Home

Agilent J-BERT N4903B High-Performance Serial BERT

1. Sample Count 1 Level Q Level Rise Time FallTime Eye Amplitude Eye Height Eye Width Jitter Peak Peak Jitter RMS Cross Voltage 3 177007E408 300 25m v 239 99m 25 63mUI 31 30mUI 600 25mwV 513 11m V 381 50mUI 8 48mUI 503 66uUl 50 11 al gt Pattern Generator im m Error Detector m E Cm 300000Gb s 272 1PRes o Mer Outputs anon Gb s 22531PRBs Enor func Data Clock This is the eye of a non distorted pattern Problems with the N4916B Problems with the N4916B Concepts When you have opened the External Instrument s Config window and have selected N4916B from the external instruments select De emphasis under Enable Function column in the list the N4916B Connection window appears 500 Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 Cannot Connect Problems An error message may appear when you now press Enable If this happens ensure that the N4916B has been correctly installed see Installing the N4916A B Procedure on page 539 Check the Power Supply Ensure that the power cord is connected The green LED close to the power switch must be illuminated when the unit is turned on Check the USB Communication Ensure that the USB cable is connected between the USB port at the rear of the N4916B and one of the USB ports at the rear of the Serial BERT 1 Open the Utility menu and click Minimize GUI to gain access to the Windows operating system 2 In the Window
2. Click the Show Error Messages button to open an Error Message dialog box with all error messages ErrorMessageForm 10 7 2009 4 11 42 PM 80814 The value Hz exceeds the physical limits of the Sinusoidal Jitter Frequency Clear amp Close Close Help On this dialog click Clear amp Close button to clear the error messages list and quit the dialog box Click the Close button to leave the dialog without clearing the error messages list This dialog box also contains a Help button that takes you to the respective section of the Online Help Elapsed Indicator This area displays the progress of the current accumulation period The accumulation period can be changed in the Accumulation Setup dialog box For more information see How to Run Accumulated Measurements on page 377 Elapsed 00 00 00 Pattern Types The PG and ED pattern types are displayed here for your reference Agilent J BERT N4903B High Performance Serial BERT 375 7 Evaluating Results Clock Rates The PG and ED measured bit rates are displayed here for your reference Accumulated Measurements Accumulated Measurements Concepts NOTE Measurement Period 376 Accumulation refers to collecting measurement data over a specific period This can be used to create test scenarios that are reproducible and comparable Also you can let tests run over long times and then evaluate the results afterwards Furthermore you can use the log
3. Clock speed Gb Ethernet 1 06250Gb s See Setting up the Pattern Generator Concepts on page 121 for more information Set up the error detector so that the input range and the termination matches the pattern generator s levels Select Normal as the Active Input an Input Range of 1 3V and set the Data Termination to 1 3V Enable the pattern generator outputs by pressing the OV Disable button Also make sure that the error detector derives its clock from the incoming data stream using Clock Data Recovery See Clock Setup Procedures on page 178 for details Agilent J BERT N4903B High Performance Serial BERT 351 6 Advanced Analysis 352 9 Press Sync Now and then Auto Align to find the optimum sampling point Check that the synchronization and the alignment were successful None of the error indicators should show red and the resulting BER should be zero How to Execute the Spectral Jitter Measurement To run the Spectral Jitter measurement 1 Switch to the Analysis area If the Spectral Jitter screen is not yet displayed press the Spectral Jitter icon 2 Press the Start button to execute the measurement The measurement software runs the measurement and displays the results A total of 65535 points are displayed for this measurement Numerical values are also displayed The values include Bit Error Rate Total Power and Noise power Additionally you can also configure up to 16 frequency power pairs The l
4. Accumulated Results In addition to the values displayed in the BER Accumulated Results see Accumulated Results on page 380 the following additional value is listed Auto Resync Counter The total number of Auto Re Sync Count received in a time interval is displayed here Agilent J BERT N4903B High Performance Serial BERT 391 7 Evaluating Results Accumulation Parameters for Bit Comparison without USB 3 1 SKPOS For detailed information on Bit Comparison without USB3 1 SKPOS accumulation parameters see Accumulation Parameters on page 382 Eye Measurements Eye Measurements Concepts The purpose of eye measurements is to measure the eye height and width of the incoming data signal at specific alignment BER thresholds This information is displayed on a representation of an eye diagram Eye Measurements Procedures Eye measurements are made each time the Auto Align Clock Data Center or 0 1 Threshold Center functions are used To analyze the results 1 View the graph and table in the Eye Results window This window provides detailed information on the calculated eye See Eye Measurements Reference on page 393 2 Analyze the results by observing changes in the eye results due to the following adjustments Make changes to your device or measurement setup Run auto search functions at different BER thresholds This will allow you to see the eye margin at specific BERs and construct basic eye
5. Continuous A Select this option to send out only pattern A repeatedly DataOut A AJAJA A AJ A AIA A AJ A A0 A1 A2 Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 Continuous B Select this option to send out only pattern B repeatedly BO B1 B2 Data Out B B B B B B B B B B B B Alternate AB Select this option to alternately send out patterns A and B A B A B A0 BO A0 BO Data Out A AJAJAIBIBIBIBIAIA A A B B B B Single Shot B Select this option in the Alternate Pattern Control dialog box to enable the Insert B button on the main display Insert B j When you press the Insert B button pattern B is inserted into the continuous output of pattern A Insert B pressed A0 Ad BO A0 Data Out AJA AJAJAJAIAJAIB B B B A JAJ JA Aux In In this section you can enable the Aux In port to control the output pattern Aux In can receive signals that control alternate patterns or blank the output The signals received at this port must be TTL compatible The granularity for signals at this port is 512 bits The following options are available for Aux In Disable With this option selected the signal at the Aux In port will be ignored Level Sensitiv
6. 1 Disable the outputs of the N4903B and use the matched cable kit M8061 61601 to connect the pattern generator to the M8061A Make sure to connect Data In 1 Data In 2 and Aux CIk In ports of M8061A to Aux Data Out Data Out and Aux Clk Out ports of N4903B respectively All the respective ports are present at the front side of the instruments Do not loop in any other equipment like N4903B s Option J20 Mountthe SMA 50 Ohm terminations on the unused DATA OUT and AUX DATA OUT ports of the pattern generator using the SMA to 2 4 mm adapters Enable the M8061A as described in How to Enable Disable M8061A Multiplexer with De emphasis Function on page 49 and set the levels such that the DUT will not be damaged Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 4 Connectthe differential outputs of the M8061A to the DUT If only one output of the M8061A is needed connect a 2 4 mm 50 Ohm termination to the unused output Installing Webserver Customizing the Web Server Concept What is a Web Server A web server is a program that operates by accepting HTTP requests from the network and providing an HTTP response to the requester Enabling a web server on an instrument allows it to accept and respond to HTTP requests from clients The web server is an industry standard for providing information and functionality of an instrument over LAN The web enabled instruments are based on two technologie
7. Agilent J BERT N4903B High Performance Serial BERT 123 4 Setting up the Pattern Generator 124 NOTE Understanding the Output Protection Circuit The generator module offers a huge flexibility for external termination schemes and external termination voltages to address common technologies For details please refer to the Technical Data Sheet An internal protection circuit continuously monitors the voltages of clock data aux data and trigger output It becomes active if the termination voltage is wrongly adjusted The output protection circuit may become active if the output voltage Vout becomes higher than Vhi 0 5V or lower than V 0 5V The output protection circuit is shared among two output drivers each Data Out and Clock Out as well as Aux Data Out and Trigger Ref Clock Out are supervised by one circuit together As a consequence if an output voltage violation occurs both outputs are disabled together for a short period of time If an output voltage violation occurs the output voltages Vhi Vlo and Vterm are re programmed to safe values typically to the externally measured termination voltage Afterwards the output drivers are re enabled The GUI shows an error message listing the effected outputs and highlights the Outputs On status indicator Fa Dutputs ON In the corresponding Output screen of the PG Setup the output settings that currently don t match the electrical output settings are highlighted in ye
8. BER Threshold To calculate the parameters for the given BER threshold This is the BER level for which output timing numerical values phase margin skew etc are calculated It is also the upper limit of the BER range for RJ DJ separation The BER threshold influences some of the parameters of the DUT Output Timing measurement You can also drag and drop the horizontal BER threshold in the graphical display to change this value Min BER for RJ DJ Separation Lower limit of the BER range for RJ DJ separation Residual BER for Estimated Total Jitter BER level for which the estimated total jitter is calculated See Estimated Total Jitter on page 245 for details on how it is used You can select the number of Decimal Places to be displayed in the table Graph Tab On the Graph tab you can use the several options to optimize the graphical display according to your needs Choose between Unit Interval and Seconds to select the timebase for the display s x axis Choose between Logarithmic and Linear to select the scale for the display s y axis For example a DUT Output Timing measurement displayed on a linear scale may look like this 6 00e 001 1Trace 207 Points Ul 2 000 ns All Errors 5 50e 001 5 00e 001 4 50e 001 4 00e 001 3 50e 001 3 00e 001 2 50e 001 2 00e 001 1 50e 001 100s 001 2 7 e e pee erem erm mm emm oe eee mse eet dee ew me mni 5 00e 002 a ALE E gt gt gt Linear 0 50 UI 0 30
9. Every sequence can be stored in a file and recalled from that file However the SequenceExpression does not include user patterns Recalling a sequence from its file may fail if the referenced user patterns are not available A recalled sequence is not automatically downloaded to the pattern generator n sequence mode the function Error Dector Pattern tracks the Pattern Generator Pattern is automatically disabled You have to download the expected payload pattern to the error detector During initialization of the sequence pure zeros are sent User Defined Sequences Procedures You can create a new sequence save it in a file or load a saved sequence from its file Creating a New Sequence If you wish to create a new sequence 1 Click the Sequence Editor menu item from the Pattern submenu E Sequence E ditor By default the Sequence Editor shows one block that is automatically repeated It has a length of 512 bits and generates Pause 0 data pure zeros 2 Click the Properties icon In the Properties window set the sequence start condition the number of blocks contained in the sequence 1 to 4 3 Click Ok to close the Properties window 4 Specify the data to be generated during execution of a block Click the Edit button of the block Choose from the list For example you can select a PRBS polynomial or browse for a pattern file A None block is completely ignored f necessary cha
10. If the error detector loses synchronization it keeps counting bit symbol frame errors and generally the measurement won t reflect the DUT s performance Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 anymore Also test results would not correlate with other error counters such as protocol testers anymore The high error ratio also serves as a trigger to perform an automatic re synchronization So on one hand automatic pattern re synchronization is desired during receiver testing but it will be necessary to mask respectively do not account the high error count which was caused by the not synchronized error detector into the measurement Side by side comparisons with other error counters show that the loss of synchronization is usually caused by only one single bit error Error counting is performed in 200ms intervals If the error ratio triggers a re synchronization of the error detector the current interval must not account into the results Furthermore it is very likely that the previous 200ms interval s also showed many errors due to lost synchronization but just did not raise the error ratio to the specified trigger level In later case we check for errors in 201 ms and if the errors are more than threshold we discard these errors The measurement pauses until the ED successfully reached pattern sync again Furthermore counters must be updated The Re Synchronization counter is increment
11. Jitter Type The standards demand and the measurement generates jitter with sinusoidal distribution In terms of the Serial BERT sinusoidal jitter SJ and periodic jitter PJ with sinusoidal characteristic are used for the test see also How the N4903 Generates Jitter on page 401 Whether SJ or PJ is used depends on the jitter frequency SJ supports wide amplitude variations up to some MHz PJ supports narrow amplitude variations over a wide frequency range The measurement switches automatically between the two sources The measurement cannot be executed if jitter generation is totally disabled SJ and PJ may be enabled or disabled NOTE SJ and SSC are mutually exclusive to rSSC so you can have SJ and SSC both on but rSSC only if neither SSC nor SJ are on If you run a Jitter Tolerance Measurement while rSSC is active PJ will be used to perform the measurement If desired you can add additional jitter of random external periodic 2 or bounded uncorrelated origin to the automatically generated jitter When the measurement has finished the jitter composition shown in the Jitter Setup window is restored Target BER The target bit error ratio must be specified The condition for moving from one jitter amplitude step to the next can be set to Agilent J BERT N4903B High Performance Serial BERT 449 8 Jitter Tolerance Tests 450 numbers of received bits and errors aconfidence level Additionally a relax time
12. Time Select this option to configure the error detector to accumulate bit errors for a specific period of time You can enter the desired time period in the Days Hr Hours Mn Minutes and Sec Seconds fields You may use the numeric keypad or front panel knob The minimum value is 1 second and the maximum value is 99 days 23 hours 59 minutes and 59 seconds When the selected time has elapsed the accumulation ends Number of Errors Select this option to configure the error detector to accumulate bit errors until a specified number of errors has been measured The number of errors can be set to 10 100 or 1000 When the selected number of errors has occurred the accumulation ends after the next full second Number of Bits Select this option to configure the error detector to accumulate bit errors until the specified number of bits have been examined The number of bits can be set from 1E7 to TE15 When the selected number of bits have been exceeded the accumulation ends NOTE The actual test period may contain more bits than specified NOTE The accumulation period cannot be changed while accumulation is progress Browse Click this button to browse for the directory where you wish to save the measurement log files NOTE Changes to the parameters of the accumulation setup do not take effect until the start of the next accumulation 222 Agilent J BERT N4903B High Performance Serial BERT Setting up the Er
13. 0o5T 0 3 1 olt a no 1 2 3 4 56 7 8 9 Delay ns Example Results The following figure shows the graphical result of a typical Eye Opening measurement 1500 mv Trace 9211 Points Ul 1 538 ns All Errors n 1 000e 0 VERANO S ROCs OUP 1200mv POSES READY De RUE NGS GO EG AUS E eet SESS DI PAE oS Mu aes 1 000e 1 900mv ES nd 60 0 mv i 1 000e 2 30 0 mv 1 000e 3 00v sordo 1 000e 4 60 0 mv 1 000e 5 90 0 mv f i A h rA 6 A EN asom EXENA E E MEL oos scale 0 850Ul 0 390Ul 0130Ul 04130UI 0 390 Ul 0 650 UI Relative 0 520Ul 0 26001 Q 000UI 0 260 Ul 0 520 UI Eye Opening Procedures This section shows how to set up and perform an Eye Opening measurement As an example we measure the eye diagram of a shielded cable This requires the following steps Preparing the measurement see How to Prepare the Eye Opening Measurement on page 290 Executing the measurement see How to Execute the Eye Opening Measurement on page 291 Agilent J BERT N4903B High Performance Serial BERT 289 6 Advanced Analysis 290 Optimizing the view of the results see How to Optimize the View of the Results on page 291 Using the color bar see How to Use the Color Bar on page 292 Adding or changing colors see How to Add or Change Colors on page 292 Changing the BER threshold see How to Change the BER Threshold on page 293 Changing the BER range of a color see How to Change the BE
14. 4 Not instaled pelay 33 096 ns 60v E Input Timing Setup 2 E File Control Setup Measure Calibrate Utilities Help 24 Mar 2011 11 56 LZ i 100 mV div 41 6 mVv div Time 28 2 ps div Trig Normal 00V 2 00 uv 2 not Installed Jno Installed Delay 32 0096 ns E l a Input Timing Setup 3 486 Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 File Control Setup Measure Calibrate Utilities Help 21 Mar 2011 12 02 LZ 100 mv div 41 6 mVv div Time 28 2 ps div Trig Normal ooy 2 oo v 3 Notnstaled 4 Not instaled pelay 33 096 ns 00V E Input Timing Setup 4 TU File Control Setup Measure Calibrate Utilities Help 21 Mar 2011 12 09 LZ i 100 mV div 41 6 mVv div Time 28 2 ps div Trig Normal 00V 2 300 uv 2 not Installed not Installed Delay 32 0096 ns ooyv l a Input Timing Setup 5 Agilent J BERT N4903B High Performance Serial BERT 487 9 Solving Problems File Control Setup Measure Calibrate Utilities Help 21Mar2011 12 15 LZ Pattem Lock 100 mVv div 41 5 mVv div Time 28 2 ps div Trig Normal ooy 2 oo v 3 Notnstaled 4 Not instaled pagys noons 00V E Input Timing Setup 6 File Control Setup Measure Calibrate Utilities Help 21 Mar 2011 1221 LZ 4 100 mV div 0 0 Vv i Input Timing Setup 1 shows timing violations at the input of the N4916B All other measurements do not show any violations during the measuremen
15. 50 96 min 100 0 ps div div 238 mU Start 1 5 Periods In detail the measurement is not that simple and more precise as it derives the contour of the eye from the bit error rates that have been measured Three Available Views The Eye Diagram can be visualized in three different ways Contour Plot Agilent J BERT N4903B High Performance Serial BERT 287 6 Advanced Analysis The contour plot shows discrete lines of equal bit error rate just like the contour lines on a map The color of a line indicates the respective BER value This graph is useful to visualize in which areas the BER changes a homogeneous BER field will give you no lines at all 154 epe HE i P aX i sub AAS w 114 lif bf aan JV ri 6d E 09 j j n _ dii s AN NS Ys ff ost WO ee ut Ee es O3 u oit 0 1 234 5067 9 Delay ns Pseudo Color Plot This plot visualizes the BER by a continuous color gradient It uses different colors for the regions between the lines of equal BER This is useful to get an immediate visual impression of the distribution of the BER in the eye diagram L7 LST 1 31 Threshold V t I t 0 123 3 4 5 6 8 9 Delay ns Equal BER at BER threshold This graph displays only one curve for the chosen bit error rate threshold 288 Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Voltage V
16. Bit Comparison without USB3 1 SKPOS results on the screen The 8B 10B Comparison includes the following measurements Symbol Error Ratio SER Agilent Technologies 361 7 Evaluating Results 362 Frame Error Ratio FER e Calculated Bit Error Ratio cBER Filler Symbol Ratio FSR Disparity Error Ratio DER e Illegal Symbol Ratio ISR In order to stress your device under test you can then for example add errors to the data stream or switch between different patterns and view the resulting BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS Bit Comparison without USB3 1 SKPOS Or you can adjust the output port parameters or the error detector s sampling point while the measurement is running and watch the resulting BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS Bit Comparison without USB3 1 SKPOS immediately Such measurements can be used to find out the benchmark data for example for a device prototype Test Methods You can analyze the behavior of your device by employing several test methods If the changes in the test setup affect the BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS Bit Comparison without USB3 1 SKPOS you can immediately notice it on the screen The resulting changes in the BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS Bit Comparison without USB3 1 SKPOS let you learn about the DUT s behavior and
17. ERRORED 1 COUNT Cumulative ERRORED 1 RATIO Cumulative ERROR SECONDS Cumulative ERROR FREE SECONDS Cumulative ERROR DECISECONDS Cumulative ERROR FREE DECISECONDS Cumulative SYNC LOSS SECONDS Cumulative BURST STATUS Cumulative BURST SYNC RATIO Cumulative TOTAL BURST COUNT Cumulative BAD BURST COUNT Cumulative 477 9 Solving Problems Other Messages 478 Other Messages Concepts Caution Level NOTE Critical Temperature This section covers additional error messages that may occure when working with the Serial BERT Overheat Protection The Serial BERT is equipped with an overheat protection function to prevent it from overheat damage If the temperature of the error detector or pattern generator exceeds a certain threshold the Serial BERT displays the following warning message respectively The temperature level is not yet critical However the Serial BERT may be damaged if operation is continued The message is intended to inform you to take the appropriate measures Save your current instrument settings shut down the Serial BERT and let it cool down before operating it again If the Serial BERT is operated at this state for a longer time the temperature may reach a critical level In this case the following message is displayed Temperature Alert ED critical above normal Shutting down instrument Immediate startup not recommended Agilent J BERT N4903B High Performance Serial BE
18. FBDIMM1_3 2G_RegularTxRev0 85 10GbEthernet_10_3125 FBDIMM1_3 2G_RxRev0 85 Desktop 5 10xFiberChannel FBDIMMI 3 2G SmalTxRev0 85 10xGbEthernet FBDIMMI 4 0G LargeTxRev0 85 gt 10xGbEthernet_12_5 gt FBDIMM1_4 0G_RegularTxRev0 85 Cei amp GlrNeTxToLrRx F amp DIMMI 4 0G RxRev0 85 Cei6GlrNeTxToSrRx F amp DIMMI 4 0G SmallTxRev0 85 x 3 Cei amp GsrFeRx amp FBDIMM1 4 8G LargeTxRev0 85 Cei GsrNeTx FBDIMM1_4 8G_RegularTxRev0 85 gt Ceit 1GlrMrTx FBDIMM1 4 8G RxRevO 85 gt Ceil 1GmrRxCJ FEDIMM1_4 8G_SmallTxRev0 85 e ili P My Network Filename Places Files of type Mask Files msk My Documents My Computer 4 Select a Mask Alignment Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 M Mask Alignment Display align to Vtop Vbase Eye boundaries align to 1 Level 5 Start m Mask Run Control Start Exit This starts the Mask Measurement You also have options for setting Mask Margins and Mask Scaling 6 Click the mask readout button on the bottom left of the screen to either see or close the detailed mask readout a How to Customize Mask Files A Mask file can be modified as per requirements Following are the sections in a mask file Mask File Identifier The mask file identifier is the first line in a mask file If the identifier string is missin
19. How to Enable the Bit Recovery Mode The bit recovery mode BRM allows you to measure the relative BER without knowing the incoming data pattern To enable the bit recovery mode 1 Click Sampling Point Setup menu item from the ED Setup submenu Agilent J BERT N4903B High Performance Serial BERT 197 5 Setting up the Error Detector 198 2 Click the Bit Recovery Mode button The Sync Now button is no longer available Note that you can also enable the bit recovery mode from the Pattern window Click the Pattern Select icon and then the Error Detector Pattern tab If this tab is not available disable the checkbox Error Detector Pattern tracks the Pattern Generator Pattern 3 Click the Auto Align button Sampling Point Setup Reference NOTE NOTE Depending on the options of your Serial BERT some of the following functions may not be valid for your instrument See Introduction to the Serial BERT Concepts on page 11 and Which Features are Available on page 14 for a description of the available options The Sampling Point Setup window provides the following elements 0 1 Threshold The 0 1 decision threshold is the reference voltage level for the evaluation of the incoming data signal At the sampling point the data signal voltage ist compared to the 0 1 decision threshold Signals greater than this threshold will be measured as a logic 1 signals below this threshold will be measured as a logic 0 0 1 Threshold
20. Measures the mean transition time taken by the data on the rising edge of the eye diagram The data crosses three thresholds lower crossover point upper and the eye transition Note Rise time is dependent on the Transition Time of the View Tab The two options are 10 90 and 20 80 Measures the mean transition time taken by the data on the falling edge of the eye diagram The data crosses three thresholds lower crossover point upper and the eye transition Note Fall time is dependent on the Transition Time of the View Tab The two options are 10 90 and 20 80 The number of points measured for the eye diagram If persistence infinite is enabled the value will gradually increase 337 6 Advanced Analysis Table 31 S No Parameter Description 4 0 Level Zero level is the measurement of the mean value of the logical 0 of an eye diagram Note This measurementis made in a section of the eye referred to as the eye window boundaries The default value for the NRZ eye window is the central 20 of the bit period You can change the eye window boundary settings in the Configure Measurement dialog box These settings determine what portion of the eye will be measured 5 1 level One Level is a measure of the mean value of the logical 1 of an eye diagram Note This measurementis made in a section of the eye referred to as the eye window boundaries The default value for the NRZ eye window is
21. Procedure 108 Sequencer Pattern Editor Reference 110 4 Setting up the Pattern Generator Setting up the Pattern Generator Concepts 121 Input and Output Ports 122 Input and Output Ports Concepts 122 Input and Output Ports Procedures 131 Input and Output Ports Reference 135 Bit Rate 139 Bit Rate Concepts 139 Bit Rate Procedures 141 Bit Rate Reference 143 Trigger Ref Clock Output 145 Trigger Ref Clock Output Concepts 145 Trigger Ref Clock Output Procedures 145 Trigger Ref Clock Output Reference 146 Aux Data Out 149 Aux Data Output Concepts 149 Aux Data Out Procedures 149 Aux Data Output Reference 154 Delay Control Input 155 Delay Control Input Concepts 155 Delay Control Input Procedures 156 Delay Control Input Reference 156 Error Addition Insertion 157 Error Addition Insertion Concepts 157 Error Addition Insertion Procedures 157 Error Addition Insertion Reference 159 Pattern Alternation 161 Pattern Alternation Concepts 161 Pattern Alternation Procedures 162 Pattern Alternation Reference 164 Agilent J BERT N4903B High Performance Serial BERT 5 Setting up the Error Detector Setting up the Error Detector Concepts 167 Inputs and Outputs 168 Inputs and Outputs Concepts 168 Inputs and Outputs Procedures 171 Data Input Setup Reference 172 Clock Setup 173 Clock Setup Concepts 173 Clock Setup Procedures 178 Clock Setup Reference 1
22. Reference The Audio dialog box contains the following elements Audio on Select this checkbox to enable the audio function Clear this checkbox if you want to switch off the audio function You can also use the Audio On Off button on the front panel to accomplish the same function Agilent J BERT N4903B High Performance Serial BERT 225 5 Setting up the Error Detector 226 TIP Main Volume Click any point on the slider or drag it to change the main volume level You will hear a test tone at the new volume level Audio on BER Alarm Click this option to make the analyzer play warning tones when a specific BER is exceeded The BER Alarm Threshold is the threshold at which tones are produced Enter the BER threshold in the following format number E exponent For example 1 5E 3 You can always adjust the BER Alarm Threshold by clicking and dragging the small yellow marker in the BER bar in the top left corner Tones on All Error Rates Click this option to make the analyzer produce warning tones when any BER greater than zero is measured The pitch will increase or decrease as the BER increases or decreases Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis Advanced Analysis Concepts DUT Output Timing Jitter Output Levels Eye Opening Error Location Capture The Serial BERT offers several different kinds of advanced measurements for various purposes This type of measurement
23. Sampling Point Setup Procedures NOTE In most cases you will set up the sampling point automatically Even if you wish to make some manual adjustments it is recommended to start with automatically aligning the sampling point Before adjusting the sampling point make sure the pattern synchronization mode is set up properly See Pattern Synchronization Procedures on page 216 for details How to Set the Optimum Sampling Point Automatically 1 Click Sampling Point Setup menu item from the ED Setup submenu 2 Select an appropriate BER threshold from the BER Threshold list 3 Press Auto Align to automatically set the optimum sampling point You can stop the search for the optimum sampling point at any time by pressing Cancel How to Adjust the Data Input Delay 1 Click Sampling Point Setup menu item from the ED Setup submenu 2 Selectthe Avg 0 1Threshold checkbox to use a good start point on the vertical voltage axis 3 Click Data Center to move the sampling point horizontally towards the center of the eye at the selected voltage You may click Cancel at any time 4 Now you can use the numeric keypad or front panel knob to manually adjust the value in the Data Delay field Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 The Sample Point in the eye diagram moves horizontally as you change the value TIP While adjusting the data input delay you can monitor the BER bar on the analyzer
24. Sequence Trigger at Block Begin After downloading the sequence to the pattern generator you can also change the behavior of the Trigger Out port for details see Sequence Trigger on page 147 In the example above the Trigger Out port has been set to Sequence Trigger mode and TrigOn is set for each block as shown by the vertical trigger indicator Edit Button of a Block To change the contents of a block click its Edit button This opens the Set Data Block Parameter dialog Sequence Block Parameters The Set Data Block Parameter dialog allows you to change the contents and the trigger generation of a sequence block If you select the Second Channel mode from the PG Aux Data Output dialog the same parameters for the Aux Data channel will also appear Choices are None The block may be present in the sequence but is completely ignored e Pause0 Pure zeros are generated when the block is executed You can change the block length if desired Pause1 Pure ones are generated when the block is executed You can change the block length if desired Block Marker The block starts with 256 1s and Os until block length You can use itto mimic the sequence marker functionality of the trigger output PRBS You can choose a PRBS of polynomial 2 n 1 The range of n is 7 10 11 15 23 or 31 You can change the block length if desired Divided Clock You can use this option to generate the signal at every nth clo
25. Set the Termination to 0 V Agilent J BERT N4903B High Performance Serial BERT 497 9 Solving Problems 498 4 5 ED Input Setup Dialog Input r Input Range 1 000 V 1 000 V Threshold Termination 7 T Data Inverted iv Averaging Iv DC coupled 0 mV 0 mY Ok Cancel Apply Help Connect the Output of the N4916A to the Data Input of the error detector Ensure that the error detector follows the generated pattern Press the Auto Align button From the Analysis panel select the Eye Diagram page Press the Start button The eye diagram should look like this after some seconds BER 0 000 E Time 0 20U1 Div Delay 0 98U1 c3le 147 11mV Div Offset Smiv Sample Count 1 Level Q Level Rise Time Fall Time Eye Amplitude Eye Height Eye Width Jitter Peak Peak Jitter RMS Cross Voltage 3 177007E408 300 25m 239 99m 25 63mUI 31 30mUI 600 25m 513 11m V 381 50mUI 8 48mUI 503 66uUl 50 11 4 Pattern Generator m Error Detector ww wm m Clock Jitter Outputs 3 00000Gb s 2723 1 PRBS o dier Oututs nanny Gbys 2 23 PRBS Enor pene Data Clock Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 Note that a de emphasis value of 6 dB causes a 50 percent amplitude reduction Check the Output of the Pattern Generator with a Scope If the N4916A does not produce a signal check the Data Out of the Serial BERT
26. This ensures that your DUT cannot be damaged due to incorrect terminations Correct clock frequency Required to recognize the bit rate in the data stream Appropriate sampling point The sampling point defines where the error detector tries to differentiate between Os and 1s in the data stream This is necessary so that the error detector recognizes the data bits correctly Synchronization to the incoming pattern The expected pattern must be synchronized to the incoming pattern so that the error detector can find any discrepancies The error detector provides the following functions to enable you to perform tests Automatic pattern synchronization The error detector shifts the incoming data stream bitwise to match it to the expected data pattern A correct BER can only be measured with matching patterns Error accumulation You can specify whether a test runs for a specified time or until a specific number of errors has occurred This lets you carry out longer tests while logging the results to a file Agilent Technologies 167 5 Setting up the Error Detector NOTE Inputs and Outputs 168 BERlocation mode You can specify whether all errors are counted or only the errors that occurred on a particular bit position or range of positions Audio warnings You can set up the audio warnings so that the instrument beeps when a certain BER is exceeded Trigger output for external measurement instruments This
27. r Pattern Type Binan cH C Symhbc Next Replace Replace All Done Help The search pattern can be entered in binary hex symbol format depending on currently selected mode in pattern editor You can continue editing the pattern while this dialog box is still open Agilent J BERT N4903B High Performance Serial BERT Next NOTE Replace Replace All Done Help Setting up Patterns 3 In addition The Find Pattern Segment dialog box also allows you to replace the searched pattern segment with the desired pattern segment Looks for the next occurrence of the pattern from the current position to the end If there is no further occurrence the current selection does not change Make sure the search patterns entered in symbol format are seperated by one white space Click this button to replace the searched pattern segment with the desired pattern segment Click this button to replace all occurrences of the searched pattern segment with the desired pattern segment Click this button to close the dialog box Takes you to the respective section of the Online Help Block Edit Configuration Dialog Box The Block Edit Configuration dialog box provides an easy way to modify parts of the pattern or the entire pattern at once This can be used when setting up a new pattern It can also be used as an optional technique for editing existing patterns The Block Edit Configuration dialog box is shown in the figure be
28. 0 INSTR ed ien Agilent Technologies N4876A Proto_004 0 1 1 0 1 d N4876A USBO Manufacturer Agilent Technologies ey N4876A Model code N4876A VISA address IDN string Serial number Proto 004 Firmware D 1 1 0 1 myinterface KZ MyInstrument Eh Myalias 32 bit Agilent VISA is the primary VISA library oy For detailed instructions refer to the Connectivity Guide which is part of the Agilent 10 Libraries Suite Documentation Connecting to Pattern Generator and DUT After the USB port has been configured you can remove keyboard and mouse and make the signal connections 1 Disable the outputs of N4903B and use the matched cable kit 2x 2 4 mm to SMA for data 1x SMA to SMA for clock with the part number N4915A 001 to 548 Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 connect the pattern generator to the N4876A Make sure to connect Data Out to Data In Aux Data Out to Aux Data In and Aux Clk Out to Aux Clk In 2 Mount the SMA 50 Ohm terminations on the unused Data Out and Aux Data Out ports of the pattern generator using the SMA to 2 4 mm adapters 3 Enable the N4876A as described in How to Enable Disable N4876A Multiplexer Function on page 47 and set the levels such that the DUT will not be damaged 4 Connect the differential outputs of the N4876A to the DUT If only one output of the N4876A is needed connect a 2 4 mm 50 Ohm termination to the unused output
29. 1 00e 3 1 00e 4 1 00e 5 1 00e 5 BER Thresholg 1 00e 7 Logarithmic 9 60 UI 0 36 UI 012UI 0 12UI 0 36 UI 0 60 Ul Relative 0 48 UI 0 24 UI 0 00 UI 0 24 UI 0 48 Ul 0 72 UI Explanation of the Numerical Results The measurement provides numerical results for Output Timing Measurement The output timing measurement parameters are defined in the following table Parameters Agilent J BERT N4903B High Performance Serial BERT 259 6 Advanced Analysis Jitter Measurement Parameters 260 NOTE NOTE Table 26 Parameter Description Definition Optimal Sample Point The average of the left A A B 2 Delay and right B bathtub BER threshold intersections Phase Margin The period of time where B A the bit error rate is lower than the BER threshold The A and B values are the left and right intersections of the bathtub curves with the BER threshold Obviously all values change if the BER threshold is modified The following illustration shows an example for a measurement I linear log scale 1 Optimum Sampling Point 0 rae ERR ERD E A TTT Pera oe Peps rare ea This graphical view is not available in the measurement software The jitter measurement parameters are defined in the following list All jitter measurement parameters except the Total Jitter RMS and Total Jitter Mean change with the BER threshold Total Jitter RMS The average of the left and right jitter histogram root mean squared
30. 173 Instantaneous Measurements 361 Interference Channel 402 Setup 431 Internal Clock Source 143 Intersymbol Interference 432 Intersymbol interference ISI 399 Interval Results 382 Introduction to the Serial BERT 11 invalid 261 Inverted Data Error Detector 172 Pattern Generator 137 ISI 399 ISR Results Window 371 ISR Status Indicators 373 J Jitter example 239 fast total 246 Generation Block Diagram 401 Total 400 Types of 397 Jitter Mean 261 Jitter Setup 404 Jitter Setup Area 415 Jitter Tolerance Reference 443 Jitter Tolerance Characterization 433 BER 435 444 Confidence level 435 Frequency 443 Relax time 435 Search 444 Search methods 436 Jitter Tolerance Compliance 446 BER 465 Confidence level 450 Frequency 464 Margin 453 Predefined Standards 451 Reference 462 Relax time 450 Standard Specification 462 Target BER 449 User defined Standards 452 Jitter Tolerance Tests Basics 395 L LAN Connectors 530 Level Results 277 Linear Markers 236 Loading Pattern 73 Pattern directly from Pattern Editor 75 Logic Families 128 Logic Levels Setup 131 loop bandwidth 181 Low Level 278 Low Level Std Dev 279 M8061A Connections 35 Installation 549 Manual Pattern Alternation 163 Manually Inserting Errors 157 Mark Density 92 Agilent J BERT N4903B High Performance Serial BERT Index Markers Gaussian 237 introduction 236 Linear
31. BER vs Threshold Graph dBER vs Threshold Graph If you compare the Low Level result with the limit we have set on the Pass Fail tab you will find that the measured result fails the upper pass fail limit for this parameter View Tah The graph shows either the BER vs Threshold the dBER vs Threshold or the QBER vs Threshold This graph shows the relationship between the analyzer decision threshold and the measured BER 800 0 my Trace 501 Points 900 0 mv 40v AAV 42V 43V 44v 45V AV ATV 48V BER Threshold 1 907e 6 Logarithmic 1 000e 7 1 000e 5 1 000e 3 1 000e 1 1 000e 6 1 000e 4 1 000e 2 1 000e The BER considers all errors It is calculated as Y Error ls SErrorOs total of Bits BER aigsrors This graph shows the relationship between the analyzer decision threshold and the absolute values of the derivative of the bit error rate BER dTh It is recommended to view this graph with a linear scale because a linear scale reveals the distribution more clearly 1 Trace 500 Points 800 0 mv 900 0 mv 40V aav 42V 43V 44V 45V ABV ATV 48V Linear 2 000e 0 6 000e 0 1 000e 1 1400e 1 1 800e 1 2200e 1 2 600e 1 3 000e 1 0 000e 0 4 000e 0 8 000e 0 1200e 1 1 600e 1 2000e 1 2400e 1 2800e 1 3200e 1 This distribution can often be approximated by a Gaussian normal distribution With a Gaussian marker you can measure the mean and standard deviation of the nor
32. CEI 6 Gb s Long Reach CEI 11 Gb s Short Reach Ingress Telecom CEI 11 Gb s Short Reach Egress Telecom CEI 11Gb s Short Reach Ingress Datacom XAUI 10G Base R 10G Base W XT Add Jitter RJ 10 7 mUl rms BUJ 150 mUl pp ISI 300 mUl recom Trace 2 350 mUl zi OK Cancel Apply Help The list contains all predefined standards including user defined standards User defined standards are preceded by an asterisk 3 Select one of the predefined standards or Press Browse to load your own standard stored in a text file This opens a dialog box which allows you to load a file By default the browser searches for files with the suffix jcs but you can access any file The required file format is described in detail in User Defined Standards on page NOTE 452 4 Press OK to load the data temporarily Use this option for example if the jitter frequency list of your standard is still subject to change or Press Apply to Predefined Standards to load the data and add the standard to the list of predefined standards For details see Adding a User Defined Standard on page 461 If you wish to increase the stress on the device under test you can enter a Margin Agilent J BERT N4903B High Performance Serial BERT 455 8 Jitter Tolerance Tests 456 TIP NOTE With the Auto Frequencies checkbox you can adjust the test frequencies automatically to the selected standard This is describe
33. Connect the Data Output of the pattern generator to an oscilloscope NOTE As long as the N4916A is in connected state after pressing the Connect button the pattern generator s output is programmed to a fixed level The amplitude is set to 1 2 V the offset to 0 This yields a signal of 0 6 V The oscilloscope should show the following picture If you see only a straight line ensure that the pattern generator is not disabled and that it is set up to generate a pattern e g PRBS Check the Output of the Pattern Generator with the Error Detector If you have connected the N4916A to a Serial BERT you can of course check the output of the pattern generator by means of the error detector 1 Setup the pattern generator as described above Set the Data Offset to 0 V and the Amplitude to 1 V 2 Set up the error detector as described above Agilent J BERT N4903B High Performance Serial BERT 499 9 Solving Problems 3 Connect the Data Output of the Serial BERT to the Data Input of the error detector 4 Ensure that the error detector follows the generated pattern Press the Auto Align button From the Analysis panel select the Eye Diagram page Press the Start button After some seconds you should see an eye diagram like the following 11 10 9 T 3 2 1 0 Eye Diagram EnorAdd Linsens 7 Eye Diagram gram Properties z Time 0 20U1 Div Delay 0 98U1 Scale 147 11mWDiv ffset 5mV
34. Installing External Instrument s Concepts on page 538 Agilent J BERT N4903B High Performance Serial BERT Setting up External Instrument s 2 Now since the physical connections are OK you can enable the external instruments through the Serial BERT s Config window 1 Click Config menu item from the External Instrument s submenu The Config window shows the list of externally connected instruments Ensure that the external instruments are properly connected and installed For more information on Config window refer to the section Config Window on page 51 Here are few examples Example 1 The following image shows the Config window when the N4916B de emphasis signal converter is connected to the J BERT N4903B BER 0 000 Remote Enable Function VISA Resource Name DATA QUT z Clock Multiplier Example 2 The following image shows the Config window when the N4876A 28Gb s Multiplexer is connected to the J BERT N4903B Agilent J BERT N4903B High Performance Serial BERT 43 2 Setting up External Instrument s BER 0 000 Remote Identify Connection Enable Function VISA Resource Name IMASzBA mal Nasza 7 Multiplexer Lu us8 0x0957 0x7518 C Example 3 The following image shows the Config window when the M8061A 28Gb s Multiplexer with De emphasis is connected to the J BERT N4903B 0m BER 0 000 Remote Identify _ Connection Enable Function _ S WR
35. Patterns can be viewed in Binary Hexa Decimal and Symbol format The Bin Hex Symbol Configuration dialog box provides option to view the patterns in Bin Hex Symbol format Bin Hex Symbol Configuration View As C Bin C Hex Symbol Running Disparity v Automatically maintains running disparity Starting Running Disparity Automatic C Manual l v Dk Cancel Help For Symbol format this dialog allows you to configure how the pattern editor handles the running disparity You can allow the editor to automatically maintain a correct running disparity while you edit by enabling the Automatic maintains running disparity checkbox In this case the Running Diparity will be applied on the full pattern i e for alternate patterns it will be applied on both A and B patterns You can also set the Starting Running Disparity to either Automatic or Manual 1 or 1 Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 Pattern Editor Canvas and Status Bar The pattern editor canvas displays the pattern and offers you to edit it 1100 0000 1100 0011 Bram eL DOCCT e S E O T REPRESENT POCO jaf al al 1100 1111 0000 1100 0011 1100 1111 0000 satin furatus ab abteyle hak ab ah ala a a ah al nie olin ahah sh al al a 1111 0000 1100 0011 1100 1111 0000 1100 tala ah alah tables ab ak duum ak al aya que boh bakata bon al 0000 1100 0011 1100 1111 0000 1100 0011 ube io ah dapes at at fej eh fmm aha
36. Some standards are more detailed as illustrated in the following figure Jitter amplitude in Ul log scale Jitter tolerance curve Jitter frequency log scale Agilent J BERT N4903B High Performance Serial BERT 447 8 Jitter Tolerance Tests 448 Measurement Results Instrument Capability During a jitter tolerance compliance test jitter with the specified jitter amplitude is sequentially applied at a number of frequency points in the frequency range of interest The receiver checks for transmission errors and measures the bit error rate The graphical display shows the results Jitter amplitude in Ul log scale Measured points Jitter frequency log scale A green circle indicates that the BER at this point did not exceed the target BER A red cross indicates that the BER at this point exceeded the target BER If this happens the test is marked as failed A table below the graph provides details about the measurement conditions like date time data pattern bit rate and added jitter components The measurement results for each point can also be exported as a numerical list in text format The maximum jitter amplitude the Serial BERT can generate is also frequency dependent as illustrated below Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Jitter amplitude in Ul log scale Maximum jitter amplitude a over frequency Jitter frequency log scale
37. The following picture illustrates the GUI layout of the Serial BERT 18 Agilent J BERT N4903B High Performance Serial BERT Planningthe Test 1 BER 0 000 mess fee zum poe BE Ze mx Toccoa me nz Ia qm Ze pu NN E feco ma Few For understanding the Serial BERT GUI can be divided into three panes namely the upper middle and the lower pane Upper Pane The following picture illustrates the upper pane of the Serial BERT 7 BER 0 000 The upper pane provides access to the Navigation Menu which allows you to launch the different controls and dialogs of the Serial BERT For more information on the Navigation Menu refer to the section How to use the Navigation Menu on page 21 In addition this pane also displays the following controls buttons and indicators BER Status Bar Displays the calculated BER Elapsed Time Indicator Displays the progress of the current accumulation period If a measurement is running it displays the measurement progress Show Error Message Button Displays error messages The Show Error Message button uses the following color codes to indicate different states Noerror messages The button will be in disabled state New unread errors The triangle icon on the button will turn red No new errors The triangle icon on the button will remain yellow Minimize Button Minimizes the GUI to the Taskbar Remote Indicator If the GUI is in remote state the remote indica
38. To find out more about appropriate pattern lengths see Pattern Resolutions and Lengths on page 67 Pattern Type Select Standard if you want to use this pattern as a standard pattern Select Alternate if you want to define two alternating patterns A and B Conversion Options Qn changing the pattern type you have the conversion options given below For converting an Alternate Pattern to a Standard Pattern Discard A Discard the A part of the Alternate pattern and use the B part for the Standard pattern Discard B Discard the B part of the Alternate pattern and use the A part for the Standard pattern Multiplex ABAB Multiplex the Alternate pattern into the Standard pattern starting with A Multiplex BABA Multiplex the Alternate pattern into the Standard pattern starting with B By default the pattern length is not changed when selecting one of the Multiplex conversions If the created Standard pattern shall contain all of the bits of the Alternate pattern then the Length in Bits has to changed manually For converting a Standard Pattern to an Alternate Pattern A Pattern fill B with 0 The content of the Standard pattern is copied to the A part of the Alternate Pattern The B part is filled with 0 B Pattern fill A with 0 The content of the Standard pattern is copied to the B part of the Alternate Pattern The A part is filled with 0 Demultiplex The even bits of the Standard pattern will be
39. Trigger Ref Clock Output Concepts The pattern generator s Trigger Ref Clock Out port can be used to send a reference signal or trigger to external devices like an oscilloscope or digital communication analyzer In pattern mode the pattern generator sends a trigger signal that is at least 32 bits long See also How the Serial BERT Generates Memory Based Patterns on page 69 for details In sequence mode the pattern generator can send a trigger signal whenever a block of the sequence starts or restarts See also Sequence Block Parameters on page 102 for details Several options are available for the trigger signal As an example you can send the trigger as a divided clock signal or as an indicator when the data pattern starts Trigger Ref Clock Output Procedures To set up the pattern generator s Trigger Ref Clock Output port 1 Connect the external instrument to the Trigger Ref Clock Out port 2 Click Trigger Ref Clock Setup menu item from the PG Setup submenu Agilent J BERT N4903B High Performance Serial BERT 145 4 Setting up the Pattern Generator 146 NOTE zn Trigger Ref Clock Setup 3 Select the trigger pattern that you want to be generated See Trigger Ref Clock Output Reference on page 146 Aux Data Output Reference on page 154 for descriptions of the available trigger signals If you have downloaded a user defined sequence of patterns to the pattern generator the pattern related settings are ignore
40. You can measure the random jitter distribution of each peak as well as the distance between the peaks which means the deterministic jitter You can also use the marker with logarithmic scale In this case it appears as a parabolic curve 1 00e 10 1Trace 43 Points Ul 8 000 ns All Errors 1 00e 8 Gaussian Marker Value Mu 0 528 Ul 1 00849 Sigma 0 003 UI 1 00e 7 Kappa 3 394e 1 1 00e 6 1 00e 5 Logarithmic 0 68 UI 0 63 UI 0 57 UI 0 53 UI 0 47 Ul Relative 0 70 Ul 0 65 UI 0 60 Ul 0 55 UI 0 50 Ul A Gaussian marker is used when the dBER vs Threshold Graph is displayed This graph shows the relationship between the decision threshold and the absolute values of the derivative of the bit error rate dABER dTh A linear scale reveals the distribution more clearly than a logarithmic scale see dBER vs Threshold Graph on page 274 In the example below p Mu and o Sigma will be the same as the Level and Standard Deviation results calculated by the measurement Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 1 Trace 500 Points 800 0 mv Gaussian Marker Value SD mV Mu ATV Sigma 16 7 mv Kappa 1018e 0 1410Y aav 1 2 Y 13 Y aav ASV ABV ATV A8V Linear 2 000e 0 6 000e 0 1 000e 1 1400e 1 1 800e 1 2200e 1 2 600e 1 3 000e 1 0 000e 0 4 000e 0 8 000e 0 1200e 1 1 600e 1 2000e 1 2400e 1 2800e 1 3200e 1 But your dBER distribution may also
41. Your device is faulty DATA LOSS This indicator turns red when no data signal is detected at the ED Data In port BERT connected looped back or to your device The PG data output is off or not connected There is no connection to the ED data input The 0 1 threshold is not in the eye limits of the incoming data signal Use Auto Align or select Avg 0 1 Threshold Your cables are faulty BERT connected to your device There is no data signal from your device Your device is faulty SYNC LOSS This indicator turns red when the measured BER is higher than the sync threshold BERT connected looped back or to your device The sampling point is incorrect Use Auto Align The data pattern is inverted Toggle Data Inverted or use Auto Align The ED pattern does not match the incoming data pattern Make the ED track the PG pattern or set ED and PG patterns independently the same Patterns have lost sync in manual sync mode Use Auto Align Sync Now or select Auto Sync mode BERT connected to your device The data eye is very small If Auto Align does not correct sync loss set the 0 1 threshold manually Your device has inverted the data pattern Toggle Data Inverted or use Auto Align Your device has changed the data pattern Set the ED and PG patterns separately If the fundamental BER of your device is higher than the sync threshold BER you can select a higher sync threshold BER Agilent J
42. and an arbitrary data signal for testing your device Pattern Generator Enor Detector os CO p Dassen r e Jem S oa Plan Mele T Oe oat ze O O o e iei jel Understanding the Pattern Generator s Input Ports The pattern generator provides the following input ports 10 MHz Ref In This connector allows you to connect a 10 MHz reference clock The Serial BERT has a 10 MHz Ref In and Out port at the rear two BNC connectors Clock In Allows you to connect a clock signal from another device Error Add Allows you to connect a device that generates a trigger for inserting a single bit error in the output stream In addition it also allows you to control the electrical idle state of Data Out and or Aux Data Out Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 Delay Control In Allows you to connect a signal for generating jitter in the data outputs Aux In The signal at the Aux In port can be used to switch between patterns A and B or to blank suppress the output signal When a user defined sequence is executed it can also be used to control the sequence execution Understanding the Pattern Generator s Output Ports The pattern generator provides the following output ports Data Out and Data Out The data outputs serve as device stimuli and can be set up so
43. and shape variations impact width height and symmetry of the eye opening An example is shown below Eye Signal amplitude amplitude Eye width Signal period 1 Ul How the N4903 Generates Jitter An Serial BERT on which the calibrated and integrated jitter injection option J10 and the interference channel option J20 are installed combines all the necessary jitter sources in one instrument Jitter Generation Block Diagram To support all present test standards the pattern generator of the Serial BERT has the following hardware architecture Ea m oe Tene D ERI ES UN ECONEEN ee 37 M 22072 rewarded Clock on e acus Clock generator The clock can be generated from the internal oscillator or an external source Jitter sources The instrument has internal sources for spread spectrum clocking SSC residual Spread Spectrum Clock rSSC sinusoidal jitter SJ two periodic jitter sources PJ 1 and PJ 2 bounded uncorrelated jitter BUJ Random Jitter RJ and spectrally distributed Random Jitter sRJ which is composed by using the BUJ Agilent J BERT N4903B High Performance Serial BERT 401 8 Jitter Tolerance Tests 402 Clock modulator Delay lines NOTE NOTE and the RJ sources i e BUJ and RJ are mut
44. bit rate sequence length The closer to a continuous spectrum the more the test signal looks like an analog noise source The benefit of this is that such a stimulus shows any frequency dependent effects in a device under test For example a clock recovery circuit Agilent J BERT N4903B High Performance Serial BERT 63 3 Setting up Patterns designed to lock onto a single frequency may have a tendency to lock onto harmonics A long PRBS may uncover such undesirable behavior When to Use Which Pattern Selecting the appropriate pattern and test setup for your application is important Below is a list of recommendations that may be helpful Tosimulate random traffic choose PRBS For functional and alarm testing choose alternating memory patterns Ifthe test pattern requires a header choose one of the Example Patterns on page 70 or use a custom pattern To stress lightwave transmitters and receivers vary the mark density of the pattern with a modulator or e o converter Toinduce baseline wander for margin testing vary the mark density and use alternating patterns Forpattern dependency testing choose one of the Example Patterns on page 70 or use a custom user pattern vary the mark density Foreye diagram measurements and mask analysis use the Serial BERT with a digital communication analyzer or oscilloscope Forreceiver sensitivity or eye contour measurements use the Serial BERT with a digital commun
45. decision threshold The location of the sampling point is the decision factor as to whether the incoming bits are identified as logic 0 s or 1 s To measure the accurate bit error ratio at the input port false readings of logic 0 s or 1 s must be avoided Therefore the sampling point must be set to the optimum location within the data eye The functions within the Sampling Point Setup window allow you to Prepare the error detector for the incoming data signal regarding the connector termination Graphically display the eye diagram in terms of voltage input delay and a BER threshold Adjust the location of the sampling point What is an Eye Diagram An eye diagram provides a way to view all possible transitions between adjacent bits of a data stream by overlaying them on the display of a high speed oscilloscope It is typically produced by triggering the oscilloscope with a synchronous clock signal When the oscilloscope is triggered by a clock signal data patterns will not be examined Instead if persistence is set high enough the shape of an eye will be Agilent J BERT N4903B High Performance Serial BERT 193 5 Setting up the Error Detector 194 observed The eye is bounded by overlaid logic 1 and 0 voltages top and bottom and multiple 0 to 1 and 1 to 0 transitions left and right f Eye diagram Overlay of all portions of the bit pattern The sampling point of the error detector must be set within the d
46. icon on the Jitter menu The following submenu will appear Jitter Setup Interference Channel B 2 1 Tolerance Characterization Tolerance Compliance Generate HTML Report 2 Select Generate HTML report le E 3 Select the folder where you want to save the report name the file and save it 460 Agilent J BERT N4903B High Performance Serial BERT Adding a User Defined Standard NOTE Deleting a User Defined Standard Managing Measurement Results Jitter Tolerance Tests 8 Managing User Defined Standards User defined standards are saved in the folder C lt instrument model NJTolStandards The extension is jcs To add a user defined standard to the list 1 Press the Properties button and switch to the Standard tab 2 Select the last entry in the list Browse This opens a dialog box which allows you to load a file The file must contain a comma separated list of frequencies and jitter values The file format is described in detail in User Defined Standards on page 452 3 Select the desired file and press Open The Properties of Imported Standard dialog is displayed 4 Enter a Name and Description for the standard 5 Press Apply to Predefined Standards The standard is added to the list of predefined standards In the list user defined standards are preceded by a To delete a user defined standard from the list Delete the corresponding file in t
47. limitations Some examples of modifications that you can make are listed below Add errors to the data stream See Error Addition Insertion Concepts on page 157 for details You can for example perform a confidence test of your measurement setup The analyzer should recognize all errors that you add If it does not you may have a problem with your measurement setup See Setup Problems Concepts on page 469 for details Usealternating patterns to switch between different data streams For example if your device is designed to turn off in response to a high BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS Bit Comparison without USB3 1 SKPOS you can check if your device behaves as expected And you can observe what the BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS is prior to data loss See Pattern Alternation Concepts on page 161 for details Manipulate the signal being generated by the pattern generator as described in Setting up the Pattern Generator Concepts on page 121 Adjust the sampling point to see the effect on the BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS Bit Comparison without USB3 1 SKPOS See Sampling Point Setup Concepts on page 193 for details Make modifications to your device setup Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 This can be used to see how changes to your device affect th
48. the Auto Align button is pressed with respect to clock rate signal voltages and offset decision threshold and so on The bits received at this sampling point are considered to be correct You can now move the visible sampling point and investigate the behavior of the DUT in detail The error detector compares the bits at your sampling point with the bits recognized at the hidden sampling point As you move the visible sampling point the indicated BER will increase The indicated BER is a relative BER It approaches the absolute BER if the error ratio at the initial sampling point is very low It is identical with the true BER if the signal at the initial sampling point is error free In bit recovery mode some restrictions apply Synchronization is impossible because no particular data is expected Sync now and auto synchronization functions are disabled Burst mode is disabled Agilent J BERT N4903B High Performance Serial BERT 195 5 Setting up the Error Detector 196 Error Location Capture cannot be used The function Error Detector Pattern tracks the Pattern Generator Pattern is disabled This function is not re enabled when the BRM is terminated BRM does not work above 11 5 Gb s when the CDR is enabled BRM indicators The mode is indicated in the status panel and in the BER Bar The mode is indicated in all screen dumps and log files The Measurement User Interface shows when this mode is used
49. you can change BackColor The background color of the graphs default is white BERMarkerColor The color of the BER Threshold indicator default is red ForeColor The foreground color of the scales and frame of the graphs default is black FreqRangesColor Color of the selected frequency ranges default is light yellow GridColor The color of the dashed grid lines default is gray PowerMarkerColor The color of the Noise Threshold marker default is red 2 Selectthe color that you want to change from the list of used colors to the left 3 Use the Color Set and the Color Palette to select the new color 4 To view the results of your changes press Apply This immediately updates the measurement window where you can see the new colors 5 Press OK to close the Properties dialog box Agilent J BERT N4903B High Performance Serial BERT 233 6 Advanced Analysis Advanced Analysis Reference 234 This section covers reference information such as definitions of important parameters and the descriptions of supplementary functions Exporting Result Data If you want to use the measurement results with other applications you can export the data to a file via Analysis Export Data The contents of the resulting file may look as follows Date 02 11 05 03 33 17 Version 1 0 Type TM Fast Eye Mask SB Electrical Ul 9 6969E 011 Threshold Units Volts Delay Relative 1 1 SerialBERT D
50. 0 260 UI 0 000 UI 0 260 UI 0 520 UI 0 780 UI 1 040 UI Several zoom factors are available When you show the zoom graph you can also allow the zoom graph to track the mouse or your finger if you are working directly on the Serial BERT If you want to see the points that have actually been measured choose Show Measured Points The MUI uses linear interpolation for the graph If you want to display the BER threshold contour line choose Show BER Threshold The BER threshold is set on the View tab or by the slider in the Color Bar If no slider for the BER threshold is displayed make sure that you have enabled the BER Threshold option on the View tab and that a suitable BER is entered Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Explanation of the Numerical Results Additionally to the graphical results the measurement provides numerical results Measurement Parameters The measurement parameters are defined in the following List Time Eye Opening This is the maximum extension of the BER threshold contour line in sample delay direction eye width TIT This value is different from the horizontal extension of the BER threshold bounding box Threshold Eye Opening This is the maximum extension of the BER threshold contour line in sample voltage direction eye height NOTE This value is different from the vertical extension of the BER threshold bounding box Optimal Sample Delay
51. 1 Here the asterisk symbol will be treated as don t care Agilent J BERT N4903B High Performance Serial BERT 191 5 Setting up the Error Detector 192 NOTE 10B Alignment Symbol The 10B Alignment Symbol contains the K28 1 K28 5 and K28 7 symbol for 10B alignment You can use the drop down list to specify symbols All standards supported by Serial BERT use the K28 5 symbol for 10B alignment However the standards differ in the way if the K28 5 is only used during some training sequence or if it comes also later in the data stream In SATA for instance K28 5 also comes in the Filler Primitive Align Symbol which occurs regularly and makes it easy to achieve 10B alignment In other standards like USB 3 0 K28 5 is not present in the Filler Primitive In USB 3 0 a pair of two K28 1 symbols is used as filler primitive SKP ordered set This means a special initial 10B alignment phase must be done To achieve 10B alignment for all cases we can configure the 10B alignment symbol also as K28 1 symbol Symbol Lock Indicator In the automatic symbol alignment mode 10B alignment is constantly done to achieve symbol alignment However if 10B symbol alignment is not done or is lost then this status is indicated by Symb Lock indicator present at the lower pane of the Serial BERT s GUI Sync Data Clock 8b10b Loss Loss Loss Error Error In the manual mode symbol alignment is done only once if alignment was not done b
52. 1 1 0 8 0 6 0 4 0 2 0 0 2 0 4 0 6 0 8 delay adjust ui Explanation of the Fast Total Jitter Measurement The Fast Total Jitter measurement is an optimized method to determine the total jitter for devices that generate a very low error density BER well below 10 10 To measure not estimate the total jitter for a device with a BER of 10 with conventional methods one usually needs to compare more than 10 bits for each sample point To measure a full eye opening this way with appropriate timely resolution takes time maybe days or weeks depending on the data rate and the probability of seeing one or no error in 101 bits is not higher than 37 If one would compare 10 bits for the same device the probability of seeing ten errors is even lower 12 but the probability of observing no error is almost zero The Fast Total Jitter measurement implements a method that reduces the measurement time considerably and provides a higher accuracy It is based on statistical and probability calculations The method was presented at the DesignCon 2005 and is described in Total Jitter Measurement at Low Probability Levels using Optimized BERT Scan Method included as 5989 2933EN pdf We will not go into the details but provide an introduction Quite often we do not need to measure the exact BER but can stop the measurement if we are sure that the BER is above or below a threshold In a jitter tolerance test for example we need ju
53. 12 5 minutes 20 minutes 80minutes 5 3 hours 1E 11 30 seconds 2 minutes 8 minutes 32 minutes 1E 10 3 seconds 12seconds 48seconds 3 2 minutes The formula for confidence level is as follows C 1 e nb Where C degree of confidence 0 95 95 e natural base number n number of bits examined without error b desired residual BER You may want to prove that your device has a certain residual BER If you know this desired BER and the desired confidence level you can calculate the number of bits that must be measured without error The number of bits in turn can be translated into a period of time at a certain bit rate That is what is done in the above table The following example shows how to calculate the number of bits if the desired BER b is 1E 10 and the desired confidence level C is 95 1 Change formula C 1 e nb to solve for the number of bits n n In 1 C b 2 Replace variables C and b n In 1 0 95 1E 10 3 Calculate the number of bits n In 1 0 95 1E 10 2 99 1E 10 2 99E 10 bits 4 Calculate the time period at the STM 16 0C 48 bit rate 2 48832 Gb s 2 99E 10 2 48832E 9 12 016 seconds Agilent J BERT N4903B High Performance Serial BERT Planning the Test 1 Connecting the DUT Connecting the DUT Concepts This section provides information on how to connect your DUT for several common test scenarios NOTE The pattern generator s Data Out Aux Data O
54. 3 Switch to the Graph tab and select Linear Scale The data remains the same but a linear scale makes it easier to see the distribution Agilent J BERT N4903B High Performance Serial BERT 269 6 Advanced Analysis 210 Output Levels NOTE 1 Trace 500 Points 800 0 mv 800 0 mv A0V Adv 42V ASV 44v ASV 16V ATV 48V Linear 2 000e 0 6 000e 0 1 000e 1 1400e 1 1 800e 1 2200e 1 2 600e 1 3 000e 1 0 000e 0 4 000e 0 8 000e 0 1200e 1 1 600e 1 2000e 1 2400e 1 2 800e 1 3 200e 1 This graph shows the absolute values of the derivative of the bit error rates over the thresholds dBER dTh It visualizes the data that forms the basis for the calculations of the level and noise values The graph provides a special marker that allows you to estimate the data distribution by approximating it by means of a Gaussian normal distribution The Output Levels measurement provides a third graphical display the 0 from BER versus Threshold graph This graph refers to the Q factor calculations For details on these calculations see QBER vs Threshold Graph on page 275 and Understanding the Q Factor Results on page 282 Reference The Output Levels measurement returns the results in a graphical and in a numerical form The following sections provide explanations of the measured parameters and the display options that are specific to this measurement Additionally some information is provided to explain the theoretical back
55. 5 Sampling point BER 05 BER 025 BER The bit error rate at the right hand side of the jitter region is 0 We therefore expect a bit error rate around 0 25 Agilent J BERT N4903B High Performance Serial BERT 347 6 Advanced Analysis 348 NOTE FFT results The Spectral Jitter measurement should only be used in conjunction with data that has an equal distribution of ones and zeros over time Otherwise the results are hard to predict and may be not reproducible Signal Processing If the error signal is obtained as explained above an analysis in the frequency domain reveals the absense or presence of deterministic jitter Dominant frequency components become visible and their contribution to the total jitter can be measured The modified error signal is subject to a fast Fourier transformation FFT FFT requires that the data record to be processed has a length that is a power of two 2 such as 217 218 219 220 and so on From that data record the FFT calculates pairs of frequency power values The number of pairs is half the number of samples If you have chosen a record length of 2 7 which means 131 072 bits or 128 Kbit the result contains 65 536 pairs The maximum frequency is half the data rate used for the test The results are displayed in the Spectral Jitter measurement graphical and numerical result window About FFT For general information about the Fourier transformation and the specia
56. 520 UI 0 2 ul This plot visualizes the BER by a continuous color gradient This is useful to get an immediate visual impression of the distribution of the BER in the eye diagram Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 1500 mv 1 Trace 13689 Points Ul 1 538 ns All Errors o 1 000e 0 120 0 mv 1 000e 1 90 0 mv 50 0 mv 1 000e 2 S mv 1 000e 3 00v 300 mv 1 000e 4 60 0 mv 1 000e 5 90 0 mv 1 000e 6 120 0 mv 150 0 mv gue 000e 7 0 130 UI 0 130 UI 0 390 UI 0 650 Ul 0 910 UI 0 260 UI 0 000 UI 0 260 UI 0 520 UI 0 780 UI 1 040 UI Equal BER at BER threshold It shows the contour line at the BER threshold 1500 mv 4 Traces 53767 Points Ul 1 538 ns All Errors 1 000e 0 120 0 mv 1 000e 1 90 0 mv 60 0 mv 1 000e 2 SANE 1 000e 3 nov 300 mv 1 000e 4 60 0 mv 1 000e 5 80 0 mv 1 000e 6 120 0 mv 150 0 mv 1 000e 7 0 910 UI 0 650 UI 0 390 UI 0 130 LII 0 130 UI 0 390 UI 0 650 UI 0 910 UI 0 780 UI 0 520 UI 0 260 UI 0 000 UI 0 260 UI 0 520 UI 0 780 UI 1 040 UI Furthermore the following parameters can be set on the View tab of the Properties dialog box Analyze You can analyze for All Errors To display all errors Errors if 0s Expected To display the errors if 0 is expected but 1 received e Errors if 1s Expected To display the errors if 1 is expected but 0 received Calculate You can calculate measurement parameters for DErrors To cal
57. 96 SER 13 SER Results Window 367 SER Status Indicators 372 Serial BERT introduction 11 Setting Date and Time 529 SL 400 Signal Noise Ratio Peak Peak 279 Sinusoidal Interference 432 Sinusoidal Interference SI 400 Software generated patterns 62 Software generated patterns 93 Software generated patterns Pattern software generated 91 Solving Problems 469 SONET SDH Receiver Connections 33 Spectral Jitter measurement procedure 351 Spread Spectrum Clocking SSC 141 Standard patterns 62 Status Indicators 372 Supported Pattern Types 65 Symbols Error Ratio 182 SYNC LOSS 471 Sync Loss Seconds 475 Sync Now button 200 T Termination Error Detector 173 Terminations Setup 131 Usage 128 Test Duration 27 Planning 11 Requirements 26 Which Test Is Appropriate 25 Test points Fast Eye Mask 308 Threshold absolute 312 offset 312 percentage 312 Threshold Eye Opening 301 Agilent J BERT N4903B High Performance Serial BERT Index Threshold Margin 278 Threshold Resolution High Level Eye Opening measurement 296 High Level Output Levels measurement 271 Low Level Output Levels measurement 271 296 Threshold Types 312 Time Eye Opening 301 Timing Unit Eye Opening measurement 300 Fast Eye Mask measurement 312 Total Jitter BER Threshold 263 Total Jitter Peak Peak 262 Total Jitter Peak to Peak 260 Total Jitter RMS 260 Total Jitter Uncertainty 263 Touchscreen Calibrati
58. After the self test has finished the results Success or Failed are displayed next to each test If a test has failed you can use the Message button to open the BIOS Messages window with detailed error messages explaining the failure reason Agilent J BERT N4903B High Performance Serial BERT 565 10 Customizing the Instrument 566 Agilent J BERT N4903B High Performance Serial BERT Index 0 1 Decision Threshold 197 0 1 Threshold 198 0 1 THReshold Center button 200 OV Disable button 138 10 MHz Ref Input 122 8B 10B Comparison Results Analyzing 365 Monitoring 364 8B 10B Comparison Results Window 367 A Absolute threshold 312 AC Coupling 130 Accumulated Measurements 376 Results 378 Running 377 Accumulated Results 380 Accumulation Parameters 382 Accumulation period 222 Adjust Output Levels 132 Advanced Analysis 227 Agilent Recovery System 525 Alternating Patterns 161 Amplifier Connections 31 Amplitude 278 Audio Signals 224 Auto Align button 199 Auto Threshold Error Detector 181 Aux Clk Out port 123 Aux Data Out port 123 AuxIn 123 Aux Out Error Detector 205 Auxiliary Out port 170 Auxiliary Output 203 Average 0 1 Threshold 201 B Bathtub curve 242 BER 12 BER Location 223 BER Results Analyzing 365 Monitoring 364 BER Results Window 366 BER Threshold Eye Opening 299 Output Levels 275 Output Timing 258 Sampling Point Setup 201 BER values 320 BER vs Threshold G
59. All these filters can be disabled or enabled by pressing the respective buttons d MHz E High Pass Spectrally Distributed Random Jitter Spectrally distributed random jitter is composed of two jitter sources low frequency jitter and high frequency jitter It is characterized by the amplitudes of the low and the high frequency jitter I me pe Spectral random jitter is characterized by Amplitude LF Amplitude HF The filter settings graph shows the actual and the possible jitter spectrum In addition the total sRJ amplitude as rms value is displayed in UI and in ps It is calculated using following formula Amp H Fe Amp_LF Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Amplitude LF rms This is the low frequency jitter amplitude as rms value Amplitude HF rms This is the high frequency jitter amplitude as rms value Filter The Spectrally Distributed Random Jitter is equipped with a 100MHz Low Pass filter to limit the spectral range which can be enabled by pressing the corresponding button External Jitter Source Parameters When you enable the External Jitter the Jitter Setup function reserves the remaining free capacity of the 220ps delay line for the signal applied to the Delay Ctrl Input Z 153 75 mUI 362 mv The parameter window informs you about Amplitude max Voltage max Slope Agilent J BERT N4903B High Performance Serial BERT 429 8 Jitter Toleran
60. BERT Setting up External Instrument s 2 Controlling N4876A Multiplexer After the multiplxer function has been enabled through the Config window the following parameter are visible in Multiplexer N4876A window 1 Provide f 2 Jitter between 10 ps to 10 ps For more information refer to the section f 2 Jitter on page 55 Controlling M8061A Multiplexer with De emphasis After the Mux with De emphasis function has been enabled through the Config window the M8061A window appears as shown in the following figure 7 BER 0 000 Remote lt Unbalanced E 0 000 ps e The M8061A window contains the following tabs 1 DataOut Tab provides parameters to set the Amplifier Deemphasis and Interference For more information refer to the section DataOut Tab on page 56 2 ClkGen Tab provides parameters to select clock source For more information refer to the section ClkGen Tab on page 58 3 Electrical Idle In Tab provides parameters to set electrical idle in For more information refer to the section Electrical Idle In on page 58 Setting up External Instrument s Reference Config Window The Config window provides an inferface which allows you to enable the functionality of the external instruments connected to the Serial BERT Agilent J BERT N4903B High Performance Serial BERT 51 2 Setting up External Instrument s Refresh Instrument List Identify NOTE Connection Enable Fun
61. BERT N4903B High Performance Serial BERT 265 6 Advanced Analysis 266 Use a shielded cable to connect the pattern generator s Data Out port and the error detector s Data In port Switch to the Pattern panel and press Pattern Select Select an appropriate pattern for this test We use a pure 2 23 1 PRBS segment For the pattern generator setup you need to specify the logic levels and the bit rate Select ECL levels and a clock speed of 1250MHz in this example This corresponds to a clock period of 0 8ns See Setting up the Pattern Generator Concepts on page 121 for more information Set up the error detector so that the input range and the termination matches the pattern generator s levels Select an Input Range from 2V to 0V Setthe Data Termination to 2V Setthe Alignment BER Threshold to 1E 6 Setthe Clock Setup to Clock Data Recovery to get the error detector s clock from the incoming data stream Enable the pattern generator outputs by pressing the 0V Disable button Press Sync Now and then Auto Align to find the optimum sampling point Check that the synchronization and the alignment were successful None of the error indicators should show red The resulting BER should be zero How to Execute the Output Levels Measurement To run the Output Levels measurement 1 2 3 Switch to the Analysis panel and then press the Output Levels icon Press the Properties button to open the Properties dialog
62. Characterization measurement offers therefore the option to set a level of confidence This approach is based on statistics Statistics tell us for example that if we receive 3 x 10 bits without any error the probability that the BER is below 10 is higher than 95 96 If an error occurred more bits must be compared to achieve the same confidence level For details see Explanation of the Fast Total Jitter Measurement on page 246 The measurement proceeds with the next amplitude as soon as the desired level of confidence is reached Setting a confidence level ensures that neither too few nor too many bits are captured Relax time If the measured BER is higher than the target bit error ratio the device under test may have lost synchronization For devices that need some time to recover from the situation a relax time can be specified This time takes effect each time the BER limit has been exceeded The next BER measurement starts after this time has elapsed Agilent J BERT N4903B High Performance Serial BERT 435 8 Jitter Tolerance Tests 436 Searching downwards Vertical Search Methods You can specify that the jitter amplitude changes in linear steps from maximum to minimum or vice versa The step size is adjustable You can also specify that the jitter amplitude changes in logarithmic steps from maximum to minimum or vice versa Minimum and ratio are adjustable Alternatively you can also specify one of two dynamic bi
63. Click this icon to download the sequence from the Sequence Editor to the pattern generator This icon appears green if the sequence has not been sent to the PG For instance if the sequence had been loaded from file or been changed since last sent to PG 105 3 Setting up Patterns 106 Table 14 Icon Name 10110 From PG PG Properties d Break 3 Reset Description Click this icon to upload the present sequence from the pattern generator to the Sequence Editor This allows you to inspect and edit a sequence that has been loaded with a program Click this icon to change the properties of the current sequence See Setting Sequence Properties on page 107 for details Click this icon to terminate an infinite loop that is set to manual break condition Sequence execution continues with the next block Click this icon to interrupt and re initialize a running sequence If the sequence start condition is IMMediate the sequence restarts immediately If the sequence start condition is a signal at the Aux In port the sequence restarts upon the specified signal If the sequence start condition is Command sequence execution starts when the Start button is clicked Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 Table 14 Icon Name Description duc CrossTalk Click this icon to open the soc CrossTalk dialog Use this dialog to fill every sequencer block of
64. Converter the Output Levels and Termination shown on the screen refer to the outputs of the N4916A B Check the indicated signal levels The output voltage range of the N4916A B is lower than the range of the Serial BERT The N4916A B is first of all meant for low voltage amplitudes If the N4916A B cannot generate the specified signal voltages they are reduced to safe levels If this happens you need to change the levels or the termination Do not remove the connection cable between the pattern generator and the N4916A B while the N4916A B is in connected state except for troubleshooting In this mode the output levels of the pattern generator are not shown on the screen and may present a hazard to your DUT they have an amplitude of 1 2 V Controlling De Emphasis Mode For De Emphasis mode you have de emphasis amplitude either pre cursor or post cursor and the corresponding unit Unit To change the de emphasis amplitude value for details see Understanding the N4916A B De Emphasis Signal Converter on page 37 click inside the text field and either e Enter the desired value directly with the numeric keyboard Usethe knob on the hardware front panel to set the value You can toggle between dB and percent Toggling does not change the value The following topics tell you how to use the pattern generator s input ports Bit Rate Procedures on page 141 Delay Control Input Procedures on page 156 e Manua
65. DER or AccumDER results Error Count Accum Error Count This area displays the actual illegal disparity change count or cumulative illegal disparity change count ISR Results Window This window displays the actual ISR or cumulative AccumISR ISR results Agilent J BERT N4903B High Performance Serial BERT 371 7 Evaluating Results Accumulated Actual Button Error Count Accum Error Count BER bar SER bar 372 RN ISR 0 000 Illegal Symbol Counts The ISR is the current ISR calculated upon a period of 200ms The AccumISR shows either the accumulated ISR of the current accumulation or if no accumulation is running the results of the most recent accumulation This enables you to monitor real time ISR behavior as you do things such as to manually adjust the sampling point to add errors or to make adjustments to your device Click this button to toggle between ISR or AccumlSR results This area displays the actual illegal symbol count or cumulative illegal symbol count Status Indicators The BER bar displays the BER calculated upon a period of 200 ms BER 0 000 You can drag the yellow alarm threshold mark to change the BER Alarm Threshold The SER bar displays the SER calculated upon a period of 200 ms SER 0 000 Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 You can drag the yellow alarm threshold mark to change the SER Alarm Threshold FER bar The
66. Diagram and the table of Automated Eye Parameter Measurement form simultaneously They become more precise with the increasing number of measured bits The following graph shows how the Eye Diagram has been integrated on the J BERT N4903A High Performance Serial BERT Agilent J BERT N4903B High Performance Serial BERT 315 6 Advanced Analysis BER 0 000 S S EN 3 2 1 0 Eye Diagram Ema 5 AJ r m Dia Properties Time 0 00U1 Div Delay 0 00U1 Scale mV Div ffset m Sample Count 1 Level Q Level Rise Time FallTime Eye Amplitude Eye Height Eye Yvidth Jitter Peak Peak Jitter RMS Cross Voltage lt No Data gt lt No Data gt No Data No Data gt No Data lt No Data gt No Data gt No Data gt lt No Data No Data lt No Data gt Pattern Generator jos en Error Detector loz Let 300000Gb s 2 231PRBS Loss ISI SSC ON 200000Gb s 2 231PRBS Enor ee Loss Loss Methods of Representations There are two methods of representing the eye diagram e Waveform Contour Waveform Waveform is the shape and form of a signal The waveform graph shows the periodical variation of voltage against time The waveform in the Serial BERT is similar to the one in the oscilloscope In this case the waveform initially gives a coarse but quick picture of the signal quality while the smooth waveform quickly generates a high resolution graph The waveform is displayed
67. Disparity in the patterns Fill Traces Rotate 8B 10B Invert Running Disparity Auto Correct Disparity NOTE The 8B 18B tab will be only be available if the Symbol Mode is enabled Status Bar The types of the patterns that are currently loaded to the pattern generator and the error detector are always displayed in the status bar at the bottom of the screen Select Pattern Dialog Box This dialog box lets you define the pattern you want to use for testing It contains the following elements NOTE The Error Detector Pattern in the Select Pattern dialog box allows the user patterns from file or memory user pattern Agilent J BERT N4903B High Performance Serial BERT 89 3 Setting up Patterns 90 Error Detector Pattern Tracks the Pattern Generator Pattern Pattern Types Pattern Size Sequence With this checkbox you can determine whether both pattern generator and error detector use the same data pattern If you clear this checkbox the Error Detector Pattern tab appears where you can select a different pattern for the error detector The checkbox is automatically cleared when you load a user defined sequence to the pattern generator See Sequence Mode Characteristics on page 98 for details You can choose between the following types of patterns User Pattern from File With this option you can load a pattern that is stored in a file Use the Browse button to select the desired file Memory User
68. How to Execute the Fast Eye Mask Measurement To run the Fast Eye Mask measurement 1 Switch to the Analysis panel and then press the Fast Eye Measurement icon 2 Press the Start button to execute the measurement Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 The measurement is run and the result window shows the bit error rates measured at six measurement points Terminal Copi 1 2 3 4 5 6 Relative Time 0 4 Ul 0 4 Ul 0 16 Ul 0 16 UI 0 16 Ul 0 16 Ul Voltage abs 22 4 mV 22 4 mV 178 mV 178 mv 222mv 222mv mE 0 0 0 0 0 0 Voltage abs 22 4 mV 22 4 mV 178 mv 178 mv 222 mY 222 mY The Relative Time refers to the current sampling point The Voltages are the decision threshold voltages for measuring the bit error rate at this measurement point The voltages of the measurement points can be set as absolute voltages as offset voltages or as percentages How to Optimize the View of the Results After you have run a measurement the resulting numerical values are displayed To improve the results you can change the measurement parameters 1 Press the Properties button to open the Properties dialog box 2 Use the different tabs in this dialog box to make the required settings Parameters tab These settings are used for data collection Changes require to run the test again See Parameters Tab on page 312 for details Pass Fail tab These settings determine whether the c
69. If this option is enabled the Serial BERT intelligently sets a resolution so that there are more sampling points at the edges of the eye This can greatly improve the results without dramatically increasing the duration of the test Set the criteria for the sample threshold Resolution Specifies how many measurement points are taken within the sample voltage range Enter the value in mV Agilent J BERT N4903B High Performance Serial BERT 295 6 Advanced Analysis Low Level Specifies the lower limit of the measurement voltage range Enter the value in mV This value should be slightly lower than the lowest expected signal voltage High Level Specifies the upper limit of the measurement voltage range Enter the value in mV This value should be slightly higher than the highest expected signal voltage Pass Fail Tab The Pass Fail tab of the Properties dialog allows you to specify the criteria to decide whether the DUT passes or fails the test You can change pass fail criteria without rerunning the test The software only uses the criteria to rate the results of a measurement NOTE The pass fail criteria do not control measurement execution The measurement run will be completed even if the measurement fails one or more of the criteria You can set several pass fail limits For an explanation of the Eye Opening Pass Fail criteria see Measurement Parameters on page 301 296 Agilent J BERT N4903B High Performance Se
