user manual - Rohde & Schwarz
Contents
1. DUT Measured i ka noise H ene Predistored seetbeeeee eet i Signal i f H DUT simulation H Expected nonlinearity Predistored oe 4 eee cece eee Signal t i H Predistortion All data which are displayed run the same signal processing path for displaying e apply selected window function to signal e transfer from time to frequency domain calculate amplitude from IQ data e calculate ACP data on linear values e apply signal detector calculate dB values from linear values All bold marked signals are display in the ACP measurement window and can be switched on and off to have a closer look at the different signals 21 Rohde amp Schwarz AmpTune version 3 0 user manual 9 Measurement results documentation You can generate a HTML document which contains all the measurement results To configure the reported results use the menu item Measurement Report report options and select the items to report and additional informations S report options x test informations Operator Je ob Meyer Company Rohde amp Schwarz DUT BGPP amplifier Note inder standard conditions items to report IV general results IV nonlinear parameters T graphical results detailed After that you can view the report using the menu item Measurement Report gt show measurement report or save it using the menu item Measurement Report gt save measuremen2. 2 3 a S x 5 a 4 5 E seeeegegese a 5 Bus T T T T T T 7 T T T T 1 0O 2 40 60 80 100 120 140 160 180 200 220 240 s N e e 5 s 20 Rohde amp Schwarz AmpTune version 3 0 user manual After the measurement is finished the histogram of the output amplitude distribution and the histogram between input and output phase is displayed It is possible to have a detailed look at the AM AM and AM PM curves by zooming rotation and moving the graphics lt STRG gt Mouse Rotate diagramm lt ALT gt Mouse Zoom diagramm lt SHIFT gt Mouse Move diagramm 8 ACP improvement measurement 1MA27_manual To understand the different curves which are displayed when running the ACP improvement measurement have a short look at the overall DSP signal flow chart and the corresponding switches in AmpTune r Measurement i i es DUT PER REEL ee eet ate t 7 H i Reference Pe ce o Measured __ eme Signal linearity Signal to MS HE DA noise floor EM reference or E lt iithaut predistortion DPS nonlinear parameters i vith AM predistortion i with PM predistortion with AM PM predistortion DUT simulation H EO i ar i nonlinearity Expected Signal i nennen Beten H fot Pre Distorer
3. Att 20 dB T 2 55 MHz Span 25 5 MHz 31 JUL 2002 12 AmpTune Version 3 0 USER MANUAL ROHDE amp SCHWARZ Subject to change 09 2003 1MA27_manual Contents 1 Overview 1MA27_manual AmpTune version 3 0 user manual Auto leveling a nennen aa Main measurement ccccceeeeeseeceeeceeecesaeseeeeceeeceaeeeseeeeeeeeaneasaeeees 4 Signal processing details 44444444Hennnnnnnnnnnnnnnnnnn nn nnn ann Timing offset calculation eccora E Polynomial calculation amp data gridding sn 5 Startup screen and basic Setup ccccceceeeeeeseceeceeeeeeeeeesncaeeeeees ACHONWINdOW nen anne naeh E een Hardware window uuessseeeneneneesennnnennnnnnnsnsnnnnnnnnnnn sn nnnnnnnnnnen seen MENU Bar ae nnd eren una eannaen 6 ENMORMESSAgES nain een ki Instruments missing u een nennen BHILICALKERFON Se een NE Non critical CrrOrs ccccccccceseeseeceeeceeeceaaeeeeeeeeecesaeeeseeeeeeeeaneaseeees 7 Nonlinear measurement screen ccccccccccccccecececeeeeeceeeeeeseseeeeeeeees Before timing COrrecti N sassen nennen After timing correction uenesnnnnennennnnnnnennnn
4. press setup to fill The Baseband to RF convertion is performed by an SMIQ You can set 2 different parameters here e Baseband filter To improve the output performance of the SMIQ you can switch on several SMIQ internal IQ filter NOTE Please take care that the test signal is not influenced by the filter e g do not use a 850 kHz filter to measure a 3GPP signal at full clockrate The filter is NOT switched off for ACP improvment measurement e Maximum output level In order to protect your DUT from overload this value specifies the maximum peak output level of the SMIQ The value is a value stored in the setup of the SMIQ and prevents the DUT from overloads generated by all type of software or hardware behavoirs e g not connected analyzer input cable which will cause the auto level function to set the level to a very high value 1MA27_manual 12 Rohde amp Schwarz 1MA27_manual AmpTune version 3 0 user manual Device under test DEVICE UNDER TEST Freq 1000 MHz Dwell 1 8 targert RMS o power value 9 dBm This element sets the main parameters for the DUT measurement e Frequency The frequency which is set here is transfered to both RF generator and RF analyzer e Dwell Due to settling effects of the amplifier you can specify a dwell time In this example every measurement will be taken 1 second after the corresponding setups are made on the generator e Target RMS power value
