It depends on the margin you have w.r.t the channel/EMC spec.
If you are able to meet spec. even with one leg of the diff pair
disconnected, that channel is "well designed"! (whether the spec. is
well designed to stress test such configurations for compliance would be
another topic).
Same for return path vias (or lack of them).
Thanks,
Vinu
On 05/31/2017 12:52 AM, Carson Au wrote:
The perennial debate... on SI-List.
SERDES differential channels 10Gbps+. For layer transitions, is it OK to
use just one return path via (both sides of the differential via sharing
one GND return path via, symmetrical/balanced)? Is it OK to use no return
path vias at all (Lee Ritchey advocates for this)?
Is there literature comparing 2x, 1x and NO return path vias?
I can imagine that the common mode conversion would be an issue. But how
much of an issue?
Anecdotal stories that one method "just works" is not sufficient proof
(i.e. for the no return vias case) I have been able to get a 10Gbps+
channel working with only one side of the differential connected - with a
BER better than 10e-15 (ran overnight with no bit-errors)!! Yet none would
say this channel is well designed.
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List forum is accessible at:
http://tech.groups.yahoo.com/group/si-list
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