70. Installing the M8061A Procedure The M8061A 28 Gb s Multiplexer with De emphasis is an optional instrument that can be connected to the J BERT N4903B For general information see M8061A 28 Gb s Multiplexer with De emphasis on page 17 The M8061A software is integrated in the J BERT N4903B installer which means that you don t require to install the M8061A software separately It automatically gets installed when you install J BERT N4903B software The J BERT N4903B software supports the M8061A starting from revision 7 50 Checking the J BERT N4903B Software Revision To check the J BERT N4903B software revision 1 Switch on the J BERT N4903B 2 Inthe Help menu click About 3 If the software revision is below 7 50 you need to update the software see Updating the Software on page 528 Connecting the M8061A with J BERT N4903B via USB The M8061A is controlled by the J BERT N4903B via USB Use the matched USB cable assembly USB 28 AWG 2 0 standard male A to B mini 4 conductor 3 mm to establish the USB connection The mini USB port is available at the front side of the AXI frame which is mounting the M8061A module Use the USB port USB 2 0 that is available in the rear side of the J BERT N4903B Follow the given steps to establish a USB connection between M8061A and J BERT N4903B Agilent J BERT N4903B High Performance Serial BERT 549 10 Customizing the Instrument 550 NOTE Connect the USB cable between the USB ports
71. Jitter 422 Periodic jitter 2 Variable Amplitude Sweep 424 Random jitter 427 Residual Spread Spectrum Clock 419 Sinusoidal jitter 420 Spectral Random Jitter 428 Spread Spectrum Clock 416 Parameters Tab Eye Opening 295 Fast Eye Mask 312 Output Levels 270 Output Timing 252 Spectral Jitter 353 570 Pass Fail tab Eye Opening 296 Fast Eye Mask 313 Output Levels 272 Output Timing 254 Spectral Jitter 355 Pattern Alternation 161 Creating New 71 Editing 72 Editor 76 102 Examples 70 from file 62 Generated Patterns 93 Hardware generated 61 Length 67 Load status indicators 89 Loading directly from Pattern Editor 75 Loading from file 73 mark density 92 memory based 61 Opening Existing 71 PRBN 63 PRBS 61 62 Properties 82 provided patterns 62 Repetition 69 Resolution 67 Saving 72 Selecting 89 Sequence 74 Sequence of patterns 96 Setup 61 software generated 62 93 Supported Types 65 Which Is Appropriate 64 Pattern Alternation Manually 163 With External Signal 163 Pattern Generator Aux Clk Out port 123 Aux Data Out port 123 Auxiliary Input 123 Clock In port 122 Clock Out port 123 Data Out port 123 Delay Control Input 123 Error Add input 122 Input and Output Ports 122 Ref Clock input 122 Setup 121 Trigger Ref Clock Out port 123 Pattern Mode 98 Pattern Synchronization 205 Peak Peak Noise 279 Percentage threshold 312 Periodic jit
72. O Configuration Tools Help ReleshAl W Und D Properties WH Interactive 10 HB Add Instrument W Add Interface XK Delete Task Guide Instrument on this PC ISA Alias Properties Tasks for a VISA Refresh All a aoos 011 9j COMI ASRL1 Sij coma ASRL4 r SA LAN TCPIPO Change propert USBD Send commands i ME N49164 USBO 2391 21000 0 0 INSTR the instrument 4 x Delete General Tasks e Refresh all L1 Add an instrume Refresh the instrument More Informatior How do I get B E MyInterface drivers c M MyInstrument f Myalias Where can I fim programming esmelac i 1 Agilent VISA is the primary VISA library The Agilent Connection Expert for N4916B shows a window like the following Ei Agilent Connection Expert DR Fie Edit View I O Configuration Tools Help RefreshAll 2 Undo I Properties E Interactivel0 E wer My Add Interface XK Delete Task Guide Instrument 1 0 on this PC VISA Alias Properties Tasks for a VISA Alias Refresh All E Refresh the instrument amp aoosssea Fe gosse EE gend eonmerss tothe d ig GPIBI VISA Alias DATA OUT SA LAN TCPIPO xX Delete 9 usso This alias refers to the following instrument a N4916B USBO 0x0957 0x5A18 Proto_R2_003 0 INSTR VISA address USBO 0x0957 0x5A 18 Proto_R2_003 0 X DATA our IDN string Agilent Technologies N49168 Proto R2 Ot g Refech al Manufacturer Agilent Technologies
73. P1 or P2 outputs with the DMI and CMI inputs of M8061A simultaneously however both will have the same modulation Agilent J BERT N4903B High Performance Serial BERT 35 1 Planning the Test 36 Agilent J BERT N4903B High Performance Serial BERT Setting up External Instrument s Setting up External Instrument s Concepts NOTE The External Instrument s menu allows you to enable external instruments connected to Serial BERT through a user interface Currently the following external instruments can be connected to Serial BERT 1 N4916A De Emphasis Signal Converter For more details refer to the section N4916A De Emphasis Signal Converter on page 16 N4916B De Emphasis Signal Converter For more details refer to the section N4916B De Emphasis Signal Converter on page 16 N4876A 28 Gb s Multiplexer 2 1 For more details refer to the section N4876A 28 Gb s Multiplexer 2 1 on page 17 M8061A 28 Gb s Multiplexer with De emphasis For more details refer to the section M8061A 28 Gb s Multiplexer with De emphasis on page 17 Understanding the N4916A B De Emphasis Signal Converter The De Emphasis Signal Converter can be connected between the Data Out port of the pattern generator and the DUT It converts the generated rectangular NRZ signal to a de emphasized differential signal The de emphasis functionality is supported by both N4916A and N4916B ee Agilent Technologies 37 2 Setting up
74. Power is calculated using the Dark Level and Conversion Gain of the View Tab 341 6 Advanced Analysis Spectral Jitter Eye Diagram Screen This graph gives the picture of the Eye Diagram Screen Error Add Properties Start a BER 0 000 on l 11 10 3 8 7 6 5 4 3 2 1 0 Eye Diagram v Dia E i t E Mask Time 0 00Ul Div Region Delay 0 00Ul Scale mV Div ffset m Automated Eye Parameter Contour Legend Sample Count 1 Level Level Rise Time FallTime Eye Amplitude Eye Height Eye Yvidth Jitter Peak Peak Jitter RMS Cross Voltage lt No Data gt lt No Data gt No Data No Data gt No Data lt No Data gt No Data No Data gt No Data No Data lt No Data 4 gt Pattern Generator m um m Error Detector 3000006b 2 231 PRBS Fock citer Outputs Data Clock 3 00000Gb s 2 23 PRBS Enor une Data Clock Spectral Jitter Concepts 342 The Spectral Jitter measurement allows you to analyze the jitter inherent in the output signals of your device under test DUT as a function of the frequency This measurement can be used for investigation of the behavior of the DUT for example to identify crosstalk effects Spectral Jitter Analysis The Spectral Jitter measurement provides a method for investigating the spectrum of the jitter i e its power distribution over frequency Deterministic jitter can be caused by repetitive dat
75. SSC response Auto Threshold If the Auto Threshold option is selected the Serial BERT continuously analyzes the incoming data stream to detect the voltage level that is optimal for the derivation of the clock signal This function is recommended for all applications where a valid data stream is continuously arriving at the error detector For measurements with partly invalid data such as burst tests this function may cause a clock loss In such cases deactivate Auto Threshold and enter a fixed voltage level manually or Agilent J BERT N4903B High Performance Serial BERT 181 5 Setting up the Error Detector Error Ratio 182 NOTE pressthe Measure button to measure the voltage level of the incoming signal while a burst signal is applied Once the level is defined it remains fixed for the following measurements Note that the entered voltage level has to be within the input range of the error detector If the value lies outside this range the change is rejected and an error message is displayed CDR Spread Spectrum Clocking This control is used to adapt the CDR to an input bit stream with SSC Enable widens the loss of lock detection window and sets the peaking to optimum SSC performance Enter the Expected Deviation and the type of deviation Upspread Downspread or Centerspread to set the locking window to an optimum SSC is mostly used down spread which means the clock signal is modulated to a lower frequ
76. Select Graphics Options then Output To and the desired output device Monitor for an external monitor Notebook for the built in display Use the same procedure to switch back to the built in display Configuring the Touchscreen Reference This section describes the options of the on screen keyboard Click Set to access the Options menu Move OSK Window Click this to enable the arrow keys on your actual keyboard They can then move the software keyboard Agilent J BERT N4903B High Performance Serial BERT 533 10 Customizing the Instrument TIP You can always move the On Screen Keyboard window by clicking its title bar and dragging it Additionally you can switch between different views of the on screen keyboard The following table lists the displayed buttons for the different views Table 37 Softkeys Letters Numeric Characters Qwerty Full X X X X Qwerty Simple X X X Qwerty X X Minimum Block Full X X X X Block Simple X X X Block X X Minimum Minimum X NOTE Qwerty options mimic the layout of an actual keyboard Block options present the letter characters in alphabetical order TIP If you select one of the larger on screen keyboard displays such as Qwerty Full or Block Full the keyboard may cover most of the instrument display If this occurs continue to make your keyboard entry then either close the on screen keyboard or change to a smaller display 534 Agilent J BERT N4903B High Performance Seri
77. Solving Problems 474 Example Log of Constant Errors Possible Cause Additional Information The error ratio during periods of constant errors was 0 00195313 This corresponds to 1 953E 3 2 49E 409 249E409 486000 0 00195313 2 49E 409 249E409 486000 000195313 2 49E 409 249E 409 485000 0 00195313 2 49E409 2 49E 09 486000 0 00195313 2 49E409 249E409 486000 000195313 2 49E409 249E409 485000 0 00195313 2 49E409 249E 4089 486000 0 00195313 2 49E 409 249E408 486000 X 0 00195313 Random Errors Accumulated Results Random errors may have been caused by a noisy waveform Noise can be caused by bad cables or connectors The output of your device may contain errors due to noise although a noisy waveform is not present at the output The noise may originate somewhere else within your device or system causing random bit errors to be passed on to other components Therefore your device may have a good output waveform and still transmit random bit errors Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 Sync Loss Seconds Accumulated Results Re Aea Raoul interval Resuts Accumulation Parameters Measurement Current Period Previcus Period Error Free Seconds Errored Seconds Power Loss Seconds Sync Loss Seconds 000006 Possible Cause The sync loss seconds in this example may have been caused by high bursts of errors or momentary clock loss The clock signal
78. Spectral Jitter Pass Fail criteria can be selected for the pass or fail decision Bit Error Rate Total Power in dB Noise Power in dB You can also define the pass fail criteria for each of the defined frequency ranges In the tabular view each of the calculated values will be marked with an e icon if it failed the test The following illustration shows a Spectral Jitter measurement that has failed the criterion for the Bit Error Rate Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 1 Trace 65535 Points 32 dB 24 dB 16 dB 8 dB 0 dB 8 dB 16 dB 24 dB 32 dB 40 dB Noise Threshold 40 dB 48 dB 56 dB Logarithmic 100 00 KHz 10 00 MHz 1 00 GHz 1 00 MHz 100 00 MHz Terminal Top Freq al Optical Hoise Power Show Color Copied Bit Err Total Power J Qo sese ol Plil I ald View Tab 30 176 dB 617 172 The graph shows the noise versus frequency The View tab provides you with various options for analyzing the data and setting frequency ranges for evaluation Power scale The linear power scale is well suited for identifying large peaks in the spectrum The View tab allows you to change the power scale from linear to dB The dB scale is a logarithmic scale The dB scale allows you to examine the whole power range The following figure shows an example of a double logarithmic display 1 Trace 65535 Points 2
79. Standard tab Checkbox is selected default The list yields frequencies that are recommended for the selected standard for example the edge frequencies Checkbox is not selected The list of frequencies yields a set of logarithmically equidistant frequencies TIT Alternatively you can set the Number of Steps on the Standard tab 464 Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Manual mode n Manual mode you can edit the automatically calculated frequencies You can also enter your own list of frequencies with a minimum of two frequencies specified For example you can jump between low and high jitter frequencies Enter the frequencies in Hz separated by commas for example 2e4 1 75e6 6 334e5 Press Apply if you want to keep the setting of frequencies for the current test run TIP A M o but if you still want to change the standard or BER specification If all properties are set as desired press OK BER Specification Target Bit Error Ratio The Target Bit Error Ratio defines the pass fail criteria of the jitter tolerance compliance test For every test point along the standard curve the test is failed if the measured BER lies above the Target Bit Error Ratio The Verification Method defines under which circumstances the measurement proceeds from one frequency to the next You can choose between two Verification Methods Confidence Level Setting a Confidence
80. Standard tab Iv Auto Frequencies e Checkbox is selected default The list yields frequencies that are recommended for the selected standard for example the edge frequencies e Checkbox is not selected The list of frequencies yields a set of logarithmically equidistant frequencies In Manual mode the automatically calculated jitter frequencies can be edited You can also enter your own comma separated list of frequencies NOTE If the frequency list is edited in Manual mode the Auto Frequencies checkbox on the Standard tab is automatically deselected The Sequence Advance checkbox advances the PG sequence during jitter tolerance compliance It will be enabled only if the sequence is send to the PG Setting the BER Limit 1 Set the Target Bit Error Ratio This defines whether a measured point is flagged green test passed or red test failed Agilent J BERT N4903B High Performance Serial BERT 457 8 Jitter Tolerance Tests 458 Avoiding Clock Loss Jitter Tolerance Properties 4 n BER Standard Frequency Target Bit Error Ratio 1e 9 X r Verification Method Confidence Level 0 95 C Absolute Bits Errors Nr of Bits Nr of Errors 1000 iv Relax Time li OS OK Cancel Apply Help 2 Choose the Verification Method Setting a Confidence Level is generally recommended See also Target BER on page 449 If you enable Absolute Bits Errors t
81. The printer properties and capabilities will vary depending on the printer you have selected Printing Problems You should check the following when your printer fails to print 1 Check if the printer is turned on and the printer cable is securely attached to both the printer and the instrument s printer connector 2 Check if the printer driver for your printer is installed 3 Check if the printer port is configured correctly For more information on printer problems see the printer s operation manual Agilent J BERT N4903B High Performance Serial BERT 557 10 Customizing the Instrument File Management File Management Procedures 558 This section provides information on saving and recalling instrument states log files patterns and screenshots You can save the files on the local disk Note that the local disk must not be re partitioned in a Windows 2000 environment You can also plug an USB stick into the rear of the instrument for saving and recalling files No driver is required because Windows has a built in driver that supports USB memory drives In the file browser you can access the USB stick via the Removable Disk drive To avoid loss of data stop the USB device and wait until the LED light of the USB stick is off before unplugging the USB stick To avoid loss of data when removing the USB stick 1 Minimize the GUI 2 Inthe system tray next to the clock in the lower right corner right click the Safely
82. Typically during tests when you adjust the amplitude you want to keep the offset constant This keeps the ideal sampling point within the eye The Serial BERT handles this by keeping V constant when Vampt is changed Changing the Output Levels On the other hand you may want to adjust the output voltage level without changing the amplitude The Serial BERT handles this by keeping Vamp constant when V is changed and V V are accordingly adjusted Voltage Level Restrictions The Serial BERT cannot generate a signal that has voltage levels out of range If you try to enter a value for one parameter that would put another parameter out of limits the Serial BERT rejects the change This could happen for example if Vio is already at the minimum and you try to lower either of Vj or Vs or increase Vampt Understanding Delay and Crossover The Serial BERT provides the possibility of modifying the output data signal by varying the signal s delay to the clock signal and the signal s crossover Delay The exact time delay through a test setup can vary The delay function allows you to compensate for this by adjusting the frequency reference for the data outputs This varies the phase relationship of the data and clock outputs causes the data pulse to have a certain time delay after the clock pulse The higher the delay the greater the time difference between the clock signal and the data signal Agilent J BERT N4903B High Performance Seri
83. Unit sus fens ELLEN ES Again from the Config window corrosponding to N4916B clear the De Emphasis function check box It will disable the de emphasis function If you could proceed until here the connection of the user software to the N4916B is ok Settings for De Emphasis Tests The following are the settings to test the output of the de emphasis Preset Instrument State Set the PRBS 2 7 1 Set the Data Rate to 3GB s Set the Data Offset to 0V Set Data Amplitude to 400mV Enable De Emphasis in the Config window Set Pre Cursor to 12dB Set Post Cursor 1 to 6dB Set Post Cursor 2 to 3dB Check the Output of the N4916B De Emphasis Part Using a Scope ci A amp w N 504 Connect the scope Data out of N4916B Connect the Trigger In of Scope to Trigger Out of PG Terminate N4916B Data Out with 500hm Terminate PG Trigger Out with 500hm Terminate PG Data Out with 500hm Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 Then the oscilloscope should show an output eye like the following If you see a display like above everything is ok If you see merely one fixed level either the N4916B or the output of the Serial BERT is defective Check the Output of the N4916B De Emphasis Part Using the Error Detector If you have connected the N4916B to a Serial BERT and have no oscilloscope at hand you can use the error detector for checking the output of the N4916B 1 Conn
84. above The option J10 includes all the jitter sources required for jitter tolerance tests For details see Jitter Tolerance Tests Concepts on page 395 The delay control input is equipped with overvoltage protection If the applied voltage is too high the input is automatically disabled and the following dialog appears Agilent J BERT N4903B High Performance Serial BERT 155 4 Setting up the Pattern Generator Delay Control Protection Input Signal at the Delay Control Input is out of hardware specification Please adjust signal levels Enable Delay Control Input again cmm ose You can enable the input after reducing the voltage to a safe level Delay Control Input Procedures The following procedure does not apply to Serial BERT on which the calibrated and NOTE LOQUERE ae integrated jitter injection option J10 is installed For such instruments refer to Jitter Setup Procedures on page 407 To set up a jitter tolerance test on an instrument without option J10 1 2 3 4 Press PG Data Setup in the PG Setup menu Disable the Delay Control In port by clearing the Delay Control Input checkbox Connect the device that supplies the jitter signal to the Delay Control In port Enable the port again by selecting the Delay Control Input checkbox While monitoring the BER you can use this checkbox to switch jitter on and off and TIP i view the effect on the BER Delay Control Input Re
85. and accessible Pattern Generator Auto Calibration This test runs a timing calibration that calibrates the clock to data alignment of the pattern generator It may take about 45 minutes to complete Error Detector Bios Self Test This test is checking all the modules chips and registers of the error detector more thoroughly than the Power Up test It may take about 5 minutes to complete Error Detector Power Up Messages This test is automatically run after switching on the instrument It mainly checks if all important modules chips and registers of the error detector are available and accessible Error Detector Auto Calibration This test runs a timing calibration that calibrates the clock to data alignment of the error detector It may take about 45 minutes to complete Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 Select All Unselect All Use the Select All button to select all tests in the Self Test Options dialog box After clicking the button it becomes the Unselect All button Clicking the button again will then unselect all tests and display the Select All button again Run Test After selecting at least one test use this button to start the test If you selected tests that need a longer time to complete a message box opens Confirm your selection by clicking OK A second message box is informing you while the test is still running and closes when finished Message Buttons
86. at the end How to Save and Recall To Save the following measurements Eye Diagram Agilent J BERT N4903B High Performance Serial BERT 329 6 Advanced Analysis Property Page Legend States Left Panel Values Automated Eye Parameter Measurement 1 Click on the Analysis Menu 2 Click on Save Measurement Eye Diagram Output Timing Output Levels Error Location Capture Fast Eye Measurement Spectral Jitter Eye Opening ee hess Load Measurement Export Data To Recall the measurements 1 Click on the Analysis Menu 2 Click on Load Measurement Eye Diagram Output Timing Output Levels Error Location Capture Fast Eye Measurement Spectral Jitter Eye Opening Save Measurement ed escena Export Data How to Export Fetched Data You can export various counters to an exported file location 1 Click on Analysis Menu 2 Click on Export Data 330 Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Eye Diagram Output Timing Output Levels Error Location Capture Fast Eye Measurement Spectral Jitter Eye Opening Save Measurement Load Measurement Export Data When you click on Export Data the following screen appears ac Export File Ic Program FilesS amp gilentNN 48035 SGUISE port txt Browse Clipboard ox Cova The Clipboard option saves the data on a clipboard The Format Options takes y
87. available on the front side of the ESM of chassis with the USB port available on the rear side of J BERT N4903B On connecting both the instruments with USB cable if the Found New Hardware Wizard appears as shown in the figure you may need to install the following drivers USB driver AX88178 driver USB to Ethernet Controller To install the USB driver choose Install the software automatically Recommended and click Next to start installing the device then click Finish once the wizard completes the process To install the AX88178 driver download the latest version of AX88178 driver from the following webpage www asix com tw After downloading simply double click on the exe file to begin installation You will receive a confirmation that the process has begun and another upon successful completion If you do not install the AX88178 driver you will still be able to use M8061A However in this case the Ethernet port over USB connectivity on the ESM will not get configured Agilent recommends you to establish a USB connection between the USB ports of M8061A and J BERT N4903B via a powered USB 2 0 hub to prevent the USB port shutdown due to excessive power drawn by the USB devices connected at the J BERT N4903B USB ports Connecting M8061A to N4903B Pattern Generator and DUT After the USB connection has been established you can connect the M8061A to the N4903B Pattern Generator and DUT Follow the given steps
88. between which the measured result turns from green to red Toreduce the measurement time without sacrificing precision the Jitter Tolerance Characterization measurement offers two additional search methods Binary search When binary search is enabled the measurement starts with the maximum jitter amplitude 100 96 If the result is red measured BER higher than target BER the search goes downwards The next amplitude is set to 50 96 If the new result is still red the next amplitude is set to 25 96 But if the new result is green the search goes upwards and the next amplitude is set to 75 96 Depending on the result the search algorithm changes the direction and continually halves the step size This continues until the specified accuracy the final step size is reached Binary is the fastest search method It approaches the amplitude of interest from both sides However binary search can lead to incorrect results if the device to be tested shows hysteresis Extended binary search This algorithm combines binary with linear downward search Extended binary search starts like the binary search Depending on the result the search algorithm changes the direction and continually halves the step size This continues until the specified coarse step size is reached Then the coarse step size is maintained until a transition from green to red is found After that has happened the algorithm changes the direction and uses fine
89. can Use Edit submenu for the cut copy and paste functionality Set Date and Time Minimize GUI View the SCPI output window Set the touchscreen feature on off Change the GPIB address Setthe onscreen keyboard feature on off Open the Self Test Option dialog Get information on the Licenses Provides help to activate Web Control For more information on Utilities refer to the section Customizing the Instrument Concepts on page 525 The Help menu launches the online help which provides information on a particular GUI item You have the following possibilities for opening this Help from the Serial BERT user interface Context sensitive Help for all elements of the GUI To get quick information on the meaning of a particular GUI item select What s This from the Help menu of the GUI and click on the item in question Help on a Serial BERT window If you need more information on one of the Serial BERT windows select Help On from the Help menu This Help opens with information on the current Serial BERT window Switching to the Help Select Contents from the Help menu to switch to this Help The last topic you visited is displayed Agilent J BERT N4903B High Performance Serial BERT Planningthe Test 1 In addition the About button on the Help displays the license information of the Serial BERT Which Test is Appropriate Which Test is Appropriate Concepts These topics help you to
90. can enter a comment The condition is automatically added by the system You can also read the full comment or condition in case only part of the text is displayed in the result table Jitter Tolerance Compliance Reference Standard Selection Jitter Tolerance Compliance determines whether a device under test maintains a desired BER at defined jitter levels The jitter levels can be derived from predefined standards Standard Specification The drop down list contains all predefined standards including user defined standards The shape of the selected standard is shown in the preview area The Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 frequency range and the jitter amplitude range of the standard is given in the preview A description is displayed at the bottom of the Standard tab The last entry in the list is Browse Select this entry to load a file that contains a user defined standard See Setting the Standard on page 454 for details Margin To aggravate test conditions the standard s test jitter amplitude can be increased by Margin percent The display shows a black line for the standard and a green line for the new jitter amplitudes Auto Frequencies Number of Steps Select the Auto Frequencies checkbox to set frequencies that are recommended for the selected standard This list contains for example the edge frequencies of the standard TIP You can change single frequency po
91. choose among Downspread Centerspread and Upspread deviation type The green LED on the button indicates the currently selected deviation type Agilent J BERT N4903B High Performance Serial BERT 417 8 Jitter Tolerance Tests 418 Profile The selected deviation type is also indicated in the jitter composition shown in the Jiter Setup area The following are the deviation type indications Downsperad fh N paz weoi Centerspread e 123 96 g 33 0 kHz i Upspread Ao soe e Arbitary fh N 0 123 Mw 33 0 kHz You can click on the corrosponding button to choose between the Triangular and Arbitrary profile The green LED on the button indicates the currently selected profile 5 For Arbitary profile you need to specify the Arbitary Waveform file To do so click on the Browse button and select the file from the shown list Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 0 1225 96 0 245 33 KHz 998 955 Mb s 996 511 Mb s CAN4903B SSCIProfiles TraA txt Residual Spread Spectrum Clock The Residual Spread Spectrum Clock rSSC is generated by modulating the clock that is used for data generation It can therefore not be enabled if SJ or SSC is active F Afer oe Residual Spread Spectrum Clock is characterized by Amplitude peak to peak Frequency There is a dependency between frequency and acceptable amplitude A graph in
92. compliance file See User Defined Standards on page 452 Optimal Measurement Setup For details see Optimal Measurement Setup on page 465 SelectPoint For details see Select Point on page 466 HTML Report For details see HTML Report on page 466 Jitter Tolerance Compliance Jitter Tolerance Compliance Concepts The Jitter Tolerance Compliance test is used to ensure that the jitter tolerance of the device under test complies with a certain standard 446 Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 The Jitter Tolerance Compliance measurement is a software option that requires a NOTE ji MP license For details on how to obtain and install such licenses refer to Obtaining a License on page 535 Measurement Principle The Optical Internetworking Forum OIF and other institutions have proposed and published standards for testing the performance of data receivers and receiver circuits in the presence of jitter These standards prescribe the required jitter tolerance over jitter frequency The jitter tolerance for each frequency point is defined by a jitter amplitude normalized to the bit rate and hence specified in Unit Intervals UI and a BER threshold which must not be exceeded Jitter Tolerance Curves A simple example is shown in the figure below Jitter amplitude in Ul log scale Jitter tolerance curve of a device schematic Jitter frequency log scale
93. connected the N4876A to a Serial BERT and have no oscilloscope at hand you can use the error detector for checking the output of the N4876A 1 Connect the ED Data In to the Data Out of N4876A 2 Connect the ED CLK In to the PG CLK Out Agilent J BERT N4903B High Performance Serial BERT 515 9 Solving Problems 516 NOTE 3 Terminate PG Data Out with 500hm 4 Terminate PG Clock Out with 500hm 5 Ensure that the error detector follows the generated pattern Press the Auto Align button The error detector should finish the auto alignment successfully and report a BER of 0 000 Manually adjust the ED sample delay by 166 ps Press Sync Now The error detector should report a BER of 0 000 Check the Output of the Pattern Generator with a Scope If the N4876A does not produce a signal check the Data Out of the Serial BERT Connect the Data Output of the pattern generator to an oscilloscope As long as the N4876A is in connected state after pressing the Enable button the pattern generator s output is programmed to a fixed level The amplitude is set to 0 8 V the offset to 0 4 V The oscilloscope should show the following picture for Data Out as well as for Aux Data Out EE A x Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 If you see only a straight line ensure that the pattern generator is not disabled and that it is set up to generate a pattern e g PRB
94. coupling is set to AC You can set the deemphasis using the following parameters De emphasis Unit Used to select the de emphasis unit of pre cursors and post cursors in dB Decibel or PERC Percentage Pre Cursor 1 2 Used to set the pre cursor 1 2 to the given value as a percentage or ratio in dB of the previous bit s amplitude Post Cursor 1 5 Used to set the post cursor 1 5 to the given value as a percentage or ratio in dB of the previous bit s amplitude You can set the interference using the following parameters CMI State Used to enable disable the Common Mode Interference CMI Gain Used to set the Common Mode Interference Gain DMI State Used to enable disable the Differential Mode Interference e DMI Gain Used to set the Differential Mode Interference Gain Agilent J BERT N4903B High Performance Serial BERT 57 2 Setting up External Instrument s 58 ClkGen Tab Electrical Idle In The ClkGen tab allows you to select clock source It provides the following parameters Clock Source Used to select the clock source to either CLK IN External clock or AUX CLK IN Internal clock Input Timing Adjustment Click Adjustment to start the input timming adjustment For more information on input timing adjustment of M8061A refer to the section M8061A System Setup Adjustment on page 492 This Electrical Idle In tab allows you to set the Electrical Idle In It provides the follow
95. delay value This knob can be used at any time while in any window Clock Falling Edge Activate this checkbox to move the sampling point from the rising edge to the falling edge of the clock signal This function is useful when the maximum data input delay adjustment is not large enough to move the sampling point to the desired location This occurs at very low frequencies only See Bit Rate Range on page 176 for more information Auto Align Click this button to automatically set the optimum sampling point This routine will not stop if the optimum sampling point cannot be found If no optimum sampling point is found after a reasonable time you can click Cancel The status message bar will indicate if this function is unsuccessful The following settings may affect the result of the auto align function Data Inverted BER Threshold NOTE In SER mode the Auto Align with CDR enabled supports the ED frequency above 11 5 Gb s and performs Data Centre with appropriate threshold settings TIP Alternatively you can also press the Auto Align button on the front panel Agilent J BERT N4903B High Performance Serial BERT 199 5 Setting up the Error Detector 200 NOTE NOTE TIP NOTE Sync Now Click this button to manually start a pattern synchronization See What Type of Synchronization Should You Use on page 208 for information on when you should use this function Note that this button is not available in bit r
96. display or listen to the BER warning tones Higher BERs may indicate that you are measuring at the edges of the data eye NOTE Pattern synchronization can be lost during sampling point adjustment If Auto Sync is enabled the Serial BERT will resynchronize the patterns automatically If the manual synchronization mode is selected you need to synchronize the error detector by clicking the Pattern Sync icon and then the Sync Now button How to Adjust the 0 1 Decision Threshold 1 Click Sampling Point Setup menu item from the ED Setup submenu 2 Clear the Avg 0 1 Threshold checkbox 3 Click 0 1 Threshold Center to move the sampling point vertically towards the center of the eye at the selected delay You may click Cancel at any time 4 Now you can use the numeric keypad or front panel knob to manually adjust the value in the 0 1 Threshold field The Sample Point in the eye diagram moves vertically as you change the value TIP While adjusting the 0 1 threshold you can monitor the BER bar on the analyzer display or listen to the BER warning tones Higher BERs may indicate that you are measuring at the edges of the data eye NOTE Pattern synchronization can be lost during sampling point adjustment If Auto Sync is enabled the Serial BERT will resynchronize the patterns automatically If the manual synchronization mode is selected you need to synchronize the error detector by clicking the Pattern Sync icon and then the Sync Now button
97. ended normal Modulates the data signal that is connected to P1 or P2 of the Interference Channel Single ended complement Modulates the data signal that is connected to P1 or P2 of the Interference Channel Jitter Tolerance Characterization Jitter Tolerance Characterization Concepts During a jitter tolerance characterization test jitter is applied sequentially at a number of frequency points in the frequency range of interest The receiver checks for the onset of transmission errors and increments decrements the jitter amplitude using a control algorithm re testing for errors at one frequency point until the target BER threshold is found This is repeated for each subsequent jitter frequency step to provide a plot of the DUT tolerance Agilent J BERT N4903B High Performance Serial BERT 433 8 Jitter Tolerance Tests 434 Jitter amplitude in UI log scale cs Jittertolerance curve of a device schematic Jitter frequency log Scale Bandwidth Green points BER below limit and red crosses BER above limit are used to display the results After a measurement has exceeded the target BER or when moving to the next frequency point a measurement settling or relax time can be applied This allows the DUT to settle after changes and before the next measurement this is important because some devices may need time to recover from error conditions caused by jitter exceeding the margin during the search for the threshol
98. enter the pre cursor value The value is interpreted as dB or depending of the specified mode NOTE This is applicable ony for N4916B Post Cursor 1 Use this text box to enter the pre cursor value The value is interpreted as dB or depending of the specified mode Post Cursor2 Use this text box to enter the pre cursor value The value is interpreted as dB or depending of the specified mode NOTE This is applicable ony for N4916B Unit Use this text box to toggle between db and percent Toggling does not change the value Bit Rate Bit Rate Concepts Bit rate is defined as bits per second The pattern generator s clock rate sets the bit rate and serves as the frequency reference for the Data Aux Data Clock and Trigger Ref Clock outputs your device and the error detector if it receives its clock from the pattern generator It can be generated internally or supplied from an external source When to Use an External Clock Source The Serial BERT s internal clock can be used for most testing purposes There are some circumstances however when an external clock source is required Synchronization with an external clock Agilent J BERT N4903B High Performance Serial BERT 139 4 Setting up the Pattern Generator NOTE 140 The Serial BERT can be connected to an external clock to allow it to run as part of a larger external system e Use of a modulated clock A frequency or delay modulated clock can be used to
99. expected output pattern of your device is internally generated in the error detector to provide a reference Selecting a pattern is the first step in setting up a BER measurement The Serial BERT provides various patterns to fulfill most standard testing needs In 8B 10B Symbol Comparison and Bit Comparison without PCle3 SKPOS modes only twelve different user patterns UPAT1 UPAT12 and user pattern files are allowed as analyser pattern How does the Serial BERT Generate Patterns The Serial BERT has two possibilities for generating patterns either by directly generating them from the hardware or by loading patterns into the memory and generating the output according to the pattern in the memory These memory based patterns can be either user patterns file based or software generated patterns The hardware generated patterns available with the Serial BERT are PRBS 2 n 1 patterns They provide long high speed random traffic patterns that can be used for testing a wide range of devices These patterns cannot be manipulated in any way The memory based approach is more flexible It makes it possible to define any pattern within the physical limits of the memory for defining the output The Serial BERT provides two possibilities for defining the output by defining user patterns or by selecting a software generated pattern Agilent Technologies 61 3 Setting up Patterns 62 User Patterns Software Generated Pat
100. eye appears with the Automated Eye Parameter Measurement and the results keep on improving in accuracy as more number of points get measured and represented on the diagram along with the deeper BER values The BER threshold in the Graph Tab decides how long the measurements will be done at the boundaries of the eye Once the measured points exceed the BER Threshold defined by the user the measurement is done deep inside the eye This is a default setup for persistence Minimum With minimum a reasonably clear eye with transition appears along with all the Automated Eye Parameter Measurement and automatically the measurement restarts showing only the refined eye diagram To stop the measurements click the abort button Secs With Secs the measurement runs for the specified time and the measurement automatically restarts showing only the refined eye diagram To stop the measurement click the abort button Sets the transition of the eye on the plot Middle of eye Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 With middle of eye the center of the eye coincides with the center of the screen Transition With transition the transition of the eye coincides with the center of the screen Number of eyes Set the criteria for the number of eyes 15ui This option displays 1 5 eyes on the graph The transitions of the complete eye are placed such that they show 0 25 eyes
101. files to determine the location of occurring bit errors for further root cause analysis While accumulation tests are running you should not modify the measurement setup as the measured bit errors do not represent the performance of your DUT under real circumstances The period of time can be set by manual start and stop commands an absolute time setting or the time it takes to measure a specified number of bits or bit errors The accumulation period should be long enough to make a statistically valid BER measurement See Test Times and Confidence Levels on page 27 for details Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 During measurement logging the Serial BERT logs data in ten second intervals NOTE c ibd ae Your log file may be missing up to the last ten seconds of data To avoid this condition accumulate for 10 seconds longer than desired Accumulated Measurements Procedures This section describes how to run accumulated measurements and how to analyze the results How to Run Accumulated Measurements To run accumulated measurements 1 Ensure that the instrument is set up correctly and that the error detector is already in synchronization 2 Click the Accumulation Setup menu item from the ED Setup submenu to check the settings made for the data accumulation 93 Accumulation Setup See Error Accumulation Procedures on page 220 for details 3 To begin accumulation press the Star
102. find the optimum sampling point if it falls in one of the gaps Clock Setup Procedures To set the clock rate 1 2 Press Clock Setup in the ED Setup menu Specify whether you are using an External Clock Source or Clock Data Recovery CDR If you are using CDR you need to approximately specify the clock rate in the Clock Rate field with an accuracy of 0 01 If you are measuring bursts deactivate Auto Threshold and set the threshold manually You can also press the Measure button to measure the current DC level at the CDR input and set the measured value as CDR threshold Clock rates that are frequently used can be saved as presets in the Preset list To select a clock rate from the Preset list click on it The clock rate appears in the Clock Rate field Specify the Loop Bandwidth within the range of 100 KHz to 12MHz Specify the CDRFine Adjustand Peaking values When using a Preset this is only required when you want to change these values Enable Spread Spectrum and adjust the deviation and deviation type as required Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 8 Either enter the transition density manually or run a measurement to update this value Clock Setup Reference The Clock Setup window contains the following elements Error Detector Bit Rate The bit rate of the data reaching the error detector is displayed in the lower right corner of the user in
103. from one jitter amplitude to the next You can choose between two Verification Methods Setting a Confidence Level is generally recommended This ensures that neither too few nor too many bits are captured See also Target BER on page 435 If you enable Absolute Bits Errors the measurement proceeds to the next jitter amplitude as soon as either the number of received bits or the number of errors is reached Ensure that you capture a number of bits that suffices to verify with the desired probability that a measured point has a BER below the Target Bit Error Ratio This is the wait time that takes effect after the measured BER is higher than the target BER It allows the DUT to recover from the situation The unit is seconds The next measurement starts after that time Search Specification On this page of the Jitter Tolerance Properties dialog you specify how the measurement proceeds upwards or downwards from one measured point to the Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 next The chosen method impacts the measurement duration The specified parameters influence both measurement duration and precision You can choose one of six search methods Downwards linear Downwards logarithmic Upwards linear Upwards logarithmic Binary search Extended binary For a description of these methods refer to Vertical Search Methods on page 438 Linear search For linear search you can set the
104. get the clock from the data stream The Synchronization Time is the time the error detector requires to synchronize to the pattern This time depends on the pattern type PRBS or memory based The Valid after Gate indicates the minimum length of data that has to arrive at Data In after the Gate In signal is deactivated There is no requirement how long a signal should be available at Data In before gating starts the data should simply be stable A pause is however required after the gating to allow the error detector to evaluate the results This time is known as Gate Passive The following is a summary of the values of these parameters for the Serial BERT Table 20 Parameter Values Bit Count Time Min 1536 bits Max 4 Gbits The optimal length is dependent on the BER threshold CDR Settling Time CDR mode 2 microseconds Non CDR mode 0 Synchronization Time Hardware PRBS 1536 bits Memory based patterns Min 15 kbits depends on when and how often the unique word occurs Valid after Gate CDR mode 1 5 microseconds Non CDR mode 1536 bits Gate Passive CDR mode 2560 bits or 1 5 microseconds whichever is larger Non CDR mode 2560 bits The following values can be derived from these parameters Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 Table 21 Parameter Values Begin Margin CDR Settling Time Synchronization Time End Margin Valid after Gate Bit Count Time Bur
105. in an incremental way showing the coarse picture quickly and then refining it further 316 Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 The BER Threshold is configurable The BER Threshold is the level up to which the signal is represented as waveform and BER values below this threshold are represented as contours Waveform v Show Waveform F Snooth Waveform Graphic BER Threshold 1e 1 BER Contour Contour is a curve connecting points where the BER has a same particular value The contour graph is plotted within the Eye Diagram and it helps to determine the Eye Opening at deep bit error rates such as 1e 10 1e 12 and so forth Depending upon the user s requirement the deep BERs can be calculated The extrapolated eye contour lines display the eye opening for lower BER levels such as 1e 15 The BER is displayed as a function of sampling delay and sampling threshold The eye opening is one of the main characteristics of a high speed device Contours iv Show Contour iv Show Legend The legend on the right side of the screen represents the contour BER values in different colors When you select a BER value from the legend it displays the contour for the selected BER in the same color Initially all the BER values on the legend are struck across When the contour is measured the BER value changes into normal font and when the contour is extrapolated the BER value changes into Agilent J BERT N49
106. in personal injury or death Do not proceed beyond a WARNING notice until the indicated conditions are fully understood and met Contents 1 Planning the Test Planning the Test Concepts 11 Introduction to the Serial BERT 11 Introduction to the Serial BERT Concepts 11 Navigating the Serial BERT GUI 18 Navigating the Serial BERT GUI Concepts 18 Which Test is Appropriate 25 Which Test is Appropriate Concepts 25 Connecting the DUT 29 Connecting the DUT Concepts 29 2 Setting up External Instrument s Setting up External Instrument s Concepts 37 Setting up External Instrument s Procedure 42 Setting up External Instrument s Reference 51 3 Setting up Patterns Setting up Patterns Concepts 61 User Patterns 65 User Patterns Concepts 65 User Patterns Procedures 71 User Patterns Reference 76 Software Generated PRBS 91 Software Generated PRBS Concepts 91 Software Generated PRBS Procedures 93 Software Generated PRBS Reference 93 Hardware Generated PRBS 93 Hardware Generated PRBS Concepts 93 Hardware Generated PRBS Procedures 95 Hardware Generated PRBS Reference 95 Agilent J BERT N4903B High Performance Serial BERT User Defined Sequences 96 User Defined Sequences Concepts 96 User Defined Sequences Procedures 99 User Defined Sequences Reference 101 Sequencer Pattern Editor 108 Sequencer Pattern Editor Concepts 108 Sequencer Pattern Editor
107. internal sources Specify the Jitter Components for PJ Sweep This section shows you how to specify the Jitter components for PJ Sweep L 10 0000 MHz You will not see the parameter Amplitude and Frequency in case the PJ Sweep is already on Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 To specify the jitter components 1 Activate a jitter source press the corresponding button 2 Press Edit button to access all parameters of the corresponding jitter type P k 4 If you want to specify the jitter components for Constant Amplitude Sweep skip steps 5 to 6 However if you want to specify the jitter components for Variable Amplitude Sweep skip steps 3 to 4 and go to step 5 3 Activate Constant Amplitude Sweep press Constant Amplitude Sweep button This opens a window which allows you to set all parameters available for the given jitter type E fikne fto fison Lm 4 4 Specify the following parameters Amplitude Frequency Range Sweep Time Nr of Steps For more information on parameters see Periodic Jitter 2 Constant Amplitude Sweep Parameters on page 423 5 Activate Variable Amplitude Sweep press Variable Amplitude Sweep button Agilent J BERT N4903B High Performance Serial BERT 411 8 Jitter Tolerance Tests 412 NOTE This opens a window which allows you to set all parameters available for the given jitter type PCle
108. is used to measure the timing and jitter behavior for a device under test DUT It uses a bit error rate BER measurement to evaluate the shape of the eye for the output signal of the DUT It also analyzes the jitter separates the random jitter and deterministic jitter components and estimates the total jitter A direct result is the determination of the optimum sampling point delay for receiving data from the DUT with maximum confidence DUT Output Timing Jitter includes the Fast Total Jitter measurement that can be used to measure the total jitter for devices which generate a very low error density in a reasonable time span The Output Levels measurement allows you to characterize the behavior of the output levels of a device under test DUT The sampling delay is fixed The error detector s decision threshold is automatically swept within a user defined range A direct result is the determination of the optimum decision threshold level for receiving data from the DUT with maximum confidence The Eye Opening measurement generates a three dimensional bit error rate BER diagram as a function of the sample delay and the sample threshold With this measurement the complete eye of the DUT output signal is measured The results comprise the voltage and timing of the eye opening and the optimum sampling point The Error Location Capture measurement allows to capture the actual position of errored bits in a memory based pattern The instru
109. linear downward steps of specified size to localize the transition precisely Extended binary search takes more time than binary search but leads to reproducible results for devices with hysteresis Agilent J BERT N4903B High Performance Serial BERT 437 8 Jitter Tolerance Tests Jitter Tolerance Characterization Procedures To access the Jitter Tolerance Characterization measurement press the corresponding menu item from the Jitter submenu To set up a Jitter Tolerance Characterization test press the Properties button Properties This opens the Jitter Tolerance Properties dialog Setting the Frequency Range Choose Auto or Manual mode Jitter Tolerance Properties Frequency Search BER Auto Start Frequency fi kHz Stop Frequency 80 000 MHz Nr of Steps 20 C Manual Sequence Advance I Show Compliance Curve Cancel Apply Help In Auto mode the jitter frequencies are automatically calculated 1 Enter the Start Frequency 2 Enter the Stop Frequency Start and stop frequency define the measurement s frequency range 438 Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 3 Setthe Number of Steps between start and stop The Number of Steps yields a set of logarithmically equidistant frequencies In Manual mode the automatically calculated jitter frequencies can be edited You can also enter your own comma separated list of frequencies The Sequence Advanc
110. markers in the marker readout Pressing the Reset Markers button will set the markers back to the default positions Linear markers are available for the BER vs Thresholdgraph and the OBER vs Threshold graph A gaussian marker is available for the dBER vs Threshold graph See Marker Definitions on page 236 for details Several zoom factors are available When you show the zoom graph you can also allow the zoom graph to track the mouse or your finger if you are working directly on the Serial BERT If you want to see the points that have actually been measured choose Show Measured Points Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 NOTE If you have a mouse connected to your Serial BERT you can access many parameters and display options conveniently from the context menu This looks as shown below iene e BER vs Threshold Display Options dBER vs Threshold QBER vs Threshold Properties Log Scale Linear Scale Linear Ratio dB Ratio v Show Measured Points Explanation of the Numerical Results The result parameters are divided into three groups Level Results The level result parameters are summarized in the following table Table 27 Parameter Name Pass Fail High Level min max Low Level min max Mean Level min max Amplitude min max Threshold Margin min For all parameters pass fail limits can be set as indicated in the table The level param
111. may be different The table in the top half of the Eye Results window additionally lists the following values Eye Voltage Center One of two components of the sampling point the 0 1 decision threshold eye voltage center is displayed here Eye Data Input Delay One of two components of the sampling point the data input delay is displayed here Alignment Threshold The alignment BER threshold is displayed here This threshold was used to define the eye edges and the above eye results Delta Error Ratio The BER measured during the previous 200 ms is displayed here Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests Jitter Tolerance Tests Concepts NOTE This chapter refers only to Serial BERT instruments on which the calibrated and integrated jitter injection option J10 is installed A jitter tolerance measurement is used to determine the ability of a device or system to maintain communication quality in the presence of jitter It comes in two varieties Jitter Tolerance Characterization determines the jitter levels where the device under test can no longer maintain a desired bit error ratio BER Jitter Tolerance Compliance verifies that the device under test is able to maintain a BER level at pre defined jitter levels and jitter frequencies as defined by a standard The basic setup of a jitter tolerance test is illustrated below Device under test Pattern gt gt Error Gen
112. not intersect with the bathtub curve not applicable indicates that the value could be calculated however is not shown because quality criteria are not met This is the case for example for RJ DJ results when the r 2 value of one or both edges is 0 75 Even though RJ DJ values can be calculated in this case they are not shown because confidence in the results is too low Understanding the Q Factor Results The variable decision threshold method used by the Output Levels measurement makes it possible to determine the Q factor of a signal The Q factor method and the related calculations have been published under Margin Measurements in Optical Amplifier Systems by Neal S Bergano F W Kerfoot and C R Davidson in IEEE Photonics Technology Letters Vol b No 3 March 1993 Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Mathematical Background Bit errors are caused by noise and the Q factor describes the signal to noise ratio at the decision circuit It is possible to calculate the Q factor from a limited number of measured BER vs threshold data points It is also possible to calculate expected bit error rates from the Q factor This is a method for predicting very low bit error rates typically below 10 11 that would take a long time to measure The Q factor is calculated as Hi Ho 0 0o where p ois the mean level of the 1 and 0 rails respectively and o 9 is the standard deviatio
113. of all instruments connected to the Serial BERT Confirm the presence of N4876A enrty in that list 5 Corresponding to N4876A entry select the Multiplexer function check box from the given options under Enable Function column It opens the Multiplexer Connection dialog Agilent J BERT N4903B High Performance Serial BERT 513 9 Solving Problems Click Enable button It enables N4876A 28 Gb s Multiplexer 2 1 that is connected between data out of Serial BERT and the DUT Enable Once the output levels of Data Out that you set on the PG Output screen will control the output levels of the N4876A s Data Output The level setting of Aux Data will be disabled on the PG Output screen when the multiplexer function is enabled Again from the Config window corrosponding to N4876A clear the Multiplexer function check box It will disable the multiplexer function and the PG Output screen will allow to control the output levels of Data and Aux Data at the connectors of the PG again If you could proceed until here the connection of the user software to the N4876A is OK Settings for Multiplexer Tests The following are the settings to testthe output of the N4876A 28 Gb s Multiplexer Preset Instrument State Set the PRBS 2 7 1 Set the Data Rate to 3GB s Set the Data Offset to 0V Set Data Amplitude to 400mV Enable Multiplexer in the Config window Check the Output of the N4876A Multiplexer Using a Scope ci A C N 514 Connec
114. of the chosen delay line Frequency The maximum Frequency of the periodic jitter source depends on the chosen waveform Refer to the technical specifications Waveform You can switch between sine rectangular or triangular jitter source waveforms Periodic Jitter 2 Parameters The Periodic Jitter 2 has the following modes Periodic Jitter 2 Constant Jitter Periodic Jitter 2 Constant Amplitude Sweep Periodic Jitter 2 Variable Amplitude Sweep Agilent J BERT N4903B High Performance Serial BERT 421 8 Jitter Tolerance Tests 422 NOTE NOTE You can click on the corrosponding button to choose among the different modes of Periodic Jitter 2 The green LED on the button indicates the currently selected mode of Periodic Jitter 2 PJ2 Constant Jitter PJ2 Constant Amplitude Sweep PJ2 Variable Amplitude Sweep The ability of sweeping the PJ settings supports testing DUTs according to the PCle 2 0 and PCle 3 0 standard without the need of providing a jitter modulation signal externally Periodic Jitter 2 Constant Jitter Parameters The Periodic Jitter 2 Constant Jitter is characterized by Amplitude Frequency e Waveform AL sur 10 0000 we You will not see the parameter Amplitude and Frequency in case the PJ Sweep is enabled The setting for the jitter Amplitude is shared with the Constant Amplitude Sweep mode A graph indicates whether the chosen setting of frequency and am
115. on page 131 Manually move the sampling point See How to Set the Optimum Sampling Point Automatically on page 196 for details Make adjustments to your device This can be used to see how changes to your device affect the BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS Bit Comparison without USB3 1 SKPOS You can determine what adjustments improve or degrade the BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS Bit Comparison without USB3 1 SKPOS Agilent J BERT N4903B High Performance Serial BERT 365 7 Evaluating Results 366 Instantaneous Measurements Reference Accumulated Actual Button Error Count Accum Error Count NOTE This section describes all elements of the BER Results window and 8B 10B Comparison Results windows and their various status indicators BER Results Window This window displays the actual BER or cumulative AccumBER BER results a BER 0 000 Remote Error Counts The BER is the current BER calculated upon a period of 200ms The AccumBER shows either the accumulated BER of the current accumulation or if no accumulation is running the results of the most recent accumulation This enables you to monitor real time BER behavior as you do things such as to manually adjust the sampling point to add errors or to make adjustments to your device Click this button to toggle between BER or AccumBER results This area displays
116. output s occurs when the amplifier detects an overload condition which is caused by the externally connected circuitry DUT In the incorrect termination detection state the Auto Range Amplitude Range Amplitude High Low Offset Coupling Termination Model Termination Voltage and CMI state are re programmed to safe values The GUI shows an error message listing the effected outputs and highlights the Outputs On status indicator In the M8061A window the parameters that currently do not match the data out settings are highlighted in yellow BER 0 000 Remote The parameters can be changed to configure the output according to the DUT s termination requirements The changes are applied only when the output protection circuit button is clicked Understanding the Output Level Parameters The following figure shows the parameters of a Data Aux Data Clock or Trigger Ref Clock output signal Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 Vii Vani 5 y 0v Y ofst V Y As shown in this figure the signal output levels have the following components Vj is the upper voltage level of the signal Vis the lower voltage level of the signal Vais the offset of the average voltage level from 0 V Vampt S the amplitude of the signal When adjusting the output levels it is important to understand the concept of how the Serial BERT handles voltages Changing the Amplitude
117. position in the pattern Depending on the type of pattern currently loaded in the pattern generator you have to specify the trigger position either according to Bit Position or N bit Trigger Pattern Bit Position If a memory based pattern is loaded in the pattern generator you have to specify the bit in the pattern that sends the trigger The bit position can be in the range from 0 to Pattern Length 1 For pattern lengths that are divisible by 512 such as 2 n PRBS patterns the trigger repetition rate is one pulse per pattern repetition Bit Pos Data Out A A A Trigger Out For pattern lengths that are not divisible by 512 such as some memory based patterns a trigger signal is sent once at the corresponding bit when the pattern reaches the 512 bit boundary If for example the pattern is 384 bits long a trigger signal is sent every 4th pattern repetition Bit Pos A0 A1 A2 DataOut A A A IAIAIAIAIAIAI A AL A Trigger Out If the bit position is outside the pattern i e the entered number greater than the pattern length 1 the entry is adjusted to the pattern length 1 i e the trigger is set to the last bit in the pattern If you load a pattern that is shorter than the entry in this field the trigger will be set to the last bit in the pattern N bit Trigger Pattern This option is used for hardware generated PRBS patterns You can specify a patt
118. that they are compatible with a variety of logic families With respect to Data Out Data Out has inverted logic Clock Out and Clock Out The clock outputs serve as frequency bit rate references and can be set up so that they are compatible with a variety of logic families With respect to Clock Out Clock Out has inverted logic Trigger Ref Clock Out and Trigger Ref Clock Out This port allows you to connect a trigger for another device for example an oscilloscope and can be set up so that they are compatible with a variety of logic families With respect to Trigger Ref Clock Out Trigger Ref Clock Out has inverted logic Aux Data Out and Aux Data Out This port allows you to generate a pulse with a frequency that is a fraction of the present clock and can be set up so that they are compatible with a variety of logic families With respect to Aux Data out Aux Data Out has inverted logic Note that this output is not phase synchronized with the clock Aux Clk Out This output port is intended as clock input for N4916B De Emphasis signal converter The complementary outputs can be used when additional output capability is needed for an instrument such as an oscilloscope or digital communication analyzer your device requires differential inputs NOTE The pattern generator s Data Out Clock Out Aux Data Out Trigger Ref Clock Out and Aux Clk Out ports must be terminated with 50 Q if they are not connected
119. that we cannot decide For example if we compared 3 x 107 bits and got two errors a measured BER of 0 667 x 10 12 we are in the uncertain white area on the graph In such a case we need to transmit more bits until the number of bits either reaches the upper limit 6 296 x 1012 or until we see more errors If the actual BER is very close to 10 2 however we are unable to apply a lower or upper limit to the BER no matter how many bits we transmit Whether such a test fails or passes depends entirely on the application Once we are able to decide for a measured point whether its BER is above or below the BER threshold we can determine the total jitter at the intersection of the BER threshold with the bathtub curve Since we are unable to find a single point on the slope where the BER is exactly 1012 we search for an interval that brackets the point at which the BER is equal to 1072 This is illustrated for the left hand slope in the following figure Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 10 10 l r1 l 10 12 10 55 10 14 i a n 12 a x x0 X We do not need to know the exact BER values at x and x It is sufficient to assure that BER x is greater than 10 1 and BER x is lower than 10 1 at a confidence level of 95 96 The algorithm then assumes that x for the left hand slope of the bathtub curve is in the middle of the bracketing interval After repeati
120. the menu item indicates the image of the last used control which means if you click somewhere on left of the seperator it will directly open the last used control Agilent J BERT N4903B High Performance Serial BERT 21 1 22 Planning the Test File External Instrument s Pattern PG Setup ED Setup Each pull down menu and its option are described in the following section The File menu allows you to perform the following tasks Generate a new pattern file Open the existing Instrument State Pattern Sequence Measurement Save the Instrument State Pattern in Editor Screen Capture in Editor Sequence Measurement e Preset Instrument State Print the screen Explore the File System Open the Text Editor e Exit the application The External Instrument s menu allows you to control the external instruments that are connected to Serial BERT The Pattern menu allows you to Select a pattern Edit a pattern Launch the Sequence Editor The PG Setup menu allows you to launch the following controls Bit Rate Setup Data Output Clock Trigger Output Trigger Ref Clock Setup Aux Data Setup Error Add Setup Alt Pattern and Aux In Setup For more information on PG Setup refer to the section Setting up the Pattern Generator Concepts on page 121 The ED Setup menu allows you to launch the following controls Clock Setup Error Ratio Agilent J BERT N4903B High Performance Seri
121. the Pattern Select Form dialog box 1 Click Pattern Select menu item from the Pattern submenu oh Pattern Select The Pattern Select Form dialog box opens 2 Select one of the following pattern types for the pattern generator and follow the next required action as listed below Table 6 Pattern Type Next Required Action User Pattern Browse for the desired pattern file Memory User Pattern Select memory slot 1 12 2 n 1 PRBS Select the pattern size 2 n PRBS Select the pattern size Mark Density Select the pattern size and the mark density ratio Zero Sub Select the zero substitution rate Agilent J BERT N4903B High Performance Serial BERT 73 3 Setting up Patterns 74 Table 6 Pattern Type Next Required Action Sequence Enable this checkbox if you intend to download a user defined sequence Patterns generated within a sequence are defined in the Sequence Editor NOTE Your instrument also contains user patterns that mimic real data packets and are designed to stress specific characteristics of your device For more information see Example Patterns on page 70 3 Ifyou want to select a different pattern for the error detector clear the checkbox Error Detector Pattern tracks the Pattern Generator Pattern The Error Detector Pattern tab appears Switch to this tab and repeat step 2 4 f you have selected 8B 10B Symbol comparison as an error ratio then ensure that the following conditions are met when
122. the calibration process It is recommended to save the instrument settings before you start the calibration process 5 Click Yes to continue if you have already saved the settings Calibration Procedure The following steps explain the calibration prodecure 1 Store current instrument setting if the current setting is required after the calibration is done 2 Load the calibration setting For more details refer Loading Calibration Settings on page 480 3 Depending on if an error detector available or not we have 2 cases Case 1 Calibration using an Error Detector Connect Data Out of N4916B to Data In of Error Detector Terminate Data Out of N4916B and Data In of Error Detector with 50 Ohm termination resistor Connect Clock Out of Pattern Generator with Clock In of Error Detector Terminate Clock Out of Pattern Generator with 50 Ohm termination resistor Foreach of the 6 timing settings do a BER measurement using the following accumulation setting Agilent J BERT N4903B High Performance Serial BERT 481 9 Solving Problems Accumulation Setup Activation Mode Period Manual C Time Single Days Hr Mn Sec C Repeat fo fo fo fi Measurement Log C Prompt for File Name Number of Errors gt C File Name ho No Logging Number of Bits gt Logging if BER ne2 H gt Log Directory Press the Auto Align key on the front panel Switch to the Accumulated Results Scre
123. the central 2096 of the bit period You can change the eye window boundary settings in the Configure Measurement dialog box These settings determine what portion of the eye will be measured 6 Eye Amplitude The difference between the logic 1 level and the logic 0 level 338 Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Table 31 S No Parameter Description 7 Eye Width The horizontal measurement of the eye opening at a specified BER Threshold Note The Eye Width is calculated according to the BER Threshold set in the View Tab Eye Width is calculated as per the selection of either Width at Crossing Point or Custom defined Width 8 Eye Height Measures the vertical opening of an eye diagram with respect to BER threshold This determines eye closure dueto noise Note The Eye Height is calculated according to the BER Threshold set in the View Tab Eye Heightis calculated with Height at in the View tab 9 Jitter P P Full width of the histogram at the eye crossing point with respect to BER threshold Note The Jitter P P is calculated according to the BER Threshold set in the BERT from Calculate Measurement Parameters of the View Tab Agilent J BERT N4903B High Performance Serial BERT 339 6 Advanced Analysis Table 31 S No Parameter Description 10 Jitter RMS A standard deviation of the crossing point histogram with respect to BER threshold Note The Jitter RMS i
124. the checkbox Error Detector Pattern tracks the Pattern Generator Pattern is selected or cleared Condition 1 When pattern tracking check box is selected Only user pattern or memory pattern that contains valid 10B symbols can be loaded in the Pattern Generator Error Detector will follow the Pattern Generator s pattern Condition 2 When pattern tracking check box is cleared Any pattern can be loaded in Pattern Generator Only user pattern or memory pattern that contains valid 10B symbols can be loaded in the Error Detector 5 Click OK to finish The selected patterns are immediately loaded to the pattern generator and error detector The types of patterns that are currently loaded to the pattern generator and the error detector are always displayed in the status bar at the bottom of the screen Pattern Generator Ead im Pu 300000 Gb s 77231 PABS toos sieup ONT Error Detector EZ lt NU 300000Gb s 27231 PRBS Enor Pe Data Clock Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 Loading Patterns Directly from the Pattern Editor After you have finished editing a pattern in the pattern editor you can load this pattern directly to the pattern generator and error detector TIT Keep in mind that your modifications in the Pattern Editor do not affect the pattern at the pattern generator or error detector until you download the pattern to the pattern generator error detect
125. the instructions prompted by the installation process Launch N4903 software Click on Utility menu item The following submenu will appear Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 Set Date and Time Minimize GUI Output Window Touchscreen Off Change GPIB Address On Screen Keyboard Self Test Licenses Web Control The WebControl dialog provides the user with the current IP address of the instrument The IP address can be used on the internet explorer to address the instrument Additionally the WebServer running on the instrument can be enabled disabled and a password can be specified Without a valid password you can not access the instrument s GUI The following screen displays the WebControl dialog Web Control OR X Insteumert IP Address 146 208 141 203 V Enable Web Server Web Server Login NowPoxmed LL OmmWe umo Agilent J BERT N4903B High Performance Serial BERT 553 10 Customizing the Instrument 554 NOTE 4 Onthe web control dialogue shown above you can request for help by pressing the Help button on the lower right hand corner The instrument IP will be automatically generated in the space labeled Instrument IP Address While adding the instrument on the network the IP address will be defined The DHCP protocol automates the assignment of the IP addresses The read only lable in the WebCont
126. the numeric keyboard to set the value You can toggle between dB and percent Toggling does not change the value NOTE It is possible to set the de emphasis ratio to a negative dB value amplification This allows you to generate a pre cursor de emphasis for N4916A CAUTION In this case there is a risk that you damage your device If you intend to generate a pre cursor de emphasized signal refer to for the characteristics and dangers Controlling the N4916B De Emphasis Signal Converter For N4916B you have two post cursor one pre cursor and the corresponding unit Pre Cur Curl F 2 Unit oe foe Set the values as following Post Cursor1 12 04db to Odb Post Cursor2 8db to Odb Pre Cursor Odb to 12 04 db To change any of the cursor s value for details see Understanding the N4916A B De Emphasis Signal Converter on page 37 click inside the text field and either Enter the desired value directly with the numeric keyboard e Use the knob by the numeric keyboard to set the value You can toggle between dB and percent Toggling does not change the value NOTE When N4916A B De Emphasis Signal Converter is connected then the X offset and Output Blanking will not be visible in the PG Data Output window Agilent J BERT N4903B High Performance Serial BERT 133 4 Setting up the Pattern Generator 134 NOTE CAUTION Once you have connected the pattern generator to the N4916A B De Emphasis Signal
127. the x axis Display Change During accumulation data will appear to move from left to right on the ratios graph When the graph is completely filling the display the x axis time scale will double The data graph is then occupying only half of the display and will continue to move to the right again This will repeat until the accumulation period has ended Zooming This graph can accumulate a large amount of data You can click any point on this graph to zoom into an area of BER data You can continue to click on the graph until you have zoomed to the maximum The point you click will appear in the middle of graph You can press the back button left arrow to zoom out again Agilent J BERT N4903B High Performance Serial BERT 379 7 Evaluating Results 380 NOTE NOTE Sizing Handles The zooming function cannot be used before the accumulation period has ended To view precise BER data for every point in time view the measurement log file While zoomed in you can move data left or right by dragging the graph or by using the front panel knob You can also zoom vertically by dragging the top and bottom sizing handles The current view can be moved vertically if you click between the handles and drag them up or down o U Vs 2 p 4 Hur Eo B U 10 EMI hes Bo Accumulated Results In the Accumulated Results table the following values are
128. used for pattern A and the odd bits of the Standard pattern will be used for pattern B The pattern length will not be adopted automatically when selecting Demultiplex By default the pattern length will be kept at the original value and the pattern is being filled with 0 bits In case that the converted Alternate pattern shall only contain bits of the original Standard pattern set the Length in Bits to half the pattern length before clicking OK Pattern Capture Dialog Box The pattern capture screen has four fields as shown in the screen below Agilent J BERT N4903B High Performance Serial BERT 83 3 Setting up Patterns Description Length in Bits Current ED pattern length Maximum Memory Depth 84 Pattern Capture Sel Description NEN C Length In Bits Curent ED Pattem Length C Maximum Memory Depth DK Cancel Help Description for the captured Pattern Pattern capture length in bits Captures the entire pattern length of currently loaded pattern in the error detector Captures the maximum 32Mbit pattern length of currently loaded pattern in the error detector Find Pattern Segment Dialog Box The Find Pattern Segment dialog box allows you to search for a specified bit sequence or symbols in the pattern If you click Next the next occurrence of this bit sequence or symbol is highlighted Find Pattern Segment Enter Find Pattern Segment 00111100 Enter Replacement Pattern Segment 11110000
129. when the BER Location Mode is set to more than one bit for example all bits or a block with a length gt 1 Error Location Capture Procedures 304 This section shows how to set up and perform an Error Location Capture measurement As an example first add a couple of errors to an alternating pattern and then capture their position This requires the following steps Preparing the measurement see How to Prepare the Error Location Capture Measurement on page 304 Demonstrating the measurement see How to Demonstrate the Error Location Capture Measurement on page 305 How to Prepare the Error Location Capture Measurement To prepare an Error Location Capture measurement to detect the location of an errored bit in an alternating pattern 1 Disable the pattern generator outputs by pressing the 0V Disable button in the PG Setup gt Data Output screen 2 Use a shielded cable to connect the pattern generator s Data Out port and the error detector s Data In port 3 Usea second shielded cable to connect the pattern generator s Clock Out port and the error detector s Clock In port 4 Terminate all non connected pattern generator output ports with 50 Q 5 Make sure to preset the instrument state before running the measurement Select Preset Instrument State from the File menu 6 Enable the pattern generator outputs by pressing the 0V Disable button 7 PressSync Now and then Auto Align to find the optimum sampling po
130. with the instrument s Welcome Page as shown in the screen below 3 4 XX Agilent Technologies NEEDED MSIE eT E lr sr i a Welcome to your Web Enabled Serial BERT Information about this Web Enabled Serial BERT NA4900AoxDE Instrument Description heme M IP Address 134 40 61 67 TCPIPO bbnfwtt 00 INSTR VISA TCPIP Connect String Ld advanced information about this Web Enabled Serial BERT Use the navigation bar on the left to access your Serial BERT and related information Click the Web Control button in the navigation bar on the leftto start the application Agilent Technologies Inc 2005 Click on the left side link to control monitor the instrument The browser returns the instrument display and control applet in the main display frame which contains the instrument display and the instrument centric controls provided in the applet How to Install the Web Server The following steps explain how to download and install the Web Server The current J BERT instrument can be upgraded to the Web Server enabled software Follow the steps below to install the web server on the instrument 1 Find webserver zip at the same location as the N4903 software and download it To download it you can right click on the link and select save target as Unzip the folder because the installation can only be started from an unzipped folder Install the Web Server software Follow
131. your hard disk This may take more than 2 hourstt Install original Recovery Image s d on the Recovery Partition of the hard di This will restore the initial fac setup 3 Exit Reboot System without doing anything Enter a Choice This user interface provides the following options 526 Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 Table 36 Options Description Run CHKDSK to resolve Disk Issues If you encounter any malfunctions you should always try this option first CHKDSK won t delete any of your user generated data buttry to recover the disk issues The CHKDSK process may take more than 2 hours Restore Factory Settings Use this option only if CHKDSK could not repairthe system This option will restore the originally installed operating system and software Remember the entire user generated data and settings will be lost and there will be no chance of recovering this data afterwards See How to Recover the System from the Hard Disk on page 527 for details Exit This option quits the user interface and reboots the instrument How to Recover the System from the Hard Disk You can use the Agilent Recovery System to restore the system to its factory setting in case of a serious malfunction caused by corrupted or inadvertently deleted files on the system s primary C partition The following steps will guide you to recover the system from the hard disk 1 Bootthe instru
132. 0 entering the required information You can remove keyboard and mouse when you are done Checking the Serial BERT Software Revision The Serial BERT software supports the N4876A on from revision 6 7 or later 1 Switch on the Serial BERT 2 Inthe Help menu click About 3 If the software revision is below 6 7 you need to update the software see Updating the Software on page 528 Checking the Agilent IO Libraries Revision New Serial BERT instruments are delivered with Agilent IO Libraries Suite Revision 15 5 or later installed 1 In the Utility menu click Minimize GUI to access the Windows operating system 2 Click the AgilentIO Control icon in the Windows task bar to open the IO Control menu 10 You see the menu item Agilent Connection Expert in the Windows task bar Agilent Connection Expert Event Viewer Interactive IO ViFind32 debug utility VISA Assistant VXI Resource Manager Agilent VISA Options Documentation Installation Information IO Config Information Hide Agilent IO Control Exit About Agilent IO Control 3 Click the About item to check the details Agilent J BERT N4903B High Performance Serial BERT 545 10 Customizing the Instrument 546 NOTE If you wish to update the Agilent IO Libraries Suite you can download the software from http www agilent com find iolib Installation instructions are given in the associated Readme file Connecting the N4876A via USB The N4876A
133. 0 1 60e 10 1 40e 10 1 20e 10 1 00e 10 8 00e 9 6 00e 9 4 00e 9 2 00e 9 Relative 240 00 ps 230 00 PS oo ps 210 00 PS coo ps 19000 P0100 ps 170 00 PS 69 00 pe 0 00 PS 49 00 ps NOTE Because the right hand slope of the tub does not provide additional information on the jitter the measurement s jitter display shows only the portion at the left hand side of the optimum sampling point The jitter histogram allows you to visually inspect the jitter components Random Jitter RJ Deterministic Jitter DJ Estimated Total Jitter TJ Jitter Measurement Parameters on page 260 describes how these components are calculated A Gaussian marker is available in this view to help analyze the jitter components See The Gaussian Marker on page 237 for details Furthermore the following parameters can be set on the View tab of the Properties dialog box Analyze You can analyze for All Errors To calculate the BER values from all bits errors e Errors if 0s Expected To calculate the BER values if 0 is expected but 1 received Errors if 1s Expected To calculate the BER values if 1 is expected but 0 received Calculate You can calculate measurement parameters for DErrors To calculate the parameters for 0 errors Agilent J BERT N4903B High Performance Serial BERT 257 6 Advanced Analysis 258 Table Number Format Timing Units Scale If this is selected RJ DJ separation is not available
134. 0 Data In port 169 Data Input Delay 196 Data input delay 199 Data Input Setup 172 DATALOSS 471 Data Out port 123 Data Termination 137 Date andTime 529 dBER vs Threshold Graph 274 DCA Connections 32 De Emphasis Basics 37 De Emphasis Signal Converter 15 Decimal Places Eye Opening 300 Fast Eye Mask 315 Output Levels 276 Output Timing 258 Decision Threshold 172 Delay 127 136 Delay Control Input 123 Delay Line display 415 Delay Line selection 409 DER Results Window 370 DER Status Indicators 373 Deterministic Jitter 261 Different In Out Rates Connections 30 Differential 172 Differential Inputs Connections 30 Disparity Error Ratio 182 Double Click Configuration 532 DUT Connecting 29 568 DUT Output Timing Jitter measurement 241 E ED CLKLOSS 470 Edge Resolution Optimization Eye Opening 295 Output Levels 271 Output Timing 253 Edit Pattern Window 76 102 Edit Pattern Window for Sequencer Pattern Editor 110 Editing Patterns 72 Equal BER at BER Threshold 288 Error Accumulation 221 Error Accumulation 220 Error Add button 157 Error Add input 122 Error Addition Insertion 157 Error Detector 1 0 threshold 172 Auto Threshold 181 Aux Out port 170 Aux Output 203 Clock In port 169 Clock Setup 173 Data In port 169 Data n setup 172 Error Out port 169 Gate In port 169 Input and Output Ports 168 Setup 167 Trigger Out port 169 Error Indicator 472 Error Insertion C
135. 0 0000 MHz n BAZ poms fonom az AA oru br mz Each jitter type can be separately turned on or off by clicking the LED button on the left side The corresponding jitter source is enabled if the LED is on The Jitter Setup area allows you to directly change the most commonly used parameters The Advanced Edit Panel allows you to change all jitter parameters available for the corresponding jitter source It is opened by pressing the Edit button A Jitter Distribution Area The Jitter Distribution area allows to Define whether the corresponding jitter source is applied to the data and or the clock output It is not possible to disconnect the jitter sources from both outputs However it can be done by disabling the corresponding jitter sources Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Specify which delay line 220ps or 610ps is used for the generation of PJ and BUJ NOTE If the 610ps delay line is selected and has any jitter source connected to it then it is not possible to disable SJ SSC rSSC on the data output Specify the delay between 220ps jitter on Clock and Data Jitter Setup Procedures Do the following to compose the desired jitter 1 Click Jitter Setup from the Jitter submenu x Jitter Setup This opens the Jitter Setup window 2 Enable jitter generation by pressing Jitter On Off button When this button is enabled the LED on the button turns green in
136. 00e 2 1 000e How to Improve the Output Levels Display You can change the display of an existing measurement for example if you wish to see more details to investigate the graph This can be done on the Graph tab of the Properties dialog box 1 Press the Properties button If you have a mouse connected to your Serial BERT you can also click the right mouse button on the graph and select Properties from the context menu Switch to the Graph tab As an example select Show Measured Points With this option selected all measured points are indicated in the graph with small squares 800 0 mv Trace 202 Points 900 0 mv 40V AAV 42V 43V 44V ASV ABV ATV BER Threshold 1 009e 6 18 Logarithmic 1 000e 5 1 000e 3 1 000e 1 1 000e 6 1 000e 4 1 000e 2 1 000e The headline of the graph tells the number of data points included in the display 101 This corresponds to the chosen threshold resolution of 100 steps per Volt With this number of measured points you do not get very precise results during the measurement but it is quickly finished Depending on your personal needs you can trade measurement precision against test time Several more options are provided on the Graph tab for you to adjust the display as you want For instance you can use markers for analyzing the graph or open a zoom window for viewing the details To find more information refer to Graph Tab on page 276 Agilent J BE
137. 01 0111 0101 1011 0110 1000 0011 0110 0110 0001 0101 0011 0111 1110 0010 0011 1100 1101 0000 0001 0100 0010 1101 1111 0111 1001 0010 1100 0001 0010 1000 1101 1011 1000 0001 1011 1111 1001 0000 1100 0101 OEA E WOO OE TOUS WE E ES aC a BS et 0001 0001 0010 0110 0000 0101 From the Pattern menu click Sequence Editor You will be now able to select edit and drop in the data you want on both the Regular Channel and the Aux Channel x 3X GUI Agilent N4903B connected to localhost Ss BER 0 000 Remote 7 Data on normal channel I Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 M Trigger Ok Cancel Edit Pattern s Create Pattem Help Send the patterns to the outputs by pressing To PG BER 0 000 Send patterns to output channels You should be now able to send the data Agilent J BERT N4903B High Performance Serial BERT 153 4 Setting up the Pattern Generator 154 Aux Data Output Reference NOTE NOTE If you wish to use the pattern generator s Aux Data Out port set the mode from the following options Set the Aux Data Out divider to a suitable value between 2 and 128 Set the Second Channel Set the Multiplexer mode Set the De Emphasis If the Divider Factor n is uneven e g 3 the clock s duty cycle will not be 50 but the signal will stay high for n 1 2 and low for n 1 2 Thi
