Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [10GMMF] TP2 waveform acquisition


Dear Launch study participants

here are minutes from the Task 2 and 4 meeting (December 8th)

please let me know of any comments or suggestions before next Wednesdays call

thanks !


tel: 1 408 524 5110
fax: 1 408 739 0568

Jonathan King
Director, Optical Systems
BigBear Networks
345 Potrero Avenue
Sunnyvale, CA 94085

-----Original Message-----
From: owner-stds-802-3-10gmmf@IEEE.ORG [mailto:owner-stds-802-3-10gmmf@IEEE.ORG] On Behalf Of Tom Lindsay
Sent: Wednesday, December 08, 2004 11:42 PM
Subject: [10GMMF] TP2 waveform acquisition


I had an action from the last TP2 call (12/2) to publish details for TP2 waveform capture, so that allow ClariPhy can analyze the waveform for LRM penalties. You may recall that Lew Aronson was the first to offer to provide waveforms. So, Lew and I exchanged a few emails and came to the following:

·         Connect TP2 directly to the scope with a short jumper. Be sure the O/E is not being overdriven.

·         For tracking information, record filename, device, OMA, average power, currents, test pattern, setup, etc. as appropriate and relevant.

o        ClariPhy does not need this information (except filename), but you should keep the rest as reference.

·         Include a 7.5 GHz Bessel-Thomson filter in the acquisition.

·         Data rate = 10.3125G.

·         Other options are possible, but for consistency at this time, use prbs9.

o        First choice is with an added zero to make 512 bits and max runs of both 1's and 0's. This 512 bit prbs9+1 pattern is attached

o        If you have a prbs9 pattern built into your pattern generator and it is not convenient to program in the attached pattern, then the standard prbs9 pattern will work. However, be sure to note which pattern was used.

§         The standard prbs9 pattern, if used, must be x^9+x^5+1. We are not set up for other prbs9 pattern options.

·         At least one full pattern repetition - include at least a dozen or so extra bits.

o        We ran our acquisition at 55 nsec across the screen (5.5 nsec per division). This is ~567 bits.

·         At least 16 averages.

o        16 is what we have been using.

·         At least 7 samples per UI (more is better, but no need to exceed 16).

o        We ran our acquisitions to acquire 4050 points across the 55 nsec. This results in a bit more than 7 samples per UI.

·         Trigger - for our work, we used the pattern trigger from the generator.

·         An sufficient experimental file format is attached.

o        Other header rows are okay, but "Data:" must immediately precede the numerical data.

o        Single column of Y data.


It may be clear that we are using an Agilent scope for our acquisition. It is an 86100A, and the waveforms are recorded in "Verbose mode".. If folks would like to use a different instrument with other headers or format, we should be able to modify the program input structure. Please let me know.


As a reminder, here are the notes in last week's minutes on why we are doing this:

·         Others should capture waveforms for penalty analysis that represent range that they would like to consider for LRM. I encourage we "stretch the envelope", as EDC brings a paradigm shift, and we're still learning what is good or not good.

·         Waveforms can be simulated or experimental.


Thanks, Tom
phone: (425) 608-0209, or (949) 480-9210
cell: (206) 790-3240

Task 2 mins-Dec8-04.ppt