RE: [SI-LIST] : diff pair routing rules

About this list Date view Thread view Subject view Author view

From: Doug Hopperstad (doug.hopperstad@qlogic.com)
Date: Tue Jan 23 2001 - 11:35:31 PST


Javin,
The +/- differential pair spacing should be kept to a minimum for
common-mode coupling. The actual spacing will depend on your board stackup
and dielectric material. For a reference, a dual-stripline using FR408 would
yield the following:

---------------- Ground plane
--(1+)-- --(1-)-- Trace layer: differential pair
--(2+)-- --(2-)-- Trace layer: Orthogonal routing to the upper trace
layer.
---------------- Ground plane

The dielectric spacing between the plane and trace layer = 4 mils
The trace-to-trace layer spacing = 14 mils

The trace width = 5 mils (1w)
The trace-to-trace spacing = (1+) to (1-) = 10 mils edge-to-edge

The above configuration will yield a 99.54 ohm differential impedance.

With regard to the second question: I try to keep the trace lengths matched
as close as possible. At the pins, the pinout should be setup such that the
+/- pins are next to each other to avoid a routing mismatches.

Doug Hopperstad
Qlogic

-----Original Message-----
From: Javin Olson [mailto:Javin.Olson@plexus.com]
Sent: Tuesday, January 23, 2001 10:15 AM
To: si-list@silab.eng.sun.com
Subject: [SI-LIST] : diff pair routing rules

All,

I have a couple of general SI/routing rules I would like to get
information on,
and thought this may be the forum to do so.

1. What is the general rule of trace to trace spacing of a differential
pair signals (positive trace to negative trace, NOT pair to pair
spacing)?
PECL and LVDS, are different spacings recommended?

2. When routing a differential pair to pins that are a few rows apart,
is it best to keep the diff. pairs close as long as possible and then
route straight to the pin; or should the net that is shorter be
(mini) wrapped so that the diff pair lengths are exactly matched? What
is
more important, keeping the diff pair nets routed togother, or breaking
off the short route near the end of the trace to wrap it and make it as
long as the long route net?

Any thoughts?

Thanks you for your comments.

Javin Olson
Plexus Technology Group
Digital Design Engineer

**** To unsubscribe from si-list or si-list-digest: send e-mail to
majordomo@silab.eng.sun.com. 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
****

**** To unsubscribe from si-list or si-list-digest: send e-mail to
majordomo@silab.eng.sun.com. 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
****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Tue May 08 2001 - 14:30:43 PDT