138. 03B High Performance Serial BERT 317 6 Advanced Analysis 318 italics A BER value for which a Contour does not exist is struck across You can select multiple BER values The lower BER values are in red and pink while the deep BER values are in green and blue The screen shot below displays the contour and the legend The outer contour represents the measured BER contour while the inner lines are extrapolated For more details on extrapolation see Estimated Total Jitter on page 245 The Minimum Number of Errors indicates the minimum number of hits required to plot the contour Eye Mask Measurement Masks are marked regions on the screen When the mask region and waveform intersect it is called a violation The Mask test shows the quality of the eye When the mask measurement begins the instrument either aligns the mask as per the eye levels of the waveform or as per the absolute voltages specified in the mask file The position of the mask can be changed by scaling the mask according to the eye Once the masks are positioned the mask test counts the number of times a particular pixel falling in the mask region has been measured The mask violations are marked in green A consolidated report of the test is available on the left bottom of the graph The user can turn off the mask computation by pressing the stop button Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 NOTE The mask te
139. 0e gus af Terminal Electrical Mean Level 430 91 mv 479 08 mV 24 085 mv 430 91 mv 479 08 mv 24 085 mv 4 Repeatthe measurement with modified parameters and observe the new results 5 To distinguish the two graphs change the color of one graph by tipping the red Color field You can now directly compare the graphical and numerical results of the two Output Levels measurements How to Rename Copied Measurements To rename a copied measurement for easy distinction 1 Highlight the copied measurement 2 Press Enter or choose Rename from the context menu 3 Change the name and press Enter How to Load Measurements To load previously saved measurements 1 Select Load Measurement from the Analysis menu 2 Use the file dialog box to locate the desired mcp file and press Open to load the measurement The respective screen of the Analysis panel opens and you can view the measurement Agilent J BERT N4903B High Performance Serial BERT 231 6 Advanced Analysis 232 How to Change Measurement Properties and the Graphical Display The various tabs in the Properties dialog box allow you to specify the parameters forthe measurement The available options are not the same for all measurements See the Reference information of the respective measurement type for details To modify the measurement properties 1 2 Press the Properties button to open the Properties dialog box Use the differ
140. 0e 1 BERT t 1 00e 2 1 00e 3 Markers Time 1 00e 4 Upper Right 0 443 LII Lower Left 0 473 Ul 1 00e 5 Delta 0 915 Ul A 1 00e 6 PI BER2 y 100e BER Thresholc Y O00e 7 T 1 e 8 Logarithmic 0 62 UI 0 37 Ul 042 UI 042l 0 37 UI 0 62 UI Relative 0 50 Ul 0 25 UI 0 00 LII 0 25 Ul 0 50 Ul gt gt x1 x2 The following values are displayed for the current marker positions Table 23 Value Description Upper Right Time and BER of the upper right intersection ofthe displayed marker lines X2 and BERT Lower Left Time and BER of the lower left intersection of the displayed marker lines X1 and BER2 Delta Distance between the two vertical marker lines on the time axis X2 X1 and distance of the two horizontal marker lines on the BER axis BER1 BER2 The markers can be moved to keep track of the data The Gaussian Marker A Gaussian marker is available for the following measurements DUT Output Timing Jitter measurement Output Levels measurement The Gaussian marker allows to measure the normal distribution of measurement results The distribution of random values is called normal if it can be described by the following formula Agilent J BERT N4903B High Performance Serial BERT 231 6 Advanced Analysis This formula describes a bell shaped Gauss curve If u is zero and o varied you would get the curves illustrated in the figure below The height and position of a normal distribution can
141. 11 00 MHz 2 00 MHz 4 00 MHz 6 00 MHz 8 00 MHz 10 00 MHz 12 00 MHz When the repetition period of the characteristic pattern in the error record increases you will also find the typical sine x over x decay of the spectral power Time Frequency Leakage and Windowing FFT assumes that the time record contains a representative section of an endless periodic signal It assumes that time records can be seamlessly concatenated If this is not the case a phenomenon called leakage occurs Agilent J BERT N4903B High Performance Serial BERT 349 6 Advanced Analysis 350 Leakage makes it impossible to detect minor adjacent spectral components The following two figures refer to a slightly disturbed sine wave The Appearance of Leakage Source Waveform Expected Toe fo 2f 3f 4f e e Tn EN A n Time record Computed SOM spectrum f 0 When you perform a Spectral Jitter measurement it is likely that some degree of leakage occurs The measurement therefore provides a choice of FFT windows that allow you to detect leakage and to reduce its impact An FFT window is a filter that sets the beginning and end of the time record smoothly to zero Such records can be seamlessly concatenated Windowing Input waveform RO e Original f smal AAAARAAAAAAAAAA DIES Spectrum Incorrectly N A A pe with leakage sampled record VVV P T N Window function Por Hanning window m S E ne le
142. 16A B simplifies remote programming and helps to make programs portable The Serial BERT software requires a specific VISA alias name to access the N4916A B The Agilent Connection Expert is available on from revision 14 0 of the Agilent IO Libraries Suite 1 Ifthe Agilent Connection Expert does not pop up automatically click the Agilent IO Control icon in the task bar and open it from the menu 2 Inthe Instrument I O panel For N4916 click UsbDevice1 UsbDevice1 is the default alias name for N4916A For N4916B click UsbInstrument1 UsbInstrument1 is the default alias name for N4916B 3 Click Change Properties and change the alias name For N4916A change the alias name to D4916 e VISA Alias Properties Set configurable properties for this VISA alias VISA alias D4916 VISA address USBO 2391 21000 0 0 1NSTR D4916 is the alias used by the software For N4916B change the alias name to DATA OUT AUXDATA OUT depending on the channel Data Out and Aux Data Out used s VISA Alias Properties et configurable properties for this VISA alias VISA alias Pata ou VISA address U5B0 0x0957 0x5A18 Proto_R2_003 DATA OUT AUXDATA OUT is the alias used by the software Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 4 Click OK The Agilent Connection Expert for N4916A shows a window like the following Agilent Connection Expert Eile Edit View I
143. 1e 10 and 1e 5 The calculation tells you the optimum decision threshold and also the BER expected when the error detector uses this threshold High level noise distribution Gaussian characteristics assumed Regions for Q factor measurements Low level noise distribution Gaussian characteristics assumed Optimum sampling delay To measure such small bit error rates you have to compare an adequate number of bits Depending on the DUT characteristics you may also have to use a very fine step size for the threshold variation Comparison of the Results On the other hand such a comparison may make it easier to understand the dependencies If you set the BER Threshold to a position that gives you only a few data points you will find The Q factor is relatively high Agilent J BERT N4903B High Performance Serial BERT 285 6 Advanced Analysis NOTE 286 The R values are excellent greater than 0 95 The mean levels and standard deviations returned by the Q factor calculations differ from the measured values The OBER vs Threshold graph may look as shown below 1 Trace 16 Points 8000mv 300 amv aay Zoom in 1111Y 4A2V ASY 44V ASY A6v ary 18y Linesr If you increase the BER Threshold to include more points you will find The mean levels and standard deviations returned by the Q factor calculations approach the measured values The Q factor decreases The R values det
144. 1x107 1E 7 Agilent J BERT N4903B High Performance Serial BERT 13 1 14 Planning the Test Understanding the Serial BERT The Serial BERT also referred to as a Bit Error Ratio Tester BERT is a powerful instrument that enables you to analyze systems and components in the telecommunication and enterprise communication industries It is an essential tool for designing and troubleshooting communications systems high speed integrated circuits ICs and photonic components Its intuitive operation and leading edge performance will help you quickly verify error performance The Serial BERT comes with advanced jitter generation capabilities for jitter tolerance testing J BERT The instrument provides Excellent waveform performance Outstanding rise times and jitter performance Achoice of logic family interfaces Easy object oriented programming with the IVI COM interface Remote commands similar to HP Agilent 71603B A variety of ways in which to accumulate and view test results Acollection of example patterns Convenient online documentation Automatic setup of optimum sampling point A variety of powerful advanced analysis tools Special support for jitter analysis and characterization Which Features are Available The Serial BERT is available with different options They differ in the range of provided bit rates and in the available features Depending on the options you bought or upgrated you can have t
145. 2 D27 3 MIPIMPhy K28 5 Filler K28 1 Display Port K28 5 Filler SAS Align K28 5 D10 2 D10 2 D27 PCle 1 amp 2 E Skip K28 5 K28 0 K28 0 4 Choose the error ratio mode from the following options a Bit Comparison b 8B 10B Symbol Comparison c Bit Comparison without PCle3 SKPOS d Bit Comparison without USB3 1 SKPOS NOTE The Bit Comparison option is selected as default NOTE Switching among the error ratio options might take some time please wait untill the switching process is completed If you select error ratio based on Bit Comparison Bit Comparison without PCle3 SKPOS or Bit Comparison without USB3 1 SKPOS skip steps 5 to 11 and go to step 12 186 Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 However if you select error ratio based on 8B 10B Symbol Comparison follow the below steps 5 Click cBER Setup cBER Setup It opens a cBER Setup dialog This dialog offers the choice to calculate cBER based on either SER FER FSR ISR or DER with either automatic conversion factor or manually specified conversion factor cBER Setup calculate cBER from SER FER FSR ISR DSR cBER based on FER Automatic C User Defined Conversion Factor 2700 OK Cancel CAUTION When cBER is calculated from FSR DER ISR in this case the cBER might be lower than the sync threshold even if the cBER based on SER or FER is higher than sync threshold This affects the au
146. 236 Mean Level 278 Measured Points 300 Measurement Log 221 476 Measurement points defaults 309 Measurements 361 Memory based Patterns 61 Min BER forQ 276 More Errored 0 s than 1 s 473 MUX DEMUX Pair Connections 32 N N4876A Connections 34 Installation 544 Troubleshooting 511 N4916A 15 Basics 37 Connections 34 Installation 539 Troubleshooting 493 N4916B 15 Basics 37 Connections 34 Installation 539 Troubleshooting 500 No of points 261 Noise Threshold 358 Normal 172 not applicable 261 Number of Compared Bits Eye Opening 295 Fast Eye Mask 312 Output Levels 270 Output Timing 252 Number of Errors Eye Opening 295 Fast Eye Mask 312 Output Levels 271 Output Timing 253 Number of Top Frequencies to Show 358 569 Index Number of Valid Points 313 Numerical Results Eye Opening 301 Output Levels 277 Output Timing 259 Spectral Jitter 359 0 Offset threshold 312 On Screen Keyboard 532 Optimal Sample Point Delay 260 263 301 Optimal Sample Point Threshold 301 Optimum Sampling Point 196 Oscilloscope Connections 32 Output blanking 138 Output Level Parameters 126 Output Levels measurement 263 Output Protection 124 Output Timing measurement 241 Output Timing Measurement Optimization 253 P Parameters Bounded uncorrelated jitter 425 External jitter 429 Periodic jitter 421 Periodic jitter 2 Constant Amplitude Sweep 423 Periodic Jitter 2 Constant
147. 294 Error Location Capture 302 Error Location Capture Concepts 302 Error Location Capture Procedures 304 Error Location Capture Reference 305 Fast Eye Mask 307 Fast Eye Mask Concepts 307 Fast Eye Mask Procedures 309 Fast Eye Mask Reference 311 Eye Diagram 315 Eye Diagram Concepts 315 Eye Diagram Procedures 320 Eye Diagram Reference 332 Spectral Jitter 342 Spectral Jitter Concepts 342 Spectral Jitter Procedures 351 Spectral Jitter Reference 353 7 Evaluating Results Evaluating Results Concepts 361 Instantaneous Measurements 361 Instantaneous Measurements Concepts 361 Instantaneous Measurements Procedures 363 Instantaneous Measurements Reference 366 Agilent J BERT N4903B High Performance Serial BERT Accumulated Measurements 376 Accumulated Measurements Concepts 376 Accumulated Measurements Procedures 377 Accumulated Measurements Reference 378 Eye Measurements 392 Eye Measurements Concepts 392 Eye Measurements Procedures 392 Eye Measurements Reference 393 8 Jitter Tolerance Tests Jitter Tolerance Tests Concepts 395 Jitter Setup 404 Jitter Setup Concepts 404 Jitter Setup Procedures 407 Jitter Setup Reference 413 Interference Channel Setup 431 Interference Channel Setup Concepts 431 Interference Channel Setup Procedures 431 Interference Channel Setup Reference 432 Jitter Tolerance Characterization 433 Jitter Tolerance Characte
148. 3 0 Y EZ Specify the following parameters Standard Sweep Time Nr of Steps Step Distance For more information on parameters see Periodic Jitter 2 Variable Amplitude Sweep Parameters on page 424 Observe the consumption of delay line capacity indicated by the delay line markers Ensure that the total jitter remains within the margins of the chosen delay line Return to the overview display by pressing either the Close button the Edit button of this jitter source or the Edit button of the next jitter source you want to change 9 Proceed with the next jitter type you wish to add When you activate an External jitter source you will find that this apparently consumes the whole free capacity of the delay line at the right hand side of the red bar But this is just an indication You can still add or change internal jitter components The external jitter range is automatically updated when you add change or switch off one of the internal sources Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Jitter Setup Reference The Jitter Setup function is used for composing the total jitter in a defined and calibrated way Jitter Enable The Jitter On Off button is used enable or disable jitter generation Jitter T When this button is disabled no jitter is generated This refers to the whole composition shown on the screen When this button is enabled the presen
149. 3 SKPOS Results window which is described in BER Results Window on page 366 3 View the Bit Comparison without PCle3 SKPOS bar and errors indicator See Status Indicators on page 372 for details How to Monitor Bit Comparison without USB 3 1 SKPOS Results There are different ways how you can monitor instantaneous Bit Comparison without USB3 1 SKPOS values 1 Listen to the Bit Comparison without USB3 1 SKPOS warning tones See Audio Signals Procedures on page 225 for details 2 View the Bit Comparison without USB3 1 SKPOS Results window which is described in BER Results Window on page 366 Agilent J BERT N4903B High Performance Serial BERT 3 Evaluating Results 7 View the Bit Comparison without USB3 1 SKPOS bar and errors indicator See Status Indicators on page 372 for details How to Analyze Instantaneous Measurements Results To analyze the behavior of your device you can do the following modifications to the test setup and monitor the effect on the resulting BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS Bit Comparison without USB3 1 SKPOS 1 Add errors to the data stream See Manually Inserting Errors on page 157 for details Use alternating patterns to switch between different data streams See Setting Up a Periodical Output on page 162 for details Adjust the parameters of the output port of the pattern generator as described in Setting Logic Levels and Terminations
150. 4 dB 32 dB 40 dB Noise Threshold 40 dB 48 dB 56 dB 64 dB 72 dB 80 dB 88 dB 96 dB 104 dB Bean aie Logarithmic 100 00 KHz 10 00 MHz 1 00 GHz 1 00 MHz 100 00 MHz Terminal Optical Show Color Copied Bit Error Rate Agilent J BERT N4903B High Performance Serial BERT 357 6 Advanced Analysis 358 Absolute vs relative values Noise Threshold The figure above shows the absolute power scale and values as calculated by the FFT You can also calculate relative values Relative values can be normalized to either the total power True Relative or to the power of a selectable frequency bin Relative The following figure shows an example where the scale and the power values have been normalized to the power measured at 1 MHz 1 Trace 65535 Points 32 dB 24 dB 16 dB Reference power 8 dB 0 dB 8 dB 16 dB 24 dB 32 dB 40 dB Noise Threshold 40 dB 48 dB 56 dB Logarithmic 100 00 KHz 10 00 MHz 1 00 GHz 1 00 MHz 100 00 MHz Terminal Optical Show Color Copied Bit Error Rate 14 Jis When you set the power values to relative this changes not only the graphical scale but also the calculated results You can thus calibrate the power values to any reference The Noise Threshold can be moved with the mouse or your finger if you are working directly on the Serial BERT It can also be set on the View page of the Properties dialog This t
151. 5 where p4 is the mean level of the 1 and 0 rails respectively and o is the standard deviation of the noise distribution on the 1 and 0 rails Q Optimum Threshold The Q Optimum Decision Threshold is calculated as Oy 0149 Optimum Decision Threshold o tO Q Residual BER The Q Residual BER is the expected BER at the Optimum Decision Threshold It is calculated as Q0 Q 2n Numbers below 1e 255 are expressed as zero BER Q High Level Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 The Q High Level is the mean calculated from the linear regression curve for the high level data Q High Level Std Dev The Q High Level Standard Deviation is the o Sigma calculated from the linear regression curve for the high level data 1 Q High Level Nr Points This is the number of data points used for the calculation of the Q High Level value It depends on the setting of the BER Threshold and also on the setting of the Min BER for Q parameter The minimum for calculating Q factor values is two points It is recommended to use more than 5 points QHigh Level R 2 The R parameter is an indicator that shows how well the converted data points fit to the straight line It is calculated as 2 XxY ae n zx x lt p 2M The R parameter should be examined before trusting the Q values Its maximum value is 1 0 It must be seen in conjunction with the number of
152. 5 Gb s data rates 12 5 Gb s when using analyzer with external clock display of the error ratio as SER or calculated BER For PCle3 PCI9 Express 3 0 the option A02 enables the error counter to ignore changes in the length of 128 130 bit coded SKPOS Skip Ordered Sets To use this functionality N4903B software revision 7 40 or higher is required gt 002 Option for N4876A 28 Gb s Multiplexer 2 1 Second Output Channel and M8061A 28 Gb s Multiplexer with De emphasis e A03 UA3 Option Upgrade with analysis of 128b 132b coded patterns To use option A03 functionality the N4903B software revision 7 60 or later is required To check the options of your Serial BERT press About in the Help menu The About dialog shows the installed options De Emphasis Signal Converter An external optional instrument which is used to convert the output of the pattern generator to a de emphasized signal as required by many standards such as PCI Agilent J BERT N4903B High Performance Serial BERT 15 1 16 Planning the Test Express SATA 3 Gb s USB3 fully buffered DIMM Hypertransport CEI or 10 GbE The following De Emphasis Signal Converter can be used N4916A De Emphasis Signal Converter The N4916A is a 2 tap de emphasis signal converter Following are the features and benifits of N4916A Generates 2 tap de emphasis Support data rates from 1 to 13 5 Gb s Transparent to data jitter Small size Programmable v
153. 55404 0 0 passed 8 1 79E403 8 50 3015843721 0 0 passed 9 3 21E 03 8 50 3015843721 0 0 passed 10 5 74E403 8 50 3015843721 0 0 passed TIP The comment can be entered with the Modify Comment button below the graphical result view Agilent J BERT N4903B High Performance Serial BERT 467 8 Jitter Tolerance Tests 468 Agilent J BERT N4903B High Performance Serial BERT e Solving Problems Solving Problems Concepts This section provides information that can help you to troubleshoot the Serial BERT in case any problems occur Setup Problems Setup Problems Concepts You may run into the following problems when setting up the Serial BERT Front Panel not Working If the front panel is not working and a hardware wizard calls on you to install the driver for an unknown USB device cancel the hardware wizard and reboot the instrument This might happen if the front panel controller is not initialized correctly at boot up Incorrect Error Detector Input Voltages The error detector s protection circuit disconnects the input amplifier if the incoming signal does not lie within the defined input range In this case a dialog box will open that lets you define the input range sample point 0 1 threshold termination voltage and state When you click OK the input amplifier will be reenabled Agilent Technologies 469 9 Solving Problems 470 If the input signal is still out of range for any of these parameters th
154. 68 15 15 23 8 388 608 23 23 Zero Substitution Patterns A potential risk to bit errors are longer rows of zeros within a data stream The longest run of zeros in a 2 n PRBS consists of n zeros The Zero Substitution function can be used to stress the DUT additionally by inserting a longer row of zeros in the data stream The length of the zero row can be freely specified up to the length of the pattern The following figure shows an example where a run of 10 zeroes is inserted into a PRBS 2 7 pattern Run of 0s ZRUN PRBN 0000000100000110 ZSUB 0000000000100110 Closing bit Mark Density Patterns The mark density of a pattern is the portion of bits that are ones A pure PRBS pattern contains an equal number of ones and zeros and thus has an average voltage in the center between high and low level Therefore the mark density of such a pattern is 1 2 Some devices may have problems processing data streams with different mark density as this results in other average voltages and thus different electrical conditions Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 The ones mark density can be varied to put the pattern out of balance in a randomly distributed way This may be useful for systems that are AC coupled This helps to check for effects such as baseline wander To test these cases the Serial BERT provides PRBS patterns with the mark densities 1 8 1 4 1 2 3 4 and 7 8 S
155. 7 ns 57 2 ns Unit interval rel 0 75 UI 0 UI 0 75 UI abs 27 1 UI 27 85 UI 28 6 UI The Unit Interval Unit interval values are a convenient way to express time values in a dimensionless form One unit interval is the equivalent of one clock period To convert a time value to a unit interval value divide it by the pulse period or multiply it by the clock frequency Agilent J BERT N4903B High Performance Serial BERT 235 6 Advanced Analysis Example Measurements at different frequencies Linear Markers 236 At 100MHz the pulse period is 1 100 x 10s 108s 10ns this is the unit interval UI For example a value in fact any parameter that is specified in time units of 37ns is equivalent to 37ns 10ns 3 7UI time values are expressed as multiples of the unit interval This makes it easy to analyze measurements at different frequencies If the measured phase margin is 8ns at 100MHz 1UI 10ns and 4ns at 200MHz 1UI 5ns The phase margin is 0 8UI in both cases and it is immediately obvious that it does not depend on the system frequency The same is true for the delay resolution parameter in the DUT Output Timing Jitter and Eye Opening measurements If you specify the delay resolution as 0 01UI the measurement software will always measure 150 points across the bathtub no matter what system frequency you choose only if edge optimization is switched off of course However if you specify the resolution in time the num
156. 79 Error Ratio 182 Error Ratio Concepts 182 Error Ratio Procedures 185 Error Ratio Reference 190 Sampling Point Setup 193 Sampling Point Setup Concepts 193 Sampling Point Setup Procedures 196 Sampling Point Setup Reference 198 Trigger and Aux Output 203 Trigger and Aux Output Concepts 203 Trigger and Aux Output Procedures 203 Trigger and Aux Output Reference 204 Pattern Synchronization 205 Pattern Synchronization Concepts 205 Pattern Synchronization Procedures 216 Pattern Synchronization Reference 218 Error Accumulation 220 Error Accumulation Concepts 220 Error Accumulation Procedures 220 Error Accumulation Reference 221 BER Location 223 BER Location Concepts 223 BER Location Procedures 223 BER Location Reference 223 Audio Signals 224 Audio Signals Concepts 224 Audio Signals Procedures 225 Audio Signals Reference 225 Agilent J BERT N4903B High Performance Serial BERT 6 Advanced Analysis Advanced Analysis Concepts 227 Advanced Analysis Procedures 229 Advanced Analysis Reference 234 DUT Output Timing Jitter 241 DUT Output Timing Jitter Concepts 241 DUT Output Timing Jitter Procedures 250 DUT Output Timing Jitter Reference 252 Output Levels 263 Output Levels Concepts 263 Output Levels Procedures 265 Output Levels Reference 270 Eye Opening 287 Eye Opening Concepts 287 Eye Opening Procedures 289 Eye Opening Reference
157. Agilent J BERT N4903B High Performance Serial BERT User Guide Es Agilent Technologies Notices Agilent Technologies Inc 2014 No part of this manual may be reproduced in any form or by any means including electronic storage and retrieval ortranslation into a foreign language without prior agreement and written consent from Agilent Technologies Inc as governed by United States and international copyright laws Manual Part Number N4903 91021 Edition Release Edition June 2014 Printed in Germany Agilent Technologies Deutschland GmbH Herrenberger Str 130 71034 B blingen Germany Warranty The material contained in this document is provided as is and is subject to being changed without notice in future editions Further to the maximum extent permitted by applicable law Agilent disclaims all warranties either express or implied with regard to this manual and any information contained herein including but not limited to the implied warranties of merchantability and fitness for a particular purpose Agilent shall not be liable for errors or for incidental or consequential damages in connection with the furnishing use or performance of this document or of any information contained herein Should Agilent and the user have a separate written agreement with warranty terms covering the material in this document that conflict with these terms the warranty terms in the separate agreement shall control
158. Agilent J BERT N4903B High Performance Serial BERT 255 6 Advanced Analysis Bit Error Rate Graph Jitter Histogram 256 1 00e 0 1 Trace 39 Points Ul 401 878 ps All Errors 1 00e 1 1 00e 2 1 00e 3 1 00e 4 1 00e 5 1 00e 6 1 00e 7 BER Threshold 4 1 000e 7 1 00e 8 Logarithmic 0 62 UI 0 37 Ul 0 12 UI 0 12 UI 0 37 UI 0 62 UI Relative 0 50 UI 0 25 UI 0 00 UI 0 25 UI 0 50 UI Terminal Show Color Copied destra eme Lem ed Delay Margin 0 006 UI af View Tab The graph shows either the bathtub curve or the jitter distribution vs time The BER graph the bathtub shows the BER vs sample delay 1 00e 0 1 Trace 160 Points Ul 333 333 ps All Errors 1 00e 1 1 00e 2 1 00e 3 BER Threshold 1 0008 1 00e 4 1 00e 5 1 00e 6 1 00e 7 1 00e 8 1 00e 8 1 00e 10 Logarithmic E 1 Relative 210 00 ps 50 00 Ps Ps 210 00 BS 180 00ps 120 00 ps 60 00 ps 0 00 s 60 00 ps 120 00 ps 80 00 ps 40 00 ps 90 00ps 3000ps 3000ps 9000ps 15000 The BER graph can be viewed in either linear and logarithmic view The logarithmic view is shown above The DUT Output Timing Jitter measurement calculates the jitter histogram as the absolute of the derivative of the measured bit error rate jitter dBER dt Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 1 Trace 79 Points Ul 333 333 ps All Errors 2 60e 10 2 40e 10 2 20e 10 2 00e 10 1 80e 1
159. BERT Solving Problems 9 7 Click the IDN button This generates the identification query in SCPI format The instrument should return Agilent Technologies N4876A and the software revision YA VISA Assistant DER File Edit View Configure Help USBO USB0 2391 29976 D Instrument Driver Formatted 1 0 Memory 1 0 Attributes N48764 Clear History Show C Code IDN Device Clear Agilent Technologies N4876A DE 50400005 0 1 3 0 10 Set Timeout Read STB RST IST SYST ERR Enter String to Print or Query Instr Lang x C Custom C IEEE 488 2 viPrintf viScanf viueryf SCPI For Help press F1 11 52 08 If you could proceed until here the N4876A power supply and the processor are fully functional If not the N4876A is probably defective Ensure that the User Software can Access the N4876A 1 From the Windows task bar restore the user interface click GUI Agilent 4900 Series 2 Press the green Preset key close to the front USB connector of the Serial BERT or from the menu bar select File Preset Instrument State 3 Disable the outputs of N4903B and use the cable kit to connect the pattern generator to the N4876A Connect the Data Out Aux Data Out and Aux Clk Out of the Serial BERT to Data In Aux Data In and Aux Clk In of the N4876A respectively 4 From the Navigation Menu select External Instrument s and then click Config sub menu The Config window shows a list
160. BERT N4903B High Performance Serial BERT 4 9 Solving Problems Measurement Problems Error This indicator turns red when errors are detected View the BER bar or BER Results to see the nature of the errors BERT connected looped back or to your device Stable errors caused by the error add function Turn error add off Stable errors caused by false sync Select a lower sync threshold BER Variable and high errors may be caused by faulty connectors cables BERT connected to your device Stable or variable errors can also be caused by your device Correct the problems with your device Ifthe problems are not clear then accumulate measurement data and analyze results Measurement Problems Concepts 472 Accumulated measurements can help you determine the cause of bit errors Bit errors can be caused by a variety of problems In addition instrument settings can affect how errors are measured and displayed The sync mode setting has the greatest effect Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 Some phenomenon you might observe with the possible causes are listed below Constant Errors More Errored 0 s than 1 s Accumulated Results NU Bit Count 00 502 034 112 Ero Ratio 16ME3 Eror Count 961 152 150 JU Ered 1 s Ratio 0000 Erored t s Count 0 ll Ernces 0 s Ratio 1634E 3 Ereoeed O s Count 81 152 150 Possible Cause A hardware failure may have occurred in your device
161. Based Patterns Introduction to Pattern Synchronization Pattern synchronization sync refers to aligning the incoming data pattern with the internal reference pattern This is accomplished in one of two ways For 2 n 1 PRBS patterns bits from the incoming data pattern seed the error detector s pattern generator causing it to generate a precisely aligned internal reference pattern For software generated and user patterns a 48 bit pattern from the pattern is used as a detect word Optimally this detect word should be unique within the entire pattern The error detector searches for this detect word within the incoming data stream and uses the point in the data stream as a reference and compares all following bits with the pattern If the measured BER is better than the synchronization BER the error detector is synchronized There are thus three possible outcomes for a synchronization Single instance of the detect word in the data stream Expected Incoming Pattern Bit stream Em iiim Detect Word coa E Detect Word Correct Sync 48 bits Detect Word 48 bits E Multiple instances of the detect word with correct synchronization Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 Expected Incoming Pattern Bit stream Em s iii qiugun Reference Point Correct Sync 3 Multiple insta
162. Data Rates Often used for multiplexers or when the data rate received by the device is not equal to the data rate sent from the device LEGEND C Clck O Data Diagram 2 Connections for Differential Inputs Application is the same as in Diagram 1 Illustrates how to connect to a device requiring differential signals 30 Agilent J BERT N4903B High Performance Serial BERT Planningthe Test 1 Diagram 3 Connections for an Amplifier Often used for amplifiers Diagram 4 Connections for a Flip Flop Often used for flip flops Agilent J BERT N4903B High Performance Serial BERT 31 1 Planning the Test Diagram 5 Connections for a MUX DEMUX Pair Used to test multiplexer demultiplexer MUX DEMUX pairs For accurate measurement results the timing of data signals between the MUX DEMUX pair must be set properly Diagram 6 Connections for a DCA or Oscilloscope PGonly Used for measuring the output waveform of your device For the least amount of jitter the clock output signal may be used as a trigger for the oscilloscope 32 Agilent J BERT N4903B High Performance Serial BERT Planning the Test 1 Diagram 7 Connections for a SONET SDH Receiver PGonly Often used with network equipment such as SONET SDH receivers The device shown may or may not require a clock signal LEGEND C Clock C Clock O Oate O Date Diagram 8 Connections for an External Data Source EDonly Used whe
163. ERT The following external instruments can be connected to Serial BERT N4916A B De Emphasis Signal Converter To install N4916A B De Emphasis Signal Converter with Serial BERT see Installing the N4916A B Procedure on page 539 N4876A 28Gb s Multiplexer To install N4876A 28Gb s Multiplexer with Serial BERT see Installing the N4876A Procedure on page 544 M8061A 28 Gb s Multiplexer with De emphasis To install M8061A with J BERT N4903B see Installing the M8061A Procedure on page 549 538 Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 Installing the N4916A B Procedure The N4916A B De Emphasis Signal Converter is an optional instrument that can be connected to the Serial BERT For general information see Understanding the N4916A B De Emphasis Signal Converter on page 37 When you install the N4916A B for the first time you need to configure the USB interface of the Serial BERT For this configuration you may wish to connect a keyboard and a mouse to the Serial BERT see Connecting Peripherals on page 529 This is not mandatory but convenient the keys provided by the instrument can be used as well for entering the required information You can remove keyboard and mouse when you are done Checking the Serial BERT Software Revision The Serial BERT software supports the N4916A on from revision 4 8 and N4916B on from revision 6 5 1 Switch on the Serial B
164. ERT 2 Inthe Help menu click About 3 If the software revision is below 4 8 you need to update the software see Updating the Software on page 528 Checking the Agilent IO Libraries Revision New Serial BERT instruments are delivered with Agilent IO Libraries Suite Revision 15 5 or later installed 1 In the Utility menu click Minimize GUI to access the Windows operating system 2 Click the Agilent 10 Control icon in the Windows task bar to open the IO Control menu IO You will see the menu item Agilent Connection Expert in the Windows task bar Agilent J BERT N4903B High Performance Serial BERT 539 10 Customizing the Instrument Agilent Connection Expert Event Viewer Interactive IO ViFind32 debug utility VISA Assistant VXI Resource Manager Agilent VISA Options Documentation Installation Information 10 Config Information Hide Agilent IO Control Exit About Agilent IO Control 3 Click the About item to check the details If you wish to update the Agilent IO Libraries Suite you can download the software from http www agilent com find iolib Installation instructions are given in the associated Readme file Connecting the N4916A B via USB The N4916A B is controlled by the Serial BERT via USB 1 Connect mains power to the N4916A B and switch the instrument on When you connect the N4916A B via USB it is important that the N4916A B is NOTE E am f switched on Otherwise it may not be ide
165. External Instrument s NOTE 38 N4916B De emphasis box is supported for Aux Data Out channel of the pattern generator as well in second channel mode The figure below shows the N4916A De Emphasis Signal Converter connected between the Data Out port of the pattern generator and the DUT The figure below shows the N4916B De Emphasis Signal Converter connected between the Data Out port of the pattern generator and the DUT De emphasized signals are used to adapt the signal waveform to the characteristics of real transmission lines as found on printed circuit boards They can considerably reduce the impact of Intersymbol Interference ISI and hence allow longer signal paths De emphasized signals are prescribed by a couple of popular standards for high speed data transmission Agilent J BERT N4903B High Performance Serial BERT Setting up External Instrument s 2 De emphasis is a method that reduces the voltage of a digital signal if the generated level is high or low for more than one clock period The principle is illustrated in the figure below 0 1 1 1 0 01 0 1 1 0 Input Signal Output Amplitude p in Vpp mplitude The de emphasis amplitude is specified as a fraction of the output amplitude in percent or dB Post cursor de emphasis The figure above refers to a so called post cursor de emphasis You may wish to know how that is generated The De Emphasis Signal Converter splits the incoming signal into two br
166. FER bar displays the FER calculated upon a period of 200 ms FER 0 000 You can drag the yellow alarm threshold mark to change the FER Alarm Threshold cBER bar The cBER bar displays the cBER calculated upon a period of 200 ms cBER 0 000 You can drag the yellow alarm threshold mark to change the cBER Alarm Threshold FSR bar The FSR bar displays the FSR calculated upon a period of 200 ms FSR 0 000 You can drag the yellow alarm threshold mark to change the FSR Alarm Threshold DER bar The DER bar displays the DER calculated upon a period of 200 ms DER 0 000 You can drag the yellow alarm threshold mark to change the DER Alarm Threshold ISR bar The ISR bar displays the ISR calculated upon a period of 200 ms ISR 0 000 You can drag the yellow alarm threshold mark to change the ISR Alarm Threshold Error Indicators These indicators inform you about the following errors Bit error in the data stream located Loss of pattern synchronization Complete loss of data Loss of the error detector s clock signal Loss of the pattern generator s clock signal 10B symbol alignment is not done or is lost Loaded patterns have some disparity errors or contains invalid symbols Agilent J BERT N4903B High Performance Serial BERT 373 7 Evaluating Results Status Indicators Remote Indicator NOTE Show Error Messages 374 The two additional error indicators Symb Lock and 8b10b Error will only appear if you se
167. GPIB Address Your instrument is set to a default GPIB General Purpose Interface Bus address You can change this address by doing the following 1 On the Utilities menu click Change GPIB Address This starts the GPIB Address Change dialog box 2 Select an address in the New GPIB Address list 3 Click OK The new address will be applied and the dialog box will close Each instrument must be set to a unique GPIB address to avoid multiple instruments transferring data at the same time NOTE The default address is 14 however addresses from 0 to 31 may be used if the default address is the same as another instrument s GPIB address Address 21 is usually reserved for the computer interface Talk Listen address and should not be used as the instrument address Connecting Peripherals You can connect an external monitor mouse and keyboard to the instrument on the Serial BERT s rear panel To connect a mouse or a keyboard you can use either the USB ports or the PS 2 ports Agilent J BERT N4903B High Performance Serial BERT 529 10 Customizing the Instrument 530 NOTE NOTE CAUTION CAUTION For more details refer to Using an External Monitor on page 533 and Using the On Screen Keyboard on page 532 If the touchscreen is disabled the external keyboard and mouse are also disabled See Turning the Touchscreen Off On on page 531 on how to enable disable the touchscreen If caps lock is pressed on the keyboard this w
168. High Performance Serial BERT Setting up External Instrument s 2 NOTE Ensure that the timing adjustment has to be run at least once when either the N4903B N4876A or the cables connecting both are being exchanged in the test setup Also ensure that the timing adjustment has to be repeated whenever the operating temperature differs by more than 5 Celsius from the temperature at the previous timing adjustment How to Enable Disable M8061A Multiplexer with De emphasis Function To enable disable the M8061A Multiplexer with De emphasis function 1 Select the Mux with DeEmphasis function from the given list It opens the M8061A Connection dialog The following image shows the an example of M8061A Connect the data aux data and aux clock output of the pattern generator to the input of the M8061A Connect both instruments via USB Help 2 Click Enable button It enables multiplexer function that is connected between Data and Aux Data Output of J BERT s Pattern Generator and the input of the M8061A Enable Agilent J BERT N4903B High Performance Serial BERT 49 2 Setting up External Instrument s 50 NOTE Once the multiplexer function is enabled a M8061A menu entry is added to the External Instrument s sub menu File gt gt lt PEE gt gt External Instrument s Config FA Pattern gt gt a M8061A 3 To disable the multiplexer function clear the Mux with DeEmphasis function check box present
169. Jitter Tolerance Compliance test requires a standard You can use one of the predefined standards Additionally you can define and use your own user defined standard In terms of the Serial BERT a standard is represented by a sequence of paired numbers Each pair consists of a frequency value Hz and the associated jitter amplitude value UI Also part of a standard specification are the standard s name and descriptive text Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Predefined Standards The Jitter Tolerance Compliance option provides a number of predefined standards Standard Specifications Predefined standards specify Jitter frequency range The jitter frequency range may depend on the bandwidth of the receiver being tested Bitrate range Some standards define the jitter tolerance curve as a function of the bit rate If this is the case the edges of the jitter tolerance curve are shifted towards higher frequencies if the bit rate is increased and vice versa Jitter amplitudes in defined sections of the jitter frequency range The slopes between adjacent sections may be double logarithmic shown as straight lines in the double logarithmic display or are calculated according to certain algorithms Data pattern to be used This can be for example a certain PRBS polynomial or a specific test pattern Signal voltage range Most standards refer to specific devices Jitter
170. Level is generally recommended This ensures that neither too few nor too many bits are captured See also Target BER on page 449 Absolute Bits Errors If you enable Absolute Bits Errors the measurement proceeds to the next jitter frequency as soon as either the number of received bits or the number of errors is reached TIT Ensure that you capture a number of bits that suffices to verify with the desired probability that a measured point has a BER below the Target Bit Error Ratio Relax Time The relax time allows the DUT to recover in case it has lost synchronization due to an excessive high bit error rate The next BER measurement starts after this time has elapsed Optimal Measurement Setup To run the Jitter measurements you need to make sure that Global Jitter Sinusoidal Jitter and Periodic Jitter are enabled Agilent J BERT N4903B High Performance Serial BERT 465 8 Jitter Tolerance Tests 466 Select Point HTML Report Understanding the Results TIP Exporting Result Data This displays the list of measured points The list classifies the points with their frequency amplitude and status The status indicates whether the test passed or failed at that point This contains the results as follows 1 Test Configuration Details gives the User Comments Device Type Last Test Date Model Number Serial Number and Software Revision number 2 Conditions Pattern Data Rate Output Level ISI Trace Number Addition
171. Matched Cable Kit N4915A 011 The N4876A is controlled by the Serial BERT via a USB cable connected to ports at the rear of the instruments Agilent J BERT N4903B High Performance Serial BERT Planning the Test 1 Diagram 11 Connections of the M8061A with J BERT N4903B The Data In 1 Data In 2 and Aux Clk In of M8061A have to be connected to Aux Data Out Data Out and Aux Clk Out ports of the J BERT N4903B respectively The connections should be made using the matched cable kit with the part number M8061 61601 The Data Output and Data Output have to be connected to the device under test TIT Any unused output of the Pattern Generator has to be terminated with 50 Ohm The M8061A provides a differential signal to the DUT The M8061A can be used in combination with J BERT N4903B as shown below Aux clock output J BERT N4903B Jittered M8061A Mux 2 1 with de emphasis De emphasized data output Device under test Clock The M8061A is a 2 slot AXle module that can be controlled via USB from J BERT s user interface For using the CMI and DMI inputs of M8061A you can connect any sinusoidal source with a voltage swing of 0 to 400 mV single ended You can for instance use the sinusoidal sources of N4903B option J20 Use a single ended connection from any of the P1 or P1 complement P2 or P2 complement outputs and terminate all unused outputs with 50 Ohm Though it is possible to connect the normal and complement of
172. Model code N49168 Serial number Proto R2 008 General Tasks More Information 9 How do I get drivers Firmware 0 7 02 2 Where can I find programming samples amp 49 myinterface c IZ MyInstrument X Myalias 32 bit Agilent VISA is the primary VISA library For detailed instructions refer to the Connectivity Guide which is part of the Agilent 10 Libraries Suite Documentation Connecting to Pattern Generator and DUT N4916A After the USB port has been configured you can remove keyboard and mouse and make the signal connections Agilent J BERT N4903B High Performance Serial BERT 543 10 Customizing the Instrument 1 Usethe N4915A 004 cable or a 2 4 mm cable m m longer than 350 mm and connect the pattern generator s Data Out port to the Input port of the N4916A 2 Mountthe SMA 50 Ohm termination on the unused Data Out port of the pattern generator using the SMA to 2 4 mm adapter 3 Enable the De Emphasis Signal Converter as described in How to Enable Disable N4916B De Emphasis Function on page 46 and set the levels such that the DUT will not be damaged 4 Usea matched pair of 2 4 mm cables to connect the differential outputs of the N4916A to the DUT If only one output of the de emphasis signal converter is needed connect a 2 4 mm 50 Ohm termination to the unused output Connecting to Pattern Generator and DUT N4916B After the USB port has been configured you can remove keyboard and mous
173. POS and Bit Comparison without USB3 1 SKPOS error ratio modes 3 f you selected Normal synchronization mode choose whether you want Automatic or Manual synchronization 4 Specify the Sync Threshold at which the pattern will be recognized at synchronized 5 n Manual synchronization mode use the Sync Now button to start the pattern synchronization This is recommended whenever you did changes to the pattern setup the voltage levels or the hardware connections for example altered cable lengths 6 Select the Mask Resync Errors option to enable the masking high error counts prior to re synchronization of the error detector feature For more details see Mask Resync Errors on page 218 Use the drop down list to specify either counting zero or one error when masking resync errors By default counting with one error is selected Unselecting the Mask Resync Errors option disables the Mask Resync Errors feature NOTE By default the Mask Resync Errors feature is enabled Agilent J BERT N4903B High Performance Serial BERT 217 5 Setting up the Error Detector 218 7 Press OK to confirm your settings Pattern Synchronization Reference NOTE The error detector can synchronize the data patterns in the following ways n Normal Sync Mode you can choose between Automatic Sync With this option selected the error detector constantly tries to synchronize the patterns when the BER threshold is exceeded Manual synchronizat
174. Pattern The Serial BERT supports 12 so called memory slots for storing user patterns The slots are stored in files named Upatl ptrn to Upat12 ptrn To load a pattern that is stored in one of these memory slots choose this option and select the desired slot from the list 24n 1 PRBS Select this option to use a hardware generated pattern of length 2 n 1 See How the Hardware Generates PRBS on page 94 for details e 2 n PRBS Select this option to use a memory based pattern of length 2 n e Mark Density PRBS Use this options to generate memory based 2 n PRBS patterns with specified mark densities Select the desired value from the list The available values are 1 8 1 4 1 2 3 4 and 7 8 Zero Sub PRBS Use the Zero Substitution function to insert a longer row of zeros into a memory based 2 n PRBS pattern The length of the zero row can be freely specified up to the length of the pattern You can select the pattern size according to your needs The pattern lengths available depend on the type of signal you have selected Enable this checkbox if you wish to use a user defined sequence The patterns used in a sequence are defined in the Sequence Editor see also User Defined Sequences Concepts on page 96 Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 Bit Recovery Mode BRM Only available on the Error Detector Pattern page Enables the error detector s bit recovery mode For details se
175. Previous Period Calc Bit Count 792 496 000 000 482 752 000 000 a Calc Bit Error Ratio CBER 0 00000000 0 00000000 Calc Bit Error Count 0 0 Compared Symbol Count 79 249 600 000 48 275 200 000 Symbol Error Count 0 0 I Symbol Error Ratio SER 0 00000000 0 00000000 Received Frame Count 19 812 400 000 12 068 800 000 NOTE The Accumulated Results window for 8B 10B Comparison does not contain tables for G 821Measurement Internal Results and Burst Results 384 Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 Ratio Graph This graph displays the delta symbol errored 1 s ratio delta symbol errored 0 s ratio and total delta symbol error ratio at data points over the entire accumulation period For more information see Ratios Graph on page 379 Accumulated Results In the 8B 10B Comparison Accumulated Results table the following values are listed NOTE If the Block Length is equal to zero then use Frame Count instead of Pattern Count for calculating Accumulated Results Calc Bit Count The total number of calculated bits considered for the accumulation period is displayed here Calculated Bit Count Error Count Error Ratio is dependent on cBER configuration mode SER FER ISR DER FSR If cBER is based on SER then Calc Bit Count Block Length x Pattern Count x Conversion Factor Where Block Length is equivalent to Pattern Length in All Bits mode and specified Block Length in Block m
176. Primitives are inserted or deleted for clock tolerance compensation These are not compared and therefore cannot be counted as errors Filler Primitives contain filler symbols A maximum of 4 alternative filler primitives can be used Each filler primitive can consist of upto 4 filler symbols Filler symbols are separated by a comma Example For SAS we have 4 filler primitives each consisting of 4 filler symbols K28 5 D10 2 D10 2 D27 3 K28 5 D7 0 D7 0 D7 0 K28 5 D1 3 D1 3 D1 3 K28 5 D27 3 D27 3 D27 3 The following are the supported standards along with their filler symbols Agilent J BERT N4903B High Performance Serial BERT Table 19 Setting up the Error Detector 5 Standard USB3 0 SATA MIPI MPhy Display Port SAS PCle 1 amp 2 Wild Cards for Filler Primitives Filler Symbols Skip K28 1 K28 1 10B Symbol alignment K28 5 Align K28 5 D10 2 D10 2 D27 3 10B Symbol alignment K28 5 Filler K28 1 10B Symbol alignment K28 5 Filler Empty String 10B Symbol alignment K28 5 Align K28 5 D10 2 D10 2 D27 3 K28 5 D7 0 D7 0 D7 0 K28 5 D1 3 D1 3 D1 3 K28 5 D27 3 D27 3 D27 3 10B Symbol alignment K28 5 Skip K28 5 K28 0 K28 0 10B Symbol alignment K28 5 The wild cards allow you to set one or more out of the maximum of four symbols of a filler primitives as don t care In this case all allowed D and K symbols will match and are removed from incomming DUT data For example K28
177. R Range of a Color on page 294 How to Prepare the Eye Opening Measurement To prepare an Eye Opening measurement to test a shielded cable 1 Disable the pattern generator outputs by pressing the 0V Disable button in the PG Setup gt Data Output screen Use a shielded cable to connect the pattern generator s Data Out port and the error detector s Data In port Terminate all non connected pattern generator output ports with 50 O Switch to the Pattern panel and press Pattern Select Select an appropriate pattern for example a pure 2 15 1 PRBS for both the pattern generator and the error detector For the pattern generator setup you need to specify the logic levels and the bit rate Select ECL levels and a clock speed of 1250 MHz in this example This corresponds to a clock period of 0 8ns See Setting up the Pattern Generator Concepts on page 121 for more information Set up the error detector so that the input range and the termination matches the pattern generator s levels Select an Input Range from 2V to 0V Setthe Data Termination to 2V Setthe Alignment BER Threshold to 1E 6 Setthe Clock Setup to Clock Data Recovery to get the error detector s clock from the incoming data stream Enable the pattern generator outputs by pressing the 0V Disable button Press Sync Now and then Auto Align to find the optimum sampling point Check that the synchronization and the alignment were successful None
178. R cBER FSR DER ISR Two additional error indicators Symb Lock and 8b10b Error will appear at the lower pane of GUI if you select the error ratio as 8B 10B Symbol Comparison from the Error Detector Error Ratio s window For more details see Symbol Lock Indicator on page 192 and 8b10b Error Indicator on page 192 NOTE Indicator at the upper pane of the GUI will change For more details see Status Indicators on page 372 Depending upon the selection SER FER cBER FSR DER ISR the Results sub menu will now show either SER Results FER Results cBER Results FSR Results DER Results ISR Results as instantaneous measurements TIT Depending upon the selection SER FER cBER FSR DER ISR the Status NOTE 11 Set the symbol alignment mode You have the following options Select Automatic mode to enable automatic alignment In automatic mode the conversion factor for SER ISR and DER is 10 For FER and FSR the automatic conversion factor is the specified block length or if not specified the expected pattern length including filler symbols Select Manual mode to disable the automatic alignment and initiate re alignment if alignment is not already done or lost Click ReAlign button to re align the symbols again in case if the symbol alignment is lost Symbol Alignment Automatic Manual 188 Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 NOTE The Automatic mode i
179. RT Solving Problems 9 Temperature Alert PG critical above normal Shutting down instrument Immediate startup not recommended The Serial BERT will be shut down your instrument settings will not be saved Troubleshooting Overheating If the Serial BERT has indicated overheating do the following Shutdown the instrument and let it cool 45minutes to 1hour Reduce the environmental temperature Make sure that the fans are running and the ventilation holes are not blocked Ifthe problem continuously recurs contact Agilent support Do not operate the Serial BERT before it has cooled down again Otherwise your CAUTION instrument may get damaged When you are switching on the Serial BERT after it was shut down due to a temperature error you will see a message boxthatinforms you about which module caused the problem Critical Error Pattern Generator Bios Power On Errors Previous instrument shutdown forced by critical temperature Offline Error Detector Bios Power On Errors Previous instrument shutdown forced by critical temperature Offline Agilent J BERT N4903B High Performance Serial BERT 479 9 Solving Problems N4916B System Setup Calibration N4916B System Setup Calibration NOTE 480 The input timing of the N4916B has to be calibrated to ensure error free operation for all data rates This calibration is required once for a dedicated setup of N4903B N4916B and the cable kit be
180. RT N4903B High Performance Serial BERT Advanced Analysis 6 How to Change the Output Levels Properties In the example measurement we have set the focus on speed 100 threshold levels and 1 000 000 compared bits per measuring point You may wish to obtain more precise results 1 Press Properties and switch to the Parameters tab 2 Increase the Number of Compared Bits to 100 000 000 Remember One failure per 1 million bits yields a BER resolution of 109 One failure per 100 million bits yields a BER resolution of 108 3 Decrease the Resolution the threshold step size to 2mV This gives us 500 steps per Volt 4 Press OK to confirm your changes 5 Press the Start button to repeat the measurement with the new parameters The measurement now takes more time than the previous but it is also much more precise 800 0 mv Trace 1002 Points 800 0 mv A0V Adv 42V 43V 44v 45V ABV ATV ABV BER Threshold 1 000e 6 Logarithmic 1 000e 7 1 000e 5 1 000e 3 1 000e 1 1 000e 8 1 000e 6 1 000e 4 1 000e 2 1 000e When you move the BER Threshold bar you will see that some of the calculated values change This is explained in Explanation of the Numerical Results on page 211 How to Use the Different Views To switch between the different views of the Output Levels measurement 1 Press Properties and switch to the View tab 2 Select dB Histogram versus Threshold This gives you another view of the measured data
181. RT s PG Datarate to 3 Gb s Set the J BERT s ED Datarate to 6 Gb s Inthe ED Sampling Point Setup window press Auto Align button The error detector should finish the auto alignment successfully and report a BER of 0 000 M8061A Status LEDs The front panel of the M8061A includes the folloiwng LEDS FAIL ACCESS FAIL ACCESS These LEDs have the following functionality e Both LEDs are on during the startup Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 After the M8061A is ready the red FAIL LED goes off and the green ACCESS LED stays on When the green ACCESS LED is on you can start the firmware When the firmware is communicating successfully with the module the green ACCESS LED starts flashing Agilent J BERT N4903B High Performance Serial BERT 523 9 Solving Problems 524 Agilent J BERT N4903B High Performance Serial BERT O60 10 e e Customizing the Instrument Customizing the Instrument Concepts The Serial BERT provides various utilities for adjusting the instrument to your personal preferences Restoring the System Agilent Recovery System Procedures The Agilent Recovery System can be used to repair the system in case of serious malfunction CAUTION When the operating system is restored all personal settings programs and user data will be lost Make sure that important data is saved to external media before restoring the system The Agile
182. Ratio Unit Marks the lower edge of the eye window boundary Right Boundary Marks the higher edge of the eye window boundary Set the criteria for eye width and eye height Eye width at crossing point This parameter gives the eye width at the crossover Total is 100 of the display screen Eye width custom This parameter gives the eye width at a specific percent of the eye amplitude Starting from low level to high level Eye height This parameter gives the eye height at a specific percent of Tui Starting from left to right Set the criteria for the decimal placing in the return values Decimal Places This parameter sets the decimal placing for the Automated Eye Parameter Measurement Set the criteria for the optical parameters These are used to convert level values to optical power values yt power value level value dark level conversion gain Dark Level Internally generated offset signals Conversion Gain Is used to convert voltage level into power dBm Output will be displayed in dBm power value dBm 10 log power value W 1mW OW Output will be displayed in Watt Set the criteria for the Extinction Ratio Unit Ratio Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Ratio of two optical power levels low level and high level dB Unit for measuring the extinction ratio ratio dB 10 log ratio 96 Extinction Ratio in percentage Gr
183. Remove Hardware icon A window containing a list of the USB devices appears 3 Click on the popup message Safely Remove Mass Storage Device 4 Confirm the acknowledgement message Save to remove hardware 5 Check that the LED light of the USB stick is off and unplug the USB stick Saving Instrument States Saving the current instrument state stores all the setup parameters for the instrument s hardware and software To save the setup with the current filename and directory path do the following e Inthe File menu click Save and then Save Instrument State The current instrument state will be immediately saved To save the setup with a new filename or directory path do the following In the File menu click Save and then Save Instrument State As This opens a dialog box which allows you to create subdirectories rename setup files or overwrite existing setup files Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 NOTE Ensure your instrument state files have the extension btz If you do not specify an extension or use a different extension the instrument may not recognize the file as an instrument state file NOTE The default pathname for user defined instrument states is C lt instrument model gt Settings The number of instrument states you can save is limited only by the available space on the internal hard disk or floppy disk Each instrument state uses approximately 6 kB of
184. S N4876A Status LEDs This following section explains about the different LEDs that are available on the front panel of the N4876A 28 Gb s Multiplexer 2 1 Output Activity amp LAN Power Table 35 LED Name Description Output Signals the state of the output amplifier Constantly off when output is disabled 0 0V Constantly green when output is enabled user defined levels Flashing orange when an overload condition has been detected Constantly orange during boot phase of the instrument Activity Flashes green when instrument is accessed on the remote interfaces Constantly orange during boot phase of the instrument Agilent J BERT N4903B High Performance Serial BERT 517 9 Solving Problems Table 35 LED Name Description LAN Constantly green when LAN is operable Constantly red when LAN is not operable Flashing green when device identification is enabled Constantly orange during boot phase of the instrument Power Constantly orange when instrument is OFF in power standby mode Constantly green when instrument is ON powered Problems with the M8061A Problems with the M8061A Concepts 518 When you have opened the External Instrument s Config window and have selected M8061A from the external instruments select Mux with DeEmphasis under Enable Function column in the list the M8061A Connection dialog appears Cannot Connect Problems An error message may appear when you now press Enable Chec
185. STB RST TST SYST ERR Enter String to Print or Query Instr Lang x C Custom C EEE 488 2 viPrintf viScanf viQueryt SCPI For Help press F1 NUM 15 01 38 If you could proceed until here the N4916B power supply and the processor are fully functional If not the N4916B is probably defective Ensure that the User Software can Access the N4916B 1 From the Windows task bar restore the user interface click GUI Agilent 4900 Series 2 Press the green Preset key close to the front USB connector of the Serial BERT or from the menu bar select File Preset Instrument State 3 Connect the Data Out of the Serial BERT to the Input of the N4916B 4 From the Navigation Menu select External Instrument s and then click Config sub menu The Config window shows a list of all instruments connected to the Serial BERT Confirm the presence of N4916B enrty in that list 5 Corrosponding to N4916B entry select the De Emphasis function check box from the given options under Enable Function column It opens the De emphasis Signal Converter Connection dialog 6 Click Enable button It enables N4916B De emphasis signal converter that is connected between data out of Serial BERT and the DUT Agilent J BERT N4903B High Performance Serial BERT 503 9 Solving Problems 7 Enable Once the de emphasis function is enabled you can see the de emphasis parameters in the PG Output screen Pre Cur Post Curl Post Cur2
186. Solving Problems 9 7 Again from the Config window corresponding to M8061A clear the Mux with Deemphasis function check box It will disable the multiplexer function and the PG Output screen will allow to control the output levels of Data and Aux Data at the connectors of the PG again If you could proceed until here you have successfully connected the application software to the M8061A Settings for Multiplexer Tests The following are the settings to test the output of the M8061A 28 Gb s Multiplexer with Deemphasis Preset Instrument State Set the PRBS 2 7 1 Inthe PG Bit Rate Setup window set the Data Rate to 3 Gb s Enable Mux with DeEmphasis function in the Config window Inthe M8061A window set the Data Offset to 0 V nthe M8061A window set Data Amplitude to 400 mV Check the Output of the M8061A Using a Scope The approach shown here is checking the output signal using an oscilloscope 1 Connect the Outputs of the M8061A to an oscilloscope 2 Connect the Trigger Output of the pattern generator to the trigger input of the oscilloscope 3 Terminate M8061A Data Out with 50 Ohm 4 Terminate PG Trigger Out with 50 Ohm 5 Terminate PG Data Out and Aux Data Out with 50 Ohm Then the oscilloscope should show an output eye like the following Eye height 380mV Eye width 160 ps If you see a display like above everything is OK Agilent J BERT N4903B High Performance Serial BERT 521 9 Solv
187. T SYST ERR Enter String to Print or Query Instr Lang j C Custom C IEEE 488 2 viPrintf viScanf viueryf SCPI For Help press F1 NUM 11 27 57 If you could proceed until here the N4916A power supply and the processor are fully functional If not the N4916A is probably defective Ensure that the User Software can Access the N4916A 1 From the Windows task bar restore the user interface click GUI Agilent 4900 Series 2 Press the green Preset key close to the front USB connector of the Serial BERT or from the menu bar select File Preset Instrument State 3 Connect the Data Out of the Serial BERT to the Input of the N4916A Agilent J BERT N4903B High Performance Serial BERT 495 9 Solving Problems 496 From the Navigation Menu select External Instrument s and then click Config sub menu The Config window shows a list of all instruments connected to the Serial BERT Confirm the presence of N4916A enrty in that list Corrosponding to N4916A entry select the De Emphasis function check box from the given options under Enable Function column It opens the De emphasis Signal Converter Connection dialog Click Enable button It enables N4916A De emphasis signal converter that is connected between data out of Serial BERT and the DUT Enable Once the de emphasis function is enabled you can see the de emphasis parameters in the PG Output screen Post Cur Unit 8 00 dB dB v 7 Again f
188. T r T r T T e 1 GBit s eM 4 3 2 5 GBit s 100000 4 7 pee HS OE 8 125 GBit s 10 3125 GBit s E E i i E i i E 10000 4 4 e EE DE eer eee M ise E H i H i x i H i i H E i 2 5 1000 4 E z 100 d d Re Mel as aa Reeve Eee ERR i t i t i t i 16KBit 32KBit 64KBit 128KBit 256KBit 512KBit 1MBit 2MBit 4MBit MBit 16MBit 32MBit Acquisition Depth NOTE A high Acquisition Depth requires a high degree of computational effort and hence time Sample Point Offset By default the sampling point for the measurement is positioned 0 5 clock periods or Ul ahead of the present analyzer sampling point If the sampling point is set in the middle of the eye this corresponds to the transition point of the incoming signal If you have set the sampling delay manually this is the time offset from that point This option allows you to fine tune the sampling point The unit is UI unit intervals This makes the setting independent of the present clock frequency Set the FFT window for the FFT Calculation The following windows are provided Uniform Hanning sometimes also called Hann Hamming Blackman The uniform window is no window at all This setting supplies the original error record to the FFT Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 The other windows have the shapes illustrated in the figure be
189. Technology Licenses The hardware and or software described in this document are furnished under a license and may be used or copied only in accordance with the terms of such license Restricted Rights Legend If software is for use in the performance of a U S Government prime contract or subcontract Software is delivered and licensed as Commercial computer software as defined in DFAR 252 227 7014 June 1995 or as a commercial item as defined in FAR 2 101 a or as Restricted computer software as defined in FAR 52 227 19 June 1987 or any equivalent agency regulation or contract clause Use duplication or disclosure of Software is subject to Agilent Technologies standard commercial license terms and non DOD Departments and Agencies of the U S Government will receive no greater than Limited Rights as defined in FAF 52 227 14 June 1987 or DFAR 252 227 7015 b 2 November 1995 as applicable in any technical data Safety Notices CAUTION A CAUTION notice denotes a hazard It calls attention to an operating procedure practice orthe like that if not correctly performed or adhered to could result in damage to the product or loss of important data Do not proceed beyond a CAUTION notice until the indicated conditions are fully understood and met A WARNING notice denotes a hazard It calls attention to an operating procedure practice or the like that if not correctly performed or adhered to could result
190. The errors became constant at the point of failure Digital circuitry often has parallel architecture in which data lines are multiplexed in stages into one serial data line This example assumed that a PRBS pattern was going through digital circuitry that had a total multiplexing ratio of 256 1 The constant errors were produced from one data line that became stuck high in the first stage of multiplexing The data line that was stuck high produced errored 0 s 5096 of the time with the PRBS pattern Therefore at the serial rate one out of every 512 bits was measured as an errored 0 1 256 50 1 512 In addition the inverse of the error ratio was a multiple of 2 Additional Information In this example the analyzer could have been in manual sync or auto sync mode with 1E 2 as the sync threshold BER If the sync mode settings were different the results would not have been measured and displayed in the same way The BER in this example approx 2E 3 was greater than the default sync threshold BER 1E 3 If the analyzer was in the default auto sync mode it would have activated the synchronization functions and continually shifted the position of the reference pattern in an attempt to find a lower BER NOTE The accumulated results table displays the total accumulated error ratio Look in the measurement log file to see the error ratio during the period of constant errors Agilent J BERT N4903B High Performance Serial BERT 473 9
191. The extra bits are ignored In symbol format Patterns lengths must be a multiple of 10 bits ASCII Pattern Files Examples Standard Pattern in Text Version EPA 2 0 Format Hex Description This is a simple standard pattern Count 1 Length 128 Data 01 02 03 04 10 20 40 80 ff 00 ff 00 ff 00 ff 00 Alternating Pattern in Text Version EPA 2 0 Format Text Description This is a simple alternating pattern Count 2 Length 96 Data 01 02 03 04 10 20 40 80 ff 00 ff 00 Data 01 02 03 04 10 20 40 80 ff 00 ff 00 Pattern Resolutions and Lengths The Serial BERT has 32Mbits 8 MB RAM for storing patterns As discussed in How the Serial BERT Generates Memory Based Patterns on page 69 all patterns are loaded into the RAM until a 512 bit boundary is reached If the pattern Agilent J BERT N4903B High Performance Serial BERT 67 3 Setting up Patterns NOTE is an odd number of bits long it is loaded into the RAM 512 times Because of the restricted memory space there are also restrictions to the pattern lengths For example a user pattern that is 256 001bit bit long would have to be loaded into the RAM 512 times to reach the 512 bit boundary Such a pattern would occupy more than 128 Mbits in the RAM The Serial BERT handles such patterns by rounding them off to the nearest length that would fit in the memory For the above example the length would be rounded off to 256 000 bits This rounding off factor is known as the pattern s r
192. The maximum number of errors that this function supports is one error per 128 bit block Agilent J BERT N4903B High Performance Serial BERT 159 4 Setting up the Pattern Generator 160 NOTE Error Add Setup Dialog Box The Error Add Setup dialog box provides the following configuration options Error Add Setup Channel Select C DATA C AUX BOTH Error Source External Error Add internal Error Insertion Rate Error Rate 1 0E 03 Average Number Of Bits Between Errors _ Off only tool bar single error add available Cancel Channel Select With this option you can apply the error insertion either on Data Aux or Both channel s Aux and Both options will be disabled in single channel mode External Error Add With this option selected when a rising edge is detected at the Error Add port an error will be generated in the output stream by flipping a single bit within a bit block of 128 bits The signals received at this port must be TTL compatible Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 Internal Select this option to use internal error insertion functions Error Rate Allows to select from a preset bit error ratio from 1 000E 02 to 1 000E 09 Errors will be added to the output pattern to produce this error ratio Average Number of Bits Between Errors Allows you to enter the average number of valid bit
193. This is the sample delay coordinate of the center of a bounding box around the BER threshold contour line Optimal Sample Threshold This is the sample voltage coordinate of the center of a bounding box around the BER threshold contour line Unavailable Values If you make a measurement that results in an incomplete eye no closed contour line at the BER threshold the numerical results cannot be calculated In this case the software will issue a warning message Ifthis happens you should first of all checkthe settings of the low and high decision threshold voltages Agilent J BERT N4903B High Performance Serial BERT 301 6 Advanced Analysis Error Location Capture Understanding the Result Parameters All result parameters are calculated from the BER threshold contour line and its bounding box So all parameters change with the BER threshold The result display of the Eye Opening measurement shows the maximum eye opening time the maximum eye opening voltage and the position of the optimum sampling point These parameters are derived as illustrated in the figure below a Ne p TA Dou pein tli Se i P Eye opening time S e 3 7 r N A 7 d F ES d Eye opening voltage LLL t Optimum sample point Error Location Capture Concepts 302 The Error Location Capture measurement allows to capture the position of an errored bit in a memory pase
194. Tolerance Tests This defines the minimum jitter amplitude step size The test for one frequency stops when this step size is reached For a description of this method refer to Binary search on page 437 Extended binary search For extended binary search you can set the coarse and fine accuracy The test for one frequency stops when the BER limit is crossed with fine accuracy amplitude steps For a description of this method refer to Extended binary search on page 437 Minimum and Maximum Jitter Curve The minimum and maximum jitter curves have been introduced to speed up the jitter tolerance characterization measurement by specifying lower and upper boundaries for the search This reduces the number of measurements needed Instead of starting and stopping the measurement at a fixed jitter level for all the frequencies it is possible to specify minimum and maximum jitter curves at which the measurement starts and stops depending on the search algorithm This avoids measuring points where the DUT is known to operate If for a particular frequency the starting pointfails then the search algorithm reverts to the minimum jitter level to start the search for that frequency The next frequency step will start based on the minimum curve again This gives a complete picture of the jitter tolerance even if the start values are not carefully selected The format used to specify the minimum and maximum jitter curves are same as that of existing jitter
195. Total Jitter Uncertainty The maximum of the uncertainties of both slopes Measured as the time between a point with a BER greater than the specified Total Jitter BER Threshold and the next point with a BER less than the specified Total Jitter BER Threshold left slope and vice versa right slope Total Jitter BER Threshold The BER threshold specified for the measurement Optimal Sample Point Delay The mean value of the left and right bathtub Total Jitter BER Threshold intersections Output Levels Output Levels Concepts The Output Levels measurement allows you to characterize the behavior of the output levels of a device under test DUT The sampling delay is fixed The error detector s decision threshold is automatically swept within a user defined range A direct result is the determination of the optimum decision threshold level for receiving data from the DUT with maximum confidence Optimum sampling delay Agilent J BERT N4903B High Performance Serial BERT 263 6 Advanced Analysis 264 Three Available Views The Output Levels measurement provides three different graphical views to visualize the calculated results BER versus Threshold This graph shows the relationship between the analyzer decision threshold and the resulting BER It presents the raw data dB Histogram versus Threshold This graph shows the relationship between the analyzer decision threshold and the derivative of the bit er
196. UI Margin 0 96 Condition 2 23 1 PRBS 2 488 Gb s Comment Passed Points frequency amplitude number of bits bit error rate number of errors result 1000 8 5 3013355404 0 0 passed 1790 61312892 8 5 3015843721 0 0 passed 3206 295377462 8 5 3015843721 0 0 passed 5741 234598079 8 5 3015843721 0 0 passed 10280 33004753 8 5 3015843725 0 0 passed 18408 09395274 8 5 3013355401 0 0 passed 22100 8 5 3015843721 0 0 passed 32961 77471018 5 69902566387 3015843724 0 0 passed 59021 78654856 3 182723041524 3015843725 0 0 passed 105685 1858862 1 777448735363 3013355405 0 0 passed 189241 2813802 0 9926481084359 3013355401 0 0 passed 338857 922973 0 5543621301573 3015843723 0 0 passed 606763 4457141 0 3095934689654 3015843724 0 0 passed 1086478 592045 0 1728980224511 3015843725 0 0 passed 1875000 0 1 3013355405 0 0 passed 1945462 831206 0 1 3015843721 0 0 passed 3483571 287383 0 1 3015843721 0 0 passed 6237728 482718 0 1 3015843722 0 0 passed 11169358 5158 0 1 3015843771 0 0 passed 20000000 0 1 3013355409 0 0 passed Failed Points frequency amplitude number of bits bit error rate number of errors result This file can be imported into spreadsheet applications such as Microsoft Excel for example Use the semicolons as delimiters for the rows or columns E 1 2 2 3 4 5 6 7 1 00E 03 8 50 30133
197. UI 0 10 UI 0 10 UI 0 30 UI 0 50 UI 0 70 UI Relative 0 60 UI 0 40 UI 0 20 UI 0 00 UI 0 20 UI 0 40 UI 0 60 UI Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Markers To analyze the graphs at a particular point you can use the markers Additionally you can display all related values for the markers in the marker readout Pressing the Reset Markers button will set the markers back to the default positions 4 00e 0 1 Trace 107 Points Ul 2 500 ns All Errors T 1 00e 1 1 00e 2 1 00e 3 1 00e 4 Markers Time BER 1 0De 5 Upper Right 0 501 UI 3 240e 1 Lower Left 0 470 UI 1 00e 6 Delta 0 970 UI 3 240e 1 M CENE RECREEPEECNEF E or sr s sm el Logarithmic _9 60 UI 0 36 UI 012UI 0 12 UI 0 36 UI 0 60 Ul Relative 0 48 UI 0 24 UI 0 00 UI 0 24 Ul 0 48 UI 0 72 UI Zoom Several zoom factors are available When you show the zoom graph you can also allow the zoom graph to track the mouse or your finger if you are working directly on the Serial BERT 100e Q 1 Trace 107 Points Ul 2 500 ns All Errors 1 00e 1 1 00e 2 1 00e 3 1 00e 4 1 00e 5 1 00e 6 BER Thresholii 1 00e 7 Logarithmic 0 60 UI 0 36 UI 012UI 0 12 UI 0 36 UI 0 60 UI Relative 0 48 UI 0 24 UI 0 00 UI 0 24 UI 0 48 Ul 072 UI Show Measured Points If you want to see the points that have actually been measured choose Show Measured Points 100e 0 1 Trace 55 Points Ul 2 500 ns All Errors 1 00e 1 1 00e 2
198. Ul 333 333 ps All Errors 2 70e 10 2 60e 10 250e 10 240e 10 230e 10 220e 10 2 10e 10 200e 10 1 90e 10 1 80e 10 1 70e 10 1 50e 10 1 50e 10 140e 10 130e 10 120e 10 140e 10 1 00e 10 9 00e 9 8 00e 9 7 00e 8 6 00e 9 5 00e 9 4 00e 9 3 00e 9 2 00e 9 1 00e 9 Linear 230 00 ps 210 00 ps 190 00 ps 170 00 ps 150 00 ps Relative 240 00 ps 220 00 ps 200 00 ps 180 00 ps 160 00 ps 140 00 ps Estimated Total Jitter The Estimated Total Jitter TJ allows you to predict the jitter expected for very low bit error rates that would take a long time to measure It is obtained by extrapolating the measured BER curves The TJ is estimated by extending the BER curves based on the points detected between the BER Threshold and the Minimum BER for RJ DJ Separation to the Residual BER for RJ DJ Separation level The estimated TJ is the period minus the width of the measured eye The following graphic is not available from the measurement software It is inserted NOTE AE here to show how the TJ period is calculated Agilent J BERT N4903B High Performance Serial BERT 245 6 Advanced Analysis The Uncertainty Band 246 Extrapolated Bathtub log scale 10 r r r 1 f 10 i BER Threshold 10 j 10 4 T Min BER for RJ DJ Gor T Separation ae I 10 1 10 ee JM Hu Residual BER for E Estimated TJ 10 i J 10 1 Li 1 1 1
199. a output of the pattern generator to the input of the N4916B Connect both instruments via USB on the rear panel Cancel Help 2 Click Enable button It enables the de emphasis signal converter that is connected between Data Out of Serial BERT and the DUT Enable Once the de emphasis function is enabled you can see the de emphasis parameters in the PG Data Output screen For more information refer to the section Controlling the N4916B De Emphasis Signal Converter on page 133 3 To disable the de emphasis function clear the De Emphasis function check box present on the Config window It will disable the de emphasis function and the respective parameters from the PG Data Output screen also disappears 46 Agilent J BERT N4903B High Performance Serial BERT Setting up External Instrument s 2 How to Enable Disable N4916B Clock Multiplier Function To enable disable the Clock Multiplier function 1 Select the Clock Multiplier function from the given list It opens the Clock Multiplier Connection dialog The following image shows the an example of N4916B Clock Multiplier Connection Dialog Connect the data output of the pattern generator to the input of the N4916B Connect both instruments via USB on the rear panel Cancel Help 2 Click Enable button It enables Clock Multiplier function that is connected between Clock Out of PG of Serial BERT and Clock In of ED of Serial BERT Enable Once the clock m
200. a patterns It can also be caused by internal or external periodic effects Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 The Spectral Jitter measurement detects even small periodic components that may be hidden in a high level of random noise It informs you about the frequencies of such components and measures their contribution to the total jitter This helps to identify jitter sources and to reduce or eliminate their influence Prerequisites for Using the Spectral Jitter Measurement It is important that the initial sampling point is in optimum position i e in the middle of the eye opening because the Spectral Jitter measurement derives its starting sampling point from that point See also General Requirements on page 228 How the Spectral Jitter Measurement Works The Spectral Jitter measurement performs a capture and compare operation in the jitter region The number of compared and captured bits is adjustable By default the analyzer sampling point is set to an offset of 0 5 analyzer clock periods or unit intervals Ul from the optimum sampling delay Jitter width at threshold Threshold level Optimum sampling delay for the signal Sampling delay for the Spectral Jitter measurement adjustable Capture and compare means that a certain amount of incoming data is captured and also compared with expected data in real time The number of captured data points is adjustable W
201. able jitter generation Press the respective checkbox i Jitter on off 3 Activate Sinusoidal jitter SAG ma mu mul 600 Hz 4 Activate Periodic jitter otl Apru amui 10 0000 Mhz 5 Set the jitter amplitude to 0 UI for both Sinusoidal jitter and Periodic jitter Running the Test Once you have specified the Frequency BER and Search criteria press the Start button to run the test Start l The bar in the lower right hand corner of the screen shows the progress Agilent J BERT N4903B High Performance Serial BERT 441 8 Jitter Tolerance Tests 442 You can also abort the test at any time by pressing the Abort button Once the measurement has finished the jitter composition shown in the Jitter Setup window is restored When the measurement is either completed or aborted you can obtain a list of measured points by clicking the following tab which appears once the measurement is stopped Select Point Clicking on the above tab you get the list of measured points 1 000 kHz 1 812 kHz 3 282 kHz 5 345 kHz 10 770 kHz 18 511 kHz 35 346 kHz 54 033 kHz 115 000 kHz 210 144 kHz 380 692 kHz 689 654 kHz 1 249 MHz 2 263 MHz 4 100 MHz 7 428 MHz 13 456 MHz 24 377 MHz 44 160 MHz 80 000 MHz Amplitude 150 0 Ul 150 0 UI 150 0 Ul 150 0 Ul 138 3 UI 76 9 Ul 42 4 Ul 23 4 Ul 12 9 Ul 7 1 UI 3 3 Ul 2 2 UI 1 2UI 663 mUI 660 mul 660 mUI 660 mUI 660 mUI 660 mUI 660 mUI Pass Fail Passe
202. age levels and the Termination voltage and enables termination 4 f necessary change the termination voltage by entering a new value You can also use the front panel knob to change the value The logic level automatically changes to Custom Selecting the wrong terminations may damage your device See Why Incorrect CAUTION ies ee Terminations Could Damage Your Device on page 129 5 Connect the DUT s input ports to the Serial BERT s output ports Do not apply external voltages to the pattern generator outputs For more CAUTION 4 A m information see AC Coupling and Bias Tees on page 130 Output ports of the pattern generator that are not connected to another device must be terminated with 50 Ohm to prevent the Serial BERT from damage Note that for unconnected ports an internal protection circuit automatically reduces the output voltages to safe levels Agilent J BERT N4903B High Performance Serial BERT 131 4 Setting up the Pattern Generator 132 NOTE Adjust Output Levels optional Data Aux Data Clock and Trigger Ref Clock offset and voltage levels can be adjusted This is typically done when you want to tune your BER measurement or stress the device You can use the knobs on the Serial BERT s front panel to fine tune the data and clock amplitudes and offsets If you want to set a specific value you can use the numeric keyboard To enter specific values for the outputs from the keyboard 1 From th
203. akage in Assumed A A A _ i REA ih spectrum periodicity N V J V V J V 8 Time record When you are using a window please note No window removes leakage completely Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Every window reduces the spectral power Results obtained from different devices can only be compared if the same window is used Spectral Jitter Procedures This section shows how to set up and use the Spectral Jitter measurement As an example we measure the spectral jitter behavior of a shielded cable This requires the following steps How to Prepare the Spectral Jitter Measurement on page 351 How to Execute the Spectral Jitter Measurement on page 352 How to Prepare the Spectral Jitter Measurement To use the Spectral Jitter measurement to test a shielded cable 1 Disable the pattern generator outputs by pressing the 0V Disable button in the PG Setup gt Data Output screen Use a shielded cable to connect the pattern generator s Data Out port and the error detector s Data In port Terminate all non connected pattern generator output ports with 50 O Switch to the Pattern menu and press Pattern Select Select an appropriate pattern for example a pure 2 15 1 PRBS for both the pattern generator and the error detector For the pattern generator setup you need to specify the logic levels and the bit rate Select Logic levels LVPECL
204. al Jitter SSC Target BER Confidence Level Relax Time and Margin 3 Graph Jitter Tolerance Graph 4 Measured Points List of points classified with their pass fail status frequency Periodic Jitter Sinusoidal Jitter and Total Jitter without ISI 5 Recommended Settings only in the Compliance Report Document Note Bit Rate Pattern Level from Standard Level at PG recommended and Additional Jitter Explanation of the Results Additionally to the graphical results the measurement provides some information below the graph You can also export the results to a txt file The following information helps you to interpret and document the test Result of the test passed or failed Date and Time when the test was finished Margin that was set for the test Further test Conditions like the selected test pattern and additional jitter components Double click Modify Comment to enter a Comment for every test If not the whole information is displayed in the Condition Comment column this is indicated by Double click Modify Comment to see the complete text in a separate window If you want to use the measurement results with other applications you can export the data to a file via Analysis gt Export Data The contents of the resulting file may look as follows Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Jitter Tolerance Measurement 4 10 2006 12 12 37 PM Standard XA
205. al BERT Customizing the Instrument 10 Installing Software Licenses Installing Software Licenses Procedures The Serial BERT has optional features that require a license The following procedures explain how to obtain and install a license Obtaining a License To obtain a license 1 In the Utilities menu click Licenses The Licenses dialog informs you about the options that require a license Record the HOST ID MAC Addr shown in the window This is the MAC Address of the Serial BERT Contact your Agilent representative and purchase a Software Entitlement Certificate Visit the website Follow the instructions on the website You will need the certificate and the MAC Address You will receive a licence file over the web The file has the suffix lic Installing a License To install a license 1 2 3 4 Copy the licence file to the Serial BERT In the Utilities menu click Licenses Double click the feature you wish to enable Follow the instructions in the window Installing Software Licenses Reference This section explains the options for installing a license Agilent J BERT N4903B High Performance Serial BERT 535 10 Customizing the Instrument NOTE NOTE Licenses Dialog The Licenses dialog informs you about the options that require a license The HOST ID MAC Addr shown in the window is the MAC Address of your Serial BERT You will be asked for this address when you request a licens
206. al BERT Planning the Test 1 Sampling Point Setup Trigger and Aux Setup Pattern Sync e Accumulation Setup Start Accumulation Stop Accumulation BER Location Audio For more information on ED Setup refer to the section Setting up the Error Detector Concepts on page 167 Analysis The Analysis menu allows you to launch the following controls Eye Diagram Output Timing Output Levels Error Location Capture Fast Eye Measurement Spectral Jitter Eye Opening In addition the Analysis menu also allows you to perform following operations Save Measurement Load Measurement Export Data For more information on Analysis refer to the section Advanced Analysis Concepts on page 227 Jitter The Jitter menu allows you to launch the following controls ditter Setup Interference Channel e Jitter Tolerance Characterization Jitter Tolerance Compliance Generate HTML Report For more information on Jitter refer to the section Jitter Tolerance Tests Concepts on page 395 Agilent J BERT N4903B High Performance Serial BERT 23 1 24 Planning the Test Results Utility Help The Results menu provides the following BER Results e Accumulated Results Eye Results For more information on Results refer to the section Evaluating Results Concepts on page 361 The Utility menu allows you to adjust the instrument as per your personal preferences Using this menu you
