not available in GENESYS Decomposition can be applied to circuits with parts which are connected via single or multiple transmission lines Some typical circuits which can be broken apart are shown below In each of the circuits the unshaded areas are simulated individually The pieces are then combined using multi mode transmission lines to connect the pieces representing the lines in the shaded area 143 Simulation 144 Spiral Inductor Interdigital Meander Line Filter For decomposition to be possible you must be able to break the circuit down into rectangular areas which are interconnected with transmission lines For example the
489. r algebraic equations using partial inversion e Resolution to Y or Z matrix relating integral grid currents and voltages in the input and lumped element regions Mapping on the Grid To map a boundary value problem for a partial differential equation on the grid basically means to substitute the problem with solution defined in a space of continuous functions by a problem with a solution defined in a discrete space The model solution must be as close to the continuous one as possible To solve the problem we approximated the partial derivatives in the signal plane by finite differences applied to grid analogues of the field components The corresponding grid is shown here 211 Simulation 212 el 087 ANE RP RD a MF dw There are L 1 equidistant cells along the x axis and M 1 cells along the y axis The grid equivalents of the electric e and magnetic h fields are defined as corresponding continuous function values in offset grid points as is shown for a grid cell above The grid functions are continuous along the z axis Grid x and y directed current variables Jx Jy are defined as integrals of the surface current in the metal plane across the grid cell Grid z directed currents Jz are defined as surface integrals of the volume current density jz across the grid cell The first offset model of Maxwell s equations was apparently proposed by G Kron 1944 The cells below show a summary of the similar models implemented by dif
490. r ed Tabea N 103 Advanced Tabs a E E EE E EO 105 Siraulati n Stau WY 1A Wy ano doc T aa 106 PERON oaan I AE ed tarp aot tides 107 EMPOWER OC De ration enkieaon a aa nd int nad cupis 107 VSS C este 107 lc eres plc atone 108 Greauno a Layout Without 42566 HIO HCL stia voce o Seit tr emo Oeo preti A 108 A IN MM rob 109 bos Dimen seio und quiu id ce Lost mmu cUm rte Ld 110 eneral LaVi a E MN ON 111 EMPOWER TAY i T tae 112 Prvine tie Lay OU ense dcm tid qubteu O mE UM tM DUI NEM IM E 114 CGS INO LING Ay OU seta tanoncatenecartasriss tos cat A A E A E 116 Table O f Contents Placing EMPOWER D OBS ie vcsatuscetessecustnsneietassinasostutebastanaeanss beatos eter tbe nde cm Mee 116 SUN AS CS TGA yO D E HE T 117 VIENA RE 01h Spar rey E E O E N Te O 119 USDE NOWE cetera rere Eero e E N rere erent ree rere 122 Creannecadayout Fronran Extstpe SCneimatG iesdstembetut os qub ness bd e pa iat 122 SHIM A ES Lay OU sites d dee bebe teeth d ot umen to etu redis 124 I umped EJeriepntSo sbesteosttade edet Nes Aedes tp eade li tei NR Den Sond eEa spit aUe 126 Reak Time TUNNO eio sto t aene debuerit O OE t27 EMRO ERTED eoe ce mee ere tiit E re ee mr em S E 128 OVS sect a ee ei NE E E A E aca he ao vac er gana E 128 SUED ee 128 Manu Cit cal Preguen oes atus tieu iut tede obiit reer rater 129 Sy AMMAN iens sc ala DC EE 129 Tinno GI
491. r many operations including FFTs it can sometimes cause one to draw incorrect accuracy conclusions 3 Click and select Output All Simulated Points This will not modify the simulation but it will cause the simulator to output all analyzed time points CAYENNE Transient Analysis 4 Click Calculate Now on the General Tab You should see the graph below You can turn on the marker symbols Click the Toggle Symbols button on the graph toolbar to easily see exactly what points were simulated Transientl_V PORT E m x r3 ce hJ x HL 5 CL j gt oF ET LL gt hA JR 5002 10005 15 007 20 04 25012 30 015 35 017 40019 45 022 50 024 Time n VPORT I vPORTE Simulation Time Steps CAYENNE is a time stepped transient simulator The basic flow of the simulator is 1 Solve the circuit at time 0 If Skip Bias is selected the initial solution at all nodes is zero volts Othewise the initial solution is the DC solution 2 Advance the time by an amount determined by the simulator Note that the entire circuit always uses the same time value 3 Solve the circuit at the new time step replacing charge based elements like capacitors with equivalent current sources and resistances If the circuit cannot 69 Simulation be solved or the error appears to be too large then return to the last good time point and go back to step two taking a different step size 4
492. r order in the table depends on the noise carrier defined by its index vector Mix1 Data variable SEMCMOISE SEIFNOISE Era Freq Era FregID 3 FreqIndexIM 23 FSPNOISE SHINAM SEIMAPM SEINCGAINM Era NCIndex Era NCMame SEINCValue SS NF 8 71 SMin SNout SEINFDSB SEIMFSSB SSINPM SSIP1 PPORT 1 SEIP2 PPORT 2 SSIPDPORT 2 Pif 32 32 hb_getspcom SEIPPORT SEIPSPNOISE SEIRPNOISE 2 SNout 145 12 Pif dbmfPP ESPORT SSIZPORT Kak liCIndex HCHame 1 y COPartl thermal chite WRF Part thermal vhite 1 flicker ticker D1 shot white D1 Rs twhitej B PFoOPart thermal white n PFOPartzthermat white B eFC Part thermal hte g PC Part thermal white 10 W 1 thermat white 269 Simulation Mix1 Data Variable FHOISE NCValue 1 NCValue 2 HCValue 3 NCValue 4 NCValue 5 NCValue 6 NCValue 7 NCValue 3 HCValue 9 NC Value 10 8 NCName 1e 6 146 146 146 6 149 95 159 18 195 69 195 91 206 2 207 62 234 38 S amp NCValue SSJNF 8 71 v iT AIENGR 13 The Checkbox Improve accuracy of Low Frequency Oscillator Phase Noise Polishing is active only in the Harbec Oscillator controller Checking it improves the low frequency phase noise solution where the phase noise would otherwise undergo flattening 14 Specify the Noise Frequency Bandwidth parameter value The output noise voltages scale with the square root of the
493. r sharp responses that are difficult to detect Data that would have remained hidden using other types of analyses can now be obtained Even in the case of a low pass filter AFS can solve for the S parameters faster and more accurately than discrete data sampling For example if you were measuring a 2 GHz low pass filter you would enable the AFS feature and then enter a start frequency of 500 MHz and a stop frequency of 4 GHz AFS would then sample the two end points construct the fitting model and then sample points in between as needed As this occurs the model is automatically refined and appropriate sample points are taken until the model and the sampled data converge When convergence is reached both sampled data and the AFS 367 Simulation 368 data are written into two separate datasets and can be presented as S parameter traces on a plot If the max sample size is reached before converging the partial results will be saved and can be reused by simulating the circuit again AFS data must converge in order to be accurate Be sure the data is o converged by checking the message in the Momentum Status window If data is not converged you will see inaccurate results AFS Convergence Convergence is determined by how close the fitting model is to the sampled data Convergence is achieved when the sampled points are accurate to 75 dB and the overall accuracy is 60 dB When AFS is used be sure that the sampled data has conver
494. r than MMTLP lines or other identical modal inputs to inputs which are modally related Connecting lumped elements to modal inputs is incorrect and will give bad results e Potts which will be modally related must have sequential numbers They must also all have the same reference shift e Ports for mode space inputs must be marked type Normal not No deembed ot Internal Correspondingly their numbers must be lower than any No deembed or Internal ports e The order of ports used must correspond between the pieces and the MMTLP lines used The lowest port number in a modally related set of inputs should connect to Mode 1 in the MMTLP line and the highest port number in the set should connect to Mode N on the MMTLP line Also port ordering should be exactly the same on both pieces connected through the MMTLP The figure below shows an incorrect numbering of the spital inductor In this example PART1 and PART2 are inconsistently numbered since on PART1 the outermost inputs numbers 2 and 6 are the lowest number while in PART2 the innermost inputs numbers 1 and 5 are the lowest number Part e Pieces can be connected directly together without using MM TLP In this case the lowest ports in each modally related set of inputs are connected to each other 152 EMPOWER Planar 3D EM Analysis EMPOWER Lumped Elements and Internal Ports Overview As described in the External Ports section every EMPOWER c
495. r the NPO8 is in WSP Simulations EMPart2 EMPOWER SS Note Some users may find it easier to write a text Netlist to combine the pieces At Eagleware we find it easier to use a schematic for this purpose but you may use whichever you feel most comfortable with 147 Simulation Whenever deembedded ports are used data files suitable for the SMTLP and MMTLP models are automatically created during the LINE portion of the EMPOWER tun For the MMTLP8 lines the file WSP Simulations EMPart1 EMPOWER L2 was used This cotresponds to the second set of inputs for PART1 You should view the listing file Right click on EMPart1 and look at the port numbers to determine which EMPOWER L file contains the line data you need Note Files with names like WSP Simulations EMPart1 EMPOWER L2 are taken from within the current workspace For a complete explanation of how these files are names see the File Formats section in this manual The substrate must also be specified but only the UNITS parameter is used by the MMTLP8 model A variable was setup LENGTH so that the lengths of line can be tuned in GENESYS simultaneously changing the size of the spiral and thus the inductance very quickly 148 EMPOWER Planar 3D EM Analysis 4 GENESYS 7 0 Graphi workspace res5d OW x File Edit View Workspace Actions Tools Synthesis Window Help 8 x SHl Rh c e dm V AA UJ 1 nmYXx A TERA AAD I LAN ILIA E
496. r the subcircuit instead of the spectral propagation engine If the system simulator finds a nonlinear behavioral model on the subcircuit the spectral propagation engine will be used For example a circuit such as the following should be moved to LINEAR ONLY subcircuit 487 Simulation 488 1 c 10pF C 50pF Intermods One of the largest time consuming operations in Spectrasys is the calculation of a large number of intermods The number of intermods generated is determined by the Maximum Order of intermods and the number of carriers used to create the intermods Besides reducing the maximum intermod order raising the Ignore Spectrum Level Below threshold will eliminate all intermods below this threshold Spectrum Analyzer Display Mode During the system simulation an analyzer trace will be created for every node in the system Consequently for systems with large number of nodes the integrated analyzer traces alone can be time consuming if the analyzer properties are not optimized The simulation speed can be reduced by a careful selection of Analyzer Mode settings If large frequency ranges are integrated with a small resolution bandwidth then the amount of data collected will be much larger and the simulation speed will decrease Furthermore enabling the Randomize Noise feature may also slow down the simulation In order to increase the simulation speed with the Analyzer Mode enabled the user can disable the Ran
497. ral Associations General Layer EMPOMWER Layers Fonts Height or Surface Imp Thickness Value or File Direc Physical DO5S DE Physical wv Tand MEI op Cover elo I c o q nw E RD BE OP METAL 3 UBSTRATE a Physical attom Cover Cancel ppp LAYOUT Properties E General Associations General Layer EMPOMWER Layers Fonts Tana Rough Surface Imp Thick Metal Sigma g Value or File Slow a feos o 0 p NEN NEN NN CNET 113 Simulation Since Air layers above and below a substrate are so common a special option has been given here to add them For more information on the individual layer options see the EMPOWER Basics section Notice that BOT METAL and Air Below are not enabled This places the box bottom at the lower substrate boundary so that it acts as a ground plane Note In almost all cases where a completely solid ground plane is used you should use the top or bottom cover to simulate it This is much more efficient than using an extra metal layer Click OK The LAYOUT editor appears The screen should look like similar to yi GENESYS V7 0 BI x File Edit View Workspace Actons Tools Layout Synthesis Window Help r m uamzaoe J y l xl BE Layout Workspace WorkSpace 1 Im x D rawing the Layout To draw the series line 1 Select the Rectangle button from the LAYOUT toolbar the sixth button on the t
498. ramietets for a Layour Thay er eurei a Pot oot DT 353 Dtecompuuns a hd Loc ossiani nana aa T AE TOE AETA TO OAA 354 Viewin Cite Mesto Unima o na a T 356 Aboutthe West eO n A E a E 357 Chapter 9 Table O f Contents Cruidelines Tor MESNO o estate ets th oigo tu eR ieu dem a arene creer n tome RN Dep TS 364 PX CAD UV Erequeney DAD DTI daa epic A d dat to unu ed i vuug t Do Mna 367 PTS CONTO ICE osiooce d ttc b d e ipeo tcu Ge bei ditte des Det e E 368 Settino Sample FON oni n ta ene rer Ero bh misto battre bet este dads 369 Viewing AFSS PANEI TS ia a dbu ebd entire daunted emi bns 370 DIMOTHOD BU C et AIO DIS aree e TEE ei timed A soi tn LEA 370 Momentum Theory of Operatoriai detou sed tela ch ek etus Deas ae MG vex ish deo Su ovt ei ugs 373 The Metbod ot Moments Leen 0 Oy siminn n E oe post E 374 The Momentum Solutioti PrOCeSS seen reno otra pne e riga raa ripa Eai EE ep dui 377 Calculation of the Substrate Green s Functions eene nennen 378 Meshing of tie Planar Signal Layer Patterns oi edo a Reto tob pto toon 378 Loading and Solving of the MoM Interaction Matrix Equation eee 378 Calibration and De embedding of the S patametets eee stetit denen en haga eR ena 379 Reduced Order Modeling by Adaptive Frequency Sampling sss 379 Special SIMU ato RTL ODIGS esscicucee teque qu eo quan ctet epe ruben dod tton dius ttt cops a ensem ota dacodt 380 Sipyulatine slotsun Ground Panes einem eie e
499. ransmission Line Mesh Use the transmission line mesh when you want to specify the number of cells between parallel lines in a layout This feature can save computation time and memory because it will create a mesh that is appropriate for straight line geometry For example the simulation results for a single transmission line with one or two cells across the width will be equal If you have coupled lines the results will differ Combining the Edge and Transmission Line Meshes Both transmission line mesh and edge mesh can be used together The total number of cells wide will be the total number of transmission line cells plus edge cells The minimum value permitted when using this combination is 3 359 Simulation 360 cellar widtn 3 Transmission Line Mesh Cell width along the length will be equal far this type of mesh calls width 3 Edge Mesh and Transmission Line Mesh cella widt s 5 The edge and transmission line mesh affect the transverse current in the circuit The more cells specified using edge mesh and transmission line mesh the more accurate transverse current approximations will be The plot below illustrates the hairpin example simulated with and without and edge mesh Although the simulation that includes the edge mesh takes more time the results are closet to the actual measured results of the filter Using the Arc Resolution The Arc Resolution parameter enables you to control the number of triang
500. rcles are shown at 0 1 2 3 4 5 and 6 dB less than optimal gain In GA and GP if K 1 then the OdB circle is at GMAX and the inside of this circle is shaded as an unstable region 12 Getting Started Note For port numbers greater that 9 a comma is used to separate port numbers For example on a 12 port device some of the S Parameters would be specified as follows S1 11 812 2 812 11 12 2 Nonlinear Measurements Tip All available measurements and their operators for a given circuit or sub circuit with their appropriate syntax are shown in the measurement wizard To bring up the measurement wizatd select measurement wizard from the graph properties dialog box Here indep is the number of points in the independent vector The short form can be used in most cases GENESYS will automatically create the appropriate formula in the dataset to derive the short form Meas Description Size Short Form VPORT Peak Voltage at all ports indep x nPorts Vport e g V1 7 PPORT RMS Power delivered at each port indep x nPorts Phart e g P2 name IProbe Peak Current through the probe name is the current indep probe designator name Operators Measurements are often combined with operators to change the data format Please read the Using Equations chapter in the User s Guide for more complex equations The general format for combining operators with measurements is standard function syntax operator measurement where operat
501. rcuit applications For most other circuit applications the using the global mesh parameters will be all that is needed to provide greater accuracy In a few special cases you may need to use the layer or primitive mesh controls For applications such as highly accurate discontinuity modeling or for geometries that have tightly coupled lines the default mesh may not be dense enough in particular areas to provide enough accuracy In such cases edge mesh should be used or the mesh density must be increased For example noise floor and dynamic range are typically small numbers In some cases geometry solutions may show a low value of S21 like 60 dB using a dense mesh Such values are different for a default mesh which may result in 40 dB for the same circuit In any one design you can use any combination of the four types of mesh control In general greater mesh density can provide more accuracy But greater density takes more computation time mote cells to solve and additionally triangles always take longer to compute than rectangles Topics in this section Meshing Thin Layers Meshing Thin Lines Meshing Slots Adjusting the Mesh Density of Curved Objects Discontinuity Modeling Solving Tightly Coupled Lines Mesh Precision and Gap Resolution Momentum GX Meshing Thin Layers Thin layers must be meshed so that the mesh cells are entirely within or entirely outside of the overlapping area If the mesh cells and object boun
502. re Verilog A Extensions Eagleware has created several extensions to Verilog A These extensions are not required in any Verilog A files but they allow more complete information to be given to GENESYS about the model making it easier to pass Verilog A files between users In GENESYS a Verilog A file gives a complete description of the model and no other files are generally necessary to share between users Parameter Descriptions First parameter descriptions and units can be included in comments parameter real Vtr 20 0 Soft breakdown model parameter V parameter real P3 0 0 Polynomial coeff P3 for channel current 1 V 3 parameter real Fnc 0 0 from 0 inf Noise corner freq Hz parameter real Cds 0 from 0 inf Zero bias D S junction capacitance F Any comments on the same line as the parameter are assumed to be a description of the parameter Additionally units can be given inside square brackets Currently supported units include Hz Ohm mho H F V A s C deg m W and DB If you use any of these units then other related units such as pF or dBm can be specified when the parts are used Unrecognized units such as 1 V 3 above are simply put into the description so that the user knows what units must be entered for the part Additionally if the comment starts with Unused or Alias then the parameter is not shown to the user in GENESYS and the default value is used If the comment starts with Re
503. re ideal for use with non planar geometries such as a coaxial T junction radar target reflections or other truly three dimensional problems 3 D simulators have the advantage that they can analyze almost any problem but they have the disadvantage that they are extremely slow 2 1 2 D SIMULATORS 2 1 2 D Simulators are simulators designed for mainly planar microstrip stripline etc circuits While they have less flexibility than true 3 D simulators they are much faster and are ideally suited for microstrip stripline and other similar geometries EMPOWER is an advanced 2 1 2 D simulator It can solve planar problems as well as problems with via holes and other z directed currents putting it in a class above true 2 1 2 D simulators which do not allow z directed currents In fact most people would consider EMPOWER to be a 3 D simulator because it can handle z directed currents Features EMPOWER incorporates many features still not present in competitive late generation EM simulators Principle features include e Benchmarked accuracy e Easy to use graphical circuit layout editor e Complete integration with the GENESYS circuit simulation synthesis and layout tools 91 Simulation 92 e Multilayer simulations with EMPOWER ML e Automatic incorporation of lumped elements e Automatic detection and solution with symmetry e Generalized S parameter support e Multi mode support for ports and lines e Tuning of EM objects in
