Some thoughts about the jitter (tolerance) discussion today -
- The channel will cause channel jitter but not transmitter jitter. Channel
jitter will be more or less linear and correctable.
- I believe there will be components of transmitter jitter that
are not correctable, either because they are non-linear or because they will
have a wider (than sine) spectrum that is asynchronous to the data. At
least, the latter transmitter jitter will combine incoherently with
channel jitter and so must be considered.
- What an EDC system does with sine jitter is a good question. I
suspect it will try to reduce its impact in some way as a tradeoff
against ISI and other noises. Sine jitter is not "natural" jitter, and so if
it causes an EDC system to adjust in an unnatural way, it is probably not a
good choice. That "if" should be answered by someone other than me. If EDC
does not adjust in response to sine jitter, then I feel it's an acceptable
choice because of relative simplicity and history.
- OIF's CEI uses jitter generated by a phase-modulating PRBS source to
represent a transmitter for tolerance testing. The PRBS output is filtered in
a way that softens the pdf into something between uniform to semi-normal, but
yet it is clearly bounded. The pattern is short enough, and its rate fast
enough that the resulting jitter spectrum is fairly broadband and above the
tracking frequency (it is high-frequency jitter). Fibre Channel already does,
and XFP is considering, using PRBS jitter also. I am not necessarily
recommending PRBS jitter (although it does have good merit and is not much
more complex than sine jitter), just pointing out what other groups have done
or are doing.
- I am in support of Lew's thought that if "other" high-frequency jitter
already exists, it can reduce requirements for jitter to be added, details
TBD.
- See XFP and CEI serdes/CDR specs for previous jitter values to
consider.
Thanks, Tom
----- Original Message -----
Sent: Monday, October 18, 2004 9:19
AM
Subject: [10GMMF] TP3 meeting Agenda Oct
19
Dear TP3'ers,
Here is the agenda for tomorrows
meeting:-
Tuesday October 19th at 9am SJ, 5pm UK, 6pm Germany Dial
in (650) 599-0374, Meeting ID:
136169
1. List
Attendees 2. Review meeting
notes from last week 3. Review
Lew's additional motions (which were not heard at Ottawa). These were
forwarded by Lew last week. 4.
TP2/TP3 link budget work. Tom Lindsay's Feedback from TP2 call
Upcoming
items of focus (for subsequent
metings):- i) Static channel
methodology
- noise
loading
- channel types and exact
characteristics
o i.e. do we pick pre, post cursor and quasi symmetric and if so where from?
108 fiber model? ii) OMA measurement
methodology.
Best Regards
Mike
|