207. al BERT 127 4 Setting up the Pattern Generator 128 NOTE Crossover CAUTION Data Out lt Delay Clock Out Note that the delay is on Data and Aux Data Crossover is the voltage level where the overlapped rising and falling edges of the logic levels intersect This adjustment varies the widths of the logic highs and lows The following figure shows examples of crossover at 50 80 and 20 50 80 96 20 Understanding how the Serial BERT Uses Logic Families and Terminations The clock and data inputs of your device have load impedances or terminations Clock and data terminations refer to the voltage levels at the end of these loads These voltage levels are related to the logic family of your device Your device should usually only receive signals that are compatible with its termination voltage and logic family When you change a logic family for the Data Aux Data Clock or Trigger Ref Clock outputs the termination voltage normally associated with the logic family is automatically selected This is assumed to be the termination of your device and will be used to determine the actual output voltage levels of the pattern generator If you manually change the termination a warning message is shown indicating that the termination is not typically appropriate for the logic family If you answer No the warning message closes with no changes to the terminations If you answer Yes the data termination voltage
208. al BERT 53 2 Setting up External Instrument s For more information on input timing adjustment of N4876A refer to the section N4876A System Setup Adjustment on page 490 Clock Multiplier Window The Clock Multiplier enables BER measurements using the forwarded clock as sampling clock for Serial BERT The Clock Multiplier window contains the following elements BER 0 000 Remote Track PG Clock Frequency You can use the Track PG Clock Frequency check box to toggle between frequency 54 and multiplier provided by Pattern Generator and the Input frequency and multiplier provided in the Clock Multiplier window If the Track PG Clock Frequency check box is not selected then you can provide the values of the clock multiplier Following are the valid values Multiplier 1 Freq Range 1GHz to to 13 5GHz Multiplier 2 Freq Range 1GHz to to 6 75GHz If the Track PG Clock Frequency check box is selected then the firmware will control both frequency as well as the multiplier value In this case the de emphasis box Clock Multiplier is clocked by a Half Rate Clock DataRate 2 provided by the Pattern Generator So the two parameters of the Pattern Generator have now influence on the clock multiplier s frequency and multiplier First it s the data rate of the instrument can be changed from the Pattern Generator Bit Rate Setup dialog and the Half Rate Clock check box can be changed from the Pattern Generator Clock Tr
209. alculated results are recognized as passed or failed However a new test run is not required when changes are made here See Pass Fail Tab on page 313 for details View tab All settings on this tab only affect the way the data is displayed You do not need to run the measurement again See View Tab on page 314 for details 3 Press OK when you have made all required changes to close the Properties dialog box Fast Eye Mask Reference The Fast Eye Mask measurement returns the results in a numerical form The following sections provide explanations of the measured parameters and the display options that are specific to this measurement Properties that can be specified on the various tabs of the Properties dialog box Agilent J BERT N4903B High Performance Serial BERT 311 6 Advanced Analysis 312 NOTE Parameters Tab Note that if you modify the parameters on this page you have to rerun the measurement to update the results Set the criteria for moving to the next sample point Number of Compared Bits After this number of compared bits the measurement stops for the current sample point and moves to the next one The default is 1 million bits That means you can measure a bit error rate down to 10 one error per million A smaller number reduces the duration of the whole Fast Eye Mask measurement A larger number increases the precision of the measured bit error rates Number of Errors After this nu
210. allows you to connect other devices for further error analysis When the Bit Recovery Mode BRM option is installed the error detector is able to detect bit errors without expecting any particular pattern This feature allows you to investigate the behavior of the DUT when the sampling point is moved Inputs and Outputs Concepts The error detector ED provides input and output ports for running tests and for connecting external equipment The error detector has the following inputs and outputs LX AglentTechmloges SRI rues uo Cei Kam avon ERROR ADD eum m Generator O wi M Dele eei co 6 6 6 Error Dotoctor D e Dain o3 eed Wy w f om sar Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 Understanding Error Detector Input Ports The error detector has the following input ports Clock In This port needs to be connected to a clock signal unless you use the error detector in Clock Data Recovery CDR mode where it derives the clock frequency from the data port See Clock Setup Concepts on page 173 for more details Data In and Data In This port is connected to the data signal and the inverted data signal Gate In This port has two different functions Innormal mode it can be used to enable and disable the error counter Errors ar
211. anches Onebranch has a programmable amplifier to produce the desired output voltage Vpop The other one has an adjustable delay automatically set to one signal clock period and a programmable inverting attenuator amplifier to produce the delayed signal with a lower voltage swing Finally the signals of both branches are added This means the delayed signal voltage is subtracted from the specified peak to peak amplitude Pre cursor de emphasis Itis also possible to convert the input signal to a pre cursor de emphasized signal This can be done by setting the output voltage swing to the desired de emphasis amplitude and specifying a negative amplitude ratio an amplification This inverts Agilent J BERT N4903B High Performance Serial BERT 39 2 Setting up External Instrument s CAUTION De emphasis on pre cursor and 2 post 40 cursors the roles of the two branches The delayed signal has now a larger amplitude than the direct signal A waveform example is illustrated in the following figure 0 1 1 1 0 01 0 1 1 0 Input Signal Output Amplitude posed Vpp mplitude When pre cursor de emphasis is generated this way the complementary Output of the De Emphasis Signal Converter becomes the normal output and vice versa Be very careful if you set the de emphasis ratio to amplification In this case there is no indication of the peak to peak voltage applied to the DUT You need to calculate or measure the ou
212. and Deterministic Jitter To understand the RJ and DJ results it is helpful to first understand how the software generates the results 1 The bathtub curve is measured Agilent J BERT N4903B High Performance Serial BERT 243 6 Advanced Analysis 244 Bathtub log scele Bathtub inear scale n 9 02 04 06 08 a 08 DA 02 02 o 02 delay adust lui delay adjust u All measurement points that have BER between the BER Threshold and Minimum BER for RJ DJ Separation are transformed into Q space The Q factor describes the signal to noise ratio at the decision circuit It is described in Understanding the Q Factor Results on page 282 Linear regression is performed for both the left and right edges The mean and sigma are calculated for both lines RJis calculated as the mean of the two sigmas DJis calculated as the period minus the difference of the two means The estimated TJ is calculated Linear regression is used to extrapolate the bathtub curve to lower BER values Theintersections of the resulting lines with the Residual BER for RJ DJ Separation are located The eye opening is calculated The estimated TJ is the period minus the width of the eye opening The illustration below shows a jitter curve where both RJ and DJ are present It also shows how the TJ peak to peak and RMS are calculated Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 1 Trace 79 Points
213. any disparity error in the pattern In case of alternate patterns the indicator will indicate the disparity error of individual pattern trace Loop Disparity Error LD This indicator turns red if the pattern causes any disparity errors when being looped The available functions to change the view and to modify the pattern are described in Edit Pattern Window on page 76 GoTo Bit Dialog Box This function allows you to set the cursor to an arbitrary position in the pattern Enter the bit position address and click Go The cursor is placed in front of the character with the selected bit You can continue editing the pattern while this dialog box is still open Pattern Properties Dialog Box The Pattern Properties dialog box is used to view or change the attributes of the currently selected pattern Pattern Properties Description New Pat File LenghinBig amp po o Pattern Type Conversion Options C Standard A pattern fill B with 0 Alternate C B pattern fill A with 0 C Demultiplex OK Cancel Help You can use your instrument s software keyboard to enter text in the dialog box For more information see Using the On Screen Keyboard on page 532 This text field allows you to enter a description of the pattern s characteristics or purpose It can be up to 256 characters long This text field allows you to enter the pattern length Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3
214. aph Timing Units Set the criteria for the Timing Units Unit Interval Unit used to measure delay relative to the eye width Seconds Unit used to measure eye width in absolute terms of time Waveform Set the criteria for the waveform Show Waveform This option shows the waveform graphic Smooth Waveform Graphic The smooth waveform gives a finer waveform output It logarithmically interpolates between the measured points to give a finer waveform BER Threshold The BER Threshold value allows the user to measure around the eye boundaries first and then deeper inside the Eye for BERs beyond the specified threshold Contour Setthe criteria for the Contour Show Contour line This option plots the contour Show Legend This shows the list of BER values Grid Show Grid The 353pxl X 289pxl screen is divided into a 10 X 8 grid This parameter adds a grid to the screen Agilent J BERT N4903B High Performance Serial BERT 335 6 Advanced Analysis 336 Markers Mask Selection Mask Alignment Mask Margins Mask Scaling Mask Run Control Set criteria for the markers Show Marker Lines This option enables the markers on the screen Reset marker This brings the marker back to its default position Mask Load Mask File With Load a mask file is loaded Once the file is successfully loaded the Start button gets activated Set Mask Alignment Display The Display mode aligns the mask us
215. asis mode an external connected passive resistor network adds two bit streams The bit streams provided at Data Out and Aux Out are identical except for the bit stream at Aux Out which is shifted by one bit The bitshift value can only be 1 Pre Cursor or 1 Post Cursor For more information see Controlling De Emphasis Mode on page 134 Delay Control Input Delay Control Input Concepts The Serial BERT features a wide bandwidth voltage controlled delay line in its pattern generator data path which can be used to delay the data relative to the clock The delay is controlled by an external signal connected to the Delay Ctrl In connector on the pattern generator front panel The delay control input is intended primarily for jitter tolerance testing where a data signal with a defined amount of jitter is needed Random jitter RJ can be created by applying high bandwidth white Gaussian noise periodic deterministic jitter DJ with sinusoidal SJ or rectangular triangle signals PJ The delay control input is most useful for the generation of high frequency low amplitude jitter Higher amounts of jitter at lower frequencies wander can be generated by supplying an FM modulated clock to the pattern generator s clock input Clk In in external clock mode With a Serial BERT on which the calibrated and integrated jitter injection option J10 NOTE ee i is installed the delay control input is rarely used as explained
216. ata eye The error detector uses the eye diagram to graphically display the location of the sampling point Additionally the appearance of the eye is useful for determining the quality of the data signal The height and width of the eye in the eye diagram are also displayed in the diagram Rise time overshoot and jitter are not displayed in the diagram Note that the diagram does not reflect the different shapes of signals in RZ or R1 pulse format What is an Eye Edge In the eye diagram the eye has 4 edges An eye edge is defined as a data input delay value or 0 1 threshold voltage value in the data eye that meets the following conditions Theinstantaneous bit error ratio BER is greater than or equal to the alignment BER threshold An adjacent point has an instantaneous BER that is less than the alignment BER threshold Agilent J BERT N4903B High Performance Serial BERT NOTE Setting up the Error Detector 5 Eye Edge BER threshold 1E 3 Adjacent Point lt BER threshold such as 1E 4 1 Eye Edge Point BER threshold such as 1E 3 1E 2 1E 1 BERs greater than BER threshold BER threshold Understanding Bit Recovery Mode In bit recovery mode BRM the error detector does not expect any particular pattern Nevertheless it detects bit errors In bit recovery mode the error detector uses two sampling points The second sampling point is not visible Both are set to the expected optimum position when
217. atter are sorted according to descending power contents These results are discussed in Explanation of the Numerical Results on page 259 1 Trace 65535 Points 0 009 0 0081 0 0072 0 0063 0 0054 0 0045 0 0036 0 0027 0 0018 0 0009 0 Noise Threshold 0 0001 Logarithmic 100 00 KHz 10 00 MHz 1 00 GHz 1 00 MHz 100 00 MHz Terminal Electrical SerialB rr How to Optimize the View of the Results After you have run a measurement the resulting graph and the calculated numerical values are displayed To improve the results you can change the measurement parameters 1 Press the Properties button to open the Properties dialog box 2 Use the different tabs in this dialog box to make the required settings Parameters tab Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 These settings are used for data collection Changes require to run the test again See Parameters Tab on page 252 for details Pass Fail tab These settings determine whether the calculated results are recognized as passed or failed However a new test run is not required when doing changes here See Pass Fail Tab on page 254 for details View tab Graph tab and Color tab All settings on these tabs only affect the way the data is displayed You do not need to run the measurement again See View Tab on page 256 Graph Tab on page 258 and How to Change the Colors of the Graph on page 232 for detai
218. atterns with the pattern editor the patterns in the pattern generator and error detector are not automatically updated For this you need to write the pattern to the pattern generator and or the error detector There the pattern is saved in the RAM of the respective module Why the Serial BERT Repeats Memory Based Patterns To generate output signals the Serial BERT can send only complete 512 bit blocks of data To maintain predictability of the output signal it is necessary to send complete patterns This applies to user patterns alternating patterns memory based PRBN patterns mark density patterns and zero substitution patterns Therefore when the Serial BERT loads a pattern into its internal memory it repeats the pattern as necessary until the pattern reaches a block boundary For example if a 1023 bit pattern is loaded this is repeated in the RAM 512 times and thus occupies 1023 blocks Or if a 192 bit pattern is loaded this is repeated 8 times and occupies 3 blocks as indicated in the following figure This applies to all memory based patterns available to the Serial BERT Block 0 bit 192 bit 384 ls ll p e oo NENNEN Block bit 64 bit 256 bit 448 9 i o CRE TE 000000 Block 2 bit 128 bit 320 Pirin 11 EE Gm 00 000010577 During test run the Serial BERT always sends out all blocks of a pattern In the case of the 192 bit pattern mentioned above the pattern is repeated 8 times when sent out in effect mak
219. be specified in terms of two parameters u and c The parameter p is the mean the parameter o is the standard deviation The Gaussian marker shows such a curve Position height and width of this curve can be changed by dragging the handles and the actual parameter values are displayed The marker has three handles that can be dragged two for adjusting its width one for changing its height and position The following figure shows an exemplary DUT Output Timing Jitter measurement result with a Gaussian marker 1 Trace 43 Points Ul 8 000 ns All Errors 4 90e 9 4 20e 9 3 50e 9 2 80e 9 2 10e 9 1 40e 9 7 00e 8 Gaussian Marker Linear 0 68 UI 0 63 UI 0 57 UI 0 47 UI Relative 0 70 Ul 0 65 UI 0 60 Ul 0 55 Ul 0 50 UI The marker readout provides the following information 238 Agilent J BERT N4903B High Performance Serial BERT DUT Output Timing Jitter Measurement Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Table 24 Parameter Symbol Mu u Meaning Mean DUT Output Timing Jitter measurement The position of the marker center on the time scale Output Levels measurement The position of the marker center on the vertical threshold scale Standard deviation The RMS value of the marked area Sigma 0 Kappa K Linear scaling factor A Gaussian marker is used when the jitter graph is displayed It is most useful if deterministic jitter is pres
220. ber of measured points will change inversely proportional to the system frequency doubling the frequency results in half the number of measured points Syntax Requirements All time related entries understand both time and unit interval notation regardless of the Ul time selection made in the View tab of the Properties dialog box You can enter 17s bns or 0 01UI at any time On run the Serial BERT will automatically convert all entries to time values using the current unit interval So the Ul is handy if you want to set values independently from the system frequency All voltage related entries understand 23mV 0 01V and so on All power related values can be entered as 50mW 0 04W 6dBm and so on All dimensionless quantities understand decimal notations e g 10000000 0 0003 and scientific engineering notation e g 1e9 1 7e 3 Marker Definitions Markers are available for all measurements except of the Fast Eye Mask They make the analysis and tracing of the results more comfortable Markers can be switched on or off at any time when results are available in the graphical view For the DUT Output Timing Jitter measurement the software provides two types of markers The following illustration shows the definitions for the markers and the values that can be derived from the marker position Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 1 00e 0 1 Trace 39 Points Ul 401 878 ps All Errors 1
221. bits and errors counted The BER is also shown in the user interface Timing of the Signal at Gate In The timing of the signal at Gate In is critical to the successful use of burst mode If the signal is not timed optimally the resulting BER will be either too high or the sync ratio good bursts total bursts will be too low The signal at Gate In has to start early enough for the error detector to get the clock the CDR settling time and synchronize to the pattern And it has to stop before the burst ends but should stay low as long as possible without hitting the maximal length limitation to extend the time where bits can be counted The following figure illustrates how the duration of the Gate In signal can influence the bit count time the time where bits are actually counted Valid after Gate CDR Settling Time Synchronization Time Data Input 4 gt Begin Margin End Margin lt gt Bit Count Time Gate Input k hi 4 uus gt Gate Active ae as Passive Gate Active is the interval in which the Gate In signal is active Agilent J BERT N4903B High Performance Serial BERT 213 5 Setting up the Error Detector 214 Interval between Bursts The Bit Count Time is the part of the burst from which the bits can actually be counted The remainder of the burst is covered by the Begin Margin and End Margin The CDR Settling Time is the time that the error detector requires in CDR mode to
222. bits from pattern A with the bits from pattern B The pattern description contains the first error the error count date and time The name of the pattern file is ELOC RESULT CURRENT ptrn for the current capture and ELOC RESULT PREVIOUS ptrn for the previous capture These patterns are saved under C lt instrument model Pattern on the machine with the firmware server Results Window The results are displayed in the window below the pattern You can compare the results of the Current Run with the results of the Previous Run 1st Bit Error Location Address of the first captured errored bit BitError Count Number of all errored bits captured during the measurement Fast Eye Mask Fast Eye Mask Concepts The Fast Eye Mask measurement is first of all meant for production and screening tests It allows to determine very quickly whether the eye opening seen at the output signal of a device is within specifications that is within certain timing and voltage limits Measuring the eye openings with an oscilloscope used to be a time consuming procedure With the Fast Eye Mask measurement of the Serial BERT pass fail information can be obtained within seconds Agilent J BERT N4903B High Performance Serial BERT 307 6 Advanced Analysis 308 NOTE This is achieved by measuring the bit error rate at a limited number of test points Up to 32 measurement points can be specified each defined by a sampling time relative to the ac
223. blishes an almost direct connection between P1 and P2 For details about the minimum trace length refer to the technical specifications Trace Number To specify the degree of intersymbol interference select a suitable trace length The user interface provides numbers from 1 to 9 The greater the number the longer is the trace Sinusoidal Interference To enable the addition of sinusoidal interference press the corresponding button Set the parameters to desired values The peak to peak amplitude of sinusoidal interference must not exceed a certain level see the technical specifications Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Frequency The frequency range of sinusoidal interference is limited see the technical specifications Mode One of four modes can be chosen for sinusoidal interference Common Requires that a differential data signal is generated and connected to P1 P1 or P2 P2 of the Interference Channel In common mode both components of the differential data signal are amplitude modulated with the same sinewave This adds common mode noise Differential Requires that a differential data signal is generated and connected to P1 P1 or P2 P2 of the Interference Channel In differential mode both components of the differential data signal are amplitude modulated one with the normal sinewave the other with the inverted sinewave This adds differential noise Single
224. bols and compares only the Block Length in Block Mode or Pattern Length in All Bits mode Symbol Error Ratio SER The ratio of the error symbol to the symbol count is displayed here to 4 significant digits The following is the expression used to calculate the Symbol Error Ratio Symbol Error Count Symbol Error Ratio Block Length x Pattern Count Received Frame Count Frame counter The number of received frames considered for the accumulation period is displayed here This may not equal the total number of frames sent from the Pattern Generator Frame Error Count Frame error counter The total number of error frames received in a time interval is displayed here Frame Error Ratio FER The ratio of the number of frame errors to the number of frames received in the current or last completed accumulation period specified by the gate period isdisplayed here The following is the expression used to calculate the Frame Error Ratio Frame Error Counter Frame Error Ratio Frame Counter Illegal Disparity Change Count Wrong disparity counter The total number of illegal disparity change count wrong disparity count received in a time interval is displayed here Filler Symbol Count Filler symbol counter The total number of filler symbol count received in a time interval is displayed here The following is the expression used to calculat
225. box On the Parameters tab set the Sample Threshold parameters to suitable values that fit to the expected signal levels for example 1 85V to 0 85V Set the Resolution to 10 mV Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Properties xi Parameters Pass Fail View Graph Color Criteria for moving to next measurement point Number of Compared Bits v Number of Errors Sample Threshold Resolution E D mv Low Level 1 bY High Level 250 mM v Edge Resolution Optimization Cancel Apply Help In this example we expect the signal voltages to be between 1 75V and 0 95V The Sample Threshold values proposed above cover this range well The Resolution is the distance between the measurement points when the threshold moves from the low to the high level A resolution of 10mV results in 100 measured points per Volt Note that we have disabled the Edge Resolution Optimization 4 Press OK to close the Properties dialog box 5 Press the Start button to execute the measurement The measurement is run and the result window shows the bit error rates measured at 100 threshold levels from 1 85V up to 850mV Agilent J BERT N4903B High Performance Serial BERT 267 6 Advanced Analysis 268 800 0 mv 1 Trace 202 Points A0V 42V AAV ABV BER Threshold 4 1 000e 6 18V Logarithmic 1 000e 5 1 000e 3 1 000e 1 1 000e 6 1 000e 4 1 0
226. button The two captured errors are illuminated red in the captured pattern grid In the result list the 1st Bit Error Location Capture indicates an error in bit 80 Use the error navigation buttons e g Next Error Prev Error to display further bit errors in the captured pattern Error Location Capture Reference The Error Location Capture measurement compares the expected pattern and the pattern that was actually received and shows the received pattern with the errored bits marked Elements of the User Interface The following buttons help you to analyze the results e Hex Bin Agilent J BERT N4903B High Performance Serial BERT 305 6 Advanced Analysis 306 Graphic View Status Bar Choose between a binary or hexadecimal display of the pattern The current setting is shown in the status bar below the pattern window Note that in hexadecimal view the captured error can only be located as being one of a group of four bits To display which of the four bits is the errored bit you have to switch to the binary display Exp Cap Expected Captured Toggle the data view between the display of the expected data i e the data you defined on the error detector and the captured data i e the data that was captured at the error detector The current state is displayed in the title bar of the pattern grid Run Select Toggle the data view between the current run and the previous run After a measurement run is successfu
227. cable R 2values The R 2 values are calculated for both slopes of the bathtub curve They are a measure of how well the transformed points between BER Threshold and Min BER for RJ DJ Separation fit to the linear regression They have to be greater than 0 75 for the RJ DJ and estimated TJ values to be applicable Unavailable Values Under certain circumstances some numerical results are not available This is indicated by lt invalid gt or not applicable in the numerical results table below the measurement graphic e invalid indicates that the value could not be calculated This is the case for example for phase margin when the BER threshold is set to high values and does not intersect with the bathtub curve e notapplicable indicates that the value could be calculated however is not shown because quality criteria are not met This is the case for example for RJ DJ results when the r 2 value of one or both edges is 0 75 Even though Agilent J BERT N4903B High Performance Serial BERT 261 6 Advanced Analysis Explanation of the Result Display NOTE Explanation of the Numerical Results 262 RJ DJ values can be calculated in this case they are not shown because confidence in the results is too low Explanation of the Fast Total Jitter Measurement Results The Fast Total Jitter measurement provides both graphical and numerical results The example below shows a copied result and the display of measure
228. cal axes according to your needs Agilent J BERT N4903B High Performance Serial BERT 345 6 Advanced Analysis 346 Youcan switch between linear and logarithmic scales Graphical markers and the zoom function assist you when you are analyzing the graph visually Thenumerical results include bit error rate total power and noise power They provide also frequency and power information about the dominant peaks in the spectrum Absolute and relative power values are available Relative values can be normalized to the total jitter power or the power of a selected tone frequency bin Pass fail limits can be set for the bit error rate total power and noise power and the allowed jitter power in user defined frequency regions Periodic Jitter and Bit Error Information When the incoming signal is sampled at the transition point periodic jitter manifests itself in the bit error record An example may be helpful to understand the phenomenon Let us assume we expect and correctly receive a simple 0 1 0 1 0 pulse signal If we would sample this signal one clock period earlier 1 Ul we would see a bit error rate BER of 1 0 This is the maximum BER for this pattern Let us also assume this signal is overlaid by a periodic jitter source with sinusoidal characteristics Now we sample the incoming signal at the transition point 0 5 UI from the optimum sampling point The result is illustrated in the figure be
229. can be specified Number of Received Bits or Errors When numbers of received bits or errors are enabled the measurement proceeds with the next frequency as soon as one of the two numbers is reached But we are often dealing with very small tolerable bit error ratios 10 2 and below Measuring such a bit error ratio precisely and just for one single point takes time Therefore the measurement provides the option to set a confidence level Confidence Level The Jitter Tolerance Compliance measurement offers the option to set a level of confidence This approach is based on statistics Statistics tell us for example that if we receive 3 x 10 bits without any error the probability that the BER is below 107 is higher than 95 96 If an error occurred more bits must be compared to achieve the same confidence level For details see Explanation of the Fast Total Jitter Measurement on page 246 The measurement proceeds with the next frequency and jitter amplitude as soon as the desired level of confidence is reached Setting a confidence level ensures that neither too few nor too many bits are captured Relax Time Before the measurement proceeds to the next frequency bin a measurement settling or relax time can be applied This allows the DUT to recover in case it has lost synchronization due to an excessive high bit error rate The next BER measurement starts after this time has elapsed About Jitter Tolerance Standards Every
230. cannot be used while Avg 0 1 Threshold is selected The 0 1 Threshold field allows the manual entry of the 0 1 threshold and displays the current value You can adjust the 0 1 threshold in the following ways lick inside the text field and use the numeric keypad or front panel knob to enter a value Click anywhere in the sampling point display The sampling point will be moved to that location Turn the Decision Threshold knob on the front panel A small window appears displaying the threshold value This knob can be used at any time while in any window Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 Data Delay The data input delay defines the point in time in reference to the clock signal at which the incoming data signal is measured Specifically it is the time delay from the active clock edge to the time at which the data is actually sampled This field allows the manual entry of the data input delay and displays the current value in picoseconds This delay can be set as high as 1 bit period or 10 ns 10 000 ps whichever is less You can adjust the data input delay in the following ways lick inside the text field and use the numeric keypad or front panel knob to enter a value Click anywhere in the sampling point display The sampling point will be moved to that location Tur the Data In Delay knob on the front panel A small window will appear displaying the
231. ce Tests NOTE Amplitude Voltage NOTE Slope 430 153 75 mUI 362 mV 400 ps V Even if the External jitter source is enabled you can still add or change internal jitter components The range of the external jitter source is automatically updated when you add change or turn off one of the internal sources This is the maximum jitter Amplitude that can be generated This is the maximum voltage that can be accepted It would generate the maximum jitter amplitude Before connecting an external jitter signal always ensure that this signal does not exceed the maximum voltage The Delay Control Input has an overvoltage protection For details see Delay Control Input Concepts on page 155 The Slope indicates the present relation between the voltage applied and the resulting delay Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Interference Channel Setup Interference Channel Setup Concepts The optional Interference Channel module can be plugged into the instrument For an introduction please refer to Interference Channel on page 402 This module is to be connected between the pattern generator s Data Out port and the DUT For installation instructions refer to Installing Hardware Options Procedures on page 536 The Interference Channel simulates intersymbol interference ISI It has backplane traces that conform to the Nelco 4000 6 FR4 specification The trace le
232. chronized with the pattern sent from the DUT output port e Ideally the sampling point should be aligned to the optimum sampling point for example with the Auto Align function Forlow bit rates below appr 620 Mbit s only the Error Location Capture measurement works Please refer to Bit Rate Range on page 176 for a detailed explanation of the limitations at low bit rates If the Serial BERT s pattern generator is used to send a data pattern to the DUT then also the following two conditions must be met e An appropriate pattern is selected The pattern generator is correctly set up For examples refer to the descriptions of the different measurements Optimizing Parameters If the presented measurement results do not answer your questions adequately you can easily modify the setup parameters in the Properties dialog box When Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 changing the measurement settings after the measurement has been run please note Parameters that affect the data capture Changes on the Parameters page take only effect if you run the measurement again Parameters that change the display of the measured data Changes on the Pass Fail View Graph and Color tabs only affect the display of the results There is no need to repeat the measurement Saving Measurements for Later Analysis You can save a completed measurement for later analysis This includes the s
233. cially if you compare a large number of bits at a low data rate Pass Fail Tab The Pass Fail tab of the Properties dialog box allows you to specify the criteria to decide whether the DUT passes or fails the test You can change pass fail criteria without rerunning a test The software only uses the criteria to rate the results of a measurement The pass fail criteria do not control measurement execution The measurement run will be completed even if the measurement fails for one or more of the criteria You can set pass fail limits individually for level noise and Q factor parameters To set the Level Pass Fail criteria see also Level Results on page 277 To set the Signal Noise Pass Fail criteria see also Noise Results on page 278 To set the Q Factor Pass Fail criteria see also Q factor Results on page 279 Your setup may finally look as shown below Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis Properties m MEE Li Signal to Noise Ratio AMS mj signal ta Noise Ratio Peak Peak J Peak Peak Noise mi G Factor Residual BER n Uptimum fihreshold In the following figure you can see how errors are flagged 1 Trace 86 Points 600 0 mv 400 0 mv 200 0 mv D 0 v 200 0 mv 400 0 mv RNV h BER Threshold 1 000e 6 Logarithmic 1 000e 5 1 000e 3 Agilent J BERT N4903B High Performance Serial BERT 273 6 Advanced Analysis 27A
234. cification so that the CDR behaves according to the standard If a standard from the preset list is selected this field is preset Loop Bandwidth This is the range of the CDR loop bandwidth In this field the user should enter the loop bandwidth value the range is within 100 KHz to 12MHz If a standard from the preset list is selected this field is preset When the BERT error detector is used to characterize a data stream instead of a receiver the loop parameters should be set according to the used standard If the CDR is used to recover the bit stream from a receiver to be characterized due to a lack of a clock output choose the loop bandwidth significantly higher than the receiver s bandwidth To characterize a DUT s CDR it is the best practice to use its recovered clock output instead of the BERT s built in CDR Choose a low loop bandwidth to measure the jitter on an incoming data stream as the CDR will track the incoming jitter up to the loop bandwidth and thus make it invisible to the error detector Fine Adjust Fine Adjust allows you to tweak the response of the CDR according to transition density changes This control can be used to reduce jitter in pattern with high changes in transition density Peaking The drop down list provides a maximum of three peaking values If a standard from the preset list is selected this field is preset A low value reduces the peaking near the loop bandwidth while a higher value improves the
235. ck pulse Pattern You can click Browse to launch Select User Pattern dialog This dialog opens a browser and search for the file that contains the pattern You can open Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 a browser and search for the file that contains the pattern You can search for files with the extentions ptrn txt or dat If the pattern file contains two alternate patterns you can also specify whether the A or B pattern shall be generated see also Pattern File Specifications on page 66 If the pattern file contains just one pattern your choice is ignored You cannot change the block length It shows the length of the pattern as stored in the file If the pattern length is not 512 bits or a multiple of 512 bits and this single block is looped the behavior depends on the loop end condition If the loop end condition is INFinite or determined by the Aux In port then the pattern is automatically repeated until 512 bits or a multiple thereof is reached In this case the duration of the block will be longer than indicated For example a pattern containing 17 bits will be repeated 512 times and the actual block length will be 8704 bits This has an impact on the trigger at block start You can also use the Select User Pattern dialog to create new pattern Block Trigger If this checkbox is enabled a trigger spike will be output at the pattern generator s Trigger Out connector whene
236. click Set As Default 4 Click Close The default printer selection is applied and the dialog box closes The printer capabilities will vary depending on the printer you have installed and configured Printing a Hardcopy Before you can print you must first have the appropriate printer driver installed The printer properties and capabilities will vary depending on the printer you have installed and configured for the instrument To print a hardcopy of the instrument s display to the default printer Click Print from the File menu on the main menu bar Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 Configuring Printer Properties You can change the properties of a selected printer 1 Onthe File menu click Print Setup 2 Click Configure Printer This opens the Printers dialog box 3 Inthe File menu in the Printers dialog box select from the following options Server Properties Click this option to change such items as printer margins spooler warnings and alarms and port configurations Properties Click this option to change such items as pre determined access times security items and LAN protocols Document Defaults Click this option to change such items as pages per sheet print quality paper select and portrait vs landscape layouts 4 Select Close in the File menu of the Printers dialog box The new settings will are applied and the dialog box closes TIT
237. click the Agilent IO Control icon 10 3 From the Agilent IO Control menu open the VISA Assistant VA VISA Assistant cog File Edit Yiew Configure Help Instrument Driver Formatted 1 0 Memory 1 0 Attributes GPIBO 11 INSTR USB0 USB0 2391 21000 0 UsbDevice1 No Instrument Driver Configured Address String ASRLI INSTR rsrcName Configure _ For Help press F1 NUM 11 26 44 4 If no VISA alias name has been assigned so far the VISA Assistant should indicate an instrument named UsbDevice1 If the N4916A has been installed 494 Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 correctly the VISA Assistant should indicate an instrument named D4916 Click UsbDevice1 or D4916 5 Click the Formatted 1 0 tab 6 In the Instr Lang box Instrument Language enable SCPI Instr Lang C Custom EEE 488 2 SCPI 7 Click the IDN button This generates the identification query in SCPI format The instrument should return Agilent Technologies N4916A and the software revision V VISA Assistant imis File Edit view Configure Help ASRLI z ASRLL SINSTR Instrument Driver Formatted 1 0 Memory 1 0 Attributes GPIBO Clear History Show C Code GPIBO 11 INSTR Seal m USBO gt IDN Device Clear 5 USB0 2391 21000 0 Agilent Technologies N49164 0 02 Saar UsbDevice1 Read STB RST TS
238. composition A standard may prescribe that additional jitter components with specific amplitudes have to be added Using Standards The standards provided by Agilent can be chosen from a list of predefined standards Every standard is accompanied by a description that states its specific requirements like applicable bit rates patterns to be used and so on NOTE Before running the test set up the instrument to ensure that these requirements are met For details on the recommendations and their dependencies see Description in the Standard Specification on page 462 If a standard specifies a tolerance curve that is bit rate dependent the curve is automatically adapted to the present bit rate If you change the bit rate the curve is automatically recalculated The standards provided by Agilent cannot be edited locally Agilent J BERT N4903B High Performance Serial BERT 451 8 Jitter Tolerance Tests User Defined Standard Specifications 452 NOTE Using User Defined Standards User Defined Standards You can create your own standard in a text file A user defined standard consists of a paired list of numbers Each pair specifies a jitter frequency Hz and the associated jitter amplitude UI for example 2e3 110 lt cr gt lt lf gt 3 24e4 50 lt cr gt lt If gt 71 234e4 10 lt cr gt lt lf gt 10 654e6 0 2 lt cr gt lt lf gt 18 333e6 0 2 lt cr gt lt If gt cr If stands for carriage return
239. continue Adjustment A message will pop up which warns you to save the instrument settings before continuing the auto adjustment process Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 N4876A The instrument settings will be lost once you start the adjustment process It is recommended to save the instrument settings Make sure to remove all cables or termination resistors at the data outputs of the N48764 before starting adjustment Do you like to start adjustment 4 Click Yes to continue if you have already saved the settings Once the auto adjustment starts you will see the adjustment progress in the following screen Auto Adjustment N4876A Input timing adjustment Data rate 11 00 Gb s Step Progress 11 x Total Progress W NOTE The adjustment process may take several minutes for completion The following message will pop up after the successful completion of adjustment process N4876A Adjustment is over No Errors were detected and the previously stored setting can be recalled Agilent J BERT N4903B High Performance Serial BERT 491 9 Solving Problems M8061A System Setup Adjustment 492 M8061A System Setup Adjustment NOTE The input timing of M8061A has to be adjusted to ensure error free operation for all data rates This adjustment is required once for a dedicated setup of N4903B M8061A and the cable kit being used and has to be repeated whenever on
240. contour information 3 Change the sampling point in the Sampling Point Setup window See Sampling Point Display on page 202 for details 392 Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 Eye Measurements Reference Eye results are displayed for the Current Align and Previous Align process They are also graphically displayed on a representation of an eye diagram Eye Width and Height Eye width is the inside horizontal opening of the data eye in an eye diagram This measurement is made by the error detector after an Auto Align or Data Center The eye width at the current 0 1 decision threshold is displayed on a simulated eye diagram The left and right eye edges are determined by the user defined BER threshold Different BER thresholds will produce different eye width results Eye height is the inside vertical opening of a data eye in an eye diagram This measurement is made by the error detector after an auto align or 0 1 threshold center The eye height at the current data input delay point is displayed on a simulated eye diagram The upper and lower eye edges are determined by the user defined BER threshold Different BER thresholds will produce different eye height results Agilent J BERT N4903B High Performance Serial BERT 393 7 Evaluating Results 394 NOTE Most oscilloscopes measure eyes differently using statistical data Therefore the eye results of a Serial BERT and an oscilloscope
241. ction 52 as BER 0 000 Identify Connection __ Enable Function _ S WRev_ VISA Resource Name __ M8061A M8061A EVO 14 21 USB 0x15BC 0x1297 0 Remote The Config window contains the following elements The Refresh button allows you to reload the external instruments list in case if you are not able to see them in the instrument list However if the Config window still do not display any externally connected instruments restart the J BERT N4903B software and firmware The instrument list contains the following columns The Identify button containing instrument s name allows you to enable and disable the identification LED on the external instrument s This is simply required to help you to connect the right things at the right connector For few instruments such as N4916A and M8061A a text showing the instrument s name will be shown instead of LED button The connection type defines the connection For example if N4916B is connected to the Data Out it will show DATA_OUT The Enable Function allows you to select and accordingly enable disable the selected functions provided by the externally connected instrument The N4916A Agilent J BERT N4903B High Performance Serial BERT Setting up External Instrument s 2 provides the De Emphasis function while the N4916B provides De Emphasis and optional Clock Multiplier functions Software Revision The Update button allows you to update the software revision of ex
242. ctual Button Error Count Accum Error Count Accumulated Actual Button Error Count Accum Error Count Click this button to toggle between SER or AccumSER results This area displays the actual symbol frame error count or cumulative error count FER Results Window This window displays the actual FER or cumulative AccumFER FER results MEA FER 0 000 _ _ I ZLZZT TTL e Frame Error Counts 0 TheFERis the current FER calculated upon a period of 200ms TheAccumFER shows either the accumulated FER of the current accumulation or if no accumulation is running the results of the most recent accumulation This enables you to monitor real time SER behavior as you do things such as to manually adjust the sampling point to add errors or to make adjustments to your device Click this button to toggle between FER or AccumFER results This area displays the actual frame error count or cumulative frame error count cBER Results Window This window displays the actual CBER or cumulative AccumcBER cBER results Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 ARAA cBER 0 000 Calc Error Counts You can click Accumulated Actual button to toggle between cBER or AccumcBER results The Error Count Accum Error Count area displays the actual error count or cumulative error count FSR Results Window This window displays the actual FSR or cumulative Acc
243. culate the parameters for 0 errors BER Threshold To calculate the parameters for the given BER threshold The BER threshold influences all parameters of the Eye Opening measurement You can also drag Agilent J BERT N4903B High Performance Serial BERT 299 6 Advanced Analysis 300 Table Number Format Timing Units Markers Zoom Show Measured Points Show BER Threshold NOTE and drop the horizontal BER threshold in the graphical display to change this value You can select the number of Decimal Places to be displayed in the table Graph Tab On the Graph tab you can use the several options to optimize the graphical display according to your needs Choose between Unit Interval and Seconds to select the timebase for the display s x axis To analyze the graphs at a particular point you can use the markers Additionally you can display all related values for the markers in the marker readout Pressing the Reset Markers button will set the markers back to the default positions 1500mv 1 Trace 13689 Points Ul 1 538 ns All Errors 1 000e 0 met egt Toc 120 0 mv Je Lu Markers Time Voltage ony lt Upper Right 0 772 UI 90 0 mv 714e T 0006 1 TME pasfedssdae saanak Lower Left 0 113 UI 90 0 mv TAES soomy lt Delta 08850 1800mV J 1 000e 2 Damy 1 000e 3 00v p 1 000e 4 60 0 mv 1 000e 5 90 0 mv 1 000e 6 1 000e 7 0 130 UI 0 130 UI 0 390 UI 0 650 UI 0 910 UI
244. d The maximum jitter amplitude the Serial BERT can generate is also frequency dependent as illustrated below Jitter amplitude in UI log scale Maximum jitter amplitude over frequency LEE SE SE SE EE EE EE 9 9 9 9 s e 5 Jitter frequency log Scale Sinusoidal jitter SJ and periodic jitter PJ with sinusoidal characteristic are used for the test Whether SJ or PJ is used depends on the jitter frequency SJ supports wide amplitude variations up to some MHz PJ supports narrow amplitude variations over a wide frequency range The measurement switches automatically between the two sources Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 When the measurement has finished the jitter composition shown in the Jitter Setup window is restored Target BER The target bit error ratio must be specified The condition for moving from one jitter amplitude step to the next can be set to numbers of received bits and errors a confidence level Additionally a relax time can be specified Number of received bits or errors When numbers of received bits or errors are enabled the measurement proceeds with the next amplitude as soon as one of the two numbers is reached But we are often dealing with very small tolerable bit error ratios 10 1 and below Measuring such a bit error ratio precisely and just for one single point takes time Confidence level The Jitter Tolerance
245. d In this case you can either set up the Trigger Ref Clock Out port to generate a subrate of the present clock or to signal the beginning of a sequence block see also Sequence Block Parameters on page 102 for details Trigger Ref Clock Output Reference NOTE You can always generate a divided clock signal at the Trigger Ref Clock Out port But if you wish to use the Trigger Ref Clock Out port for controlling external equipment please note that you can also generate a divided clock signal at the pattern generator s Aux Data Out port The alternate trigger signals refer to patterns and are not generated in sequence mode To support the generation of a trigger spike at the beginning of a sequence block the Trigger Ref Clock Output can be put into Sequence mode Clock Divided by n Select this option to send a trigger signal from the Trigger Ref Clock Out port at every nth clock pulse Note that the trigger signal itself consists of n 2 bits high followed by n 2 bits low For example Clock divided by 8 works as shown below c IA If the Divider Factor n is uneven For example 3 the clock s duty cycle will not be 50 but the signal will stay high for n 1 2 and low for n 1 2 This results in a Duty Cycle Distortion DCD of 0 5 UI Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 Alternate Pattern Trigger Level NOTE This option applies for alternating patterns only Select this op
246. d Passed Passed Passed Passed Passed Passed Passed Passed Passed Passed Passed Passed Passed Passed Passed Passed Passed Passed Passed Up Down Set Close Selecting a point and clicking on Set updates the Jitter Setup page with the jitter setting of that point Generate HTML Report The measurement results can be saved as HTML file To generate the HTML report 1 Click Jitter menu from list of menu option provided on the drop down menu The following submenu will appear Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 ry Jitter Setup EJ Interference Channel se Tolerance Characterization E Tolerance Compliance Generate HTML Report 2 Select Generate HTML report 3 Select the folder where you want to save the report name the file and save it Jitter Tolerance Characterization Reference Jitter tolerance characterization determines the actual jitter levels where the device under test can no longer maintain a desired BER Frequency Specification You can choose Auto or Manual mode Auto mode n Auto mode the jitter frequency is automatically swept from Start Frequency to Stop Frequency Start Frequency This is the minimum and first jitter frequency to be generated It determines the left hand border of the graphical result display Stop Frequency This is the maximum and last jitter frequency to be generated It determines the right han
247. d belowin Setting the Frequency Range on page 456 Setting the Frequency Range You can freely specify the jitter frequency range and the number of points to be measured By default the test points are suitably distributed within the frequency range according to the loaded standard But you can change the distribution according to your needs 1 Switch to the Frequency tab Jitter Tolerance Properties Standard Frequency BER Auto Start Frequency fi 000 MHz Stop Frequency 80 000 MHz Nr of Steps 20 C Manual Sequence Advance OK Cancel Apply Help The measurement s frequency range and the list of test frequencies are displayed If the Auto Frequencies checkbox on the Standard tab is selected start and stop frequency are by default set according to the selected standard When you change the selection state of the Auto Frequencies checkbox press Apply on the Standard tab to accept the change and to update the frequency list Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 2 If desired change the default values For example if only a certain frequency range has to be tested Enter the Start Frequency and the Stop Frequency 3 Setthe Number of Steps between start and stop TIP When Auto Frequencies is activated the Number of Steps can also be set on the Standard tab The resulting frequency list depends on the setting of the Auto Frequencies checkbox on the
248. d border of the graphical result display The measurement always includes the start and stop frequencies The frequency scale is logarithmic Number of Steps The Number of Steps between start and stop defines the test frequencies To determine the individual test frequencies an algorithm divides the horizontal logarithmic scale into a set of logarithmically equidistant frequencies Agilent J BERT N4903B High Performance Serial BERT 443 8 Jitter Tolerance Tests 444 Manual mode Show Compliance Curve Target Bit Error Ratio Confidence Level Absolute Bits Errors NOTE Relax Time In Manual mode you can edit the automatically calculated frequencies You can also enter your own list of frequencies with a minimum of two frequencies specified For example you can jump between low and high jitter frequencies Enter the frequencies in Hz separated by commas for example 2e4 1 75e6 6 334e5 Enabling this compliance curve helps in ensuring that the device under test complies with a certain standard This feature is available only if the compliance jitter measurement option is installed The standard is taken from the compliance measurement BER Specification The Target Bit Error Ratio defines the required BER The jitter tolerance curve is constructed from the points where the measured BER crosses the Target Bit Error Ratio The Verification Method defines under which circumstances the measurement proceeds
249. d pattern The instrument saves the position of the errored bit and writes a bit sequence neighbouring the errored bit to a file This feature can be used to find rare or random errors A DUT could have problems handling long series of zeroes Error Location Capture can be used to locate the bit errors in such cases Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 NOTE The Error Location Capture measurement is disabled in 8B 10B Comparison Bit Comparison without PCle3 SKPOS and Bit Comparison without USB3 1 SKPOS error ratio modes NOTE The measurement run is aborted by various actions like selecting a new pattern or starting synchronization or alignment Example Results The following figure shows the results of a typical Error Location Capture measurement 1st Bit Error Location 80 Bit Error Count 128 Important Requirements In order to perform Error Location Capture measurements the following requirements have to be met in addition to the ones listed in General Requirements on page 228 Only memory based patterns with a unique 48 bit pattern detect word are allowed No alignment features can run during error location capture Auto Align 0 1 Threshold Center Data Center Agilent J BERT N4903B High Performance Serial BERT 303 6 Advanced Analysis No other advanced measurement Output Timing Output Levels etc can be running Error Location Capture can only run
250. d points was enabled The results of the recent measurement are disabled because they refer to a standard bathtub measurement that provides many more results By actuating the Show indicator you can inspect both alternatively 1 Trace 129 Points Ul 141 642 ps All Errors gt TJ BER close to TJ BER x TJ BER 044 UI 0 42 UI Absolute 0 85 Ul 0 28 UI Optimal Sample Delay Terminal SerialBERT xdisabled 0 135 UI disabled 0 0141 ul disabled disabled 1E 009 0 865 Ul disabled 0 251 ul B The result graph shows you the points in time that have been investigated and whether the actual BER at these points was higher or lower than the BER threshold specified for the measurement The grey area in the middle indicates the uncertainty band see Explanation of the Fast Total Jitter Measurement on page 246 The test will fail if the BER floor is not well below the BER threshold specified for the measurement The Fast Total Jitter measurement result parameters are summarized in the following list Phase Margin The period of time where the bit error rate is lower than the Total Jitter BER Threshold Total Jitter Peak Peak Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Peak to peak value of the total jitter Calculated as the pulse period unit interval minus the Phase Margin at the Total Jitter BER Threshold
251. d tests you can view detailed error messages in a separate window If a self test is started remotely the GUI is not able to inform the user about that NOTE n DUNS most likely it will not even show that it is in remote state but will be blocked until the self test is finished Self Test Procedures Running Self Test To run a self test 1 Select Self Test from the Utility menu Agilent J BERT N4903B High Performance Serial BERT 563 10 Customizing the Instrument 564 This opens the Self Test Options dialog box Select the desired tests and click the Run Test button After the self test has finished the results are displayed next to each selected test If a test failed you can view detailed error messages by clicking the respective Message button Self Test Reference This section contains information about the Self Test Options dialog box used for setting up and running a self test and for viewing the results Self Test Options The Self Test Options dialog box offers to select the following tests Pattern Generator Bios Self Test This test is checking all the modules chips and registers of the pattern generator more thoroughly than the Power Up test It may take about 5 minutes to complete Pattern Generator Power Up Messages This test is automatically run after switching on the instrument It mainly checks if all important modules chips and registers of the pattern generator are available
252. data However to perform tests that can be repeated and compared you can collect the measurement data in several ways and save it to a log file The available options are to test for a specified time until a certain number of errors occurred until a specified number of bits have been sent until you stop it manually Error Accumulation Procedures To set up the Serial BERT for error accumulation 1 6 Click the Accumulation Setup menu item from the ED Setup submenu 9 Accumulation Setup Select the desired Activation Mode for the type of test you want to run Manual Single or Repeat test If you selected a Single or Repeat test you need to specify the accumulation Period Time Number of Errors or Number of Bits Use the Measurement Logs area to specify whether you want save the results in a log file The log files will be saved in the Log Directory displayed at the bottom of the dialog box Use the Browse button to select a different directory Click OK to apply your selections and close the Accumulation Setup dialog box For more information on how to run error accumulation tests see How to Run Accumulated Measurements on page 377 220 Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 Error Accumulation Reference The Accumulation Setup window contains the following elements Activation Mode Manual Activate this option to configure the error detector t
253. data points Re For example Two data points always fit perfectly well but the resulting Q factor calculations are not reliable On the other hand 50 data points may reveal a poor R value This tells you that the linearization is prone to errors If the R value falls below 0 75 the Q factor calculations are not applicable QLow Level The Q Low Level is the mean calculated from the linear regression curve for the low level data zA H B Agilent J BERT N4903B High Performance Serial BERT 281 6 Advanced Analysis NOTE 282 Q Low Level Std Dev The Q Low Level Standard Deviation is the c Sigma calculated from the linear regression curve for the low level data Q Low Level Nr Points This is the number of data points used for the calculation of the Q Low Level value It depends on the setting of the BER Threshold and also on the setting of the Min BER for Q parameter The minimum for calculating Q factor values is two points It is recommended to include more than 5 points Q Low Level R 2 See the definition of Q High Level R 2 Unavailable Values Under certain circumstances some numerical results are not available This is indicated by invalid or not applicable gt in the numerical results table below the measurement graphic lt invalid gt indicates that the value could not be calculated This is the case for example for phase margin when the BER threshold is set to high values and does
254. data from one trace to another Swap Traces Choose this option to swap data between the traces The Copy and Swap features are enabled only when there are two traces Pattern A and Pattern B when alternate pattern is used for single channel or Data and Aux Data pattern for second channel For normal pattern or when PG is in multiplexer mode Copy and Trace will be disabled as there will be only one pattern trace available Reverse Traces Choose this option to reverses the data in the selected range 4 Rotate This tab provides the following options Fill Traces Rotate Rotate Left by o Rotate Right by o Align to Sequence 001 1111010 Rotate Left Treats the pattern data in the specified range as a circular buffer and rotates the bits to the left by the specified amount No data will be lost and what is at the start of the buffer will be at the end of the buffer after the rotation Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 Rotate Right Treats the pattern data in the specified range as a circular buffer and rotates the bits to the right by the specified amount No data will be lost and what is at the end of the buffer will be at the start of the buffer after the rotation Align to Sequence Aligns the pattern data in the specified range to a specified pattern sequence 5 8B 10B This tab provide you an option to either Invert Running Disparity or Auto Correct
255. decide on the appropriate test setup for your device Determining How to Test Your Device The Serial BERT can help you test the performance of components and systems for high speed digital transmission equipment An important parameter of a digital system is the rate at which errors occur in the system The analyzer will compare the bits sent to your device with the bits received from your device It will record the number of bit errors and display results such as the bit error ratio BER In addition accumulated results can be analyzed to help you determine what caused the bit errors The analyzer can be set up to work with your device in many different ways To determine the best measurement setup you must first decide how you want to test your device Before actually setting up and making measurements you may want to consider the following questions Whatis the purpose of your test What are the testing requirements of your device What is the Purpose of Your Test Does the purpose of your test fall under one of the following categories To monitor instantaneous BER during measurement setup experiments and adjustments See Instantaneous Measurements Concepts on page 361 To measure and log error performance over time See Accumulated Measurements Concepts on page 376 To measure the eye height and width of the incoming data signal at specific alignment BER thresholds See Eye Measurements Concepts o
256. der a pattern of 1000 ones and 1000 zeros as shown in the following figure With reference alignment 1 the patterns are totally out of phase and the error detector is measuring 100 errors But as the reference moves closer to optimum alignment the percentage of errors gradually approaches zero reference alignment 2 and 3 For exact alignment the sync threshold must be set lower than the BER caused by a 1 bit misalignment in this case 1E 3 1000 ones 1000 zeros Incoming data Reference alignment 1 Reference alignment 2 Reference alignment 3 Agilent J BERT N4903B High Performance Serial BERT 209 5 Setting up the Error Detector 210 NOTE How Can You Tell if Your Synchronization is False You may suspect false synchronization under the following conditions You are using a pattern other than PRBS and the error detector gains sync but it measures a constant fixed error ratio You are using a pattern other than PRBS and the error detector gains sync but auto search functions Auto Align Clock Data Center 0 1 Threshold Center repeatedly fail In a false sync the sync threshold BER of 1E 3 may be met but eye edges at BER 1E 3 required by an auto search function may not be found This is because BERs less than 1E 3 do not exist within the data eye Possible error profile after fase synchronization WZ Possible error profile after true synchronization If you suspect a false sync try re sy
257. dialog box allows you to specify the criteria to decide whether the DUT passes or fails the test You can change pass fail criteria without rerunning a test The software only uses the criteria to rate the results of a measurement The pass fail criteria do not control measurement execution The measurement run will be completed even if the measurement fails for one or more of the criteria You can set pass fail limits individually for the output timing parameters and the jitter measurement parameters For information on the DUT Output Pass Fail criteria see also Output Timing Measurement Parameters on page 259 For information on the Jitter Pass Fail criteria see also Jitter Measurement Parameters on page 260 Agilent J BERT N4903B High Performance Serial BERT 3 72UI ul Advanced Analysis 6 Properties Ej Phase Margin fa Uptinal Sample Delay iv E iv iv LI a E If the results of a Fast Total Jitter measurement are displayed only the appropriate parameters are compared and flagged e Phase Margin e Optimal Sample Delay Total Jitter Peak to Peak e Fast Total Jitter Uncertainty Other pass fail limits may be enabled but are ignored In the tabular view each of the calculated values will be marked with an e icon if it failed the test The following illustration shows a DUT Output Timing Jitter measurement that has failed the criterion for the phase margin
258. dicates whether the chosen setting of frequency and amplitude is acceptable Agilent J BERT N4903B High Performance Serial BERT 419 8 Jitter Tolerance Tests 420 Amplitude p p The maximum peak to peak Amplitude is limited by the chosen frequency Frequency The maximum Frequency of the rSSC jitter source depends on the amplitude Amplitude Frequency Sinusoidal Jitter Parameters Sinusoidal jitter SJ is independent from periodic jitter It is generated by modulating the clock that is used for data generation It can therefore not be enabled if SSC or rSSC is active 7 100 UI 600 Hz Sinusoidal jitter is characterized by Amplitude peak to peak Frequency There is a dependency between frequency and acceptable amplitude A graph indicates whether the chosen setting of frequency and amplitude is acceptable The maximum peak to peak Amplitude is limited by the chosen jitter frequency Note that high frequency sinusoidal data jitter can be generated by the Periodic Jitter function or by applying an External jitter source Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Periodic Jitter 1 Parameters ks foom The Periodic Jitter is characterized by Amplitude Frequency Waveform A graph indicates whether the chosen setting of frequency and amplitude is tolerable 10 0000 MHz a9 un a Amplitude The maximum peak to peak Amplitude is limited by the free capacity
259. dicating the jitter active state When this button is disabled no jitter is generated This refers to all the jitter types shown on the screen even if they seem to be connected The LED on the buttons will turn orange to indicate this NOTE The Jitter On Off button has no impact on SSC If spread spectrum clocking is enabled it continues even if the Jitter On Off button is off Agilent J BERT N4903B High Performance Serial BERT 407 8 Jitter Tolerance Tests Set Jitter Configuration Use Jitter Configuration area to specify whether you want to use SSCorrSSC e RJ and BUJ or sRJ a stt R s Set Jitter Distribution 1 Specify whether the corresponding jitter source is applied to the data and or the clock output by clicking the corresponding LED button 2 Specify the delay line 220ps or 610ps used for the generation of PJ and BUJ NOTE The delay has to be defined in steps of 100ps The given value will automatically round to the next valid value 408 Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Select the Delay Line To select the delay line for sinusoidal periodic jitter 1 and 2 and BUJ press 610ps button The LED will turn green if the 610ps delay line is selected Note that sRJ can not be enabled if the 610ps delay line is selected 610ps NOTE The 610ps delay line cannot be used for bit rates above 3 37 GHz Specify the Jitter Components In this section an examp
260. disk space TIP Remember to always make a backup copy of your files If your original files are damaged or lost you can restore the files from the backup copy Recalling Instrument States Recalling the instrument state restores all the setup parameters for the instrument hardware and software To recall the setup do the following In the File menu click Open and then Open Instrument State This opens a dialog box which allows you to select the setup file you want to recall NOTE Loading the instrument state might take some time please wait untill the instrument state restore process is completed TIP Remember always to make a backup copy your setup files If your original files are damaged or lost you can restore the files from your backup copy NOTE The default pathname for user defined instrument states is C lt instrument model gt Settings Saving and Recalling Measurement Logs Measurement Logs are saved as CSV comma separated values files Follow the steps below to view results in a measurement log file 1 Copy the measurement log file from the Serial BERT to your PC By default log files are saved in the folder C lt instrument model gt log Agilent J BERT N4903B High Performance Serial BERT 559 10 Customizing the Instrument 560 NOTE 2 Open a spreadsheet application on your PC 3 Import the measurement log file If your application has an import wizard you may need to indicate that data is delimi
261. e A green area in the display indicates the voltage range specified in the Input Range fields for non differential modes Normal and Complement The Serial BERT is searching this range to find the eye when calculating the optimum sampling point Return to Results Click this button to return the 0 1 Threshold and Data Delay to the values measured during the last auto alignment The Data Inverted and Avg 0 1 Threshold checkboxes will be returned to their previous state Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 Trigger and Aux Output Trigger and Aux Output Concepts The error detector has a Trigger Out and an Aux Out port that you can use to send signals to other external devices such as an oscilloscope The Trigger Out can be setup to senda trigger to an external device either according to the clock or according to the data pattern being generated The Aux Out can be set up to either send the error detector s clock signal directly from Clk In port or the error detector s clock signal recovered from Data In port or the input data passed over a comparator Trigger and Aux Output Procedures To set up the trigger and auxiliary outputs 1 Press Trigger and Aux Out in the ED Setup menu 2 Forthe Trigger select one of the threeoptions Clock Divided by n Sends a trigger signal every 4 8 16 20 32 64 or 128 bits Pattern A trigger is sent every time the complete patter
262. e If Aux In is logic high pattern B is sent f Aux In is logic low pattern A is sent Agilent J BERT N4903B High Performance Serial BERT 165 4 Setting up the Pattern Generator Aux In Data Out A A B A B B B e Edge Sensitive Pattern A is sent out until a rising edge at Aux In is detected Then at the next complete repetition of pattern A a single occurrence of pattern B is inserted into the output Data Out A A B A B A A Output Blanking Pattern A is sent when the signal at Aux In is low Pattern B is not sent in this mode Aux In Data Out A A AA If Aux In is pulled to logic high and a standard pattern is selected no output pattern will be sent NOTE Due to the pattern generator behavior the patterns may be repeated up to 512 times before alternating See How the Serial BERT Generates Memory Based Patterns on page 69 for details 166 Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector Setting up the Error Detector Concepts The error detector analyzes an incoming bit stream compares it to the expected pattern and locates any inconsistencies The error detector requires the following settings to work correctly The expected pattern The error detector needs to know which data to expect so that it can detect bit errors Correct input and output terminations to the DUT
263. e Consider these recommendations when setting up the pattern generator see Setting up the Pattern Generator Concepts on page 121 Additional Jitter components you have to add for the compliance test if the standard requires non sinusoidal jitter components such as random jitter RJ or intersymbol interference ISI The recommended ISI setting is only valid if both pattern and bit rate are set according to the standard s recommendation see the corresponding description items Consider these recommendations when composing the desired total jitter see Jitter Setup Procedures on page 407 Note that the sinusoidal jitter is automatically generated and no setup is required Frequency Specification You can choose Auto or Manual mode In Auto mode the jitter frequency is automatically swept from Start Frequency to Stop Frequency Start Frequency This is the minimum and first jitter frequency to be generated It determines the left hand border of the graphical result display Stop Frequency This is the maximum and last jitter frequency to be generated It determines the right hand border of the graphical result display The measurement always includes the start and stop frequencies The frequency scale is logarithmic Number of Steps The Number of Steps between start and stop defines the test frequencies The location of the individual test frequencies depends on the setting of the Auto Frequencies checkbox on the
264. e BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS Bit Comparison without USB3 1 SKPOS You can determine what adjustments improve or degrade the BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS Bit Comparison without USB3 1 SKPOS Instantaneous Measurements Procedures This section describes how to run monitor and analyze instantaneous measurements Running Instantaneous Measurements The Serial BERT s error detector immediately starts calculating the error rate as BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS Bit Comparison without USB3 1 SKPOS as soon as it receives a valid signal and the respective error mode settings are done Both the pattern generator and the error detector must be correctly set up and NOTE connected to the DUT To run an instantaneous measurement 1 Check if all the values in the status line are correct pattern generator pattern and bit rate error detector pattern and bit rate 2 Viewthe BER SER FER FSR ISR DER cBER display in the top left corner of the GUI If the measurement is running properly a certain BER SER FER FSR ISR DER cBER should be displayed here In most cases the BER SER FER FSR ISR DER cBER is nota constant value but slightly varying over time If no BER SER FER FSR ISR DER cBER is displayed because there are no errors detected you can manually insert an error by pressing the Error Add button in the top right corne
265. e The browser shows all files with the suffix seq in the selected directory The sequence is shown in the Sequence Editor but NOT automatically downloaded to the pattern generator In contrast to loading a pattern from file the download of a sequence requires an additional command Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 User Defined Sequences Reference The elements of the Sequence Editor and the dialog boxes that can be opened from the Sequence Editor are described below Sequence Block Display A new sequence consists of one block that is infinitely repeated looped By default this block has a length of 512 bits and generates a Pause 0 signal a continuous stream of zeros All this is shown on the display The vertical waveform at the left hand side of the block indicates the setting of the pattern generator s Trigger Ref Clock Out port By default it generates a pulse a subrate of the present clock For information on how to change the trigger behavior see Sequence Trigger at Block Begin on page 102 Number of Blocks In the Blocks field you can directly specify the number of blocks up to 120 Agilent J BERT N4903B High Performance Serial BERT 101 3 Setting up Patterns 102 NOTE Block Contents In this example the number of blocks has been set to three A None block generates no pattern and is completely ignored indicated by a straight vertical blue line
266. e Understanding Bit Recovery Mode on page 195 Software Generated PRBS Software Generated PRBS Concepts Software generated patterns are PRBS patterns that can be modified They are calculated by the Serial BERT software and then loaded to the memory from which the necessary output is generated As such they are closely related to user patterns with the exception that they cannot be edited in the Pattern Editor window When to Test with Software Generated PRBS You can use these patterns for any application where random data and an even number of bits is required In addition you can vary the mark density or zero substitution to test the influence on the BER This allows potential problems like baseline wander for instance to be fully characterized How the Software Generates PRBS When you set up a software generated pattern the Serial BERT calculates the required pattern and loads this into the memory From here the memory is read bit by bit identically as for memory based patterns and the corresponding output is generated See also How the Serial BERT Generates Memory Based Patterns on page 69 Range of Software Generated PRBS Patterns The Serial BERT supports the following 2 n PRBS patterns Agilent J BERT N4903B High Performance Serial BERT 91 3 Setting up Patterns 92 Table 10 n Sequence Length Longest Run of 1 s Longest Run of 0 s 7 128 7 7 10 1 024 10 10 11 2 048 11 11 13 8 192 13 13 15 32 7
267. e make sure that the software matches the instrument s hardware An update can be carried out even if the software does not match the hardware In this case the instrument may go into a corrupt state or not all features may be available The applicable hardware versions are listed with the software update version You can find the instrument s hardware version on a label at the rear panel When updating the firmware server on the instrument make sure you select the Hardware installation option After the firmware server has been updated you are prompted to start a BIOS update Allow the BIOS to be updated the firmware server requires the modifications Do not interrupt the BIOS update Note that you must reboot the instrument after updating the firmware If you do not reboot the instrument the firmware may not have access to the instrument s hardware Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 Configuring the Instrument Procedures Follow the instructions below to configure the instrument according to your personal needs Setting Date and Time Your instrument uses the date and time when Saving files Saving self test information Saving calibration information To set the date and time of your instrument do the following 1 On the Utilities menu click Set Time and Date This opens the Date Time Properties dialog box 2 Setthe date and time 3 Click OK Setting the
268. e reciprocal of the present data rate For example if the data rate is set to 2 Gbit s one UI corresponds to 500ps Sinusoidal Jitter This kind of jitter can be picked up from any adjacent signal or clock The jitter histogram has a U shape Jitter amplitude Triangular Jitter Triangular jitter is always generated by a spread spectrum clock but can also be caused by other periodic sources Triangular jitter has a uniform jitter distribution Jitter amplitude Intersymbol Interference Intersymbol interference ISI is caused by bandwidth limitations of cables and backplanes It cuts off higher frequencies and hence changes the shape of the eye opening Agilent J BERT N4903B High Performance Serial BERT 399 8 Jitter Tolerance Tests 400 ios 8 8x Intersymbol interference can be simulated by inserting a defined transmission path between the pattern generator and the DUT Crosstalk Crosstalk between adjacent signal paths modulates the vertical eye amplitude This narrows the voltage range of the receiver needed for capturing data correctly For jitter tolerance tests amplitude modulation is most often simulated by modulating the data output with a sinewave signal This is called Sinusoidal Interference SI Total Jitter Total jitter is the sum of all kinds of jitter Remaining eye opening in presence of jitter Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Edge level
269. e the Bit Count Time will be shortened and thus the burst sync ratio will be low which forces the measurement of more bursts to obtain the necessary level of reliability CDR Settling Time Synchronization Time Start too late Data Input Gate In End Ending too soon If the Gate In is deactivated too soon the same problems will occur as with starting the gate too late see previous point CDR Settling Time Synchronization Time End too soon Data Input End Bitceuntiims Margin c HM Ending too late If the Gate In signal is deactivated too late the last received bits will not be part of the burst Thus the BER will increase If the BER is higher than the Burst Sync Threshold the burst will be considered a bad burst CDR Settling Time End too late Synchronization Time Data Input Bit Count Time Gate Input Pattern Synchronization Procedures To synchronize the incoming pattern to the expected pattern 1 Press Pattern Sync in the ED Setup menu 216 Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 Pattern Sync Setup Sync Mode Normal Sync Type Automatic Sync Manual C Burst Sync Threshold 1E 3 iv Mask Resync Errors Count 1 Error OK Cancel Apply Help 2 Select the synchronization mode Normal or Burst sync mode NOTE The burst sync mode is disabled in 8B 10B Comparison Bit Comparison without PCle3 SK
270. e N4903B Serial BERT The De Emphasis option works only if there is no external De Emphasis box connected to the N4903B Serial BERT Setting the Data Pattern on the Aux Data Out Port Opt 002 To set up both patterns using Second Channel mode Opt 002 begin by setting up the Aux Data From the PG Setup menu click Aux Data Setup to open PG Aux Data Output dialog Select Second Channel C Clock Divided by 2 1 500 GHz Second Channel Multiplexer Mode C De Emphasis Bit Offset 1 Cancel Apply Create your pattern for the Aux Channel in the Pattern Editor by opening the Edit Pattern on File even if you don t have a pattern on file Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 BER 0 000 Pattern geese Create the pattern then save it 1 Click Properties icon to set up the properties of the current pattern including parameters and length 2 Use the pattern editor canvas to set up your pattern 3 Click Save icon to save your pattern Agilent J BERT N4903B High Performance Serial BERT 151 Setting up t h e Pattern Generator 7 E e GUI Agilent N4903B connected to localhost BER 0 000 Remote G ED Jolie H Hh 435 En de 3 5 g En e A srungelceedinezz 1111 1100 0000 0111 0000 1111 0100 11113 1000 1100 1311 1010 0010 0100 1000 0000 0001 0000 0100 0001 1001 0011 0100 0010 0011 11
271. e PG Setup menu press PG Data Setup to modify Data and Aux Data and PG Clock Trig Setup to modify Clock and Trigger exl dec Data Output ER Clock T rigger Output 2 Click into the field you want to change 3 Enter the desired voltage using the numeric keyboard or the respective front panel knob only for Data and Clock Ifthe termination voltage is wrongly adjusted or the output ports are not connected an internal protection circuit automatically sets the output voltage levels to safe levels Connect the output ports or change the termination voltage before adjusting the output levels See Understanding the Output Level Parameters on page 126 for information about how the Serial BERT works with output level parameters Controlling the De Emphasis Signal Converter After the De Emphasis function has been enabled using the External Instrument s Config menu the following parameters are visible in PG Data Output screen Controlling the N4916A De Emphasis Signal Converter For N4916A you have de emphasis amplitude post cursor and the corresponding unit Post Cur Unit sme pos Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 To change the de emphasis amplitude ratio value for details see Understanding the N4916A B De Emphasis Signal Converter on page 37 click inside the text field and either Enter the desired value directly with the numeric keyboard Usethe knob by
272. e and make the signal connections 1 Usethe N4915A 010 matched cable pair 1x 2 4 mm to SMA for data 1x SAM to SMA for clock longer than 350 mm 2 Connect the pattern generator s Data Out port to the Data Input of N4916B and the pattern generator s Aux Clk Out to the Clock In of N4916B 3 Mountthe SMA 50 Ohm termination on the unused Data Out port of the pattern generator using the SMA to 2 4 mm adapter 4 Enable the De Emphasis Signal Converter as described in How to Enable Disable N4916B De Emphasis Function on page 46 and set the levels such that the DUT will not be damaged 5 Use a matched pair of 2 4 mm cables to connect the differential outputs of the N4916B to the DUT If only one output of the de emphasis signal converter is needed connect a 2 4 mm 50 Ohm termination to the unused output Installing the N4876A Procedure 544 The N4876A 28 Gb s Multiplexer 2 1 is an optional instrument that can be connected to the Serial BERT For general information see N4876A 28 Gb s Multiplexer 2 1 on page 17 When you install the N4876A for the first time you need to configure the USB interface of the Serial BERT For this configuration you may wish to connect a keyboard and a mouse to the Serial BERT see Connecting Peripherals on page 529 This is not mandatory but convenient the keys provided by the instrument can be used as well for Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 1
273. e checkbox advances the PG sequence during jitter tolerance characterization It will be enabled only if the sequence is send to the PG The Show Compliance Curve checkbox enables the compliance curve on the Jitter Tolerance Characterization graph This feature is available only if the compliance jitter measurement option is installed The standard is taken from the compliance measurement Setting the BER Limit 1 Setthe Target Bit Error Ratio This defines whether a measured point is flagged green or red Jitter Tolerance Properties Frequency Search BER Target Bit Error Ratio 1e 9 Verification Method Confidence Level 95 00 96 C Absolute Bits Errors Nr of Bits Nr of Errors V Relax Time Mos i OK Cancel Apply Help 2 Choose the Verification Method Setting a Confidence Level is generally recommended See also Target BER on page 435 If you enable Absolute Bits Errors the measurement proceeds with the next amplitude as soon as one of the two numbers is reached Agilent J BERT N4903B High Performance Serial BERT 439 8 Jitter Tolerance Tests 3 If desired set a Relax Time For details see Relax time on page 435 Setting the Search Mode 1 Choose one of six search modes Downwards linear Downwards logarithmic Upwards linear Upwards logarithmic Binary search Extended binary For details refer to Vertical Search Methods on page 436 Jitter Tolerance Properties Up
274. e error detector will be immediately disconnected again Error Detector Does not Count Bits If the analyzer does not count any bits press Sync Now This might happen after the Error Detector had glitches at the clock input DLL Alignment Failed If an error message containing the text DLL alignment failed is displayed increase or decrease the bit rate by 1bit s See Bit Rate Procedures on page 141 for details on setting the bit rate The instrument provides several error indicators to help you identify setup problems Pattern Generator bees o Piet 300000Gb s 2 231PRBS eer siese GN Error Detector LJ lee ion Es 300000Gb s 2 231 PRBS Enor hr Lass Loss The following list contains possible causes for setup problems sorted from the most severe to the least severe PG CLK LOSS This indicator turns red when no clock signal is detected at the PG Clock In port External Clock Input is selected and the external clock is off or not connected Your cables or the external clock is faulty ED CLK LOSS This indicator turns red when no clock signal is detected at the ED Clock In port BERT connected looped back or to your device The PG clock output is off or not connected e There is no connection to the ED clock input Your cables are faulty Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 BERT connected to your device There is no clock signal from your device
