Re: [SI-LIST] : 20-H Rule for Power Planes

Douglas McKean ([email protected])
Thu, 27 May 1999 14:42:20 -0700

At 03:31 PM 5/27/99 -0500, [email protected] wrote:
>
>
>I've read every post on this subject so far. I still only have a vague
>idea of what the 20H rule is and the assumed conditions under which it is
>applied. Any one want to try a one or two paragraph description?

How about a picture?

Edge of PCB
|
Sig Plane --------------+
|
|
|
|
Gnd Plane --------------+ -----
| Seperation
| Between
| Planes
| H
Pwr Plane ---- + -----

|<- 20H ->|

Several versions of this Rule of Thumb exist.
Here's two.

1.) ANY edge of the power plane is cutback
back away from the edge the corresponding
ground plane by 20 times the seperation
between the planes.

2.) The edge of the power plane is cutback
back away from the edge the corresponding
ground plane only around the perimeter
of the pcb by 20 times the seperation
between the planes.

Version #1 has the power plane cutback
everywhere on the poer plane. Strict
adherance to this version produces a
swiss cheese effect thus possibly
compromising the intended EMC effect.
And, in my opinion, the basis for counter
arguments as to its true effectiveness.
In some instances, well substantiated
counter arguments.

One other effect of Version #1 is to shield
any slot in the power plane by the ground
plane. Circulating currents about the
slot and associated radiation fields are
thereby hidden withing the ground planes.
And thus, counter arguments as to the
seperation distance between the planes being
extremely small (good argument by the way
for that type of mechanism) is insignificant
compared to the slot mechanism.

I have to emphasize here that I only
recommend the 20-H with a stackup that
has a power plane sandwiched between
two ground planes.

Version #2 has the power plane cutback
ONLY around the outer perimeter of the
power plane.

Regards, Doug McKean

**** To unsubscribe from si-list: send e-mail to [email protected]. 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 ****