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

RE: Optical Connectors

As an open question, what is the impact of having more than one optical connector specified in MDI of the 802.3 standard? Is this an interoperability question or can it be left flexible without adversely affecting the standard. Is this similar to the Power over DTE decision to define two different wiresets that could be used to "transmit power", and every receiver had to accept both, or is this a bogus analogy?
Throwing my personal opinion into this:
I can see near-term implementations in SC, moving to LC in the next year, so I would prefer to have a standard that specified only LC, and I would not be adverse, if it were possible, to include both connectors as "possible MDIs".
David Kabal
Photonics Engineer, OPTera Metro Solutions, Nortel Networks
Phone: 613.270.5953  Fax: 613.591.2035
e-mail: dkabal@xxxxxxxxxxxxxxxxxx
-----Original Message-----
From: Jonathan Thatcher [mailto:Jonathan.Thatcher@xxxxxxxxxxxxxxxxxxxx]
Sent: Tuesday, July 18, 2000 11:49 AM
To: HSSG_reflector (E-mail)
Subject: Optical Connectors

I have opened this thread to continue the discussion on optical connectors. So far (what has come into my reader), we have the following comments:
"Bill Wiedemann: Regarding 850CWDM we are planning to make first implementations with duplex SC moving to LC with small form factors. Our expectation is that small form factor with LC could be available a year from today. "
"Jim Tatum: I would assume that 802.3ae would do the same as 802.3z, and NOT specify conectors. "
"Ed Chang: There are so many different form factors, and connectors, which even the GbE and Fibre Channel market can not get consensus."
If we review the 802.3 Ethernet specification, we see that we have identified connectors for each variant (I don't remember an exception). For example:
7.6.2 AUI Configuration cable Optical for repeaters
38.11.3 MDI = Duplex SC for GigE Optics
39.5.1 MDI = Style 1 (DB9) and Style 2 for GigE Cu
While I remember no rules that require us to do so, it seems obvious that there exists a precedent which should guide our decision.
In 802.3z, we specifically took a vote to avoid connector discussions ("connector wars")**. We could do the same in 802.3ae. If we did, I would argue that we would, effectively, be retaining the duplex SC optical connector specified in clause 38.
My PERSONAL preference would be to specify the LC connector. Rationale:
1. There seems to be an overall inclination to move in that direction.
2. It sets the stage for some kind of "Small Form Factor" 10 Gig transceiver.
3. I don't think that it would negatively impact the cost of the transceiver in the 2002 (standard completion time frame).
As CHAIR, I don't want to use up any cycles on this. If there isn't sufficient consensus to agree on an alternative to the SC, we should just adopt the SC and move on.
** In reality, this was bumped up to 802.3 because neither I (sub-chair for PMD) nor Howard (802.3z chair) wanted to use precious committee time for the discussion.

Jonathan Thatcher,
Chair, IEEE 802.3ae (10 Gigabit Ethernet)
Principal Engineer, World Wide Packets
PO BOX 141719, Suite B; 12720 E. Nora, Spokane, WA 99214
509-242-9000 X228; Fax 509-242-9001; jonathan@xxxxxxxxxxxxxxxxxxxx