275. e factory presets consider the parameters as given in the selected standard Click a preset to select it for the error detector clock rate Double click a preset to modify it Update with measured data rate Use the Update with measured data rate button to copy the measured data rate into the Clock Rate field Add Preset Use the Add Preset button to add a new preset to the Preset list You can then enter a value a unit and a short description for this new preset Changes made in this dialog are effective immediately So if you press Cancel the settings are still valid Edit Preset Use the Edit Preset button to modify the currently selected preset from the list You can only edit the presets that are user defined Delete Preset Use the Delete Preset button to delete the currently selected preset from the list You can only delete presets that are user defined Transition Density The transition density is defined by the number of transitions in the incoming data divided by the total number of bits transmitted In this field enter the transition density in 96 Click measure to measure and update the transition density This parameter affects the loop bandwidth and thus must be either entered correctly or measured Some standards specify the loop bandwidth for a given Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 transition density In such a case enter the value given in the spe
276. e file for an option from Agilent The MAC Address is used for generating the license key code That means the key stored in the license file refers to just one instrument If you wish to enable a licensed feature on a different instrument you need another license file License handling is only available if both NIC Network Interface Card Adapter are enabled Install License The Install License dialog is used for installing a license It indicates the feature that has been selected before opening the window and allows you to browse for the associated license file The suffix of a license file is lic If the license file has not been copied to the Serial BERT and is not accessible via LAN click the Enter button Enter License Key This dialog allows you to enter the license key manually That means you need to read and probably print the license file with a text editor and type the key code Installing Hardware Options Installing Hardware Options Procedures 536 The Serial BERT has a slot for installing optional hardware Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 At the present time the J20 Interference Channel is available This module can be used for adding intersymbol interference and sinusoidal interference to the signal generated by the Pattern Generator Installing the J20 Option To install the J20 Interference Channel 1 Perform the regular shutdown proc
277. e level and crossing point parameters Example to X and Y Coordinates 0 0 1 0 Mask Setup You can list various remote commands in the mask setup area to define the relative vertical and horizontal positions of the 0 and 1 coordinates of the mask X and Y axes Commands listed in the setup area are executed as part of the mask data load process Example to Mask Setup MTES SCAL YTRACK OFF MTES SCAL Y1 0 65 MTES SCAL Y2 0 65 MTES SCAL XDEL 312 5E 12 MTES AMET NRZEYE Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 MTES INP MODE DIFF In the above example the Y1 and Y2 values define the mask with an absolute voltage value In some mask files Y1 and Y2 values are not specified and so the voltage value is relative to either Vtop and Vbase or to the eye boundaries specified by the user in the GUI A mask can be set in two modes either differential or single The above example MTES INP MODE DIFF sets the mode of the mask file to differential The mask mode when defined should match the mode set in ED Input Setup Dialog In case it doesn t match an error message will indicate this mismatch How to Enter Notes There is a tab on the top left corner of the graph for entering notes To enter notes 1 Click on the Note Tab a 2 Enter your notes in the drop down text box This gives the timestamps of the comments entered It s a scrollable editor where your latest comments are entered
278. e of these are changed in the setup or the operating temperature is changed by more than 5 Celsius compared to the previous adjustment Input timing alignment can be started from Adjustment button The type of alignment depends on the clock source selected 1 When the clock source is an external clock clean clock mode using the CLK IN connector on the front panel realignment needs to take place each time the frequency cable set or setup is changed Results are not saved 2 When clock source is AUX CLK IN internal clock the results are saved once the alignment is done Ensure that the normal and complement data outputs of the M8061A are terminated with 50 Ohm before starting the alignment Performing M8061A System Setup Adjustment Setting Follow the given steps to perform system setup adjustment setting 1 From the Navigation Menu select External Instrument s and then click Config sub menu The Config window shows a list of all instruments connected to the N4903B 2 Select the external instrument from the given list For example here you will select M8061A 3 Click Adjustment to continue the system setup adjustment setting Adjustment A message will pop up which warns you to save the instrument settings before continuing the auto adjustment process 4 Click Yes to continue if you have already saved the settings Once the auto adjustment starts you will see the adjustment progress in the following scree
279. e only counted when this signal is low The error counter can only count errors in 128 bit data blocks Thus the signal at the Gate In port must be low for at least 128 bits to enable the counter The following figure illustrates the effect of error gating Errors are only counted in the gray blocks Gate In Expected Data on Data Input A A only 128bit minimum signal granularity length not matched The burst sync mode is a special operating mode for measuring data in bursts of bits rather than one continuous stream of bits In this case the signal at the Gate In port controls the timing of synchronization and error counting for each burst Understanding Error Detector Output Ports The error detector provides several output ports that are used to connect it to other devices for further error analysis Trigger Out Use this port to send a trigger signal to another connected device such as an oscilloscope kError Out This signal can be used to trigger an external instrument to help in error analysis When any number of errors are located within a 128 bit data block a Agilent J BERT N4903B High Performance Serial BERT 169 5 Setting up the Error Detector single trigger pulse is sent If continuous errors are detected the error out signal would be a clock signal n Errors within Qu Usganuin EOS ceding ei jns 28 3 DatalN A A Error Out LU single error marking A
280. e periodical data output 1 Click Alt Pattern and Aux In menu item from the PG Setup submenu to open the Alternate Pattern Control dialog box R Alt Pattern and Aux In E Setup 2 Select one of the following options Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 Continuous A Only pattern A is output repeatedly Continuous B Only pattern B is output repeatedly Alternate AB The output stream switches between pattern A and pattern B Single Shot B With this option selected the Insert B button is enabled 3 Select OK when your selection is complete Manually Inserting Pattern B into the Output Stream You can insert a single shot of pattern B into the continuous stream of pattern A 1 Click Alt Pattern and Aux In menu item from the PG Setup submenu to open the Alternate Pattern Control dialog box A Alt Pattern and Aux In E Setup 2 Select Single Shot B and click OK to close the dialog box 3 During the test run press the Insert B button on the screen Insert B This inserts a single occurrence of pattern B into the output stream of pattern A Inserting Pattern B via External Signal You can configure the Serial BERT to insert pattern B into the continuous output stream of pattern A according to an external signal 1 Connect an external instrument to the Aux In port The signals received at this port must be TTL compatible The granularity for the
281. e the Filler Symbol Count Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 Filler Symbol Count Filler Symbols in Pattern x Pattern Count e Illegal Symbol Count Illegal symbol counter The total number of illegal symbol count received in a time interval is displayed here e Disparity Error Ratio DER The ratio of the number of illegal disparity change count to the number of symbols received in the current or last completed accumulation period specified by the gate period is displayed here The following is the expression used to calculate the Disparity Error Ratio Wrong Disparity Counter Disparity Error Ratio parii Received Symbol Count Filler Symbol Ratio FSR The ratio of the number of filler symbol count to the number of symbols received in the current or last completed accumulation period specified by the gate period is displayed here The following is the expression used to calculate the Filler Symbol Ratio Filler Sumbol Ratio Filler Symbol Counter lier 5ymoo xm al sited e Illegal Symbol Ratio ISR The ratio of the number of Illegal Symbol count to the number of symbols received in the current or last completed accumulation period specified by the gate period is displayed here Illegal Symbol are invalid K and D symbols i e 0000000000 it does not belong to any valid K or D symbols The following is the expression used to calc
282. e the incoming pattern to the expected pattern For more information see Pattern Synchronization Procedures on page 216 Evaluating Results The BER results can be viewed in the BER Results window For more information refer BER Results Window on page 366 The 8B 10B Symbol Comparison results can be viewed in the Symbol Comparison Results window For more information refer 8B 10B Comparison Results Window on page 367 The Bit Comparison without PCle3 SKPOS results can be viewed in the BER Results window For more information refer BER Results Window on page 366 The Bit Comparison without USB3 1 SKPOS results can be viewed in the BER Results window For more information refer BER Results Window on page 366 Agilent J BERT N4903B High Performance Serial BERT 189 5 Setting up the Error Detector Error Ratio Reference 190 8B 10B Presets Dialog The 8B 10B Presets dialog allows you to enter a short description values for Filler Primitives definition and Alignment Symbol for the new preset This dialog contains the following elements 8B 10B Presets Description sas Filler Primitives Definition Filler Symbols Name Align Filler Primitive 1 K28 5 D1 0 2 D10 2 D27 3 Filler Primitive 2 K28 5 D7 0 D7 0 D7 0 Filler Primitive 3 K28 5 D1 S D1 3 D1 3 Filler Primitive 4 K28 5 D27 3 D27 3 D27 3 10B Alignment Symbol Comma k235 OK Cancel Filler Primitives Definition Filler
283. e the voltage range within which the eye is located Both the high and low level of the data signal must be within this range to find the eye NOTE The input voltage range is 2V When you modify either the high or low voltage the other voltage is automatically adjusted See Sampling Point Setup Reference on page 198 for a complete description of NOTE all parameters of the Sampling Point Setup window Clock Setup Clock Setup Concepts To measure the Bit Error Rate with the error detector the bit rate of the data stream must be known Depending on the options the instrument is delivered with you could use either an external clock source for the error detector for example the Agilent J BERT N4903B High Performance Serial BERT 173 5 Setting up the Error Detector 174 clock from the pattern generator or extract the clock signal from the incoming data CDR mode CDR mode does not work for all kinds of data patterns For example if the device under test sends only blocks of ones and zeros there are no transitions in the data stream and the Serial BERT cannot recover the clock Also if you are testing bursts there are some special considerations for setting up CDR See the following sections for details How does Clock Data Recovery Work In CDR mode the CDR has to recover the clock from the incoming data To do this the hardware has to decide whether the voltage at the input connector is a logical 1
284. e these areas To do SO Move the handles of a color with the mouse or your finger if you are working directly on the Serial BERT This is particularly useful if you have set the display to show the Pseudo Color Plot By default you may see a graph like the one below 1 Trace 9510 Points Ul 401 878 ps All Errors 000e 0 600 0 mv 000e 1 400 0 mv 0006 2 200 0 mv DODe 3 00v 000e 4 200 0 mv 000e 5 400 0 mv 000e 6 600 0 m 0006 7 lt scale gt 0 448 UI 0 149 UI 0 149 Ul 0 446 Ul 0 746 Ul Relative 0 597 Ul 0 299 UI 0 000 Ul 0 299 Ul 0 597 Ul Now you may wish to know more precisely what happened between yellow and red After extending the range of yellow such information becomes available 1 Trace 9510 Points Ul 401 878 ps All Errors n 1 000e 0 600 0 mV 1 000e 1 4pm 1 000e 2 200 0 mv 1000 3 0 0 v 1 000e 4 200 0 mv 1 000e 5 400 0 mV 1 000e 6 600 0 mv 1 000e 7 scale 0 448 Ul 0 149 UI 0 149 UI 0 448 UI 0 746 Ul Relative 0 597 Ul 0 299 UI 0 000 UI 0 299 UI 0 597 Ul Eye Opening Reference 294 The Eye Opening measurement returns the results in a graphical and in a numerical form The following sections provide explanations of the measured parameters and the display options that are specific to this measurement Additionally some information is provided to explain the theoretical background behind Properties that can be specified on the various tabs of the Properties dialog bo
285. easurement point the Fast Total Jitter measurement can for example determine the total jitter at a BER of 10 in less than 20 minutes at a data rate of 10 Gbit s For details refer to Explanation of the Fast Total Jitter Measurement on page 246 Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Example Results The following illustration shows the resulting graph of a typical DUT Output Timing measurement 4 00e 0 1 Trace 55 Points Ul 2 500 ns All Errors 1 00e 1 1 00e 2 1 00e 3 1 00e 4 1 00e 5 1 00e 6 BER Threshold 1 000e 1 00e 7 Logarithmic 1 56 Ul 1 80 UI 2 04 UI 2 28 UI 252UI 2 76 Ul Absolute 1 68 Ul 1 92 U1 246Ul 2 40 UI 2 64 Ul 2 88 UI The following figure shows a typical jitter histogram with two peaks indicating the presence of random and deterministic jitter 1 Trace 79 Points Ul 333 333 ps All Errors 2 60e 10 2 40e 10 2 20e 10 2 00e 10 1 80e 10 1 60e 10 1 40e 10 1 20e 10 1 00e 10 8 00e 9 6 00e 9 4 00e 9 2 00e 9 Linear 230 00 ps 210 00 ps 190 00 ps 170 00 ps 150 00 ps Relative 240 00 ps 220 00 ps 200 00 ps 180 00 ps 160 00 ps 140 00 ps For a detailed explanation of the Fast Total Jitter measurement results see Explanation of the Fast Total Jitter Measurement Results on page 262 Understanding the Jitter Calculation After the output timing behavior of the DUT is measured the Serial BERT calculates the different jitter components Random
286. ecovery mode 0 1 THReshold Center Starts an auto search function that sets the 0 1 threshold to the optimum point of the incoming data eye on the vertical voltage axis without changing the data input delay This function can be used for determining the optimum 0 1 threshold for asymmetric data eyes or for patterns with an unequal mark density The status message bar will indicate if this function is unsuccessful This function clears the Avg 0 1 Threshold checkbox This function uses the alignment BER Threshold to determine the top and bottom eye edges Data Center This button starts an auto search function that aligns the data signal with the clock signal so that the error detector samples at the optimum point of the data eye in the time axis This automatically compensates for delays in the clock data paths preventing unnecessary errors The 0 1 decision threshold is not changed The status message bar will indicate if this function is unsuccessful Ensure that the received clock frequency is stable before using Data Center The clock data alignment process time is pattern dependent and with some large user patterns the alignment can take several minutes If you encounter such a long time with a user pattern it may be possible to first perform clock data alignment on a pure PRBS pattern This generally does not affect alignment accuracy and can minimize measurement time This tip does not apply in cases of severe pattern depend
287. ect the ED Data In to the Data Out of N4916B 2 Connect the ED Clock In to the PG Clock Out 3 Terminate PG Data Out with 500hm 4 Terminate PG Clock Out with 500hm 5 Ensure that the error detector follows the generated pattern Press the Auto Align button From the Analysis panel select the Eye Diagram page Press the Start button Agilent J BERT N4903B High Performance Serial BERT 505 9 Solving Problems NOTE 506 The eye diagram should look like this after some seconds BER 0 000 Error Add 11 10 9 8 7 6 5 4 3 2 1 0 Eye Diagram Eroradaf o Cw s E n sar Dis Properties E E Time 0 20U1 Div Delay 1 37U1 cale 63 22mW Div ffset 32mV zee se Sample Count 1 Level Q Level Rise Time FallTime Eye Amplitude Eye Height Eye Width Jitter Peak Peak Jitter RMS Cross Voltage 5 440563E 11 127 35mW 94 30m No Data No Data 221 65mV 40 23mV 306 51mUI 33 48mUI 23 34mUI 47 79 gt Pattern Generator EX m F3 Error Detector m m m m Clock Jitter Outputs A Sync Data Clock 3 00000 Gb s 2 7 PRBS Loss ISI SSC QN 299995Gb s 2771 PRBS Eror l Loss Loss Check the Output of the Pattern Generator with a Scope If the N4916B does not produce a signal check the Data Out of the Serial BERT Connect the Data Output of the pattern generator to an oscilloscope As long as the N4916B is in connected state after pressing the Enable button the pattern generator s ou
288. ed if the amplifier coupling is set to AC High Used to set the upper voltage level of the signal Low Used to set the lower voltage level of the signal Offset Used to set the offset of the average voltage level from 0 V Coupling Used to set the amplifier coupling to AC or DC Selecting AC coupling will remove the DC component from the signal This parameter can only be changed if the output is off disabled Termination Model Used to setthe termination type to balanced or unbalanced according to DUT s termination requirements This parameter can only be changed if the output is off disabled Termination Voltage Used to set the termination voltage according to DUT termination requirements This parameter can only be changed if the output is off disabled Cross Over Allows you to modify voltage level where the overlapped rising and falling edges of the logic levels intersect This adjustment varies the widths of the logic highs and lows and the allowed range is between 30 70 This parameter is enabled if the amplifier coupling is set to AC Electrical Idle State Used to set the electrical idle state to Off Idle or External You can either switch it Off or make it Idle power saving mode or can be externally controlled This parameter is enabled if the amplifier coupling is set to AC F 2Jitter Used to set f 2 jitter at the DATA OUT port in units of seconds This parameter is enabled if the amplifier
289. ed by one This counter is also incremented if high error ratio masking is turned off In BER BER Comparison without PCle3 SKPOS and Bit Comparison without USB3 1 SKPOS modes one bit error shall be counted and the number of bits is advance by one as well As it would not be possible to identify the cause of the synch loss it may not be possible to account this into either errored 0 s or 1 s However displaying the re synch counter also in BER mode is essential and clarifies the mismatch of error count and sum of errored 0 s and 1 s n 8B 10B Symbol Comparison mode the frame error counter and symbol error counter are advanced by one Since an error is counted also one received frame is counted The compared symbol counter is to be advanced by the block length in symbols without fillers Received symbols is incremented by pattern length in symbols with fillers TIT For the applications e g MIPI M Phy in which the synch loss is not caused by a bit error there is an option to count no errors instead of one when masking resynch errors Other counters such as disparity error illegal symbol counter and filler symbol counter are discarded for 200ms during re synchronization of the error detector Agilent J BERT N4903B High Performance Serial BERT 219 5 Setting up the Error Detector Error Accumulation Error Accumulation Concepts The Serial BERT instantly starts measuring the BER when the error detector receives
290. ed on the data path will always occur on the Data and Aux Data path Interference Channel The option J20 Interference Channel is available for simulating intersymbol interference ISI with or without sinusoidal interference SI This is a hardware module that can be plugged into the instrument It has two ports P1 P1 and P2 P2 Both can be connected single ended or differentially Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 The Interference Channel provides a switchable signal path with Nelco 4000 6 FR4 characteristics This path can be connected between the pattern generator s Data Out port and the DUT For instructions refer to Connecting the J20 Option on page 537 The Interference Channel provides also a sinewave generator and a resistive combiner This makes it possible to add sinusoidal interference amplitude modulation to the data signal The Interference Channel is built up as shown in the following figure Sinusoidal interference Variable trace lengths P1 The figure above shows the connections for a single trace segment combined with sinusoidal interference NOTE Note that all these relays and traces exist twice to support single ended and differential operation When intersymbol interference is switched off there is a minimum trace length between P1 and P2 ISI must be enabled if sinusoidal interference shall be added to the signal In this case the minimu
291. ed output pattern and select it as the error detector reference pattern What logic family is your device compatible with The pattern generator has preset voltage levels to work with ECL SCFL LVPECL LVDS and CML Be sure to select the correct logic family See Understanding how the Serial BERT Uses Logic Families and Terminations on page 128 What is the input termination voltage of your device Make sure it is the same voltage as displayed in the pattern generator termination list What termination voltage is the output of your device designed to work with Make sure it is set for the error detector s data port termination See Data In Port Termination on page 170 What bit rate do you need to test your device at Agilent J BERT N4903B High Performance Serial BERT NOTE Planningthe Test 1 The Serial BERT allows you to run tests from 150 Mb s up to 12 5Gb s depending on the options with which you bought the instrument See Error Detector Bit Rate on page 179 for more information Does your device require differential inputs As complementary outputs Data Clock Out and Data Clock Out may satisfy this requirement See Diagram 2 Connections for Differential Inputs on page 30 e Is the output of your device noisy or very small in amplitude You may need to manually set up a sampling point See Sampling Point Setup Concepts on page 193 Do you want to check the BER in bursts for example for loop
292. edure provided by the operating system 2 Disconnect the instrument from mains CAUTION Never plug in or remove a module while the instrumentis connected to mains power 3 Unscrew and remove the slot cover 4 Insert the module into the slot 5 Press the lever until the module rests firmly in place the front plate aligned with the Error Detector s front panel 6 Insert and tighten the fixing screw 7 Power up the instrument The software will automatically recognize the new component and provide the Interference Channel Icon and according setup functions in the Jitter window Connecting the J20 Option The J20 Interference Channel has two ports P1 and P2 They can be connected single ended or differentially For a description see Interference Channel on page 402 If you connect P1 to the Data Out of the Pattern Generator P2 provides the output that is meant to be connected to the DUT This is ilustrated in the following figure Agilent J BERT N4903B High Performance Serial BERT 537 10 Customizing the Instrument If you connect P2 to the Data Out of the Pattern Generator P1 provides the output CAUTION The module has golden 2 4 mm connectors Be sure to use the specific cables or adapters for the connections Improper connection can damage the module Installing External Instrument s Installing External Instrument s Concepts This section describes the procedure to install the external instruments with the Serial B
293. efore or is lost Subsequent manual re alignment can be done by pressing Re Align button The Symb Lock indicator uses the following colors to indicate the different states Symbol alignment not done Default indicator The indicator appears grey Symbol alignment done The indicator appears green Symbol alignment lost The indicator appears red For Bit Comparison without USB3 1 SKPOS as an error ratio mode the Symb Lock indicator indicates the EIEOS alignment 8b10b Error Indicator If the incomming data have some disparity errors or contains an invalid symbols itis indicated by the 8b10b Error indicator Invalid Symbols in ED are shown as N A in pattern editor Running disparity errors in ED data mark the corresponding symbol with red color in the pattern editor Agilent J BERT N4903B High Performance Serial BERT Sampling Point Setup Setting up the Error Detector 5 Sync Data Clock Symb 8 Error Loss Loss Loss Lock Sampling Point Setup Concepts This section provides basic information on the sampling point setup and eye diagrams How Does the Sampling Point Setup Work The sampling point of a data signal is defined by two values a point in time and a voltage level Each bit of the data signal is sampled at this point in time and in reference to this voltage level The point in time in reference to the clock signal is referred to as the data input delay and the voltage level is referred to as the 0 1
294. el of the terminal Low Level Std Dev The Low Level Standard Deviation is calculated as x threshald Mean V Ber x dBer where Mean is the Low Level of the terminal e Peak Peak Noise SaDev The peak to peak Noise is calculated as Peak Peak Noise HighLevel LowLevel Threshatd Margin Note that the Threshold Margin depends on the position of the BER Threshold e Signal Noise Ratio RMS The RMS Signal to Noise Ratio is calculated as HighLevel Lo wLey el SNR RMS StdDev 1 StdDev O Signal Noise Ratio Peak Peak The peak to peak Signal to Noise Ratio is calculated as HighLevel LowLevel P eakPeakNoise SNR PeakPeak Note that the Peak Peak Noise depends on the position of the BER Threshold Q factor Results The Q factor result parameters are summarized in the following table Table 29 Parameter Name Pass Fail Q Factor min Q Optimum Threshold min max Q Residual BER max Q High Level Q High Level Std Dev Agilent J BERT N4903B High Performance Serial BERT 279 6 Advanced Analysis 280 Table 29 Parameter Name Pass Fail Q High Level Nr Points Q High Level R 2 Q Low Level Q Low Level Std Dev Q Low Level Nr Points Q Low Level R 2 For some of these parameters pass fail limits can be set as indicated in the table The numerical Q factor parameters are defined as follows QFactor The Q factor is calculated as H Ho Q 0 0
295. elay Threshold BER All 3 878757E 011 1 504954E 001 2 497699E 001 3 878757E 011 1 504954E 001 2 499503E 001 1 551503E 011 3 504954E 001 4 922570E 001 1 551503E 011 3 504954E 001 4 952636E 001 1 551503E 011 4 950459E 002 0 000000E 000 1 551503E 011 4 950459E 002 0 000000E 000 This file can be imported into spreadsheet applications such as Microsoft Excel for example Use the semicolons as delimiters for the rows or columns Grianna Version B 16 010306 Type DUT Output Timing Jitter Ul 2 50E 05 2 1 Data Delay BER AII Compared Bits Errors All 3 60E 09 0 500044 1 OOE 06 5 00E 05 3 73E 09 0 499992 1 00E 06 5 00E 15 3 85E 09 0 499947 1 00E 06 5 00E 5 3 98E 09 0 500007 1 00E 06 5 00E 15 Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Timing Unit Definitions The timing unit specifies the timebase for the measurements It is possible to switch between the unit interval Ul or seconds for the time scale The timebase is set on the View page of the Properties dialog box The following figure and table shows the relationship between UI and seconds for a 500MHz signal 4 BER log scale 1 0 0 5 BER thresh ld M D s gt t An example of the measured range of time expressed in the different units is listed below Table 22 Timebase Earliest Optimum Latest Sampling Sampling Sampling Point Time rel 1 5 ns 0 ns 1 5 ns abs 54 2 ns 55
296. en Press the Start Accum key on the front panel to start the BER measurement For each measurement write down the number of errors at the end of the measurement Thecorrect Input Timing is selected by the following rule Look for the BER measurement that produced no error and that has the biggest distance to a measurement with errors and select the Input Timing setting that was used in this measurement Example Input Timing Setting 2 has the biggest distance to a setting that produced errors no matter if checking in forward or backward direction through the table So this is the setting that has to be used 482 Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 Table 33 Input Errors Forward Backwar Timing distance d to bad distance setting to bad setting 1 0 3 1 2 0 2 2 3 0 1 3 4 50 0 0 5 1000 0 0 6 100 0 0 Case 2 Calibration using an Oscilloscope Connect Data Out of N4916B to the oscilloscope Terminate Data Out of N4916B Connect Trigger Out of PG with Trigger In of oscilloscope Terminate Trigger Out of PG with 50 Ohm termination resistor Select each of the 6 Input Timing Settings and measure the output of the N4916B with the oscilloscope Measurement has to be done checking for erroneous sampled bits so the oscilloscope has to be adjusted to measure at a xx0 to 111 transition of the pattern The oscilloscope has to be configured to display ever
297. ency and back Thus the average frequency is lowered by half of the maximum deviation The CDR is adapted to that value Enter the maximum deviation as most standards specify Although the CDR will lock even without the presence of SSC it is strongly recommended to uncheck this box when SSC is not present otherwise the CDR will unnecessarily loose performance in lock detection and lock time Error Ratio Concepts The Serial BERT provides the enhanced capability to perform a receiver RX jitter tolerance test where it analyses 10 bits symbols as coded data to measure Symbols Error Ratio SER Frame Error Ratio FER Filler Symbol Ratio FSR Illegal Symbol Ratio ISR Disparity Error Ratio DER along with the calculated BER cBER You need to set the error ratio mode from Comparison the J BERT s Error Detector Error Ratio window For more information see Setting up Error Ratio on page 185 Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 NOTE All measurements except Jitter Measurement in 8B 10B Symbol Comparison Bit Comparison without PCle3 SKPOS and Bit Comparison without USB3 1 SKPOS modes automatically switch to BER mode and work in Bit Recovery Mode These changes are reflected in GUI too Jitter measurements also works in 8B 10B Symbol Comparison mode and uses calculated BER cBER to check against Target BER In case of Jitter Measurement the auto re sync automatically
298. enerator 4 For the pattern generator the following rules apply Below 620 Mbit s the pattern generator can only be operated with an external clock source because the internal clock source can only produce signals higher than 620 Mbit s The trigger output cannot be set up to trigger on certain pattern positions or pattern sequences If this option is enabled for example in the user interface the trigger is sent once for each pattern butthe bit position cannot be specified The option to trigger on the divided clock signal is supported as usual There are restrictions to the available clock to data delay values The pattern generator can vary the clock to data delay only within a range of 750 ps to 750 ps relative to the clock signal For frequencies above 666 Mbit S this range is sufficient to cover the complete clock cycle 1 unit interval For lower frequencies the valid data delay range is smaller than the clock cycle The pattern generator cannot generate signals with a delay outside this range Therefore the data delay cannot be set to all values within the clock cycle Clock 500 Mbit s Valid Data Delay Valid Range for i Data Delay Spread Spectrum Clocking A Spread Spectrum Clock SSC is widely used for reducing the peak electromagnetic radiation at the nominal clock frequency With SSC the clock pulse is modulated with a relatively low frequency triangle waveform This broadens t
299. ent In this case the jitter graph shows more than one peak see also Estimated Total Jitter on page 245 The jitter values Mean RMS and Peak Peak refer to the whole jitter The Gaussian marker allows to measure the normal distribution of individual jitter peaks You can thus determine how parts of the jitter histogram contribute to the overall jitter values You may have measured a bathtub curve like the one shown below 1 00e 0 Ul 8 000 ns 1 00e 1 1 00e 2 1 00e 3 1 00e 4 1 00e 5 1 00e 6 1 00e 7 Logarithmic 0 63 UI 0 38 Ul 043Ul 043UI 0 38 UI Relative 0 50 UI 0 25 Ul 0 00 UI 0 25 Ul 0 50 UI 1 Trace 84 Points All Errors BER Threshold 1 09 0 63 UI 239 6 Advanced Analysis Output Levels measurement 240 As the bathtub borders are not uniform both have two edges the linear derivative the jitter will show two peaks 100e 10 1 Trace 43 Points Ul 8 000 ns All Errors 1 00e 9 1 00e 8 1 00e 7 1 00e 6 1 00e 5 Logarithmic 0 68 UI 0 63 UI 0 57 Ul 0 53 UI 0 47 UI Relative 0 70 UI 0 65 UI 0 60 UI 0 55 UI 0 50 UI If you switch to linear scale and enable the marker you can see its bell shape 1 Trace 43 Points Ul 8 000 ns All Errors horae Gaussian Marker Value uen Mu 0 551 Ul 3 50e 9 Sigma 0 013 Ul 2 80e 9 Kappa 1 410e 0 210e 9 1 40e 9 7 00e 8 Linear 0 68 UI 0 63 Ul 0 57 Ul 0 53 Ul 0 47 Ul Relative 0 70 UI 0 65 UI 0 60 UI 0 55 UI 0 50 UI
300. ent jitter or with devices that do not work with PRBS patterns This function uses the alignment BER threshold to determine the left and right eye edges Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 Cancel Click this button to cancel the Auto Align 0 1 Threshold Center or Data Center functions while they are in progress The following parameters will be returned to their previous value or status Auto Align Canceled Data Delay and 0 1 Threshold values returned to previous Status of Data Inverted and Avg 0 1 Threshold checkboxes returned to previous Q 1Threshold Center Canceled 0 1 Threshold value returned to previous Status of Avg 0 1 Threshold checkbox returned to previous Data Center Canceled Data Delay value returned to previous BER Threshold In this list select an alignment BER threshold that is appropriate for your application The alignment BER threshold is the pre defined threshold used by the Data Center 0 1 Threshold Center and Auto Align functions to define the edges of the data input eye in the time and voltage axes You may wish to change the threshold for the following reasons Choosing smaller alignment BER thresholds will cause the auto search functions to set more accurate sampling points However if the BER threshold is set lower than the residual BER of the measurement the auto search functions will fail 1E 7 is the smalle
301. ent tabs in this dialog box to make the required settings Parameters tab The settings made on the Parameters tab define how the data is collected If you make changes here you need to run the measurement again Pass Fail tab On the Pass Fail tab you can specify the conditions that must be met by the DUT to pass a test Changes here do not require a new test run The data will be newly calculated and the results will be displayed View tab Graph tab and Color tab With the functions on these tabs you can modify the graphical display of the results at any time without rerunning the measurement This allows you to analyze the results in different ways and with different preferences After you have made your changes on the various tabs press Apply to employ the settings without leaving the Properties dialog box Or press OK to confirm the modifications and close the Properties dialog box How to Change the Colors of the Graph You can customize the colors of the graphical display This function is common to all measurements 1 In the Properties dialog box switch to the Colors tab Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 I Parameters Pass Fail View Graph Color Properties Color Set E BackColor Standard Colors Aw Color Palette Edit Custom Color Cancel Apply Help Depending on the current measurement type
302. er defined loop Loops always define the transition from the end of a block to the beginning of the same or a previous block It is not possible to jump into an existing loop It is also not possible to specify loops within loops exept the default overall loop Loop end condition Default is an INFinite endless loop Other options are AuxinHi AuxInLo Loop continues until Auxiliary Input is high or low AuxinRising AuxlnFalling Loop continues until Auxiliary Input receives a rising or falling edge Manual Loop continues until the Break button is clicked Counted loop Loop repeats a specified number of times Sequence Mode Characteristics When a sequence is executed the pattern generator is in Sequence Mode This mode differs from usual Pattern Mode Pattern related triggers cannot be generated In sequence mode the Trigger Ref Clock Out port can produce either a continuous pulse a fraction of the current clock pulse or a trigger spike at the beginning of each block that is set to TrigOn The duration of that spike depends on several conditions and is therefore undefined Software generated PRBS with polynomial 2 n cannot be referenced in a sequence If several PRBS patterns shall be included all have to use the same polynomial Every sequence has to be downloaded to the pattern generator before it can be executed Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3
303. erator Detector Jitter Generator Understanding Jitter Tolerance The behavior and the passed failed classification of data receiving devices or circuits are determined by their ability to withstand Jitter Agilent Technologies 395 8 Jitter Tolerance Tests 396 ntersymbol interference Level noise This refers particularly to the performance of phase locked loops PLLs or clock data recovery circuits CDR Jitter tolerance can be measured by applying a distorted data signal to the DUT and measuring the resulting bit error ratio To make jitter tolerance tests reproducible the signal distortion must also be reproducible This requires some definitions Jitter Receivers react differently on different types of jitter e the jitter composition the jitter frequency spectrum the combination of jitter frequencies and data rate Various test standards specify the jitter composition to be used for jitter tolerance tests The Serial BERT provides the following means for generating artificial reproducible jitter Two types of voltage controlled signal delay lines Aphase shifter for modulating the generated clock For more information see Understanding the Types of Jitter on page 397 How the N4903 Generates Jitter on page 401 Intersymbol Interference Conductors on PC boards have a limited bandwidth This causes intersymbol interference which changes the shape of the received eye o
304. ere Comparing SATA FER counter with the Serial BERT shows that due to errors at the receiver input DUTs may drop a complete symbol In such situation a protocol analyzer which only calculates CRC Cyclic Redundency Check reports Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 one single frame error In contrast to this the Serial BERT loses pattern synchronization and must re synch before the test can be continued Currently the detection mechanism for loss of synchronization is a high error ratio This causes the Serial BERT to show many errors in the Accumulated Results window though there was only one single error The serial BERT shows with such DUTs high error number the number is not accurate The Auto Resync Counter events are shown in the Accumulated Results window and the same are logged in file Log settings can be changed in Error Detector s Accumulation Setup dialog Accumulation Parameters for 8B 10B Comparison For detailed information on 8B 10B Comparison accumulation parameters see Accumulation Parameters on page 382 Accumulated Measurements for Bit Comparison without PCle3 SKPOS The Accumulated Results window for Bit Comparison without PCle3 SKPOS contains a graph and two tables Accumulated Results Accumulation Parameter Measurement Current Period Previous Period Bit Count 792 496 000 000 482 752 000 000 Wi Bit Error Ratio BER 0 00000000 0 00000000 Bit Err
305. ere are more sampling points at the edges This can greatly improve the results without dramatically increasing the duration of the test In the following example a step width of 2 mV was used This has led to 502 data points per Volt 800 0 mv 1 Trace 502 Points 900 0 mv 40V AAV 42V 43v 44v 45V AS6V ATV BER Threshold 3 000e 3 18v Logarithmic 1 000e 7 1 000e 5 1 000e 3 1 000e 1 1 000e 8 1 000e 6 1 000e 4 1 000e 2 1 000e If Edge Resolution Optimization is enabled the software first scans the threshold range in relatively coarse steps and determines the areas where the signal produces no edges that means where the BER is nearly constant Then the algorithm uses the specified step width for measuring the regions where the BER changes and a much lower resolution about one tenth for measuring the regions where the BER is constant This reduces the number of data points to 145 as shown in the figure below Agilent J BERT N4903B High Performance Serial BERT 271 6 Advanced Analysis 272 NOTE 800 0 mv Trace 145 Points 800 0 mv A0V 1141Y 42V ASV A4 V ASV ABY ATV 1418Y Logarithmic 1 000e 7 1 000e 5 1 000e 3 1 000e 1 1 000e 8 1 000e 6 1 000e 4 1 000e 2 1 000e As you can see from the figures above Edge Resolution Optimization does not sacrifice the precision of the measurement But it can reduce the measurement duration considerably espe
306. eriorate The latter is caused by increasing nonlinearity This can be inspected in the OBER vs Threshold graph 1 Trace 31 Points c EL Zoom in hum 42V i 800 0 mv 900 0 mv 10y AAY Aa EFI 45V AV Er rob ATN ABY Linear 4 _ 12 2 28 36 44 08 6 88 76 H The Q factor calculations are disabled and flagged as not applicable if the R value falls below 0 75 The R value can also be seen as an indicator of how well the noise distribution fits to Gaussian shape It will not fit for example if the received signal is dominated by cross talk or modal noise Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Eye Opening Eye Opening Concepts The Eye Opening measurement generates a three dimensional bit error rate BER diagram as a function of the sample delay and the sample threshold With this measurement the complete eye of the DUT output signal is measured The results comprise the voltage and timing of the eye opening and the optimum sampling point To getthe result the sampling delay and the input threshold of the signal are shifted within 1 5 signal periods and a user defined voltage range The results comprise the voltage and timing of the eye opening and the optimum sampling point An example that illustrates the general principle of the Eye Opening measurement is shown in the figure below Threshold Optimum sampling point max
307. ern with n bit length A trigger pulse is sent when this pattern is found in the PRBS The trigger repetition rate is one pulse per 512 pattern repetitions Example Pattern length 2 7 1 127 bits N bit pattern 1110001 Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 Data Out 10111000100 10111000100 Trigger Out Shift Trigger Position Clicking the left and right buttons will move the bit trigger position back and forward by one bit Aux Data Out Aux Data Output Concepts The Aux Data Output port on the Pattern Generator s Output ports allows you to generate a pulse with a frequency that is a fraction of the present clock Note that this output is not phasesynchronized with the clock Aux Data Out Procedures To setup the pattern generator s Aux Data Output port Click Aux Data Setup in the PG Setup menu A PG Aux Data Output dialog opens as shown in the figure below PG Aux Data Output C Clock Divided by 2 1 500 GHz Second Channel Multiplexer Mode De Emphasis BitOffset 1 Select the mode for the Aux Data Output You can select from the following options Agilent J BERT N4903B High Performance Serial BERT 149 4 Setting up the Pattern Generator NOTE NOTE 150 Clock Divider Second Channel Multiplexer mode e De Emphasis The options on the Aux Data Output dialog are disabled when the N4876A Multiplexer is connected to th
308. erns To start a new pattern 1 Click Edit Pattern on File menu item from the Pattern submenu zA Edit Pattem on File In the toolbar click the New icon This opens the Pattern Properties dialog box where you can Enter a Description for the pattern Specify the pattern s Length in Bits Specify the pattern s Pattern Type Standard or Alternate After you have entered these parameters click OK The pattern is then opened in the Edit Pattern window You can now continue with modifying the pattern according to your needs Opening Existing Patterns To open an existing user pattern 1 Click Edit Pattern on File menu item from the Pattern submenu A Edit Patter on File Agilent J BERT N4903B High Performance Serial BERT 71 J Setting up Patterns 72 2 In the toolbar click the Open icon Da The Select User Pattern dialog box opens Use this dialog box to locate and open the desired pattern You can select files of the following types 71612A BERT Pattern Files dat 86130A Pattern Files binary hex symbol 01 ptrn ASCII Text Pattern Files txt See Supported Pattern Types on page 65 for descriptions of these files Click New File to create new pattern D Editing Patterns To edit a user pattern 1 Click Edit Pattern on File menu item from the Pattern submenu A Edit Pattern on File If no pattern is loaded in the editor first open a pattern f
309. errors were measured is displayed here Error Free Milliseconds The number of milliseconds in which no errors were measured is displayed here Errored Seconds The number of seconds in which errors were measured is displayed here Errored Deciseconds The number of deciseconds in which errors were measured is displayed here Errored Centiseconds The number of centiseconds in which errors were measured is displayed here Errored Milliseconds The number of milliseconds in which errors were measured is displayed here Sync Loss Seconds The number of seconds in which the error ratio was greater than the sync threshold is displayed here Accumulation Parameters Elapsed Time The time elapsed during the accumulation period is displayed here Percent Complete For single and repeat accumulation modes the percentage of the accumulation period completed is displayed here to one decimal place For repeat mode this percentage is recalculated each time the accumulation period repeats Accumulation Mode The type of accumulation mode is displayed here The mode can be Manual Single or Repeat Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 Accumulation Period For single and repeat accumulation modes the accumulation period is displayed here The period will be displayed in elapsed seconds error count or bit count depending on the selected method Burst Results The recommended application for bu
310. esolution The following table lists all pattern resolutions Table 5 Pattern length standard Pattern length alternate Resolution lt 64 kbit lt 32 kbit 1 bit gt 64 kbit gt 32 kbit 2 bit gt 128 kbit gt 64 kbit 4 bit gt 256 kbit gt 128 kbit 8 bit gt 512 kbit gt 256 kbit 16 bit gt 1 Mbit gt 512 kbit 32 bit gt 2 Mbit gt 1 Mbit 64 bit gt 4 Mbit gt 2 Mbit 128 bit gt 8 Mbit gt 4 Mbit 256 bit gt 16 Mbit gt 8 Mbit 512 bit The above table is not valid for 8B 10B Comparison mode In 8B 10B Comparison mode the error message is generated if a pattern occupies more than 32 Mbits after reaching 512 bit boundary Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 How the Serial BERT Generates Memory Based Patterns The basic concept of how the Serial BERT generates memory based patterns is relatively simple The Serial BERT has 32Mbits of internal RAM memory used to store data patterns The RAM is organized in 64k of 512 bit blocks During test run the pattern generator reads this memory bit by bit and generates the corresponding voltage at the output port 1 high 0 low The error detector compares the incoming signal voltage data bits with the bits in the memory and counts the errors The Serial BERT has actually three RAM memory locations where patterns are stored in the Pattern Editor memory in the pattern generator and in the error detector When you modify p
311. eters are defined as follows High Level The High Level is the mean of the upper dBER dTh distribution It is calculated as Mine Y dBer threshold gt dBer Agilent J BERT N4903B High Performance Serial BERT 271 6 Advanced Analysis 278 Noise Results Low Level The Low Level is the mean of the lower dBER dTh distribution It is calculated as 5 dBer threshold bj dBer Mean MeanLevel The Mean Level is the middle between the High and Low Levels calculated as Mean Level LowLeve Highlevel Amplitude The Amplitude is the difference between its High and Low Levels Threshold Margin The Threshold Margin is the distance between the upper and the lower BER curves at the position given by the BER Threshold setting The noise result parameters are summarized in the following table Table 28 Parameter Name Pass Fail High Level Std Dev Low Level Std Dev Peak Peak Noise max Signal Noise Ratio RMS min Signal Noise Ratio Peak Peak min For some of these parameters pass fail limits can be set as indicated in the table The noise parameters are defined as follows High Level Std Dev The standard deviations are derived from the dBER dTh histogram The High Level Standard Deviation is calculated as x threshold Mean V dBer gt dBer StdDev Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 where Mean is the High Lev
312. etting of 28 2ps and a measurement time of 338 seconds Agilent J BERT N4903B High Performance Serial BERT 489 9 Solving Problems When using lower target BER values the required measurement time per Input Timing Setting is getting longer very quickly The required measurement time for a target BER of 10 is approximately 338e6 seconds per Input Timing Setting which is not practicable in a real measurement setup N4876A System Setup Adjustment 490 N4876A System Setup Adjustment NOTE The N4876A has to be adjusted to ensure error free operation for all data rates This adjustment is required once for a dedicated setup of N4903B N4876A and the cable kit being used and has to be repeated whenever one of these are changed in the setup or the operating temperature is changed by more than 5 Celsius compared to the previous adjustment When the N4876A is enabled from the N4903B the jitter amplitude in time is decreased by a factor of 2 to make the jitter amplitude in UI at the N4876A output match the setting in Ul in the N4903B GUI Performing System Setup Adjustment Setting Follow the given steps to perform system setup adjustment setting 1 From the Navigation Menu select External Instrument s and then click Config sub menu The Config window shows a list of all instruments connected to the N4903B 2 Select the external instrument from the given list For example here you will select N4876A 3 Click Adjustment to
313. ettings you selected for the measurement and the measured data This allows you to load the measurement at a later stage and inspect all data in detail You can also change display parameters to evaluate this data even more thoroughly NOTE When saving the measurement the following settings are not saved Selected pattern e Settings of the pattern generator Settings of the error detector To save these settings see Saving Instrument States on page 558 Besides saving a measurement as a file you can also export the data to any spreadsheet application Advanced Analysis Procedures This section describes some basic functions that apply to all measurements If you want to store the measurement results for later use you can either save them as an mcp file to be opened in the Serial BERT software or as a txt file for spreadsheet applications How to Save Measurements To save a measurement as an mcp file for later use in the Serial BERT user interface 1 After your measurement has finished and the results are displayed select Save Measurement from the Analysis menu 2 Select a proper file name and location and press the Save button Agilent J BERT N4903B High Performance Serial BERT 229 6 Advanced Analysis 230 How to Export Measurements To export the measurement data into a txt file for later use in external spreadsheet applications 1 After your measurement has finished and the results are displayed
314. ev_ VISAResourceName__ USB 0x15BC 0x1297 0 2 Choose the external instrument from the instrument list shown on the Config window 44 Agilent J BERT N4903B High Performance Serial BERT Setting up External Instrument s 2 Use Refresh button to reload the external instruments list in case if you are not NOTE f MOA f able to see them in the instrument list However if the Config window still do not display any externally connected instruments restart the J BERT N4903B software and firmware 3 Click Update button to update the software revision of external instrument If required TIT This button will be either disabled or not available in case of M8061A if the software updates are not required 4 Onthe System Setup area perform the calibration adjustment settings depends upon the external instrument selection For more information on calibration settings refer to the section Loading Calibration Settings on page 480 For more information on N4876A adjustment settings refer to the section N4876A System Setup Adjustment on page 490 For more information on M8061A adjustment settings refer to the section M8061A System Setup Adjustment on page 492 5 Enable the function by selecting the corresponding check boxes To enable N4916B De emphasis function refer to the section How to Enable Disable N4916B De Emphasis Function on page 46 To enable N4916B Clock Multiplier function refer to
315. external clock that must be connected to the Clock Input This clock is routed internally through a PLL You can choose between Automatic and Manual Mode on page 143 Automatic and Manual Mode The Manual Adjustment checkbox is available if you have chosen an external clock source Clock Source External v Divider fi Manual Adjustment In automatic mode the external clock frequency is continually measured When the external clock frequency changes the bit rate changes as well In manual mode you can enter the expected frequency or press the Measure amp Set button Measure amp Set requires that the external clock generator is connected and active Its frequency is measured once and determines the bit rate When the external clock frequency changes PG Clock Loss is signaled External Clock Divider The internal clock i e the bitrate is the external clock divided by the value specified in the clock divider field The external clock divider field is available if you have closen an external clock source Agilent J BERT N4903B High Performance Serial BERT 143 4 Setting up the Pattern Generator 144 External PLL Clock Divider and Multiplier The internal clock i e the bitrate is the external PLL divided by the value specified in the divider field and multiplied with the value specified in the multiplier field The external PLL clock divider and multiplier field are available if you have chosen an external PLL clock
316. f every burst NOTE The burst sync mode is disabled in 8B 10B Comparison Bit Comparison without PCle3 SKPOS and Bit Comparison without USB3 1 SKPOS error ratio modes Agilent J BERT N4903B High Performance Serial BERT 211 5 Setting up the Error Detector 212 How Burst Sync Mode Works The following figure presents the basic order of events that make up a burst mode measurement It also illustrates how the instrument operates while in burst mode no do em v CDR Settling Time Synchronization Time 3 een success Register Incoming Bits y yes S ie 3 eme no Y BER Bad Burst 1 no yes y Total Burst 1 4 34A Incr Bit Counters e NOTE Burst mode requires a complete measurement setup including an external input for the Gate In port This input is used for the sync alignment 1 Burst sync is started when the Gate In signal goes low indicating that a burst is arriving at the Data In port If the error detector is in CDR mode it first tries to generate a clock from the data A short time interval is required to stabilize the clock the CDR settling time Note that if the error detector is in CDR mode it is necessary to set a CDR threshold This is the BER threshold that the error detector uses to verify that the input pattern is synchronized See How does Clock Data Recover
317. f these lines a linear regression is performed This is illustrated in the figure below Threshold L d Gradient 1 0 K opt Th R Ly Gradient 1 59 Q factor opt Q A straight line can be expressed as Y A BX where Y is the inverse error function of BER and X is D the decision threshold The following calculations are performed for the high and low level data tye ae 2 X 2X n DY XX A n B n where n is the number of respective data points The results of the linear regression are displayed in the QBER vs Threshold graph 1 Trace 16 Points 800 0 mv 900 0 mV 410V Wae AAV 42V 43V 44v 45v ABV me n ATV 418V Linear 4 12 20 28 36 44 52 60 68 76 84 Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 This graph shows two straight lines The intersection of these lines marks the Q factor and the Q optimum Threshold Mathematically the standard deviation and mean values are calculated as 1 B Oo A H B This calculation leads to the values of ji o and o 9 Notes on the Q Results You can specify the range of data points used for these calculations by means of the Min BER for Q and BER Threshold parameters see also View Tab on page 214 Use of the Q factor Calculations For example you can calculate the Q factor from relatively few data points with measured bit error rates between
318. f your device inverts data Threshold Only available if the input is set to normal or complement mode Choices are 172 Averaging The 0 1 threshold voltage is derived from the input signal via a low pass filter This will work fine for most applications Manual input Enter a value for applications that do not provide a continuous data stream at the input for example any application using bursts because the averaged 0 1 threshold voltage will drift from the correct level when there is no input The manually set threshold voltage must of course be within the input range Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 Termination In this field enter the termination voltage that is appropriate for the incoming data signal This selection should be made before the device is connected to the analyzer If the input is set to differential mode a termination voltage can only be set for a device that is DC connected CAUTION Selecting the wrong terminations may damage your device The Data In port is connected to a 50 Ohm load impedance or termination within the error detector Data termination refers to the voltage level at the end of this load The logic output from a device requires any connected equipment including the error detector to have a specific termination voltage Input Range Before you can synchronize the error detector to the incoming data stream you need to defin
319. fault the error detector is in automatic sync mode with a sync threshold of 1E 3 This setting is recommended for most applications and usually allows the synchronization function to be transparent requiring no attention However for special applications changes can be made to the sync mode and sync threshold What Type of Synchronization Should You Use The type of synchronization you use affects how errors are measured and displayed A Sync Loss is recognized when the BER is greater than the sync threshold This can be caused by a high error rate pattern misalignment or clock loss Choose the sync mode setting that is appropriate for the type of errors you anticipate Automatic Sync with a sync threshold BER of 1E 3 is recommended for most applications With this mode selected the synchronization algorithm starts whenever the BER exceeds the threshold However it is not possible to make accurate BER measurements higher than the sync threshold Manual sync can be used for synchronizing once confirming proper pattern alignment and then measuring BERs higher than the sync threshold This is useful for the following applications To monitor the integrity of clock signals You may wish to measure BERs that exceed the sync threshold to confirm clock slip To collect data for constructing eye contour information You may wish to move the sampling point to locations in the data eye that have BERs exceeding the sync thres
320. ference This checkbox in the PG Data Setup window is used to disable or enable the Delay Control In port where you can supply an external signal to add jitter to the output signal 156 Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 NOTE This checkbox is not available on instruments on which the calibrated and integrated jitter injection option J10 is installed On such instruments the Delay Control In port is enabled or disabled from the Jitter Setup window Error Addition Insertion Error Addition Insertion Concepts To test error correction algorithms alarms and other functions that are imbedded in the data pattern you can insert logic errors flipped bits into the pattern The instrument provides several options in the pattern generator s Error Add dialog box for inserting error bits manually or automatically Error Addition Insertion Procedures You have the following options for inserting errors into the output data stream Manually Inserting Errors To manually insert a single error into the output stream 1 Click the Error Add button at the top of the screen to insert one single bit error in the output stream Error Add This button is active unless you define a fixed internal error rate The maximum number of errors that this function supports is one error per 128 bit block 2 f you want to disable all other error insertion functions so that only manual err
321. fferential case but it matches both single ended cases Threshold Voltage Input e Low Pass DC tracking on off recovered Connector Filter switches Clock Threshold NE O Voltage Data Processing Comparator gt P Bit Rate Range The Serial BERT provides bit rates from 150Mbit s up to 12 5 Gbit s depending on the instrument s options However several specific properties and limitations need to be taken into account when working at low bit rates The limitations apply to the instrument according to the following hysteresis curve If the bit rate falls below 615 Mbit s the limitations apply e If the bit rate exceeds 620 Mbit s the limitations no longer apply Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 The following figure clarifies the behavior in the range between 615Mbit s and 620 Mbit s Limitations do not apply in this range 620 615 Bitrate Mbit s Limitations apply in this range For the error detector the following rules apply Forlow frequencies you cannot use the automatic data alignment functions Auto Align and Data Center Instead you need to align the error detector manually For instructions please refer to the Online Help go to the Error Detector Setup section select Sampling Point Setup and switch to the Procedures information There are restrictions to the available sa
322. from a clock recovery module may have momentarily ceased To confirm periods of momentary clock loss check the measurement log file Additional Information In this example the analyzer was in auto sync mode If the analyzer was in manual sync mode the results would not have been measured and displayed in the same way When the clock signal resumed the received and ED reference patterns could only be aligned if the synchronization functions were activated In manual sync mode this would not have happened because synchronization functions are only activated manually In manual sync mode sync loss and high BER would have been measured from the start of the momentary clock loss to the end of the accumulation period Example Log of Sync and Clock Loss Data in this measurement log file confirms that there were periods of momentary clock loss M N ES Ss e SYNC LOSS PG CLOCK LOSS ED CLOCK LOSSIDATA LOSS Agilent J BERT N4903B High Performance Serial BERT 475 9 Solving Problems 476 NOTE O 22222225224 cOoooooooooooocooo ecoooooooooo cooooooooooocoococcoo Using the Measurement Log to Identify Problems Measurement Logs are saved as CSV comma separated variable files Follow the steps below to view results in a measurement log file 1 Copy the measurement log file from the analyzer to your PC By default log files are saved in the folder C lt instrument model Mog 2 Open a spreadsheet application on
323. g or incorrect the file will not be loaded and an error dialog box will be displayed MASK FILE N49xxx Mask Title The mask title is a quoted string of up to 15 characters The title is displayed on the mask readout when the file is loaded Example to Mask Title 2xGB Ethernetaa Region Number The region number is an integer that defines a mask violation area or polygon You can specify 1 to a maximum of 3 regions in a mask file Example to Region Number Agilent J BERT N4903B High Performance Serial BERT 327 6 Advanced Analysis 328 Region Number 2 Region Type You can define three types of regions for each region number you have specified STD defines the actual mask region MARGIN MAX defines the maximum margin area when test margins are set to 10096 MARGIN MIN defines the minimum margin area when test margins are set to 100 Example to Region Type Region Type STD Region Type MARGIN MAX or Region Type MARGIN_MIN Number of Vertices The number of vertices is an integer that specifies the quantity of X and Y coordinates needed to define a mask region or polygon Example to Number of Vertices Number of vertices 4 X and Y Coordinates These are the floating point numbers that define the locations of the mask polygon vertices for each region and region type defined in the mask file Mask Files use a relative coordinate system based on an eye diagram s zero level on
324. g for the jitter Amplitude is shared with the Constant Jitter mode Periodic Jitter 2 Variable Amplitude Sweep Parameters The Periodic Jitter 2 Variable Amplitude Sweep is characterized by Standard e Waveform Sweep Time Nr of Steps Step Distance A graph indicates whether the chosen setting of frequency and amplitude is tolerable Agilent J BERT N4903B High Performance Serial BERT Standard NOTE Waveform Sweep Time Nr of Steps Step Distance Jitter Tolerance Tests 8 You can use this drop down list to specify whether you want to select a pre defined standard or a user defined standard All the available pre defined standards will be shown in this list However if you select the user defined standard press Load button to locate the Jitter Tolerance standard The user defined standard uses the same file format like the Jitter Tolerance Compliance measurement See Jitter Tolerance Compliance Concepts on page 446 Selecting the Standard does allocate the required amount of jitter modulation on the delay line being used To avoid errors when changing the selection it is recomended to either select the corresponding bit rate first or enable the PJ2 source after setting the correct bit rate You can switch between sine rectangular or triangular jitter source waveforms You can specify the duration for sweeping the selected jitter profile standard once You can specify the number of steps
325. g on the maximum data rate of the PG and the N4916B Set Trigger Out Pattern Trigger at Bit Position 0 for the scope measurement Set Trigger Out Offset 0V Set Pattern PRBS 2 7 1 for PG and ED Set Error Detector Alignment BER Threshold 1E 1 Set Error Detector Nomal Input Single Ended Set Error Detector Pattern Sync Normal Manual Example Calibration Using a Sampling Oscilloscope The following example shows the Input Timing Calibration for a target BER of 10 and a confidence level of 95 Lower target BER values will lead to more reliable results for the settings that are close to the worst Input Timing Setting But lower target BER values also lead to significantly longer measurement times Please refer to the notes at the end of the example for details about the required measurement time at different target BER values The example is using the following instruments 1 N4903B with option D14 2 N4916B specified for up to 14 2Gb s 3 Infiniium DCA 86100A with a 83484A module Connect the N4916B with the N4903B using the N4915A 010 matched cable pair Agilent J BERT N4903B High Performance Serial BERT Calibration Preparation 1 Solving Problems 9 Connect the N4903B s Trigger Output with the oscilloscope s Front panel Trigger input Connect the N4916B s Data Output with channel 1 of the oscilloscope Terminate the N4916B s complement output using a 500hm termination resistor N4903B a Open the co
326. g up Patterns 112 Table 15 Icon Name d Find Block Edit E Select All Description Click this icon to open the Find Pattern Segment dialog box and perform the search and replace operation for a specified segment in the pattern See Find Pattern Segment Dialog Box on page 84 for details Click this icon to edit a block of data in the pattern s See Block Edit Configuration Dialog Box on page 85 for details Click this icon to select all the bits in the pattern Agilent J BERT N4903B High Performance Serial BERT Table 15 Setting up Patterns 3 Icon E Agilent J BERT N4903B High Performance Serial BERT Name INS OVR Go To Bin Hex Symbol Description Click this icon to toggle among the insert dynamic insert and overwrite editing mode With Insert mode you insert bits at the cursor position shifting the rest of the pattern to the right The pattern length is not changed on insertion For every inserted bit at the cursor position one bit at the end of the pattern will be removed With Dynamic Insert mode you can change insert or delete bits at the cursor position without deleting previous bits This will either increase or decrease the length of the pattern by the number of entered bits In case of deleting the bits atleast one symbol 10bits 4bits 1 bit should be present With Overwrite mode you overwrite the bits at the cursor position Click
327. gh Performance Serial BERT Setting up Patterns 3 Fill with Ones Invert The Invert does the following functions Symbol Mode off Invert the bits Symbol Mode OOB 0 becomes 1 1 becomes 0 Z remains unchanged Symbol Mode PAMA Invert bits on per channel basis Clock Pattern 0101 0101 e Clock Pattern divide by Symbol NOTE All the above options are not available in all modes The Symbol fill is available if the Symbol Mode is enabled NOTE In PAM4 mode the symbol that is used for a 0 is 00 and for a 1 it is 11 This will be used for Fill with 0 Fill with 1 and for the Clock pattern Fills 2 Recode This tab is used to change the coding of a selected range in the pattern Fill Recode Traces Rotate Symbol Bits Change To 0 00 0 00 1 i 1 11 X 01 Zz 01 undefined 10 Z 01 Convert Undefined to Z Execute The button Convert undefined to Z provides an easy access to fix all undefined symbols to the coding of a Z symbol It is only accessible when symbol mode is set to OOB as there is no undefined symbol in PAM4 mode Agilent J BERT N4903B High Performance Serial BERT 87 3 Setting up Patterns NOTE The Recode tab is not accessible if Symbol Mode is set to OFF NOTE 3 Traces This tab provides the following options Fil Recode Traces Rotate Copy Data to Aux Data Out C Swap Traces Reverse Traces Copy Choose this option to copy
328. ground behind The properties of the Output Levels measurement can be specified on the various tabs of the Properties dialog box Parameters Tab If you modify the parameters on this page you have to rerun the measurement to update the results Set the criteria for moving to the next sample point Number of Compared Bits After this number of compared bits the measurement stops for the current sample point and moves to the next one Number of Errors Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 After this number of errors the measurement stops for the current sample point and moves to the next one This allows you to speed up the measurement You can switch off this option if only the number of compared bits is important NOTE The measurement moves to the next sample point when the first of the two criteria is reached Set the criteria for the sample threshold Resolution Specifies the distance between the sampling points The lower this value is the more sampling points you have in the selected voltage range The minimum step width is hardware dependent At the time being the minimum is 1 mV Low Level This is the lower end of the measured voltage range High Level This is the upper end of the measured voltage range Edge Resolution Optimization Turns the resolution optimization on or off If this option is enabled the Serial BERT intelligently sets a resolution so that th
329. gt lt line feed or just Return The list must be sorted according to the frequencies in ascending order The jitter amplitude for a given frequency must not exceed the instrument s capability The list of predefined standards has a Browse option that allows you to identify and load such a text file If you have selected a text file you can append itto the list of predefined standards You can assign a name to your standard and add a description Then your standard appears directly in the list like the standards provided by Agilent and you don t have to browse any more Your standard is stored in a file with the suffix jcs This is still a text file which can be edited The directory is C lt instrument model NJTolStandards User defined standards are taken as is That means neither the bit rate nor any other conditions are taken into account The instrument uses double logarithmic interpolation to generate the jitter amplitudes between two successive amplitude values This is indicated by linear slopes in the double logarithmic graph Special Features You can freely specify the jitter frequency range and the number of points to be measured You can also adjust the frequency range and the point positions to the loaded standard In addition you can add a margin to the standard Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Jitter Frequency Test Points You can specify the jitter fre
330. h optional de emphasis to extend the rate of J BERT N4903B pattern generator The M8061A window is shown in the following figure AN BER 0 000 Remote Unbalanced The M8061A window contains the following tabs 1 DataOut Tab provides parameters to set the amplifier deemphasis and interference For more information refer to the section DataOut Tab on page 56 ClkGen Tab provides parameters to select clock source For more information refer to the section ClkGen Tab on page 58 Electrical Idle In Tab provides parameters to set electrical idle in For more information refer to the section Electrical Idle In on page 58 The range of the parameters provided by M8061A window are listed in M8061A Parameters and Range on page 58 The DataOut tab provides parameters to set the Amplifier Deemphasis and Interference You can set the amplifier using the following parameters Auto Range Used to enable disable Auto Range Enabling the Auto Range will allow J BERT N4903B software to switch to amplitude ranges automatically without user intervention Amplitude range can be between 50 mV 1 2 V Amplitude Range Used to select one of the 8 pre defined amplitude ranges manually This parameter is disabled if the Auto Range option is enabled Agilent J BERT N4903B High Performance Serial BERT Setting up External Instrument s 2 Amplitude Used to set the amplitude of the signal This parameter is enabl
331. ha al fo fo at ah eh al iiaiai 1100 0011 1100 1111 0000 1100 0011 1100 ibas ajiaiioiel abe a upubitan kal tayo ahah al al elu tetia ak alo io 0011 1100 1111 0000 1100 0011 1100 1111 eUin ab alata iohtey eh ak abul fait e e e a e a a ELSE 1100 1111 0000 1100 0011 1100 1111 0000 a ak ye te fa a shal see eta apo all fo ie 0 CO CO CO CO CO CO CO CO CO CO CO CO CO OJ CO CO CO CO CO CO CO CO CO CO CO CO CO QO CO CO CO W CO CO CO O CO CO CO CO CO 0 CO CO CO CO CO CO CO O CO CO CO CO CO The symbols marked with amp denotes symbol disparity while the symbols without signs denotes neutral disparity The symbols with red color signifies the error in runnning disparity The symbols with N A signifies an invalid symbols Not a valid 8B 10B symbol The status bar below the editor canvas provides the following information from left to right Status Offline indicating that this is the display of the editor not the pattern generator nor the error detector Current Position in bits of the cursor in the pattern Editing mode Overwrite Dynamic Insert Insert Display mode Bin Hex Symbol Mark Density of the pattern s Total Length of the pattern in bits Error Indicators Agilent J BERT N4903B High Performance Serial BERT 81 3 Setting up Patterns 82 TIP Description Length in Bits Disparity Error D This indicator turns red if there is
332. has it s own comparator for the incoming data This comparator also needs to know the threshold voltage 0 1 decision threshold Agilent J BERT N4903B High Performance Serial BERT 175 5 Setting up the Error Detector 176 The threshold voltage can be derived from the input signal via a low pass filter This will work fine for most applications But applications that do not provide a continuous data stream at the input for example any application using bursts cannot use this low pass filter because the threshold voltage will drift from the correct level when there is no input In such cases the threshold can be specified manually It is then no longer derived from the input signal see the following figure The manually set threshold voltage must of course be within the input range The difference between the data path and the CDR path is that the comparator of the CDR is always single ended Thus this comparator always needs a threshold voltage that lies between the high and low levels of the incoming signal The differential threshold of the data path comparator has no relation to the single ended threshold of the CDR path comparator This means that in differential mode the two thresholds will be different and in single ended mode either normal and complement they will can be equal except during measurements The following figure shows a simplified block diagram It does not reflect the different input modes especially the di
333. he C lt instrument model NJTolStandards directory If you want to store the measurement results for later use you can either save them as an jcp file to be opened in the user interface or as a txt file for spreadsheet applications Saving Measurements To save a measurement as an jcp file for later use in the user interface 1 After your measurement has finished and the results are displayed select Save Measurement from the File gt Save menu 2 Select a proper file name and location and press the Save button Agilent J BERT N4903B High Performance Serial BERT 461 8 Jitter Tolerance Tests 462 Exporting Measurement Results To export the measurement results into a txt file for later use in external spreadsheet applications 1 After your measurement has finished and the results are displayed select Export Data from the Analysis menu 2 Inthe Save as dialog box select the path and file name for the txt file 3 Press Save to export the data to the specified destination Loading Measurements To load previously saved measurements 1 Select Open Measurement from the File gt Open menu 2 Use the file dialog box to locate the desired jcp file and press Open to load the measurement Accessing Comment and Condition Text To access the comment or condition text of a measurement Double click Modify Comment below the graphical display Modify Comment This opens the Jitter Measurement Comment dialog box You
334. he J BERT N4903B or from the menu bar select File Preset Instrument State 3 Disable the outputs of N4903B and use the matched cable kit to connect the pattern generator to the M8061A The Data In 1 Data In 2 and Aux Clk In of M8061A has to be connected to Aux Data Out Data Out and Aux Clk Out ports of the J BERT N4903B respectively 4 From the Navigation Menu select External Instrument s and then click Config sub menu The Config window shows a list of all instruments connected to the J BERT N4903B Confirm the presence of M8061A entry in that list However if the Config window do not display any externally connected instruments restart the J BERT N4903B software and firmware Agilent J BERT N4903B High Performance Serial BERT 519 9 Solving Problems A BER 0 000 Identify Connecion Enable Function S WRev VlSAResouceName EVO 14 21 USB 0x15BC 0x1297 0 5 Corresponding to the M8061A entry select the Mux with Deemphasis function check box from the given options under Enable Function column It opens the M8061A Connection dialog 6 Click Enable button It enables M8061A Multiplexer with Deemphasis that is connected between data out of J BERT N4903B and the DUT Enable Once the Mux with Deemphasis function is enabled you will see the Mux with Deemphasis parameters in the M8061A window BER 0 000 Remote Unbalanced 2 mY p 520 Agilent J BERT N4903B High Performance Serial BERT
335. he Numerical Results The measurement provides numerical results for Bit Error Rate Total Power Noise Power The Frequency and total power for up to the 16 frequencies with the highest total power You can set the number of top frequencies to be evaluated in the View tab see View Tab on page 256 for details Agilent J BERT N4903B High Performance Serial BERT 359 6 Advanced Analysis 360 Agilent J BERT N4903B High Performance Serial BERT Evaluating Results Evaluating Results Concepts The Serial BERT offers several different kinds of tests that can be run Instantaneous Measurements This type of measurements is used to monitor the instantaneous BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS Bit Comparison without USB3 1 SKPOS during measurement setup experiments and adjustments e Accumulated Measurements This type of measurements is used to measure and log error performance over time Also you can capture the location of bit errors for further analysis Eye Measurements This type of measurements is used to measure the eye height and width of the incoming data signal at specific alignment BER thresholds This information is displayed on a representation of an eye diagram Instantaneous Measurements Instantaneous Measurements Concepts With instantaneous measurements you can directly view the current BER 8B 10B Comparison Bit Comparison without PCle3 SKPOS
336. he clock signal spectrum and reduces the peak energy The Serial BERT has a built in SSC generator This is a licensed option For information on how to enable this feature see Installing Hardware Options Procedures on page 536 For more information on SSC see Jitter Setup Concepts on page 404 Bit Rate Procedures You can use an external clock source or the Serial BERT s internal clock to control the bit rate Agilent J BERT N4903B High Performance Serial BERT 141 4 Setting up the Pattern Generator 142 To set the bit rate 1 If the external clock source has a frequency of 10 MHz connect the clock source to the pattern generator s 10 MHz Ref In port If the clock has another frequency connect it to the Clock In port Click Bit Rate Setup menu item from the PG Setup submenu a Bit Rate Setup Select the appropriate clock source Internal Clock Source This setting uses the internal clock oscillator 10 MHz Reference Clock This locks the clock generator to an external 10 MHz reference External Clock Source This connects the external clock directly to the clock generator If you have chosen an External Clock Source decide between manual or automatic mode In automatic mode the clock rate of the external source is automatically measured and used If the source frequency varies the generated clock varies as well n manual mode you can measure and set the rate of the external c
337. he following installed e C07 Bitrate range 150 Mbit s to 7 Gbit s pattern generator and error detector with built in clock data recovery CDR C13 Bit rate range 150 Mbit s to 12 5 Gbit s pattern generator and error detector with built in CDR A01 U01 Bit Recovery Mode measures BER without expecting a particular pattern J10 Agilent J BERT N4903B High Performance Serial BERT Planningthe Test 1 Option Upgrade for jitter injection Total jitter can be composed from random periodic sinusoidal and bounded uncorrelated sources J11 U11 Spread Spectrum Clock SSC and residual Spread Spectrum Clock rSSC generation J12 U12 Option Upgrade for automated jitter tolerance compliance tests Use this test to ensure that the jitter tolerance of the device under test complies with a certain standard gt J20 User installable hardware module Interference Channel for simulating intersymbol and sinusoidal interference 003 U03 Option Upgrade for half rate clocking with variable duty cycle A02 The symbol error ratio SER analysis allows error counting of coded packetized and retimed data streams SATA and USB3 are popular examples of serial bus standards using retimed loopback mode for receiver tolerance testing SER analysis includes the automatic handling of the running disparity of 8B 10B coded patterns filtering of up to 4 user definable filler symbols filtering without any dead times up to 11
338. he measurement proceeds with the next amplitude as soon as one of the two numbers is reached 3 If desired set a Relax Time For details see Relax Time on page 450 Before Running the Test Before running the test set up the instrument to ensure that the requirements of the selected standard are met Recommended test pattern and additional jitter are displayed in the standard s description field on the Standard tab of the Jitter Tolerance Properties dialog For details on the recommendations and their dependencies see Description in the Standard Specification on page 462 For DUTs that require a training sequence it is essential to prevent clock loss at the beginning or end ofthe test To avoid clock loss the sinusoidal jitter component has to be enabled Since the sinusoidal jitter for the test is generated internally the jitter amplitude must be set to zero 1 Press Jitter Setup from the Jitter submenu Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Jitter Setup el This opens the Jitter Setup window 2 Enable jitter generation Press the respective checkbox ES Jitter on off i 3 Activate Sinusoidal jitter zZ 4 Activate Periodic jitter Z Jamur 10 0000 mHz 5 Set the jitter amplitude to 0 UI for both Sinusoidal jitter and Periodic jitter Bil NOTE When rSSC is enabled the measurement does not use SJ The maximum available jitter amplitude is reduced for the jitter
339. he output port is enabled See Setting up Patterns Concepts on page 61 for more details Alternating patterns consist of two patterns one of which is pattern A the A half the other is pattern B the B half Both patterns are of equal length each up to 16Mbit The Alternate Pattern Control dialog box lets you control when which pattern is sent See Pattern Alternation Procedures on page 162 When to Use Alternating Patterns Alternating patterns can be used To systematically insert errors in the bit stream at particular positions The error detector expects pattern A only If you set up pattern B so that there are only minor differences when you run the test the error detector will detect the changed bits as errors To see how long it takes your DUT to settle You can set up an easy to process output stream as the standard pattern and a difficult output stream as the alternate pattern for example all Os for the standard pattern and alternating 1s and Os as the alternate pattern You can then track how long it takes for your DUT to recover after the alternate pattern has been sent Pattern Alternation Procedures 162 The Alternate Pattern Control dialog box controls how the output stream switches between the standard pattern and the alternate pattern You have the following possibilities for setting up alternating patterns Setting Up a Periodical Output To configure the pattern generator to generat
340. here Data Rate Ratio The ratio of the number of SKPOS bit count to the Expected SKPOS bit count received in the current or last completed accumulation period specified by the gate period is displayed here Auto Resync Counter The total number of Auto Re Sync Count received in a time interval is displayed here Accumulation Parameters for Bit Comparison without PCle3 SKPOS For detailed information on Bit Comparison without PCle3 SKPOS accumulation parameters see Accumulation Parameters on page 382 Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 Accumulated Measurements for Bit Comparison without USB 3 1 SKPOS The Accumulated Results window for Bit Comparison without USB 3 1 SKPOS contains a graph and two tables s Accumulation Parameter it Ci a Bit Error Ratio BER 0 00000000 0 00000000 Bit Error Count 0 0 B Errored 1 s Ratio 0 00000000 0 00000000 Errored 1 s Count 0 Errored 0 s Ratio 0 00000000 0 00000000 Errored 0 s Count 0 Auto Resync Counter t NOTE The Accumulated Results window for Bit Comparison without USB 3 1 SKPOS does not contain tables for G 821 Measurement Internal Results and Burst Results Ratio Graph This graph displays the delta symbol errored 1 s ratio delta symbol errored 0 s ratio and total delta symbol error ratio at data points over the entire accumulation period For more information see Ratios Graph on page 379
341. hile the measurement is running correct and incorrect data is captured The resulting records contain the captured data and the corresponding error information Agilent J BERT N4903B High Performance Serial BERT 343 6 Advanced Analysis 344 These records are automatically processed The error information is subject to a fast Fourier transform FFT FFT reveals the spectral components and their power Several window algorithms are provided to reduce the influence of leakage Jitter Distribution Over Time Jitter has a more or less characteristic distribution over time The histogram of pure random jitter shows its Gaussian distribution as illustrated in the following figure Deterministic jitter periodically adds and subtracts a delay to from the received signal Jitter caused by a square wave or on off signal has produced the following histogram Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Jitter caused by a triangle signal shows an even distribution Last but not least the histogram of jitter modulated by a sinusoid exhibits two significant peaks The jitter histogram of a real world signal shows most often a mixture of these characteristic distributions Measurement Results The Spectral Jitter measurement provides A graphical view of power vs frequency This makes it easy to identify prominent spectral jitter components You can adjust the horizontal and verti
342. his section you will learn about How to Prepare the Eye Diagram Measurement on page 320 How to Initiate an Eye Diagram Measurement on page 321 How to Change the Default Settings of an Eye Diagram on page 322 How to Set Markers on page 325 How to Run the Mask Measurement on page 326 How to Customize Mask Files on page 327 How to Enter Notes on page 329 How to Save and Recall on page 329 How to Export Fetched Data on page 330 How to Prepare the Eye Diagram Measurement An example to prepare an Eye Diagram measurement 1 Disable the pattern generator outputs by pressing the OV Disable button in the PG Setup gt Data Output screen Connect the DUT to the pattern generator s Data Out port and the error detector s Data In port Refer to the Getting Started Guide Terminate all non connected pattern generator output ports with 50 O Switch to the Pattern panel and click Pattern Select Select an appropriate pattern for example a pure 2 15 1 PRBS for both the pattern generator and the error detector Specify the logic levels and the bit rate for the pattern generator setup For example select ECL levels and a clock speed of 1250 MHz This corresponds to a clock period of 0 8 ns Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 7 Setupthe error detector to match the input range and the termination with the pattern generator s level
343. hold This mode doesn t allow the analyzer to automatically synchronize if the BER becomes greater than the sync threshold For example the analyzer will not re synchronize after momentary clock loss Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 Adjusting the data input delay may cause momentary clock loss If you select NOTE Manual Sync mode this may also result in sync loss Burst sync mode is a special operating mode for measuring data in bursts of bits rather than one continuous stream of bits For more information refer to Introduction to Burst Sync Mode on page 211 NOTE If the error detector is in Manual Sync mode it is recommended that you keep an eye on the SYNC LOSS light at the top of the instrument There are various actions that can lead to loss of synchronization Check the SYNC LOSS any time you make changes to the instrument What is False Synchronization For patterns other than PRBS the error detector may gain sync at a point in the pattern that meets the sync threshold but is not the correct point where the internal reference pattern and the received data pattern match This is called false synchronization False synchronization cannot occur with PRBS patterns because a 1 bit NOTE nae 9 misalignment would cause a measurement of 50 or more errors Thus the BER during a misalignment would always be greater than the sync threshold BER For example consi