504. re information about e The Method of Moments Technology e The Momentum Solution Process e Special Simulation Topics e Limitations and Considerations The Method of Moments Technology 374 The method of moments MoM technique is based upon the work of R F Harrington an electrical engineer who worked extensively on the method and applied it to electromagnetic field problems in the beginning of the 1960 s It is based on older theory which uses weighted residuals and variational calculus More detailed information on the method of moments and Green s theorem can be found in Fre d Computation by Moment Methods 1 In the method of moments prior to the discretization Maxwell s electromagnetic equations are transformed into integral equations These follow from the definition of suitable electric and magnetic Green s functions in the multilayered substrate In Momentum a mixed potential integral equation MPIE formulation is used This formulation expresses the electric and magnetic field as a combination of a vector and a scalar potential The unknowns are the electric and magnetic surface currents flowing in the planar circuit Using notations from linear algebra we can write the mixed potential integral equation in very general form as a linear integral operator equation Jase rr Arn Pr Equation 1 Momentum GX Here J t represents the unknown surface currents and E r the known excitation of the problem The Green s dyad
505. ree times returning the mode to teal but resetting the time The resulting picture in the main viewer window is a 3D plot of the surface current density shown with the grid generated to solve the problem The axes in the metal plane grid plane correspond to the X and Y axes in the box The origin of the coordinates X and Y correspond to the geometrical origin of the box 0 0 in LAYOUT The z axis perpendicular to the metal plane corresponds to the plotted current voltage values The ted color on the axis is for high values and dark blue is for zero The color coded scale makes it possible to evaluate actual values of current density The plotted values are an additive function of interpolated X and Y components of the current density The current components are calculated along the cell sides not at the corners of the cells The X and Y current components are interpolated to the grid corners and are then added The X Y current display provides general insight into circuit behavior Again consider the view given above The dominant eigenwave of the stripline 1s excited at the left input of the structure Observe the typical current distribution in the cross section X 0 click the side view button for a better look at this At this time the current declines to almost zero at the right output click the Front view button This confirms a line length of 90 degrees Next animate the response by clicking on the Animate button again Notice how the do
506. rential port go to the part selected and search the Eagleware part library for Balanced Getting Started New in GENESYS 2007 Most ports also place an associated EM port in the associated layout This can be bypassed by using for the footprint If ports are used on the master design they must be placed in sequential order Top level ports are required for linear simulation and are strongly recommended for SPECTRASYS Ports are optional for DC simulation HARBEC and CAYENNE Other than the symbols there is no difference between standard input and output ports To use a complex port impedance you can either directly use a complex number in the port termination or you can use a 1 port data file Device Data Linear vs Nonlinear Device Models S parameters for RF and microwave devices are commonly available and easy to measure with a network analyzer They are the most accurate way to model the small signal performance of circuits However they are only valid at a particular operating point bias level Nonlinear device models are also commonly available from manufacturers but they are harder to extract from measurements The advantage of nonlinear models is that they model circuit performance at all bias levels and frequencies Moreover the model characterizes the complete linear and nonlinear performance of the devices including effects such as compression and distortion Linear Data O verview Within GENESYS are a wide range of
507. requency resistance substrate height and DC supply level are just a few of the parameters that are typically swept To create a new Analysis 1 Click the New Item button A on the Workspace Tree toolbar and select Analysis 2 Select which Analysis you want to add 3 Define the analysis properties and click OK or Calculate Now to run the simulation W hich Simulator Should U se Often we at Eagleware are asked which simulation method should be used in a particular circuit Linear Analysis Nonlinear HARBEC SPICE by exporting Electromagnetic EMPOWER SPECTRASYS Simulation For most circuits you will use a combination of the different simulations We have developed several guidelines that should simplify the decision for most applications First each method has benefits and drawbacks CAYENNE Momentum GX EMPOWER Linear HARBEC Steady State Time domain Nonlinear Extremely fast Extremely accurate Extremely accurate Does not require an intimate knowledge of the circuit Does not require an intimate knowledge of Study mixing Schematic or p the circuit simulator compression and Schematic or netlist netlist l l l simulator figures out figures out coupling intermodulation coupling etc etc Real time Starting waveforms Can predict l s i 5 P Can predict radiation DC biasing tuning of e g oscillator radiation current E i on Mp current distribution information
508. res For best results the highest simulation frequency for slotline structures should satisfy Substrate thickness lt 0 15 effective wavelength Simulation of narrow slot lines are accurate at frequencies somewhat higher than the high frequency limit determined by this inequality but wide slotlines will show deteriorated accuracy at this frequency limit Via Structures Limitation In Momentum the current flow on via structures is only allowed in the vertical direction horizontal or circulating currents are not modeled As via structures are metallization patterns the modeling of metallization losses in via s is identical to the modeling of metallization loss in microstrip and slotline structures Via Structures and Substrate Thickness Limitation The vertical electrical currents on via structures are modeled with rooftop basis functions In this modeling the vertical via structure is treated as one cell This places an upper limit to the substrate layer thickness as the cell dimensions should not exceed 1 20 of a wavelength for accurate simulation results Momentum simulations with via structures passing through electrically thick substrate layers will become less accurate at higher frequencies By splitting the thick substrate layer into more than one layer more via cells are created and a more accurate solution is obtained CPU Time and Memory Requirements Both CPU time and memory needed for a Momentum simulation increase with the
509. rily along the line segment as expected Note however the component visualization modes X Y or Z ate mote accurate because the values displayed correspond directly to the values calculated by EMPOWER no interpolation is necessary for these modes The absolute value of the current density is currently displayed Switch to the Real mode using the menu View Menu Switches Absolute Value Display and select Real mode Animation should be turned on also Animation camera button A snapshot of the plot is shown below The Real mode displays both current density values and direction Current flows in the positive X direction if the displayed values are above the metal layer the color coded axis direction The current flows in the opposite direction if the displayed values ate below the metal plane EMPOWER Planar 3D EM Analysis empower Viewer Y6 5 File View x 9 Real Solid Freq GHa 16 Front Side Oblique To obtain even additional insight the phase of the signal along the line may be displayed Stay in X component mode turn off animation and switch to the Angle mode by clicking the Display Option button until it reads Ang You may view the wireframe mode by clicking the Wire Solid button until it reads Wire At the initial time t 0 and with a matching rotation you will a display similar to the one below It displays delay of the current densities along the structure in terms of a complex vector rotation angle 360 degrees
510. rious Free Dynamic Range Receiver 518 oo eee ME ER 86 SS 184 suere rry NC m 11 14 Sta Dit y CU CIOS mn EN 25 stabi TACO cec c c eis 25 Sta DY MCAS UT Oase decreta ttti qii deii dosiel 25 SAO Sy NAMIC EDI OC arrira 519 Stage Equivalent Input Noise Voltage 528 Stao eA A N aia T aa 519 Stage Input 1 dB Compression Point 520 Since Papal Patet e Ede oe erento 520 Stage Input Saturation Powet 520 STAGE MOISE OUTS Go uon EAn 519 stage output 1 db compression point 521 Stave Output Titer Ce ptasia SER 521 stage output saturation POWEL eee 521 stage output second order intercept 521 Si DOS top DRM A NM RUE 91 92 Sb CULE CLOVES aa te Ee eee Pid eiit 149 SUL Strate Pay CE notet eed dts 92 100 substtate thickness esseist 131 substrates BASIC B COMO se cecccssusinnsnsnvnns steccstscscasasdedererorss 92 SUD RCO MC UCLONS qs cic EEan Sa 92 SEA Ce TOUS INC iaa 92 Suspefideci CE OSEPI D meses nette tette do oss 92 SVO A um E 530 va uU M 528 SWCD EEO ooo RR NN eee 257 SWESPS m M 440 STEE M T 129 206 symmetry DEOCOSSIDIE vni 181 S EEEE 481 SS e100 93 08 1010 els ia E T 423 system simulation parameters 536 538 545 548 552 Systern simulatioh ApS eiii 438 iO c 92 Qt E 502 TOP ea TNES 510 FEPEMPERATUR P eer ette 225 OL YON EEE EEEE EEEE E 6 CETAN asonansi 6 ST CRAG MG UG LOE e EM tee cesses see ERRE
511. rm steps from DC to the maximum frequency Initially the minimum number of points is used If the solution quality is not good enough the number of points is doubled up to the maximum specified The truncation values are used to determine if the solution is acceptable If in the second half of the fft impulse waveform the voltage level is above the absolute truncation value or is compared to the maximum at any time relatively larger than the relative truncation then the solution is not acceptable 80 CAYENNE Transient Analysis If Save Impulse Responses is checked then impulse responses for all convolution based elements will be saved into the dataset beginning with the letter H Accuracy Testing is used to determine if convolution is necessary All linear elements are sampled across the specified band using the specified number of points If the resistance and capacitance or inductance for an element do not vary across the band by more than the tolerance amount then the value at the Most Accurate Frequency from the General tab is used for all frequencies If the variance is greater than the tolerance typically 1 as shown above then full convolution is used for that element This accuracy test does not generally work as expected for loss resistances since they are often zero at DC which would be a 100 error when compared to higher frequencies The RLOSS and GLOSS models instead switch to a constant resistance admittance mo
512. roughly proportional to the cube of the number of frequencies So doubling the number of frequencies will take about 8 times longer to simulate However if not enough frequencies are present to adequately model the signals then the results will not be accurate Moreover the simulator may have difficultly converging if not enough of the energy in the circuit is modeled The best practice in selecting order is to start with a reasonable number of harmonics of each signal typically 5 is a good point then increase the number until the results stop changing Order and Maximum Mixing Order on the HARBEC Options dialog box control the number of terms In this way you can make tradeoffs of speed versus accuracy Amplitude Stepping To start the search for convergence HARBEC analyzes the circuit at DC this 1s with all independent AC signal turned off Using DC as a first guess it turns on the signals to Maximum Amplitude Step percentage of full signal If convergence is reached at this step it takes another equal step If convergence is not reached it decreases the step size and tries at the lower signal level Some circuits will converge in a single 100 step Others will require a smaller step to find the solution If a smaller step is required it will be faster to start with that step If the step size is too small the simulator may waste time calculating intermediate steps to find the final solution Convergence speed can be improved
513. round references to the surfaces of object The object must be on strip metallization layers Multiple ground reference ports can be associated with the same port To 5 be associated with a single port the ground reference port should be a port attached to an edge of an object in the same reference plane as the single port To add a ground reference 1 Placea new EM pott in layout ot select any port that you want to assign this type to Note the port numbet 2 Open the port dialog window double click or clicking Enter for selected port 3 Under Associate with port number enter the number of the single or internal port that you want to associate with this ground reference Make sure that the distance between the port and ground reference 1s electrically small 4 Click OK In this design ports 3 and 4 are ground references for internal port 1 and ports 5 and 6 are ground references for internal port 2 Ground reference ports may be associated the only with Normal single 1 mode calibrated ports with same reference plane or with an internal port Any of the ports may be associated with more than one ground reference pott Momentum GX mE EM Port Properties Draw Size E Ref Plane Shift Pork Number Location 250 Layer I TOP METAL Current Direction width in Line Direction Height Port Type jJ PartL ist bp Associate with park number Cz Polarity Ground reference For
514. rrent date is always written e Unit Setup Always output in MHz and mm In manual mode these units can be switched in Sonnet for easier editing e Reference planes Shifted inside the box only Sonnet Interface e Top and Bottom Cover Waveguide load free space normal resistive e Metal Types Normal Resistor TMM e Box Setup Dimensions and cell count come from layout Box Settings e Ports Standard via and ungrounded internal ports Layout X and Y directed ports turn into ungrounded internal ports e Symmetry e Polygons All entities on the layout including text and pours are supported e Via Polygons Supported both by drawing viaholes and by specifying XYZ or Z directed current for a layer e Frequency setup e Both Adaptive Sweep ABS and standard sweeps number of points are supported e ADS Caching Level Target for Manual Frequency Resolution for ABS Sweep and Target for Automatic Frequency Resolution for ABS Sweep are supported e Simulation Options e De embedding e Generate current density e Box Resonance Info e Memory Saver e Multi Frequency Caching e Speed memory accuracy tradeoff e Subsections per lambda e Estimated Epsilon Effective e Maximum Subsectioning Frequency e Polygon Edge Checking e Response file entry Always outputs GENESYS yp file e Quick Start Guide Info Outputs info about tasks completed by Sonnet Interface Sonnet features not supported
515. rrier power This common representation is applicable only to small phase deviations Ps fo Phase noise definition KA M dBc Hz a Oscillator Phase Noise Analysis The simulator computes the noise at an offset from the unknown oscillation frequency After the normal harmonic balance noise analysis has determined the steady state 281 Simulation 282 oscillation frequency and amplitude the phase noise analysis computes the noise amplitude as an offset from the carrier With this noise spectral density Sv f the phase noise can be computed as 4 where Ps is the amplitude of the carrier of oscillator output for which the phase noise is measured There are four distinct regions of phase noise as shown next Note that not all oscillators will show all four regions Phase Noise dependence for frequency offset from a carrier Lif Flicker FIM Noise T F 3 White FIM Noise 1 f 2 Flicker Pil Noise Tif White Pit Noise fa fb fc fnoisa The lowest frequency is dominated by flicker FM noise which is device flicker noise that causes a random frequency modulation This has a slope of 1 f 3 White FM noise is white noise that causes a random frequency modulation This has a slope of 1 f 2 Flicker PM noise is modeled by flicker noise that mixes up to the oscillation frequency This has a slope of 1 f White PM noise is simply white noise that mixes up to the oscillation frequency This
516. rs linear measurement c cccscestecasasasasasasusnconensnens 11 oae aN A 1 partial dielectric loading su assets 160 D E E IUe n OMNIA tn teet 136 i Ido M P 426 538 Pescent Dutethriods t ooa Ge o OD I 517 percent olse fioU e cue dachte tidie 516 percent third order intermod 517 Performing a Nonlinear Noise Simulation 265 CEI CAD HIE Y soupe ida a d o eld 208 PETANI oen NUM M 174 175 DOLIDICUVIOV oen taceam e de 92 208 PRASE 410186 raea rot nde 1 271 475 Phase Noise Channel Powet ote 517 DIE uc E E 155 206 208 ur cR 182 PE wep a cet alc a een e RUNS 180 er rrr reer 517 DISC VEEE ESA 528 ports TCA AICO a A 183 E a secu EST ar er Tey Cm RECA CID 153 RUE a1 8c gereret 11 jeg 308 POSE PLOCESSING cci iio alite a 15 Powers Voltages and Impedances 540 1440119 MR 236 pteterred cell oBtlbusses de Etuis 133 PRIM ee ee hte e d tetas 517 Ig dI E O 517 PRISE ote rcc 516 problem fOPHIldtiOf eee d evo tetas 208 provided device tabac e AA Cpe cones 6 FORNO E riin 271 Q a een 13 R1183 te Douation LUC ON eanais 15 PAULING TUNAO aea een tacito initiis 127 recalculate NOW eei ceu 255 Simulation 562 Receiver mace Channel os tees 540 record Reepinm TIES ery e HH basate 8 REC Porrero AAAA ia 11 13 POCHAN OCULAR Cavity aA 158 159 rectangular w veguldesu anii bts 208 Reduce imaterial node
517. rt Chapter 8 Momentum GX Momentum G X Basics Momentum is an electromagnetic simulator that computes S parameters for general planar circuits including microstrip slotline stripline coplanar waveguide and other topologies Vias and airbridges connect topologies between layers so you can simulate multilayer RF microwave printed circuit boards hybrids multichip modules and integrated circuits Momentum gives you a complete tool set to predict the performance of high frequency circuit boards antennas and ICs This section discusses the Typical Use Process for Momentum GX and how to Select the Correct Mode to use Momentum Major Benefits Momentum enables you to e Simulate when a circuit model range is exceeded or the model does not exist e Identify parasitic coupling between components e Go beyond simple analysis and verification to design automation of circuit performance Momentum Major Features Key features of Momentum include e An electromagnetic simulator based on the Method of Moments e Adaptive frequency sampling for fast accurate simulation results e Comprehensive data display tools for viewing results e Equation and expression capability for performing calculations on simulated data Typical U se Process The following steps describe a typical process for creating and simulating a design with Momentum 1 Create a physical design You start with the physical dimensions of a planar design such as a
518. rt I Summary of results Part II Theory IEEE Trans v MI T 23 1975 N 5 p 421 433 W T Weeks Exploiting symmetry in electrical packaging analysis IBM Journal of Research and Development v 23 1979 N 6 p 669 674 A B Mironov N I Platonov Yu O Shlepnev Electrodynamics of waveguiding structures of axisymmettical microwave integrated circuits Journal of Communications Technology and Electronics 1990 N 7 p 71 76 originally published in Radiotekhnika 1 Elektronika v 35 1990 N 2 p 281 286 E V Zakharov S I Safronov D P Tarasov Abelian Groups of finite order in numerical solution of potential theory boundary value problems in Russian GVM amp MF Journal of Computational Mathematics and MathematicalPhysics v 32 1992 N 1 p 40 58 EMPOWER Engine Theory and Algorithms B V Sestrotetzkiy V Yu Kustov Yu O Shlepnev Analysis of microwave hybrid integrated circuits by informational multiport network method in Russian Voprosi Radioelektroniki ser OVR 1988 N 12 p 26 42 B V Sestrotetzkiy V Yu Kustov Yu O Shlepnev Technique of electromagnetic analysis of microstrip devices using general purpose programs in Russian Voprosi Radioelektroniki ser OVR 1990 N 1 p 3 12 Yu O Shlepnev Method of lines in mathematical modeling of microwave integrated circuit planar elements in Russian Ph D Thesis NEIS Novosibirsk 1990 221 Simulation 222 V Yu Kustov
519. rtsint Nportsext number of external ports Nportsijnt number of internal ports including internal ports automatically created for all pads of the circuit lumped elements 2 Generation of the co simulation design consisting of a A board modeling multiport object defined by Yraw board parameters b Allcircuit external ports and lumped elements connected to the board multiport After the co simulation design has been created Momentum GX e creates a linear solver e calculates the DC operating point of the circuit 370 Momentum GX e linearizes nonlinear elements at the DC point and e simulates the co simulation design at all co simulation frequencies as defined in the Momentum analysis properties tab Finally the linear co simulation creates S Nports Nports Nfreqs S parameters of the full co simulation design ZPORT Npotts Nfreqs reference impedances of the co simulation design ports F Nfreqs array of the co simulation design linear simulation frequencies The co simulation sweep frequencies are defined in the Momentum GX Options dialog General page 371 Simulation Momentum GX Options General Simulatien Options Mesh Name ce Automatic Recalculation Derion Layout Ampifier Automatically save workspace after calculate oai e ERE pem Description Simulation Frequencies Frequency Range Type OF Sweep CO near Number of Points i
