[802.3ae_Serial] From serial PMD call 23 Apr: still would like more experimental evidence
We still hope to progress the standard next week. We will hold another
phone conference on Friday (see bottom). Anyone intending to submit a
comment that relates to Clause 52 (especially something new) is asked to
inform Dave before then so that the group has a chance to discuss it on this
call.
I think the concerns about the transversal filter implementation have
subsided.
The issue list is now:
1. Lack of experimental reports on interoperability, TDP and stressed
sensitivity;
2. Relation between VECP and transmitter penalty of stressed eye
generator is not well known, could over-burden the receiver especially for
BASE-S, BASE-E where degree of stress (VECP) is higher;
3. Tolerances of Bessel Thomson responses allow error;
4. 20/100ppm clock tolerances seen as inconsistent;
5. #99101 Propose reference ITU-T G.783
6. Detail of eye mask timing alignment ("at the average power") makes
for a less reproducible measurement;
7. Is 10GBASE-S transmitter power window very narrow?
8. OMA vs. average power
9. A couple of editorial comments
Apologies to anyone whose concerns I have forgotten!
Going through the issue list:
1. It looks like we may have to assume that interoperability and TDP
are OK, on the grounds that several organisations claim they have parts and
in the main they aren't complaining.
1/2. On stressed sensitivity (stressed eye generation, VECP...), an
update from people who have tried to do this test for BASE-S, BASE-E would
be very helpful for the standard's progress. I have been trying to simulate
various stressed eye generators. In summary, the simulations predict that
the allowed variability in pulse shrinkage can cause pass/fail variations
for receivers. Can we reduce the variability yet still allow the
flexibility for implementers to actually make stressed eye generators,
bearing in mind that real hardware is not the same as simulation?
3. No change. Not a fatal problem.
4. No change. Latest is Juergen's email
http://www.ieee802.org/3/10G_study/public/serial_adhoc/email/msg00616.html .
5. No discussion: see above.
6. As I understand it, measuring the average timing of the edges away
from the crossing level (waist) introduces a new form of error, because the
scope will sample a random proportion of rising edges vs. falling edges,
which then occur at different times. For typical sample sizes, this creates
a random timing error which largely negates any benefit of moving the
expected timing to the desired place. In the example I looked at we were
talking 1 ps.
The mask dimensions are not chosen to 1ps precision.
Mask measurements are disappointingly inaccurate already. This would make
it worse.
The mask is not the primary measure of transmitter quality; TDP is.
There is an industry standard way of mask alignment already. It adds cost
and confusion to all users, on an ongoing basis, to create another way of
doing it.
7. BASE-S maximum mean power is set by safety limits. Say it is -1.1
dBm (this may be debatable). Take minimum OMA at -4.2 dBm. For e.g. 6 dB
extinction ratio (2.2 dB penalty), minimum mean power is -4.2-3+2.2=-5,
giving a window ~3.9 dB wide. This is tight, but the key question is: have
we understood the safety limit right?
8. No change since changes in D4.3
9. The RIN test, BASE-S variant, deletes the wrong step in the recipe.
Some conditionally mandatory PICS entries are missing their N/A
check boxes.
Present
-------
Piers Dawe Agilent SPG
Petar Pepeljugoski IBM
Greg LeCheminant, Stretch Camnitz Agilent T&M
Tom Lindsay Stratos
Peter Öhlén Optillion
David Kabal Picolight
Next call
---------
Friday 26 April, at the usual time and coordinates:
4:15 pm BST = 17:15 CET = 11:15 am EDT = 8:15 am PDT
+1(816)650-0631 Access code 39209
Piers