5. 3 Rohde amp Schwarz AmpTune version 3 0 user manual 3 Overview of operation To show the details of AmpTune and how each measurement step is done you may have a look at the flowchart shown below Start of the measurement v lt SETUP gt v lt AUTO LEVELING gt y lt MAIN MEASUREMENT gt AX Predistore no w YES display spectrum for source signal and raw DUT output signal v lt PREDISTORE gt each selected v predistortion display spectrum for predistored DUT output signal A Measurement finished Setup lt SETUP gt attaching all devices set operation mode reference frequency source set RF frequency generate test signal internal or load from file load test signal to baseband generator start test signal switch RF source signal off set RF settings on analyzer perform noise measurement During the setup process AmpTune attach all instruments to the computer switch the SMIQ to ARB mode if SMIQ is selected as basband signal source set up the reference frequency master and slave devices check the reference frequency connections and set the RF frequency for both generator and analyzer 1MA27_manual 4 Rohde amp Schwarz AmpTune version 3 0 user manual In the next step the test signal is generated This is done by loading the selected WinIQSim file or calcul
6. ain 1 stage difference be i and analyser ween peak and RMS level tween generator RMS level RMS level 1 Step RMS output level at 2 level mean output power 2 stage estimation step ain 2 stage difference be i and analyser tween generator RMS level RMS level 2 Step mean output power at target level RMS output level at target level gain at target level diference be and analyser noise level with RF generator OFF RMS level with RF signal source OFF analyzer dynamic margin Patentamt Is 0 2d a g sample amplitude and max possible value crest factor of reference signal difference bet tween generator RMS level RMS level target level tween highest measured ween peak and RMS level crest factor of measured signal difference be tween peak and RMS level red 6 2 2250 a dB dB tin ad d setti 0 5dB 5 15dB 15 dB SP Ser uN ned menor difference difference difference depending on DUT 0 30dB 30 50dB 50 dB tin ad d setti 0 5dB 5 15dB 15 dB BISCHEN difference difference difference depending on DUT 0 30dB 30 50dB 50 dB as set in advanced settings 0 9 dB 319 dB 15 dB 9 difference difference difference depending on DUT 0 30dB 30 50dB 50 dB 120 120 100 100 dBm dBm dBm 0 2 if BoB dB HS Se 2 0 eee dB dB 2 2 0 dB dB dB above 5 dB first timing estimation result timing difference between reference and measured sign
7. 2 MHz The baseband generator can be a AMIQ or a SMIQ B60 option within the SMIQ Both can be controled by the software NOTE The maximum memory depth of the SMIQ B60 is limited to 512 kSamples for a output waveform and 1408 kSamples for the total memory AmpTune does NOT check whether there is enough free space in the memory of the SMIQ B60 option or not There are only 2 different parameters which can be set for the baseband generator e Depth For band limited noise signals and dual tones this value sets the length of the signal For external WinlQSim files this value is ignored It can be set but has no influence on the hardware setup For WinIQSim files all values are transfered to the baseband generator 11 Rohde amp Schwarz AmpTune version 3 0 user manual e Clock For ALL types ot output signal this value sets the clockrate for the signal generation NOTE For WinlQSim files the clock rate stored in the file is ignored You can e g use a 3GPP signal which was generated with 32 MHz clock rate and set the generator to 8 MHz This will lead to a signal bandwith of 1 MHz You can record the nonlinear parameters with this signal and afterwards increase the clockrate to 32 MHz in order to check the relation between clockrate and nonlinear effects To check the clockrate the WinlQSim files was generated with simply us the info button BB gt RF convertion BB gt RF CONVERTION