344. hreshold is used to separate between total power and noise power All components below the Noise Threshold are considered noise When the threshold is changed the calculated Noise Power values change If desired change the Number of Top Frequencies to Show This determines the number frequency power pairs in the numerical section Up to 16 pairs can be calculated and displayed Graph Tab On the Graph tab you can use the several options to optimize the graphical display according to your needs Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Frequency Scale Choose between logarithmic and linear scale for displaying the frequencies Power Unit Display the power either in a logarithmic dB or linear scale Frequency Axis Range Show the entire frequency range or zoom in one part of it Markers To analyze the graphs at a particular point you can use the markers Additionally you can display all related values for the markers in the marker readout Pressing the Reset Markers button will set the markers back to the default positions Zoom Several zoom factors are available When you show the zoom graph you can also allow the zoom graph to track the mouse or your finger if you are working directly on the Serial BERT 1 Trace 65535 Points 28 dB 21 dB 14 dB 7 dB 0 dB 7 dB ene rey eS at WIE CT E EI 14 dB 21 dB 28 dB 35 dB 42 dB Logarithmic 100 00 MHz Explanation of t
345. ia J BERT N4903B or stand alone USB and LAN connectivity N4916B De Emphasis Signal Converter The N4916B is a 4 tap de emphasis signal converter with optional clock multiplier It is intended to accurately characterize your multi gigabit serial interfaces Following are the features and benifits of N4916B Generates 4 tap de emphasis with variable de emphasis levels Supports data rates from 660 Mb s to 10 5 Gb s Tolerates non balanced patterns Transparent to jitter Optional clock multiplier Option 001 Small size Agilent J BERT N4903B High Performance Serial BERT Planningthe Test 1 Programmable via J BERT N4903B or stand alone USB and LAN connectivity Once the N4916A B is connected it can be conveniently controlled via the Serial BERT N4876A 28 Gb s Multiplexer 2 1 The Agilent s N4876A is a 2 1 multiplexer with an output data rate of up to 28 4 Gb s SE Agilent Technologies N4B76A 2na0 Py Motoren 24 cem eew OUT caour ara GUT TA vy e Dy o DE aman Following are the features and benefits of N4876A 2 1 multiplexer driven by Serial BERT data and aux data or by ParBERT Data rate up to 28 4 Gb s DCcoupled tolerates unbalanced patterns Transparent to Serial BERT generated jitter Small box can be located closely to DUT Controlled via the local user interface of Serial BERT Remote Programming is transparently done via Serial BERT M8061A 28 Gb s Mult
346. ial BERT Customizing the Instrument 10 Configuring the Instrument Reference The Serial BERT has the following utilities for configuring the instrument Set Date Time Dialog Box This function opens the Date Time Properties dialog box of Windows Use this dialog box to set the date and time as well as the time zone of your location GPIB Address Change Dialog Box Your instrument is set to a default GPIB General Purpose Interface Bus address You can change this address by doing the following 1 On the Utilities menu click Change GPIB Address This starts the GPIB Address Change dialog box 2 Select an address from the New GPIB Address list 3 Click OK The new address will be applied and the dialog box will close Each instrument must be set to a unique GPIB address to avoid multiple instruments transferring data at the same time TIT The default address is 14 however addresses from 0 to 31 may be used if the default address is the same as another instrument s GPIB address Address 21 is usually reserved for the computer interface Talk Listen address and should not be used as the instrument address Configuring the Touchscreen Configuring the Touchscreen Procedures The following procedures help you to configure the touchscreen according to your personal needs Turning the Touchscreen Off On This function disables the instrument touchscreen external keyboard and mouse Agilent J BERT N4903B High Perfo
347. ication analyzer or oscilloscope e Fora clock recovery circuit stress test vary the mark density or use CID Consecutive Identical Digit patterns that contain long runs of 1 s and 0 s Alternatively use the Serial BERT or a DCA with an ITU T specified pattern for testing STM n systems This pattern is to verify the adequacy of timing recovery and low frequency performance of STM n systems It consists of a user pattern that is comprised as follows The first row of overhead bytes for the STM n system all ones zero timing content high average signal amplitude pseudo random data with 50 96 mark density ratio This is followed by a repeat of the pattern overhead all zeros zero timing content low average signal amplitude and the PRBS For full details see ITU T Recommendation G 958 For a regenerator test set up the bit rate according to the device specification TIP If your device converts the input pattern the error detector needs a different pattern than the pattern generator See Loading Patterns to the Pattern Generator and Error Detector on page 73 for details 64 Agilent J BERT N4903B High Performance Serial BERT User Patterns Setting up Patterns 3 User Patterns Concepts NOTE NOTE User patterns are user editable patterns that are written to the Serial BERT s memory for pattern generation see How does the Serial BERT Generate Patterns on page 61 for details User patterns can contain up to 32Mb
348. ick Error Add Setup menu item from the PG Setup submenu Tii Error Add Setup 2 Select Internal 3 Use one of the following options to define the Error Insertion Rate Error Rate allows you to select a BER from a list Average Number of Bits Between Errors allows you to specify the average number of regular correct bits that are sent without errors before inserting the next error 4 Click OK to close the dialog box 158 Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 To find out how your DUT reacts on very small bit error rates set up the pattern uir generator to enter errors once every 10 bits and run a longer accumulative test You can then find the DUT s true error rate by calculating the difference between the bit error rate set up in the pattern generator and the accumulated bit error rate found by the error detector If you set up too high an error rate the error detector will not be able to synchronize to the incoming pattern When setting up an error rate always make sure that the synchronization threshold is higher than the bit error rate Error Addition Insertion Reference The following functions are available to insert errors in the bit stream Error Add Button The Error Add button in the top right corner is used to insert a single bit error in the data stream Error dd This button is always active even if other error insertion mechanisms are running
349. icon to send the pattern from the editor to the pattern generator and the error detector These functions follow Microsoft Windows conventions Click this icon if you want to jump to an arbitrary bit position See GoTo Bit Dialog Box on page 82 for details 71 3 Setting up Patterns Table 9 Icon Name Description INS Click this icon to toggle 2 among the insert dynamic OVR insert and overwrite editing mode With Insert mode you insert bits at the cursor position shifting the rest of the pattern to the right The pattern length is not changed on insertion For every inserted bit at the cursor position one bit at the end of the pattern will be removed With Dynamic Insert mode you can change insert or delete bits at the cursor position without deleting previous bits This will either increase or decrease the length of the pattern by the number of entered bits In case of deleting the bits atleast one symbol 10bits 4bits 1 bit should be present With Overwrite mode you overwrite the bits at the cursor position Bin Clickthis icon to open Bin Hex Symbol d Configuration dialog See Symbol Bin Hex Symbol Configuration Dialog Box on page 80 for details 78 Agilent J BERT N4903B High Performance Serial BERT Table 9 Setting up Patterns 3 Icon Agilent J BERT N4903B High Performance Serial BERT Name Properties Find Block Edit Select All Alt Pat V
350. iew Description Click this icon to change the properties of the current pattern See Pattern Properties Dialog Box on page 82 for details Click this icon to open the Find Pattern Segment dialog box and perform the search and replace operation for a specified segment in the pattern See Find Pattern Segment Dialog Box on page 84 for details Click this icon to insert a block of data in the pattern at the current cursor position See Block Edit Configuration Dialog Box on page 85 for details Click this icon to select all the bits in the pattern Click this icon repeatedly to view alternating patterns in the following Ways Pattern A only Patterns A and B in interleaved rows Patterns A and B in horizontally split windows If only a standard pattern is selected this icon has no function 79 3 Setting up Patterns TIP Table 9 Icon Name Description ig Pat Click this icon to open the Pattern Select Form dialog box For more information see Loading Patterns to the Pattern Generator and Error Detector on page 73 Sel If you are working on a remote PC you can use the keyboard shortcuts for Cut Copy and Paste Also you can use the clipboard function to copy strings in either binary or hex format to and from other applications When pasting them into the pattern in binary view a 1 sets a single bit in hex view a 1 sets 4 bits Bin Hex Symbol Configuration Dialog Box
351. igger Output Agilent J BERT N4903B High Performance Serial BERT Setting up External Instrument s 2 Ifthe Half Rate Clock check box is selected the multiplier of the clock multiplier will be set automatically to 2 and the frequency sent to the clock multiplier is Data Rate 2 If the Half Rate Clock check box is not selected the multiplier of the clock multiplier will be set automatically to 1 and Data Rate itself is sent to the box Termination You can use the Termination check box to toggle between true differential DC coupling and center tapped termination In center tapped mode you can specify the termination voltage The Termination fields are used to set the termination voltage Multiplexer N4876A Window The N4876A multiplexer doubles the pattern generator data rate by multiplexing two pattern generator channels It extends the variable generator data rate of J BERT N4903B up to 28 4 Gb s The Multiplexer N4876A window contains the following elements BER 0 000 Remote f 2 Jitter The f 2 jitter defines how much mismatch the period of odd even bits at the output of the MUX have If f 2 jitter is at 0 then the observed output period at the multiplexer is half the input period Agilent J BERT N4903B High Performance Serial BERT 55 2 Setting up External Instrument s 56 DataOut Tab M8061A Window The M8061A is a 2 1 multiplexer to characterize serial interfaces of up to 28 4 Gb s wit
352. ile as described in Opening Existing Patterns on page 71 or create a new one as described in Creating New Patterns on page 71 To change the pattern length or description or to switch between standard and alternating patterns press the Properties button The Pattern Properties dialog box appears Define the pattern length whether it is alternate or standard and give it a description Click OK when finished Edit the pattern as required For the function of the different tool buttons see Edit Pattern Window on page 76 Saving Patterns To save the current user pattern In the toolbar click the Save icon Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 If the pattern has already been saved earlier the saved file is updated Otherwise the Windows standard Save As dialog box opens where you can define the file s path and name The available file formats to save the file in are Pattern File ptrn in binary 8 bits byte hex 4 bits byte 01 1 bit byte or symbol format e ASCII Pattern Files txt See Pattern File Specifications on page 66 for descriptions of these file types If you want to save the pattern file in one of the 12 memory slots use the file names from UPAT1 ptrn through UPAT12 ptrn respectively Loading Patterns to the Pattern Generator and Error Detector The recommended way to load the data patterns to the pattern generator and error detector is to use
353. ile the BLUE background denotes the pattern for Aux Data Out channel If any of the channel has anything beside pattern file the corresponding channel NOTE NICA would not be visible in the canvas The status bar below the editor canvas provides the following information from left to right Status Offline indicating that this is the display of the editor not the pattern generator nor the error detector Current Position in bits of the cursor in the pattern Editing mode Overwrite Dynamic Insert Insert Display mode Bin Hex Symbol Mark Density of the pattern s Total Length of the pattern in bits Error Indicators Disparity Error D This indicator turns red if there is any disparity error in the pattern In case of alternate patterns the indicator will indicate the disparity error of individual pattern trace Loop Disparity Error LD This indicator turns red if the pattern causes any disparity errors when being looped The available functions to change the view and to modify the pattern are described in Edit Pattern Window for Sequencer Pattern Editor on page 110 Create New Pattern The Create New Pattern dialog is shown in the figure below Agilent J BERT N4903B High Performance Serial BERT 115 J Setting up Patterns 116 Create New Pattern Select Trace Data Out C Aux Data Out Description New Pattern File name cAn4903blpatterridemotusb 3 0 upat_seq_ Bro
354. ill also affect the frontpanel keys i e M ega and m illi will be exchanged Enabling the LAN Connectors You can enable the LAN connection on your Serial BERT This will allow you to connect to network printers and remote servers with access to shared folders and files The Serial BERT has two RJ 45 LAN connectors on the rear panel They support 10 Base T and 100 Base T Ethernet networks using TCP IP and other Microsoft supported networking protocols The Serial BERT uses Microsoft Windows XP Professional or Microsoft amp Windows XP Professional for Embedded Systems For detailed instructions on how to enable the LAN connectors please refer to the Microsoft Windows XP Professional manual Because your network settings are unique to your IT infrastructure Agilent Technologies will not be able to assist you with connecting your instrument to your network Please contact your network administrator or IT department for assistance For more information refer to the MS Windows resource kit available from Microsoft that is appropriate for your computer system Do not connect the Serial BERT to a network that is configured to automatically install software on network devices Installing or overwriting files on the Serial BERT computer system may impact the operation of the instrument Please contact your network administrator or IT department to find out whether you have this type of network Agilent J BERT N4903B High Performance Ser
355. ing points at the edges This can greatly improve the results without dramatically increasing the duration of the test The following illustration shows a diagram with default resolution and no edge resolution optimization 100e Q 3 Traces 453 Points Ul 2 500 ns All Errors 1 00e 1 1 00e 2 1 00e 3 1 00e 4 1 00e 5 1 00e 6 1 00e 7 Logarithmic 4 Ul 2 04 Ul 2 28 Ul 2 52 Ul 2 76 Ul 3 00 UI 3 24 UI 3 48 UI Absolute 1 92 UI 216 Ul 2 40 UI 2 64 UI 2 88 UI 312Ul 3 36 UI 3 3 72 Ul 60 Ul The following illustration shows a diagram with high resolution and edge resolution optimization For the areas of the diagram with edges additional sampling points are taken resulting in a much better display of the signal shape Agilent J BERT N4903B High Performance Serial BERT 253 6 Advanced Analysis 254 NOTE 100e 0 3 Traces 3888 Points Ul 2 500 ns All Errors 1 00e 1 1 00e 2 1 00e 3 1 00e 4 1 00e 5 1 00e 6 1 00e 7 Logarithmic 21 Absolute 1 920 I 2 28 UI 2 52 Ul 2 76 Ul 3 00 UI 3 24 UI 3 48 Ul 246 Ul 2 40 UI 2 64 Ul 2 88 UI 3412UI 3 36 UI 3 60 04 Ul Fast Total Jitter at BER This enables the Fast Total Jitter measurement Before enabling this measurement you need to know the BER floor of the device and to specify a BER threshold that is above that floor For details see Explanation of the Fast Total Jitter Measurement on page 246 Pass Fail Tab The Pass Fail tab of the Properties
356. ing Problems 522 If you cannot see a clear eye diagram then check if the M8061A is connected properly to the J BERT N4903B The M8061A must be connected to Aux Data Out Data Out and Aux CIk Out to work properly If the connection to the J BERT N4903B is setup correctly then check the output signal of the J BERT N4903B at the Data Out Aux Data Out both should show a eye diagram with an offset of 70 mV and a amplitude of 800mV Check the output signal of Aux CIk Out it should give a clock signal of set set data rate with an offset of OV and an amplitude of 700 mV if one of the used J BERT N4903B outputs is not providing the expected signal then the J BERT N4903B is defective If the J BERT N4903B is providing the correct signals then connect the M8061A again and run the adjustment procedure from the Config screen Check the output of the M8061A again after the adjustment has been done If the adjustement does not fix the problems then the M8061A is defective Check the Output of the M8061A using the Error Detector If you have connected the M8061A to a J BERT N4903B and have no oscilloscope at hand you can use the error detector for checking the output of the M8061A The approach shown here is checking the output signal using the error detector of the J BERT N4903B nthe ED Clock Setup window set the clock as internal CDR Set the J BERT s PG and ED patterns to PRBS 15 1 Set the M8061A Output to 500 mV Set the J BE
357. ing it a 1536 bit pattern How the Serial BERT Handles The same behavior of repeating a pattern to fill up complete blocks also applies to Alternating Patterns alternating patterns In this case each half of the pattern is loaded into a separate location in the RAM When an alternating pattern is generated first all blocks of pattern A are sent then all blocks of pattern B are sent and so on Agilent J BERT N4903B High Performance Serial BERT 69 J Setting up Patterns CAUTION 70 A0 Al A2 BO B1 B2 AO A1 A2 Implications of Using Memory Based Patterns There are several implications to the way Serial BERT handles memory based patterns a few of which are Ifyou set up an alternating pattern where pattern B is a replica of pattern A except for the addition of one error and the pattern is an odd number of bits long the error will be repeated 512 times when B is sent because pattern B is sent 512 times before switching back to A Ifyou want to senda periodic pattern you only have to define a single repetition For example if you want to send alternating 1 s and 0 s you only have to define the first two bits This pattern is automatically loaded 256 times to fill up a complete block of RAM Ifyou set up a trigger to be generated whenever a pattern is sent the trigger is sent whenever all blocks that sto
358. ing parameters Termination Voltage Used to set Electrical Idle termination voltage when externally controlled Threshold Used to set Electrical Idle threshold when externally controlled Table 4 M8061A Parameters and Range Parameters Range Amplitude Based on amplitude range selected High 975 mV to 3 3 V Low 1 V to 3 275 V Offset 975 mV to 3 275 V Cross Over 30 to 70 F 2 Jitter 20 ps to 20 ps Termination Voltage 1 V to 3 3 V PreCursor2 6 02 dB to 6 02 dB PreCursor1 12 04 dB to 12 04 dB PostCursor1 20 dB to 20 dB PostCursor2 12 04 dB to 12 04 dB Agilent J BERT N4903B High Performance Serial BERT Setting up External Instrument s 2 Table 4 M8061A Parameters and Range Parameters Range PostCursor3 12 04 dB to 12 04 dB PostCursor4 6 02 dB to 6 02 dB PostCursorb 6 02 dB to 6 02 dB Threshold Voltage 1Vto3V Termination Voltage 1Vto3V Agilent J BERT N4903B High Performance Serial BERT 59 2 Setting up External Instrument s 60 Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns Setting up Patterns Concepts NOTE Hardware Generated Patterns Memory Based Patterns The purpose of data patterns is to simulate the type of data that your device might receive in the real world Different patterns present different data loads to your device which can cause variations in the bit error ratio A bit pattern is sent from the pattern generator to your device At the same time the
359. ing the Instrument 10 Printing Procedures NOTE This section provides information on setting up and using a printer Connecting a Printer to the Instrument The instrument supports parallel and serial printers as well as USB and LAN connected printers To connect your printer to the instrument 1 Turn off power to the instrument and printer before connecting 2 Connectthe printer to the instrument with the cable that came with your printer Once you have connected the printer to the instrument you will need to add the printer driver before you can use the printer Adding a Printer Use the following procedure to add a printer to the instrument 1 On the File menu click Print Setup then Add Printer This opens the Add Printer Wizard dialog box 2 Click My Computer for a local printer installation or click Network printer server for a printer connected to another server Before you can add a network printer you must connect to a Local Area Network LAN In addition your login must provide you with access to the network printers that you wish to add Please contact your network administrator or IT department for assistance 3 Follow the instructions in the Add Printer Wizard until you have selected a local printer and configured the printer port 4 f your printer is not in the list of available printers you will need a copy of the printer driver Use the USB Stick to copy the printer driver into the local drive of y
360. ing the Vtop and Vbase values of the displayed waveform All the threshold coordinates of the mask will be relative to this voltage This is the default mode for mask alignment Eye Boundaries The Eye Boundary mode aligns the mask using the 1 level and 0 level of the waveform The 1 level and the 0 level are calculated between the Eye Boundary 1 and Eye Boundary 2 Set Mask Margins Margins on Mask Margins allows the user to enlarge or shrink a mask in order to see the margin size and how much of it is violated The maximum and minimum margins are defined within the mask files Set Mask Scaling With mask scaling the mask can be manually enlarged shrunk and moved around With the logic levels track selected the mask regions are bound together and move together with a fixed amplitude Position pos gives the timing displacement and dT gives the delta time offset Set Mask Run Control Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 The Mask Run Control aligns the mask switches it on starts displaying the number of waveforms and the number of mask violations per mask region and the worst BER per Mask region Exit stops the computation hides the mask and the violations Automated Eye Parameter Measurement Automated Eye Parameter Measurement Table 31 S No 1 Agilent J BERT N4903B High Performance Serial BERT Parameter Rise time Fall time Sample Count Description
361. ing used and has to be repeated whenever one of these are changed in the setup The Input Timing setting that is determined during this calibration is only valid for all data rates when using the N4915A 010 cable kit For all other cables this calibration has to be done for each data rate that has to be set using the N4916B De Emphasis Signal Converter Maintaining a list of proper Input Timing Settings to data rates and use ofthe correct setting for a dedicated data rate is up to the user when not using the cable kit Loading Calibration Settings Follow the given steps to load the calibration setting 1 From the Navigation Menu select External Instrument s and then click Config sub menu The Config window shows a list of all instruments connected to the N4903B 2 Select the external instrument from the given list For example here you will select N4916B 3 Use the drop down list to provide the Input Timing Setup Input Timing Setup 2 Load Calibration Setting 4 Click Load Calibration Settings to continue A message will pop up which warns you to save the instrument settings before continuing the calibration Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 N4916B The instrument settings will be lost once you start the calibration process It is recommended to save the instrument settings Do you like to start calibration WARNING The instrument settings will be lost once you start
362. insertion of pattern B via external signal is 512 bits 2 Click Alt Pattern and Aux In menu item from the PG Setup submenu to open the Alternate Pattern Control dialog box A Alt Pattern and Aux In E Setup Agilent J BERT N4903B High Performance Serial BERT 163 4 Setting up the Pattern Generator 164 3 4 In the Aux In section define whether the Serial BERT should insert pattern B in Level Sensitive or Edge Sensitive mode See Aux In on page 165 for more information on the different available modes Click OK to close the dialog box Suppressing the Outputs via External Signal You can configure the Serial BERT to suppress the output according to an external signal 1 4 Connect an external instrument to the Aux In port The signals received at this port must be TTL compatible The granularity for the insertion of pattern B via external signal is 512 bits Click Alt Pattern and Aux In menu item from the PG Setup submenu to open the Alternate Pattern Control dialog box A Alt Pattern and Aux In E Setup In the Aux In section select Output Blanking See Aux In on page 165 for more information on the different available modes Click OK to close the dialog box Pattern Alternation Reference NOTE The Serial BERT offers the following options for alternating between pattern A and pattern B These functions are only available if Alternate Pattern is selected in the Pattern Properties dialog box
363. int Check that the synchronization and alignment were successful None of the error indicators should show red The resulting BER should be zero Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 How to Demonstrate the Error Location Capture Measurement To demonstrate the Error Location Capture measurement 1 Switch to the Pattern panel and press the Open button Load a Fiber Channel random data pattern from the demo patterns Demo gt Fiber Channel gt RPAT ptrn Press the Properties button and set the Pattern Type to Alternate To display both halves of the alternate pattern press the Alt pat view button Ensure also that the error detector tracks the pattern generator 2 Insert two errors into pattern B To do so edit bits 80 and 100 of pattern B in overwrite mode 3 Write the modified pattern to the pattern generator and the error detector by pressing the ToPG amp ED button 4 Forthe pattern generator setup you need to specify the Alternate Pattern and Aux In Set the Internal Alternate Pattern Mode to Single Shot B and disable Aux In 5 Set up the error detector so that the Clock Setup is set to Clock Data Recovery to get the error detector s clock from the incoming data stream 6 Switch to the Analysis panel and then press the Error Location Capture icon 7 Press the Start button to execute the measurement Status information is displayed in the lower section of the display 8 Press Insert B
364. ints Press Apply to accept the selected Auto Frequencies setting and switch to the Frequency tab In Manual mode you can edit the automatically calculated frequencies If you do so the Auto Frequencies checkbox is automatically deselected Enter the number of test frequencies Number of Steps either on this tab or on the Frequency tab Description This is the creator s description of the selected standard Standards provided by Agilent are described by the following items TIP Use the scroll bar if not all information is displayed Document that describes the standard in detail The document name usually indicates the date or revision of the standard Recommended settings for the test pattern like Bit Rate range Pattern types and voltage level ranges Voltage Level from Standard This is the peak to peak differential input level range for the DUT as given in the standard Voltage Level at PG recom This is the recommended output level at the pattern generator to achieve the minimum voltage Level from Standard This recommendation depends on the simulated intersymbol interference ISI The NOTE i minimum voltage Level from Standard is only met if the trace to simulate the ISI is selected as in the description item Additional Jitter Other settings of the ISI can change the DUT s input level up to four times Agilent J BERT N4903B High Performance Serial BERT 463 8 Jitter Tolerance Tests NOTE Auto mod
365. io mode supports the bit rate up to 10 35 Gbit s However if you want to use bit rate greater than 10 35 Gbit s then switch to 8B 10B Symbol Comparison or Bit Comparison without PCle3 SKPOS error ratio mode and then go to Bit Comparison mode Error Ratio Procedures This section explains how to set up an Error Ratio Setting up Error Ratio Follow the below steps to set up an Error Ratio 1 Load the appropriate patterns in Pattern Generator and Error Detector For more information see User Patterns Procedures on page 71 NOTE Patterns with Invalid symbol s when loaded in ED display an error message However if the same pattern in loaded in PG the 8b10b Error indicator turns red and no error message is displayed Patterns with Illegal runing disparity when loaded in ED will turn the respective symbol red when current running disparity is same as previous running disparity If same pattern is loaded in PG it will turn 8b10b Error indicator red 2 f you want to test your device in CDR mode then set the Clock Data Recovery CDR from the Clock Setup window For more information see Clock Setup Procedures on page 178 3 From the main menu choose ED Setup and then click Error Ratio from the list of menu items E Error Ratio This will launch Error Ratio window Agilent J BERT N4903B High Performance Serial BERT 185 5 Setting up the Error Detector Mia cBER 0 000 K28 5 Align K28 5 D10 2 D10
366. ion can be selected for example if the signal delay is very unstable and you want to avoid that the re synchronization process affects the measurement results Burst Sync Mode Click this button to enable the Burst Sync Mode Burst mode operates as follows For each burst of data BER measurements are only made after the error detector synchronizes to the incoming pattern The signal at the Gate In port controls the timing of synchronization and error counting for each burst The burst sync mode is disabled in 8B 10B Comparison Bit Comparison without PCle3 SKPOS and Bit Comparison without USB3 1 SKPOS error ratio modes The Sync Threshold defines when the error detector assumes the patterns to be synchronized The Serial BERT shifts the received pattern bitwise until the measured BER is below this threshold Mask Resync Errors The masking feature is available in all error ratio modes Bit Comparison 8B 10B Symbol Comparison Bit Comparison without PCle3 SKPOS and Bit Comparison without USB3 1 SKPOS modes Some DUTs may drop bits respectively symbols due to high stress during receiver test For example SATA drops complete symbols if the received symbol is illegal because of a bit error Other cases of unexpected more data are added symbols because of a bit error in a filler symbol so that the filler won t be recognized anymore All cases of dropped or additional bits causes the error detector to lose synchronization
367. iplexer with De emphasis The Agilent s M8061A is a 2 1 multiplexer to characterize serial interfaces of up to 28 4 Gb s with optional de emphasis to extend the rate of J BERT N4903B pattern generator Data rates up to 28 4 Gb s are used for Optical transceivers such as 100GBASE LR4 SR4 and ER4 32G Fibre Channel SERDES and chip to chip interfaces such as OIF CEI Backplanes cables such as 100GBASE KR4 CR4 Next generation computer buses such as PCle4 Following are the features and benefits of M8061A Agilent J BERT N4903B High Performance Serial BERT 17 1 Planning the Test 2 slot AXle module that can be controlled from J BERT N4903B user interface via USB Expands data rate of J BERT N4903B pattern generator up to 28 4 Gb s Integrated and calibrated 4 tap de emphasis expandable to 8 taps Internal superposition of interference for common mode and differential mode Transparent to jitter generated by J BERT Clock 2 jitter can be added Supports electrical idle The Agilent M8061A is available with the following two options Option M8061A 004 provides 4 tab de emphasis feature Option M8061A 008 provides extension to 8 tab de emphasis feature These features are only operational when the respective options are installed during manufacturing Navigating the Serial BERT GUI Navigating the Serial BERT GUI Concepts The Serial BERT GUI provides a central launching point for its various controls
368. is applied and the Logic Level is changed to CUSTOM See also Why Incorrect Terminations Could Damage Your Device on page 129 Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 The following table lists the characteristics generally associated with some logic families All values are nominal Table 18 ECL 0 95 1 7 750 2 SCFL 0 0 9 900 0 LVPECL 2 35 1 6 750 1 3 LVDS 1 425 1 075 350 1 250 CML 0 0 4 400 0 Terms ECL Emitter Coupled Logic SCFL Source Coupled FET Logic LVPECL Low Voltage Positive Emitter Coupled Logic PECL Lite e LVDS Low Voltage Differential Signaling CML Current Mode Logic Why Incorrect Terminations Could Damage Your Device Choosing wrong terminations may cause your device to output voltage levels that are not as expected It may also cause excessive current or current flow in the wrong direction which can damage your device Agilent J BERT N4903B High Performance Serial BERT 129 4 Setting up the Pattern Generator 130 Note that an internal protection circuit becomes active if the termination voltage is wrongly adjusted The protection circuit sets the output voltages to safe levels typically Vhi Vlo Vterm externally measured termination voltage AC Coupling and Bias Tees The pattern generator s outputs are normally DC coupled even when AC termination is selected For this reason extreme caution must be taken
369. is controlled by the Serial BERT via USB 1 Connect mains power to the N4876A and switch the instrument on When you connect the N4876A via USB it is important that the N4876A is switched on Otherwise it may not be identified correctly by the operating system 2 Connect the USB cable between the USB ports at the rear sides of the N4876A and the Serial BERT You can use any rear port of the Serial BERT 3 The Found New Hardware Wizard will pop up and ask whether he should connect to Windows Update 4 Click No not this time Found New Hardware Wizard Welcome to the Found New Hardware Wizard Windows will search for current and updated software by looking on your computer on the hardware installation CD or on the Windows Update Web site with your permission Read our privacy policy Can Windows connect to Windows Update to search for software Yes this time only O ery time connect a device Click Next to continue 5 Click Next and let the Wizard locate the appropriate driver The USB driver is part of the Agilent IO Libraries Suite 6 Step through the Wizard by clicking Next The Wizard will finally set up a USB device named USB Test and Measurement Device Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 Found New Hardware Wizard Completing the Found New Hardware Wizard The wizard has finished installing the software for USB Test and Measurement De
370. is formulated in its own language checked by the Sequence Editor The SequenceExpression specifies the sequence start and break conditions the blocks their contents and trigger output theloops You can inspect the contents of the SequenceExpression in the Properties dialog of the Sequence Editor The SequenceExpression uses the following keywords Version optional Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 Description optional Start optional Block repeated for each block numbered Loop repeated for every loop not numbered Example of a SequenceExpression Version 1 0 Start IMM Block 1 PRBS11 1024 TrigOn Block 2 C lt Instrument Model gt Pattern Upat 0 ptrn Block 3 P0 512 TrigOff Loop B1 B1 2 A sequence is in common for both channels Data Out Channel 1 and Aux Data Out Channel 2 However the sequence for both channels will have a different version and pattern in a block specification Example of a SequenceExpression for both channels Version 1 1 Start IMM Block 1 P0 P1 311040 TrigOn Block 2 c n4903b pattern sdh_stm 16_hp bip_alt ptrn B c n4903b pattern sdh_stm 16_hp bip_alt ptrn A TrigOff Loop B1 B1 INF Start Options Default is IMM immediately after downloading the sequence to the pattern generator Other options are e AuxlnHi AuxInLo Sequence starts when Auxiliary Input is high or low e AuxlnRising AuxI
371. it of any arbitrary data pattern They may be generated by the Serial BERT software or defined in a pattern file and loaded into the memory The Serial BERT reads the memory bit by bit and generates the output accordingly The following topics explain how the Serial BERT works with user patterns and also explain how user patterns are generated The way software generated patterns distorted PRBS PRBN are output is identical to that of user patterns The patterns are written to the memory and the output is generated from there Some of the information in these topics covers how the patterns are actually generated and thus also applies to software generated patterns When to Test with User Patterns Creating or editing patterns is necessary for quality testing because different patterns present different data loads to the device This can cause variations in the bit error ratio With user patterns you can define exactly the pattern used in the test Supported Pattern Types The Serial BERT can import the following pattern types HP Agilent 71612A series patterns dat file type Patterns binary hex symbol 01 saved as ASCII files by the Serial BERT or the Agilent 86130 ptrn and txt file type These are described in detail in Pattern File Specifications on page 66 The Serial BERT can save files as either dat ptrn or txt files Downloading text format pattern into your instrument may take significantly lo
372. ition Comment 2005 04 1204 28 50 passed 2 23 31 PRBS 92 488 Gb s RJ 18 171 mul Modify Comment Adaptation of frequency points margin added Agilent J BERT N4903B High Performance Serial BERT 453 8 Jitter Tolerance Tests Jitter Tolerance Compliance Procedures To access the Jitter Tolerance Compliance measurement click Jitter Compliance from the Jitter submenu te Tolerance Compliance To set up a Jitter Tolerance Compliance test press the Properties button Properties j This opens the Jitter Tolerance Properties dialog Setting the Standard Select either a predefined standard or load a standard from a text file 1 Switch to the Standard tab Jitter Tolerance Properties Standard Frequency BER CEI 6 Gb s Short Reach M Margin 0 0 96 iv Auto Frequencies Nr of Steps 20 Description log Frequency MHz Document X OIF CEI 02 0 4 976 5 375 Gb s PRBS31 recom Level from Standard Vdiff pp 125 750 mV Level at PG recom Vampt 290 mV Add Jitter RJ 7 10 7 mUl rms BUJ 150 mUl pp Cancel Apply Help A preview and a short description of the selected standard are displayed gt lin epmiduas 5o 2 To load a different standard open the list 454 Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Jitter Tolerance Properties Standard Frequency BER cei 6 Gb s Short Reach CEI 6 Gb s Short Reach
373. itor are described below Edit Pattern Window The Edit Pattern window provides the following convenient pattern editing functions Table 9 Icon Name Description Toolbar Ctrl Click this icon to switch the toolbar from two rows to a single row appearance leaving more space for the pattern data k New Click this icon to create a new pattern See Creating New Patterns on page 71 for details r Open Click this icon to open a pattern from a file See Opening Existing Patterns on page 71 for details B5 Save Click this icon to save the current pattern If it was not previously saved a Save As dialog box opens for you to specify name and path for the file Agilent J BERT N4903B High Performance Serial BERT Table 9 Setting up Patterns 3 Icon Name k Export Ec To PG 10110 To PG 10110 iw To ED To ED 10510 Capture ED Capture FED To PGED 10110 To PGED P Cut Cut Copy Paste Sopy Delete e Undo Paste x Delete Lel Undo EDS Go To Agilent J BERT N4903B High Performance Serial BERT Description Click this icon to to open Export Trace dialog See Export Trace on page 118 for details Click this icon to send the pattern from the editor to the pattern generator Click this icon to send the pattern from the editor to the error detector Click this icon for Pattern Capture See Pattern Capture Dialog Box on page 83 for details Click this
374. k the Power Supply Ensure that the power cord is connected Check for the steady green STATUS light that appears on the ESM of the chassis indicating a power ready status of the chassis which mounts the M8061A module Check for the status LEDs present Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 on the front panel of M8061A For more information on the functionality of these LEDs refer to the section M8061A Status LEDs on page 522 Check the USB Connections Ensure that the USB cable is properly connected between the USB ports of the M8061A and one of the USB ports of the J BERT N4903B Use the matched USB cable to establish the USB connection The mini USB port is available at the front side of the AXI frame which is mounting the M8061A module Use the USB port USB 2 0 that is available in the rear side of the J BERT N4903B For more information on connecting the M8061A with J BERT N4903B via USB refer to the section Connecting the M8061A with J BERT N4903B via USB on page 549 NOTE Agilent recommends you to establish a USB connection between the USB ports of M8061A and J BERT N4903B via a powered USB 2 0 hub to prevent the USB port shutdown due to excessive power drawn by the USB devices connected at the J BERT N4903B USB ports Ensure that the User Software can Access the M8061A 1 Start the J BERT firmware and the user interface 2 Press the green Preset key close to the front USB connector of t
375. l characteristics of the fast Fourier transform please refer to the standard literature For details see the following documents Frederic J Harris On the use of Windows in Harmonic Analysis with the Discrete Fourier Transform Proceedings of the IEEE Vol 66 January 1978 The Fundamentals of Signal Analysis Agilent Application Note 243 Publ No 5952 8898E Fibre Channel Methodologies for Jitter Specifications National Committee for Information Technology Standardization NCITS T11 2 Project 1230 Rev 10 June 1999 e Yi Cai Bernd Laquai Kent Luchman Jitter Testing for Gigabit Serial Communication Transceivers IEEE Design and Test of Computers Jan Feb 2002 The error record contains a bipolar rectangular signal If such a signal is periodic you can expect a spectrum as illustrated below Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 T 1 IETS gt Frequency Time 1 T The fundamental frequency and its harmonics appear Such spectra have been measured with the Spectral Jitter measurement 1 Trace 65535 Points Linear 3 00 MHz 5 00 MHz 7 00 MHz 9 00 MHz 11 00 MHz 2 00 MHz 4 00 MHz 5 00 MHz 8 00 MHz 10 00 MHz 12 00 MHz A logarithmic power scale shows the details 1 Trace 65535 Points 10 dB 20 dB 30 dB 40 dB 50 dB 60 dB 70 dB 80 dB k Threshold 40 dB 90 dB 100 dB Linear 3 00 MHz 5 00 MHz 7 00 MHz 9 00 MHz
376. le of periodic jitter 1 is shown to specify the jitter components P1 7 10 0000 MHz To specify the jitter components 1 Activate a jitter source press the corresponding button PJA 2 Set the most commonly used parameter typically Amplitude and Frequency directly 10 0000 MHz 3 Press Edit button to access all parameters of the corresponding jitter type This opens a window which allows you to set all parameters available for the given jitter type Agilent J BERT N4903B High Performance Serial BERT 409 8 Jitter Tolerance Tests 410 NOTE NOTE A parameter that must be set for any jitter type is Amplitude This defines the timely width of the jitter component 4 Observe the consumption of delay line capacity indicated by the delay line markers Ensure that the total jitter remains within the margins of the chosen delay line 5 Return to the overview display by pressing either the Close button the Edit button of this jitter source or the Edit button of the next jitter source you want to change 6 Proceed with the next jitter type you wish to add When you activate an External jitter source you will find that this apparently consumes the whole free capacity of the delay line at the right hand side of the red bar But this is just an indication You can still add or change internal jitter components The external jitter range is automatically updated when you add change or switch off one of the
377. lect the error ratio as 8B 10B Symbol Comparison from the Error Detector Error Ratio s window For more details see Symbol Lock Indicator on page 192 and 8b10b Error Indicator on page 192 For troubleshooting information see Setup Problems Concepts on page 469 The LEDs on these indicators turns green if any of Jitter ISI or SSC is enabled the outputs are enabled yellow if output protection is active on any of Data Clock Aux Data or Trigger Ref Clock Output This indicator turns blue when the instrument is under remote control Remote The instrument does not enter the remote lock state if the N4917A Optical Receiver Stress Test Software is running on the instrument This is because the N4917A ORST Software also remotely controls the J BERT The Show Error Messages button displays the error messages It uses the color codes to indicate the following possible states No error messages The button will be in disable state New unread errors The triangle icon on the button will turn red Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 No new errors The triangle icon on the button will remain yellow J If an error occurs the detailed error message is displayed for a few seconds on the top of the status indicator area You can click the message to make it disappear immediately 80814 The value 0Hz exceeds the physical limits of the Sinusoidal Jitter Frequency
378. lements of the pattern editor and all dialog boxes that can be opened from the sequencer pattern editor are described below Edit Pattern Window for Sequencer Pattern Editor The Edit Pattern window for sequencer pattern editor provides the following convenient pattern editing functions Table 15 Icon Name Description Lx Toolbar Ctrl Click this icon to switch the toolbar from two rows to a single row appearance leaving more space for the pattern data k Export Trace Click this icon to open Export Trace dialog See Export Trace on page 118 for details Agilent J BERT N4903B High Performance Serial BERT Table 15 Setting up Patterns 3 Icon Name v Apply m Discard X Cut Cut Copy Paste Sopy Delete 3 Undo Paste x Delete Le undo Properties Agilent J BERT N4903B High Performance Serial BERT Description Click this icon to apply the pattern to the hardware in case the sequence is already active or simply update the corresponding pattern file s if the sequence is not yet active Applying closes the editor and brings the user back to the sequence editor Click this icon to close the editor without updating anything and bringing the user back to the sequence editor These functions follow Microsoft Windows conventions Click this icon to view change the properties of the current pattern s See Pattern Properties on page 119 for details 111 3 Settin
379. ler3 setting 1 2 or4 symbols Filler 4 setting 1 2 or4 symbols Symbol lock llegal symbol or disparity erro i Pattern Pointer Patt z Automatic P n Seca can cay Se Pas at See 300000 Gb s fromEdtor Enor FINE Loss Los Lock Enor Agilent J BERT N4903B High Performance Serial BERT 183 5 Setting up the Error Detector 184 NOTE Understanding the Error Free Receiving in PCI Express 3 0 128B 130B Encoded Data Comparison When RX detects incoming data correctly the pattern looped back has the same content except the length of the Skip Ordered Set SKPOS primitives The change in the SKPOS length by DUT is to compensate for the speed differences of the clock domains DUT Slave receives SKP Ordered Set of 16 symbol and if clock compensation is required DUT must add or remove multiple of four skip symbols per Skip Ordered Set Hence received Skip Ordered Set can be 8 12 16 20 or 24 Symbols The SKP END Symbol indicates the last four Symbols of SKP Ordered Set The three Symbols received after loopback following the SKP END Symbol can be different Although bit wise different in content and number of bits in order to report this different looped back bit stream as error free the ED has to recognize the different length SKPOS and ignore these symbols The Bit Comparison without PCle3 SKPOS option is only applicable for PCle3 testing PCle3 is an abbreviation for PCI Expre
380. listed Bit Count The number of bits considered for the accumulation period is displayed here This may not equal the total number of bits sent from the PG Bit Error Ratio The ratio of the error count to the bit count is displayed here to 4 significant digits Bit Error Count The number of errored bits measured during the accumulation period is displayed here Errored 1 s Ratio The ratio of the errored 1 s count to the bit count is displayed here to 4 significant digits Errored 1 s Count The number of logic 1 s measured as logic 0 s during the accumulation period is displayed here Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 Errored 0 s Ratio The ratio of the errored 0 s count to the bit count is displayed here to 4 significant digits Errored 0 s Count The number of logic 0 s measured as logic 1 s during the accumulation period is displayed here Auto Resync Counter The total number of Auto Re Sync Count received in a time interval is displayed here G 821 Measurements Availability The ratio of the available seconds to the total gating period expressed as a percentage e Unavailability The error ratio is calculated over 1 second timed intervals during the gating period An unavailable period begins when the error ratio is worse than 1 x 103 for 10 consecutive seconds These 10 seconds are considered part of the unavailable time The unavailable period e
381. llow Data Whi M Termination Vampt Vof Logic Level Custom x W Electrical Idle Output Blanking Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 The parameters can be changed to configure the output according to the DUT s termination requirements Be aware that the changes are not applied immediately but only when the output protection circuit button is clicked Output Protection on M8061A Output s The output protection state on M8061A output s occurs when the amplifier detects an overload condition which is caused by the externally connected circuitry DUT In the output protection state the Auto Range Amplitude Range Amplitude High Low Offset Coupling Termination Model Termination Voltage and CMI state are re programmed to safe values The GUI shows an error message listing the effected outputs and highlights the Outputs On status indicator In the M8061A window the parameters that currently do not match the data out settings are highlighted in yellow RES BER 0 000 The parameters can be changed to configure the output according to the DUT s termination requirements The changes are applied only when the output protection circuit button is clicked Agilent J BERT N4903B High Performance Serial BERT 125 4 Setting up the Pattern Generator 126 Incorrect Termination Detection on M8061A Output s The incorrect termination detection state on M8061A
382. lly Inserting Errors on page 157 e Setting Up a Periodical Output on page 162 Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 Input and Output Ports Reference The pattern generator produces clock and data outputs that serve as frequency reference and device stimulus for the device under test NOTE The Serial BERT will not allow you to adjust a voltage beyond its limits The limit is determined by the Serial BERT s internal hardware If a limit is encountered the Serial BERT sends a message to the status bar For more information about voltage limits see the pattern generator s technical specifications NOTE The parameters listed in the Pattern Generator s Data Out window will be disabled if you select M8061A from the External Instrument s Vhi This text field allows manual entry of the logic high voltage level and displays the current value To modify the value click inside the text field and either Enter the desired value directly with the numeric keyboard Use the knob by the numeric keyboard to fine tune the value See Understanding the Output Level Parameters on page 126 for information about how the Serial BERT modifies the output levels Vof This text field allows manual entry of the voltage level halfway between logic high and logic low the offset and displays the current value To modify the value click inside the text field and either Enter the desired va
383. lly finished the display is automatically switched to the current run The current state is displayed in the title bar of the pattern grid First Error Jumps to the first errored bit in the pattern Prev Error Jumps to the previous errored bit in the pattern Next Error Jumps to the next errored bit in the pattern Last Error Jumps to the last errored bit in the pattern Explanation of the Results The results are shown in various ways The graphic view highlights the located bit errors The currently selected bit is highlighted in red all other errored bits are written in red The status bar below the graphic view provides the following information Status The status indicates if the measurement is running or has ended If the measurement has ended it indicates if the measurement ended successfully or was aborted Position The position of the current errored bit is shown Hex Bin This field indicates how the data is shown bin or hex format Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Length This field indicates the length of the captured data Note that the value here does not equal the length of the pattern Compare Pattern File The captured data is saved as an alternating pattern Pattern A contains the expected data Pattern B contains the errored data Os if the expected bits were also received 1s for errored bits To calculate the captured pattern XOR the
384. lock source by clicking the respective button If the source frequency varies you will be alerted Ifyou need SJ and SSC simultaneously use an external modulated SSC signal from 6 75Gb s to 12 5Gb s and use the J BERTs SJ for additional jitter generation To generate data rates below 6 75Gb s the external clock must be divided Select a clock rate For Internal Clock Source you can either select a clock rate from the Presets list or enter a clock rate in the Value and Units field For 10 MHz Ref Clock Source the clock rate connected to the 10 MHz Ref In port must be 10 MHz The selected clock rate applies to the pattern generator This is also the clock rate generated at the pattern generator s Clock Out port If the error detector receives its clock from the pattern generator it runs at the same clock rate Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 Bit Rate Reference The Bit Rate Setup window contains the following elements Clock Source Clock source can be one of the following Internal The clock generator uses the internal oscillator 10 MHz Ref The clock generator uses the 10 MHz Reference that must be connected to the 10 MHz Ref Input External The pattern generator uses the external clock that must be connected to the Clock Input You can choose between Automatic and Manual Mode on page 143 External PLL The pattern generator uses the
385. look like this 1 Trace 316 Points 10V 900 0 mv 800 0 mv 700 0 mv 500 0 mv 500 0 mv 400 0 mv 300 0 mv 200 0 mv 100 0 mv 00v Linear 3 000e 0 9 000e 0 1 500e 1 24100e 1 2 700e 1 3 300e 1 3 900e 1 4 500e 1 5100e 1 5 700e 1 6 300e 1 0 000e 0 5 000e 0 1200e 1 1 800e 1 2400e 1 3 000e 1 3 600e 1 4200e 1 4 800e 1 5 400e 1 5 000e 1 The measurement will calculate the Level and Standard Deviation results from all data points The marker allows you to measure the u Mu and o Sigma of the individual peaks DUT Output Timing Jitter DUT Output Timing Jitter Concepts This type of measurement is used to measure the timing and jitter behavior for a device under test DUT It uses a bit error rate BER measurement to evaluate the shape of the eye for the output signal of the DUT It also analyzes the jitter separates the random jitter and deterministic jitter components and estimates the total jitter A direct result is the determination of the optimum sampling point delay for receiving data from the DUT with maximum confidence DUT Output Timing Jitter includes the Fast Total Jitter measurement that can be used to measure the total jitter for devices which generate a very low error density in a reasonable time span Agilent J BERT N4903B High Performance Serial BERT 241 6 Advanced Analysis 242 Output Timing Characteristics The sampling point is swept automatically within a 1 5 clock period t
386. low Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Signal without jitter 1 I i Expected and correctly d i sampled data i Periodical sinusoid jitter 1 Na i i i i i 1 i i 1 H increasing right shift decreasing right shift H increasing left shift j decreasing left shift A H r Y Y Y Signal with jitter T p 0 LLL LLL fee ee a ee apo po Ss ee Pol cub uk al his eS sb ub leak Mer ling cd o dax cd Captured data 1 0 1 0 1 0 1 0 T 1 0 1 0 1 0 1 0 1 Expected data 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 Error data 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 The jitter source moves the received signal to the right and to the left but the sampling point is fixed dotted lines show the undisturbed signal Because we are sampling at the transition point we get errors when the signal is shifted to the right and no errors when the signal is shifted to the left The error signal shows a very characteristic and reproducible pattern The resulting BER for this kind of signal is 0 5 Other jitter frequencies yield different but also characteristic error patterns and also a BER which is half the maximum BER The maximum BER and hence the actual BER are pattern dependent Considering the BER bathtub curve the sampling point is in the middle of the descending line If random data is used the average bit error rate at the left hand side of the jitter region is 0
387. low Agilent J BERT N4903B High Performance Serial BERT 85 3 Setting up Patterns Block Edit Operations Range Fill Recode Traces Rotat gt Fill with Zeros Fill with Ones C Select All Range From lo Sod To fo Clock Pattern 0101 0101 Fill From Cursor to End Clock Pattern divide by C Fill From Start to Cursor C Symbol 00 gt Range Applies to Pattern A Execute ses With the Block Edit Configuration dialog box you define the Range that is to be modified The available options for the Range are Select All Choose this option to select the entire pattern for editing Range Choose this option to select the range of bits specified by the bit positions entered in the From and To fields Fill From Cursor to End Choose this option to select the bits from the cursor location to pattern end Fill From Start to Cursor Choose this option to select the bits from pattern start to the cursor The line at the bottom displays to which part of pattern the current range applies The Block Edit Configuration dialog contains the following tabs NOTE All the 5 tabs are visible in Symbol and OOB symbol mode In Bin Hex mode the Recode and 8b 10B tab are not available In PAM4 the 8b 10b tab is not available 1 Fill This tab allows filling the given range in the pattern with the specified value The available options for the Fill are Fill with Zeros 86 Agilent J BERT N4903B Hi
388. low Uniform Hanning Hamming Blackman win You can see from the figure that the Blackman window is the strongest filter The windows are based on the following formulas Uniform w niz 1 0 Hanning n 2m wi n 0 5 cos N jj Hamming mn wini 0 54 0 46 cos OON 7 2n wia 0 42 0 5 cos Black ian ackman 0 08 cos NOTE In case of leakage FFT windows improve the spectral resolution FFT windows generally reduce the measured spectral power For an introduction to FFT windows see Leakage and Windowing on page 349 Pass Fail Tab The Pass Fail tab of the Properties dialog box allows you to specify the criteria to decide whether the DUT passes or fails the test You can change pass fail criteria without rerunning a test The software only uses the criteria to rate the results of a measurement Agilent J BERT N4903B High Performance Serial BERT 355 6 Advanced Analysis 356 NOTE The pass fail criteria do not control measurement execution The measurement run will be completed even if the measurement fails for one or more of the criteria Properties Frequency Range 1 ZA D sana E FTequency mange 2 Fre H a Bd Frequency Range 3 Frequency Range 4 eee Danan F Frequency Range amp pU EILEEN J Frequency Range 6 Frequency Range 7 Frequency Range 8 The following
389. ls 3 Press OK when you have made all required changes to close the Properties dialog box Spectral Jitter Reference The Spectral Jitter measurement returns the results in a graphical and in a numerical form The following sections provide explanations of the measured parameters and the display options that are specific to this measurement Additionally some information is provided to explain the theoretical background behind the jitter calculations Properties that can be specified on the various tabs of the Properties dialog box Parameters Tab NOTE If you modify the parameters on this page you have to rerun the measurement to update the results Set the data acquisition parameters e Acquisition Depth Suitable numbers can be chosen from the drop down list The Acquisition Depth determines the length of the time record used for the FFT A smaller number reduces the duration of the Spectral Jitter measurement and its precision A larger number increases the measurement duration but also the frequency resolution of the measured spectral components The frequency resolution of the measurement is Agilent J BERT N4903B High Performance Serial BERT 353 6 Advanced Analysis 354 data rate Acquisition Depth For example If you have a data rate of 2 5 GHz and an Acquisition Depth of 128 Kbit the frequency resolution is 19 0735 kHz The relations are illustrated in the following figure 1000000 4
390. lue directly with the numeric keyboard Usethe knob by the numeric keyboard to fine tune the value TIP You can also adjust Data and Clock V with the Data Offset and Clock Offset knob on the front panel See Understanding the Output Level Parameters on page 126 for information about how the Serial BERT modifies the output levels Agilent J BERT N4903B High Performance Serial BERT 135 4 Setting up the Pattern Generator TIP TIP 136 Vlo This text field allows manual entry of the logic low voltage level and displays the current value To modify the value click inside the text field and either Enter the desired value directly with the numeric keyboard Use the knob by the numeric keyboard to fine tune the value See Understanding the Output Level Parameters on page 126 for information about how the Serial BERT modifies the output levels Vampt This text field allows manual entry of the voltage amplitude and displays the current value To modify the value click inside the text field and either Enter the desired value directly with the numeric keyboard Use the knob by the numeric keyboard to fine tune the value You can also adjust Data and Clock Vamp with the Data Amplitude and Clock Amplitude knob on the front panel See Understanding the Output Level Parameters on page 126 for information about how the Serial BERT modifies the output levels Delay This text field allows manual entry
391. ly is automatically selected NOTE When you select Custom there are no changes in the expected terminations See Understanding how the Serial BERT Uses Logic Families and Terminations on page 128 for additional information Terminations for Data Aux Data Clock and Trigger Ref Clock Use the Termination checkboxes to toggle between true differential DC coupling and center tapped termination In center tapped mode you can specify the termination voltage The Termination fields are used to set the termination voltage When you select a Logic Level the terminations are set accordingly It is recommended that you keep the default values of the terminations See Understanding how the Serial BERT Uses Logic Families and Terminations on page 128 for additional information Agilent J BERT N4903B High Performance Serial BERT 137 4 Setting up the Pattern Generator 138 CAUTION In Pattern Mode In Sequence Mode NOTE 0V Disable Before clicking this button ensure that your device s inputs tolerate ground potential Use this button to clamp the pattern generator s output ports Data Clock Aux Data and Tigger Ref Clock to ground You can see the status of the outputs in the display area of the pattern generator Note that you can also use the buttons present below the display to enable or disable the output ports This function can be useful in case you want to connect or disconnect a device to any of the o
392. m trace is longer refer to the technical specifications Data can be input by connecting the pattern generator s Data Out port to either P1 or P2 Agilent J BERT N4903B High Performance Serial BERT 403 8 Jitter Tolerance Tests Jitter Setup If P1 is used as input sinusoidal interference is or can be added after the signal has passed the chosen trace Seen from the DUT this is called near end injection If P2 is used as input sinusoidal interference is or can be added at the beginning of the trace This is called far end injection Jitter Setup Concepts 404 The Jitter Setup function is used for composing the total jitter in a defined and calibrated way This function reflects the present setup of the Serial BERT s internal jitter generation hardware see also Jitter Generation Block Diagram on page 401 The Jitter Setup display is divided into the following four sections Jitter on off button and delay line status area Jitter configuration area Jitter setup area Jitter switch area Jitter On Off Button and Delay Line Status Area The Jitter On Off button allows you to globally enable or disable the jitter generation Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 The delay lines show markers which indicate the used and the remaining jitter amplitude in UI Colored bars allow you to judge the impact of each jitter component on the total jitter Ul is the ac
393. mal distribution see also The Gaussian Marker on page 237 and Graph Tab on page 276 Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 QBER vs Threshold Graph This graph shows the extrapolation of the optimum Q factor and the optimum threshold level from a limited number of measured points The measured data points to be used for the calculation have to be within a contiguous BER range This range is defined by specifying the Min BER for Q lower threshold and the BER Threshold upper threshold Both thresholds can be set in the lower section of the View tab The Q factor can only be calculated if for both high and low level rails two or more points fall within the defined BER range For reliable results use at least five measured points To ensure proper settings activate the Show Measured Points function on the Graph tab and switch to the BER vs Threshold graph Move the upper BER threshold marker vertical line so that a sufficient number of measured points is included for the calculation 800 0 mv 1 Trace 501 Points 900 0 mv A0Vv AAV 42v A3V 44v ay ABV ATY 48V BER Threshold 2 194e 2 Logarithmic 1 000e 7 1 000e 5 1 000e 3 1 000e 1 1 000e 6 1 00064 1 000e 2 1 000e From these points the following QBER vs Threshold graph is generated 1 Trace 16 Paints Q factor This graph illustrates the calculation of the Q factor 65 in the figure ab
394. mber of errors the measurement stops for the current sample point and moves to the next one This allows you to speed up the measurement You can switch off this option if only the number of compared bits is important The measurement moves to the next sample point when the first of the two criteria is reached Set the criteria how the sampling points of the eye are displayed Timing Unit Choose between Unit Interval and Seconds to set the timing values for the measurement points on the Parameters tab This setting does not affect the display of results Threshold Type Choose between Absolute Offset or Percentage to set the thresholds for the measurement points Absolute You specify the thresholds for the measurement points as absolute voltages Offset You specify the thresholds for the measurement points relative to the threshold voltage of the actual sampling point Percentage You specify the thresholds for the measurement points as a percentage of the current eye opening voltage of the error detector Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Threshold voltage 4 016 40 16 04 Actual sampling point Depending on the quality and characteristics of the eye opening the resulting 50 threshold may deviate from the decision threshold defined in the loaded setting Number of Valid Points Change the Number of Valid Points if you wish to use less
395. ment from the Agilent Recovery System See How to Boot from the Agilent Recovery System on page 526 2 From the Agilent Recovery System user interface choose Restore Factory Setting option WARNING Remember with this option the entire user generated data and settings will be lost and there will be no chance of recovering this data afterwards 3 Follow the on screen instructions to continue the restore process from hard disk The Agilent Recovery System will now restore the factory settings You must not turn off the instrument or remove the power before the process completes Doing so may result in system corruption and the application may not operate Agilent J BERT N4903B High Performance Serial BERT 527 10 Customizing the Instrument Configuring the Instrument Configuring the Instrument Concepts CAUTION General Notes to Updating the Firmware Server 528 This section covers information on the configuration of the Serial BERT Updating the Software New software features are offered for the Serial BERT from time to time you may be able to enhance the capabilities of your instrument by updating the software See the Agilent Serial BERT web site to see if updates are available for your instrument http www agilent com find SerialBERT Besides the latest versions of the user interface and firmware server you will also find instructions for the update process Before you update the instrument s softwar
396. ment searches for the first bit errored in the incoming bitstream and marks it in the pattern The address of the errored bits can be displayed after the error is located Agilent Technologies 227 6 Advanced Analysis 228 Fast Eye Mask Eye Diagram Spectral Jitter The Fast Eye Mask measurement is first of all meant for production and screening tests It allows to determine very quickly whether the eye opening seen at the output signal of a device is within specifications that means within certain timing and voltage limits This is achieved by measuring the bit error rate at a limited number of test points The Eye Diagram allows a quick check for the DUT s signal output and determines the signal quality Due to the higher sampling depth of a BERT value the eye contour lines display the measured eye at a deeper BER level for accurate results The Spectral Jitter measurement allows you to analyze the jitter inherent in the output signals of your device under test DUT as a function of the frequency This measurement can be used for investigating the behavior of the DUT for example to identify crosstalk effects Before you can run any of these advanced measurements you have to properly set the required parameters After the test is complete you can view the results General Requirements For any measurement to yield reasonable results the following requirements apply Theerror detector must be correctly set up and syn
397. mmended way to load the data patterns to the pattern generator and error detector is to use the Pattern Select Form dialog box See Loading Patterns to the Pattern Generator and Error Detector on page 73 for details Hardware Generated PRBS Reference Use the Pattern Select dialog box to select the pattern s you want to write to the pattern generator and the error detector See Select Pattern Dialog Box on page 89 for details Agilent J BERT N4903B High Performance Serial BERT 95 J Setting up Patterns User Defined Sequences User Defined Sequences Concepts 96 This section describes the basics of user defined sequences A sequence is created and maintained by means of the Sequence Editor that can be enabled from the Pattern window A sequence consists of up to 120 blocks than can be looped Each block can generate a pause signal constant 0 or 1 a divided clock signal a 2 n 1 PRBS or a user pattern Single or multiple blocks can be looped The sum of the blocks and the counted loops must not exceed 120 An overall loop restarts the sequence after it has come to its end When to Use a Sequence You may wish to test a device that uses a certain protocol for processing data For example the device might expect synchronization data a preamble payload data and a suffix All this can be provided by a user defined sequence How a Sequence is Defined The sequence is defined by a SequenceExpression which
398. mpling point delay values The error detector can vary the sampling point delay only within a range of Ops to 1 6129ns relative to the clock signal For frequencies above 620 Mbit s this range is sufficient to cover the complete clock cycle 1 unit interval For lower frequencies the maximum sampling point delay is smaller than the clock cycle Therefore the sampling point cannot be set everywhere within the clock cycle Clock 500 Mbit s Valid Range for Sampling Point i H 0 1 6129 ns lt v a n D 3 T a D 9 a If you cannot find the optimum sampling point in the valid range of sampling points you can switch from the rising to the falling clock edge For this purpose activate the Clock Falling Edge checkbox on the Sampling Point Setup screen of the error detector setup i 2ns i Clock 500 Mbit s Sampling Point Delay Rising Edge Sampling Point Delay Falling Edge Covered Range Agilent J BERT N4903B High Performance Serial BERT 177 5 Setting up the Error Detector 178 With this method you can still place the sampling point anywhere in the clock cycle to find the optimum sampling point even at low frequencies i 3 226 ns A Clock 310 Mbit s Covered Range However if the bit rate falls below appr 310 Mbit s even with this method gaps occur in the range of possible sampling points 6 66 ns Clock 150 Mbit s Covered Range with Gaps As a result you might not be able to
399. n Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 Auto Adjustment M8061A Input timing adjustment Data rate 8 77 Gb s Step Progress 6 MEE Total Progress MEE Cancel NOTE The adjustment process may take several minutes for completion The following message will pop up after the successful completion of adjustment process M80614 Adjustment is over No Errors were detected Problems with the N4916A Problems with the N4916A Concepts When you have opened the External Instrument s Config window and have selected N4916A from the external instruments select De emphasis under Enable Function column in the list the N4916A Connection window appears Cannot Connect Problems An error message may appear when you now press Enable Agilent J BERT N4903B High Performance Serial BERT 493 9 Solving Problems If this happens ensure that the N4916A has been correctly installed see Installing the N4916A B Procedure on page 539 Check the Power Supply Ensure that the power cord is connected The green LED close to the power switch must be illuminated when the unit is turned on Check the USB Communication Ensure that the USB cable is connected between the USB port at the rear of the N4916A and one of the USB ports at the rear of the Serial BERT 1 Open the Utility menu and click Minimize GUI to gain access to the Windows operating system 2 In the Windows task bar
400. n correctly installed see Installing the N4876A Procedure on page 544 Check the Power Supply Ensure that the power cord is connected The green LED close to the power switch must be illuminated when the unit is turned on Check the USB Communication Ensure that the USB cable is connected between the USB port at the rear of the N4876A and one of the USB ports at the rear of the Serial BERT 1 Open the Utility menu and click Minimize GUI to gain access to the Windows operating system 2 In the Windows task bar click the Agilent IO Control icon 10 Agilent J BERT N4903B High Performance Serial BERT 511 9 Solving Problems 3 From the Agilent IO Control menu open the VISA Assistant WS VISA Assistant File Edit Yiew Configure Help R merere instrument Driver Formatted 1 0 Memory 1 0 Attributes N48764 No Instrument Driver Configured Address Sting USB0 2391 29976 DE50400005 0 INSTR rsrcName Configure For Help press F1 11 51 04 4 fnoVISA alias name has been assigned so far the VISA Assistant should indicate an instrument named UsblInstrument1 If the N4876A has been installed correctly the VISA Assistant should indicate an instrument named N4876A Click UsbInstrument1 or N4876A 5 Click the Formatted 1 0 tab 6 Inthe Instr Lang box Instrument Language enable SCPI m Instr Lang C Custom C IEEE 488 2 G so 512 Agilent J BERT N4903B High Performance Serial
401. n external PC 1 Copy the screen image file from the Serial BERT to your PC By default image files are saved in the folder C lt instrument model gt User 2 Open the file in Microsoft Paint or another graphics program Preset Instrument State Preset Instrument State Procedures Steps to Preset Instrument State To reset the instrument to the preset state select one of the following options Press the Preset button on the front panel or Select Preset in the File menu or Send an RST command to the GPIB controller 562 Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 Self Test Self Test Concepts The self test checks specific hardware components for basic functionality At initialization the following self tests are automatically run Pattern Generator Power Up Test Error Detector Power Up Test NOTE The current system state is saved and the instrument is re initialized before performing the selected tests The saved system state is restored upon completion of the self test When Self Test is selected from the Utility menu the Self Test Options dialog box opens In this dialog box you can select additional self tests that will run when you click Run Test NOTE During accumulation or while auto search functions are in progress Self Test is not available The pass fail results of the self test are displayed in the Self Test Options dialog box In case of faile
402. n of the noise distribution on the 1 and 0 rails The p o and o 9 values are calculated from a selected range of data points This calculation is correct if the noise distribution has Gaussian characteristics Then the bit error rate can be expressed as erfc 2 o BER D je ncm l where D is the decision threshold y o and 04 are the mean and standard deviation of the 1 and 0 rails and erfc x is the complementary error function This formula is the sum of two terms It considers the probabilities of deciding that a 0 has been received when a 1 was sent and that a 1 has been received when a 0 was sent For the following calculations the assumption is made that the BER is dominated by only one of the terms noted above depending on whether the threshold is closer to the 1 or 0 rail For the complementary error function oo 1 p 2 1 72 erfc x Ju e dx Wr e X an inverse logarithmic approximation exists 1 Log gt orfeco 2 1 192 0 6681x 0 0162 x where x Log BER Agilent J BERT N4903B High Performance Serial BERT 283 6 Advanced Analysis 284 This function applied to the high level and low level data points yields new threshold vs value combinations In the area of low BER typically below 10 4 these new data pairs should fit to two straight lines although a couple of assumptions and approximations have been made To determine the gradient and offset o
403. n page 392 Agilent J BERT N4903B High Performance Serial BERT 25 1 26 Planning the Test The following references may also be helpful in determining how to test your device For a list of recommended patterns for your device and application see When to Use Which Pattern on page 64 For a list of recommended connection diagrams for your device and application see Connecting the DUT Concepts on page 29 What are the Testing Requirements of Your Device Considering the requirements of your device will help you choose which instrument functions to use during the test setup and when making measurements Do you need to use a custom test pattern If you have a long series of 1 s or 0 s in your custom pattern be sure to avoid false sync See What is False Synchronization on page 209 Do you want to use alternating patterns Then you must first import or create a custom pattern that contains an A half pattern A and B half pattern B See Pattern Alternation Concepts on page 161 Are you aware of how custom pattern size can affect the pattern output See Why the Serial BERT Repeats Memory Based Patterns on page 69 Do you need more than two patterns or a mix of custom patterns and PRBS Does your device change the input pattern it receives If your device inverts logic then select Data Inverted in the Sampling Point Setup window If your device changes patterns in other ways then you must know the expect
404. n the device is stimulated by another data source Illustrates the use of an oscilloscope in addition to the error detector DATA SOURCE NOTE As in all setups the pattern sent to the error detector must be the same as the internal reference pattern of the error detector Dynamic patterns and or patterns with live payloads cannot be used NOTE Diagrams 1 through 8 can use an external clock reference connected to the Clock In port of the pattern generator if desired Agilent J BERT N4903B High Performance Serial BERT 33 1 34 Planning the Test Diagram 9 Connections of the De Emphasis Signal Converter The Data Output of the Pattern Generator is connected to the Input of the N4916A B The unused output of the Pattern Generator has to be terminated with 50 Ohm The N4916A B provides a differential signal to the DUT The N4916A B is controlled by the Serial BERT via a USB cable connected to ports at the rear of the instruments Diagram 10 Connections of the N4876A 28 Gb s Multiplexer The Data In Aux Data In and Aux CIk In of N4876A has to be connected to Data Out Aux Data Out and Aux CIk Out ports of the Serial BERT s respectively The connections should be made using the matched cable kit with the part number N4915A 011 The Data Output has to be connected to the device under test The unused output of the Pattern Generator has to be terminated with 50 Ohm The N4876A provides a differential signal to the DUT
405. n the test again See Parameters Tab on page 295 for details Pass Fail tab These settings determine whether the calculated results are recognized as passed or failed However a new test run is not required when doing changes here See Pass Fail Tab on page 296 for details Agilent J BERT N4903B High Performance Serial BERT 291 6 Advanced Analysis 292 View tab Graph tab and Color tab All settings on these tabs only affect the way the data is displayed You do not need to run the measurement again See View Tab on page 298 Graph Tab on page 300 and How to Change the Colors of the Graph on page 232 for details 3 Press OK when you have made all required changes to close the Properties dialog box How to Use the Color Bar The color bar at the right hand side of the diagram shows the assignment of BER thresholds to colors These colors can be changed In addition the color bar provides some useful options for controlling the display Color bar 1 Trace 9510 Points Ul 401 878 ps All Errors 600 0 mv 400 0 mv 200 0 mv Handle for the oo v BER range 200 0 mv 400 0 m 600 0 mv Slider for the BER LI scale 0 448 UI 0449 ul 0149 LI 0 446 LII 0 746 LI threshold Relative 0 597 LII 0 299 UI 0 000 LII 0 299 LII 0 597 LII Terminal Show Color Copied Isa ices c eee eee pening Opening Del e iv 0 945 LII 880 mv oul xj ils of How to Add or Change Colors If you have a mouse co
406. n was received Filler Word SKPOS A trigger is sent on reception of programmed Filler Word or SKPOS depending on 8B 10B Symbol Comparison Bit Comparison without PCle3 SKPOS or Bit Comparison without USB3 1 SKPOS mode respectively Next trigger is sent only when first trigger is processed 3 For the Aux Out select whether the Clock signal or the Data signal is supplied If Clock is selected the clock signal used by the error detector is output directly at Aux Out If Data is selected the input data is output over a comparator at Aux Out The comparator is controlled by the 0 1 threshold This lets you use an oscilloscope to determine if the 0 1 threshold is correctly set If the 0 1 Agilent J BERT N4903B High Performance Serial BERT 203 5 Setting up the Error Detector threshold is set below or above the data eye the output at Aux Out will be constant high or low respectively 4 Click OK to finish the Trigger and Aux Out setup Trigger and Aux Output Reference The available options in the ED Trigger amp Aux Output window are Trigger You have the following options to configure the signal at the Trigger Out port e Clock Divided by n The trigger is set every n bits zs mnt Pattern The trigger signal is sent every time when the complete pattern has been received This refers to the single pattern repetitions not the 512 bit blocks Filler Detected The usage of the enhanced error detector
407. nFalling Sequence starts with a rising or falling edge at the Auxiliary Input Manual Sequence starts when the Start button is clicked Block Options Defaultis pattern PO Pause0 a length of 512 bits and Trigger Ref Clock On Other options are Patterns P1 Pause undistorted PRBS a user pattern from a file or a fraction of the clock pulse divider range 2 127 Length Must be 512 bits or a multiple thereof Shorter or longer user patterns are automatically repeated until the next 512 bit boundary is reached see also How the Serial BERT Generates Memory Based Patterns on page 69 Agilent J BERT N4903B High Performance Serial BERT 97 3 Setting up Patterns 98 NOTE Length is related to master channel Data Out while the other channel Aux Data will always follow this length Trigger Refers to the pattern generator s Trigger Ref Clock Out port If the Trigger Ref Clock Out is set up to generate a divided clock pulse the block setting is ignored But if the Trigger Ref Clock Out is set to Sequence Trigger and the block to TrigOn a trigger pulse is generated whenever the block starts or restarts No trigger is sent if the block is set to TrigOff See Trigger Ref Clock Output Procedures on page 145 for information on how to set the mode of the Trigger Ref Clock Out port LOOP Options Default is one overall endless loop for the whole sequence B Bm Define the start and end block of a us
408. nary search methods Searching Down Up Both search directions upwards and downwards are supported This helps to find out whether the device exhibits some degree of hysteresis The BER measured at a specific jitter amplitude may be different if this amplitude is approached downwards or upwards When you search in downward direction from maximum to minimum the test for one frequency stops as soon as the measured BER falls below the target bit error ratio The following figure shows an example of a logarithmic search downwards 1 1 kUI 10 UI 10 UI Hee Bie tho e 0 M cooks S K x 100 mUI Select Point 10 mUI 1 i 1 1 900 Hz 10 kHz 100 kHz 1 0 MHz 10 MHz 88M Frequency The maximum amplitude indicated by the dotted blue line is given by the capability of the instrument Searching downwards is usually faster than searching upwards because the measurement of a single point is finished as soon as the BER is found to be above the limit Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Searching upwards When you search in upward direction from minimum to maximum the test for one frequency stops as soon as the measured BER exceeds the target bit error ratio Dynamic Search If you perform the measurement over a wider frequency range you measure many useless points with the downwards upwards methods To construct the tolerance curve we actually need only the two points
409. nce tests prescribe combinations of certain jitter types This section gives an overview of these types Random Jitter Random jitter RJ is always present Resistors transistors and other components generate noise that impacts transmitters and receivers This kind of jitter has a Gaussian distribution over time Theoretically the Gaussian distribution stretches from to In practice the amplitude is defined by the crest factor p p amplitude sigma x crest factor Agilent J BERT N4903B High Performance Serial BERT 397 8 Jitter Tolerance Tests Bounded Uncorrelated Jitter Similar to random jitter bounded uncorrelated jitter BUJ has also a Gaussian distribution but this distribution is cut bounded at both sides This kind of jitter can be caused for example by crosstalk on a parallel bus or by intersymbol interference of random or long pseudo random binary bit sequences To simulate bounded uncorrelated jitter for jitter tolerance tests it can be generated from a filtered frequency limited PRBS Periodic Jitter Periodic jitter PJ is generally picked up from other periodic sources Rectangular Jitter Rectangular jitter may be caused for example by a switching power supply It has a distribution as shown below Jitter amplitude 398 Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 The jitter amplitude is usually specified in UI One UI Unit Interval is always th
410. nces of the detect word with false synchronization Expected Incoming Pattern Bit stream Em 0 Reference Point False Sync If the error detector attempts to synchronize on the incorrect detect word the BER will be unacceptably high and if automatic synchronization is selected the error detector attempts another resync Agilent J BERT N4903B High Performance Serial BERT 207 5 Setting up the Error Detector 208 Further Considerations The detect word on which the error detector attempts to resync is chosen strictly by chance So if there are two instances of the detect word in the pattern the error detector has a 50 chance of selecting the correct one The more instances of the detect word exist in the pattern the higher are the chances for incorrect synchronization The software attempts in any case to identify a 48 bit pattern that occurs as seldom as possible in the pattern For very large patterns this can unfortunately take a very long time and the software ends the search if it expects that it would take longer to find an adequate detect word than it would to attempt to synchronize If the search for a detect word is ended the most unique detect word identified is used Patterns must always be synchronized in order to do accurate BER testing If patterns are out of alignment by just one bit errors can be as high as 50 bE 1 for PRBS patterns and 100 1E 0 for custom patterns By de
411. nchronizing at a sync threshold BER lower than the fixed error ratio If sync is acquired without the problems listed above then your previous sync was false Your current sync should be on an exact pattern alignment While auto search functions are in progress the sync threshold BER is changed to the same value as the alignment BER threshold If you are using these functions and want to consistently re synchronize at a lower sync threshold you must set the alignment BER threshold to the same value as the sync threshold BER Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 Introduction to Burst Sync Mode The burst sync mode is a special operating mode for measuring data in bursts of bits rather than one continuous stream of bits The burst sync mode measures bit error rates for each burst of data after the error detector synchronizes to the incoming pattern The signal at the Gate In port controls the timing of synchronization and error counting for each burst This function is useful for example to analyze recirculating loop data With a recirculating loop you can simulate very long cable connections like transoceanic cables Data is sent into a looped cable drum and after a certain number of loops it is diverted into the error detector This way only a short burst of valid data reaches the Serial BERT followed by a long break Therefore the error detector needs to be synchronized at the beginning o
