RE: [SI-LIST] : differential pair routing

Heyfitch, Vadim (vadim.heyfitch@intel.com)
Fri, 13 Aug 1999 16:43:03 -0700

I would vote with Jim. My argument for maintaining equal path length
over keeping constant impedance profile goes like this. Differential pairs
on a typical PCB stackup are only 30 to 40% differential (i.e. only 30 to
40% of the return current flows through the complementary trace of the
pair). Thus, where the traces in the pair are slightly diverged
(=uncoupled), there this fraction of the return current going through the
other trace drops down to some 5-10% for a short while until the traces come
close together again. In cases when this differential pair is driven by a
couple of single-ended drivers 180 degrees out of phase (as opposed to a
truly differential driver) this impedance bump due to the routing should be
even less of an issue.

Vadim

-----Original Message-----
From: Knighten, Jim L [mailto:JK100005@exchange.SanDiegoCA.NCR.COM]
Sent: Friday, August 13, 1999 2:27 PM
To: si-list@silab.eng.sun.com
Subject: RE: [SI-LIST] : differential pair routing

Mario,

While "bumps" in traces do present an impedance change, differences in path
lengths have a direct bearing on common-mode current generation leading to
radiated EMI. Delay skew increases common-mode current generation linearly
(for small delay skew values).

So, I would vote to maintain equal path lengths as closely as possible, even
at the expense of the impedance bumps.

Data on the bad EMI effects of delay skew was presented at last week's IEEE
EMC Symposium in Seattle.

Jim Knighten
________________________________________________________
Dr. Jim Knighten e-mail: Jim.Knighten@SanDiegoCA.NCR.com
<mailto:Jim.Knighten@SanDiego.NCR.com>
Senior Consulting Engineer
NCR
17095 Via del Campo
San Diego, CA 92127 http://www.ncr.com <http://www.ncr.com>
Tel: 858-485-2537
Fax: 858-485-3788

***** Notice the Area Code change from 619 *****

-----Original Message-----
From: Mario Appiani [mailto:mappiani@attotech.com]
Sent: Tuesday, August 10, 1999 12:18 PM
To: si-list@silab.eng.sun.com
Subject: [SI-LIST] : differential pair routing

From: Mario Appiani@ATTO TECHNOLOGY on 08/10/99 03:18 PM

To: si-list@silab.eng.sun.com
cc:
Subject: differential pair routing

Hello all,

I have a question about routing differential pairs. Is it
more important
to maintain trace lengths between the pairs (one spec says
to keep trace
lengths within .100") or to maintain the trace spacing
throughout the trace
run? I 've noticed on several boards that in order to
maintain trace
lengths between pairs, many designers have had to add loops
or bumps in one
of the pairs' lines. When I see this, I wonder about how the
impedance
change, due to the increased trace pair separation, has on
the signal.

_____
_____/ \________________ + leg

___________________________ - leg

Do I need to worry more about jitter or impedance
mismatches? If loops are
ok, does it matter where they are placed (i.e. in the middle
of the run or
at the end) ?

Thanks in advance,

Mario Appiani ext. 130
Sr. Digital Systems Engineer
ATTO Technology, Inc.
40 Hazelwood Drive
Suite 106
Amherst, NY 14228
Tel. 716.691.1999
Fax. 716.691.3724
Email. mappiani@attotech.com

**** To unsubscribe from si-list: send e-mail to
majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
si-list, for more help, put HELP. si-list archives are accessible at
http://www.qsl.net/wb6tpu/si-list ****

**** To unsubscribe from si-list: send e-mail to
majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
si-list, for more help, put HELP. si-list archives are accessible at
http://www.qsl.net/wb6tpu/si-list ****

**** To unsubscribe from si-list: send e-mail to majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****