TP2 Waveform Capture - PowerPoint PPT Presentation

About This Presentation
Title:

TP2 Waveform Capture

Description:

Background Information Three methods of Jitter/Noise Analysis Analyze an LF square-wave +Simplest Measurement equipment setup Signal far from natural ... – PowerPoint PPT presentation

Number of Views:25
Avg rating:3.0/5.0
Slides: 7
Provided by: Pavel81
Learn more at: https://www.ieee802.org
Category:

less

Transcript and Presenter's Notes

Title: TP2 Waveform Capture


1
TP2 Waveform Capture
  • Pavel Zivny with input from Greg LeCheminant

2
TOC
  • Proposal for the TP2 CR and Trigger method
  • Issues for jitter/noise analysis
  • TP2 uncorrelated jitter/noise measurements

3
TP2 - CR and Trigger methodProposal
  • Record the single valued data pattern waveform
    (rather than the eye diagram) using a
    synchronizing timing reference obtained with a
    clock recovery unit having a characteristic as
    described in 802.3ae 52.9.7 Transmitter optical
    waveform. If a sampling oscilloscope is used,
    the pattern waveform is acquired by means of
    triggering on divided recovered clock (the
    recovered clock is divided by the pattern length
    to produce a trigger edge occurring once per
    pattern repetition).
  • Only bits ltxgt through ltygt of pattern ltzgt are
    analyzed. The bits are recognized by their
    position relative to the longest transition-less
    sequence in the pattern. In particular the bits
    are located ltkgt UIs after the sequence lt?gt.

See next page
4
TP2 - CR and Trigger methodSupplementary info
CRU per 802.3ae
From 802.3ae 52.9.7 Transmitter optical
waveform A clock recovery unit (CRU) should be
used to trigger the scope for mask measurements
as shown in Figure 52-9. It should have a
high-frequency corner bandwidth of less than or
equal to 4 MHz and a slope of -20 dB/decade."
5
Issues for jitter/noise analysis
  • Measured uncorrelated jitter and noise are
    assumed Gaussian or other high probability
    jitter. Low probability, high PkPk value
    jitter/noise escapes detection through RMS test.
    Mask test captures this, if thered be no mask
    test this would need addressing in another way.
  • Pj, Pn are both bundled into R(andom) component
  • Thats a pessimism. Comment from Xmitter
    designers desireable is P ever a problem?
  • Always measure with the ORR on. Is there a
    consensus on this?

6
Background InformationThree methods of
Jitter/Noise Analysis
  • Analyze an LF square-wave
  • Simplest Measurement equipment setup
  • - Signal far from natural
  • - Dedicated pattern not useful for other
    measurements
  • Analyze Dedicated bit(s) within a Multi-use
    348-bit Stress Pattern
  • Useful, near-natural pattern
  • Rj and Rn measurements simple easily
    verifiable
  • - Needs Trigger on Pattern (with e.g. a bit
    countdown trigger)
  • - Needs a bit search (not implemented today)
  • Perform complete Jitter/Noise analysis (Rj and
    Rn, DDj and DDn) on the whole Pattern (e.g. the
    Multi-use 348-bit Stress Pattern)
  • Useful, near-natural pattern
  • Complete Jitter/Noise measured (rather than just
    one, possibly special, bit)
  • - Needs Trigger on Pattern (with e.g. a bit
    countdown trigger)
  • - Needs complete jitter/noise analysis package
    (single source, marginal availability today)

Current choice
Write a Comment
User Comments (0)
About PowerShow.com