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

Michael E. Vrbanac ([email protected])
Thu, 27 May 1999 18:43:58 -0700


re: my proprietary "excuses"... ouch!
Sorry you feel that way about that... it takes walking a fine line to contribute
constructively to a forum like this and not reveal too much which might be
a problem for the company I work for. To tell you the truth, I quite often pick
up stuff on this forum which tells me a lot more than the writer intends to tell......
not a good thing if we want to work together here and yet do our best for the
firms we work for. I am trying tell as much as I can... others may contribute
more details even from Cisco but then perhaps they don't feel as responsible
about it as I do and I suppose that's my problem. Honestly, I get paid to
know this stuff and I need to be careful. I hope that you can understand that.
Otherwise, I cannot contribute at all.

In the very best sense that this can be said, please remember it can also be
thought a "convenient answer" to ask for simulation data for certain "field
problems" in their "real setting" which are extremely difficult or impossible to
do on available tools just to make a point that a doubter's position is correct
in the absence of evidence acceptable to them. Unfortunately, too many folks
do this and perhaps to their own hurt. There should always be a constructive
skepticism in scientific inquiry but reports and evidence outside the realm of
computer simulation can at least be circumstantial evidence pointing
to the reality of a matter and perhaps in some cases, can be more real than
the simulation itself. (Haven't you found that once in a while some of the
simulations you've done do not "exactly" match reality? I have... especially
with high frequency phenomena. What does one do? One adjusts to reality...
not one's presuppositions.)

As a comment to all.... some final remarks... before I close my comments on
the 20H rule.

My posts were intended to explain that the fringing fields do exist and what
the 20H rule was intended to do about that and to explain that at times there
needs to be something done about these fields. I do not need to defend the
20H rule (which I think some thought I was doing). It'll be shown that it has
validity;in that I have no doubt.

Yes, I have encountered the need to apply the 20H rule (or some form of it) on
a number of occasions and it has been very carefully verified through certain
strict methods (and many were performed on digital systems, some with
performance levels much lower than today). I and many others have
performed structured tests to validate specific solutions and have implemented
them (like the 20H rule) with good success. I have performed these
experiments and demonstrated them in the presence of "staunch unbelievers"
who are some the most intelligent people around and they left as "believers".
Its too bad that some have not had this experience so that they can be
convinced as well. I would encourage you all to prove this to yourselves
however it might be helpful to you.


Michael E. Vrbanac

At 06:25 PM 5/27/99 -0500, you wrote:

>Apparently this is proprietary data only (a la Cisco). When pressured
>to show results it's a convenient answer to state proofs are proprietary.
>I too would love to see simulation results. I have talked to several
>respected folks in the industry who did not see a correlation to the
>20H rule (when as Vrbanac says "just doing digital"). Of course, no one
>was addressing microwave antenna design. Just hoping that their digital
>didn't become an effective radiator. Right?
>Has anyone designed boards with and without the 20H rule and tested them
>each other? Or are we happy with several of the "proofs" that are out
>there? I'm
>all for an experiment....
>Bill Dempsey
>DNA Enterprises, Inc.
>**** 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 ****

**** 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 ****