520. s 3 hax critical freg 1000 Turn off physical losses Faster Only check errors topology and memory do not simulate Advanced Options Lo simulation sweep Setup Layout Port Modes M Use EM simulation frequencies m Thinning out subgrid Start freq MHz 2000 bor l L dx I 1 I I 1 I F Stop freq MHz 1000 p Number of points 3 Fort number to excite NEN Mode number to excite f M Thin out electrical lossy surfaces Solid thinning aut slower accurate v Use planar ports for one port elements Add extra details to listing file Show detailed progress messages Command line We are starting with 3 sample points in the range 8 11 GHz This will place 1 point at 8 9 5 the supposed resonance and 11 GHz Click the Recalculate Now button This launches EMPOWER to simulate the layout Note While EMPOWER is calculating any simulation which takes more than two seconds a window similar to the one in below will be shown This window shows the current status throughout the calculation mode For more details on the contents window see the Basics Console section 118 EMPOWER Planar 3D EM Analysis EEMI EMPOWER Log Running Workspace layoniy Press Escape to stop the EMPOUER run EMPOWER Planar 3D EM Simulator Version 7 66 C gt 1998 99 Eagleware Corp FREQC 11668 MHz Modet DISC UiewtH gt Losst Thintk gt Summt YZ MIRR Estim time 66 68 61 Each frq 66
521. s The step size can be reduced until the maximum number of simulation points is reached Number of Simulation Points The number of simulation points used for the graph is determined internally in SPECTRASYS This parameter cannot be 551 Simulation 552 changed by the user Since SPECTRASYS can deal with large frequencies ranges the amounts of data collected for a single spectrum analyzer trace could be enormous Furthermore the analyzer function is not a post processing function and the number of simulation points cannot be changed without rerunning the simulation In order to better control the amount of data collected which is proportional to the simulation time SPECTRASYS internally determines the number of simulation points to use Simulation Speed Ups During the system simulation the analyzer will create an analyzer trace for direction of travel for every node in the system Consequently for systems with large number of nodes the convolution routines used to calculate the analyzer traces alone can be time consuming if the analyzer properties are not optimized If simulation speed is important then using the narrowest filter shape will have the best simulation speed File Size The size of the data file will increase when the analyzer mode is enabled Furthermore the file size can grow rapidly depending on the settings of the analyzer mode For example the smaller the resolution bandwidth the more data points are neede
522. s The thinning out procedure decreased the number of the currents in the Combination of these two procedures makes it possible to overcome restrictions of the MoL with a regular grid while keeping the main advantages of the equidistant grid Simulation 218 The described procedure with total elimination of some currents inside the solid metal regions is called the wire model It basically substitutes a problem with another one with removed small metalization pieces It certainly gives an additional error but fortunately this error is opposite to the regular grid model error In other words the wire thinning out model actually increases the solution accuracy if the structure 1s thinned out propetly However if too much metal is removed the thinning out error dominates Thus a solid thinning out model procedure was introduced to avoid it The solid model can be represented as a simple modification of the wire model To explain it we start from the pseudo non equidistant grid of currents formed for the filter and shown above Instead of complete elimination of the currents inside the enlarged grid cells we leave some of them to keep metal surface solid Those currents left are also replaced with just two variables by means of linear re expansion The solid model is more correct but gives a larger number of variables for similarly thinned out problems in comparison with the wire model The solid model is actually a way to form a non equidistant grid
523. s E21 V2 Desired to 21 501 Simulation 502 SystemS21 V2 Desired 1 Linear1_Data S11 setunits SystemS21 dB Setthe units to dB GroupDelay gd SystemS21 Get the group delay NOTE See Spectrasys group delay examples that ship with the product for illustration of this Interferer Channel Frequency ICF This measurement is the frequency of the interferer used for intermod measurements such as IIP OIP SFDR etc The Interferer Channel Frequency is determined set on the Calculate Tab of the System Analysis Dialog Box As with other frequency measurements Spectrasys is able to deal with frequency translation through mixets frequency multipliers etc Image Frequency IMGF This measurement is the image frequency from the input to the first mixer Since Spectrasys knows the Channel Frequency of the specified path it also has the ability to figure out what the image frequency is up to the 1st mixer After the 1st mixer the Image Frequency measurement will show the main channel frequency This measurement will show what that frequency is For example if we designed a 2 GHz receiver that had an IF frequency of 150 MHz using low LO side injection then the LO frequency would be 1850 MHz and image frequency for all stages from the input to the first mixer would be 1700 MHz Power Adjacent Channel Power ACP U or L n This measurement is the integrated power of the specified adjacent
524. s and are relatively slow However the GENESYS AMK includes a Verilog A compiler that creates C code which is compiled yielding simulation times similar to hand coded models Additionally since the derivatives symbolically calculated by the AMK are often more accurate convergence of circuits using Veriloe A models is generally better This often results in a speed up not slow down when using Verilog A models U sing the Additional AMK Models The GENESYS AMK includes additional nonlinear models As of the time of this writing the additional models are Angelov NFET PFET BSIM4 NMOS PMOS EKV NMOS PMOS HiSIM NMOS PMOS Philips JUNCAP 45 Simulation e MEXTRAM NPN PNP e Philips MOS9 NMOS PMOS e Philips MOS11 NMOS PMOS e DPatker Skellern NFET e TFT NMOS PMOS e UCSD HBT NPN To access these models simply place the appropriate part from the schematic nonlinear toolbar or from the part library Creating New Verilog A Models The basic steps in creating a new Verilog A Model are 1 1 We recommend placing new Verilog A models into your GENESYS User Model directory This is generally the My Models directory in your My Documents folder Open the Tools Options Dialog go to the Directories Tab and examine or change the User Model location listed here If you do not place your Verilog A source files in the directory listed here you must specify a full directoty path when referring to the model Note You may add a re
525. s and it is not surprising that a matching network would be less beneficial N oise Circles To achieve the best available noise figure from a device the correct impedance must be presented to the device The impedance resulting in the best noise performance is in general neither equal to 50 ohms or the impedance which results in minimum reflection at the source 298 Linear Analysis The Avantek AT10135 GaAsFET transistor S parameter data given earlier includes noise data This data is comprised of four numbers for each frequency These numbers are NFopt dB the optimum noise figure when correctly terminated Gopt magnitude and angle the terminating impedance at the device input which achieves NFopt and Rn Zo a sensitivity factor which effects the radius of the noise circles Noise circles plotted by GENESYS for the AT10135 at 2500 MHz ate given below Circles of increasing radius represent noise figure degradations of 0 25 0 5 1 1 5 2 2 5 3 and 6 dB In this case direct termination of the device with a 50 ohm source results in a degradation of the noise figure of 1 dB The arc orthogonal to the circles is the locus of Gopt vetsus frequency Response SPARAMISU DB MCI DE MCI DE MCI DE NCI 2000 4000 12000 E20356 3 09804 8 17871 4 43697 6 20356 3 09804 8 17871 4 43697 Smith Chart In 1939 Philip H Smith published an article describing a circular chart useful for graphing and solving proble
526. s associated data from the system Properties Opens the HARBEC Options dialog box Calculate Now Starts a simulation 233 Simulation 234 Automatically Recalculate Toggles on or off the state that starts a simulation any time a change is made to the design Mark results up to date Changes the status of a simulation to current Use this feature when a change has been made to the design that does not affect the simulation results such as changing a value and then changing it back Solving Convergence Issues The simulator searches for a solution until the user specified accuracy is reached or until a specified number of searching steps Sometimes you might run into convergence issues Below are a few steps that you can use to improve convergence results Each of the parameters below is changed on the Harmonic Balance HARBEC Options dialog box 1 Increase the number of frequencies the order used in analysis If not enough frequencies ate used the data is being undersampled and cannot accurately represent the solution For example modeling a square wave with three harmonics will ignore a lot of energy in the circuit often leading to convergence issues Increasing the number of frequencies analyzed will more accurately model the signals at the expense of more time 2 Try Always and Never options for calculating the Jacobian If a Jacobian is calculated the simulator will search in a different direction from the F
527. s into a port in the final data file Ex It E Ey Ez Jy Y p BxJx 98 EMPOWER Planar 3D EM Analysis Note It is possible to make a line so narrow that it maps to one border between cells zero cells wide This is legal but is not normally recommended and should be used only for very high impedance lines where accuracy is not important such as DC power lines The grid and the box are controlled with parameters in the Preferences box from the LAYOUT File menu The Dimensions Tab shown below is as it was setup for the microstrip bend above Create New Layout x General Associations General Layer EMPOMWER Layers Fonts Units Box Settings The UNITS box at left show units Designs to include Desin s Peete Grid Spacing X 20 W Show Box ee Grid Spacing Y 20 Iv Show Grid Dots M Show EMPOWER Grid Bos width Fs 300 ih 5 Cells Box Height Mr 300 ii p Cells Origin E o Layout J Object Dimensions Line width 30 Pad width 80 Drill Diameter 50 Drawing Options Port Size 20 widths Rat Snap Angle 30 10 2n Remove Multi Place Parts 3l Default Viahale Layers Top Layer TOP METAL Bottom Layer ER Bottom Cover 3n Add New The following entries are especially relevant to EMPOWER Show EMPOWER Grid Turning on this checkbox forces LAYOUT to display the rectangular EMPOWER grid It also allows different grid spacings
528. s is adequate However if one desires to examine an undesired intermod or harmonic though a filter then more points may be needed to accurately represent the shape of the signal When unchecked the average number of points from all input signals is used to represent the undesired intermod or harmonic Noze Desired intermods and harmonics like the principle signal coming out of a frequency multipher or divider will always be represented by the number of points of the input spectrum for a harmonic or the average number of points from all input signals in the case of an intermod Maximum Order This parameter is used to limit the maximum order of the spectrums created in the simulation This limit applies to all non linear elements Each model has a limitation on the maximum order that it can generate Please refer to the element help to determine the order limit for each model CALCULATE NOISE This section controls calculation parameters for thermal noise When checked noise is calculated The option must be enabled for path noise measurements Every component in the schematic will create noise A complex noise correlation matrix is used to determine the noise power for each element at every node Unchecking this option can improve the simulation speed drastically during troubleshooting of a block diagram System Temperature This is the global ambient temperature of the entire design under simulation This is the temperature needed to determ
529. s not zero a signal path exists from the output to the input This feedback path creates an opportunity for oscillation The stability factor K is ES be oul sioe pum psu where D S11822 S12821 From a practical standpoint when K gt 1 S11 lt 1 and S22 lt 1 the two port is unconditionally stable These are often stated as sufficient to insure stability Theoretically K gt 1 1s insufficient to insure stability and an additional condition should be satisfied One such parameter is B1 which should be greater than zero B1 1 S11 S22 DD 40 Linear Analysis Stability circles may be used for a more detailed analysis The load impedances of a network which ensure that S11 1 are identified by a circle of radius R centered at C on a Smith chart The output plane stability circle is Cow S22 DS S22 D 2 Rout S2521 S22 D2 This circle is the locus of loads for which S11 1 The region inside or outside the circle may be the stable region The input plane stability circle equations are the same as the output plane equations with 1 and 2 in the subscripts interchanged Shown in the figure below are the input plane stability circles on the left and the output plane stability circles on the right for the Avantek AT10135 GaAsFET The shaded regions are potentially unstable At the input the stability circle with marker 1 indicates sources with a small resistive component and inductive reactance
530. s spectral domain approach Note that a backwatd process is impossible and a simple truncation of the series does not give the same answet as the grid technique The finite sums and the grid corrections are the most important things for monotonic convergence of the algorithm To construct the GGF matrix in the grid spectral domain the impedance form of the solution for a layer was used The base of the solution is a layer admittance matrix in the grid spectral domain This matrix relates the grid analogues of the tangential electric and magnetic field components at opposite surfaces of the layer z directed currents and integrals of z directed grid electric field along the z directed current inside the layer All of these are in the basis of the grid eigenwaves thus we have a set of independent matrices for each pair of grid eigenwaves Uniting those matrices for all layers in a structure gives a grid spectral GGF representation The construction procedure is completely automated for arbitrarily layered configurations This technique is similar to the impedance approach in the spectral domain Uwano Itoh 1989 The grid spectral GGF representation was also called a GGF eigenvalue vector but that term is not quite correct The dimension of the vector is about 3 L M if there 1s only one signal layer All we need now to get the GGF matrix in the initial space is to perform a backward transformation of the GGF eigenvalue vector from the grid spectral dom
531. s weenie 225 Tet pane shifts UE PIS 136 Tel Crence PANE ous a battere 139 145 151 reflection coefficient 292 296 297 299 relative dielectric Constante cc et teet 92 Vela ve IGEEOP oc det EE M 225 Relative Error Jacobian recalculation 225 Relative Numeric Derivation Step 225 relative PEE VIL cossera iaaa 92 telative toleranceu e Bae tence seen 225 234 KAO 86 Residual INGtim Ty pesani 225 eC ee 8 E A rrr Tr errs seams 11 FOIS TD E AE LU iE 92 LOS Gia ACG sss beatae ee MALI IS APA 131 157 160 Reuse Jacobian At Most sss 225 RGT rU 183 doo 92 Richardson s Extrapolation 221 R N ENCANTAN E 8 11 IOUDD OsS cused aad 92 RPNOIS E T Y 244 eT UP 183 NE TL CORPS 509 poaae lc n 515 lo OID RR P 518 sample pieds Eee ES edd eR Etre 14 Save solution forall nodes totes 225 BB M A 25 PECE E Mp nnn DIE EE 11 SBA e CIOS ES H DID op D ER eR e 519 semi infinite waveguide uuo cat tariis 92 SEDSE VIL oec osa pu AH RA DA eT 298 VID I HH 518 IN eicusd erect UE 519 sional metal efe ets tees vesc mnsaetectese tede 160 eee HMM DM MT tees andaare 520 SUN m 529 Sa SiS EN ET TT 1 simulation VEES EAE EEEE E 289 poti defauftuss Qaae a 14 SU AON S pee edt ico Ness MR ve citi 486 simultaneous match impedance 11 SitibIe patt AOE llo osossaracecasencuctasisasdtstacs
532. sabled 484 Spectrasys System Filter 3 Properties General Parameters Simulation Custom Schematic Element Netlist Simulation Parameter Override Use Parameters and Model as Entered CO Disable Part For All Simulations Open Circuit CO Disable Part For All Simulations Short Circuit ALL terminals together Filker3 Schematic A Use Dataset Port Filter3_Analysis_Data a Layout Options Use Standard Part in Layout CO Replace Part with Open CO Replace Part with Short See the specific synthesis section for more information about each synthesis tool Directional Energy Node Voltage and Power When three or more connections occur at a node a convention must be established in otder to make sense of the information along path NOTE The path value reported for a node along a path that has more than three or more elements is the value seen by the series element in the path entering the node For example in the following example we have defined two paths Path1_2 which is the path from node 1 to node 2 and Path3_2 which is the path from node 3 to node 2 On a level diagram or in a table the value reported at node 5 for Path1_2 would be the value of the measurement leaving terminal 2 of the resistor R1 entering node 5 Likewise the impedance seen along this path 1s that seen looking from terminal 2 of the resistor R1 into node 5 Consequently the impedance seen by R1 is the L1 to port 3 network in
533. se Relative FNOISE power of AM PM SSB noise correlation Relative FNOISE power of Noise contributors Noise NCIndex Ncontrib contributors names The number of noise carriers A carrier is a spectral component of HB analysis solution Ncarriers 1 if noise calculated for one carrier and Noise lt indep gt contributors indexes noise sources of all circuit elements HARBEC Harmonic Balance Analysis Npoints Ncontrib real real real complex real String Integer 275 Simulation Some examples of the plotted dataset variables are shown below Oscillator SSB noise measurements for 1st harmonic of output spectrum dBc HBOSCI NPM osB noise of 1st Harmonic of Oscillator Output Spectrum 40 54 A9 co he co T i i o EL E 3 LL 1000 10000 100000 Frequency Hz RPNOISE 4 RAPI Notice that the noise is at least 45 dB below the fundamental resonant frequency The simulated noise sidebands of 1st harmonic of the oscillator are shown next The spectrum extends from Fc 1 MHz to Fc 1 MHz and is assumed to be centered relative to the carrier frequency where Fc is the carrier frequency In this case Fe Fosc Fosc is the steady state oscillations frequency Oscillator noise spectrum around 1st harmonic 276 HARBEC Harmonic Balance Analysis HBOSCT PSPHOISE PSPNOISE a e i kJ LU bt db Lu FE fpf
534. se to calculate If this happens simply right click the Sonnet simulation on the Workspace Window and select Delete Internal Simulation Data Manual Mode In addition to the normal design flow described above Manual Mode design flow is available Manual Mode should only be used if you need to use a Sonnet feature not available through the GENESYS interface Before attempting to use Manual Mode you should be very familiar with the operation of Sonnet The design flow in Manual Mode is as follows 1 Create a layout for electromagnetic analysis In this step you must at least place EMPOWER ports down on the layout as GENESYS counts these ports to determine how many ports to expect in the Y Parameter data 2 Create a Sonnet analysis 3 If desired recalculate the Sonnet analysis using the normal flow mode 4 Right click the Sonnet simulation on the Workspace Window and select Manual Mode 5 Right click the Sonnet simulation on the Workspace Window and select Manual Mode Create Sonnet Files and Export to Disk This step exports the Genesys son and Genesys yp files to the WorkSpace Sonnet VSonnet1 directory after optionally recreating the Genesys son file from the current layout 6 Right click the Sonnet simulation on the Workspace Window and select Manual Mode Edit in Sonnet Native Editor Make any desired changes to the Sonnet file The following restrictions apply a Do not change the number of ports in the Sonnet file
535. sed Same as DCV Travel Direction Same as DCV Node Noise Voltage NNV This measurement is the peak average noise voltage at the node along the specified path This includes all noise signals both in and out of the channel Default Unit dBV Channel Used No channel is used for this measurement Types of Spectrums Used ONLY NOISE Travel Direction Only spectrums traveling in the FORWARD path direction Offset Channel Voltage OCV The Offset Channel is a user defined channel relative to the main channel The Offset Channel Frequency and Offset Channel Bandwidth are specified on the Options Tab of the System Analysis Dialog Box As with the Channel Frequency measurement Spectrasys automatically deals with the frequency translations of the Offset Channel Frequency through frequency translation devices such as mixer and frequency multipliers For example if the Channel Frequency was 2140 MHz Offset Channel Frequency was 10 MHz and the Offset Channel Bandwidth was 1 MHz then the OCV is the average voltage from 2149 5 to 2150 5 MHz This measurement is simply a Channel Voltage measurement at the Offset Channel Frequency using the Offset Channel Bandwidth 527 Simulation 528 Channel Used Offset Channel Frequency and Offset Channel Bandwidth Types of Spectrums Used Same as CV Travel Direction Same as CV Phase Noise Channel Voltage PNCV This measurement is the average phase noise vol
