Resending per Chris’ request
From: Geoff Thompson [mailto:thompson@xxxxxxxx]
Sent: Thursday, October 25, 2012 11:46 AM
To: STDS-802-3-RTPGE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_RTPGE] Channel definitions ad hoc survey
Chris-
What I heard on the channel definitions ad hoc call today plus what I have from history comes down to something like the following which I would like to put forth as a strawman for discussion as the connector requirement for RTPGE.
- There is no particular desire to reuse the RJ-45 as a twisted pair Ethernet connector in the automotive environment. The only possible exception to this might be "external" connections for use by end users for connecting consumer Ethernet equipment to the vehicle. That requirement (if it exists) probably has nothing to do with RTPGE.
- RTPGE provides a "moment in history" point for us to escape the tyranny of the installed base for twisted-pair Ethernet and the well-known problems of the RJ-45.
- Automotive manufacturers and suppliers have their own requirements (driven by a number of factors) for the physical configuration of connectors to be used for RTPGE.
- Automotive manufacturers and suppliers have requirements for RTPGE to pass through in-harness connectors that are very likely to have a significant number of non-ethernet pins carrying a wide variety of signals.
- The specs for the MDI attachment and in-line connectors should be performance based but not mechanically constrained.
- The above not withstanding, it would be useful to be able to standardize the circuit board interface for RTPGE. (I have problems with this concept reconciling edge connectors vs. headers, etc.)
- Power transmission over the same conductors and connectors was not discussed today. It is a real consideration and might come into play at multiple power levels.
- It would be very useful to have a fully specified (i.e. electrically and mechanically) test connector for RTGE. This would be used for establishing baseline performance and for interfacing test equipment. This connector would not have as stringent cost constraints as the connectors used in production volumes.
If all of the above is not correct then I hope it is at least useful as a starting point for discussion.
Best regards,
Geoff Thompson