8. In the first measurement step the output level of the amplifier is set to the value specified with this input box NOTE The level reading is taken at the input of the analyzer So additional cable attenuations after the DUT must be taken into account EXAMPLE Target DUT output level 4 9 dBm Cable attenuation 1 2 dB Target RMS power level 6 1 dBm RF gt BB RF gt BB CONVERTION amp BASEBAND ANALYZER Depth 32k samples Clock 2 MHz press setup to fill For RF to baseband conversion FSIQ FSP and FSU are supported There are 3 different parameters for the setup 13 Rohde amp Schwarz 1MA27_manual AmpTune version 3 0 user manual Depth This value sets the memory depth for the analyzer NOTE For data analysis the size of generated and analyzed signal sequence must be equal So the specified number of samples for the analyzer is used to get a part of this length from the generated data Baseband filter To improve the output performance of the analyzer you can switch on several analyzer internal digital IQ basband filters NOTE Please take care that the test signal is not influenced by the filter e g do not use a 850 kHz filter to measure a 3GPP signal at full clockrate The filter is NOT switched off for ACP improvment measurement The filter is not available for FSIQ Clock This value specifies the clock rate for recording IQ data
9. and AM PM on or off with the corresponding switches In the example above AmpTune will record the AM AM and AM PM curves and perform a combined AM AM and AM PM predistortion NOTE It is possible to record the AM AM and AM PM curves and predistore any WinlQSim file by using the menu item Data Management gt predistore existing file It is NOT possible to predistore internal generated files after recording the AM AM and AM PM curves without including predistortion 1MA27_manual 10 Rohde amp Schwarz 1MA27_manual AmpTune version 3 0 user manual Hardware window BASEBAND BB gt RF GENERATOR CONVERTION DEYICE UNDER RF gt BB CONVERTION amp BASEBAND TEST ANALYZER Depth 32k samples targert RMS power value 0 dBm p z signal type 10 MHz reference band limited noise noise init 0 signal BW 2 MHz baal dual tone SMiQ z WinlQSim file F S LOKAL WEISS AppN 1MA27 AmpTune _ info Within this window the main settings for the hardware are done There are different windows for all instruments For every window which represents an measurement instrument you can see a selection list at the bottom This list shows all available instruments on the IEEE bus When the program is started this list is empty You have to press setup in order to show the available instruments Baseband generator BASEBAND GENERATOR Depth 32k samples Clock
10. 60 dBm or less Perhaps the generator was switched to local and did not receive the RF ON command Switch the generator RF output on manually The measurement produces some non critical errors and the measurement crashes The analyzer shows OVL or IFOV overload There may be a problem with the optimum mixer level Use Settings advanced settings and set the optimum mixer level for the used instrument to a higher level The ACP of the distored signal is worse then the reference signal There is an problem with the polynomial fit of the measured data This is a theoretic problem and can not be solved There might be an future version of this software which fits data by spline interpolation Please re run the measurement to get better results 23 Rohde amp Schwarz AmpTune version 3 0 user manual Q The ACP of the reference signal is much more better then the ACP of the distored reference signal A The ACP of the reference signal generated by AmpTune is only a theoretic value Without numeric resultions problems is should be infinity in real live it is about 300 dB or less The analyzer only has ACP values of about 60 dB Q The measured ACP is about 60 dB but have seen up to 70 dB then measuring ACP on the analyzer A AmpTune uses the IQ signal processing of the instrument in order to measure ACP Because this are complete other internal signal pathes ACP in normal analyzer operation and ACP in IQ operation
