RE: [SI-LIST] : 20H rule: a theory?

About this list Date view Thread view Subject view Author view

From: Jeremy Plunkett (jeremy@serverworks.com)
Date: Wed Jan 12 2000 - 12:24:15 PST


One aspect of the 20H rule which I never saw mentioned in the discussion
earlier was whether it applies to power planes on the outside surface
vs.
those sandwiched between ground planes in a many-layered board. I am
not an
EMI expert and I don't have the tools to do any measurements or
analysis.
But, by analogy to the rate at which coupling between microstrip traces
drops off with distance, it seems reasonable that for a power plane
sandwiched between ground planes, moving the edge of the power plane in
from
the edge of the ground planes should result in a serious drop in the
fields
that reach the edge of the board and propagate out as EMI.

For a power plane which is the last solid plane under the surface of the
board(which would always occur in boards with only two planes) it is a
different story, since in that case pulling the edge back will just
shift
the direction of the fields and change the impedance at the edge of the
plane a little bit, and if the new impedance happened to be higher(a
better
match to the impedance of free space) you could get more EMI leaving the
board instead of less.

Based on my current understanding I plan to continue using the 20H rule,
but
if my understanding is flawed I would like to know about it...

Any comments?

Jeremy Plunkett
Signal Integrity Engineer
Serverworks
(formerly Reliance Computer Corp.)

jeremy@serverworks.com

-----Original Message-----
From: owner-si-list@silab.eng.sun.com
[mailto:owner-si-list@silab.eng.sun.com]On Behalf Of Abd ul-Rahman Lomax
Sent: Wednesday, February 02, 2000 1:28 PM
To: si-list@silab.eng.sun.com
Subject: [SI-LIST] : 20H rule: a theory?

A little while back there was a discussion on this list of the so-called
20H rule, which suggests backing off a power plane from the edges of its
co-extensive ground plane by 20 times the distance between the planes,
in
order to reduce EMI.

There seemed to be some scepticism that this was actually effective, but
one writer noted that this would change the radiation pattern, which
brings
me to my suspicion as to why this rule might be effective in some
situations.

I would expect that co-extensive planes would show peak radiation in the
plane of the planes; i.e., out the edges. Where one is concerned about
coupling between one section of a PCB and another, which is often the
case,
setting one of the planes back might move the radiation peak out more
toward normal to the planes, and thus result in less coupling to a
sensitive on-board area.

Thus the rule might work to reduce interference between one section of
the
board and another, even if overall emission was not reduced, or even if,
possibly, it even increased a little.

Comments? Thanks....

marjan@vom.com
Abdulrahman Lomax
P.O. Box 690
El Verano, CA 95433

**** 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
****

**** 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
****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Thu Apr 20 2000 - 11:34:53 PDT