536. sed in the AFS process In general this is not necessary and is in fact discouraged but there may be instances where it is beneficial An application where this may be beneficial 1s simulating a structure that has a distinct variation in response at some point over a frequency range such as a resonant structure To set up the simulation 1 Select the Single sweep type enter either the value of the resonant frequency ot a value neat it in the Frequency field and add this to the frequency plan list 2 Select the Adaptive sweep type set up the plan then add it to the list of frequency plans For this situation the sample point aids the AFS process by identifying an area where there is clearly a variation in the response of the circuit Applying extra sample points may be necessary for visualization or far field calculations It is not beneficial to attempt to add sample points for other purposes such as attempting to force smoothing to occur at specific points without taking into consideration the response at these points For example 1 Selecting the Linear sweep type adding sample points spread linearly across the frequency range and adding the frequency plan to the list 2 Adding an Adaptive sweep type selecting the same frequency range and adding it to the list For this situation the sample points from the linear simulation will be included in the adaptive simulation This impedes the AFS process of selecting the opti
537. shown below NOTE The path name may be different because this is based on the dataset and path name RIM D maa l I 5 zb System Data New Power Plot at Node 2 Add New Graph Table gt Soure pe System1 Data New Voltage Plot at Node 2 Ema F INE 4000NHe L 10dB System1 Data New Phase Plot at Node 2 Beatie System1_Data_Path1 New Level Diagram of CP Channel Power new PTone 20dBm System1 Data Pathi New Level Diagram of CGAIN Cascaded Gain bss dares System1_Data_Path1 New Level Diagram of GAIN Stage Gain v Keep Connected System1_Data_Path1 New Level Diagram of CNDR Carrier to Noise and Distortion Ratio wv Show Part Text System1 Data Pathi New Level Diagram of CNP Channel Noise Power System1 Data Pathi New Level Diagram of CNF Cascaded Noise Figure ParlList gt Schematic Systemi Data Pathi Mew Level Diagram of SDR Stage Dynamic Range Properties Systemi1 Data Pathi New Table of Measurements Schematic Properties An The following default table will appear aystemd Data Path1 Measurements E a E Wemwmes Poris Cro cP CMP GAN Coa GMR On Sm L sows 4X0 sass aae ofo ol e O mesel 8 Ti w me naa 08 soe neoa 120M 107927 8 Ai 40 verze esra 4 sore BON 120M Misa Fz 7 RFAmp2 4000 66 451 106 502 9 994 15 072 18 415 12 252 21 932 8 32 A3 4000 1451 111 485 10 072 18 415 12 269 106 932 Identifying Spectral O rigin Since each spectrum is tracked
538. sics section for more information on cells and the problem geometty Maximum Critical Frequency This parameter is set in the EMPOWER dialog box when starting a simulation Changing this parameter has three and only three effects 1 The maximum amount of thinning out is affected EMPOWER will thin out until an area is 1 20th of a wavelength at this frequency in the default thinning mode 2 Thelength of line analyzed for deembedding is 1 2 wavelength at this frequency in automatic mode 3 Many parameters in the listing file are based on this frequency The most important thing to know about maximum critical frequency is to keep it the same between runs of the same problem even if you are changing the frequency range which you are analyzing If it is changed then the thinning out 1s changed and the entire problem geometry is slightly different As an example if you are analyzing a filter with a passband from 5 1 to 5 5 GHz with a reentrance mode additional passband around 15 GHz you should probably set the maximum critical frequency to 5 5 GHz This is because the exact characteristic of the reentrance mode probably is not important critical you just want to know approximately where the filter re enters On the other hand you want to know precise y where the passband is so you set the maximum critical frequency above it The effect of maximum critical frequency is generally secondary Most of the other choices in the tabl
539. sing GENESYS has been performed on a schematic and an EMPOWER simulation is desired or when any lumped elements are needed in the EMPOWER Simulation In addition to the schematic objects any desired LAYOUT objects can be added to the board before simulation For example linear simulation would normally not include ground pours power supply rails and lumped element pads However these are included in the EMPOWER run allowing inspection of their effects Box Dimensions Note In EMPOWER the layout s box dimensions are used to define the bounding box Double click on the layout to open the LAYOUT Properties dialog box The box dimensions are shown below Box Width was chosen as 425 the width of the filter since there are two 200 mil lines and a stub width of 25 mils The filter height is 275 mils including the stub length and series line width The box height was chosen as 600 mils to give plenty of spacing on either side of the filter This minimizes wall interference in the filter s frequency response EMPOWER Planar 3D EM Analysis Create New Layout BEN TOP METAL Beton Cover The EMPOWER grid settings for this example are shown in the upper right above EMPOWER simulation time is greatly reduced if dimensions are chosen so that metal lies exactly on as large a grid size as possible The grid width and height settings for this filter were chosen as 12 5 since the filter dimensions 425x275 are exactly divisible by this v
540. sistors this is almost always 2 When the datafile part is used in a design and a simulation takes place the datafile is imported into the workspace it takes the short name of the datafile and that imported dataset becomes the cache for the datafile To have the datafile re imported delete the cache dataset The NPOD and NEGD elements use in workspace datasets as input These parts are recommended when using results of one linear simulation in another design Search for dataset in the Eagleware part library to find the dataset part s Provided D evice D ata GENESYS includes over 25 000 data files for many different device types Device data was provided directly by the manufacturers in electronic format Note To reduce the size of the installation all S Parameter files are zipped to begin with The user should unzip the needed libraries from the Genesys SData directory which is by default C V Program Files Genesys20XX XX SData these files can then be referenced by N Port linear blocks Caution Eagleware could not test every file that was provided Through random sampling we edited errors found in some files It is the user s responsibility to test each file for accuracy Creating New Linear Data Files You may easily add other devices to the library by using a text editor such as NOTEPAD to type the data into a file with the name of your choice Be sure to save the file in standard ASCII format The first line in the fil
541. sition as shown here We strongly recommend that you write a similar plan on paper when you setup a problem for multi mode analysis Parte Part 1 Part2 The first step is to create workspace with a layout for each unique piece In this example there are two unique pieces The lower left corner is the first and each of the other three corners which are identical There are two basic methods for creating these pieces e Create the pieces individually drawing only the part that will be simulated in each piece In this case each individual layout will look like the parts shown above Or 5 e Create a complete layout of the entire problem first Then make the box smaller so that only the desired piece is simulated This is the method we will use for the spiral 145 Simulation We have created a layout of the entire spiral inductor as a starting point EAGLE EXAMPLES DECOMP FULL WSP This file was created by starting with an MRIND element so that the layout was created mostly automatically The only addition was the extra leneth leading to port 1 and the EMPorts Notice that the reference plane for port 1 is shifted to the actual start point of the spiral model Port 2 is an internal port This circuit can be analyzed directly but it requires minutes per frequency point and 37 megabytes of RAM This file was then saved as COMBINE WSP The box was shrunk and the circuit was moved so that only the bottom left quarter of the circui
542. spiral inductor above is broken down into four rectangular areas one for each corner These sections ate then connected with multi mode transmission lines In each of the circuits above the three main advantages of decomposition can be seen e The lengths of the connecting transmission lines can be varied In the spiral inductor this allows the size of the spiral and the inductance to be tuned or optimized in GENESYS e Far fewer points need to be analyzed This is because each of the pieces is simpler and interpolation works well For example in the edge coupled filter each of the pieces contain only open ends and small sections of lines which do not resonate As a result this filter only needed 5 frequency points for a good analysis e With any of these circuits The grey areas can easily get so large that the problem requires hundreds of megabytes to analyze In the meander line if the lengths of the coupled lines grey areas gets very long the EMPOWER simulation could take a long time When the circuit is decomposed simply changing one length value in GENESYS gives a virtually instant analysis no matter how long the coupled sections ate EMPOWER Planar 3D EM Analysis Spiral Inductor Example NOTE The Single and Multi Mode transmission lines required to use decomposition are not available in GENESYS As a first decomposition example we will analyze a spiral inductor The first step is to come up with a plan for decompo
543. splayed They can display Individual pieces of spectrum including signals intermods and harmonics thermal noise and phase noise The total spectrum comprised of all individual pieces of spectrum in every direction through the node Spectrum analyzer trace for each total spectrum Spectrum like signals intermods and harmonics thermal noise and phase noise can be grouped and shown instead of individual spectrums Click here for additional information on controlling what types of spectrums displayed Easiest Way to Add a Spectrum Plot The easiest way to add a spectrum plot 1s to right click the node of interest then select the desired plot from the Add New Graph Table submenu as shown below 431 Simulation Systeml Data New Power Plot at Mode5 System Data Mew Voltage Plot at Made 5 ut picis ae SURE RFA l Systemi Data Mew Phase Plot at Node 5 VIEW Tc 4uB L uunm Find Part In Layout Add Mew Graph Table w Keep Connected w Show Park Text Properties Schematic Properties The spectrum plot will then appear System PWR_at_Node_45 1500 s000 4500 6000 500 YOUU 10500 12000 13500 15000 Frequency MHz Easiest Way to Add a Table 432 Spectrasys System The easiest way to add a path table is to right click THE NODE WHERE THE PATH ENDS then select the System1 Data Path1 New Table of Measurements from the Add New Graph Table submenu as
544. ss Fileer Bu Bandpass Butter Sede 4 ED o amp Of Create the following system schematic default parameters for all models will be used For additional help creating a schematic click here Couplerl 1 IL 0 5dB CPL 20 0dB Port_ CVYSoOurce_1 Perm F 100MHz G 2 dB L 3 0dB ges Sa NF 3dE Port 3 Isalator 1 zO 500 IL SdE 1 Select the RF Amplifier 2nd 3rd Order from the system toolbar or part selectot 2 Move the cursor and click inside the schematic window to place the part 3 Use the prior steps to place a fixed attenuator single directional coupler and isolatot 415 Simulation 4 Place a CW source at the input 5 Place a output port on the output of the isolator and the coupler HINT Press the O key on the keyboard to place an output port 6 Make sure each element output is wired to the subsequent element input HINT Use the F4 key when a part is highlighted to repeatedly move the part text to default locations around the part Note The node numbers seen on your schematic may vary due to the order of the parts placed on the schematic To Renumber Nodes select the schematic then select Renumber Nodes from the Schematic menu The following dialog box will appear Renumber nodes Select Rename all nets only rename nets connected to ports CO Rename all nets that have default names Select the desired options and click OK Back to Overview
545. ssed data Now add two parameters sweep and check the propagate all variables checkbox to make LSy sweep along with the built in variables See the below example where the above steps are done to verify that for an RLC circuit the Large Signal S Parameters match Sij perfectly Large Signal S Parameter Linear Test The Large Signal S Parameter Linear Test 2 tones shows up using the 2 nd method of the LS parameters calculation Intercept Point Calculation NEW Genesys 2005 and later now has special functions calculating IP n for any 2 components of a spectrum for any number of signal tones of Harmonic Balance analysis for input hb iipn SpectrPout FreqIndexIM IndexS1 IndexS2 PindBm or output powet hb oipn SpectrPout FreqIndexIM IndexS IndexIM To propagate these in sweeps the functions need to be called directly in declared variables of the HB analysis dataset and the flag Propagate All Variables When Sweeping of Parameter Sweep analysis must be set For example the tested circuit has input signal port with 2 tones signal of equal powet 35 Simulation F1 F0 delta F 2 F2 F0 delta_F 2 PO 35 dBm F0 850 MH and delta_F 10 RHz P1 P2 P0 3 01dBm F F1 F2 P P1 P2 i a By x Amplifier Poit 2 PORT 2 ZO 50 Q 20 U1 Pt d DUT 0 I PORT 1 ID R 50 0 z0 F 849 995 850 005 MHz F PAC 38 01 38 01 dBm F ELI CET IM1 27 025 hb getspcompdbm P 2 Freq
546. sstse totae am ee ertrne tais etat teas pe bietet 540 System Simulation Parameters Calculate Labssiisseenni n btt ditis dts 545 System Simulation Parameters Composite Spectrum Tab ees 548 System bumulation Parameters pons Dabo nevera e teri a 552 System Simulation Parameters Output Tab ous eiue io ien EUH Spe e bU I Ue 595 M 557 xiii Chapter 1 Getting Started Simulations GENESYS supports several different types of analysis allowing the exploration of a complete range of circuit performance A simulation run is when you run an analysis on a circuit e DC Simulation nonlinear a part of HARBEC e Linear S Parameter Simulation e Planar 3D Electromagnetic EM Simulation EMPOWER e Planar 3D Electromagnetic EM Simulation Momentum GX e Harmonic Balance Simulation nonlinear HARBEC e Spectral Domain System Simulation SPECTRASYS e Transient Simulation nonlinear time domain CAYENNE Additionally the following items are available e Parameter Sweep e TESTLINK Covered in the User s Guide Several of these capabilities work together EM co simulates directly with the linear and DC circuit simulators and indirectly with HARBEC SPECTRASYS and CAYENNE combining the accuracy of EM analysis with the generality and speed of circuit simulation Parameters sweeps can be used with any analysis type as well as with other sweeps F
547. st be reduced to the spreadsheet case That is 1 Remove VSWR and frequency effects a Behavioral filter have return loss which is a function of the ripple Set the ripple to something really small like 0 001 dB b Set all ports and stages to the same impedance c Replace S Parameters elements or other frequency dependent elements with attenuators or amplifiers of the equivalent gain 2 Remove sneak path effects a Set isolations very high 100 dB Spectrasys System b Set reverse isolation very high 100 dB 3 Remove gain compression effects a Gain compression is based on total node power not channel power All unwanted signals including noise will contribute to the total node power b Increase the P1dB PSAT IP3 and IP2 points of all non linear stages 4 Remove image noise effects a Set the image rejection high 100 dB in all mixers be sure to reject the image frequency band not the desired channel band After making these changes you will get excellent correlation See Cascaded Noise Figure Equations and Cascaded Intermod Equations for additional information No Attenuation Across a Filter There are two kinds of loss produced by a filter 1 dissipative resistive and 2 impedance mismatch The dissipative loss of a filter is the same as its insertion loss For in band frequencies the dissipative loss is typically more significant than the mismatch loss However for out of band frequencies the m
548. stance and an appropriate internal inductance to form the complex Momentum GX surface impedance At high frequencies the current flow is dominantly on the outside of the conductor and Momentum uses a complex surface impedance that closely models this skin effect Horizontal currents on the side metallization of finite thickness conductors and conductors with arbitrary height width ratio can be accurately modeled with Momentum The surface impedance model for thick conductors includes mutual internal coupling for currents at the top and bottom plane of the thick conductor The meshing density can affect the simulated behavior of a structure A more dense mesh allows current flow to be better represented and can slightly increase the loss This is because a more uniform distribution of current for a low density mesh corresponds to a lower resistance Losses can be defined for ground planes defined in the substrate definition This uses the same formulation as for loss in microstrips Le through a surface impedance approximation It should be noted however that since the ground planes in the substrate description are defined as infinite in size only HF losses are incorporated effectively DC losses are zero by definition in any infinite ground plane DC metallization losses in ground planes can only be taken into account by simulating a finite size ground plane as a strip metallization level For infinite ground planes with a loss conductivity s
549. step are used as the initial guess Time Step Methods Fixed step mode forces the simulator to always use the maximum step size Robust mode uses a sophisticated test of predicted vs actual values to determine if the time step needs to be decreased The Truncation Factor and Relative Tolerance values on the General tab are multiplied together to determine acceptable 79 Simulation error If the error is too large the solution point is discarded and a smaller step from the previous solution is used In Approximate mode if accuracy is not acceptable then the next time step is decreased no solution is ever discarded Transient Analysis Properties General Integration Time Step Convolution Output Miscellaneous Impulse Responses Maximum Frequency Minimum Number of Points Maximum Number of Points Convolution Relative Truncation Convolution Absolute Trunckion Save Impulse Responses Accuracy Testing Maximum Test Frequency Most Accurate Frequency 2 or ETUR Number of Points Tolerance 0 01 Always Use Constant Loss Not Frequency Dependent Losses Impulse Responses are used to calculate results when only frequency domain models are available The maximum frequency should correspond be at frequency where every device needing convolution like transmission lines is at least two wavelengths long An FFT is performed using unifo
550. sys System SIPIDB Stage Input 1 sip1db Same as SIPIDBhi SOPIDB dB SOPIDB SOP1DB amp SGAINI Compression SGAIN SGAIN Point SIIP siip SOIP SIIP i SOIP i SGAIN i SGAIN SIIP2 sipn X 2 SIIP for 2 4 Order SIIP3 sipn X 3 SIIP for 3 4 Order SIPSAT sipsat Same as SIPSAT SOPSATTi SOPSAT SOPSAT amp SGAINT SGAIN SGAIN SOP1dB SOP1dB Stage entered value SOIP SOIP Stage entered value SOIP2 Stage 2 d Order SOIP2 None Stage entered value Output SOIP3 SOIP3 Stage entered value SOPSAT Stage Output SOPSAT None Stage entered value Saturation CIMCP Intermod CIMCP Same as CIMCPR TIMP 1 Channel Power TIMP amp g2zin DCP i Where CIMCP 0 CIMCP2 cimcpn Same as CIMCP for 274 Order CIMCP 2 CIMCP 495 Simulation 496 CIMCP3 cimcpn CIMCP 3 Same as CIMCP GIMCP Generated Intermod GIMCP2 gimcpn GIMCP 2 GIMCP2 gimcpn GIMCP 3 TIMCP Total Intermod Order N TIMCP2 Same as TIMCP TIMCP3 Same as TIMCP Channel Power TIMP VTCP irtual Tone Same as Channel Power ICP and CGAIN otal Intermod TIMP Total Power Intermod TNP otal Node TNP Total Power TCP CIMCP for 3 d Order Generated intermod power at CF for all orders GIMCP for 2 d Order GIMCP for 3rd Order Total intermod power for all orders at CF TIMCP for 2 4 Order TIMCP for 3 Order Total intermod power at CF Power of entire spectrum at
551. system is to interference In a cascade of RF behavioral models a diagram similar to that above can be derived to determine the overall system performance The cascaded intercept point 1s generally referred to the input ot output for convenience Transmitters and amplifiers generally have their intercept points referred to the output and receivers have them referred to their input Intermod and Harmonic Basics This section will help the user understand fundamental relationships between intermods harmonics and intercept points When Calculate Intermods and Calculate Harmonics are enabled intermods and harmonics will always be created by nonlinear behavioral models The Maximum Order parameter on the Calculate tab of the system analysis determines the maximum intermod and harmonic order used in the simulation Here is an example of the output spectrum of an amplifier with a two tone input 445 Simulation 446 2 Tone Nonlinear Spectrum E ZEN z Output Spectrum dBm Er LLLI Texestl FTT Enem eaae 20 100 150 200 250 300 350 400 450 500 Frequency MHz The 2 tones are located at 100 and 125 MHz Notice that the bandwidth of 2nd order products is twice that of the fundamentals and the 3rd order products are 5 times the bandwidth of the fundamentals The amplifier OIP3 is 30 dBm and the OIP2 is 40 dBm NOTE The channel measurement bandwidth must be set to at least 3 times the bandwidth of the fundamental t
