Hi Peter,
my apologies if I dind’t addressed completely your comment yesterday due to lacks in my headsets.
Below I’m articulate better my statements, since I believe some of my comments didn’t gone through at all.
Still showing a characterization of a SiP transmitter across the worst case chromatic dispersion limits for 10 Kms application that makes feasible it, I want to re-state
that this not represent the best of the SiP transmitter quality in terms of SECQ, neither represent the absolute best optimization one can do (as stated in the presentation).
- I cannot share any data as you can imagine, but the SECQ experimental data shown
yesterday are quite close to the average data we’re getting from EML suppliers – from here I strongly encourage people presenting propagation data about, this would be very helpful for future discussions.
Also said I’m against the introduction of a further gate for chromatic dispersion penalty (TDECQ-SECQ) over PAM4 interfaces, for below reasons:
- It would not address a significant problem or lack in the standard we didn’t noticed or addressed before.
- Absolute TDECQ is used to build the overall budget: together with TDECQ-10*Log(Ceq) helps to protect about interoperability
over short and long reaches.
- I do not expect it will be a good representation of real propagation conditions.
- PAM4 real receivers implement >1% threshold optimization, while TDECQ reference receiver is limited to 1% (we
know all effort done to get to 1dB to 1dB correlation between TDECQ and RX sensitivity - this allowed to generate such linear equations and charts into IEEE drafts).
- TDECQ-SECQ penalty would more easily diverge from the real receiver under propagation condition. This because
I expect SECQ (TDECQ at km=0) would have a better fit with real receivers, while TDECQ at maximum dispersion would probably need more than 1% threshold adjustment. Thus TDECQ-SECQ can be a loose metric.
- It would probably penalize ‘virtuous’ transmitters with low SECQ, which can still meet the absolute TDECQ at
the worst case fiber propagation conditions, but fail this new parameter.
- To me the added Noise/Distorsion ‘equivalent’ sigma generated by chromatic dispersion will affect more transmitters
having the merit to have a lower Noise/Distortion sigma (corresponding to lower SECQ), while probably having less impact on trasmitters starting with higher SECQ.
- If not carefully defined, what can then happen then is the absurde situation that after propagation tests a ‘good’
transmitter ends with lower TDECQ with respect a worse one, but just fails the new TDECQ-SECQ metric ...
- The introduction of any TDECQ-SECQ still requires considerable effort and contruibutions (especially from people
developing ‘good’ EMLs TX) to be applied ‘at least’ to the existing PAM4 interfaces which can suffer more propagation DGD/penalties (like 400GBASE-ER8 and 400GBASE-LR4).
- In manufacturing it will require a controlled environment in which, not only the TDECQ at the worst chromatic
dispersion will have to be tested (and one can choose it own corner case test, once defined it in case), but ALSO the transmitter transmitter SECQ, which is actually a parameter not included into the standard.
Thanks and ciao
Marco
From: Mark Nowell (mnowell) <00000b59be7040a9-dmarc-request@xxxxxxxx>
Sent: martes, 20 de agosto de 2019 22:34
To: STDS-802-3-100G-OPTX@xxxxxxxxxxxxxxxxx
Subject: [802.3_100G-OPTX] Proposed Agenda for 802.3cu ad hoc - Wed 8/21
Dear Colleagues,
The proposed agenda for the IEEE P802.3cu Task Force ad hoc on Wed Aug 21st is as follows. Due to travel changes I will be unable to attend and chair, but Kenneth Jackson has kindly
offered to chair the meeting in my absence.
- Approve agenda
- Approve previous ad hoc minutes
- Patent reminder
- Participant reminder
- P802.3cu Task Force Ad hoc:
- “Task Force Update” – Ken Jackson (5 mins)
- “802.3cu: 400GBASE LR4 fiber propagation penalty test completion - positive dispersion tests” – Marco Mazzini (15 mins)
- “400GBASE-LR4 Engineered Link Specification Approach Proposal” – Chris Cole (25 mins)
Call information is posted at:
http://www.ieee802.org/3/cu/public/cu_adhoc/
*Slides to be posted in Ad Hoc area as soon as possible.
http://www.ieee802.org/3/cu/public/cu_adhoc/cu_archive/index.html
Regards, Mark
To unsubscribe from the STDS-802-3-100G-OPTX list, click the following link:
https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-100G-OPTX&A=1
To unsubscribe from the STDS-802-3-100G-OPTX list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-100G-OPTX&A=1