From: Boris Yost (firstname.lastname@example.org)
Date: Fri Nov 10 2000 - 06:45:58 PST
[This was sent once yesterday, but seems to have been routed to the trash.
If you received it twice my apologies--Boris]
1) Yeah, you need to specify the differential impedance between 2 traces.
2) No, the issues are how close is the ground plane and how close is the
aggressor. If the ground plane is far away and the aggressor is close, one
line will get more noise than the other. But you're not supposed to do
3) Thicknesses and dimensions scale with each other. 5 mils to the plane
means very thin traces. Giant dielectrics means aggressors must be kept
very far away.
Since you got me to wonder about it I ran it through Hyperlynx. See
attached. Orange trace (LVDS minus) is closer than the red trace to a
ultrafast CMOS aggressor source-terminated with 56 ohms. Because they don't
track in parallel, you have common-mode.
[mailto:email@example.com]On Behalf Of Gaboian, Jerry
Sent: Thursday, November 09, 2000 9:32 AM
Subject: [SI-LIST] : Routing Differential Pairs as 100 differentially Vs
individually 50 ohm lines
I am looking for some feedback on the topic routing LVDS differential pairs
as 100 ohms
with respect to each other Vs. 2 separate 50 ohm lines.
At the PCB East conference several keynote speakers referenced the following
"No need to specify differential impedance between two lines. Routing traces
side by side
does not guarantee noise immunity from common mode noise. On the contrary,
noise from an
adjacent line will couple as a differential signal"
The length of the pairs should be matched and kept close to a reference
plane (about 5 mils)is the must important criteria.
**** To unsubscribe from si-list or si-list-digest: send e-mail to
firstname.lastname@example.org. In the BODY of message put: UNSUBSCRIBE
si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu
This archive was generated by hypermail 2b29 : Tue May 08 2001 - 14:30:03 PDT