11. Signal type signal type band limited noise noise init 0 signal Bw 2 MHz dual tone WinlOSim file F S LOKAL WEISS AppN 1M amp 27 AmpTune i _ 2info You can select 3 different types of measurement signals band limited noise This type of signal generates a noise signal with a selectable bandwidth To generate the signal a FFT based signal generation method is used Example If a signal with 2 MHz bandwidth is generated using a generation clockrate of 8 MHz and a memory depth of the signal generator of 16384 16k the signal is generated in the frequency domain 16384 2 MHz 8MHz 4096 single carriers are placed near the carrier 2048 at the left side 2048 at the right side with 8 MHz 16384 488 Hz spacing and constant amplitude but random phase is applied to each charrier The other 12288 carriers are set to 0 With this method a signal with defined bandwidth can be generated Signal BW sets up the selected bandwidth noise init sets the initialising number for the random phase generator The same noise init value generated the same random phase sequence different init values different signals NOTE To generate your own custom noise signal with defined notches noise level please use the free of charge software NPR 2 which can be downloaded from the R amp S homepage 14 Rohde amp Schwarz 1MA27_manual AmpTune version 3 0 user manual dual tone This type of signal g
12. Wei Entwicklung und Verifikation eines Verfahrens zur Ermittlung der dynamischen Kennlinie von Hochfrequenzverst rkern bei breitbandigen Eingangssignalen Diplomarbeit Institute for Communication Engineering TU M nchen 12 Additional Information Please contact TM Applications rsd rohde schwarz com for comments and further suggestions amp ROHDE amp SCHWARZ ROHDE amp SCHWARZ GmbH amp Co KG M hldorfstra e 15 D 81671 M nchen P O B 80 14 69 D 81614 M nchen Telephone 49 89 4129 0 Fax 49 89 4129 13777 Internet http www rohde schwarz com This application note and the supplied programs may only be used subject to the conditions of use set forth in the download area ofthe Rohde amp Schwarz website 1MA27_manual 25 Rohde amp Schwarz
13. connection types between the instruments and the instruments and your computer 1 IEEE bus connection Please make sure that all instruments needed for the AmpTune measurements are connected to your computer via IEEE bus and all instruments have different IEEE bus address RF connection Please make sure that signal generator output and signal analyzer input are connected using the DUT The maximum DUT output level must be lower than the maximum input level of the signal analyzer Please refer to the manual for the maximum input level of the analyzer IQ signal connection only AMIQ Connect I and Q output of AMIQ to and Q input of AMIQ Please make sure that and Q are not swaped Reference frequency connection For the measurements AmpTune carries out all frequencies for the instrument internal frequency synthesis must be derived from 1 source or master clock Please connect the reference frequencies of all instruments with 1 master clock Select the master clock instrument in the software setup too Trigger connection To start waveform recording with the baseband analyzer an external trigger signal must be provided Please connect for AMIQ operation MARKER 1 for SMIQ B60 operation TRIGGER OUT 1 to the analyzer input EXT TRIGGER for SMIQ B60 operation you need option SMIQ Z5 BNC Adapter for rear panel You could also build an adaptor yourselve by using real panel pin 11 TRIGOUT1 and pin 1 ground
14. diverge about 10 dB Better results are not possible with the current hardware To get best IQ performance use an FSQ which has a much more better IQ performance than any analyzer before Q The software hangs up while taking the main measurement A You might have no trigger connection between AMIQ SMIQ and analyzer Connect the MARKER 1 TRIGGER OUT 1 output of the AMIQ SMIQ with the EXT TRIGGER input of the analyzer Q The measurement report does only contain a blank page A You have not selected any item in the report option window Please use Measurement Report report options and select some items 1MA27_manual 24 Rohde amp Schwarz AmpTune version 3 0 user manual 11 Literature and additional free software 1 2 3 4 5 6 IQWizard IQ Signal Measurement and Conversion Application Note 1MA28 www rohde schwarz com Application Notes IQWizard NPR Noise Power Ratio Signal Generation and Measurement Application Note 1MA29 www rohde schwarz com Application Notes gt NPR Simulation Software WinlQSIM Product Portfolio Test amp Measurement Signal Generation gt Products gt WinIQSIM Martin Wei Measuring the Dynamic Characteristic of High Frequency Amplifiers with Real Signals European Wireless 2000 September 2000 Dresden Martin Wei Check Amplifier Dynamic Behaviour With True Test Signals Microwaves amp RF December 2000 Martin