552. t above Set the Current Direction of the EMPort to Along Y along the y axis if the current along the component flows from top to bottom as if the capacitor were turned 90 degrees from the one on the layout above e Connecting a lumped element from the port to ground when you use the resulting data is equivalent to connecting the lumped element accross the length of the port in the layout This does not mean that the component is gtounded It simply means that the component is connected across the port This concept is key to understanding X and Y directed ports When the S Parameters of MYNET are displayed in a graph you see the resulting S Parameters of the entire circuit Resonance Often when a circuit contains lumped elements you can use very few frequency points for the EMPOWER runs Since the lumped elements are not included in the EMPOWER data there are generally many fewer resonances and the data interpolates much mote accurately In this case you may want to only use 2 or 3 points in the electromagnetic analysis while showing the results of the entire network with 100 points or more specified in the Co Simulation Sweep in the EMPOWER Options Dialog box For a complete example which takes advantage of this property see the Narrowband Interdigital example EMPOWER Box Modes O verview A fully enclosed rectangular box acts as a cavity resonator At frequencies near each resonance mode significant coupling ex
553. t at Node 4 CWWSource 1 idin System1_Data New Phase Plot at Node 4 View F 100MHz Find Part In Layout Pwr 2 0dBm wv Keep Connected w Show Part Text ParlList gt Schematic Place Another OUT OUTPUT Properties Schematic Properties The following graph will appear 420 Spectrasys System System PWR_at_Node_4 oystem1 PWR at Node 4 Frequency MHE 4i P4 To add a level diagram a path number be defined first right click on the ending node of the path and selecting System1 Data Path1 New Level Diagram of CGAIN Cascaded Gain from the Add New Graph Table submenu Coupler1 1 IL 0 5dB CPL 20 0dB 1 System1 Data New Power Plot at Node 7 Systemi Data New Voltage Plot at Node 7 System1 Data New Phase Plot at Node 7 CVV Source_1 System1 Data Path1 New Level Diagram of CP Channel Power F 100MHz System1 Data Pathi Mew Level Diagram of CGAIN Cascaded Gain Pwr 2 dBm System1 Data Path1 New Level Diagram of GAIN Stage Gain v Keep Connected System1 Data Pathi New Level Diagram of CNDR Carrier to Noise and Distortion Ratio wv Show Part Text System1 Data Pathi Mew Level Diagram of CNP Channel Noise Power Systemi Data Pathi Mew Level Diagram of CNF Cascaded Noise Figure System1 Data Pathi New Level Diagram of SDR Stage Dynamic Range Properties Systemi Data Pathi New Table of Measurements Schematic Properties Format View Find Park
554. t is in the box The number on the internal port on the end of the spiral was changed to 10 Ports 2 5 on the right and 6 9 were added Since these ports are in the middle of a line instead of on the end their width must be set manually Also the reference planes on the ports were shifted in The resulting layout for the first piece is shown below EMPOWER was run for Part1 The settings are as shown below Noze that only 5 points are needed since the individual parts are not resonant The Setup Layout Port Modes button was 146 EMPOWER Planar 3D EM Analysis clicked and the checkboxes in the Setup Modes dialog box were set to indicate that those inputs are modally related Caution Do not forget to setup the modes when you are analyzing by decomposition The Mode Setup box turns red if any inputs are modally related Improper mode setup is one of the most common errors in decomposition E etup Modes A similar set of steps was followed for Part2 The final step in decompositional analysis is to combine the pieces The Schematic COMBINE which does this is shown here The pieces used are NPO10 and NPO8 blocks under DEVICE in schematic for the data in PART1 and PART2 plus MMTLP8 models multi mode physical transmission lines found under T LINE for the interconnecting lines The data for the NPO10 is in WSP Simulations EMPart1 EMPOWER SS and the data fo
555. t is not converged and iterations continue Additionally even if the criteria are met if the value of the solution at any node changed by more than VoltTol RelTol Value then iterations continue since significant improvements are still being made After convergence has been achieved the solution is tested for accuracy using criteria defined above in Simulation Time Steps Integrator and Predictor 76 The Integration Time Step tab controls the operation of the integrator time step control and predictor The integrator is used for all charge based elements which are not using convolution By default the settings are trapezoidal for the integrator and third order Gear for the predictor In general RF amp microwave circuits will simulate with no problems using the default settings If you have problems with your simulation the integrator should be the last thing you change after you adjust the time step and convolution settings If you do change the integrator you should ensure that the predictor is set to at least one order higher than the integrator If you set the predictor to Auto it will automatically be set to Gear and one order higher than the integrator Generally using a higher order integrator will allow CAYENNE to take larger timesteps and will speed simulation However higher order integrators are not as stable and can cause false oscillations generally growing exponentially with time If you are going to ma
556. t node 2 will follows its normal course including through frequency translation devices NOTE When using multiple frequencies node names are a better path specification since part generally have more than 1 node and the system analysis doesn t know which terminal of the part to use OUTPUT This section will determine which measurements are added to the dataset and some of the parameters needed by some of the measurements Intermods Along a Path When checked with calculates intermod measurements and add them to the path dataset This is not to be confused with intermods created during a simulation Those will always be created as long as calculation of intermods has been selected on the Calculate Tab of the system analysis Tone Interferer Frequency This is the frequency where the tone or interferer is located that will be used to determine the intercept point In order to determine the intercept point the system analysis must measure the intermod power and the power of the tone or interferer that created the intermod The path channel frequency is set to the frequency of the intermod and an additional tone interferer channel must be created at the frequency of the tone or interferer The intercept measurement technique used in the system analysis is the generally the same used in the laboratory using signal generators and spectrum analyzers See Intermod and Harmonic Basics Intermod Path measurement Basics Cascaded Intermod Equatio
557. t node and adding this value to the cascaded gain In this case cascaded gain source mismatch will equal S21 Cascaded Noise Figure CNF Spectrasys System This measurement is the cascaded noise figure in the main channel along the specified path The Cascaded Noise Figure is equal to the Channel Noise Power measurement at the output of stage n minus the Channel Noise Power measurement at the path input and the Cascaded Gain measurement at stage n as shown by CNF n CNP n CNP 0 CGAIN n dB where n stage number Caution When wide channel bandwidths are used channel noise power and cascaded gain are affected more by VSWR and frequency effects In this case is it extremely important that sufficient noise points are used to represent the noise in the channel of interest Furthermore it is very possible because of these frequency effects that the channel noise power and the cascaded gain can change in a nonlinear way so that cascaded noise figure appears to drop from a prior node Additionally looking at cascaded noise figure through a hybrid combining network may also be deceptive since the cascaded gain used to determine the cascaded noise figure is from the current path and not all paths in the system See the Broadband Noise section for more information Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as CNP and CGAIN Travel Direction Same as CNP and CGAIN
558. t of Frequencies MHz 290 Linear Analysis e N Simulation Linear ore TT Number of points in entire simulation Equations may Number of A be used here no needed Points Log Number of points in every decade Equations may be Points Decade used here no needed Type of Sweep Allows specification of start and stop frequencies and Linear Step i space between points Equations may be used here no Size MHz E needed Allows the explicit specification of analysis frequencies List of These points are entered into the List of Frequencies Frequencies MHz box separated by spaces Equations may be used here no needed S Parameters 292 Overview The purpose of this section is to summarize network analysis concepts and to define some of the parameters plotted by GENESYS For further details on measurements see the Measurements section of this manual Networks are considered as black boxes Because the networks are assumed to be linear and time invariant the characteristics of the networks are uniquely defined by a set of linear equations relating port voltages and currents A number of network parameter types have been developed for this purpose including H Y Z S ABCD and others These parameters may be used to compute and display network responses and to compute quantities useful for circuit design such as Gmax maximum gain and gain circles Each parameter type has ad
559. t or voltage values at each frequency EMPOWER creates an EMV file whenever Generate Viewer Data is checked or the In option is specified EMV files can only be read by the EMPOWER viewer If you want to generate viewer data for import into other programs you should generate a PLX text file For more information on viewer files the Viewer section L1 L2 Ln Line Data Files Written by EMPOWER Type Binaty Can be safely edited No Average size 1 to 5Kbytes but may be larger Use Internal file for EMPOWER but can also be used in the SMTLP and MMTLP models in GENESYS EMPOWER must perform a separate line analysis for all external ports If no filename is specified by the user then the results from the line analysis are stored in Ln files These files also store all information about the box and port and are intelligent They are only recalculated if necessary and even then only at frequencies necessary Even if the circuit changes they are only recalculated if the change affects the line analysis Notes When these files are numbered modally related groups of ports are counted as one Also if two ports are identical then only the first one will create a Ln file LST Listing Files Written by EMPOWER Type Text Can be safely edited Yes Average size 50K to 200K but may be larger Use Gives all calculated data and grid mapping from EMPOWER in human readable form This file is overwritten whenever EMPO
560. t the Channel Bandwidth to a value wider than the intermod but narrow enough to exclude any power from the tone 6 Add the RX_IIPn or RX_OIPn measurement to a graph or table when n is the intercept order Troubleshooting Intermod Path Measurements Here are a couple of key points to remember when troubleshooting and intermod measurement problems Using a table is generally much better at troubleshooting than level diagrams Look at the Channel Frequency CF measurement in a table This must be the frequency of the intermods of interest Make sure there are intermods within the channel by looking at the Total Intermod Channel Power TIMCD If there are no intermods in the channel look at the spectrum and verify that an intermod of the order of interest has been created at the Channel Frequency of the path If there are no intermods at the channel frequency make sure Calculate Intermods has been enabled If there are still no intermods make sure the nonlinear models have their nonlinear parameters set correctly If the Total Intermod Channel Power doesn t seem to be too low verify that the Channel Measurement Bandwidth is wide enough to include the intermod order of interest If the Total Intermod Channel Power still doesn t seem to be correct then verify that the Channel Power CP measurement is showing the approximate expected power The Channel Measurement Bandwidth may be set so wide that other interferer fre
561. t to the view destination Before any of the phase noise path measurements are made the phase noise is scaled to the appropriate bandwidth before spectrum identification begins Spectrum Analyzer Display The spectrum analyzer mode is a display tool to help the user visualize what the simulation would like on a spectrum analyzer This mode is extremely useful when out of phase signals may cause the total spectrum to cancel NOTE This mode is NOT used for any path measurement data and is for display purposes only The parameters used in this mode have NO bearing on the accuracy of the real simulation results The spectrum analyzer mode performs a convolution of a 5 pole gaussian filter on the total spectrum trace A gaussian filter is used because this is the type of filter used in real spectrum analyzers Enable Analyzer Mode Resolution Bandwidth REM MHz Limit Frequencies Defaults to channel bandwidth Start 100 MHz Filter Shape Stop 2000 MHz Randomize Moise 4dd Analyzer Noise ep The challenges associated with convolving the spectrum is as follows e Because of broadband noise frequencies can literally go from DC to daylight Spectrasys System e Frequencies are not evenly spaced e Spectrums generally appear in groups with larges spacing between groups Because of practical issues associated with performing a convolution several parameters have been added to this mode to decrease the
562. tage in the main channel along the specified path Phase noise is displayed on the graphs in V sqrt Hz Channel Used Main Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used ONLY PHASE NOISE Travel Direction Only spectrums traveling in the FORWARD path direction Stage Equivalent Input Noise Voltage SVNI This measurement is the stage equivalent input noise voltage entered by the user Stage noise figure source resistance and temperature can be converted to stage equivalent input noise voltage using the following equations Nomeless LMA Equivalent Input Moise Rs Voltage sart Hz Noiseless Source Noise Voltage V sqrt Hz ens Rs sqrt 4k T Rs for example ens 50 0 895 nV Stage Equivalent Input Noise Voltage Vni sqrt F 1 ens Stage Noise Factor F Vni ens 2 1 Stage Noise Figure dB NF 10 Log F where Rs is the input source resistance to the stage Spectrasys System Default Unit nV sqrt Hz Channel Used No channel is used for this measurement Types of Spectrums Used None Travel Direction N A Stage Input 1 dB Compression Voltage SIVIDB This measurement is the stage 1 dB compression voltage point calculated by using the Stage Output 1 dB Compression Voltage Point and the Stage Voltage Gain When a stage doesn t have this parameter 100 kV 1s used SIVIDB n SOVIDB n SVGAIN n dBV where n stage number Channel Used No channe
563. td which allows specification of additional model options EAGLEWARE NAME keyword Normally the GENESYS model name is the same as the Verilog A model with any device class options added to this base name The base name can be overridden by VY oEEAGLEWARE NAME wodelnameYoVo EAGLEWARE SWAP12 keyword Advanced keyword which reverses pins 1 and 2 in GENESYS Can override DEVICE CLASS swapping if placed later in the file EAGLEWARE_NOSWAPI12 keyword Advanced Modeling Kit Advanced keyword which prevents pins 1 and 2 from being swapped in GENESYS Can override DEVICE CLASS swapping if placed later in the file EAGLEWARE IGNORE keyword Tells GENESYS to ignore a parameter For example YYEAGLEWARE_IGNORE x will cause the parameter x to not be displayed in GENESYS and the default value will be used in Verilog A 63 Chapter 3 CAYENNE Transient Analysis CAYENNE Overview CAYENNE like SPICE simulates the time response of a circuit to an arbitrary input waveform In order to accomplish this a time stepped DC analysis is performed to compute samples of the output waveform All elements whether energy storing linear or non linear are broken down into a form that allows Nodal Analysis to be performed Nodal analysis based on Kirchoff s laws describes the currents in a circuit as an equation involving the conductances and voltages in the circuit The number of nodes in a circuit determines the number of variables
564. ted from the S parameter solution This process of adding or subtracting line length is generally referred to as de embedding This 1s the basic process During the solution process the impedance and propagation constant has been calculated for the ports based on their physical location in the circuit When you know the impedance propagation constant and the distance of de embedding you can cancel out the extra lengths of line from the S parameter results by compensating for the loss and phase shifts of those lines The net result is a set of S parameters calculated as if the extra line lengths were not there De embedding Considerations It is possible to de embed right up to the discontinuity itself However make sure that you do not shift the reference offset beyond the first discontinuity This would yield incorrect simulation results as there is another linewidth beyond that discontinuity which means that there is another set of impedance and propagation values that applies there You can de embed away from the circuit by placing reference offsets Gy beyond the edges of the layout This enables you to simulate the effect of a long feed line that was not drawn in the simulated structure Allowing for Coupling Effects If you have two or more single ports that lie on the same reference plane the calibration process will take into account the coupling caused by parasitics that naturally occurs Momentum GX between these
565. tered around the box resonance frequency which cannot be represented by a smooth function no smooth adaptive S parameters will be available in this frequency band When simulating such a circuit Momentum will inform you of 345 Simulation Mesh 346 these box resonance frequencies of the frequency bands where there will not be smooth S parameters available and of the quality factor a measure for the sharpness of the box resonance depending on the losses e Propagating modes may be present You may have a situation where there are no metal sidewalls in the structure but the substrate definition is bounded by other material for example you have a finite size substrate where the dielectric material abruptly ends so you have a dielectric air transition Although Momentum boxes are defined as perfect metals and not dielectric material you may decide that defining a metallic enclosure in the simulated circuit may be more representative of the real structure than using no enclosure Adding Absorbing Layers under a Cover You may want to model your box as having absorbing layers between the covers and the layout Use the air above and air below layers to define absorbing layers for the box If the top cover is open then the height of the air layer 1s ignored but all other parameters are still used for free space absorption Similarly for the bottom cover Boxes and Radiation Patterns If you have a structure enclosed in a box and y
566. ternal file from the EMPOWER run WSP Simulations EM1 EMPOWER SS An input and output were added on nodes one and two of the FOU block the ground was added to the ground node and a capacitor was connected across ports 3 and 4 This has the effect of putting the capacitor into the EMPOWER simulation This capacitor can then be tuned and optimized just like any other element in GENESYS When the S Parameters of MYNET are displayed you see the resulting S Parameters of the entire circuit Automatic Port Placement One advantage of EMPOWER is its true integration In most electromagnetic simulators you would have no choice but to go through the complicated steps above Imagine how tedious this would be if you had 10 lumped elements 2 transistors and an op amp chip in EMPOWER Planar 3D EM Analysis the box Fortunately the internal ports and lumped elements can be generated and added automatically The circuit below uses automatic port placement Initially the circuit on the left is drawn The layout on the right of the figure was then created The footprint for the chip capacitor was automatically placed The lines and EMPorts were then manually added When EMPOWER is invoked internal ports are automatically added so the circuit simulated is virtually identical to the one on the left below and the result is a 4 port data file Hy tity Te HH MYNET 2 EMPOWER then automatically creates a network whi
567. th S parameters can be calculated During a simulation all current directions on the sidewalls of the box are taken into account Momentum GX Calibrated ports in the circuit 1 e Single Differential Coplanar and T Common mode ports must be located on and perpendicular to the box edge or an error will occur Internal uncalibrated ports may be placed anywhere in the circuit and point in any direction GENESYS Layout always creates the sidewalls of a box using specifications in the Layout properties dialog LAYOUT Properties General Associations Layer Fonts m MEN m Box Settings Uses UNITS as set in box to left Designs to include Design Units Grid Spacing 0 075 Show Grid Dots SPIRAL Grid Spacing Y 0 075 Layout IL Object Dimensions DUE Show EM Box Show EMPOWER Grid Box Width X Cell Box Height 3 7 5 Cells Drill Diameter 1 Drawing Options Show Momentum Mesh Line Width Pad Width Multi Place Parts until Esc Port Size 0 254 Addn Rotation Snap Angle Solid Opaque Default viahale Layers 9 X Ray mode Top Layer BE Top METAL v Hollow Bottom Layer I Bottom Cover v sv ee _ Drawing Style The top and bottom characteristics of the box ate set in the Layer tab of the Layout Properties dialog The Momentum GX simulator can ignore the box sidewalls or use them Select box usage on the Simulation Op
