key培訓(xùn)示波器產(chǎn)品scope product

1、S800 Training SharingScope Products and Probes update Zhidong TianDigital AEZhidong_March 29, 2011Agilent Confidential1Internal Terminology“Infiniium Classic” = historical Infiniium GUI “Kraken” = InfiniiView (offline Infiniium app running on a PC)“Magneto” = new GUI running on an Infiniium scopeLyn

2、x = GREX-class MSORaptor H = 9000-class with 3X noise reduction (high-res)Brick = new current/voltage probe for down to sub mA/mV sensitivityPage 2Group/Presentation TitleAgilent Restricted6 March 2022Infiniium Classic Win XP & Win 7Kraken Win XP & Win 7Infiniium SW RoadmapMarch 6, 2022Agile

3、nt Confidential3Q1FY14MagnetoRaptor, TREX, GREX, QREX + future frames$500 For pay upgradeNowInfiniium Classic Win XP & Win 7Infiniium Classic SW ReleasesMarch 6, 2022Agilent Confidential4Q1FY14Nov124.1 (released) LLI & UniPro decodes, BUJ, Raptor H4.2 Brick probing support4.5 Lynx MSO suppor

4、tSupport Critical defect fixesNo enhancements anticipatedQ213Q313Kraken Win XP & Win 7Kraken ReleasesMarch 6, 2022Agilent Confidential5Q1131.1 Defect fixes5.0Align with Magneto feature set including MSO and API supportQ1FY14Magneto ReleasesMarch 6, 2022Agilent Confidential6MagnetoRaptor, TREX, G

5、REX, QREX + future framesKraken GUI on Infiniium +lots of Infiniium features not yet in KrakenQ1FY14Infiniium HW RoadmapPage 7Group/Presentation TitleAgilent Restricted6 March 2022Infiniium Classic FY14Next gen 9000Focus on signal quality and higher BWNext-gen 90000A Focus on signal quality and high

6、er BWFY15MagnetoQ113Q213G8, G13, Lynx IntroLynx MSOReleasing 4.10- Tailfit algorithm- Sine wave cal- InfiniiSim with jitter free- Deeper memory on roll mode- G-REX 13 supportMarch 29, 2011Agilent Confidential8Help Us PrioritizeMarch 29, 2011Agilent ConfidentialLimit to top 5 choices1,2,3,4,5 where 5

7、=most important Competitive BUJ MeasurementsMarch 29, 2011Agilent Confidential10ABUJ: Crosstalk or Ground bounceAmplitude interference uncorrelated with data and not periodic in nature.Crosstalk Interference ModelNo crosstalkWith crosstalkDvVictimAggressorDtVictim OutDt = Dv/SlopevictimTotal Jitter

8、(TJ)DeterministicJitter (DJ)Random Jitter(RJ)Correlated with Data (DDJ)Uncorrelated with Data (BUJ)DutyCycle Distortion (DCD)InterSymbol Interference (ISI)Periodic(PJ)Non Periodic (ABUJ)Gaussians (s, RJRMS)Jitter ComponentsXtalkNon Linear CREventsThermalShot1/fBurstTr, Tf DSettling TimeReflectionsCl

9、ocksBoundedUnBoundedNon flat Freq ResponseXtalk-2-101200.10.20.30.40.50.60.70.80.9How is cross-talk jitter detected?Yes: it shows up as Data-Dependent Jitter (jitter that is correlated with the pattern). No: it will appear random, but it is not. (Bounded, Uncorrelated). Is the aggressor signal corre

10、lated with the victim signal?e.g. PRBS7e.g. PRBS15Random jitter is modeled as Gaussian. It is unbounded.Crosstalk jitter is bounded. -4-3-2-10123400.10.20.30.40.50.60.70.80.90.20.40.60.811.21.41.61.82x 104-0.500.511.5fourier coefficientsEZJit Today:Estimates RJ as noise floor of jitter spectrumGet R

11、MS value of noiseWhen BUJ 0, however, this approach over-estimates RJ. It returns RJ + BUJ.Very accurate when BUJ=0 This makes BER look worse than it is.ABUJ Observations and Measurement(ABUJ= Aperiodic Bounded Uncorrelated Jitter)1. View ABUJ in time domain2. Techniques to Evaluate Something is wro

12、ng here.Using the slope continuity concept we expect the extrapolated curve to look like this.The RJ/PJ spectral extraction doesnt deal with ABUJ well. The RJ is overestimated severely. New Jitter AlgorithmAgilents 4.1 software will feature new tail fitting algorithm for both noise and jitterThree j

13、itter separation methods:-Spectral Wide-Spectral Narrow-Tail FitSelect Tail Fit in the EZJIT+ or EZJIT Complete menuGaussian Extraction MethodRJ Extraction4-30-20-10010203000.20.40.60.811.21.4Histogram and Gaussian fit to right tailjitter, psFit a Gaussian characteristic to the right and left extrem

14、es of the RJ/PJ distribution.246810121400.10.20.30.40.50.60.70.80.9Histogram Fits. True RJrms = 2, PJmax = 5 Actual Data is Never smoothMeasurement DetailWhat makes Tail fitting hard051015-0.200.20.40.60.811.2Histogram ObjectHigh PrecisionLow accuracyLow PrecisionHigh accuracy6789101112131415-0.8-0.

