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

Re: XAUI update from Interim meeting




Mike - hate to do this, but at the end of the day 6/4 at FC, the discussion
was heading to the conclusion that floating the mask meant the jitter budget
was not self-consistent with the belief that receivers use the median (or
mean) of ???, and that this issue was by no means resolved. (I promise I did
not bring this up again - Dave and Bill did).

It WAS resolved that, due to schedule, PI will not consider any more
changes, and that floating the mask will stay in that document.

Tom

----- Original Message -----
From: "Mike Jenkins" <jenkins@xxxxxxxx>
To: "Kesling, Dawson W" <dawson.w.kesling@xxxxxxxxx>
Cc: "Serial PMD reflector (E-mail)" <stds-802-3-hssg-serialpmd@xxxxxxxx>
Sent: Friday, June 08, 2001 6:28 PM
Subject: Re: XAUI update from Interim meeting


>
> Dawson, et al,
>
> Thanks much for your summary on the XAUI spec changes.  I have
> a comment/question concerning the summary item stating:  "The
> templates were centered in the data eye and not allowed to float."
>
> Unless I misunderstand, this change causes an inconsistency in
> the specification.  In Fibre Channel, at least, eye masks (or
> templates) have been defined such that the left-most vertex is
> one half the total jitter (i.e., TJ/2) from the left edge of the
> bit period (i.e., at 0 UI).  The right-most vertex is symmetrically
> TJ/2 from the right edge of the bit period (i.e., 1 UI), as shown
> in my crude character graphics below (if you have fixed-width font).
>
>       _______                               ___________
>              \                             /           \
>               \                           /             \
>                \ :<- TJ/2 ->:<- TJ/2 ->: /               \
>                 \:          :          :/                 \
>                  \          :          /                   \
>                  /          :          \                   /
>                 /           :           \                 /
>                /           0 UI          \               /
>               /            (bit           \             /
>      ________/           boundary)         \___________/
>
> If the further constraint is added that 0 UI must be the mean of
> the jitter (i.e., zero crossing) distribution, then that FORCES a
> mask failure if the distribution is skewed, as it generally is in
> copper transmission media.
>
> The practice in Fibre Channel (for copper media) has been to freely
> adjust the X position of the waveform.  This permits a waveform with
> jitter approaching the spec limit of TJ to fit legally on the template.
>
> The proposal to constrain the mean of the distribution to be at 0 UI has
> been voted down several times in the Fibre Channel jitter working group.
> If I am missing something, I would appreciate being clued in.
>
> Regards,
> Mike Jenkins
>
> "Kesling, Dawson W" wrote:
> >
> > Attached is a summary of XAUI progress and issues for those of you who
could
> > not come to the Interim meeting last week. All changes and refinements
will
> > be included in draft D3.1 due to be posted to the 802.3ae web site by
6/11.
> >
> > If you have questions that aren't answered in the attached summary and
can't
> > wait until the draft is posted, please send them to the reflector
audience
> > in general so that other attendees can help me answer them. I probably
can't
> > respond quickly to ALL the questions that might otherwise be sent to me
> > personally!
> >
> > Thanks to all who participated so helpfully in St. Louis last week.
Special
> > thanks to Anthony Sanders of Infineon and Tom Lindsay of Stratos
Lightwave
> > for their extra effort on XAUI jitter leading up to (and after hours
during)
> > the Interim.
> >
> > -Dawson Kesling, Intel
> >  Editor, Clause 47
> >
> >  <<St_Louis_summary.ZIP>>
> >
>
  ------------------------------------------------------------------------
> >                            Name: St_Louis_summary.ZIP
> >    St_Louis_summary.ZIP    Type: application/zip
> >                        Encoding: Base64
>
> --
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>  Mike Jenkins               Phone: 408.433.7901            _____
>  LSI Logic Corp, ms/G715      Fax: 408.433.7495        LSI|LOGIC| (R)
>  1525 McCarthy Blvd.       mailto:Jenkins@xxxxxxxx        |     |
>  Milpitas, CA  95035         http://www.lsilogic.com      |_____|
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>