568. the unilateral transducer gain S12 is set to zero Note See the section on S Parameters for a detailed discussion of Gain Circles Values Complex values versus frequency Simulations Linear Default Format Table center magnitude angle radius Linear Graph none Smith Chart Circles Commonly Used Operators None Examples Getting Started Measurement Result in graph Smith chart Result on table optimization or yield GU1 unilateral gain circle at port 1 center magnitude angle radius Linear GU2 unilateral gain circle at port 2 center magnitude angle radius Linear Available on Smith Chart and Table only Stability Factor K Stability Measure B1 The Stability Factor and Measure parameters are real functions of frequency and are available for 2 port networks only These parameters aid in determining the stability of the 2 port network If 512 of a device is not zero a signal path will exist from the output to the input This feedback path creates an opportunity for oscillation The stability factor K is ke t jsn pa 2 9 7 2 r2 San where D S11822 S12821 From a practical standpoint when K gt 1 S11 lt 1 and S22 lt 1 the two port is unconditionally stable These are often stated as sufficient to insure stability Theoretically K gt 1 by itself is insufficient to insure stability and an additional condition should be satisfied One such parameter is the stability measure B1 which should be
569. the GENESYS built in nonlinear models and then customize these as you see fit Note You cannot change the built in models Instead you must create a new model and must use this new model in your schematic The source code is in the Examples VerilogA directory normally installed to C Program Files AGENESYS 2005 11 Examples VerilogA To use these files you should copy them to a new directory as described in step 1 of Creating New Verilog A Models above The example modules have va added to the end of the name to keep them from conflicting with the built in models After you have copied the Verilog A source file you should follow the steps in Creating New Verilog A Models above Verilog A Tutorial Verilog A is a procedural language with constructs similar to C and other languages While the language does allow some knowledge of the simulator most model descriptions should not need to know anything about the type of analysis being run Perhaps the simplest possible Verilog A file is a resistor the line numbers are not part of the verilog file include disciplines vams module resistor p n inout p n electrical p n parameter real r 50 from O inf exclude 7 analog begin V p n lt r I p n 10 end 11 endmodule You can use this resistor as a starting point for your own Verilog A files or you may start with a more complex file such as the built in nonlinear models d Line 1 include disciplines va
570. the Workspace Window and select Run EMPOWER Viewer A top down view has been selected and the notch frequency has been specified Port 1 is at the left of the image and port 2 is at the right The plot is color coded to the scale given in the lower left of the figure Notice that port 2 is nearly black This indicates that very little energy is being delivered to that port at 9 2 GHz as we d expect Creating a Layout From an Existing Schematic The file used in this example is FiltersV Tuned Bandpass wsp This example demonstrates the following topics e Creating a layout from an existing schematic e Tuning with EMPOWER data e Using lumped elements with EMPOWER In GENESYS open the example Filters Tuned Bandpass wsx Double Click the F2000 design in the Workspace Window to display the schematic for this filter shown below EMPOWER Planar 3D EM Analysis C2000 pF ji E C2000 pF This is the schematic of a 2nd order microstrip combline bandpass filter with 50 Ohm terminations and transformer coupling on the input and output The lumped capacitors are gang tuned to adjust the resonant frequency of the two center lines Tuning in this manner affects only the center frequency and keeps the passband bandwidth constant Double Click Layout1 under Designs in the Workspace Window to display the layout for this schematic shown below 123 Simulation 124 L z a E a E ui x A 0402 Ch
571. the path dataset Frequency Offset from Channel The is the frequency spacing between the main channel center frequency and the center frequency of the offset channel Bandwidth This is the bandwidth of the offset channel power measurement System Simulation Parameters Calculate Tab This page controls calculation of Intermods Harmonics Noise and Phase Noise Tabs General Paths Calculate Composite Spectrum Options Output 945 Simulation System Simulation Parameters General Paths Calculate Composite Spectrum Options Output Harmonics and Intermads Maximum Order Calculate Intermads From Sources Only Odd Orders Only Coherent Addition Fast Intermod Shape These settings affect the amount of data generated by Spectrasys as Well as simulation speed Calculate Moise System Temperature 27 0 Pus Thermal Moise 173 83 dBm Hz Moise Points For Entire Bandwidth 11 Add extra noise points 10 MHz bandwidth at each signal Frequency Defaults to channel bandwidth Calculate Phase Noise Parameter Information HARMONICS AND INTERMODS This section controls calculation parameters for harmonics and intermods See the Calculate Intermods Harmonics section for more information on intermods and harmonics Calculate Harmonics When checked the system analysis will calculate and save harmonic data to the system analysis dataset Calculation time for harmonics is typically ver
572. the path such as impedances etc VSWR Related Issues Occasionally gains and noise figures may be unexpected This is generally related to VSWR issues across the channel SPREADSHEETS generally ignore all VSWR effects and assume power is always delivered to a perfectly matched load Cascaded equations 532 Spectrasys System make the same erroneous assumptions The best way to determine VSWR issues is as follows 1 Compare the GAIN calculated gain and SGAIN user specified gain measurements in a table If these measurements are not in close agreement then there is a VSWR issue at the stages with the largest discrepancies All filter blocks should be replaced with attenuators to coincide with spreadsheet measurements Long Simulation Time and Convergence Issues The number of passes can be observed on the simulation status widow to determine convergence issues A high pass count indicates continued spectrum propagation To troubleshoot a convergence or long simulation time problem follow the given steps Ie 2 3 4 5 Increase isolation for components having more than 2 ports Isolation can be increased to eliminate any loop Increase reverse isolation of active devices such as amplifiers Reduce the maximum simulation order Reduce number of carriers Disable noise calculations Once the problem has been isolated the root issue can be determined Reduce the Number of Unknowns The more unknowns there
573. the sum of node currents must sum to zero to achieve harmonic balance convergence The simulator is converged if the magnitude of the vector sum of the currents entering all node at all frequencies is less than the specified absolute tolerance Absolute Voltage Tolerance absolute tolerance of solution changes between 2 consequent Newton iterations Relative Tolerance The relative accuracy to which the sum of node currents must sum to zero to achieve harmonic balance convergence The simulator is converged if for all HARBEC Harmonic Balance Analysis frequencies and all nodes the ratio of the vector sum of the currents into a given node currents to the sum of magnitudes of the current entering that node is less than the specified relative tolerance Min Rel RHS Norm Change minimal change of residual norm for a Newton iteration when current solution method is accepted as successful In case when the relative error change is less then the value Harbec will switch the current solution method to another one Default 0 any decreasing of the residual norm 1s accepted as successful Oversampling Factor Sets a factor for additional time points to be calculated during nonlinear device simulation which can improve convergence but will take additional time The factor should be set gt 1 Default value is 2 Allow 1 D FFT The simulator will normally convert frequency spectrums to time waveforms and back usin
574. thermore spectral amplitudes can have large dynamic ranges These dynamic ranges can be restricted to ranges of interest As a general rule the more data collected the longer the simulation time is Spectrasys supports several different spectrum types The user can select which types of spectrum to simulate Controlling Frequency Ranges There are 2 parameters that control the simulation frequency range ALL frequencies outside this window will be ignored by the simulator under any condition Thermal noise is automatically generated in this frequency range as long as Calculate Noise has been enabled e Ignore Frequency Below e Ignore Frequency Above Controlling Spectrum Amplitude Ranges There is 1 parameter used to control the simulation amplitude range ALL spectrums whose power levels fall below the given amplitude value will be ignored e Ignore Amplitude Below Controlling Spectrum Types The following spectrum types can be ignored during simulation e Intermods e Harmonics 439 Simulation e Thermal Noise e Phase Noise Controlling Path Data Every measurement is dependent on one or more types of spectrum All measurements whose spectrum types have been enabled are added to the path dataset Furthermore path powers voltages and impedances can also be added to the dataset Sweeps of a Path Sweeps of path measurements can be plotted in one of two ways e Ona Level Diagram e Ona rectangular graph showing a S
575. this Manual Values Complex matrix versus frequency Simulations Linear EMPOWER Default Format Table dB angle Graph dB Smith Chart dB angle Commonly Used Operators Operator Description Result Type ang S 1 1 Angle in range 180 to 180 degrees Real ed S 2 2 Group Delay Real ql S 2 1 Loaded Q Real Other Operators db mag re im Examples Measurement Result in graph Smith chart Result on table optimization or yield S 2 2 dB Magnitude of 22 dB Magnitude plus angle of S22 ql S 2 1 Loaded Q of S21 Loaded Q of S21 mag S 2 1 Lineat Magnitude of S21 Lineat Magnitude of S21 S Shows dB Magnitude plus angle of all S Parameters gd S 2 1 Group delay of S21 Group delay of S21 Note For port numbers greater than 9 the array must be indexed For example use 5 12 33 fot S 12 33 H Parameters This H parameter or hybrid parameter measurements are complex functions of frequency The frequency range and intervals are as specified in the Linear Simulation dialog box The H parameters are only defined for a two port network and are of the form Hi for j equal 1 2 Getting Started The equations relating the input voltage V1 and current I1 to the output voltage V and current I2 ate Vi Hii h H12 V2 Ip Ha h Hz V2 Values Complex matrix versus frequency Simulations Linear Default Format Table RECT Graph RE Smith Chart none Commonly Used Operators Operator Descrip
576. ting from the left with order 0 See the Intermods Along a Path section for information on how to configure these tests Remember intermod bandwidth is a function of the governing intermod equation For example if the intermod equation is 2F1 F2 then the intermod bandwidth would be 2BW1 BW2 Note Bandwidths never subtract and will always add The channel bandwidth must be set wide enough to include the entire bandwidth of the intermod to achieve the expected results The Automatic Intermod Mode will set the bandwidth appropriately Caution This method used to determine the intercept point is only valid for 2 tones with equal amplitude Channel Used Main Channel Frequency Interferer Channel Frequency and Channel Measurement Bandwidth Types of Spectrums Used Same as OIP and CGAIN Travel Direction Same as OIP and CGAIN Input Intercept Receiver All Orders RX IIP This measurement is the receiver input intercept point along the path This is an out of band type of intermod measurement RX_IIP n RX_OIP n CGAIN n dBm where n stage number 509 Simulation 510 This measurement simple takes the computed Output Intercept and references it to the input by subtracting the cascaded gain The last IIP value for a cascaded chain will always be the actual input intercept for the entire chain Each column in this measurement is for a different intermod order up to the Maximum Order specified on the Cal
577. tion Result Type rect H11 real imaginary parts Real re H22 real part Real magang H21 Lineat magnitude and angle in range of 180 to 180 Real Examples Measurement Result in graph Smith chart Result on table optimization or yield H22 RE H22 real part of H22 rect H Shows teal imaginary parts of all H Parameters mag H21 Lineat Magnitude of H21 Linear Magnitude of H21 H Shows teal imaginary parts of all H Parameters Not available on Smith Chart Y Parameters This Y parameter ot admittance parameter measurements are complex functions of frequency The frequency range and intervals are as specified in the Linear Simulation dialog box The Y parameters for an n port network are of the form YP fori j equal 1 2 n For a two port network the equations relating the input voltage V1 and current I1 to the output voltage V2 and current 12 are li YPi1 Vi YPi2 V2 In YP Vi YP22 Vo Values Complex matrix versus frequency 17 Simulation 18 Simulations Linear Default Format Table RECT Graph RE Smith Chart none Commonly Used Operators Operator Description Result Type re Y P22 real part Real magang YP21 Lineat magnitude and angle in range of 180 to 180 Real Examples Measurement Result in graph Smith chart Result on table optimization or yield YP22 re Y P22 real part of YP22 mag YP21 Linear Magnitude of YP21 Linear Magnitude of YP21 YP Shows real imaginary p
578. tions can either be strong or weak depending on their relative position and their length scale The matrix filling process is essentially a process of order N2 i e the computation time goes up with the square of the number of unknowns In the solving step the interaction matrix equation is solved for the unknown current expansion coefficients The solution yields the amplitudes of the rooftop basis functions which span the surface current in the planar circuit Once the currents are known the field problem is solved because all physical quantities can be expressed in terms of the currents For large problem sizes the iterative matrix solve performs as an order N2 process Momentum still uses a direct matrix solve process if the structure is small or when convergence problems are detected in the iterative matrix solver process Calibration and De embedding of the S parameters Momentum performs a calibration process on the single type port the same as any accurate measurement system to eliminate the effect of the sources connected to the transmission line ports in the S parameter results Feedlines of finite length typically half a wavelength at high frequencies short lines are used at low frequencies are added to the transmission line ports of the circuit Lumped sources are connected to the far end of the feedlines These sources excite the eigenmodi of the transmission lines without interfering with the circuit The effect of the feedli
579. tions page of the dialog 343 Simulation Momentum Options General Simulation Options Mash Simulation Mode O RF Faster no radiation effects G Microwave recalculates substrate for every Frequency Calculate CO al substrate mesh and S Parameters 3D Metal Expansion Hone thin metal Op O Down Via Model OLumped COD wire 8 2D planar no horizontal currents def sult CO 3D spatial include horizontal currents LI cal Use horizontal side currents thick conductors e Reuse results of ask simulation Lo JD ee Ji Viewing Layout Layer Settings of a Box To view the box layer specifications 1 Choose Layout Properties and go to the Layers tab 2 The box top cover and bottom cover are defined there 344 Momentum GX LAYOUT Properties LW General Associations Layer Fonts Show Columns Metal Substrate General Laver Number and Color C Show all EM EMPOWER Momentum Momentum Slot Type ame Type Height Er Tand Ur Surface Imp Current Thick Metal Element Sigma Value or File Direction Slow Ports Top Cover Lossless Air Shove Air 6 55 1 1 TOP METAL Sub Default Mortal Thin Down SUBSTRATE Sub Default BOT METAL Air Below 50 1 0 1 Bottom Cover Electrical 50 About Boxes There are a variety of reasons why you would want to simulate a circuit in a box e The actual circuit is enclosed in a metal box Circuits are often encased
580. to wait the additional time while doing most of your analyses There is an additional caveat regarding loss described in the section on Slot type structure See the Narrowband Interdgital example for an example of the effect of loss on an interdigital filter 131 Simulation 132 Viewer Data Looking at currents in the viewer is a great way to get insights into circuit performance However generating this viewer data requires additional time increasing the length of a run by a factor from two to ten and sometimes requiring additional memory also Generating viewer data has no effect whatsoever on the solution given so you should not have this option turned on unless you actually intend to run the viewer You can turn this option on and off by using the checkbox labeled Generate Viewer Data Slower when starting an EMPOWER run You will not normally need viewer data and when it is needed you will not normally need viewer data at every frequency Our recommendation 1 Run all problems the first time without generating viewer data If the answer is completely unexpected check for errors in your description of the file This can save a lot of time in the experimenting stage 2 If you decide you want viewer data open the EMPOWER Options dialog box Reduce the number of frequency points to be analyzed and turn on Generate Viewer Data Slower Recalculate the EMPOWER simulation and you will now have viewer data at some po
581. tours Example wsx located in the Amplifiers subdirectory of your Examples directory GENESYS can import Focus and Maury Microwave format load pull data files Using this data contours may be generated on a Smith Chart using the contour function which utilizes a thin plate spline of the data By selecting the Fz menu and choosing the Import submenu followed by Load Pull Data File one may import Focus format data files by selecting Focus Format or Maury Microwave format Maury Format and selecting the appropriate file from the dialog box This will produce a data set in the root workspace folder with the name set to the name of the file without the file extension In this data set three variables will be present DATA GAMMA and MeasNames DATA contains the numeric data of the measurements specified in the load pull data file GAMMA contains the gamma values at which each measurement is taken and is the independent variable of DATA MeasNames is a string array containing the name of the measurement of each column of DATA as specified in the load pull data file The Load Pull Contours example Load Pull Contours Example wsx located in the Amplifiers subdirectory in Genesys Examples loads a Focus Microwave data file The file contains several columns of amplifier measured data In the plot the Gain column from the data file is used to plot load pull contours from the two equation statements shown below contours contout GAIN
582. ts by exp jw where w cycles from 0 to 2pi and plotting snapshot graphs for sequential time moments What is animated is controlled by the Display Option Button see E below E Display Option Button This button selects the current display option Real Displays the real portion of the current values Mag Displays the magnitude or time averaged values of the currents Ang Displays the phase delay of the current values F Solid Wire Button This button toggles the type of surface plot to display Wire Displays a wireframe version of the current patterns A wireframe is created by drawing the outlines of the EMPOWER grid currents without filling the resulting polygons Solid Displays a solid surface plot of the current patterns This 1s created by filling the wireframe polygons G Freq GHz This box shows the simulation frequency in GHz for which the current image data is being displayed This box is restricted to frequencies that EMPOWER has created data for The value can be increased by clicking the button see I below and decreased by clicking the button see H below H Decrease Frequency Button EMPOWER Planar 3D EM Analysis Decreases the current frequency see G above If you are already at the lowest calculated frequency then this button has no effect I Increase Frequency Button Increases the current frequency see G above If you are already at the highest calculated frequency