412. nds when the error ratio is better than 1 x 10 for 10 consecutive seconds These 10 seconds are considered part of the available time Percent 96 Unavailability is the ratio of the unavailable seconds to the total gating period expressed as a percentage Severely Errored Seconds The ratio of the total number of available seconds with an error ratio worse than 1 x 103 to the total number of available seconds expressed as a percentage Degraded Minutes Severely errored seconds are discarded from the available time and the remaining seconds are grouped into blocks of 60 seconds Blocks which have an error ratio worse than 1 x 105 are called Degraded Minutes and Percent 96 Degraded Minutes is the ratio of the total number of degraded minutes to the total number of 60 second blocks in the available time expressed as a percentage Incomplete blocks of less than 60 seconds are treated as complete blocks of 60 seconds Errored Seconds The ratio of the errored seconds in the available time to the total number of seconds in the available time expressed as a percentage Agilent J BERT N4903B High Performance Serial BERT 381 7 Evaluating Results 382 Interval Results Error Free Seconds The number of seconds in which no errors were measured is displayed here Error Free Deciseconds The number of deciseconds in which no errors were measured is displayed here Error Free Centiseconds The number of centiseconds in which no
413. nfiguration page for external instruments b Enable the De Emphasis function Click the Load Calibration Setting button d Setthe PG s data rate to 14 2 Gb s e Open the external instruments configuration screen again 86100A a Enable Channel 1 setthe vertical scale to 100 mV div and the offset voltage to 0 V b Select the appropriate trigger input and set the threshold voltage to 0 V Select Trigger Mode Normal DC 2 5 GHz Set the time base to 28 2 ps d Adjust the Delay to show the start of the 27 1 PRBS at the center of the screen 0000001111111 bit sequence e Setthe Waveform Persistence to infinite f Enable Eye Mask Mode Calibration Sequence For each Input Timing Setting 1 to 6 execute the following steps Set the Input Timing on the N4903B Clear the oscilloscope s display Wait for waveforms to accumulate for approximately 6 minutes Check the waveform on the oscilloscope for erroneous measurements that indicate a timing violation at the input of the N4916B When all 6 settings have been measured select the optimum input timing setting keep in mind that one erroneous sample on the screen has to be treated as an error Example screen shots Input Timing Setup 1 Agilent J BERT N4903B High Performance Serial BERT 485 9 Solving Problems File Control Setup Measure Calibrate Utilities Help 21Mar2011 11 50 LZ 100 mV div 41 6 mVv div Time 28 2 ps div Trig Normal ooy 2oy 3 notinstaled
414. ng an output signal Providing a wide range of clock frequencies You can use the pattern generator s internal clock or an external clock for defining the frequency of the outgoing stream Distorting the signal by adding jitter You can connect an external delay control device for example a function generator to add jitter to the generated signal Adding errors to the output stream The Serial BERT can be set up to insert errors into the outgoing stream either internally according to an external signal or manually from the operator Run time switching between two patterns You can set up two patterns and switch between them during runtime either automatically according to an external signal or manually Suppressing the output stream The output signal can be suppressed according to an external signal The pattern generator also provides output ports that let you connect an external instrument such as an oscilloscope Agilent Technologies 121 4 Setting up the Pattern Generator Input and Output Ports 122 Input and Output Ports Concepts Input Ports Output Ports The pattern generator s input ports are used to set the pattern generator s clock frequency and to manipulate the output signal with respect to jitter error insertion and signal output The pattern generator s output ports are used to supply a clock signal and trigger for another device for example for the Serial BERT s error detector
415. ng patterns Standard patterns have a value of 1 alternating patterns have a value of 2 Any other value is not valid Syntax Count lt 1 2 gt This is the length of the pattern in bits For a standard pattern of 128 bits for example the value is 128 For an alternating pattern it is the length of one of the patterns only The Data tag identifies the start of the pattern data itself For standard patterns there is only one Data tag Alternating patterns require two Data tags Note that the actual data starts on the ine after the tag Syntax Data lt CR LF gt lt data contents gt n dual format each character represents one bit only the numbers 1 and 0 are allowed Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 In hex or text format each character represents four bits only the characters 0 to 9 and A to F are allowed not case sensitive n binary format each character represents eight bits all ASCII characters are allowed n symbol format each symbol represents 10 bits Example X in binary format or 88 in hex or text format produce the same bit sequence as 10001000 in dual format In hex or text format the data must be separated by one white space The white space can be a tab a space or a new line In hex or text format For pattern lengths that are not multiples of 8 bits the data has to be padded to provide the correct number of hex characters
416. ng the procedure to determine xg for the right hand slope it calculates the total jitter peak to peak like in the standard timing jitter measurement Measurement Duration The duration of a Fast Total Jitter measurement depends on the BER threshold the bit rate the sample delay step size the contribution of random jitter the contribution of deterministic jitter Compared to a conservative bathtub measurement the Fast Total Jitter measurement can reduce the measurement time by more than a factor of 40 Typical test times are approximately 20 minutes at 10Gbit s and slightly more than one hour at 2 5 Gbit s for a measurement at the 10 1 BER threshold with an overall confidence level better than 9096 Agilent J BERT N4903B High Performance Serial BERT 249 6 Advanced Analysis 250 DUT Output Timing Jitter Procedures This section shows how to set up and use the DUT Output Timing Jitter measurement As an example we measure the output timing and jitter behavior of a shielded cable This requires the following steps Preparing the measurement see How to Prepare the DUT Output Timing Jitter Measurement on page 250 Executing the measurement see How to Execute the DUT Output Timing Jitter Measurement on page 251 Optimizing the view of the results see How to Optimize the View of the Results on page 251 How to Prepare the DUT Output Timing Jitter Measurement To use the DUT Output Timing Jitter meas
417. nge the block length The block length of user patterns is derived from the file Remember that such patterns are automatically repeated until a multiple of 512 bits is reached Agilent J BERT N4903B High Performance Serial BERT 99 3 Setting up Patterns 100 NOTE Decide on Trigger On or Off This setting becomes effective if the Trigger Ref Clock Out port is set to Sequence Trigger trigger at block begin See Sequence Trigger on page 147 5 Create loops if desired Click the out arrow of a block A loop condition box labeled INF appears Click the in arrow of the same or a previous block This closes the loop Click the condition box and choose the appropriate break condition 6 Download the sequence to the pattern generator and test it PG 10110 7 Save the sequence in a file for later re use Saving the Sequence in a File To save the present sequence under a new name 1 Inthe toolbar click the Save As icon 2 Select a suitable directory for example C lt instrument model gt Sequences 3 Enter a file name The filename extension is seq by default If the present sequence was loaded from a file and edited you can also click the Save icon In this case the saved file is updated Loading a Sequence From a File To recall a sequence from a file 1 In the toolbar click the Open icon Da a The Open Sequence dialog box opens 2 Use this dialog box to locate and open the desired sequenc
418. nger than a binary pattern Agilent J BERT N4903B High Performance Serial BERT 65 3 Setting up Patterns NOTE Version Format Description Count Length Data If you save a modified text format pattern to the same file name it is always saved in binary format If you wish to preserve the original text format file use the Save As file command and save the data to a different file Pattern File Specifications For your instrument to download a pattern file the pattern must be composed of at least 7 lines The first 6 lines start with a keyword to identify the field followed by an sign There should be no spaces around the sign and no blank lines Use the following keywords and examples to construct your ASCII file Note that the keywords must appear in the given order The version Field is used to identify different file versions For this software version the field should be set toEPA 2 0 An exact match for this field is required for the pattern to load correctly Syntax Version EPA 2 0 The format field identifies whether the pattern is in dual hexadecimal or binary format Syntax Format lt Dual Hex Text Bin Symbol Hex and text formats are identical This is the creator s description of the pattern The data after the sign is read until the end of the line and used directly as the pattern description The maximum number of characters is 256 This identifies standard and alternati
419. ngth can be set According to the chosen length the module generates a reproducible amount of intersymbol interference The Interference Channel provides also a sinewave generator and a modulator If sinusoidal interference amplitude modulation is enabled a minimum trace length is used Data can be input at P1 or P2 by connecting the pattern generator s Data Out to P1 or P2 respectively The output is meant to be connected to the input of the DUT The module has golden 2 4 mm connectors Be sure to use the specific cables or CAUTION adapters Improper connection can damage the module Interference Channel Setup Procedures To set up the Interference Channel module click Interference Channel from the Jitter submenu El Interference Channel To use the Interference Channel module 1 Click the checkbox that enables the module 2 Select a trace refer to the technical specifications for details 3 If desired enable sinusoidal interference Agilent J BERT N4903B High Performance Serial BERT 431 8 Jitter Tolerance Tests 432 If you have enabled sinusoidal interference enter appropriate values for Frequency Amplitude and Mode Interference Channel Setup Reference Amplitude The Interference Channel module must be enabled before it can be used This is done by clicking the corresponding button eme cece 300 0 mV 500 000 MHz Single Ended Normal Sd When the Interference Channel is disabled it esta
420. nization for example is very fast When you compare the sequence with itself you get 50 errors in all positions except when the pattern is exactly aligned Hardware generated PRBSs are non inverted PRBS 2 n 1 patterns that allow high speed measurement of random traffic They enable stressful testing of devices because large patterns contain long strings of 0 s and 1 s Software generated PRBSs are non inverted PRBS 2 n patterns They are also called PRBN patterns These patterns are useful for applications where the pattern length must be an even number The Serial BERT provides variations of software generated PRBS patterns allowing you to set a pattern s mark density ratio of 1 s and zero substitution These features make it easier for you to run specific stress tests on your DUT The following figure illustrates the PRBS based patterns available with the Serial BERT PRBS 2 n PRBS 2 n 1 PRBS Software Generated Hardware Generated Pure 2 n PRBS Mark 2 n PRBS 2 n PRBS Density Zero Substitution PRBS Frequency Spectrum Digital non return to zero NRZ data has a sin x x 2 characteristic frequency spectrum The PRBS frequency spectrum follows this envelope with line spectra of spacing related to the sequence repetition rate The longer the sequence before it repeats the narrower the line spacing closer to a continuous spectrum Line spacing
421. nnected to your Serial BERT you can add a color to the color bar or change a color From the context menu of the color bar choose from the following display options Table 30 Option Description Gradient Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Table 30 Option Description Add color To add an additional color to the color gradient at the cursor position The BER range for this new color will be assigned automatically Update Plots Dynamically Rainbow To display a large variety of colors for the bit error ranges Yellow Blue To display a color gradient from color1 to color2 for the bit error Green Blue ranges Red Green Black White White Black How to Change the BER Threshold The red slider indicates the present BER threshold To use another value as threshold Move the slider with the mouse or your finger if you are working directly on the Serial BERT This is particularly useful if you have set the display to show Equal BER at BER Threshold By moving the slider you can investigate the eye openings at destinct BER thresholds NOTE If no slider for the BER threshold is displayed make sure that you have enabled the BER Threshold option on the View tab and that a suitable BER is entered Agilent J BERT N4903B High Performance Serial BERT 293 6 Advanced Analysis How to Change the BER Range of a Color The bit error ranges are set automatically but you can chang
422. ns that may contain more than one instance of the used detect word Statistically every other burst would be correctly synchronized In this case it is recommended that you redefine the pattern This error can only occur with memory based patterns Unknown Agilent J BERT N4903B High Performance Serial BERT 383 7 Evaluating Results The status is unknown This can occur if accumulation has not been started or if Burst Sync mode has not been activated Burst Sync Ratio Percentage of time in burst while the Gate In signal is low that patterns are in sync and error counting is active A higher percentage is desirable because it indicates that more data is measured during each burst Bit Count Time Burst Sync Ratio CDR Settling Time Synchronization Time Bit Count Time Total Burst Count Total number of bursts received while in burst mode during the accumulation period Bad Burst Count Total number of bursts where sync is never achieved during the accumulation period Bit Count The number of bits considered for the accumulation period is displayed here This may not equal the total number of bits sent from the PG Accumulated Measurements for 8B 10B Comparison The Accumulated Results window for 8B 10B Comparison contains a graph and two tables Accumulated Results Accumulation Parameter Measurement Current Period
423. nt Recovery System requires the following steps How to Create the Agilent Recovery Partition on page 526 How to Boot from the Agilent Recovery System on page 526 How to Recover the System from the Hard Disk on page 527 Agilent Technologies 525 10 Customizing the Instrument How to Create the Agilent Recovery Partition When booting the instrument for the first time the Agilent Recovery System will automatically set up the recovery partition This process is indicated by a small window that pops up on first boot and will take less than one minute to complete Agilent Recovery Image Setup Setup will now create a recovery image This process will take less than a minute to complete Creating a recovery image provides a way to restore the instrument in the event of a failure How to Boot from the Agilent Recovery System The following steps will guide you to boot from the Agilent Recovery System 1 Plug in a keyboard and reboot the instrument 2 From the boot screen select Agilent Recovery System boot option NOTE This screen will be shown only for few seconds You will be now able to see the Agilent Recovery System user interface System Recovery for the following product N4961A N4962A N4961B N4982B N4963A N4983B N4986B 811418 81142A Choose one of the following options i Run CHKDSK to resolve any potential disk or file system issues Always try this first CHKDSK will also try to locate bad sectors on
424. ntified correctly by the operating system 2 Connectthe USB cable between the USB ports at the rear sides of the N4916A B and the Serial BERT You can use any rear port of the Serial BERT 3 The Found New Hardware Wizard will pop up and ask whether he should connect to Windows Update 4 Click No not this time 540 Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 Found New Hardware Wizard Welcome to the Found New Hardware Wizard Windows will search for current and updated software by looking on your computer on the hardware installation CD or on the Windows Update Web site with your permission Read our privacy policy Can Windows connect to Windows Update to search for software O Yes this time only Click Next to continue 5 Click Next and let the Wizard locate the appropriate driver The USB driver is part of the Agilent IO Libraries Suite 6 Step through the Wizard by clicking Next The Wizard will finally set up a USB device named USB Test and Measurement Device Found New Hardware Wizard Agilent J BERT N4903B High Performance Serial BERT Completing the Found New Hardware Wizard The wizard has finished installing the software for e USB Test and Measurement Device Click Finish to close the wizard 541 10 Customizing the Instrument Using the Agilent Connection Expert 542 Assigning a VISA Alias Name Assigning a unique VISA alias name to the N49
425. nts Select the Eye Measurements to be displayed Rise Time Fall Time Eye Amplitude Eye Height Eye Width Jitter Peak Peak Jitter RMS Cross Voltage Signal to Noise Ratio Duty Cycle Distortion Extinction Ratio Optical Modulation Amplitude Average Power amp I amp amp K amp XXX amp amp x KIKI KIKI Select All Cancel How to Set Markers Markers help in tracking the BER at specific points so to set the markers 1 Click on the Graph tab 2 Select Show Maker Lines Markers Show Marker Lines 3 ApplySetting This enables the markers and a marker readout button appears on the top left corner You also have an option to Reset Markers to their original position Agilent J BERT N4903B High Performance Serial BERT 325 6 Advanced Analysis 326 The Marker reading is shown on the left side of the graph 4 Click the marker readout button on the top left corner to see the detailed marker readout a How to Run the Mask Measurement To run the Mask Measurement the Eye Diagram measurement must be running 1 Click on the Mask tab 2 Click Load button Mask Selection 3 Select a Mask File Select mask file Look in 8 MaskFiles X E3 01xGbEthernet E Ceit 1GsrRxCI 02xGbEthernet 5 Ceit 1GsrRxNJ My Recent i0GbE 9 953 May02 5 Ceit 1GsrTx Documents 10GbE 10 3125 May02 FBDIMM1_3 2G_LargeTxRev0 85 Ei 10GbEthernet
426. o generate a bathtub curve The resulting graph is centered around the optimum sampling point of the port In addition the results are available in a tabular view If a clock signal is defined the software measures the data to clock alignment and displays the absolute delay Jitter Characteristics The DUT Output Timing Jitter measurement calculates the jitter histogram as the absolute of the derivative of the measured bit error rate jitter d BER dt The jitter histogram allows to separate and calculate the different components making up the jitter Random Jitter RJ Deterministic Jitter DJ Estimated Total Jitter TJ The peak to peak RMS and mean values are displayed in the result table They can also be displayed graphically Jitter Measurement Parameters on page 260 describes how these components are calculated A dedicated Gaussian marker allows to investigate the jitter graph in detail and to measure the contribution of certain sections or jitter peaks to the overall results You can set several signal parameters the values to be displayed and the output format according to your needs Fast Total Jitter Measurement Characteristics This measurement is an alternative to the standard jitter measurement It can be used to measure the total jitter at very low bit error ratios Whereas usual jitter measurements at a BER around 10 can take days due to the huge number of bits that has to be compared for each m
427. o start and stop accumulation when the front panel buttons are pressed Single Activate this option to configure the error detector to accumulate over one accumulation period and then stop Repeat Activate this option to configure the error detector to accumulate over repeated accumulation periods whereby one period follows another immediately There is no dead time between the end of one period and the start of the next NOTE Repeat will not create multiple measurement log files The data for all repeating periods will be saved to the same log file The accumulation mode cannot be changed while accumulation is running Measurement Log Prompt for File Name Click this option if you want to be prompted for a measurement log file name before each accumulation period begins With this option you can select a new file name for each test run File Name Click this option if you want to enter a file name for measurement logs before starting the accumulation With this option you will overwrite the measurement log file with each new test run However in Repeat mode the results of all repetitions will be appended to this file and no information is lost e No Logging Click this option to turn off measurement logging Serial BERT will not save the test results Agilent J BERT N4903B High Performance Serial BERT 221 5 Setting up the Error Detector Period The accumulation period cannot be specified for Manual tests
428. ode with filler symbols Block Length is equivalent to Pattern length in All Bits mode and specified Block Length in Block mode without filler symbols Pattern Count is equivalent to Pattern Count Symbol Counter Block Length Conversion factor is equivalent to 10 in case of SER ISR DSR in Auto and User mode and Block length in case of FER FSR e Calc Bit Error Ratio cBER The cBER is based on i e SER FER DER ISR FSR is displayed here The following is the expression used to calculate the Calc Bit Error Ratio BER Symbol Error Count A Calc Bit Count e Calc Bit Error Count The Calc Bit Error Count is is displayed here which is equal to Symbol Error Count Agilent J BERT N4903B High Performance Serial BERT 385 7 Evaluating Results 386 Compared Symbol Count Symbol counter The number of compared symbols considered for the accumulation period is displayed here This may not be equal to the total number of symbols sent from the Pattern Generator Compared Symbol count excludes any filler symbols i e K28 1 K28 1 for USB 3 0 and considers block length in Block Mode or Pattern length in All Bits mode during comparison with Expected Data Symbol Error Count Symbol error counter The number of errored symbols measured during the accumulation period is displayed here Comparison results between incoming Compared Symbols Count with Expected Symbol Count Both Compared Symbols and expected Symbols excludes Fillers Sym
429. of Data and Aux Data output delay and displays the current value in picoseconds To modify the value click inside the text field and either Enter the desired value directly with the numeric keyboard Usethe knob by the numeric keyboard to fine tune the value You can also adjust delay with the Data Out Delay knob on the front panel Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 Half Rate Clocking and Clock Duty Cycle A checkbox allows to enable half rate clocking the clock at the clock output runs at half the bit rate In half rate clocking mode the duty cycle of the clock can be adjusted in the range of 40 to 6096 Xover This text field allows manual entry of the data s crossover percentage and displays the current value To modify the value click inside the text field and enter the desired value directly with the numeric keyboard or use the front panel knob to fine tune the value Data Polarity Inverted Select this checkbox to invert the logic of both data outputs Data Out and Aux Data Out Data Aux Data Clock and Trigger Ref Clock Logic Levels You can quickly set up an output signal that is compatible with your device by selecting a logic family from the Logic Level list Select the logic family you want to use for the Data Aux Data Clock and Trigger Ref Clock outputs When you select a family the corresponding signal voltage levels and the Termination for this fami
430. of the error indicators should show red The resulting BER should be zero Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 How to Execute the Eye Opening Measurement To run the Eye Opening measurement 1 Switch to the Analysis panel and then press the Eye Opening icon 2 Press the Start button to execute the measurement The measurement is run and the display is continually updated The following illustration shows the eye diagram of the measured bit error rate and in the tabular view the calculated results for the Time Eye Opening the Threshold Eye Opening the Optimal Sample Delay and the Optimal Sample Threshold 1 Trace 9510 Points Ul 401 878 ps All Errors 600 0 mv 400 0 m 200 0 mv 0 0 v 200 0 mv n o B B a QO 400 0 mv 600 0 m 0 149 UI 0 448 UI 0 746 UI I 0 299 UI 0 597 UI Threshold Eye Optimal Sample Opening Delay sscalez 0 448 UI 0 149 UI Relative 0 597 Ul 0 299 UI 0 0 Doo LI Time Eye Opening How to Optimize the View of the Results After you have run a measurement the resulting graph and the calculated numerical values are displayed To improve the results you can change the measurement parameters 1 Press the Properties button to open the Properties dialog box 2 Use the different tabs in this dialog box to make the required settings Parameters tab These settings are used while the measurementis running Changes require to ru
431. oftware Generated PRBS Procedures The recommended way to load the data patterns to the pattern generator and error detector is to use the Pattern Select Form dialog box See Loading Patterns to the Pattern Generator and Error Detector on page 73 for details Software Generated PRBS Reference Use the Pattern Select dialog box to select the pattern s you want to write to the pattern generator and the error detector See Select Pattern Dialog Box on page 89 for details Hardware Generated PRBS Hardware Generated PRBS Concepts This section describes the basics of hardware generated PRBS 2 n 1 patterns When to Test with Hardware Generated PRBS Patterns You can use the Serial BERT s hardware generated PRBS patterns if you require simple random traffic for testing your device The 2 n 1 odd pattern lengths are not related to the divided clock rate When using an oscilloscope this will ensure that the random contents of the bit stream is sampled The spectral contents of PRBS 2 n 1 patterns and the special property of decimation may also be useful in many applications for example for multiplexer or demultiplexer testing Agilent J BERT N4903B High Performance Serial BERT 93 3 Setting up Patterns NOTE 94 How the Hardware Generates PRBS 2 n 1 PRBS is an inverted hardware generated pattern that is created by a series of shift registers with adjustable feedback The example below shows the register config
432. on 532 Off On 531 Transition Density 180 Trigger Error Detector 203 Trigger Out Error Detector 169 Trigger Ref Clock Pattern Generator 145 Trigger Ref Clock Out port 123 Troubleshooting 469 U Uncertainty Band 246 User pattern 62 User Defined Sequences 96 V Vampt 136 Variable Mark Density 92 Vhi 135 View Tab Eye Opening 298 Output Levels 274 Output Timing 256 Spectral Jitter 357 571 Index View tab Web Server 551 Fast Eye Mask 314 Vlo 136 Vof 135 X Voltage of measurement point 313 Xover 137 W Waveform 320 572 Z Zero SubstitutionPattern zero substitution 92 Zoom Function Eye Opening 300 Output Level 276 Output Timing 259 Spectral Jitter 359 Agilent J BERT N4903B High Performance Serial BERT
433. on both sides So the total number of displayed eyes is 1 5 Qui This option displays 2 0 eyes on the graph The transitions of the complete eye are placed such that they show 0 5 eyes on both sides So the total number of displayed eyes is 2 0 View Calculate Measurement Setthe criteria for calculating the eye width eye height JPP JRMS cross voltage Parameters according to the BER threshold selected Errors To see 0 errors select the option This gives the results according to the last measured contour for the current measurement This is more significant in the case of deep BERs BER Threshold Enter the BER threshold at which the five results eye width eye height JPP JRMS and cross voltage will be calculated The show checkbox displays the contour at which the results are calculated Transition Time Set criteria for transition time 10 90 Measures horizontal scan from 1096 to 9096 of the signal amplitude 20 80 Measures horizontal scan from 20 to 80 of the signal amplitude Eye Boundaries Set the criteria for eye boundaries Total is 100 of the display screen These are applied for the level related parameters for example 0 level 1 level Signal to Noise ratio Eye Amplitude and so forth Left Boundary Agilent J BERT N4903B High Performance Serial BERT 333 6 Advanced Analysis 334 Eye Width Height Table Number format Optical Parameters Power Unit Extinction
434. on the Config window It will disable the multiplexer function and the M8061A menu item disappears Ensure that the timing adjustment has to be run at least once when either the N4903B M8061A or the cables connecting both are being exchanged in the test setup Also ensure that the timing adjustment has to be repeated whenever the operating temperature differs by more than 5 Celsius from the temperature at the previous timing adjustment Controlling N4916B Clock Multiplier After the Clock Multiplier function has been enabled through the Config window the following parameters are visible in Clock Multiplier DATAOUT window 1 Select Track PG Clock Frequency check box to toggle between frequency and multiplier provided by Pattern Generator and the input frequency and multiplier provided in the Clock Multiplier window If the Track PG Clock Frequency check box is selected the frequency and multiplier text boxes will be disabled and the values will be provided by Pattern Generator If the Track PG Clock Frequency check box is not selected then you can provide the values of the input frequency and multiplier For more information refer to the section Track PG Clock Frequency on page 54 If necessary change the termination voltage by entering a new value You can use the front panel knob to change the value For more information refer to the section Termination on page 55 Agilent J BERT N4903B High Performance Serial
435. on value specifies 75 p p value Change the downspread value the upper frequency remains unchanged while the bitrate is adjusted The deviation value specifies the p p value For some setups the I 0 modulator s range limitations require to use centerspread instead of downspread and adjust the bitrate and deviation accordingly To ease the handling of this setup the SSC Advanced Edit Panel allows to edit the upper center and lower frequency Adjust Buttons 998 955 Mb s 0 1225 96 If the upper or lower frequency is change the bitrate will be adjusted according to the selected deviation The center frequency corresponds to the configured bitrate The adjust buttons allow to adjust the corresponding frequency by the specified ppm value Please note that both the upper and lower frequency and the frequency adjustment of all three frequencies are not reflected in the firmware The GUI just calculates the resulting bitrate and writes it to the firmware Deviation The Deviation of the clock rate The range of SSC deviation are as following Downspread with deviation of 0 0 5 96 Centerspread with deviation of 0 0 5 96 Upspread with deviation of 0 0 5 96 The deviation in Upspread and Downspread is the p p value while in Centerspread MOTE it is 1 2p p Frequency The deviation Frequency The SCC provides the frequency range of 10 kHz 100 kHz Type You can click on the corrosponding button to
436. onnecting the DUT Concepts on page 29 tells you how you should connect your DUT to the Serial BERT Introduction to the Serial BERT Introduction to the Serial BERT Concepts These topics explain some of the basics of BER Serial BERT and BER testing with the Serial BERT Important Information about the Pattern Generator Outputs The pattern generator output ports must be terminated with 50 O if they are not connected Data Out Data Out Clock Out iie Agilent Technologies 11 1 12 Planning the Test Clock Out e Aux Data Out Aux Data Out Trigger Ref Clock Out Trigger Ref Clock Out Termination of output ports improves the test performance Important Information Regarding Security The Serial BERT is a PC based instrument with a standard Windows operating system As such it is subject to the same security protection measures as any other PC See the Microsoft web site for more information regarding data security http www microsoft com security default mspx Every user must have administrator privileges to run the Serial BERT firmware Otherwise the firmware does not have access to the instrument s hardware In case of any serious malfunction in the operating system the Agilent Recovery System can be used to repair the system See Agilent Recovery System Procedures on page 525 Understanding BER Bit error ratio BER is the most fundamental measure of system performance It is
437. onstant Rate 158 External Signal 158 Manually 157 Error Location Capture measurement 302 Error Out port 169 Estimated Total Jitter 261 Example of an Output Level Measurement Changing Properties 269 Improving the Display 268 Running the measurement 266 Example Patterns 70 External Clock Source 143 External Data Source Connections 33 Eye Diagram 193 Eye Edge 194 Eye Height 393 Eye Measurements 392 Eye Opening measurement 297 Eye Width 393 F False Synchronization 209 Fast Eye Mask measurement 307 Fast Total Jitter at BER 254 Fast Total Jitter measurement 241 Explanation 246 Numerical results 262 Result Display 262 FER Results Window 368 FER Status Indicators 373 File Management 558 Filler Symbol Ratio 182 Flip Flop Connections 31 Frame Error Ratio 182 FSR Results Window 369 FSR Status Indicators 373 G G 821 Measurements 381 Gate In port 169 Gate In signal timing 213 Gating 138 Gauss curve 238 Gaussian Marker introduction 237 Output Levels 274 Output Timing 242 Generated Patterns 93 GPIB Address 529 Graph Tab Eye Opening 300 Output Levels 276 Output Timing 258 Spectral Jitter 358 Agilent J BERT N4903B High Performance Serial BERT Grid 320 Hardware generated patterns 61 High Level 277 High Level Std Dev 278 Illegal Symbol Ratio 182 Input and Output Ports Error Detector 168 Pattern Generator 122 Input Range Error Detector
438. or All Errors To display all errors Errors if 0s Expected To display the errors if 0 is expected but 1 received e Errors if 1s Expected To display the errors if 1 is expected but 0 received You can set the Timing Unit for the display of the relative time of the measurement points on the Parameters tab Choose between Unit Interval and Seconds to select the unit Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Table Number Format You can select the number of Decimal Places to be displayed in the table Eye Diagram Eye Diagram Concepts The Serial BERT provides quick design analysis with the Eye Diagram capability The Eye Diagram allows a quick check for the DUT s signal output and determines the signal quality Due to the higher sampling depth of a BERT value the eye contour lines display the measured eye at a deeper BER level for accurate results The Eye Diagram feature provides the following measurement capabilities Eye Diagram Waveform BER Contour Lines Automatic Measurement for the most relevant eye parameters Eye Height Eye Width Jitter P P Jitter RMS Cross Voltage Markers EyeMask The Eye Diagram generates a three dimensional graph of the bit error rate BER This measurement helps in determining and analyzing the quality of the DUT s signal output The Eye Diagram results comprise of voltage y time x and BER z The Eye
439. or 1 First determine whether the pattern generator and error detector should use the same pattern Click Pattern Select menu item from the Pattern submenu Pattern Select The PatternSelectForm dialog box opens 2 Select Error Detector Pattern tracks the Pattern Generator Pattern if you want both modules to receive the same pattern 3 Click OK to close this dialog box 4 The pattern editor now displays the following buttons in the toolbar If you chose that both the error detector and pattern generator use the same pattern FG Ebj Loads the current pattern to both PX ium the pattern generator and error detector If you allowed different patterns for the two modules Fs Loads the current pattern to the SOT pattern generator To PG His Loads the current pattern to the ED d To ED error detector 10110 Captures data of specified length ELE from the error detector Capture Agilent J BERT N4903B High Performance Serial BERT 75 3 Setting up Patterns NOTE Once the pattern of a specific length is captured from the error detector it can be sent back to the error detector and the pattern generator For later use of this captured pattern it should be saved using SaveAs option The user will be prompted to save the captured pattern if a new pattern is loaded and the same pattern is recaptured User Patterns Reference 76 The elements of the pattern editor and all dialog boxes that can be opened from the pattern ed
440. or 0 and then recover the clock from the detected transitions Clock Data Recovery CDR is a special kind of Phase Locked Loop PLL which recovers clock signal from of a data stream It is a regulatory loop which synchronizes the local oscillator with an external reference in this case the incoming data stream Phase Locked Loop A PLL has three parts a phase detector a loop filter and a voltage controlled oscillator VCO The phase detector has two inputs and one output which is proportional to the phase difference of the inputs The loop filter is a low pass filter which attenuates the higher frequencies from the output of the phase detector The VCO is an adjustable oscillator which changes the output frequency depending on its input voltage The diagram below shows a simple PLL PHASE DETECTOR LOW PASS FILTER IN OUT One of the most important characteristics of a PLL is its loop transfer function The loop bandwidth is defined as the integrated magnitude of the PLL s frequency transfer function over the entire frequency spectrum The loop bandwidth describes how the regulatory loop tracks the VCO to a sine wave FM modulated input signal Above the bandwidth the loop cannot track such a modulation completely and thus the response to the modulation is attenuated The other loop parameter is peaking This describes how much a modulation is exaggerated mostly close to the loop bandwidth Agilent J BERT N4903B High Pe
441. or indicators at the top of the user interface should show red and the resulting BER should be zero How to Execute the DUT Output Timing Jitter Measurement To run the DUT Output Timing Jitter measurement 1 Switch to the Analysis area If the Output Timing screen is not yet displayed press the Output Timing icon 2 Press the Start button to execute the measurement The measurement software runs the measurement and displays the results The following illustration shows the bathtub curve of the measured bit error rate and in the tabular view the calculated results for the Optimal Sampling Delay the Skew and the Phase Margin 1 00840 1 Trace 39 Points Ul 401 878 ps All Errors 1 00e 1 1 00e 2 1 00e 3 1 00e 4 1 00e 5 1 00e 6 1 00e 7 Logarithmic 0 62 UI 0 37 UI r 0 12 U Relative 0 50 UI 0 25 UI 0 120 0 37 UI 0 62 UI I 0 25 Ul 0 50 Ul Total Jitter Total J 0 00 UI How to Optimize the View of the Results After you have run a measurement the resulting graph and the calculated numerical values are displayed To improve the results you can change the measurement parameters 1 Press the Properties button to open the Properties dialog box 2 Use the different tabs in this dialog box to make the required settings Parameters tab Agilent J BERT N4903B High Performance Serial BERT 251 6 Advanced Analysis 252 These settings are used for data collection Changes here require the tes
442. or Count 0 m Errored 1 s Ratio 0 00000000 0 00000000 Errored 1 s Count a o a Errored O s Ratio 0 00000000 0 00000000 Errored O s Count 0 The Accumulated Results window for Bit Comparison without PCle3 SKPOS does NOTE not contain tables for G 821Measurement Internal Results and Burst Results Agilent J BERT N4903B High Performance Serial BERT 389 7 Evaluating Results 390 Ratio Graph This graph displays the delta symbol errored 1 s ratio delta symbol errored 0 s ratio and total delta symbol error ratio at data points over the entire accumulation period For more information see Ratios Graph on page 379 Accumulated Results In addition to the values displayed in the BER Accumulated Results see Accumulated Results on page 380 the following additional values are listed Modified SKPOS Count The total number of SKPOS count whose length has been changed by DUT in a time interval is displayed here Expected SKPOS Count The total number of Expected SKPOS Count received in a time interval is displayed here SKPOS Bit Count The total number of SKPOS bits received in a time interval is displayed here Expected SKPOS Bit Count The total number of Expected SKPOS bits received in a time interval is displayed here SKPOS Bit Ratio The ratio of the number of SKPOS bits to the number of bits received in the current or last completed accumulation period specified by the gate period is displayed
443. or insertion is possible click Error Add Setup menu item from the PG Setup submenu to open the Error Add Setup dialog box J TL Error Add Setup 3 Select Off and click OK to close the dialog box Agilent J BERT N4903B High Performance Serial BERT 157 4 Setting up the Pattern Generator Using an External Signal for Inserting Errors To configure the Serial BERT to insert a single error into the output stream according to an external signal 1 Connect an external instrument to the Error Add port The signals received at this port must be TTL compatible 2 Click Error Add Setup menu item from the PG Setup submenu Tii Error Add Setup 3 Select External Error Add and click OK to close the dialog box Now upon a rising edge of the signal at the Error Add port an error will be generated in the output stream by flipping a single bit within a bit block of 128 bits resulting in a maximum BER of 1 128 or 10 You can use an external signal to set up a bit error rate to fit your needs The frequency of the external signal can be calculated by multiplying the desired bit error rate by the clock frequency TIP If for example you want a bit error rate of 3x107 and you are running a test at 12 5 G b s the frequency of the external signal is 3x107 x 12 5x10 3750 Hz Using the Serial BERT to Insert Errors To configure the Serial BERT to automatically insert errors in the output stream to achieve a desired bit error rate 1 Cl
444. or more measurement points for the measurement Relative Time and Voltage In the table at the bottom of the Parameters tab you can change measurement points Enter the Relative Time and Voltage according to the above settings Pass Fail Tab The Pass Fail tab of the Properties dialog box allows you to specify the criteria to decide whether the DUT passes or fails the test You can change pass fail criteria without rerunning a test The software only uses the criteria to rate the results of a measurement NOTE The pass fail criteria do not control measurement execution The measurement run will be completed even if the measurement fails one or more of the criteria The Fast Eye Mask measurement has only one pass fail condition The bit error rate If desired enable the BER Threshold and set a suitable threshold Agilent J BERT N4903B High Performance Serial BERT 313 6 Advanced Analysis 314 NOTE Analyze Grid The BER Threshold will usually be 0 since you want the measurement to fail if a single bit was received in error The pass fail threshold applies to all measurement points An e icon indicates all measurements where the bit error rate is higher than this threshold as shown in the figure below Terminal 1 2 3 4 Relative Time Vottage abs E Osem Quis o o 0o View Tah The following parameters can be set on the View tab of the Properties dialog box You can analyze f
445. ou to the following screen LogFileFormatDialog m Format Locale English United States Character Delimiter Save to File BER C All BER Compared bits Errors m Error Options All of expected Os and 1s of expected 1s of expected Os x m You can select the Format Save to File and Error Options as per requirements The Save to File option specifies either saving only the BER values or saving the BER the Compared Bits and the Errors Agilent J BERT N4903B High Performance Serial BERT 331 6 Advanced Analysis 332 Eye Diagram NOTE Persistence Center to The Error option gives the user the choice of saving the BER the Compared Bits and the Errors for different combinations of expected 1s Os and all Reference The Eye Diagram measurement returns the results in graphical and numerical forms The following section explains the measured parameters and the displayed options that are specific to these measurements Additionally some information is provided to explain the theoretical background to these measurements Parameters that can be set on the various tabs of the Properties pop up Parameters If the parameters on this page are modified run the measurement again to update the results While the measurement is running the parameter settings cannot be changed Set the criteria for Persistence Infinite With Infinite the
446. our system and click Have Disk button The instrument will then prompt you to locate the drive for printer driver Type the path where the printer driver is copied 5 Follow all other instructions on the screen to finish adding a printer Agilent J BERT N4903B High Performance Serial BERT 555 10 Customizing the Instrument 556 NOTE NOTE Removing a Printer When you disconnect a printer from the instrument and no longer intend to use it you can remove the printer driver and increase the free disk space available on the instrument s internal hard disk drive To remove a printer driver 1 On the File menu click Print Setup then Configure Printer This opens the Printers dialog box 2 Click the printer driver you want to remove 3 Inthe File menu in the Printers dialog box select Delete A dialog box will appear to confirm the removal of the printer driver 4 Click Yes to confirm your selection Setting the Default Printer You can set the printer you use most often as the default printer Then all printouts will be sent to the default printer unless you explicitly select another printer To specify the default printer 1 On the File menu click Print Setup then Configure Printer This opens the Printers dialog box 2 Choose one of the currently mapped printers If the printer you want to use is not yet included in this list see Adding a Printer on page 555 3 Onthe File menu in the Printers dialog box
447. ove as the best fit line through the calculated points This method makes the calculation reproducible For details see Q factor Results on page 279 Furthermore the following parameters can be set on the View tab of the Properties dialog box Calculate You can calculate measurement parameters for BER Threshold Agilent J BERT N4903B High Performance Serial BERT 215 6 Advanced Analysis Table Number Format Scale Ratio Unit Markers Zoom Show Measured Points 216 This is the bit error rate threshold at which the Threshold Margin is determined It is also the upper threshold for the Q factor calculations The BER Threshold is displayed in the BER vs Threshold graph There it can be positioned with the mouse or your finger if you are working directly on the Serial BERT Min BER for Q This is the lower threshold for the Q factor calculations You can select the number of Decimal Places to be displayed in the table Graph Tab On the Graph tab you can use the several options to optimize the graphical display according to your needs Choose between Logarithmic and Linear to select the scale for the display s x axis The scale of the OBER vs Threshold graph is always linear Ratios in the numerical results such as SNR can be displayed linearly or logarithmically in dB To analyze the graphs at a particular point you can use the markers Optionally you can display all related values for the
448. ows you to determine more than just the actual levels The Output Levels measurement also calculates the Q factor a measure that describes the quality of the received signal and derived values These results can assist you in characterizing the device They can also enable you to predict very low bit error rates that would take a long time to be measured Output Levels Procedures This section shows you how to set up and perform an Output Levels measurement As an example we measure the output level behavior of a shielded cable This requires the following steps Preparing the measurement see How to Prepare the Output Levels Measurement on page 265 e Executing the measurement see How to Execute the Output Levels Measurement on page 266 Improving the Display see How to Improve the Output Levels Display on page 268 Changing the Properties see How to Change the Output Levels Properties on page 269 Using the different views see How to Use the Different Views on page 269 How to Prepare the Output Levels Measurement To prepare an Output Levels measurement to test a shielded cable 1 Disable the pattern generator outputs by pressing the OV Disable button in the PG Setup gt Data Output screen 2 Use a shielded cable to connect the pattern generator s Data Out port and the error detector s Data In port 3 Terminate all non connected pattern generator output ports with 50 Q Agilent J
449. pecified Start Bit and Block Length are counted The following legend applies for the above graphics Evaluated bits Not evaluated bits Audio Signals Audio Signals Concepts The use of warning tones is convenient when you want to monitor the instantaneous BER over a longer time without constantly watching the display With the Audio dialog box you can adjust the sound volume as well as some conditions when to play a sound 224 Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 Audio Signals Procedures To set up the warning sounds do the following 1 Click Audio menu item from the ED Setup submenu 2 Check the Audio on checkbox to switch on the sounds To switch the audio signals on and off you can also use the Audio On Off front TIP panel button 3 Use the Main Volume slider to adjust the volume After adjusting the slider a test tone is played at the new volume NOTE If no sound is heard please check whether the system volume is muted 4 Click the respective radio button if you want to play the sound whenever the BER exceeds the threshold specified in the BER Alarm Threshold field Enter an appropriate value in this field You can always adjust the BER Alarm Threshold by clicking and dragging the small TIP yellow marker in the BER bar in the top left corner onAll Error Rates greater than zero 5 Click OK to finish the audio setup Audio Signals
450. pening The eye becomes unsymmetrical The Serial BERT provides the Interference Channel option J20 to simulate ISI For more information see Intersymbol Interference on page 399 Level Noise Level noise affects the voltage amplitude of the eye opening Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 The Serial BERT provides the Interference Channel option J20 to simulate level variations by adding sinusoidal interference For more information see Crosstalk on page 400 PLL Performance Test Testing a PLL or CDR is not complete until the entire frequency range of the PLL has been checked under worst case conditions For this purpose the Serial BERT provides the Jitter Tolerance Characterization measurement For more information see e Jitter Tolerance Characterization Concepts on page 433 PLL Standards and Compliance Tests The Optical Internetworking Forum OIF as well as other institutions have proposed standards for testing the performance of data receivers and receiver circuits in the presence of jitter For this purpose the Serial BERT provides the Jitter Tolerance Compliance measurement This measurement is available as an option option J12 which can be ordered together with the instrument or retrofitted For more information see e Jitter Tolerance Compliance Concepts on page 446 Understanding the Types of Jitter The standards for jitter tolera
451. plitude is tolerable Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests Amplitude The maximum peak to peak Amplitude is limited by the free capacity of the chosen delay line Frequency The maximum Frequency of the periodic jitter source depends on the chosen waveform Refer to the technical specifications Waveform You can switch between sine rectangular or triangular jitter source waveforms Periodic Jitter 2 Constant Amplitude Sweep Parameters The Periodic Jitter 2 Constant Amplitude Sweep is characterized by Amplitude Frequency Range Waveform Sweep Time Nr of Steps A graph indicates whether the chosen setting of frequency and amplitude is tolerable Agilent J BERT N4903B High Performance Serial BERT 423 8 Jitter Tolerance Tests 424 Amplitude Frequency Range Waveform Sweep Time Nr of Steps NOTE 15000 MHz u c z The maximum peak to peak Amplitude is limited by the free capacity of the chosen delay line The stop frequency has to be higher than start frequency and the range should be in accordance with the selected waveform You can switch between sine rectangular or triangular jitter source waveforms You can specify the duration for sweeping the specified frequency range once You can specify the number of steps to fulfill a complete sweep The start and stop values are included The valid range is between 2 to 100 The settin
452. provide a small amount of jitter to the clock signal Use of a precision clock A precision clock with very low phase noise can be used to enhance the instrument s performance This is especially interesting for long term measurements Testing at bit rates lt 620 Mb s The Serial BERT s internal clock cannot be used for bit rates below 620 Mb s Bit Rate Range The Serial BERT provides bit rates from 150 Mbit s up to 12 5 Gbit s depending on the instrument s options The Bit Comparison without USB3 1 SKPOS error ratio mode supports the bit rate up to 10 35 Gbit s However if you want to use bit rate greater than 10 35 Gbit s then switch to 8B 10B Symbol Comparison or Bit Comparison without PCle3 SKPOS error ratio mode and then go to Bit Comparison mode For details refer to Setting up Error Ratio on page 185 However several specific properties and limitations need to be taken into account when working at low bit rates The limitations apply to the instrument according to the following hysteresis curve e Ifthe bit rate falls below 615 Mbit s the limitations apply e Ifthe bit rate exceeds 620 Mbit s the limitations no longer apply The following figure clarifies the behavior in the range between 615 Mbit s and 620 Mbit s Limitations do not apply in this range 620 615 Bitrate Mbit s Limitations apply in this range Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern G
453. quency range to be used for the test from 1 kHz up to some hundred MHz You can hence measure only a portion of the standard or points outside the standard Points outside the standard will always pass the test If a standard prescribes a certain range for example from 2 kHz to 20 MHz you can select a checkbox to adjust the jitter frequency range to the loaded standard You can also specify the number of points to be measured By default the measurement points are evenly distributed along the logarithmical frequency axis You may wish to measure the BER exactly at those frequency points where the specified jitter amplitude changes By selecting a checkbox the distribution of the measurement points can be adapted to the loaded standard and the BER at the points of interest will be measured Margin It is possible to increase the jitter amplitudes of any standard by adding a margin This increases the stress on the device under test The margin is set in percent A margin of 2596 for example has the effect that each amplitude value is multiplied by 1 25 The following figure shows an example of a measurement with a margin of 50 and sample point adaptation Properties Start 1 kUI CEl amp Gb s Short Reach Standard Margin LETTA z O Passed cr x Failed 100 UI e ee suaka a J BERT Capability Amplitude 10 mUI 1 1 1 lkHz 10 kHz 100kHz 1 MHz 5 MHz 20M Frequency Date Time Margin Result Cond
454. r Pattern dialog 4 Use this dialog box to locate and open the desired pattern You can select files of the following types 71612A BERT Pattern Files dat e 86130A Pattern Files binary hex symbol 01 ptrn ASCII Text Pattern Files txt Editing Patterns in Sequencer Pattern Editor To edit a user pattern for second channel 1 Click Sequence Editor menu item from the Pattern submenu Sequence E ditor 2 Click Edit to launch Set Data Block Parameter dialog for second channel 3 If no pattern is loaded in the editor first open a pattern file as described in Opening Existing Patterns in Sequencer Pattern Editor on page 109 or create a newone as described in Creating New Patterns in Sequencer Pattern Editor on page 108 4 Click Edit Pattern s to launch Pattern Editor for second channel 5 Use the standard keyboard functions to edit patterns at the position of the cursor 6 In the toolbar click Apply to confirm the new changes Agilent J BERT N4903B High Performance Serial BERT 109 J Setting up Patterns 7 Inthe toolbar click Discard to reject all the new changes tj Exporting Patterns in Sequencer Pattern Editor To export the current pattern under a new name 1 In the toolbar click Export icon The Export Trace dialog opens 2 Use this dialog to locate and save a copy of the current pattern See Export Trace on page 118 for details Sequencer Pattern Editor Reference 110 The e
455. r moving to the next sample point Number of Compared Bits After this number of compared bits the measurement stops for the current sample point and moves to the next one Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Number of Errors After this number of errors the measurement stops for the current sample point and moves to the next one This allows you to speed up the measurement You can switch off this option if only the number of compared bits is important NOTE The measurement moves to the next sample point when the first of the two criteria is reached Both numbers are ignored if the Fast Total Jitter measurement is selected Set the criteria for the sample delay Resolution Specifies the time distance between sampling points A smaller value yields more sampling points in a unit interval You can enter the resolution in UI or ps ns s The timebase of the display is set on the View tab If the resolution used for the measurement is not high enough the bathtub curve does not clearly show the edges For example you may wish to change the Resolution from 0 01 to 0 005 and run the measurement again Optimization You can choose between Edge Resolution Optimization Fast Total Jitter at BER or None Edge Resolution Optimization Turns the resolution optimization on or off If this option is enabled the Serial BERT intelligently sets a resolution so that there are more sampl
456. r of the screen Error Add j When inserting an error the BER SER FER FSR ISR DER cBER should increase for a short time and then return to the initial value Agilent J BERT N4903B High Performance Serial BERT 363 7 Evaluating Results 364 How to Monitor BER Results There are different ways how you can monitor instantaneous BER values 1 Listen to the BER warning tones See Audio Signals Procedures on page 225 for details 2 View the BER Results window which is described in BER Results Window on page 366 3 View the BER bar and errors indicator See Status Indicators on page 372 for details How to Monitor 8B 10B Comparison Results There are different ways how you can monitor instantaneous 8B 10B Comparison values 1 Listen to the 8B 10B Comparison warning tones See Audio Signals Procedures on page 225 for details 2 View the 8B 10B Comparison Results window which is described in 8B 10B Comparison Results Window on page 367 3 View the 8B 10B Comparison bar and error indicators See Symbol Lock Indicator on page 192 and 8b10b Error Indicator on page 192 for details How to Monitor Bit Comparison without PCle3 SKPOS Results There are different ways how you can monitor instantaneous Bit Comparison without PCle3 SKPOS values 1 Listen to the Bit Comparison without PCle3 SKPOS warning tones See Audio Signals Procedures on page 225 for details 2 View the Bit Comparison without PCle
457. rameters View Graph Mask Eye Measurements Persistence Infinite Center ta Number of Eyes 2 Dui Cancel Help View tab 322 Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 The view tab options set the parameters to measure and choose units for the displayed results in the numerical table Refer to View on page 333 for the detailed options Properties Parameters View Graph Mask Eye Measurements M Calculate Measurement Results for C DEnors BER Threshold 12 3 Show M Transition Time M Eye Boundaries C 10 90 Left Boundary 40 x 20 80 RightBoundary 60 M Eye Width Height Width at crossing point C custom Height at 50 z M Table Number Format Decimal Places 2 M Optical Parameters Power Unit Dark Level O 0 mv Conversion Gain 0 0010 wy dBm C W ee Ratio Unit Ratio CE Cancel Apply Help Graph tab The graph settings give the display parameters like timing units waveform contour grid and markers Agilent J BERT N4903B High Performance Serial BERT 323 6 Advanced Analysis Properties Eye Measurements tab The Eye Measurements tab lists out all the measurements The user can select specific results 324 Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Properties Parameters View Graph Mask Eye Measureme
458. raph 274 Bias Tees 130 Bit Comparison without PCle3 SKPOS Results Monitoring 364 Bit Comparison without USB 3 1 SKPOS Results Monitoring 364 Bit Rate Different In Out Rates 30 Pattern Generator Setup 139 Bit Rate Range Error Detector 176 Bit Recovery button 202 Bit Recovery Mode enable 197 explanation 195 Bounded uncorrelated jitter BUJ 398 Bracketing Approach 248 BRM 195 BUJ 398 Burst Results 383 Burst Sync Mode 211 Agilent J BERT N4903B High Performance Serial BERT C Cancel button 201 cBER Results Window 368 cBER Status Indicators 373 CDR 174 Clock Data Recovery 174 Clock Falling Edge button 199 Clock In port Error Detector 169 Clock In termination 171 Clock Input 122 Clock Multiplier 54 Clock Out port 123 Clock Rate Error Detector 179 Clock Rate Indicators 145 Clock Setup Error Detector 178 Clock Source External 143 Internal 143 Clock Termination 137 Color Tab 232 Complement 172 Confidence Levels 27 Connecting the DUT 29 Connections Amplifier 31 DCA 32 Different In Out Rates 30 Differential Inputs 30 External Data Source 33 Flip Flop 31 M8061A 35 MUX DEMUX Pair 32 N4876A 34 N4916A 34 Oscilloscope 32 SONET SDH Receiver 33 567 Index Constant Errors 473 Contour 320 Contour plot 288 Creating New Pattern 71 Criteria for Moving to the Next Measurement Point 312 Crossover 127 CrossTalk 107 Crosstalk 400 D Data Center button 20
459. re this pattern have been sent and not every time the pattern is sent So for example if you have a standard pattern that is 384 bits long this pattern must be loaded into the memory four times to reach the 512 bit boundary A trigger would then be generated every 4x384bits or 1536bits Example Patterns The Serial BERT provides a set of example user patterns These patterns mimic real data packets and standard stress patterns The directory for the example user patterns is C lt instrument model PatternsNDemo Any changes you make to the patterns in this folder are non recoverable there is no backup folder To avoid inadvertent changes it is recommended that you copy the demo files to a different directory before loading them For a complete list of available patterns and their description see C lt instrument model gt Patterns Demo Description pdf For more information on the recommeded pattern for specific purposes see When to Use Which Pattern on page 64 Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 User Patterns Procedures You can do the following tasks in the Edit Pattern window Creating New Patterns on page 71 Opening Existing Patterns on page 71 Editing Patterns on page 72 Saving Patterns on page 72 Loading Patterns to the Pattern Generator and Error Detector on page 73 Loading Patterns Directly from the Pattern Editor on page 75 Creating New Patt
460. red PAM4 Mode Displays bit combinations of both channels as shown in the table below Table 17 Symbol Data Out Aux Data Out 0 0 0 1 0 1 2 1 0 3 1 1 The PAM4 Mode is only valid in Bin representation The Hex and Symbol NOTE E representations are disabled in this mode You need to press the described keys to configure Z in the hardware front panel NOTE for 00B Mode Agilent J BERT N4903B High Performance Serial BERT 117 3 Setting up Patterns 118 NOTE Export Trace The Export Trace dialog saves a copy of the currently edited pattern The Export Trace dialog is shown in the figure below ExportT race Select Trace Data Out C Aux Data Out C Both Description BERC 2SKP DR File name C1N4903BiPatterniMyHeader ptrn Browse Export Cancel Help The Export Trace dialog will look different for the different cases in respect to the selection of what shall be exported Data Out Aux Data Out Both Pattern A Pattern B Both Even Bits Odd Bits All Bits This dialog provides the following options Select Trace Depending on the type of the pattern that is currently being edited the Export Trace dialog is offering different options Standard single pattern Export either only the even or odd bits in the pattern or all bits in the pattern Alternate pattern Export either Pattern A or B as standard pattern or pattern A and B as an Alternate pattern Dual channel pat
461. rformance Serial BERT Setting up the Error Detector 5 bandwidth 3dB point The frequency is the modulation frequency and amplitude is the tracking of NOTE the PLL Data rate and loop parameters The data stream contains multiple frequencies and the CDR needs to know the expected data rate The user entered information is necessary to lock the CDR and to detect a loss in lock condition The entered frequency should be accurate and better than 100 ppm N4903B provides a CDR which has additional input parameters loop bandwidth transition density peaking and SSC deviation Loop bandwidth is the input parameter to set the characteristics of the loop Additionally the Peaking can be chosen from a drop down list Transition density affects the loop parameters and it must be either entered or measured Some standards define a loop bandwidth for a specific transition density The SSC deviation parameter widens the window for loss in lock detection and detunes the center of detection because SSC is generally asymmetric down spread CDR is not available together with Bit Recovery Mode BRM at data rates above NOTE 11 5GHz Threshold The regular threshold voltage is not only used to determine the optimum sampling for the data but also to perform measurements such as eye diagram or output level measurements it is not possible to use it for the clock recovery For this reason the clock recovery circuitry
462. rial BERT Advanced Analysis 6 Properties In the tabular view each of the calculated values will be marked with an e icon if it failed the test The following illustration shows an Eye Opening measurement that has failed the criteria for the eye opening voltage Agilent J BERT N4903B High Performance Serial BERT 297 6 Advanced Analysis Contour plot Pseudo Color Plot 298 1 Trace 9510 Points Ul 401 878 ps All Errors 600 0 mv 400 0 m 200 0 mv 0 0 v 200 0 mv 400 0 mv 600 0 mV scale 0 448 UI 0 149 UI 0 149 Ul 0 446 Ul 0 746 Ul Relative 0 597 Ul 0 299 Ul 0 000 Ul 0 299 Ul 0 597 Ul I E i Threshold Eye Optimal Sample 9 Serial Ki View Tab The graph shows either a Contour Plot a Pseudo Color Plot or only one curve for the selected bit error rate threshold The contour plot shows discrete lines of equal bit error rate just like the contour lines on a map The color of a line indicates the respective BER value This graph is useful to visualize in which areas the BER changes a homogeneous BER field will give you no lines at all 1500mv 1 E Ul 1 538 ns All Errors D 1 000840 ju 1200 mv d 1 000e 1 90 0 mv 60 0 mv a 1 000e 2 om d 1 000e 3 00v anon ayy 0002 4 60 0 mv Jam 1 000e 5 90 0 mv 1 000e 6 120 0 mv PRESS um E Leomar 150 0 mv Ls tem eo ia 0 at alii 000e 7 0 650 Ul 0 390 UI 0 130 UI 0 130 Ul 0 390 UI 0 550 UI 0
463. ring the Touchscreen Procedures 531 Configuring the Touchscreen Reference 533 Installing Software Licenses 535 Installing Software Licenses Procedures 535 Installing Software Licenses Reference 535 Installing Hardware Options 536 Installing Hardware Options Procedures 536 Installing External Instrument s 538 Installing External Instrument s Concepts 538 Installing the N4916A B Procedure 539 Installing the N4876A Procedure 544 Installing the M8061A Procedure 549 Installing Webserver 551 Customizing the Web Server Concept 551 Printing 555 Printing Procedures 555 Agilent J BERT N4903B High Performance Serial BERT File Management 558 File Management Procedures Preset Instrument State Preset Instrument State Procedures Self Test 563 Self Test Concepts Self Test Procedures Self Test Reference Index 567 Agilent J BERT N4903B High Performance Serial BERT 10 Agilent J BERT N4903B High Performance Serial BERT Planning the Test Planning the Test Concepts The following topics provide some information that can help you in planning tests with the Serial BERT Introduction to the Serial BERT Concepts on page 11 explains the basics of bit error rates BER and BER testing with the Serial BERT Which Test is Appropriate Concepts on page 25 provides you with some basic information about the types of testing you can do with the Serial BERT C
464. rization Concepts 433 Jitter Tolerance Characterization Procedures 438 Jitter Tolerance Characterization Reference 443 Jitter Tolerance Compliance 446 Jitter Tolerance Compliance Concepts 446 Jitter Tolerance Compliance Procedures 454 Jitter Tolerance Compliance Reference 462 9 Solving Problems Solving Problems Concepts 469 Setup Problems 469 Setup Problems Concepts 469 Measurement Problems 472 Measurement Problems Concepts 472 Other Messages 478 Other Messages Concepts 478 N4916B System Setup Calibration 480 N4916B System Setup Calibration 480 Agilent J BERT N4903B High Performance Serial BERT N4876A System Setup Adjustment 490 N4876A System Setup Adjustment 490 M8061A System Setup Adjustment 492 M8061A System Setup Adjustment 492 Problems with the N4916A 493 Problems with the N4916A Concepts 493 Problems with the N4916B 500 Problems with the N4916B Concepts 500 Problems with the N4876A 511 Problems with the N4876A Concepts 511 Problems with the M8061A 518 Problems with the M8061A Concepts 518 10 Customizing the Instrument Customizing the Instrument Concepts 525 Restoring the System 525 Agilent Recovery System Procedures 525 Configuring the Instrument 528 Configuring the Instrument Concepts 528 Configuring the Instrument Procedures 529 Configuring the Instrument Reference 531 Configuring the Touchscreen 531 Configu
465. rmance Serial BERT Planning the Test 1 How to use the Navigation Menu The navigation Menu comes with a pull down menu which provides a central launching point for the various controls of the Serial BERT NOTE The availability of the menu items depends on the available hardware and licenses The navigation Menu includes single menu items like File Utility and Help and group menu items like External Instument s Pattern PG Setup ED Setup Analysis Jitter and Results Each menu entry on the Menu button has a separator and an arrow sign gt gt on it While the separator indicates that the menu item can be directly opened by clicking on the group menu item the arrow sign indicates a horizontally expanding submenu associated to it The following figure explains the usage of the navigation menu Image indicates a Sepreator indicates group menu entry that the menu entry can be opened directly Arrow indicates a submenu Pattern Select gt A Edit Pattern on File e Sequence E ditor Instrument s Pattem PG Setup ED Setup No arrows indicates no submenu Analysis Jitter Clicking here opens the last viewed control of this group For example for Pattern it would be Sequence Editor Results Utility Help gt gt Each group menu items has an image assigned to it The image displayed on
466. rmance Serial BERT 531 10 Customizing the Instrument 532 On the Utilities menu click Touchscreen Off The Enable Touchscreen icon is placed in the upper right hand corner of the instrument display Ca Click the Enable Touchscreen icon to enable the touchscreen keyboard and mouse again Calibrating the Touchscreen In normal usage you should not need to calibrate the touchscreen display But if you feel that it needs to be realigned with the video display do the following 1 If the title bar of the Serial BERT user interface is not visible select Title Bar from the Utility menu Minimize the Serial BERT user interface window In the system tray next to the clock in the lower right corner right click the Touchscreen Settingsicon to open the menu with the available options Q ey 11 47 AM Select Calibrate then Device 1 and Normal A blank screen with a target appears Click the target to start the calibration When the calibration is complete the calibration window will close Configuring the Double Click You can adjust the behavior of the touchscreen display to simulate the double click action of a typical personal computer or trackball To configure the touch settings 1 Select Display Utility from the Utility menu This opens the Configuration Utilities dialog box Click Display to open the Display Area Configuration dialog box Click the Timed Double Touch slide bar to adjust the amo
467. rol dialog shows the current IP address of the instrument and this can be used to address the instrument on the internet explorer by the web client Enter a password which will be used for authentication Follow the steps below to access the instrument through the web server 3 You need to know the instrument IP address and the password Additional to the WebBrowser a JAVA runtime environment has to be installed on the client computer If the Java runtime environment is not installed on the client computer you cannot access the GUI of the instrument Open a web browser either Internet Explorer 4 0 or higher or Netscape 4 0 or higher Type in the instrument IP and enter the password when it prompts for it XX Agilent Technologies NCAR Wall IT f rea DOT STE s sa Welcome to your Web Enabled Serial BERT Information about this Web Enabled Serial BERT Instrument Serial BERT N4900A00DE NA4900AoxDE Instrument Description 13440 61 67 VISA TCPIP Connect String TCPIPO bbnfwtt 00 INSTR Id advanced Information about this Web Enabled Seria BERT Use the navigation bar on the left to access your Serial BERT and related information Click the Web Control button in the navigation bar on the left to start the application Agilent Technologies Inc 2005 From this page you can access all the functionalities of the J BERT Agilent J BERT N4903B High Performance Serial BERT Printing Customiz
468. rom the Config window corrosponding to N4916A clear the De Emphasis function check box It will disable the de emphasis function If you could proceed until here the connection of the user software to the N4916A is ok Check the Output of the N4916A Using a Scope 1 Set the Data offset to 0 V and the amplitude to 1 V Set the De Emphasis to 6 dB Connect the Output of the N4916A to an oscilloscope Connect the trigger output of the Serial BERT to the trigger input of the oscilloscope Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 Then the oscilloscope should show an output eye like the following f i f 3 reee iuf leu 2i m med tente If you see a display like above everything is ok If you see merely one fixed level either the N4916A or the output of the Serial BERT is defective Check the Output of the N4916A Using the Error Detector If you have connected the N4916A to a Serial BERT and have no oscilloscope at hand you can use the error detector for checking the output of the N4916A 1 Setupthe pattern generator as described above Set the Data Offset to 0 V and the Amplitude to 1 V Set the De Emphasis to 6 dB 2 Press ED Setup and open the Sampling Point Setup page 3 Set up the error detector Press the Edit button to enter the ED Input Setup Dialog Set the Input to Normal Set the Input Range to 1 V to 1 V Setthe Threshold to Averaging
469. ronym for Unit Interval One UI is the present clock period Sinusoidal Jitter is not generated by a delay line but by modulating the clock Sinusoidal Jitter and Spread Spectrum Clock cannot be used together with residual Spread Spectrum Clock because they share the same modulator NOTE The delay line display for SU SSC rSSC is set accordingly with the selected configuration Jitter Configuration Area The Jitter Configuration area allows you to select between the following jitter configurations the corresponding jitter sources are mutually exclusive Spread Spectrum Clock SSC or residual Spread Spectrum Clock rSSC NOTE That SSC and rSSC are licensed options J11 Spectrally distributed Random Jitter sRJ or Random Jitter RJ and Bounded Uncorrelated BUJ Configuration Sse Jitter Setup Area The Jitter Setup area consists of seven colored rows that keep you informed about the present jitter composition of Sinusoidal Jitter Spread Spectrum Clock Residual Spread Spectrum Clock Periodic Jitter 1 Periodic Jitter 2 Bounded Uncorrelated Jitter Agilent J BERT N4903B High Performance Serial BERT 405 8 Jitter Tolerance Tests 406 NOTE Random Jitter Spectrally Distributed Random Jitter External Jitter Source Connected to the Delay Ctrl input It depends on the jitter configuration which seven out of these nine jitter sources are displayed mE ipee c 20 UI 500 Hz 7 1
470. ror Detector 5 BER Location BER Location Concepts The Serial BERT can be run in BER location mode In this mode the errors are not counted on all bits of the data signal but on a particular bit or block of bits in the pattern You can specify which bit position s are considered during the measurement This allows for example to calculate the BER for the header bits or the payload of a data stream exclusively BER Location Procedures To set up the BER location mode do the following 1 Click the BER Location menu item from the ED Setup submenu 2 Select whether you want to locate errors on All Bits on a particular Block of bits or on a Single Bit in the pattern NOTE The Single Bit option is disabled in 8B 10B Comparison Bit Comparison without PCle3 SKPOS and Bit Comparison without USB3 1 SKPOS error ratio modes If you select Block or Single Bit enter the bit positions in the respective fields 3 Click OK to finish the BER location mode setup BER Location Reference The Bit Error Location Mode dialog box offers the following options NOTE The Single Bit option is disabled in 8B 10B Comparison Bit Comparison without PCle3 SKPOS and Bit Comparison without USB3 1 SKPOS error ratio modes Agilent J BERT N4903B High Performance Serial BERT 223 5 Setting up the Error Detector All Bits With this option all errors are counted Block With this option only errors occuring in the block between the s
471. ror rate BER dTh A Gaussian marker allows you investigate the peaks of this graph OQ from BER versus Threshold This graph shows the extrapolation of the Q factor and the optimum threshold level from a limited number of measured points Example Results The following illustration shows the BER versus Threshold graph of a simple Output Levels measurement 1 Trace 86 Points 600 0 mv 400 0 mv 200 0 mv DOv 200 0 mv 400 0 mV d um BER Threshold 1 000e 6 Logarithmic 1 000e 5 1 000e 3 1 000e 1 1 000e 6 1 000e 4 1 000e 2 1 000e Changing Properties If you change the measurement settings after the measurement has been run please note Parameters that affect the data capture Changes on the Parameters tab take only effect if you run the measurement again To remind you that the present results have not been obtained with the modified settings and that you should repeat the measurement the result display shows a yellow bar Parameters that change the display of the measured data Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Changes on the Pass Fail View Graph and Color tabs only affect the display of the results There is no need to repeat the measurement Variable Decision Threshold Method The method used by this measurement is commonly known as Variable Decision Threshold Method It provides a vertical analysis of the eye opening seen by the receiver This method all
472. rst mode is to measure multiple data bursts in quick succession In this way you can accumulate data for all bursts together You may wish to accumulate the number of bits necessary for a statistically valid BER measurement Refer to Introduction to Burst Sync Mode on page 211 and Test Times and Confidence Levels on page 27 for details When the accumulation period lasts for the duration of all bursts the gaps in error counting while the Gate In signal was high will not be included in BER calculations The gaps in error counting will not be considered error free periods If the gaps between data bursts are greater than 1 second or 100 ms in the NOTE DM measurement log file they may be visible in the accumulated results graph Burst Status The following states may be returned No Error No burst errors have occurred any of those listed below Gate Signal Too Long This error can occur if there are too many bits within a burst The limit is 4 Gbit At 13 Gb s this occurs roughly after 0 3 s slower frequencies have a higher gate in period This error has a higher priority than no unique 48bits No Unique 48 Bits Found For reliable synchronization a pattern must contain a unique 48 bit pattern the detect word If the current pattern does not have a detect word this error occurs If this status occurs the synchronization may be incorrect as could also the measured bit error rate There are standard patter
473. s Select an Input Range from 2 V to 0 V Set the Data Termination to 2 V Set the Alignment BER Threshold to 1E 6 Setthe Clock Setup to Clock Data Recovery to get the error detector s clock from the incoming data stream 8 Enable the pattern generator outputs by pressing the 0V Disable button How to Initiate an Eye Diagram Measurement To run the eye diagram measurement 1 Click the Eye Diagram menu item from the Analysis submenu TII Eye Diagram 2 Click the Start button to execute the measurement Start l The measurement runs and the eye diagram along with the Automated Eye Parameter Measurement get updated constantly The following graph shows the eye diagram 3 To stop the measurement click on Abort Abort l Agilent J BERT N4903B High Performance Serial BERT 321 6 Advanced Analysis How to Change the Default Settings of an Eye Diagram To achieve desired results you can change the measurement parameters 1 Click the Properties button to open the Properties dialog box Properties 2 Use the different tabs in this dialog box to make the required settings Parameters Graph View Mask Eye Measurements Parameters tab The Parameter settings used for the measurement can be changed only when the measurement is not running and if the settings are changed then you will have to re run the measurement Refer to Parameters on page 332 for the detailed options Properties Pa
474. s The first technology is the instrument itself The typical user scenarios for Web Enabled Instruments identify the user as being familiar with the instrument involved They normally use the instrument to gather data and make measurements The second technology is that of the World Wide Web and the web browser User data gathered across target user groups support the hypothesis that the great majority of Electrical Engineers and supporting technicians are familiar with the web and web browsers They frequently use one or more web browsers An Instrument Embedded Web Page is defined as a web page that is served up to a Web Browser from an instrument or product The server may reside either internally in the instrument or externally on a computer that is connected to and interacts with the instrument A Web Enabled Instrument or product is defined as one that is enabled to provide a Graphic User Interface GUI of the instrument or product in a World Wide Web browser The GUI provides the user an interface to view and control the instrument or product remotely J BERT provides the option of web server and installing this program enables users to access the instrument remotely over LAN The four basic steps to using a web browser 1 Open the web browser and enter the URL for the target instrument Agilent J BERT N4903B High Performance Serial BERT 551 10 Customizing the Instrument NOTE 552 2 The instrument web server responds
475. s calculated according to the BER Threshold set in the BERT from Calculate Measurement Parameters of the View Tab 11 Cross Voltage Crossing percentage is a measure of the amplitude of the crossing points relative to the 1 level and 0 level Note The Cross Voltage is calculated according to the BER Threshold set in the BERT from Calculate Measurement Parameters of the View Tab 12 Signal to Noise Ratio In signal to noise the signal is the information power of the signal indicated by the difference between the 1 and 0 level The noise is the combined standard deviations of the 1 level spread and the 0 level spread 13 Extiction Ratio Extinction ratio is the ratio of the 1 level power and the 0 level power of an eye diagram Note The Extinction Ratio is calculated using the Dark Level and Conversion Gain of the View Tab 340 Agilent J BERT N4903B High Performance Serial BERT Table 31 Advanced Analysis 6 S No 14 Agilent J BERT N4903B High Performance Serial BERT Parameter OMA Duty Cycle Distortion DCD Average Power Description The Optical Modulation Amplitude OMA is the difference between the 1 level power and 0 level power Note The OMA is calculated using the Dark Level and Conversion Gain of the View Tab This value is the difference between the period of a 1 bit and a 0 bit The mean of level 0 power and the level 1 power Note The Average
476. s delay line The PRBS polynomial can be chosen from a list Available are eight polynomials from 27 1 up to 2 1 Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Filter The bounded uncorrelated jitter source is equipped with three low pass filters with cut off frequencies at 50 100 and 200 MHz One of these filters is always active Random Jitter Parameters Random Jitter is characterized by Amplitude Filter Settings RAA Be The graph shows the possible and the current jitter spectrum which is defined by filter settings Amplitude The random jitter Amplitude must be entered as an rms root mean square value The rms value corresponds to the sigma factor o of the bell shaped Gauss distribution curve see also The Gaussian Marker on page 237 To modify the value click inside the text field and enter the desired value directly with the numeric keyboard or use the knob close to the numeric keyboard to fine tune the value Amplitude p p The peak to peak amplitude is calculated by multiplying the rms value with the crest factor which is currently fixed to 14 The Amplitude p p box shows how much of delay line capacity the setting consumes Agilent J BERT N4903B High Performance Serial BERT 427 8 Jitter Tolerance Tests 428 Filter The random jitter source is equipped with a 10 MHz high pass and a 100 MHz and 500 MHz low pass filter to limit the spectral range
477. s between two errors This is the inverse of the bit error ratio Off Select this option turn off all internal add rates and external error inputs This allows you to manually add single errors by pressing Error Add on the main display This flips a single bit in the output stream Pattern Alternation Pattern Alternation Concepts This section explains the basic purpose of alternating patterns Two alternate patterns can be stored in one pattern file Note that you can also generate a sequence of up to four alternating patterns see User Defined Sequences Concepts on page 96 Standard and Alternating Patterns With the Serial BERT you have full control on the data stream that is sent by the pattern generator to your device under test You can use standard patterns to specify a test data pattern that is sent to your DUT repeatedly Or you can use alternating patterns to specify two patterns that are sent to the DUT alternately To implement alternating patterns the instrument s user pattern memory is split into two portions one for pattern A and one for pattern B The differences between these patterns are Standard patterns consist of one single pattern that is repeatedly sent Any arbitrary pattern up to 32Mbit length can be set up However there are no Agilent J BERT N4903B High Performance Serial BERT 161 4 Setting up the Pattern Generator further configuration possibilities This pattern is sent as long as t
478. s results in a Duty Cycle Distortion DCD of 0 5 UI Aux Data Clock Divider This setting controls the pattern generator s Aux Data Out port The Aux Data Out port is differential always active and generates a fraction of the current clock The range of the divider is 2 3 4 5 128 Second Channel In Second Channel mode Data Out and Aux Out data channel can be setup with an independent selection of data and PRBS patterns including free choice of PRBS type and polynomial Multiplexer Mode In multiplexer mode Aux Data Out no longer generates a Sub Rate Clock signal but instead generates the required second data stream for a multiplexer that is connected to Data Out and Aux Data Out The multiplexer mode is automatically selected when N4876A is connected and the Multiplexer is enabled from the Config window Once the multiplexer is enabled itis necessary to perform system setup adjustment before it is practically used For more information see N4876A System Setup Adjustment on page 490 In multiplexer mode the pattern for the PG is automatically split between Data Out and Aux Data Out This mode is intended to stimulate multiplexing devices without Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 having to care for the correct setup of the required data streams The PG pattern is considered to be the serialized output of the connected multiplexer De Emphasis In De emph
479. s selected by default for symbol alignment The ReAlign button is disabled in Automatic mode During the Jitter Tolerance Characterization Compliance measurement the Automatic symbol alignment mode switches to Manual mode and restore back to Automatic mode when the measurement is finished aborted 12 From the Sampling Point Setup window set the sampling point For more information see Sampling Point Setup Procedures on page 196 13 For SER as an error ratio mode achieve the Symbol Lock condition The successful symbol alignment is indicated by a green Symb Lock indicator present at the lower pane of the Serial BERT s GUI For more information see Symbol Lock Indicator on page 192 14 For Bit Comparison without USB3 1 SKPOS as an error ratio mode achieve the EIEOS alignment condition The successful EIEOS alignment is indicated by a green Symb Lock indicator present at the lower pane of the Serial BERT s GUI However if the EIEOS alignment is lost or not done the manual re alignment can be done by pressing Re Align button as described in step 11 NOTE The Bit Comparison without USB3 1 SKPOS error ratio mode supports the bit rate up to 10 35 Gbit s However if you want to use bit rate greater than 10 35 Gbit s then switch to 8B 10B Symbol Comparison or Bit Comparison without PCle3 SKPOS error ratio mode and then go to Bit Comparison mode 15 If required use the Pattern Sync Setup dialog to synchroniz
480. s task bar click the Agilent IO Control icon 10 Agilent J BERT N4903B High Performance Serial BERT 501 9 Solving Problems 502 3 From the Agilent IO Control menu open the VISA Assistant 4 5 6 VAR VISA Assistant File Edit View Configure Help R USB0 5 spo ox09S7 0xSate Instrument Driver Formatted 1 0 Memory 1 0 Attributes DATA_OUT No Instrument Driver Configured Address String DATA OUT rsrcN ame Configure For Help press F1 15 01 16 If no VISA alias name has been assigned so far the VISA Assistant should indicate an instrument named UsblInstrument1 If the N4916B has been installed correctly the VISA Assistant should indicate an instrument named DATA OUT Click UsbInstrument1 or DATA OUT Click the Formatted 1 0 tab In the Instr Lang box Instrument Language enable SCPI Instr Lang C Custom C IEEE 488 2 SCPI Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 7 Click the IDN button This generates the identification query in SCPI format The instrument should return Agilent Technologies N4916B and the software revision YA VISA Assistant File Edit View Configure Help USB0 USB0 0x0957 0x5A18 Instrument Driver Formatted 1 0 Memory 1 0 Attributes DATA_OUT Clear History Show C Code IDN Device Clear Agilent Technologies N4916B Proto R2 003 0 8 0 2 20 z Set Timeout Read
