28G-&-56G-NRZ&PAM4-设计中的性能验证

VerifyingNRZandPAM4Performancein28G&56GDesigns余洋OceanYuApplicationEngineerofTektronixChinaNovember4,2014November4,2014AgendaPAMTechnologyOverview–PAMvsNRZmethodsofcommunication–WherePAMfitsinthestandardscommunity–ChallengeswithPAMsignalingRecommendedmethodsforvalidatingPAMsignaling–TransmitterTest–Oscilloscope-basedAnalysistools–ReceiverTest–SignalGenerationandErrorDetectiontoolsQ&A/ResourcesforyouVerifyingPAM4Performance-ALightwaveWebinar2November4,2014WhatisPAM?VerifyingPAM4Performance-ALightwaveWebinar3PulseAmplitudeModulation–PAM4combinestwobitstreamsanduses4levelstoencode2bitsinto1UI–ForExample,56Gbit/sPAM4runsatasymbolrateof28GBaudLSBstreamMSBstreamCombinedPAM4streamNovember4,2014WhatarethedifferencesbetweenPAM4andNRZ?PAM4–4Levels-3Eyes–SensitivetoSNR(eyessmaller)–2bitsinto1UI–½BaudRateforsamedatathroughput(28GBaud=56Gbps)–Addscomplexity/costtoTx/RxVerifyingPAM4Performance-ALightwaveWebinar4NRZ–2Levels-1Eye–LessSensitivetoSNR–1bitin1UI–2XBaudRateforsamedatathroughput(28GBaud=28Gbps)–LessexpensiveTx/RxNovember4,2014200G/400GMarketSituationCurrentNRZdesigns(25GBd)havetwooptionsforincreasingthroughput–56GBdNRZ–Targetedtowardsupportingabroadsetofreachobjectivesrangingfromdietodie(10mmUSR),chiptoopticalengine(50mmXSR),chiptopluggablemodule(100mmVSR)andchiptochip(500mmMR).–28GBdPAM4,withearlydiscussionsof56GBdPAM4–Extendbeyond3”channelsVerifyingPAM4Performance-ALightwaveWebinar5November4,2014NextGenerationStandardsUpdateIEEE802.3bj–PAM4(100GBASE-KP44x13.8GBd)isdefinedintheIEEE802.3bj–Goalwastoleveragelegacybackplanes,howeverre-designedbackplanesusingNRZ(100GBASE-KR44x25.78GBd)waswidelyadoptedOIFCEI4.0DraftSpecification–56GBdNRZforUSRapplications–28GBdPAM4forCEI-56G-LRVerifyingPAM4Performance-ALightwaveWebinar6November4,20147PAMSignalingChallengesVerifyingPAM4Performance-ALightwaveWebinarNovember4,2014PAM4MeasurementChallengesWhatdomeasurementsonindividualeyesmeanrelativetooveralllinkBERExtendNoiseandJitterdecompositionconceptstoPAM4–Forexample,whatdoesaBERcontourmeanon3eyesrelativetotheoveralllinkTroubleshootandunderstandtheeffectsofISIduetomultipletransitionsConceptofaggregatemeasurements–linearity,noise,etc.HorizontalPAM4signalmeasurementVerticalPAM4signalmeasurementVerifyingPAM4Performance-ALightwaveWebinar8November4,2014PAM4MeasurementChallengesPAM4BERmeasurementmethodologies?–ContourplotopeningreducedPAM4(vsNRZ)duetoaddeddeterministicjitter/noise–Industryspecsmovingto~1E-5BERforPAM4,asopposedto1E-12/1E-15forNRZ–LowBERmeasurementsarecriticaltoensuredetectionoflow-probabilityevents/floors9VerifyingPAM4Performance-ALightwaveWebinarNovember4,2014TestMethodologiesforPAMSignalingValidationDeviceUnderTestTransceiverchipset,GearboxTestSignalSignalGenerationSignalAcquisitionAnalysisSoftwareBuildateststrategythatworkswithyourmeasurementtask10•Optical&ElectricalSignals?•AdequateDynamicRange?•EventCapture/Trigger?•MeasurementsystemB/W?VerifyingPAM4Performance-ALightwaveWebinar•Threshold/leveldetection?•Renderamulti-leveleye?•Providejitter&noiseresults?•MakeaBERmeasurement?•Comparetospecreference?•Fastenoughsignalrisetime?•DeepPatternLength(PRBS)?•Flexiblepatterndefinition?•Jitter•AmplitudeConsiderationsforaPAM4SignalAcquisitionSystemVerification/ComplianceManufacturingChipAdd-InCardsSystemR&DSamplingScopesReal-timeScopesThemostversatiletoolforallareasofhigh-speeddigitalandanalogapplicationsSingleshotacquisitionidealforpostprocessingMostadvancedtriggersystemtoidentifyuniqueeventsMostflexiblesoftware-basedclockrecoveryDebuggingandTroubleshootingForapplicationsthatplacetoppriorityonwaveformprecisionOver60dBofdynamicrange,idealforPAMHighBWto100GhzRepetitivewaveformsVeryLowJitterNoiseFloorVerifyingPAM4Performance-ALightwaveWebinar11November4,2014PAM4TXMeasurementsInheritedfrom100GBASE-KP4(PAM4for100GbEbackplane)Measurementsacrosstheentiresignal–Levelseparationmismatch-RLM–TimeandLevelDeviation–LevelThickness–EyeCenterDeviations–CompositeEyeWidthIndividualEyeMeasurements–Jitter–Noise–EyeHeight–EyeWidthFuturedirection:LinearFit,SNRoverallvia‘COM’(ChannelOperationalMargin),andMasktestarealsobeingconsideredVerifyingPAM4Performance-ALightwaveWebinar12November4,2014TXMeasurementsPAM4LevelsandThresholdsLevel3Threshold2-3Level2Level1Level0Threshold1-2Threshold0-1Levelsareusedfordeterminingthethresholds,calculatingnoise,linearity,timedeviationThresholdsareusedtocalculatejitter,eyecentersandothereyeparametersVerifyingPAM4Performance-ALightwaveWebinar13November4,2014TXMeasurementsLinearityMeasurementLinearityestimatecanbecalculatedforarbitrarypatternatthecenteroftheUI(averagedISIeffectsareincluded)MeasureshowevenlythefoursignallevelsaredistributedMeasuredusingtheLinearityPattern802.3bjrequiresRLM0.92Measurementmadeonbits7and8oneachlevelNovember4,2014TXMeasurementsTimeandLevelDeviationCharacterizestimingpropertiesofthesignalandnon-linearartifactsinthedriverLooksattheskewbetweenthenarrowestpointoneachlevelDoneonQPRBS13patternVerifyingPAM4Performance-ALightwaveWebinar15November4,2014TXMeasurements