15. power 3 Asa last step the final caluclated level is applied to the DUT input This power should result in a very accurate output power which is only 1 10 dB or less away from the target power 1MA27_manual 5 Rohde amp Schwarz AmpTune version 3 0 user manual Main measurement lt AUTO LEVELING gt display raw amplitude and phase data display raw AM AM and AM PM data correct time offset display corrected AM AM and AM PM data AM AM and AM PM data gridding calculate polynomial factors See Signal processing details for more informations 4 Signal processing details 1MA27_manual Timing offset calculation Timing offset is calculated in 2 steps Raw timing estimation The raw timing estimation does a correction of the timing offset with a resolution of 1 sample The reference and the measured signal are converted to amplitude data and the standard correlation method is applied to get the timing difference y k Dix G k Calculation is done by using FFT in both frequency and time domain This calculation give a first raw value of the timing offset With this offset value the reference signal is generated by taking the corresponding values from the test signal 6 Rohde amp Schwarz AmpTune version 3 0 user manual Fine timing estimation To eliminate the remaining time offset both signals are subject to FFT the phases are deducted from each other in the frequency domain and a
16. shape as the applied test signal On the bottom right side the result of the conjugate complex multiplied phase spectrum of input and output signal is shown In regions there the test signal has significant spectral components the gradient of the phase spectrum shows the timing difference between input and output signal Steep spectrum large timing offset flat spectrum small timing offset 19 Rohde amp Schwarz 1MA27_manual AmpTune version 3 0 user manual After timing correction Be RD Ds w oO input amplitude input amplitude amplitude histrogramm amplifier output phase difference histrogramm amplifier output eel m T T T T T T 1 1 i 1 1 1 1 o 5000 10000 15000 20000 25000 30000 35000 0 5000 10000 15000 20000 25000 30000 35000 There meaning of the 4 displays is the same but now timing corrected data are displayed 1 Linear input amplitude versus linear output amplitude should be a curve like graphic 2 Linear input amplitude versus phase difference between input and output should also result in a curve like graphic showing some phase noise at the lower values 3 Conjugate complex multiplied amplitude spectrum of input and output signal is the same as before 4 Conjugate complex multiplied phase spectrum of input and output signal should be flat in areas where the measurement signals has significant spectral components After finished nonlinear measurement
17. al none software will correct it timing difference between reference and measured signal timing difference between reference and measured signal before correction 0 2 samples 2 5 0 1 sample samples samples timing difference between signals after correction timing difference between reference and 0 0 1 0 1 1 1 0 samples measured signal after correction 1MA27_manual 18 samples samples samples Rohde amp Schwarz AmpTune version 3 0 user manual 7 Nonlinear measurement screen During the measurement some additional information graphics will come up Before timing correction p amplitude histrogramm amplifier output T T T T T T 1 1 1 1 1 1 f 1 1 1 5000 10000 15000 20000 25000 30000 35000 40000 o 5000 10000 15000 20000 25000 30000 35000 40000 There are 4 different graphics shown 1 1MA27_manual On the top left side linear input amplitude versus linear output amplitude is shown If the values are not corrected in timing you will get some sort of cloud showing no correlation between the data On the top right side linear input amplitude versus phase difference between input and output signal is shown If the values are not corrected in timing most of the data are off screen On the bottom left side the result of the conjugate complex multiplied amplitude spectrum of input and output signal is shown It should have the same