15、6-0.4-0.200.20.40.60.8errorFit WindowDJ endNoisy dataCurve fit errorRJ Extraction4Measurement DetailNew Tail Fit ChartActual Rj/Pj DataSpectral fit to Rj/Pj dataTail fit to Rj/Pj dataBathtub curve comparison-Pink curve shows the actual Rj/Pj histogram data-Gray shows how well the tail fit method tra

16、cks to the pink-Black shows how well the spectral method tracks to the pinkUse this tool to find the best jitter method to useNeed tailfit method to separate BUJ from RjIn this experiment, crosstalk has been added.Notice that the black no longer fits well to the pinkWhy?The Reason spectral method: I

17、s unable to separation Rj from bounded uncorrelated jitter. Over reports Rj Has a difficult time fitting the data in this caseSo how does it compare to the tailfit?Spectral MethodTail Fit MethodSpectral method:Tj 158.7ps Rj 11.24 psPj 790 fsTail fit method:Tj 82.73ps Rj 3.1 psPj 30.96ps BUJ 10.80 ps

18、Notice: The spectral method over reports the Tj, due to over reporting the RjSo lets run an experimentChannel 1of 81134A used to generate 2.5 Gpbs signalChannel 2 of 81134A plus TDR board used to simulate crosstalk on the deviceNo cross talk addedSpectral method:Tj 32.72 ps Rj 1.53 psPj 520 fsTail f

19、it method:Tj 32.15 Rj 1.43 psPj 1.3ps BUJ 530 fsNotice: Slight BUJAdding 50mV of cross talk to the signalSpectral method:Tj 38.07ps Rj 1.98 psPj 300 fsTail fit method:Tj 36.12 Rj 1.7 psPj 2.2ps BUJ 1.06 psNotice: Even with small cross talk additions, the spectral method begins to over report RjAddin

20、g 100mV of cross talk to the signalSpectral method:Tj 47.5 ps Rj 2.85 psPj 180 fsTail fit method:Tj 40.12 Rj 1.7 psPj 5.90ps BUJ 2.29 psNotice: The delta begins to grow between spectral and tail fit modeAdding 200mV of cross talk to the signalSpectral method:Tj 68.5 ps Rj 4.85 psPj 410 fsTail fit me

21、thod:Tj 49.9 Rj 2.0 psPj 12.01ps BUJ 4.42 psNotice: The delta continues to grow between spectral and tail fit modeAdvantage: Improved Estimation of RJrms in presence of BUJ. 010203040506000.511.522.533.544.55RJrms Accuracy when BUJ 0experimentRJ true RJNew RJEZJit RJRJ Estimation with no BUJ:EZJIT i

22、s very accurate010203040506000.511.522.533.544.55RJrms Accuracy when BUJ=0experimentRJ true RJNew RJEZJit RJRJ Estimation with BUJ:New algorithm tracks better.RJRJDisadvantage: Less accuracy when no BUJRemoval of the RjMarch 29, 2011Agilent Confidential28Remove scope random jitterAutomatic mode Meas

23、ure ACVrms at screen position relative to measurement threshold level. Measure average slew rate on every acquisition between upper and lower thresholds. Apply Jitter Measurement Floor equation for given product utilizing measured noise, slew rate and acquisition time span. Compute and report Scope

24、RJrms. Remove Scope RJrms from reported RJrms.Manual mode Can also directly specify Scope RJrms.Put a ? on the reported RJrms if 60dB at 1 MHz)Superb Usability InfiniiMode probing for making differential, single-ended and common mode measurement with a single probe Built-in multi-function scope cont

25、rol for quick access to a variety of scope functions Built-in “Headlight” to see where you are probing Includes solder-in, browser and socketed tips standard AutoProbe interface for auto configuration, auto control and probe power via scope At Industry-leading Affordable Prices Page 62S800 OSC_S12_2

26、Agilent ConfidentialNov 2012InfiniiMax III 25 GHz PC board ZIF Tips (N2838A) More robust, PCB-based ZIF tip Compatible with InfiniiMax III amp and N5439A ZIF head 25 GHz, 100k input Rdiff ( DC) 6:1 attenuation ratio, normal sensitivity 5 tips in a package Will continue to sell ceramic ZIF tips (N544

27、0A, N5447A) N2838-60001 replacement resistor kit (10/kit)Page 63N5439AN2838AS800 OSC_S12_2Agilent ConfidentialNov 2012InfiniiMax III 26 GHz Solder-in Head (N2836A) 26 GHz bandwidth (vs. 16 GHz with N5441A) Compatible with InfiniiMax III probe ampPage 64Frequency Response (Vout/Vin)S800 OSC_S12_2Agil

28、ent ConfidentialNov 2012InfiniiMax III for extreme temperature testing Full probe bandwidth (up to 16 GHz) with slightly higher noise floor -55C +150C with N5441A solder-in headN5450BInfiniiMax III N2800A-03AN5441A+New “B” versionS800 OSC_S12_2Agilent ConfidentialNov 2012E2655C Deskew and PV fixture

29、Page 66Replacing discontinued E2655BE2655BE2655CClear pincher RoHS compliantS800 OSC_S12_2Agilent ConfidentialNov 2012N5380B InfiniiMax II SMA probe headPage 67N5380A, $2,836N5380B, $2,836Fully assembled with a plastic case RoHS compliantBuilt-in strain reliefS800 OSC_S12_2Agilent ConfidentialNov 2012N2894-60002/N2894A 700MHz passive probe 700MHz system BW when used with 4000X 1 GHz/1