583. tt 1e 06 5000 1e 04 DB S21 DB S21 DB S11 NN 3000 7000 10000 1e 06 3000 7000 10000 0 40224 8 97009 3 40433 1 33748 14 4258 0 746666 3 64981 10 1519 0 40224 8 97009 3 40433 1 33748 0 0 0 0 Tune 5 E Error 0 Losses NOTE The Single and Multi Mode transmission lines required to use decomposition are not available in GENESYS A current limitation of decomposition is that losses are not taken into account in multi mode transmission line sections or in reference plane shifts For the spiral inductor this means that the losses as calculated are accurate for the nominal dimensions but any modification to the lengths using the multimode lines will not affect the calculated loss In general if the decomposed pieces cover the circuit completely as is the case in the spiral inductor then the losses will be accurate If the pieces do not completely cover the circuit if sections of line are left out of the EMPOWER analysis and are added with MMTLP sections then the losses will not include these sections This is true regardless of the reference plane shifts used since these shifts do not affect the loss Port Numbering NOTE The Single and Multi Mode transmission lines required to use decomposition are not available in GENESYS You must be very careful when setting up and numbering ports for decompositional analysis The following rules must be followed 151 Simulation e Never connect anything othe
584. ttom walls of the box can be ideal electric amp magnetic walls or walls with surface impedance The structure can also be terminated by semi infinite rectangular waveguides in the planes of the box top and bottom walls A clarification of the boundary conditions for the media layer interfaces A 2 are given in the following table A 2 Region without metalization Lossless metalization Surface Impedance Port Region along X Axis or Internal Port Lumped Element Region along X Axis the same for y axis C is region cross section l is region length Internal Port along Z axis Input ports in the structure are modeled by line segments approaching the outer boundaries line conductors and surface current sources in the regions where line conductors approach the walls of the volume It is assumed that the currents inside the input and the lumped element regions are constant in the direction of current flow and the 209 Simulation 210 corresponding electric field component along the region is constant across it Thus the integral of current across the region gives an integral current and integral of the electric field along the region gives an integral voltage for the region The desired solution of the electromagnetic problem is an immitance matrix relating the integral voltages and currents in the port and lumped element regions This is actually a kind of Green s function contraction on the port and lumped element r
585. tworks only The noise figure is defined as the ratio of input signal to noise power ratio SNR to the output signal to noise ratio SNRour NF SNRw SNRour The noise figure is related to the minimum noise figure NFMIN by the expression NF NFMIN Rn Gs Ys Yorr where Ys Gs j Bs Source Admittance Rn Normalized Noise Resistance The minimum noise figure represents the noise figure with ideal match of source impedance i e Ys Yopr Values Real value versus frequency Simulations Linear Default Format Table dB Graph dB Smith Chart none Constant Noise Circles NCI A noise circle is a locus of load impedances for a given noise figure as a function of frequency This locus which is specified via a marker is plotted on a Smith chart with noise figure degradations of 0 25 0 5 1 0 1 5 2 0 2 5 3 0 and 6 0 dB from the optimal noise figure Note See the section on S Parameters for a detailed discussion of noise circles Values Complex values versus frequency Simulations Linear 21 Simulation 22 Default Format Table center MAGI ANGJ radius Linear Graph none Smith Chart Circles 6 Commonly Used Operators None Examples Measurement Result in graph Smith chart Result on table optimization or yield NCI noise circle locus of load impedances for optimal noise figure for each circle center MAG ANGI radius Linear Available on Smith Chart and T
586. u should follow the first example To briefly review An external port is placed in LAYOUT by selecting EMPort from the toolbar These ports are generally placed on the edge of the box at the end of a line This figure shows a comparison between a pott in circuit theory and a port in EMPOWER In the circuit theory schematic on the left there are two ports Each port has two terminals with the bottom terminal generally being ground In the EMPOWER illustration shown on the right the figure the section of line stops before the edge of the box generally one cell width away and a port begins in its place See the Grid discussion in the Basics section to see how this is mapped onto the grid As in the circuit theory schematic there are two ports and each port has two terminals However in EMPOWER instead of the ground plane being modeled as a simple short circuit the effect of currents traveling through the box 1s taken into account 135 Simulation 136 EM Port O ptions When you first create a port it is automatically configured to be an external port with the proper characteristics to be placed on the end of a transmission line For many applications you will want to modify these characteristics when you place the port These characteristics ate shown in the EM Port Properties dialog box which comes up automatically when the port is placed and which can be accessed later by either double clicking on the port or by selecting the
587. ucing matrix Status Only lt lt i 5 Parameters Parameters Response Data I Pi F 8700 0 Eeff 1 89143528 5 8283 amp e 3 20 48 4692904 0 0 le PZ F 8700 0 Eeff 1 89143528 5 8283 amp 3 Z0 48 4692904 0 0 Frequency 68600 HHZ De embedded 50 Ohm S Params Mag Ang Touchstone Formar 311 321 Erraors Warnings i 4 800 00000 0 942510 41 783 0 225888 1313 31 0 258958 133 31 0 942 l Pl F 8e00 0 Eetr 1 89168017 5 80552e 31 j z 45 456758521 0 07 l Pz F 6800 0 Eetf 1 89165013 5 8 055e 3 j Z20 145 4675531 0 07 Timing Info Frequency 8900 MHZ De embedded 50 Ohm 5 Params Bag Ang Touchstone Format 311 3231 900 00000 0 956453 376 964 0 246063 125 584 0 246063 125 54 0 595t Pl F BSDD D Eeff 1 9919z2983 5 783Zze 343 Z 453 465659336 0 0 Batch List Pe FzsBHSODU 0 Eefi 2 1 99192983 5 78903Z286 343 AUS 40 4659436 U Uz While Sonnet is running GENESYS will also display the simulation status window indicating the Sonnet is running When Sonnet is finished calculating GENESYS will automatically load the simulation results and update the screen Viewing Results After Sonnet simulation of the layout the data must be displayed in GENESYS This is done by creating a Data Output such as a Rectangular Graph To create a rectangular graph in this workspace 1 Click on the Create New icon in the Workspace Window and select Output Add Rectangular Graph from the menu Accept the d
588. ucture is not symmetrical The coordinates specified refer to the terminal map shown above LINE ANALYSIS MODE RESULTS This area of the listing contains sections identical to those described above which pertain to the line analysis Below these sections you will find a table of line parameters for each frequency The entries are Nm port number Type impedance type real re or imaginary im Normal lines should have a real impedance Zo ohm Line impedance Gw rad m propagation constant Gw Go propagation constant relative to free space Comp Phase Compensation Admittance value of phase and impedance compensation for deembedding S MATRIX TABLES Each table gives the circuit s s parameters at one frequency For normal non multimode inputs as an example S21 is found in the row with input numbers 2 and 1 in that order PLX Current Viewer D ata Files Written by EMPOWER Type Text EMPOWER Planar 3D EM Analysis Can be safely edited Yes Average size 200 Kbytes to 2Mbytes but may vary Use Importing current data from EMPOWER into another application such as Matlab ot Excel This file contains two tables per frequency one each for x and y directed currents Each table contains 4 columns containing the x and y coordinates followed by the real and imaginary part for each current These tables could be edited but it would be best to leave them alone since they would be very tedious and error pro
589. ude ranges of interest can speed up the calculation process drastically especially when calculating intermods However take caution when setting these limits so that intentional spectrums are not ignored Frequency Below default 0 Hz All spectral components whose frequency is below this threshold will be ignored Spectrums falling below this limit will not continue to propagate However there are several cases where negative frequencies may be calculated at interim steps 1 e through a mixer which will be folded back onto the positive frequency axis This parameter will only affect the final folded frequencies and not the interim frequency steps Likewise this is the lower noise frequency limit Frequency Above default Max Order 1 times the highest source frequency All spectral components whose frequency is above this threshold will be ignored and will not be created Spectrums falling above this limit will not continue to propagate Likewise this is the upper noise frequency limit MAXIMUM NUMBER OF SPECTRUMS TO GENERATE This group is used to limit or restrict the maximum number of spectrums that will be created Max Spectrums Limits the maximum number of spectrums that are created Once this limit is reached during a simulation no additional spectrums will be created Nov This option must be used with care since a premature limitation of the number of total spectrums will more than likely affect the accuracy of all th
590. uding the output port Noise figure SSB calculated from the formulas 2 uses large signal conversion gain while formula 1 uses small signal conversion gains calculated at the steady state regime of the circuit The results calculated using formula 1 and 2 will be close each other for small nonlinearity of the circuit relatively to the input signal Harbec Equations block including calculation of Nzn HARBEC Harmonic Balance Analysis Equation MUI ee Dew F 250e6 variable ue2tTPITF i dicim29 18 12 uzzu w I dcza11 1e 12 Ci 29 103e 12 d Fe z50e 5 Lisi wz cl IL 113 92 3 C2711 141 amp 12 Tog 2m36 4e 3 L2 1 wa CZ T oisedila ris t Id PLOS 1 PRF 7 2D dEm iS IPNIne 173 83 PLO 210 dBm ifpeF 2n NoiseBand 1 1Hr fea Sinn 15 3 853 Tle S7e 5 Input signal to noise racio calculation I uz2 4784 a PHinsPnoisedBmi MoiseBand T TEMPEEATUBE SMin PRF FNHin The thermal noise in the equations is calculated using function PnoisedBm NoiseBand Te 10 le k Tet 273 15 NoiseBand where Ic ambient temperature Celsius NoiseBand noise frequency bandwidth Harbec mixer analysis output dataset including equations that calculate PIF POUT FIF SNOUT and NF NF SSB 285 Simulation 286 jx _FHOISE NFSSB 1 aes 8 87 SEIFNOISE Ha Freq Ha FregID Ea FregIndexIM SSIFSPMOISE SSIMAM SEINAPM SEIMCGAIM Ha NCIndex aa NCMame a NCValue SEIMFDSE SSINFSSB S
591. uencies simulated is double the harmonic frequencies The result is that a nonlinear noise simulation requires four times the memory of a normal harmonic balance simulation Use low values for the parameter Maximum Order ander the General tab to limit the demands on computer memory 270 HARBEC Harmonic Balance Analysis If computer memory is insufficient for a noise simulation reduce the number of tones in the harmonic balance simulation component by 1 The results of the noise figure simulation should not change significantly If the noise figure is needed add an input port to the circuit In addition because the single sideband definition of noise figure is used the correct input sideband frequency must be specified The frequency is associated with input port name To provide the noise figure measurements the input port must be a one tone signal port Oscillator Noise Simulation This section focuses on setting up noise simulations for oscillators It describes how to calculate Phase noise around fundamentals and harmonics Absolute noise voltage spectrum around a harmonic Relative noise voltage spectrum around a fundamental If you are not familiar with the general procedures for simulating oscillators refer to Harmonic Balance for Oscillator Simulation before continuing with this chapter Simulating Phase Noise To determine oscillator phase noise 1 Open the Harmonic Balance Oscillator Analysis Options dialog box
592. uential and in order For example if you swapped ports 1 and 2 above you could not use a 3 mode port because the ports would be in the order 2 1 3 along the sidewall Running the circuit above in EMPOWER will give 6 port data as would be expected by glancing at the picture However the fourth port is the only normal single mode port In the data file the first three ports of data are in mode space and the last two ports of data are in mode space For example in the data file e S41 represents the transmission of energy from mode 1 of multi mode port 1 2 3 to port 4 e 25 represents the transmission of energy from mode 1 of multi mode port 5 6 to mode 2 of multi mode port 1 2 3 e S66 represents the reflection of energy in mode 2 of multi mode port 5 6 Multimode data should be carefully connected Multimode ports should be connected only to other identical multi mode port or line configuration same box line widths spacings etc Otherwise the connection is non physical and the results are meaningless See the Spiral Inductor example in the Decomposition section for more information on the use of Multimode lines Setup Modes x These boxes are for modal use only For normal parts clear all boxes 4 1 Iv 6 gt empower F F Ports 5 j OF Cancel 141 Simulation Generalized S Parameters When normal circuit theory analysis is performed the ports are often terminated with a standard imped
593. ular cells that appear in the mesh Triangular cells are used to generate a mesh on the curved area of an object The larger the value in the Arc Resolution field the fewer the number of triangular segments will be used on the curved object The maximum value fot the Arc Resolution is 45 degrees The minimum value depends on the original faceting used when the curved object was drawn For example the Double Patch Antenna is drawn with an Arc Circle Resolution of 30 degrees and the Arc Resolution is set to 45 degrees The actual value used to refacet the object will be in between these values The angle is chosen so that the facet length is smaller or equal to the cell size so refacetting will not affect the cells per wavelength that is specified for computing mesh density Momentum GX In order for the mesh generator to produce an accurate mesh the mesher must be able to recognize the polygon translation it receives as an atc D Therefore a poorly defined layout resolution may prevent arc recognition and by that arc refacetting For more information on layout resolution refer to Mesh Precision and Gap Resolution The VIASTUB example is shown below the Arc Resolution is set to 45 degrees the coarsest setting and 0 01degrees which effectively turns off refacetting The mesh for a stub based on each angle setting is displayed The simulation for the denser mesh was longer required more memory and produced little improvement in the results
594. ulation by specifying the frequency sweep the Simulation Options and the Mesh options The simulation process computes the Green s functions for the layer stack then the mesh pattern and finally the currents in the design S parameters are computed based on the currents If the Adaptive Frequency Sample sweep type is chosen a fast accurate simulation 1s generated based on a rational fit model For more information refer to Adaptive Frequency Sampling 5 View the results The data from a Momentum simulation is saved in the analysis dataset It includes board S parameters feeding T lines wave impedances and results of full circuit co simulation Selecting the Correct Mode 306 Momentum can operate in two simulation modes microwave or RF Select the mode based on your design goals Use Microwave mode for designs requiring full wave electromagnetic simulations that include microwave radiation effects Use RF mode for designs that are geometrically complex electrically small and do not radiate You might also choose RF mode for quick simulations on new microwave models that can ignore radiation effects and to conserve computer resources RF Mode provides accurate electromagnetic simulation performance at RF frequencies At higher frequencies as radiation effects increase the accuracy of the Momentum RF models declines smoothly with increased frequency RF Mode addresses the need for faster more stable simulations down to DC whil
595. um Through nonlinear models the phase noise may remain the same or increase or decrease in amplitude relative to its parent spectrum For example through a frequency doubler the phase noise will increase 6 dB telative to its parent spectrum Phase noise is also processed by the mixer Mixed output spectrum inherit the phase noise of the LO When input spectrum have phase noise and there is no LO phase noise specified then the mixed spectrum will retain the phase noise of the input spectrum Phase Noise Coherency The coherency of the phase noise is the same as the coherency of the parent spectrum 477 Simulation 478 Viewing Phase Noise Data In graphs phase noise is always displayed in dBm Hz even though the channel measurement bandwidth is something other than 1 Hz However when the mouse is placed over the phase noise bandwidth scaling and other information it provided as shown below IF Output P24 219 1MHz 170 176dBm Phase Noise at Carrier 70 0 dBc Hz Bandwidth Adjustment 53 0 dB 3h CInterferer Osc Interferer Splitter Mixer Filter The absolute frequency point with its power in a 1 Hz bandwidth is displayed The phase noise power at the carrier frequency is displayed in dBc Hz The bandwidth scaling factor 20 Log Channel Bandwidth is also displayed The last line shows the spectrum coherency number in braces then the phase noise equation in square brackets followed by the path that the phase noise took to ge
596. unications Technology and Electronics v 42 1997 N 1 p 13 16 originally published in Radiotekhnika 1 Elektronika v 42 1997 N 1 p 13 16 Yu O Shlepnev A new generalized de embedding method for numerical electromagnetic analysis Proceedings of the 14th Annual Review of Progress in Applied Computational Electromagnetics Monterey CA March 16 20 1998 v II p 664 671 Yu O Shlepnev Extension of the method of lines for planar 3D structures Proceedings of the 15th Annual Review of Progress in Applied Computational Electromagnetics Monterey CA 1999 p 116 121 Test Examples and Comparisons E G Farr C H Chan R Mittra IEEE Trans v MTT 34 1986 N 2 p 307 G Gronau I Wolff A simple broad band device de embedding method using an automatic network analyzer with time domain option IEEE Trans v MI I 37 1989 N 3 pp 479 483 D J Swanson Grounding microstrip lines with via holes IEEE Trans v MTT 40 1992 p 1719 1721 J C Rautio An ultra high precision benchmark for validation of planar electromagnetic analysis IEEE Trans v MTT 42 1994 N 11 p 2046 2050 EMPOWER Planar 3D EM Analysis T Kawai I Ohta Planar circuit type 3 dB quadrature hybrids IEEE Trans v MTT 42 1994 N 12 p 2462 2467 Y Gao I Wolff Miniature electric near field probes for measuring 3 D fields in planar microwave circuits IEEE Trans v MTT 46 1998 N 7 p 907 913
597. unless you also change the number of ports on the Layout b Do notchange the configuration of ports used for lumped elements otherwise the results could be incorrect Sonnet Interface c Do not remove the Genesys yp file output However you may change the Touchstone format options of this file d Ifyou save the Sonnet file to a new name or location GENESYS will not be able to read it back If you want to do this you can read the data back into GENESYS by creating a link to the resultant Y Parameter data file 7 While in the Sonnet editor save and analyze your file 8 Right click the Sonnet simulation on the Workspace Window and select Manual Mode Load Calculation Results into GENESYS This loads the Genesys son and Genesys yp files into the GENESYS workspace and updates all output graphs 9 If additional changes to the simulation are desired repeat steps 6 and 7 In Manual Mode your updated Genesys son file is saved back into the workspace step 8 above eliminating the need to store the Sonnet file separately Passing this file to cowotkers will allow them to export edit analyze and update your modified Sonnet file 397 Simulation Translation D etails Sonnet O ptions Dialog Box Sonnet Interface Options General Sonnet Advanced Port impedance Use ports From schematic Necessary For HARBEC co simulation Electromagnetic simulation Frequencies Co simulation sweep enn SEHE Wed edd Use
598. urate if only a few points are used to represent either the sional or noise through a device whose impedances vary across the channel Click here for more information about adding additional data points to signal sources As can be seen in the following figure a wideband signal is slightly larger than the narrowest filter bandwidth If only 2 points are used to represent the signal then the channel power will only be as accurate as those 2 points If additional data Spectrasys System points are add to the signal a better representation of the signal and noise will be achieved Remember cascaded noise figure in not only dependent on the channel noise power but also the cascaded gain signal Wider than Filter SLE TIE ALL Lil LL LL ETE amp 10 Point Signal M 2 Point Signal s E i 7 m L En un 4 E CL T a FR ju r CL c ae HE Ltt 2405 2450 5 249 249 5 20 9 291 291 5 2 ca T More accurate noise pedestals would like to be seen on spectrum plots By default Spectrasys uses a very small number of noise points to represent the entire noise spectrum The following figure shows what noise would look like in was only represented by 4 points These 4 noise points are evenly spaced between 0 and 500 MHz 0 166 67 333 33 and 500 MHz You see additional noise points around 250 MHz because of smart noise points insertion which 1s discussed later The number of noise points can be ch