18. ating the bandlimited noise signal dual tone After that the signal is loaded to the baseband generator clock rate and trigger lines are set and the test signal is started The last setup step is switching the RF source signal off setting reference level and input attenuation of the analyzer to get optimum levels at both mixer and A D converter and taking a measurement with no signal applied in order to check the noise floor of the analyzer to estimate the maximum measurement performance Auto leveling lt AUTO LEVELING gt set RF output level x dB below target switch source on measure output level and gain g set RF output level x dB below target consider gain g measure output level and gain g set RF output level to target level consider gain g measure target level and gain level margin The only power parameter which must be set for the DUT is the output RMS level AmpTune takes care that this output level is set with a very high accuracy This is done dureing the so called auto leveling process The auto leveling is done in 3 steps 1 To get a first very rough value of the DUT gain the output RMS level of the RF source is set to 50 dB below target RMS level With this settings the level and gain is calculated 2 Taking this first gain into account the RMS output level ot the RF source is set to 5 dB below target RMS level This second measurement gives a very good estimation of amplifier gain and output
19. blocks can hold 65 627 and Q sample pairs and every waveform does occupy a least 1 block Because of this limitations please consider this maximum file sizes in and Q sample pairs only test signal 1 441 594 samples test signal 1 predistored signal 720 797 samples test signal 2 predistored signals 458 689 samples test signal 3 predistored signals 327 635 samples All values are only valid for an empty arbitraty waveform generator memory 15 Rohde amp Schwarz 1MA27_manual AmpTune version 3 0 user manual 10 MHz Reference source Please select the instrument which shall provide the 10 MHz reference clock frequencies All other instruments used in the test setup must be connected as slave units Menu bar The menu bar is used for standard windows actions load save exit help info and settings with are not used very often Advanced settings Instrument setup optimum analyzer mixer level instrument FSIQ z 20 AmpTune will set up the internal attenuators of the analyzer for an optimum dynamic range This optimum point can be characterized by the so called mixer level For each instrument class different mixer levels are predefined and will result in best performance peak mixer level dBm If due to some special reasons this levels must be changed you can do the settings here The predefined values are 20 dBm for FSIQ 15 dBm for all other instruments Signal p
20. enerates a dual tone the spacing of the tones is set up using the signal BW If e g 2 MHz is selected 2 tones are generated at 1 MHz and 1 MHz left and right from the carrier NOTE Keep in mind that the quotient between sampling rate and tone spacing is whole numbered If it is not the signal will generate additional harmonic frequencies due to an additional rectangular component WinIQSim file It is possible to load WinIQSIM IQ signals which are stored in WinlQSIM using the multi charrier store option shown below To select a file double click at the file text and a selection window will appear to select a file to use For measurement the whole waveform is loaded into the baseband IQ signal generator but for analysis only a part of the waveform with length of the analyzer memory is taken from the beginning of the waveform Eile System Data Modulation Impairments New Ctrl N Open Settings Ctl O Save Settings As Etrl S Save for Add Multi Carrier Mixed Signal Save Mod Mapping Save Mod Filter Save Receiver Filter Save Bandpass Filter Show Report Exit NOTE To import user defined waveforms which are generated in ADS MathLab please use the free of charge software IQWizard 1 which can be downloaded from the R amp S homepage Size Limitations SMIQ B60 The arbitraty waveform generator SMIQ B60 is organized in 22 blocks each of this
