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

[802.3_NGAUTO] AW: number of inline connectors



Title: Default Disclaimer Daimler AG

Hi Georg,

 

agree on your argumentation, that’s why I also have 4 inline connectors in my objective proposals.

From my perspective the 10G has even less effort with 4 inline connectors as it is intended to clearly be a shielded approach, where the connector is of less influence if he is designed appropriate.

 

Regards,

Stefan

 

Von: George Zimmerman [mailto:george@xxxxxxxxxxxxxxxxxxxx]
Gesendet: Freitag, 17. Februar 2017 18:22
An: STDS-802-3-NGAUTO@xxxxxxxxxxxxxxxxx
Betreff: [802.3_NGAUTO] number of inline connectors

 

I think we’ve come a long way building consensus and thinking through the issues in the ad hoc.  In reviewing notes to put together my ad hoc report, I noticed that we really hadn’t discussed and reached consensus on one item – how many connectors are we required to support in the channel? (for any speed)

 

We have had discussions of 2 connectors or 4 connectors. 

The proposal we have for 2.5Gb/s seems clearly for 4 connectors, but the other speeds still leave an open question –

 

Please respond to the reflector on your thoughts so we can try to get some consensus before Warren:

-          Does 10Gbps need to support 4 connectors to have broad market potential and technical feasibility?

-          Does 5Gbps need to support 4 connectors to have broad market potential and technical feasibility?

 

My feeling is that either one works for technical feasibility.  This is due to the fact that we have not nailed down the other characteristics of the link segment, we have flexibility in insertion loss of the wiring to compensate for extra connectors.  The return loss situation remains similar with 2 or 4 connectors, as in either case the connectors can be near the ends of the channel.  If I’m right about that, then the issue comes down to broad market potential – existing IEEE standards are written for 4 connectors, adding flexibility in cabling architecture.  There is almost surely greater market potential in supporting the same architecture already used.  Therefore, I would tend towards 4 connectors.

 

Is there objection to 4 connectors? (note, that for cost-optimized cabling solutions, you don’t HAVE to put in 4 connectors, it is just that our link segment models will need to support a 4 connector model).

 

 

George Zimmerman, Ph.D.

President & Principal

CME Consulting, Inc.

Experts in Advanced PHYsical Communications

george@xxxxxxxxxxxxxxxxxxxx

310-920-3860

 


If you are not the addressee, please inform us immediately that you have received this e-mail by mistake, and delete it. We thank you for your support.