599. urement Result in graph Smith chart optimization Result on table or yield PPORT 1 dbm PPORT 1 RMS power delivered to dbm PPORTT1 port 1 Not available on Smith Chart Port Voltage VPORT This voltage measurement array is the RMS Voltage at the ports Values Real value in specified units Simulations Nonlinear dc analysis Default Format Table MAG Graph MAG Smith Chart none Short Form Vport Example V1 is equivalent to VPORT 1 Examples Measurement Result in graph Smith chart optimization Result on table or yield ICP1 magnitude of ICP1 current through current magnitude of ICP1 probe 1 Not available on Smith Chart Node Voltage Vnode 31 Simulation 32 This voltage measurement is the peak voltage at the specified node The node is the name of the node net as seen on the schematic or in the part netlist Values Real value in specified units Simulations Nonlinear dc analysis Default Format Table MAG Graph MAG Smith Chart none Examples Measurement Result in graph Smith chart optimization Result on table or yield VTP2 mag VTP2 voltage at test point TP2 mag VTP2 Not available on Smith Chart Large Signal S Parameters The Large Signal S Parameters LS parameters of an N port non linear device can be calculated using 2 different methods 1 Exciting each of ports of the Device Under Test DUT separately To calculate the full set of LS parameters this meth
600. urns the index of first element of array lt array gt which has value lt value gt Gain db10 P2 Freq 2 850 P1 Freq 22850 243 Simulation 244 SweepPowerl Data Variable ie 1 fone HB input RFpower 5wp zm 1 tone HB input RFpower 5 amp wp Freg FH 1 tone HB input RFpower Swp Time fA FPreq fe a a rrequ ejFregindexIM EET Gain Gain 1 db 10 P 2 Freq 850 P 1 Freq sso r2 Gain TA i mH 1 3n 2 LI S E te Prout if PPORT i Tiri 4 7 861 7 657 O aj 766 EE 7597 e 74 s 5 798 l 1 842 S 2 511 8 758 Table 1 Basic HB analysis Measurements o 4 5 Measurement Genesys 2004 and earlier Md E ang Voltage Pp CUNA V lt node name gt V lt node name gt the node lt node name gt Voltage Spectrum at lt gt lt gt the port lt por iY port number VPORT lt port number gt number gt Current Spectrum through X Icvoltage source name gt gt branch I lt branch name gt Trae node branch name gt Voltage W_V lt node name Waveform at or gt the node TIME LV node nime time V lt node lt node name gt name gt Freq Time 5 Current TIME I branch name gt W_I lt branch name gt HARBEC Harmonic Balance Analysis Spectrum TIME I voltage source name gt or through the time I lt branch bran
601. urrent voltage viewer program Selecting this box will increase the amount of time required to solve the problem This box must also be checked in order to generate far field radiation data See the Viewer section for more information Port number to excite This option is available if Generate viewer data above is checked It specifies which EMpott to excite for viewer data By default mode one is excited but if the input is multi mode then you can add the option Imj to excite mode j instead 103 Simulation Mode number to excite This option is available if Generate viewer data above is checked It specifies which mode to excite for viewer data Generally mode one is excited but if the input is multi mode then you can add excite any mode number up to the number of modes at that input Generate Far Field Radiation Data Checking this box causes EMPOWER to generate data for the radiated electric fields of a structure in the far field region The data generated is specified by the sweeping theta and phi coordinates of the spherical coordinate system Sweep Theta This option is available if Generate Far Field Data above is checked It generates data for varying theta in the spherical coordinate system Theta is the angle formed from the z axis to a point in 3 space If Sweep Theta is unchecked a fixed angle will be specified and far field data will be produced only at this theta angle Sweep Phi This option is ava
602. used If you are calculating your mesh at low frequencies the estimated cell size is high and you may not want thin layer overlap extraction In this case disable the Thin layer overlap function Mesh Generator Messages When a mesh is generated you may see a message similar to this On some layers the mesher used the maximum allowable cell size for the given substrate You will not be able to generate a coarser mesh on these layers 363 Simulation 364 This means that the coarsest mesh has been achieved for the objects on certain layers This is because the maximum cell size has been limited due to the accuracy of the Green s functions that were calculated for the substrate For example consider a layout with 20 cell mesh when the mesh frequency is 10 GHz Generally you would get a 10 cell mesh if you set the mesh frequency to 5 GHz But if with the lower mesh frequency the cell size would be larger that the maximum allowed the limit being based on the accuracy of the Greens function of the substrate the mesh would contain more than 10 cells Another message that may occur at meshing reports POLYLINE requires two unique points This occurs if the layout mesh resolution set in the Momentum GX Mesh tab is inadequate for slanting the meshing line Increase the resolution to a finer value to make the error disappear Guidelines for Meshing The default mesh will provide an adequate and accurate answer for many ci
603. ut used to be Note These sources are also found in the Input button on the schematic Basic toolbar Double click the input and set the pulse width to 500 ns This will give a 50 duty cycle pulse at the default Frequency for the part of 1 MHz We are now ready to add the transient analysis Click the New Item button on the Workspace Tree and select Analyses Add Transient Analysis Click the Factory Defaults button to be sure we are starting with good settings This is also a good practice is an analysis begins to have unexpected behavior The only parameters we need to modify are Stop Time and Maximum Step Size Set the Stop Time to 5000 ns and the maximum step size to 10 ns This will show us 5 cycles at 100 MHz with 100 time points in each cycle Note that the simulator will simulate at more points as necessary to maintain accuracy Click the Calculate Now button This will close the dialog box and run the transient analysis Double click the newly created Transient1 Data dataset in the workspace tree Right click on the VPORT variable and select Graph Rectangular Graph You should now see a graph similar to the one below CAYENNE Transient Analysis Transient _VPORT mnm z Ed ci bh gt C l z cL Oo a7 LL ka A 500 1000 1500 2000 2500 3000 3500 4000 4500 S000 Time n PORTE PORTE Note that the output has only small spikes at each transition This is consistent with our cir
604. vantages e Fast simulation speed e Identification of every spectrum e Signals can be seen underneath other signals e True in channel signal to noise ratio measurements e Spectral directionality e Bandwidths for all spectrums e Broadband noise e Phase noise e Path VSWR effects 413 Simulation e Multiple path analysis for single block diagram e Restrictive assumptions from traditional cascaded equations are removed e Flexibility for future growth G etting Started Spectrasys Walkthrough O verview Spectrasys uses a new simulation technique called SPARCA that brings RF architecture design to a whole new level This walkthrough will help you design a simple RF chain and measute the architectures noise and gain performance The basic steps to analyze an RF system is 1 Create a schematic 2 Adda system analysis 3 Run the simulation 4 Add a graph or table Create a System Schematic Spectrasys supports all linear models and behavioral non linear models The behavioral models can be found on the system toolbar or in the part selector 414 Spectrasys System Part Selector A Current Library Description T Amp 2nd amp 3rd Or RF Amplifier T Amp High Order High Order RF A J Amp Variable Gain Variable Gain Am 3 Antenna Path Antenna Path LE Attenuator Attenuator ET Attenuator DC Co Attenuator TF ArtenuatoriVariable Attenuator Vari T Bandpass Fileer Be Bandpass Bessel T Bandpa
605. vantages and disadvantages Carson 1 and Altman 2 provide additional information S Parameter Basics S parameters have earned a prominent position in RF circuit design analysis and measurement Parameters used earlier in RF design such as Y parameters require opens ot shorts on ports during measurement This is a nearly impossible constraint for high frequency broadband measurements Scattering parameters 3 4 S parameters are defined and measured with the ports terminated in a characteristic reference impedance Linear Analysis Modern network analyzers are well suited for measuring S parameters Because the networks being analyzed are often employed by insertion in a transmission medium with a common characteristic reference impedance S parameters have the additional advantage that they relate directly to commonly specified performance parameters such as insertion gain and return loss Two port S parameters are defined by considering a set of voltage traveling waves When a voltage wave from a soutce is incident on a network a portion of the voltage wave is transmitted through the network and a portion is reflected back toward the source Incident and reflected voltage waves may also be present at the output of the network New variables are defined by dividing the voltage waves by the square root of the reference impedance The square of the magnitude of these new variables may be viewed as traveling power waves a1 inci
606. ve does not in general result in an excellent return loss match Balanced amplifiers and isolators are sometimes used to achieve both the optimum noise figure and a good match Getting Started Losses in the input network feedback networks around the transistor emitter feedback and multiple stages all effect the noise figure of the circuit All of these effects are accurately simulated in GENESYS using the noise correlation matrix technique 5 6 Noise parameters can be added to the two port data files after the S Y G H or Z parameters See the section Creating New Data Files for information about entering S Y G H or Z parameters Each line of a noise parameter has the following five entries Frequency NF dB Mag Gamma Opt Ang Gamma Opt Rn Zo Frequency Frequency in units NF dB Minimum noise figure in dB Mag Gamma Opt Magnitude of the optimum source reflection coefficient for minimum noise figure Ang Gamma Opt Angle of the optimum source reflection coefficient for minimum noise figure Rn Zo Normalized effective noise resistance Here is an example of noise data in a file along with the device S parameters BFP620 Si NPN RF Transistor in SOT343 Vce 2 V Ic 8 mA Common Emitter S Parameters 01 February 2000 GHz S MA R 50 f S11 S21 512 522 GHz Mag Ang Mag Ang Mag Ang Mag Ang 0 010 0 8479 1 3 21 960 179 3 0 0024 27 9 0 9851 0 4 0 020 0 8424 1 9 21 606 178 2 0 0021 34 2 0 9676 1 5 0 050 0 8509 5 7 2
607. very Frequency Calculate gt Substrate O All substrate mesh and 5 Parameters 2 In the Calculate section select Mesh and substrate 3 Click OK Run Momentum analysis when computations of substrate and mesh are complete the mesh will be displayed on the layout Viewing Mesh Status After mesh computations have started any messages regarding the computations will appear in the Simulation Status window Messages usually refer to any errors or warnings ot indicate when the mesh is complete e All messages sent by the Momentum analysis are saved in the Momentum Analysis dataset text variable LogOutput The value of LogOutput is shown in the Simulation Log window e You may stop Momentum simulation at any time by clicking the Stop button on the Momentum Simulation status window 355 Simulation W Simulation Status Running Momentum press the Stop button to end calculations Substrate is electrically small below 6 63 GHz surface wave radiation Extracting layout Adaptive Frequency sweep started RF mode amp amp PORTZ 50 Simulation Frequency 1 0 Hz Fast View Rate loading Green Functions loading quasi static matrix Redraw after each sweep OOo OE e It may take a few moments for Momentum to stop Viewing the Mesh Summary If the mesh computation is successful you can view mesh statistics in the in the Simulation Log window the Momentum Analysis dataset text variable
608. wept Variable vs Measurement at a given node Background The following schematic will be used for the discussion of this topic FEFilter IF Filter Notice that 3 of the nodes have been manually renamed to help illustrate concepts that will be shown hereafter To rename a net right click the net then select Net and then Rename from the sub menu In this example the source frequency is swept from 2110 to 2170 MHz An equation is used to slave the LO frequency to the input frequency so the IF frequency will remain constant across the frequency sweep Click here for additional information on setting up parameter sweeps When a sweep is performed on a system analysis the spectrum data is swept and a new swept main system dataset will be created If the system analysis contains paths then each path will be swept and a new swept path dataset will be created for each path Swept Level Diagram To create a swept level diagram do the following 440 Spectrasys System e Adda rectangular graph e Set the Default Dataset or Equations to the swept path dataset of interest e Type in or select the path measurement Swept Cascaded Noise Figure CMF dB In T L 7 gt mn OS F H o FE Fihn brat M ode blo iu Swept Variable versus Measurement at a given node Many times users need a measurement at an input output or some other intermediate node For example the user may want to examine the cascaded noise figure at
609. wer displays absolute values only If not selected an actual value with information about flowing direction is displayed The difference is that absolute value is always positive whereas the actual current values can be positive for forward directed currents and negative for backward directed currents Negative amplitudes are drawn below the x y plane This option has a checkmark beside it when selected Toggle Animation When selected the viewer animates the image in real or angle mode This is accomplished by multiplying the individual currents by exp jw where w cycles from 0 to 2pi and showing a sequence of snapshot images for increasing w This option has a checkmark beside it when selected Toggle Scale When selected the viewer displays the scale in the lower left of the viewer window This option has a checkmark beside it when selected Toggle Value Mode Real Mag Ang This option selects the current display option The options include the Real current value for current distribution snapshots and animation Magnitude for time averaged current values and Angle for the current phase delay distribution snapshots Toggle Wireframe When selected the viewer displays a wireframe version of the current plots A wireframe is created by drawing the outlines of the EMPOWER grid currents without filling the resulting polygons When this option is not selected the viewer fills the polygons resulting in a solid surface plot
610. with the grid function re expansion in a discrete space The GGF matrix of a symmetrical problem could be reduced to a centrosymmetrical matrix with centrosymmetrical blocks in the case of two plane symmetry and it is treated in the way similar to described in Weeks 1979 This reduces required CPU memory from 4 to 16 times serial allocation of partial matrices and speeds up calculations from 4 to 16 times One plane two plane and 180 rotational symmetries are included in the program Thereafter the classic Gauss inversion algorithm is used with a few changes The result of this stage of solution is a matrix Y or Z matrix relating the grid currents and voltages in the input source regions and thus we need to get only a small part of the inverted matrix corresponding to these variables A partial inversion procedure performs it and gives an additional acceleration D e Embedding Algorithm The method of simultaneous diagonalization MoSD Shlepnev 1990 1998 is used to extract a multimode or generalized S matrix The MoSD 1s based on the electromagnetic analysis of two line segments corresponding to an MIC structure port to be de embedded The segments have different lengths and the same surface current source regions as in the initial structure The result of the EM analysis is two Y matrices relating integral grid currents and voltages in the source regions These matrices transformed from the space of the grid functions to
611. xing terms and will not override the order of individual soutces specified in the frequency table Temperature The temperature in degrees Celsius at which to perform nonlinear analysis Maximum Analysis Frequency Frequency above which no nonlinear analysis is performed If not checked all frequency points in the analysis input frequencies their specified number of harmonics and intermods will be used Frequency accuracy The minimum difference in frequencies before the simulator will merge frequency terms If the difference between two calculated frequencies usually mixed frequency terms is less than the frequency resolution they will be considered a single frequency term for simulation Calculation defines basic convergence criteria and contents of output data 227 Simulation 228 Harmonic Balance Analysis Options General Calculate Noise Advanced Convergence Criteria Absolute Current Tolerance Absolute Voltage Tolerance Relative Tolerance Minimum Relative RHS Norm Change Oyversampling Factor Jacobian Harmonic Factor jallow 1 D FFT use Man Binary FFT Options Automatic Recalculation dutoSave Workspace After Calculation adds Backup to Filename Use Previous Solution 4s Starting Point Mutou Wave Daka Save Solution For All Modes CO Mane Da Not Calculate Wave Data Ports Only All Nodes Absolute Current Tolerance The absolute accuracy to which
612. y quick Calculate Intermods When checked the system analysis will calculate and save intermod data to the system analysis dataset Intermod simulation time depends on the number of input signals levels of the resulting intermods number of non linear stages and how tightly coupled loops are Unchecking this option can improve the simulation speed drastically during troubleshooting of a block diagram From Sources Only When checked harmonics and intermods will only be created from source sionals All undesired products created along the path will be excluded from the calculation of harmonics and intermods When unchecked harmonics 546 Spectrasys System and intermods will be created from all signals appearing at the input to the non linear element This includes intermods harmonics and other undesired signals This option typically requires longer simulation time since more spectral components are being created Odd Order Only When checked only odd order intermod and harmonics will be created Coherent Addition When checked intermod harmonics and mixed signals will be added coherently Generally cascaded intermod equations assume coherent intermod addition Noz Desired signals will always be added coherently regardless of this setting See the Coherency section for mote information Fast Intermod Shape When checked undesired intermods and harmonics will be represented by only 2 data points In most cases thi
613. y spreadsheets Spreadsheets typically never account for mismatch loss so it is assumed that all forward traveling signals equal the transmitted signals which is really not the case Example 535 Simulation The following figure shows a 3 dB attenuator followed by a low pass filter The source frequency is 1 GHz and which frequency the filter produces about 50 dB The channel power and voltage is displayed on the following graph As expected the voltage changes very little across the 3 dB attenuator However most of the power drop occurs across the attenuator The input impedance to the filter at 1 GHz is about 336 ohms For this reason there is more attenuation that appears across the attenuator due to the mismatch loss than the anticipated 3 dB Out of Band Filter Path D D 5 0 45 12 D 4 18 0 35 5 24 na 5 D i Le T T 3 30 0 25 L c z z E 35 D2 t i x 42 0 15 48 D 1 5A 0 05 BD D Allri FII ler Dialog Box Reference System Simulation Parameters General Tab This page sets the general settings for a Spectrasys Simulation Tabs General Paths Calculate Composite Spectrum Options Output 536 Spectrasys System System Simulation Parameters General Paths Calculate Composite Spectrum Options Output Design To Simulate v Dataset Systeml Data amp ubamatic Recalculation Frequency Units Measurement Bandwidth Calculate Mow Nominal Impedance 50 Ohms Channel MHz P
614. you are looking at the circuit from the top along the z axis A box can be used only where the top and bottom layers in the substrate definition are groundplanes or impedance termination The four vertical metal walls plus the top and bottom groundplanes result in a box hence the name the walls are the sides and the top and bottom ground planes are lid and base of the box A box can represent amp highly resonant metallic enclosure ts effect will be taken into account during airnutation The next section describes how to apply a box to a circuit For more information about applications for this item refer to About Boxes U sing a Box 342 A box defines the boundaries on four sides of the circuit substrate One box can be applied to a circuit at a time A box can be applied to a circuit only if the top and bottom layers of the substrate definition are defined as groundplanes or impedance termination The walls of the box are perfect metal The ground planes can be defined either as perfect metals or a lossy metal Using a box in the circuit enables you to analyze the effects of enclosing the circuit in metal for example to identify box resonance Box resonance can have a significant effect on S parameters in a small band centered around the box resonance frequency When a simulation is performed the resonance frequencies will be noted in the status window when the circuit is simulated along with the frequency bands where no smoo

Simulation - Agilent Technologies

Contents

Download Pdf Manuals

Related Search

Related Contents