21. ge between 0 7 and 0 8 All samples in this area are taken into a seperat array and represented by the mean value in x and y to forma single point 1MA27_manual 7 Rohde amp Schwarz AmpTune version 3 0 user manual Data gridding detail of range 0 7 0 8 1MA27_manual 8 Rohde amp Schwarz AmpTune version 3 0 user manual 5 Startup screen and basic setup When starting AmpTune the software will come up with this screen w Rohde amp Schwarz AmpTune Y 2 9 beta 27 from Jul 16 2002 File Setup DataManagement Utils Help FREQ Gen LEVELGen C measurement ie redaction FREQ Ana LEVELAna making measurement signal processing SIGNALBW verification setup stop am PM__ AMPM STATUS tead BASEBAND BB gt RF GENERATOR CONVERTION DEVICE UNDER RF gt BB CONVERTION amp BASEBAND TEST ANALYZER targert RMS E power value pomas 10 MHz reference band limited noise noise init I 0 signalBW 2 MHz Suite dual tone SMIG z WinlQSim file FAS LOKALWEISS AppN 1MA27 AmpTune _2info 2 The screen is splitted up into 4 main regions e The Action window which starts all setup and measurement activities searching the IEEE bus starting measurements e The Hardware window which is used to configure the hardware of the test setup and some parts of the signal processing e g which
22. nnnennnnnnnnnnnnnnnn After finished nonlinear measuremernt nneenenen 8 ACP improvement measurement 4444ssnsnneneennnnnennnennenn nn 9 Measurement results documentation nnnnnnnnenenn 10 EAO Se ee esse 11 Literature and additional free software uununeneneneneneneneneen This document describes how to operate the software AmpTune As a basic rule operating the software is done like shown below 1 Connect your instruments via GPIB bus to the computer 2 Connect reference frequencies and triggers 3 Start the software and press setup in order to connect your instruments 4 Select the type of signal you want to use Setup your generator and analyser parameters 6 Select the type of measurement measurement or measurement predistortion and which type of predistorion should be used 7 Start the measurement The software comes up with setting suitable for normal operation and demonstration of the software In order to check if everything is wirking fine simply connect the generator output to the analyzer input press SETUP and press START 2 Rohde amp Schwarz AmpTune version 3 0 user manual 2 Instrument connection 1MA27_manual Modulation Source IQ memory AMIQ or SMIQ B60 FSIQB70 or build in Trigger 10 MHz reference signal PC with IEEE card You need to take care of 5 different
23. regression calculation is carried out over the linear phase obtained by the time offset of the two measurements If two measurements x t and x2 t only show a different timing that means x t x 7 then the timing difference can be derived by 2 XN corr x x gt ANENE er gt are FFT corr x t x t 27 f T This method does not depend on the nonlinear effects of the amplifier because these effects are too small compared with the effect of timing offset The timing offset thus determined is corrected in the frequency domain so that phase locked IQ data are available by retransformation into the time domain Polynomial calculation amp data gridding If the nonlinear data are recorder and the timing correction is applied the software calculates a polynomial which fits the recorded data best Polynomial caluclation is done by using the least squares method which tries to minimiz the error sum where x are the input data and y are the output data while f discribes the fit function and a are the curve coefficients Computing this sum for the complete set of input and output data take quit long fot huge input sequences In order to minimized this time data gridding is applied to the measured data Data gridding means to collect data in between a certain range together and represent them with a single data point The 2 images below show a linear curve where data gridding is applied in the ran
24. rocessing setup Signal precessing setup AM AM AM PM polynom order amamo polynom order Combot grid points 10 grid points 10 lower limit 10 lower limit 10 T ignore 1 set linear T ignore set zero For each AM AM and AM PM curve calculation there are 5 parameters for the curve calculation e polynomial order Set the order of the polynomial representation If you only want to use k and ks use 5 here e g e grid points As shown on page 6 raw data are gridded The number of grid points can be set here e lower limit ignore and set linear For very low signal amplitude the AM AM and especially the AM PM data are very noise To avoid that the polynomial calculation is influenced too much from this points you can delete this points or set them to ideal levels 16 Rohde amp Schwarz 6 Error messages 1MA27_manual AmpTune version 3 0 user manual e Example for 50 grid points and 10 lower limit If ignore is selected the lowest 5 x y points are deleted The curve calculation is performed only on 45 points If set linear is selected the lowest 5 y points are AM AM set to the value of the 5 x points AM PM set to the 0 phase difference between input and output signal This leads to a ideal curve representation in this area AmpTune will report most of the common errors which can occure during the measurement Instruments missing ERROR missing IQ analyzer mi