481. s to test clock multiplier 2 using a scope 1 Connect PG Clock Out to Clock Multiplier Clock In 2 Terminate PG Clock Out with 50 Ohm 3 Termiate PG Trigger Out with 50 Ohm 4 Terminate Clock Multiplier Clock In with 50 Ohm 5 Set Clock Multipiler to 2 BER 0 000 Agilent J BERT N4903B High Performance Serial BERT 509 9 Solving Problems 510 N4916B Status LEDs This following section explains about the different LEDs that are available on the front panel of the N4916B De Emphasis Signal Converter Activity LAN Q POWER Q Table 34 LED Name Description Activity Flashes green when instrument is accessed on the remote interfaces Constantly orange during boot phase of the instrument LAN Constantly green when LAN is operable Constantly red when LAN is not operable Flashing green when device identification is enabled POWER Constantly red when instrument is OFF in power standby mode Constantly green when instrument is ON powered Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 Problems with the N4876A Problems with the N4876A Concepts When you have opened the External Instrument s Config window and have selected N4876A from the external instruments select Multiplexer under Enable Function column in the list the Multiplexer Connection dialog appears Cannot Connect Problems An error message may appear when you now press Enable Ifthis happens ensure that the N4876A has bee
482. select Export Data from the Analysis menu In the Export dialog box select the path and file name for the txt file Alternatively you can activate the Clipboard checkbox to copy the data to the clipboard In this case you can then paste it into any other application Press the Format Options button to open the Export Logfile Format dialog box Make further selections for the format of the exported data Press OK to close the dialog box Press OK to export the data to the specified destination How to Compare Measurement Results If you intend to repeat a measurement with different settings you may wish to preserve the current results for easier comparison This can be done with all measurements except Error Location Capture by copying the results 1 In the lower left hand corner of the measurement window select the measured terminal by tipping on it Select Copy from the Edit menu If you have a mouse connected to your Serial BERT you can directly copy the measurement via the context menu Open the Edit menu or context menu once more and select Paste This inserts a copy of the chosen measurement results For an Output Levels measurement this looks as shown below Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 2 Traces 172 Points 600 0 mv 400 0 mV 200 0 mv 00v 200 0 mv 400 0 mv 600 0 mV Logarithmic 1 000e 5 1 000e 3 1 000e 1 1 000e 6 1 000e 4 1 000e 2 1 00
483. source Value and Units This text field allows manual entry of the clock rate value Click inside this text field and use the numerical keypad or front panel knob to enter a clock rate value Complete the entry with a frequency unit by pressing the G n k m M p or X1 front panel key Preset List The Preset list contains clock rate values that are commonly used by the pattern generator Click a preset to select it for the pattern generator clock rate User presets can be modified To do so double click the preset System presets can not be modified Add Preset Button Click this button to open the Enter Bit Rate dialog box and add a new clock rate to the preset list In this dialog box you can enter a value unit and text description of the new clock rate Delete Preset Button Select a preset clock rate in the Preset list and click this button to remove it from the list You can only delete presets that are user defined Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 Clock Rate Indicators The PG Pattern Generator and ED Error Detector bit rate indicators shown in the lower pane display their current bit rate Pattern Generator m Id dm 300000Gb s 2 231PRBS Loss ISU SSC ONO Error Detector be ash 3 00000 Gb s 2 23 1 PRBS Eror pre aa Tides The ED bit rate is measured from the incoming clock signal or derived from the data signal Trigger Ref Clock Output
484. splay for SSC SJ shows the usage of the modulator in percent Jitter Setup Area The Jitter Setup area shows the present jitter composition 2 0 2 33 0kHz l Aou 500 Hz l 2 10 0000 MHz p mima hoc me E AZo mom Mz To become active jitter sources must be switched on by enabling the corresponding button and must at least have an Amplitude greater than zero Agilent J BERT N4903B High Performance Serial BERT 415 8 Jitter Tolerance Tests 416 The Jitter Setup area allows you to enable a jitter source to change its most commonly used jitter parameters directly or to go into advanced edit mode by pressing the corresponding Edit button Spread Spectrum Clock The Spread Spectrum Clocking setting controls the pattern generator s spread spectrum SSC clocking feature When the SSC is enabled it impacts the Data Out Clock Out Aux Data Out and Trigger Ref Clock Out ports A ien tore The spread spectrum clock is characterized by Deviation Frequency Type Profile 0 1225 96 0 245 33 kHz 998 955 Mb s 996 511 Mb s If the deviation type centerspread is selected the deviation can be changed in two different ways 0 1225 96 0 245 33 kHz Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Change the centerspread value the bitrate remains unchanged while the upper and lower frequency changes according to the selected deviation The deviati
485. ss 3 0 and is referred at several places in this document PCle and PCI Express are the registered trademarks of the PCI SIG Understanding the Error Free Receiving in USB 3 1 128B 132B Encoded Data Comparison When RX detects incoming data correctly the pattern looped back has the same content except the length of the Skip Ordered Set SKPOS primitives The change in the SKPOS length by DUT is to compensate for the speed differences of the clock domains DUT Slave receives SKP Ordered Set of 16 symbol and if clock compensation is required DUT must add or remove multiple of two skip symbols per Skip Ordered Set Hence received Skip Ordered Set can be 4 6 8 40 Symbols The SKP END Symbol indicates the last four Symbols of SKP Ordered Set The three Symbols received after loopback following the SKP END Symbol can be different Although bit wise different in content and number of bits in order to report this different looped back bit stream as error free the ED has to align with EIEOS Electrical Idle Exit Ordered Set and recognize the different length SKPOS and ignore these symbols For recognition of these special symbols one bit error correction is performed on header SKP and SKP END symbol each The Bit Comparison without USB3 1 SKPOS option is only applicable for USB 3 1 testing Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 NOTE The Bit Comparison without USB3 1 SKPOS error rat
486. st BER threshold available Running the auto search functions and viewing eye results at different BER thresholds will allow you to see the eye margin at specific BERs You can also construct basic eye contour information Avg 0 1 Threshold Activate this checkbox to make the error detector continuously set the 0 1 threshold level to the average DC voltage level of the incoming data signal NOTE Manual 0 1 threshold changes are impossible while this function is selected Agilent J BERT N4903B High Performance Serial BERT 201 5 Setting up the Error Detector 202 This function is useful for providing a starting point for sampling somewhere within the data eye on the vertical voltage axis It is recommended to use this prior to clock data alignments When used with the following patterns this function may provide a 0 1 threshold that requires no further adjustment e 2 n 1 PRBS patterns Patterns with 50 mark densities large amplitudes and no excessive noise Bit Recovery Mode Click this button to put the error detector in bit recovery mode BRM For details on BRM see Understanding Bit Recovery Mode on page 195 Then click Auto Align to set the sampling point Sampling Point Display This area of the Sampling Point Setup window graphically illustrates the parameters of the current sampling point and data eye Click anywhere in the sampling point display to set the sampling point to that location suggested ey
487. st Length Begin Margin End Margin Gate Active Begin Margin Bit Count Time Optimizing the Timing There are three things to watch when optimizing the timing for burst sync mode BER The bit error rate increases when the gate closes too late The duration of the signal should typically be reduced Burst sync ratio The burst sync ratio is an indication of how much of the burst signal is in synchronization and can therefore be used for counting bit errors The higher the value the better Bad burst count The bad burst count counts the number of bursts that are invalid for example because synchronization failed or the BER of a synchronized burst exeeds the sync threshold The following describes how changing the Gate In start and end points can affect the evaluation of bursts Gate In Start Starting too soon If the Gate In signal is applied too soon the error detector will either not be able to recover the clock or synchronize the pattern and will thus mark the burst as a bad burst thus increasing the bad burst count You can typically reduce the bad burst count by starting the gating period later CDR Settling Time Synchronization Time Start too soon Data Input Gate Input Settling Synchronization fail here Attempted Settling Time ends here Agilent J BERT N4903B High Performance Serial BERT 215 5 Setting up the Error Detector Starting too late If the Gate In signal is applied too lat
488. st is enabled when center is set to middle of eye and persistence is set to infinite in the Parameter Tab Automated Eye Parameter Measurement Automated Eye Parameter Measurement characterizes an Eye Diagram by measuring the rise time the fall time the eye amplitude and so forth These measurements are called Automated Eye Parameter Measurement they display the Eye behavior and they depend on the parameters set in the View Tab These results are displayed in a tabular form as shown below Sample Count 1 Level Level RiseTime Fall Time Eye Amplitude Eye Height Eye Width Jitter Peak Peak 1 304922E408 300 00m 300 00 m 0 03Ul 0 04UI 500 00 m 516 11 mV 0 9901 0 0101 Ki INNEN Jitter Peak Peak Jitter RMS Cross voltage Signalto Noise Ratio Duty Cycle Distortion Extinction Ratio 0 0101 0 0101 49 67 252 17 0 0101 lt No Data 4 Optical Modulation Amplitude Average Power lt No Data lt No Data gt In case you want to see a specific set of Eye Measurements you can select it on the Eye Measurements tab Agilent J BERT N4903B High Performance Serial BERT 319 6 Advanced Analysis 320 NOTE The scroll bar at the bottom of the table shows the entire table and its contents For more details refer to Eye Diagram Reference on page 332 Eye Diagram Procedures This section lists out some of the generic processes to set up measurements to run them and to create Eye Diagrams In t
489. st to assure that the device under test Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 operates with a BER better than let us say 10 12 whether the true BER is 1 1 x 10 or 2 7 x 1075 is irrelevant To abort the measurement for a single point and proceed to the next we need two limits that tell us whether the BER is above or below the given threshold These limits have been calculated from the error probability density functions applicable to BER measurements The equations were solved for a level of confidence of 95 The calculations have led to the following table Table 25 95 confidence level lower limits BER 95 confidence level upper limits BER gt 10 12 lt 10 12 Min number of Max number of Max number of Min number of errors compared bits x errors compared bits x 1012 1012 1 0 05129 0 2 996 2 0 3554 1 4 144 3 0 8117 2 6 296 4 1 366 3 7 154 5 1 970 4 9 154 6 2 613 5 10 51 7 3 285 6 11 84 The BER threshold of 101 and hence the setting of the multiplication factor to 10 2 is just an example by changing the exponent the table applies analog to other thresholds The following figure shows a plot of this table Agilent J BERT N4903B High Performance Serial BERT 247 6 Advanced Analysis 248 The Bracketing Approach N BER 1e 12 10 Number of transmitted bits 10 Number of errors Note that there is a gap where the BER is so close to 10 1
490. step size Downwards linear starts from the maximum which depends on the jitter frequency A step size of 100 mUI for example may result in a sequence of 1000 UI 999 9 UI 999 8 UI and so on The test for one frequency stops when the BER limit is met or zero amplitude is reached Upwards linear starts from the minimum A step size of 100 mUI for example will result in a sequence of 0 1 UI 0 2 UI 0 3 UI and so on The test for one frequency stops when the BER limit is exceeded or the maximum amplitude is reached Logarithmic search For logarithmic search you can set the minimum amplitude and the ratio The ratio is the factor by which the present amplitude value is multiplied to calculate the next amplitude Downwards logarithmic starts from the maximum which depends on the jitter frequency A ratio of 0 666 for example may result in a sequence of 1000 UI 666 UI 444 UI and so on The test for one frequency stops when the BER limit is met or the specified minimum amplitude is reached Upwards logarithmic starts from the specified minimum amplitude A ratio of 1 5 for example may result in a sequence of 0 1 UI 0 15 UI 0 225 UI 0 3375 UI and so on The test for one frequency stops when the BER limit is crossed or the maximum amplitude which depends on the jitter frequency is reached Binary search For binary search you can set the accuracy Agilent J BERT N4903B High Performance Serial BERT 445 8 Jitter
491. switches to manual sync mode and then restores it back when the measurement is finished aborted Understanding the Error Free Receiving in 8B 10B Symbol Comparison When RX detects incoming data correctly the pattern looped back has the same content and number of symbols except for filler primitives These can be removed or inserted by DUT to compensate for the speed differences of the clock domains Furthermore 8B 10B coding defines two different bitwise complementary 10B symbols for the majority of 8bit words based on a running disparity It may therefore happen that the data looped back by the DUT is transmitted with a different than the incoming disparity depending on the DUT s starting disparity Both effects represent desired operation of the DUT and the pattern looped back although bit wise different in content and number of bits is an indicator for error free data detection of the RX under test so that consequently in order to report this different bit stream as error free the ED has to recognize the 10B Data and filler symbols of both disparities and to ignore the insertion or deletion of filler symbols The following figure shows the functional block diagram of Option A02 Functional Block Diagram of Option A02 Input seria bit stream Clock input voltage threshold Input voltage threshold No clock 10B domain Comma setting Filler 1 setting 1 2 or4 symbols Filler2 setting 1 2 or4 symbols Fil
492. t to be run again See Parameters Tab on page 252 for details Pass Fail tab These settings determine whether the calculated results are recognized as passed or failed However a new test run is not required when doing changes here See Pass Fail Tab on page 254 for details View tab Graph tab and Color tab All settings on these tabs only affect the way the data is displayed You do not need to run the measurement again See View Tab on page 256 Graph Tab on page 258 and How to Change the Colors of the Graph on page 232 for details 3 Press OK when you have made all required changes to close the Properties dialog box DUT Output Timing Jitter Reference TIP NOTE The DUT Output Timing Jitter measurement returns the results in a graphical and in a numerical form The following sections describe the results They provide explanations of the measured parameters and the display options that are specific to this measurement Additionally some information is provided to explain the theoretical background behind the jitter calculations You have access to the user documentation including User s Guide over the instrument s Start menu Start gt Programs gt N4900 Series Documents Properties that can be specified on the various tabs of the Properties dialog box Parameters Tab If you modify the parameters on this page you have to rerun the measurement to update the results Set the criteria fo
493. t Accum button on the front panel This causes the following events If Prompt for File Name has been selected during setup you will be asked to enter a name for the log file before accumulation begins The data display in the Accumulated Results window is reset The recording of data to the measurement log file is reset How to Stop Accumulated Measurements If you specified a single or repeattest the test runs until the specified end condition is reached If you selected a manual test it will run until you stop it manually Press the Stop Accum button on the front panel to stop the test immediately NOTE Errors are only counted when the signal at the Gate In port is low Agilent J BERT N4903B High Performance Serial BERT 377 7 Evaluating Results 378 How to View the Results You can view data from accumulated measurements in several ways 1 Click the Accumulated Results menu item from the Results submenu to view the test results For easy comparison the results of the current and previous accumulation are listed here See Accumulated Measurements Reference on page 378 for details on the reported values 2 Examine the measurement log files on an external PC This allows to access all data for detailed processing and analysis on a PC Also you can track the status of all indicators such as DATA LOSS and CLK LOSS during the accumulation period If you want to view results from an accumulation period prior
494. t interval 6 minutes 488 Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 Notes on the measurement duration The example above was done for a target BER measurement of 10 and a confidence level of 95 As a rule of thumb this requires to measure approximately 3 times the number of bits given by the target BER The sampling frequency of the 86100A is specified with 40kHz which gives a sampling period of t 25us The timing interval resolution is given as screen diameter At lt max 62 5 fs record length The time base needs to be set to amp bits time base 10 Data Rate The number of waveforms that are required to be measured for achieving the confidence level of 95 is 3 Waveforms Target BER Each measured waveform contains the following number of samples time base bits Samplesy 10 divs a pesswaveform At At Data Rate The overall required measurement time can be estimated as Tmeas Waveforms Sampleswaverorm ts 3 bits t Target BER At Data Rate T meas The above formula shows that choosing At as 62 5fs leads do longer measurement times than required So it is safe to assume the highest possible resolution for this estimation In the example above the following values have been used bits 4 Target BER 10 Data Rate 14 2 Gb s This gives a time base s
495. t status becomes active NOTE The Jitter On Off button has no impact on SSC If spread spectrum clocking is enabled it continues even if the Jitter On Off button is off Jitter Configuration Use these radio buttons to select SSC or rSSC RJ and BUJ or sRJ Configuration os Jitter Distribution Use these switches to define whether the jitter shall be applied to clock and or data Agilent J BERT N4903B High Performance Serial BERT 413 8 Jitter Tolerance Tests Defining Jitter Delay on Clock and Data Use this entry to define the delay between 220ps jitter on clock and data Delay Line Switch This switch can be pressed to switch the source of periodic jitter 1 and 2 and BUJ to either the 610ps or the 220ps delay line NOTE The 610ps delay line cannot be used for data rates above 3 37 GHz 414 Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 Delay Line Display The delay line display shows red markers that indicate the consumption of delay line capacity by the present jitter setup Colored bars indicate the consumption by the various jitter components The consumption is shown in UI Unit Intervals and is calculated from the present data rate Because the delay lines have a fixed time capacity the scales change when the bit rate is changed NOTE The total consumption of the delay line in Ul is indicated on top of the delay line display The delay line di
496. t the Scope to Data Out of N4876A Connect the Trigger In of Scope to Trigger Out of PG Terminate N4876A Data Out with 50 Ohm Terminate PG Trigger Out with 50 Ohm Terminate PG Data Out and Aux Data Out with 50 Ohm Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 Then the oscilloscope should show an output eye like the following If you see a display like above everything is OK If you cannot see a clear eye diagram then check if the N4876A is connected properly to the Serial BERT The N4876A must be connected to Data Out Aux Data Out and Aux Clk Out to work properly If the connection to the Serial BERT is setup correctly then check the output signal of the Serial BERT at the Data Out Aux Data Out both should show a eye diagram with an offset of 400mV and a amplitude of 800mV Check the output signal of Aux CIk Out it should give a clock signal of set set data rate with an offset of 0V and an amplitude of 400mV if one of the used Serial BERT outputs is not providing the expected signal then the Serial BERT is defective If the Serial BERT is providing the correct signals then connect the N4876A again and run the adjustment procedure from the Config screen Check the output of the N4876A again after the adjustment has been done If the adjustement does not fix the problems then the N4876A is defective Check the Output of the N4876A Multiplexer Using the Error Detector If you have
497. tart Click this button to start the sequence execution Sequencer Pattern Editor Concepts NOTE The sequencer pattern editor displays the pattern for Data Out and or Aux Data Out Channel and offers you to edit it There is no dedicated main menu entry to open sequencer pattern editor Access to this editor is only possible from within the sequence editor Sequencer Pattern Editor Procedure 108 You can do the following tasks in the sequencer pattern editor window Creating New Patterns in Sequencer Pattern Editor To create a new pattern standard or alternate 1 Click Sequence Editor menu item from the Pattern submenu EX Sequence E ditor Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 By default the Sequence Editor shows one block that is automatically repeated It has a length of 512 bits and generates Pause 0 data pure zeros 2 Click Edit to launch Set Data Block Parameter dialog for second channel 3 Click Create Pattern to launch Create New Pattern dialog Use this dialog to create new patterns See Create New Pattern on page 115 for details Opening Existing Patterns in Sequencer Pattern Editor To open an existing user pattern 1 Click Sequence Editor menu item from the Pattern submenu Sequence E ditor 2 Click Edit to launch Set Data Block Parameter dialog for second channel 3 Select Pattern from the provided options and then click Browse to launch Select Use
498. tector trigger capability is only available when the 128B 130B coding support is activated Only the trigger pulse timing is compatible with the pattern generator Aux input for sequence advance It takes 2048 bit time period to process the trigger and the signal is active only for half of this period Hence next trigger will be processed only after this period NOTE The SKPOS Detected option will be available only if you have selected the error ratio based on Bit Comparison without PCI3 SKPOS Note that the length of the trigger signal is at least 40 bits and divisible by 4 Aux Out You have two options to set up the signal at the Aux Out port e Clock This option supplies the clock signal to the Aux Out port Data This option supplies the data signal via a comparator to the Aux Out port The comparator is controlled by the 0 1 threshold This lets you use an oscilloscope to determine if the 0 1 threshold is correctly set If the 0 1 threshold is set below or above the data eye the output at Aux Out will be constant high or low respectively Pattern Synchronization Pattern Synchronization Concepts The Serial BERT calculates bit error rates by comparing the received data with the expected data patterns To do this it needs to know where the start of the pattern is located in the data stream Agilent J BERT N4903B High Performance Serial BERT 205 5 Setting up the Error Detector 206 Hardware Generated Patterns Memory
499. ted with commas 4 Once the file has been imported you may need to resize columns The imported log file should appear similar to the following example During measurement logging the Serial BERT logs data in ten second intervals Your log file may be missing up to the last ten seconds of data To avoid this condition accumulate for 10 seconds longer than desired A BER c D E 1 Description test 2 Log Time 05 29 2000 12 14 58 334 3 Activation Mode Single 4 Accumulation Measurement ELAPSED SECONDS S Accumulation Limi 60 000000 6 7 Elapsed Time BIT COUNT BIT FREQUENCY ED FREQUENC 8 0 1 2 49E 08 2 49E 09 249E BI 0 2 2 49E 08 2 49E 09 2 49E 10 03 2 49E 08 2 49E 03 249E 11 0 4 249E408 2 49E 03 2 49E 12 0 5 2 49E 08 2 49E409 2 49E You can view and analyze the following information in the measurement log file Table 38 Instantaneous Cumulative BIT COUNT BIT COUNT Cumulative ERROR COUNT ERROR COUNT Cumulative ERROR RATIO ERROR RATIO Cumulative ERRORED 0 COUNT ERRORED 0 COUNT Cumulative ERRORED 0 RATIO ERRORED 0 RATIO Cumulative ERRORED 1 COUNT ERRORED 1 COUNT Cumulative ERRORED 1 RATIO ERRORED 1 RATIO Cumulative SYNC LOSS ERROR SECONDS Cumulative PG CLOCK LOSS ERROR FREE SECONDS Cumulative Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 Table 38 Instantaneous Cumulative ED CLOCK LOSS ERROR DECISECONDS Cumulative DATA LOSS ERROR FREE DECISECONDS Cumula
500. ter PJ 398 PGCLKLOSS 470 Phase Margin 260 262 PJ 398 Planning the Test 11 Post cursor de emphasis 39 Power scale 357 PRBN 63 PRBS 62 PRBS Patterns 61 Pre cursor de emphasis 39 Printing 555 Problems 469 Constant Errors More Errored 0 s than 1 s 473 Random Errors 474 Sync Loss Seconds 475 Properties dialog box DUT Output Timing Jitter 252 Eye Opening 294 Fast Eye Mask 311 Output Levels 270 Spectral Jitter 353 Pseudo Color Plot 288 Q Q Factor 280 Q High Level 280 Q High Level Nr Points 281 Q High Level R 2 281 Agilent J BERT N4903B High Performance Serial BERT Q High Level Std Dev 281 Q Low Level 281 Q Low Level Nr Points 282 Q Low Level R 2 282 Q Low Level Std Dev 282 Q Optimum Threshold 280 Q Residual BER 280 Q Factor Calculations mathematics 283 Notes 285 QBER vs Threshold Graph 275 R 2values 261 Random Errors 474 Random jitter RJ 397 Random Jitter RMS 261 Ratios Graph 379 Relative Time 313 Requirements 26 Resolution Eye Opening 295 Fast Eye Mask 295 Output Levels 271 Output Timing 253 Results see Numerical Results 259 Return to Results button 202 RJ 397 Running Measurements 361 S Sampling Point Setup 193 Saving Patterns 72 Self Test 563 Sequence creating 99 definition 96 loops 103 properties 107 recalling 100 saving 100 Sequence Editor 101 Toolbox 104 Sequence Mode 98 SequenceExpression 96 Sequences
501. terface Error Detector 3 00000 Gb s 2 23 1 PRBS The bit rate of the error detector is set by the clock signal received by the error detector If the output bit rate of your device is not equal to its input bit rate the pattern generator and error detector will be operating at different frequencies Clock Mode The error detector supports two different clock modes e External Clock Source If this mode is selected the Serial BERT uses the clock signal that is connected to the Clock In port e Clock Data Recovery If this mode is selected the Serial BERT derives the exact bit rate from the incoming data stream To make this function work the approximate clock rate must be specified in the Clock Rate field with an accuracy of 0 0196 Clock Rate This field is used to specify the approximate clock rate for CDR mode Serial BERT needs this value to be about 0 0196 accurate to find the actual bit rate in the data stream Use the numeric keypad to enter the value followed by the unit that you can select with the G n M y k m and X1 keys on the front panel Alternatively you can select a value from the Preset list Agilent J BERT N4903B High Performance Serial BERT 179 5 Setting up the Error Detector 180 NOTE Preset List The Preset list contains clock rate values that are commonly used by the error detector These presets also modify the loop parameters of the tunable CDR according to the selected standard Th
502. tern editing from sequencer Export either Data Out or Aux Data Out as Standard pattern or both as an Alternate pattern Description Add a description for the captured Pattern File name Provide a name to the file Export Click this button to export the file at the given location Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 Pattern Properties The Pattern Properties dialog is used to view change the properties of the current pattern s Pattern Properties Data Channel Description Pattern A Pattern Type Alternate Pattern A Aux Data Channel Description Pattern B Pattern Type Alternate Pattern B Length in Bits 2560 Cancel Help This dialog provides the following options Description Description of the pattern This text field allows you to enter a description of the pattern s characteristics or purpose It can be up to 256 characters long Pattern Type Pattern type of pattern This field is non editable Length in Bits Pattern length in bits This field is editable Agilent J BERT N4903B High Performance Serial BERT 119 J Setting up Patterns 120 Agilent J BERT N4903B High Performance Serial BERT e 4 e e Setting up the Pattern Generator Setting up the Pattern Generator Concepts The Serial BERT s pattern generator generates an output signal based on a data pattern It has the following possibilities for generati
503. ternal instrument s NOTE This button will be either disabled or not available in case of M8061A if the software updates are not required Visa Resouce Name The Visa Resouce Name displays the unique symbolic name of external instrument s System Setup Area The System Setup area allows you to perform the following settings Load Calibration Settings when the Serial BERT is connected to N4916A B Input Timing Adjustment Adjustment when the Serial BERT is connected to N4876A Load Calibration Settings The Load Calibration Settings button allows you to provide input timing setup and load calibration settings Input Timing Setup 2 Load Calibration Setting NOTE The Load Calibration Settings is applicable for N4916B For more information refer to Loading Calibration Settings on page 480 Input Timing Adjustment The Adjustment has to be done once for a dedicated setup of J BERT N4903B M8061A N4876A and the cable kit being used whenever one of these components in the setup is changed or if the operating temperature changes by more than 5 Celsius compared to the previous adjustment Click Adjustment to start adjustment if there is any change in the connection NOTE The Input Timing Adjustment is applicable for M8061A N4876A For more information on input timing adjustment of M8061A refer to the section M8061A System Setup Adjustment on page 492 Agilent J BERT N4903B High Performance Seri
504. terns Sequences User patterns are file based editable patterns These may be patterns that you set up using the Serial BERT s Pattern Editor or any of the patterns delivered with the Serial BERT You can easily define your own patterns for any special requirements The Serial BERT can also generate various PRBS 2 n patterns with a range of polynomes from 7 to 23 These patterns can be further modified by setting special mark densities ratio of 1 s to total bits or by defining the zero substitution It is possible to combine several patterns in a sequence A sequence consists of up to 120 blocks that can be looped The sum of blocks and counted loops must not exceed 120 For example if 100 blocks are used a maximum number of 20 counted loops can be defined Each block can generate a pause signal constant 0 or 1 a divided clock a 2 n 1 PRBS or a user pattern What Patterns are Available The Serial BERT offers various industry standard and customized patterns for many BER testing scenarios The following graphic illustrates what patterns are available on the Serial BERT Note that there are two main branches corresponding to the hardware generated and memory based patterns and the memory based patterns are further broken down into software generated patterns and user patterns Pattern Family Tree Hardware Generated Memory Based Patterns Patterns Software Generated Patterns User Pat
505. terns 2 n 1 PRBS Custom User Example Patterns Patterns in C N490X Pattern Demo PRBS Zero Substitution Mark Density Fiber Channel Gigabit Ethernet 2 n PRBS Mark Density FDDI 2 n PRBS SDH SONET Industry Standard Pattern or modified Industry Standard Pattern See Example Patterns on page 70 for a complete listing of the example patterns What is PRBS The most common type of industry standard pattern is the Pseudo Random Binary Sequence PRBS PRBSs have a number of desirable properties including the ability to simulate random data with a balanced number of 1 s and 0 s Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 They are referred to by their pattern lengths which are powers of 2 Longer pattern lengths give a better approximation of random data and provide a more rigorous test for the device However the pattern repetition time is significantly longer PRBS patterns provide a means to simulate the type of traffic that a system is likely to see random traffic They also enable you to stress your system to its limits because large patterns that include long strings of 1 s or 0 s are more stressful The patterns are easy to generate and can measure at high speeds because they are sent repeatedly and can be predicted Pattern synchro
506. testing applications The Serial BERT provides capabilities that help you set up and run burst tests See Introduction to Burst Sync Mode on page 211 Are you aware of how the sync mode setting can affect the way errors are measured and displayed See What Type of Synchronization Should You Use on page 208 Test Times and Confidence Levels A true BER measurement must be statistically valid Because it is not possible to predict with certainty when errors will occur your device must be tested long enough to have confidence in its BER performance The table below lists the test times required for different BERs at specific bit rates Thetesttimes in this table are valid for continuous measurements If measurements are made in bursts the test times will be longer continuous test time divided by the burst duty cycle times the burst sync ratio See Introduction to Burst Sync Mode on page 211 for background information The test times for 9596 confidence level are Table 3 BER STM 64 STM 16c STM 4c STM 1 0C 3 0C 192 0C 48c 0C 12c 155 52 Mb s 9 95328 Gb 2 48832 Gb 622 08 Mb s s s 1E 14 8 4 hours 1 4 days 5 6 days 22 4 days 1E 13 50 minutes 3 3 hours 13 hours 2 2 days Agilent J BERT N4903B High Performance Serial BERT 27 1 Planning the Test 28 Table 3 BER STM 64 STM 16c STM 4c STM 1 0C 3 0C 192 0C 48c 0C 12c 155 52 Mb s 9 95328 Gb 2 48832 Gb 622 08 Mb s s s 1E
507. the second channel Aux Out with a selected PRBS automatically Note This icon will only appear in the second channel mode Setting Sequence Properties For a sequence you can specify the following Sequence Start Condition Choices are Immediate Sequence starts immediately after downloading it to the pattern generator Aux In High Aux In Low Sequence starts when Auxiliary Input is high or low Aux In Rising Edge Aux In Falling Edge Sequence starts with a rising or falling edge at the Auxiliary Input Command The Sequence Editor shows a Start button Sequence execution starts when this button is clicked Number of Blocks A sequence consists of up to 120 blocks than can be looped Description A field for entering and editing descriptive text Sequence Expression A window that shows the present SequenceExpression for details see How a Sequence is Defined on page 96 You can edit the SequenceExpression in this field Any changes override the previous settings Agilent J BERT N4903B High Performance Serial BERT 107 3 Setting up Patterns NOTE Sequencer Pattern Editor If you edit the SequenceExpression manually take care to adhere to the syntax and order of the keywords Remember that you must download the updated sequence to the pattern generator if your changes shall take effect Start Button The Start button of the Sequence Editor indicates that the sequence has been set up for manual s
508. the 0V Disable button in the PG Setup gt Data Output screen Use a shielded cable to connect the pattern generator s Data Out port and the error detector s Data In port Terminate all non connected pattern generator output ports with 50 Q Use a shielded cable to connect the pattern generator s Data Out port and the error detector s Data In port Switch to the Pattern panel and press Pattern Select Select an appropriate pattern for example a pure 2 15 1 PRBS for both the pattern generator and the error detector For the pattern generator setup you need to specify the logic levels and the bit rate Select ECL levels and a clock speed of 1250MHz in this example This corresponds to a clock period of 800ps See Setting up the Pattern Generator Concepts on page 121 for more information Set up the error detector so that the input range and the termination matches the pattern generator s levels Select an Input Range from 2V to 0V Setthe Data Termination to 2V Setthe Alignment BER Threshold to 1E 6 Setthe Clock Setup to Clock Data Recovery to get the error detector s clock from the incoming data stream Enable the pattern generator outputs by pressing the OV Disable button Press Sync Now and then Auto Align to find the optimum sampling point Check that the synchronization and the alignment were successful None of the error indicators should show red The resulting bit error rate should be zero
509. the actual error count or cumulative error count The BER bar below the main menu bar also displays the actual BER You can drag the small yellow alarm threshold mark to change the BER Alarm Threshold Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 8B 10B Comparison Results Window This window displays the actual or cumulative results of the following measurements Symbol Error Ratio SER See SER Results Window on page 367 Frame Error Ratio FER See FER Results Window on page 368 Calculated Bit Error Ratio CBER See cBER Results Window on page 368 Filler Symbol Ratio FSR See FSR Results Window on page 369 Disparity Error Ratio DER See DER Results Window on page 370 Illegal Symbol Ratio ISR See ISR Results Window on page 371 SER Results Window This window displays the actual SER or cumulative AccumSER SER results om SER 0 000 Remote Symbol Error Counts The SER is the current SER calculated upon a period of 200ms The AccumSER shows either the accumulated SER of the current accumulation or if no accumulation is running the results of the most recent accumulation This enables you to monitor real time SER behavior as you do things such as to manually adjust the sampling point to add errors or to make adjustments to your device Agilent J BERT N4903B High Performance Serial BERT 367 7 Evaluating Results 368 Accumulated A
510. the filler primitives contained in the original test pattern this indicates that the DUT clock is faster than the PG clock 3 If the number of filler primitives is lower than the filler primitives contained in the original test pattern this indicates that the DUT clock is slower than the PG clock 4 Ifthe number of filler primitives matches the number of filler primitives contained in the original test pattern the DUT is likely a tracking device that doesn t do any re timing It also likely passes all jitter that is imposed to its receiver to its transmitter and the test may be impacted by the Error Detector s Jitter Tolerance The following is the expression used to calculate the Data Rate Ratio Received Symbol Count Data Rate Ratio Expected Symbol Count Where Expected Symbol Count Pattern Count X Block Length with Filler Symbols Received Symbol Count The total number of Received Symbol Count all incoming symbols including incoming filler symbols received in a time interval is displayed here The following is the expression used to calculate the Received Symbol Count Received Symbol Count Pattern Count x Expected Pattern Length x Filler Symbol Counter Where Expected Pattern Length is without Filler Symbols Pattern Count Symbol Counter Block Length without Filler Symbols Auto Resync Counter The total number of Auto Re Sync Count received in a time interval is displayed h
511. the ratio of the number of errored bits to the total number of bits received For example 1 error in 1000 bits corresponds to a BER of 1 1000 or 1x102 This would be displayed on the instrument as 1E 3 Refer to the following table to familiarize yourself with how BERs are displayed Notice that smaller exponents correspond to higher BERs and larger exponents correspond to lower BERs Table 1 Fraction Exponent Instrument Display 1 1 1x 10 1E 0 1 10 1x 107 1E 1 1 100 1x10 1E 2 1 1000 1x10 1E 3 Agilent J BERT N4903B High Performance Serial BERT Planning the Test 1 Table 1 Fraction Exponent Instrument Display 1 10 000 1x 10 1E 4 1 100 000 1x10 1E 5 1 1 000 000 1x108 1E 6 1 10 000 000 1x107 1E 7 Understanding SER Symbol error ratio SER is the symbol ratio of the number of errored symbols to the total number of symbols received For example a single error will result in a one symbol error in 1000 symbols where each symbol consists of 10 bits that corresponds to a SER of 1 1000 or 1x10 This would be displayed on the instrument as 1E 3 Refer to the following table to familiarize yourself with how SERs are displayed Notice that smaller exponents correspond to higher SERs and larger exponents correspond to lower SERs Table 2 Fraction Exponent Instrument Display 1 1 1x 10 1E 0 1 10 1x101 1E 1 1 100 1x10 1E 2 1 1000 1x10 1E 3 1 10 000 1x10 1E 4 1 100 000 1x10 1E 5 1 1 000 000 1x108 1E 6 1 10 000 000
512. the section How to Enable Disable N4916B Clock Multiplier Function on page 47 To enable N4876A Multiplexer function refer to the section How to Enable Disable N4876A Multiplexer Function on page 47 To enable M8061A Multiplexer with De emphasis function refer to the section How to Enable Disable M8061A Multiplexer with De emphasis Function on page 49 6 Setthe parameters for De emphasis Clock Multiplier Multiplexer and Multiplexer with De emphasis functions To set the parameters for De emphasis function refer to the section Controlling the De Emphasis Signal Converter on page 132 To set the parameters for Clock Multiplier function refer to the section Controlling N4916B Clock Multiplier on page 50 To set the parameter for Multiplexer function refer to the section Controlling N4876A Multiplexer on page 51 To set the parameter for M8061A Multiplexer with De emphasis function refer to the section Controlling M8061A Multiplexer with De emphasis on page 51 Agilent J BERT N4903B High Performance Serial BERT 45 2 Setting up External Instrument s How to Enable Disable N4916B De Emphasis Function To enable disable the de emphasis function 1 From the Config window select the De Emphasis function check box It opens the Deemphasis Signal Converter Connection dialog The following image shows the an example of N4916B Deemphasis Signal Converter Connection Dialog Connect the dat
513. this icon if you want to jump to an arbitrary bit position See GoTo Bit Dialog Box on page 82 for details Clickthis icon to open Bin Hex Symbol Configuration dialog See Bin Hex Symbol Configuration Dialog Box on page 80 for details 113 J Setting up Patterns Table 15 Icon Name Description 021 Symbol Codings Click this icon to open az Symbol Settings dialog See Symbol Settings on page 116 for details end Pattern View Click this icon repeatedly to view patterns in the following ways Normal Mode Data Out or Aux Data Out or Both when symbol mode is enabled pattern only Interleaved Mode Data Out and Aux Data Out patterns in interleaved rows Split Mode Data Out and Aux Data Out patterns in horizontally split windows This icon will only work when both the channels are using pattern files Pattern Editor Canvas and Status Bar for Sequencer Pattern Editor The pattern editor canvas for sequencer pattern editor displays the pattern for Data Out and or Aux Data Out Channel and offers you to edit it Access to this editor is only possible from within the sequence editor The pattern editor canvas for sequencer pattern editor is shown in the figure below 114 Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 It uses a different set of background color to distinguish between the two channels The RED background denotes the pattern for Data Out chanel wh
514. tion to set the trigger level high whenever alternate pattern B is sent The trigger output for 1024 bit patterns looks as follows Data Out A A B A B B A Trigger Out Jj e EE TS PL NOTE This pattern must be at least 1024 bits long Alternate Pattern Trigger Pulse Select this option to send a trigger pulse whenever the pattern being sent changes A to B or vice versa For patterns with a length divisible by 512 there is one trigger pulse sent for every pattern repetition Data Out A B A B B Trigger Out If the length of the pattern is not divisible by 512 a trigger is sent whenever the pattern reaches a 512 bit RAM boundary Sequence Trigger This checkbox becomes available after a user defined sequence has been downloaded to the pattern generator Click this checkbox to switch the Trigger Ref Clock Out port to Sequence mode In Sequence mode the Trigger Ref Clock Out can generate a spike whenever the execution of a block starts or restarts Whether that happens for a particular block or not is defined for each block individually in the SequenceExpression See also Sequence Block Parameters on page 102 Agilent J BERT N4903B High Performance Serial BERT 147 4 Setting up the Pattern Generator 148 NOTE Pattern Trigger Position This option is not available for alternate patterns Select this option to send a trigger signal that is synchronized to a certain
515. tional de emphasis to extend the rate of J BERT N4903B pattern generator For the most accurate receiver characterization results the M8061A provides four calibrated de emphasis taps which can be extended to eight taps built in superposition of level interference and Clock 2 jitter injection Agilent J BERT N4903B High Performance Serial BERT 41 2 Setting up External Instrument s De emphasis on pre cursor and 2 post cursors with M8061A and J BERT 28 4 Gb s with M8061A and J BERT N4903B 42 N4903B The Data In 1 Data In 2 and Aux Clk In of M8061A have to be connected to Aux Data Out Data Out and Aux Clk Out ports of the J BERT N4903B respectively The connections should be made using the matched cable kit with the part number M8061 61601 The Data Output has to be connected to the device under test The following example illustrates how to emulate transmitter de emphasis up to Pre Posti Past Xd U TTTU00011 Setting up External Instrument s Procedure NOTE This section describes how to use the External Instrument s menu to enable external instruments connected to the J BERT N4903B Before continuing to next step make sure that the physical connection among the Serial BERT and the external instruments are properly done and then turn them on Also ensure that you have properly installed the external instruments For a complete list of external instruments and their installation refer to the section
516. tive SYNC LOSS SECONDS Cumulative BURST STATUS Cumulative BURST SYNC RATIO Cumulative TOTAL BURST COUNT Cumulative BAD BURST COUNT Cumulative Saving and Recalling Patterns You can save your pattern in either of the following ways lick the Save icon on the Pattern Editor toolbar If the pattern has not been saved earlier a file dialog box opens that allows you to navigate the file system and name your pattern Alternatively you can use the following procedure 1 From the File menu select Save 2 Then select Save Pattern in Editor to save the pattern with the current filename Or click Save Pattern in Editor as to open a file dialog box that allows you to save the pattern with a new name NOTE The default pathname for patterns is C lt instrument model gt Patterns Saving and Recalling Screen Images The instrument allows you to save an image of the current display screen You can save this image to the internal hard disk or a USB stick You can later print this file or import it to a word processing or graphics editing program Agilent J BERT N4903B High Performance Serial BERT 561 10 Customizing the Instrument Saving To save an image of the display screen do the following 1 In the File menu select Save 2 Select Save Screen Capture This opens a dialog box which allows you to save the image as a bmp 256 color bitmap file format file Recalling Follow the steps below to view screen images on a
517. to fulfill a complete sweep The start and stop values are included The valid range is between 2 to 100 You can use this drop down list to specify whether the frequency steps are log equidistant EQUidistant along the periodic jitter curve or a frequency step matches a corner frequency on the periodic jitter curve Bounded Uncorrelated Jitter Parameters Bounded uncorrelated jitter is characterized by Data rate Agilent J BERT N4903B High Performance Serial BERT 425 8 Jitter Tolerance Tests 426 Setting Data rate Amplitude PRBS Amplitude PRBS polynomial Low pass filter ma A graph indicates whether the chosen setting of frequency and amplitude is tolerable 2 000 Gb s 100 MHz Y You can specify your own setup or choose one of the predefined settings You can choose one of four predefined settings CUSTom The individual parameter values apply CEI6G Meant for CEI 6 Gbit s tests PRBS data rate is 1 1 Gbit s the PRBS polynomial is 29 1 the low pass filter is 100 MHz CEI11G Meant for CEI 11 Gbit s tests PRBS data rate is 2 Gbit s the PRBS polynomial is 27 1 the low pass filter is 200 MHz Gaussian Preset values for Gaussian distribution PRBS data rate is 2 Gbit s the PRBS polynomial is 23 1 the low pass filter is 100 MHz Enter an appropriate Data Rate Refer to the technical specifications The maximum peak to peak Amplitude is limited by the free capacity of the 20 p
518. to the current and previous test you can view the respective log file too How to Analyze the Results Accumulated measurements can help you determine the cause of bit errors Bit errors can be caused by a variety of problems In addition instrument settings can affect how errors are measured and displayed The sync mode setting has the greatest effect Analyzethe basic accumulated results and log files Possible observations that may occur are Errors became constant and remained constant for a period of time or number of errored 0 s was greater than errored 1 s see Constant Errors More Errored 0 s than 1 s on page 473 Errors were random see Random Errors on page 474 Sync loss seconds were measured see Sync Loss Seconds on page 475 Accumulated Measurements Reference NOTE The Accumulated Results window presents data from the current and previous accumulation period Data from the current and all previous accumulation periods is available in the measurement log files if correctly set up This window contains a graph and five tables Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 Ratios Graph This graph displays the delta errored 1 s ratio delta errored 0 s ratio and total delta error ratio at data points over the entire accumulation period The error ratios on the y axis are set to a range of 1E 0 100 errors to 1E 12 The accumulation period is on
519. tolerance measurement Running the Test Once you have specified the Standard Frequency and BER press the Start button to run the test Start The bar in the lower right hand corner of the screen shows the progress You can also abort the test at any time by pressing the Abort button When the measurement is either completed or aborted you can obtain a list of measured points by clicking the following tab which appears once the measurement is stopped Select Point Clicking on the above tab you get the list of measured points Agilent J BERT N4903B High Performance Serial BERT 459 8 Jitter Tolerance Tests Frequenc Amplitude Pass Fail 1 000 kHz 150 0 UI Passed 1 812 kHz 150 0 UI Passed 3 282 kHz 150 0 Ul Passed 5 945 kHz 150 0 UI Passed 10 770 kHz 139 3 Ul Passed 19 511 kHz 76 9UI Passed 35 346 kHz 8642 4 Ul Passed B4 033kHz 23 4Ul Passed 116 000kHz 129Ul Passed 210 144 kHz 7 1 Ul Passed 380 692kHz 3 9Ul Passed 689 654kHz 2 2UI Passed 1 243 MHz 1 2 Ul Passed 2 253 MHz 663 mUl Passed 4 100 MHz 660 mUl Passed 7 428 MHz 660 mUI Passed 13456 MHz 660 mUl Passed 24 377 MHz 660 mUl Passed 44 160 MHz 660 mUl Passed 80 000 MHz 660 mUl Passed Up Down Set Close Selecting a point and clicking on Set updates the Sinusoidal and Periodic jitter of that point on the Jitter Setup page Generate HTML Report The measurement results can be saved as HTML file To generate the HTML report 1 Click on the gt gt
520. tomatic re sync behavior and a manual Sync Now is required 6 Select the predefined standards listed in the Setup window In order to achieve Symbol Lock the pattern loaded in PG must have the same NOTE Comma Symbol as selected from the standard 7 Click Add Preset button to open 8B 10B Presets dialog It allows you to add a new preset to the Setup list You can then enter a short description values for filler symbols and align symbols for the new preset For more details see Error Ratio Reference on page 190 Add Preset NOTE Changes made in this dialog are effective immediately So if you press Cancel the settings are still valid 8 Select a preset in the Setup list and click Delete Preset button to remove it from the list You can only delete presets which are defined by the user Agilent J BERT N4903B High Performance Serial BERT 187 5 Setting up the Error Detector Delete Preset 9 You can view the details of the selected presets To do so either click View Edit Preset button or double click the preset Remember the details of the system presets will be in the disable mode User presets can be modified To do so either click View Edit Preset button or double click the preset System presets can not be modified View Edit Preset 10 f you choose error ratio based on 8B 10B symbol comparison then use the drop down list to specify whether you want to show the 8B 10B symbol comparison results as SER FE
521. tor turns blue and the mouse is catched within Agilent J BERT N4903B High Performance Serial BERT 19 1 20 Planning the Test Control s Title Displays the title of the control window Middle Pane The following picture illustrates the middle pane of the Serial BERT Lower Pane NOTE s P FZ en pz 10 0000 MHz The middle pane allows you to launch the different controls and dialogs of the Serial BERT These will be described in detail in the subsequent sections The following picture illustrates the lower pane of the Serial BERT The lower pane provides information about the Pattern Generator and the Error Detector this include Setup information like Bit Rate and Selected Pattern State information like Outputs ON or Jitter ISI SSC Indicated by green LED Error indicators like Error Sync Loss Data Loss Clock Loss Indicated by red LED Two additional error indicators Symb Lock and 8b10b Error will appear if you select the error ratio as 8B 10B Symbol Comparison from the Error Detector Error Ratio s window For more details see Symbol Lock Indicator on page 192 and 8b10b Error Indicator on page 192 Warning indicators like Output Protection Indicated by orange LED at the Outputs ON indicator Temperatur warnings Only visible if the temperature exceeds a certain threshold value For more information see Overheat Protection on page 478 Agilent J BERT N4903B High Perfo
522. tput is programmed to a fixed level The amplitude is set to 1 2 V the offset to 0 This yields a signal of 0 6 V Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 The oscilloscope should show the following picture DUPO etuer If you see only a straight line ensure that the pattern generator is not disabled and that it is set up to generate a pattern e g PRBS Settings for Clock Multiplier Test Following are the settings to test the output of the clock multiplier Preset Instrument State Set Data Rate to 3GB s Set Clock Offset to 0V Set Clock Out Amplitude to 400mV Set PG Trigger Out Offset to 0V Enable Clock Multiplier in the Config window Set the following for Clock Multiplier Input Frequency 3GHz Termination ON at 0V Check the Output of the N4916B Clock Multiplier Part Using a Scope Clock Multiplier 1 Following are the steps to test clock multiplier 1 using a scope Agilent J BERT N4903B High Performance Serial BERT 507 9 Solving Problems Connect PG Clock Out to Clock Multiplier Clock In Terminate PG Clock Out with 50 Ohm Termiate PG Trigger Out with 50 Ohm Terminate Clock Multiplier Clock In with 50 Ohm a A Ww N Set Clock Multipiler to 1 BER 0 000 508 Agilent J BERT N4903B High Performance Serial BERT Solving Problems 9 Check the Output of the N4916B Clock Multiplier Part Using a Scope Clock Multiplier 2 Following are the step
523. tput signal voltage precisely Otherwise you might damage your device The following example illustrates the output of the differential signal with variable de emphasis on pre cursor and 2 post cursors generated by De Emphasis Signal Converter N4916B Agilent J BERT N4903B High Performance Serial BERT Setting up External Instrument s 2 Understanding the N4916B Clock Multiplier The clock multiplier option enables error counting and error analysis of devices using half rate clocking The full rate clock is needed to use the error eye or jitter analysis capabilities of Serial BERT NOTE The clock multiplier functionality is only supported by N4916B Understanding the N4876A 28 Gb s Multiplexer 2 1 The Agilent Technologies N4876A is a 2 1 multiplexer with an output data rate of up to 28 4 Gb s The N4876A multiplexer doubles the pattern generator data rate by multiplexing two pattern generator channels It extends the variable generator data rate of J BERT N4903B up to 28 4 Gb s The Data In of N4876A has to be connected to Data Out of the PG Aux Data In to Aux Data Out of PG and Aux Clk In to Aux Clk Out of PG using the using the matched cable kit with the part number N4915A 011 Its Data Output has to be connected to the device under test Understanding the M8061A 28 Gb s Multiplexer with De emphasis The Agilent Technologies M8061A is a 2 1 multiplexer to characterize serial interfaces of up to 28 4 Gb s with op
524. trigger is based on the 8B 10B data coding The filler removal function is configured with up to four different detect words which can be 1 2 or 4 symbols and also allows for wildcards This flexibility is used to load detect words that is typically required for DUT handshakes The filler removal unit triggers upon a detected word which is used with the pattern generator AUX input to advance a pattern sequence and thus implements a simple handshake The enhanced error detector trigger capability is only available when the 8B 10B coding support is activated Only the trigger pulse timing is compatible with the pattern generator Aux input for sequence advance It takes 2048 bit time period to process the trigger and the signal is active only for half of this period Hence next trigger will be processed only after this period The Filler Detected option will be available only if you have selected the error ratio NOTE based on 8B 10B symbol comparison SKPOS Detected 204 Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 The usage of the enhanced error detector trigger is based on the PCle3 SKPOS Skip Ordered Set The 130 bit SKPOS is used as detect word and is not configurable The SKPOS removal unit triggers upon a detected word of 66 98 130 162 and 194 bit SKPOS which is used with the pattern generator AUX input to advance a pattern sequence and thus implements a simple handshake The enhanced error de
525. tton This opens the ED Input Setup dialog Differential Input Range 1 000 V 1 000 V Termination 0 000 V Edit 3 Inthe Termination field enter the termination voltage that is appropriate for the incoming data signal You can use the front panel knob or the numeric keypad to enter the value CAUTION Selecting the wrong termination may damage your device 4 Select the Data Inverted checkbox if your device inverts data You can now physically connect the DUT to the error detector Agilent J BERT N4903B High Performance Serial BERT 171 5 Setting up the Error Detector Data Input Setup Reference The ED Input Setup dialog is accessed from the Sampling Point Setup window It is used to set up the error detector s data input port Input The selection in this list defines how the signals arriving at the Data In and Data In connectors are interpreted The following options are available Differential If differential mode is selected both input ports need to receive a signal The actual data signal is measured as the voltage difference between the two incoming signals Normal In normal mode only the Data In port receives the data signal the Data In port is inactive Complement In complement mode only the Data In port receives the data signal the Data In port is inactive Data Inverted Activate this checkbox to invert the polarity of the error detector reference pattern This function is required i
526. tual sampling point which can be the optimum sampling point and a threshold voltage which is adaptive In practice six measurement points will often suffice to approximate the shape of the eye Six measurement points are preset by default The Fast Eye Mask can be run by using the appropriate SCPI commands making it possible to integrate the Serial BERT into a testing environment See the Programming Guide for details Example Results The Fast Eye Mask measurement measures the bit error rate at certain measurement points These points have to be positioned inside the expected eye opening Threshold voltage Measurement points ui Time Actual optimum sampling point The results of a Fast Eye Mask measurement are displayed in numerical form Terminal Copi 1 2 3 4 5 6 Relative Time 0 4 UI 0 4 Ul 0 16 Ul 0 16 UI 0 16 Ul 0 16 Ul Voltage abs 22 4 mv 22 4 mv 178 mv 178 mv 222 mw 222 mv o o o o o Voltage abs 22 4 mY 22 4 mV 178 mY 178 mY 222 mY 222 mY The rows Relative Time and Voltage define the positions of the measurement points The last row shows the measured bit error rates at these points The Relative Time of the measured points refers to the actual sampling point If synchronization and auto alignment were successful the current sampling point is the optimum sampling point Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Instead of UI one Unit Inter
527. ually exclusive with sRJ In addition an external jitter signal can be applied at the Delay Ctrl Input Note that all these sources can be individually switched on or off This is a special modulator that modulates the generated clock with a low frequency signal The resulting jitter superimposes all other jitter components Three sources are provided for modulating the clock Spread Spectrum triangular residual Spread Spectrum and sinusoidal Residual Spread Spectrum and sinusoidal can not be used simultaneously Two delay lines are provided for generating data jitter a 610ps delay line and a 220ps delay line Considering the data output they are connected in series The 610ps delay line is provided for generating large jitter amplitudes at low bit rates It cannot be used for bit rates higher than 3 37Gbit s The source for periodic jitter PJ 1 and PJ 2 and BUJ can be connected to the 610ps delay line or to the 220ps delay line All other sources can only be connected to the 220ps delay line One 220ps delay line is provided for generating clock jitter This delay line can be used like the corresponding one in the data path The two 220ps delay lines in the Clock and Data path can be enabled independently however they share the same jitter sources If the jitter sources are routed to the 220ps delay lines then at least on of them has to be enabled The Aux Data Output always follows the data output Any jitter that is appli
528. ulate the Illegal Symbol Ratio tca Saoi eae Illegal Symbol Counter Cga oymtao Received Symbol Count Expected Filler Symbol Count The total number of Expected Filler Symbol Count received in a time interval is displayed here The following is the expression used to calculate the Expected Filler Symbol Count Expected Filler Symbol Count Filler Symbols in Pattern x Pattern Count Data Rate Ratio The ratio of the number of filler symbol count to the expected filler symbol count received in the current or last completed accumulation period specified by the gate period is displayed here Agilent J BERT N4903B High Performance Serial BERT 387 7 Evaluating Results 388 The raw number of received symbols by the error detector equals the sum of filler primitives which are being dropped before comparison and the actually expected symbols The actually expected symbols may be further distinguished into compared symbols and not compared symbols when using block mode The number of filler primitives is a key indicator for DUT behavior This number must also visible if there is NO pattern lock Furthermore number of filler primitives that were contained in the expected memory before it was loaded to the error detector is also shown in Accumulated Results window 1 If the filler primitive counter remains zero this indicates that a USB3 or SATA product is in the wrong test mode 2 Ifthe number of filler primitives is higher than
529. ultiplier function is enabled a Clock Multiplier DATAOUT menu entry is added to the External Instrument s sub menu 3 To disable the Clock Multiplier function clear the Clock Multiplier function check box present on the Config window It will disable the clock multiplier function and the Clock Multiplier DATAOUT menu item disappears How to Enable Disable N4876A Multiplexer Function To enable disable the multiplexer function Agilent J BERT N4903B High Performance Serial BERT 47 2 Setting up External Instrument s 1 Select the Multiplexer function from the given list It opens the Multiplexer Connection dialog The following image shows the an example of N4876A Multiplexer Connection Dialog m F r IE ro r Ez me Connect the data and aux data output of the pattern generator to the input of the N4876A Connect both instruments via USB on the rear panel Cancel Help 2 Click Enable button It enables multiplexer function that is connected between 3 48 Data and Aux Data Output of Serial BERT s Pattern Generator and the input of the N4876A Enable Once the multiplexer function is enabled a Multiplexer 4876A menu entry is added to the External Instrument s sub menu To disable the multiplexer function clear the Multiplexer function check box present on the Config window It will disable the multiplexer function and the Multiplexer 4876A menu item disappears Agilent J BERT N4903B
530. umFSR FSR results Agilent J BERT N4903B High Performance Serial BERT 369 7 Evaluating Results m FSR 0 000 Ear Filler Symbol Counts The FSR is the current FSR calculated upon a period of 200ms The AccumFSR shows either the accumulated FSR of the current accumulation or if no accumulation is running the results of the most recent accumulation This enables you to monitor real time FSR behavior as you do things such as to manually adjust the sampling point to add errors or to make adjustments to your device Accumulated Actual Button Click this button to toggle between FSR or AccumFSR results Error Count Accum Error Count This area displays the actual filler symbol error count or cumulative filler symbol count DER Results Window This window displays the actual DER or cumulative AccumDER DER results 370 Agilent J BERT N4903B High Performance Serial BERT Evaluating Results 7 ZR DER 0 000 Illegal Disparity Change Counts 0 The DER is the current DER calculated upon a period of 200ms The AccumDER shows either the accumulated DER of the current accumulation or if no accumulation is running the results of the most recent accumulation This enables you to monitor real time DER behavior as you do things such as to manually adjust the sampling point to add errors or to make adjustments to your device Accumulated Actual Button Click this button to toggle between
531. unt of time that can pass as you touch the display screen You can preview your adjustment by clicking touching the Test Box Using the On Screen Keyboard You can use the on screen keyboard to enter text in any text field It can be used along with the standard keyboard or keypad 1 Click in a text field to position the cursor in it Agilent J BERT N4903B High Performance Serial BERT Customizing the Instrument 10 2 Press the On Screen Keyboard button to open the on screen keyboard 3 Click letters or numbers as desired For numbers you must add the unit by pressing the G n k m M y or x1 front panel button 4 Press the On Screen Keyboard button again to close the on screen keyboard 5 Continue with your work NOTE You can also access the on screen keyboard from the Utilities menu A check mark appears next to the menu entry indicating that it has been selected Using an External Monitor To switch between the Serial BERT s built in display and an external monitor do the following 1 Connect the external monitor with an appropriate power supply and with the VGA connector on the Serial BERT s rear panel 2 If the title bar of the Serial BERT user interface is not visible select Title Bar from the Utility menu 3 Minimize the Serial BERT user interface window 4 Inthe system tray next to the clock in the lower right corner right click the Displayicon to open the menu with the available options V 11 50 AM 5
532. uration for a 2 7 1 127 bit pattern 4555188 BI The following table describes the operation of XOR ing two points for the different patterns Table 11 Sequence Length Shift Register Configuration 27 1 D7 D 1 0 inverted 210 1 D D 1 0 inverted 211 1 D D 1 0 inverted 215 1 D 5 D amp 1 0 inverted 223 1 DZ D 8 1 0 inverted non inverted 231 1 D D 1 0 inverted 223 1p using D D D 8 D D D 1 0 produces the same bit sequence like the PCle 3 0 scrambler when fed with zeroes only The PCle 3 0 scrambler is using D D p 6 p8 D D 1 0 with bit numbering being reverse compared to the N4903B Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 Range of Hardware Generated PRBS Patterns PRBS Lengths The following table lists the hardware generated 2 n 1 PRBS pattern lengths that correspond to different n values This is before the pattern is finally inverted Table 12 n Sequence Length Longest Run of 1 s Longest Run of 0 s 7 127 7 6 10 1 023 10 9 11 2 047 11 10 15 32 767 15 14 23 8 388 607 23 22 31 2 147 483 647 31 30 Decimation of PRBS Patterns A special property of 2 n 1 PRBS patterns is that they can be demultiplexed into the same patterns at slower speeds with different phases This is also true for multiplexing The demux or mux must have n ports Hardware Generated PRBS Procedures The reco
533. urement to test a shielded cable 1 Disable the pattern generator outputs by pressing the OV Disable button in the PG Setup gt Data Output screen Use a shielded cable to connect the pattern generator s Data Out port and the error detector s Data In port Terminate all non connected pattern generator output ports with 50 O Use a shielded cable to connect the pattern generator s Data Out port and the error detector s Data In port Switch to the Pattern menu and press Pattern Select Select an appropriate pattern for example a pure 2 15 1 PRBS for both the pattern generator and the error detector For the pattern generator setup you need to specify the logic levels and the bit rate Select LVPECL levels and Gb Ethernet 1 06250Gb s as clock speed See Setting up the Pattern Generator Concepts on page 121 for more information Set up the error detector so that the input range and the termination matches the pattern generator s levels Select an Input Range from 1V to 3V Setthe Data Termination to 1 3V Setthe Clock Setup to Clock Data Recovery to get the error detector s clock from the incoming data stream Enable the pattern generator outputs by pressing the OV Disable button Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 9 Press Sync Now and then Auto Align to find the optimum sampling point Check that the synchronization and the alignment were successful None of the err
534. ut Clock Out and Trigger Ref Clock Out ports must be terminated with 50 Q if they are not connected NOTE The error detector s Clock In connector must be terminated with 50 Q if they are not in use The following procedure is recommended when setting up a test 1 If your DUT can handle OV press the OV Disable button in the PG Setup Data Output or PG Setup Clock Trigger Output screen The pattern generator s Data Out Aux Data Out Clock Out and Trigger Ref Clock Out outputs are set to OV This is indicated in the PG Setup Data Output or PG Setup Clock Trigger Output screen 2 Connect the DUT as necessary 3 Terminate any non connected Data Out Aux Data Out Clock Out and Trigger Ref Clock Out ports normal and complementary 4 Ifthe outputs are disabled press the OV Disable button to enable them To avoid damaging your device set up the devices properly before making connections CAUTION ESD can damage or destroy electronic components Coaxial cables with both ends unconnected may store electrostatic charges Before connecting any coaxial cable of this sort to a device or instrument momentarily short the center and outer conductors of the cable When making connections ensure the proper use of a grounded resistor isolated wrist strap Connect your device according to the diagrams below Agilent J BERT N4903B High Performance Serial BERT 29 1 Planning the Test Diagram 1 Connections for Different In Out
535. utput ports to prevent any damage Output Blanking When the generator is set up to generate a pattern use this checkbox to activate the Aux In port of the pattern generator You can then connect a signal to this port and operate the pattern generator in output blanking mode If a logic low signal is supplied to the Aux In port pattern A is sent to the output port If a logic high is supplied no data is sent to the output port This function is the same as the output blanking function in the Alternate Pattern Control dialog box see Aux In on page 165 for details However this function also works with standard patterns When the generator is set up to generate a sequence this checkbox is disabled You can connect a signal to the Aux In port of the pattern generator The reaction on this signal for example start sequence execution or break loop is defined in the SequenceExpression For details see User Defined Sequences Concepts on page 96 Electrical Idle Use this checkbox to activate Electrical Idle Out Of Band Signaling OOB Electrical Idle is controlled via the existing Error Add Input which will loose its error insertion functionality when used for Electrical Idle The Electrical Idle checkbox is invisible when the N4876A Multiplexer is connected to the N4903B Serial BERT Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 De Emphasis Pre Cursor Use this text box to
536. ux Out This port can be used to output the clock signal or the data signal to another connected device The following figure shows how the clock signal is directed to Aux Out in Clock mode CLOCK MODE CLK IN AUX OUT T CDR Recovered CLK DATA IN Comparator 7T e internal Signal 0 1 Threshold Analyzer Data In Port Termination To ensure a valid setup and to protect the devices from damage proper termination must be specified for both Data In connections You can specify the termination by entering the termination voltage in the respective field CAUTION Selecting the wrong termination may damage your device 170 Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 Why Can Wrong Terminations Damage Your Device Choosing wrong terminations may cause your device to output voltage levels that are not as expected It may also cause excessive current or current flow in the wrong direction which can damage your device NOTE The clock termination is set to 50 Ohm AC coupled and cannot be changed Inputs and Outputs Procedures To select the termination for the error detector Prerequisite You must know the termination voltage of the data signal that your NOTE DUT sends to the error detector 1 Inthe ED Setup menu press Sampling Point Setup to access the Sampling Point Setup window 2 Click the Edit bu
537. val is equal to one system clock period the relative time can also be specified in seconds The Voltage is the decision threshold voltage at this measurement point The voltages of the measurement points can be set as absolute voltages as offset voltages or as percentages This is done on the Parameters page of the Properties dialog By default the six measurement points are symmetrically placed as illustrated in the figure below Threshold voltage Analyzer threshold Actual sampling point These settings can be changed and up to 32 measurement points can be defined NOTE Critical areas at the error detector are generally close to 0 5 UI Fast Eye Mask Procedures This section shows how to set up and perform a Fast Eye Mask measurement As an example we measure the eye of a shielded cable This requires the following steps Preparing the Fast Eye Mask Measurement See How to Prepare the Fast Eye Mask Measurement on page 310 Executing the Fast Eye Mask Measurement See How to Execute the Fast Eye Mask Measurement on page 310 Agilent J BERT N4903B High Performance Serial BERT 309 6 Advanced Analysis 310 Optimizing the View of the Results See How to Optimize the View of the Results on page 311 How to Prepare the Fast Eye Mask Measurement To prepare a Fast Eye Mask measurement to test a shielded cable 1 Disable the pattern generator outputs by pressing
538. values Total Jitter Peak to Peak Peak to peak value for total jitter Calculated as the pulse period unit interval minus the Phase Margin Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Jitter Mean Mean value for total jitter Calculated as the weighted average of the left edge jitter histogram Random Jitter RMS The total jitter component with Gaussian distribution After transforming a contiguous range of measured points into Q space and performing a linear regression it is calculated as the mean of the sigmas of the two straight lines The contiguous range is limited by the the BER Threshold and the Min BER for RJ DJ Separation threshold Deterministic Jitter The total jitter component with non Gaussian distribution After transforming a contiguous range of measured points into Q space and performing a linear regression it is calculated as the period minus the difference between the means of the two straight lines Estimated Total Jitter A forecast of the expected jitter for very low bit error rates After extrapolating the measured BER curves it is calculated as the period minus the expected width of the eye opening No of points This is the number of points that has been measured between the BER Threshold and the Min BER for RJ DJ Separation threshold It is displayed for both slopes This number has to be greater than 2 for the RJ DJ and estimated TJ values to be appli
539. ver the execution of the block begins provided that the Trigger Out port is set to Sequence Trigger mode see also Sequence Trigger on page 147 Create Pattern Click this button to launch Create New Pattern dialog See Create New Pattern on page 115 for details Edit Pattern s Click this button to open the Edit Pattern Window for Sequencer Pattern Editor See Edit Pattern Window for Sequencer Pattern Editor on page 110 for details Loops Within a Sequence A loop defines the transition from the end of a block to the beginning of the same or a previous block It is not possible to jump into an existing loop It is also not possible to specify loops within loops exept the default overall loop For information on how to create a loop see Creating a New Sequence on page 99 Agilent J BERT N4903B High Performance Serial BERT 103 3 Setting up Patterns Loop Delete Button Loop End Condition Button NOTE 104 m Length 2560 rc 2skp rd ptrn A rc 4skp rd ptr A Every loop has a delete button x and a loop condition button Clicking the delete button x removes the loop from the sequence The initial default setting of the loop end condition is INF inite which means this loop will continue until the instrument is switched off Clicking the loop end condition button allows you to specify the loop end condition Choices are Counted The loop ends after the specified number of repetitions Sequence e
540. vice Click Finish to close the wizard Assigning a VISA Alias Name Assigning a unique VISA alias name to the N4876A simplifies remote programming and helps to make programs portable The Serial BERT software requires a specific VISA alias name to access the N4876A Using the Agilent Connection Expert The Agilent Connection Expert is available on from revision 15 5 of the Agilent IO Libraries Suite 1 Ifthe Agilent Connection Expert does not pop up automatically click the Agilent 10 Control icon in the task bar and open it from the menu 2 Inthe Instrument 1 0 panel For N4876A click UsbInstrumentl UsbInstrument1 is the default alias name for N4876A 3 Click Change Properties and change the alias name For N4876A change the alias name to N4876A Agilent J BERT N4903B High Performance Serial BERT 547 10 Customizing the Instrument VISA alias NA876A VISA address US80 0x0957 0x751 8 Proto_004 0 IN N4876A is the alias used by the software 4 Click OK The Agilent Connection Expert for N4876A shows a window like the following Ej Properties Interactive IO ME Add Instrument W9 Add Interface RefeshAl Undo Delete used fotonatinarrogan eme e emm M A N4903B ENVT S COMI ASRL1 Change Properties EJ cow ASRL2 cpio VISA Alias N4876A e pi GPIB1 in GPIB2 This alias refers to the following instrument SA LAN TCPIPO USBO 0x0957 0x7518 Proto_004
541. wards Logarithmic Y Ratio fi 5000 Min Value Start fi 00 0 mUI iv Minimum Jitter Curve Load N4903B J TolStandards MaxCurve jcs Maximum Jitter Curve Load N4903B J TolStandards MaxCurve jcs iv Show Jitter Curve OK Cancel Apply Help 2 If desired change the related parameters 3 Check the Minimum Jitter Curve and Maximum Jitter Curve checkboxs For details see Minimum and Maximum Jitter Curve on page 446 4 Press the respective Load buttons to load your own defined minimum and maximum jitter curve stored in a text file 440 Agilent J BERT N4903B High Performance Serial BERT Jitter Tolerance Tests 8 This opens a dialog box which allows you to load a file By default the browser searches for files with the suffix jcs but you can load any file The required file format is described in detail in User Defined Standards on page NOTE 452 5 Check the Show Jitter Curve checkbox It shows the jitter curve on the Jitter Tolerance Characterization graph Before Running the Test For DUTs that require a training sequence it is essential to prevent clock loss at the beginning or end of the test To avoid clock loss the sinusoidal jitter component has to be enabled Since the sinusoidal jitter for the test is generated internally the jitter amplitude must be set to zero 1 Click the Jitter Setup menu item from the Jitter submenu l Jitter Setup This opens the Jitter Setup window 2 En
542. when connecting your instrument to a device or test setup The diagram below shows a device that is AC coupled Notice that the capacitor is part of the test setup DUT Dutput You can use an external bias network also known as a bias tee to power your device You must ensure however that the network is oriented correctly If it is not damage may occur to your device or instrument The diagram below shows a bias tee that is positioned correctly Notice that the pattern generator s outputs are protected by the blocking capacitor DC Bias Bias Tee Agilent J BERT N4903B High Performance Serial BERT Setting up the Pattern Generator 4 Input and Output Ports Procedures You have to do the following to set up the pattern generator s output ports Setting Logic Levels and Terminations Before you can start sending signals to your device you have to set the logic levels and terminations 1 From the PG Setup submenu click the Data Outputmenu item to modify Data and Aux Data and the Clock Trigger Output menu item to modify Clock and Trigger I Soe Data Output an Clock T rigger Output 2 In the Data Logic Level list select the logic family that is appropriate for the data input of your device 3 In the Clock Logic Level list select the logic family that is appropriate for the clock input of your device When you modify the Logic Level of an output the instrument automatically sets the appropriate volt
543. wse Length in Bits 2560 Using this dialog you can Choose Select Trace as Data Out Aux Data Out to create a Standard pattern or as Both to create an Alternate pattern Enter a Description for the pattern Provide a File Name and click Browse to locate the new pattern Specify the pattern s Length in Bits Symbol Settings The Symbol Settings dialog is used to select the symbol mode It displays the calculated output levels based on the current Data Out and Aux Data Out levels The Symbol Settings dialog is shown in the figure below Symbol Settings 00B with External adder between Data Out and Aux Da OR Off OOB with External adder between Data Out and Aux Data Out PAM4 with External adder between Data Out and Aux Data Out Symbol DataOut Aux Data Out Output Level 0 0 0 my 0 1 150 mv 1 1 300 mV Cancel Apply Help Agilent J BERT N4903B High Performance Serial BERT Setting up Patterns 3 This dialog contains a drop down list which provides the following options Off Mode Displays the pattern editor without any symbol coding 00B Mode Displays bit combinations of both channels as shown in the table below Table 16 Symbol Data Out Aux Data Out 0 0 0 Z 0 1 1 1 1 The OOB Mode is only valid in Bin and Symbol representations The Hex NOTE P WEN representation is disabled in this mode NOTE If the bit combinations are not matched in the table then U will be shown in
544. x Agilent J BERT N4903B High Performance Serial BERT Advanced Analysis 6 Parameters Tab NOTE If you modify the parameters on this page you have to rerun the measurement to update the results Set the criteria for moving to the next sample point Number of Compared Bits After this number of compared bits the measurement stops for the current sample point and moves to the next one The default is 1 million bits That means you can measure a bit error rate down to 10 one error per million A smaller number reduces the duration of the whole Eye Opening measurement A larger number increases the precision of the measured bit error rates e Number of Errors After this number of errors the measurement stops for the current sample point and moves to the next one This allows you to speed up the measurement You can switch off this option if only the number of compared bits is important NOTE The measurement moves to the next sample point when the first of the two criteria is reached Set the criteria for the sample delay Resolution Specifies the distance between sampling points The lower this value the more sampling points you have in a unit interval You can enter the resolution in UI or ps ns S The timebase of the display is set on the View tab The default is 0 01 UI that means 100 points per unit interval will be measured Edge Resolution Optimization Turns the resolution optimization on or off
545. xecution continues with the next block Infinite An infinite loop can be broken by Aux In High Aux In Low The loop ends when Auxiliary Input is high or low Sequence execution continues with the next block Aux In Rising Aux In Falling The loop ends when a rising or falling edge at the Auxiliary Input is detected Sequence execution continues with the next block Manual The loop ends when the Break button of the Sequence Editor is clicked Sequence execution continues with the next block The pattern generator reacts on the end condition as soon as the pattern has finished Sequence Editor Toolbox The Sequence Editor toolbox provides the following functions Agilent J BERT N4903B High Performance Serial BERT Table 14 Setting up Patterns 3 Icon LI m o e Agilent J BERT N4903B High Performance Serial BERT Name New Open Save Save As To PG Description Click this icon to create a new sequence This opens the Properties dialog See Creating a New Sequence on page 99 for details Click this icon to open a sequence from a file See Loading a Sequence From a File on page 100 for details Click this icon to save the current sequence in its original file Click this icon to save the current sequence in a new file A Save As dialog box opens for you to specify name and path for the file See Saving the Sequence in a File on page 100 for details
546. y Work on page 174 for more information If an external clock is used there is no delay The clock signal must however be continuous Agilent J BERT N4903B High Performance Serial BERT Setting up the Error Detector 5 3 The error detector can then synchronize to the incoming signal For PRBS the first received bits are used to seed the synchronization If there is an errored bit in this phase the synchronization fails bad burst For memory based patterns a unique 48 bit detect word is used for the synchronization This pattern should be available one time only in the pattern If the detect word is not found synchronization fails bad burst 4 If synchronization has been established the received data is analyzed until the signal at Gate In goes high indicating the end of the burst The data should continue arriving at Data In slightly longer than the Gate In signal 5 Incase of failed synchronization the total burst counter and bad burst counter are incremented The bit counters total count error count etc are not incremented 6 Atthe end of the burst the data is analyzed the bits are counted as are the errors total errored Os errored 1s If the BER is higher than the Burst Sync Threshold in the Pattern Sync Setup dialog box the burst is considered a bad burst and the total burst count and bad burst count are incremented 7 Forgood bursts the total burst count is incremented and the bits are analyzed
547. y measured sample infinitely Measurement time has to be at least long enough for the required target BER e g 1 12 at a confidence level of 95 A measurement is considered error free when there are no samples taken outside the regular signal trace Selection of the optimum Input Timing setting is done as shown above for the BER measurement 4 Selection of the optimum Input Timing setting is done as shown above for the BER measurement 5 Optionally reload the previously stored instrument state Agilent J BERT N4903B High Performance Serial BERT 483 9 Solving Problems 484 Setup Description How to Create the Calibration Settings Use the following calibration settings Preset Instrument State Enable De Emphasis Set all cursors to 0dB Set Output Offset at N4916B gt 0V Set Output Amplitude at N4916B gt 400mV Set Data rate to the maximum possible data rate of your setup This depends on the maximum data rate of the PG ED and N4916B Note that a N4916B that is specified for 10 5Gb s should be calibrated at 12 5Gb s this makes selecting the optimum input timing setting easier A N4916B that is specified up to 14 2Gb s shall bei either calibrated at the ED s maximum data rate 12 5 Gb s or the PG s maximum data rate 7Gb s 13 5Gb s or 14 2Gb s Use either the PG s or ED s maximum data rate whatever is less If calibrating using an oscilloscope then use either 7Gb s 12 5Gb s or 14 2Gb s dependin
548. your PC 3 Import the measurement log file If your application has an import wizard you may need to indicate that data is delimited with commas 4 Once the file has been imported you may need to resize columns The imported log file should appear similar to the following example During measurement logging the Serial BERT logs data in ten second intervals Your log file may be missing up to the last ten seconds of data To avoid this condition accumulate for 10 seconds longer than desired A B c D E 1 Description test 2 Log Time 05 29 2000 12 14 58 334 3 Activation Mode Single 4 Accumulation Measurement ELAPSED SECONDS Accumulation Limit 60 000000 6 7 Elapsed Time BIT COUNT BIT FREQUENCY ED FREQUENC 8 0 1 249E 08 249E 09 2 49E 9 0 2 249E 08 2 49E 09 249E 10 0 3 2 49E 08 2 49E 09 249E 11 04 249E408 249E409 2 49E 12 0 5 2 49E 08 2 496409 2 49E You can view and analyze the following information in the measurement log file Agilent J BERT N4903B High Performance Serial BERT Table 32 Solving Problems 9 Instantaneous BIT COUNT ERROR COUNT ERROR RATIO ERRORED 0 COUNT ERRORED 0 RATIO ERRORED 1 COUNT ERRORED 1 RATIO SYNC LOSS PG CLOCK LOSS ED CLOCK LOSS DATA LOSS Agilent J BERT N4903B High Performance Serial BERT Cumulative BIT COUNT Cumulative ERROR COUNT Cumulative ERROR RATIO Cumulative ERRORED 0 COUNT Cumulative ERRORED 0 RATIO Cumulative

Agilent J-BERT N4903B High-Performance Serial BERT

Contents

Download Pdf Manuals

Related Search

Related Contents