25. ssing IQ generator missing RF generator If some instruments are missing an error message will pop up How to solve the problem e Check your IEEE bus cable setup e Press setup again perhaps some instruments did not respond during the first scan Critical errors Measurement aborted D Analyzer reference unlocked and 3 seconds later Measurement aborted Some errors can not be solved by the software You have to adjust your setup in order to make the program running How to solve critical errors e Check your reference connections e Check your trigger connections 17 Rohde amp Schwarz AmpTune version 3 0 user manual Non critical errors Level error 2 stage too high Continue s and 3 seconds later if No is selected Measurement aborted The software will check some parameters during the measurement If any of this parameters are out of range the user can select to abort or continue the measurement How to solve non critical errors e Check if the attenuation or the gain of the measurement path is not too high e g cable not connected amplifier voltage source off The list below shows a list of parameters to be checked and the pre definded ranges for the values Limit name male _ green yellow red crest factor input signal liveness 2 P g of complete test signal waveform RMS output level at 1 level mean output power 1 stage estimation step
26. t report The report will be displayed using the Internet Explorer lolx Datei Bearbeiten Ansicht Eavorten Extras 274 General Results Report Informations Sep 26 2003 13 42 07 a Device undertest cable Comments Instrument Setup aseband generator SMIQ B60 aseband Analyser FSQ Baseband Generator Setup Test Signal Type Bandlimited Noise clock rate 16 0 MHz baseband filter OFF Parameters Bandwidth 1 0 MHz init value 0 16384 samples Baseband Analyzer Setup True Ale 16 000 MHz sampling rate 16384 samples 10 MHz baseband filter Numeric Measurement Results RMS Level Gain 30 0 below target Bi 30 0 belowtargetlevel 1054 Fertig 133 Arbeitsplatz 1MA27_manual 22 Rohde amp Schwarz 10 FAQ s 1MA27_manual AmpTune version 3 0 user manual The AM AM curve is ok but AM PM produces only random points There is a problem with the reference frequency You are feeding in 2 reference sources on the reference line Please check the reference frequency line Both AM AM and AM PM cuve are worse There is a common reasons for this typ of error You use an FSP FSU and have select AMIQ clock rates which are not integer multiple of the sampling rate of the FSP the FSP FSU has fixed sampling rates compared with the FSIQ FSQ which can measure at any sampling rate you want to Select the AMIQ clock as an integer multiple of the FSP sampling rate The amplitude is very low about
27. typ of test signal is used e The Status Window which informs about the current actions and settings in use e The Menu bar where all advances settings take place 1MA27_manual 9 Rohde amp Schwarz AmpTune version 3 0 user manual Action window C measurement it measurement predistortion setup start stop AM PM AM PM This window is used to setup the basic operation mode of the software and starts the main measurement tasks e SETUP By pressing this button the program will scan the IEEE bus and fill the list of available lists in the hardware window The selection lists for available sampling rates memory depth will be set e START This button will start the main measurement task e STOP This button will stop the current measurement task NOTE The software does not check the button all the time It will perform some actions and check the button status from time to time So by pressing the STOP button the software will not break at the same time but run some seconds before it stops e MEASUREMENT MEASUREMENT PREDISTORTION This radio buttons switch between the 2 main measurement methods MEASUREMENT will only record the AM AM and AM PM curves MEASUREMENT PREDISTORTION will record the AM AM and AM PM curves predistore the signal and test the resulting waveform for improved ACP performance You can switch the different predistortion technics only AM AM only AM PM combined AM AM
Download Pdf Manuals
Related Search