Re: [SI-LIST] : 20H Revisited

About this list Date view Thread view Subject view Author view

From: Michael Vrbanac ([email protected])
Date: Fri Jan 14 2000 - 10:31:36 PST


Sainaith,

Ah, truly said like a "modern man"!!!! <grin> Contradiction? Probably not.
But I think we might be missing something. I wonder if we aren't confusing
inter-relationship with unity. The explanation is somewhat complicated
but I'll take a shot at it.

That unity you speak of underneath the diversity is appreciated. I (and at
least one other person I know) have created a team that did both EMC and
SI at the same time. That means, though, I (we) also know the differences
and what it takes to competently move over and "operate" in the other
realm in spite of that diversity.

Hence, this reveals to me reason why I think there has been a distinct
disagreement among folks on the 20H rule. It wasn't even originally
created to deal with an SI issue but some folks thought so because the
20H rule has been used for other things since (perhaps rightly or wrongly)
and got labeled a "screwy SI rule". So I think it very poignantly shows
us that this diversity is very real and that there are things in electromagnetic
theory outside the immediate scope of SI which are problematic. If we
take a lesson from that, we should not only be careful in our reactions to
things we think are outrageous but also allow for those who disagree so
that they can do so without fear of recrimination. We want to stimulate
creative thought, not squelch it! (at least I hope we do! <grin>)

Actually, much of this diversity can be bridged at an individual level.
(I don't think we are quite there to do it more globally.) Mostly
it is a matter of training, a bit more to broaden the scope for the strictly
SI person moving to EMC. For the strictly EMC person (strong design/
theoretical emphasis, not regulatory only), it is similar but different. The EMC
person focuses in on a distinct subset of already acquired skills and needs
training on the "simulator du jour". After that, for both, its altering one's
viewpoint, focus, and goals. Not necessarily hard for either but it must be
done. So again, they inter-related but not the "same" thing.

Knowing that, I would have to draw the Venn diagram differently i.e.
(without graphics. of course)

SI is a subset of EMC is a subset of electromagnetic theory.

or said another way

SI is a specialty of EMC which is a specialty using electromagnetic theory

And that's not a bad thing... I'd hate to think if we weren't focusing on the
SI issues and nothing ever worked to begin with. And the same goes
for the EMC folks who are looking at the larger "system issues" and
are interested in many pertinent SI issues to do their job so that the system
level issues are resolved. But again, they are not the "same" thing...at
least not now.

I am looking for the day when both disciplines work as one in a more global
sense but that day is not here. I figure it won't happen until at least two things
allow us to do that: when the SPICE and IBIS (et al) model issues become
resolved and common "corrected" libraries are widely available AND the
simulation tools relevent to the two disciplines continue to improve and begin to
merge... to allow folks to begin to focus energy elsewhere ... especially to bring
the two disciplines together. So until then, to me, the Venn diagram is different
and therefore the "coin" has two joined faces, each a different size. SI is a
very important subset of EMC. Maybe someday soon, hopefully, I can
change my model.... to unity.

Michael E. Vrbanac

Sainath Nimmagadda wrote:

> Michael, I don't see any contradiction. Consider the underlying theory as the coin and the disciplines as the sides. Of course, the two sides appear different for
> normal coins. I believe system/signal integrity gets better as we begin to appreciate the unity in the diversity of various disciplines. On the other side of the
> coin, we need diversity in disciplines for technology growth and, as you mentioned, steady paychecks. Oops! did I spin the coin again?
>
> Sainath
>
> Michael Vrbanac wrote:
>
> > Sainaith,
> >
> > Thanks! Perhaps I should have originally said... "it was a 'non-SI' issue" to
> > be more technically correct and be better understood.
> >
> > I hate to be a bit contrarian here but I don't think they are" two sides of the
> > same coin". The underlying theory is the same but the "disciplines" aren't.
> > I think there might be a bunch of folks who haven't yet figured that out. I'll
> > bet you know of some brilliant SI folks who would not enjoy working with
> > EMC for instance. They'd be out of their league. I believe there is a reason
> > for this.
> >
> > Electromagnetics and field theory are the general labels for all the underlying
> > physics for what we do. EMC would be a bit more specific to better define
> > the specific use of electromagnetics and field theory to make an electrical
> > device not only compatible with its environment but also with itself so that
> > proper operation can be insured for the subject machine and all those in its
> > environment. SI is even more defined and a subset of that. It is the use of
> > electromagnetics and field theory to insure that the subject machine can
> > properly operate and deal with its own internal issues. It is usually involving
> > basic functional and reliability issues and essentially works with signal levels,
> > thresholds, and purity as a specialty. Maybe that's not the best definitions
> > but at least its a first cut at it. (BTW, there's always room for some exceptions.)
> >
> > When one considers the "language" of SI and the "language" of EMC and also
> > the views, focus, and goals along with that, we find that they are different things.
> > Even the simulation tools for the two are different. I have managed a group that
> > did both EMC and SI so I know. People can bridge that distinction as I have
> > but make no mistake. They aren't the same thing. Related in some ways but
> > not the same.
> >
> > Michael E. Vrbanac
> >
> > Sainath Nimmagadda wrote:
> >
> > > Thank you Ray and Michael, for quickly picking up my comment and clarifying. For people who do not know that fields and SI are two sides of the same coin, I was
> > > afraid Michael's 'side note' might mislead them that they are not related subjects. Hence the comment.
> > >
> > > Sainath
> > >
> > > Michael Vrbanac wrote:
> > >
> > > > Sainath,
> > > >
> > > > Everything to do with SI is ultimately a field problem. (T-line work & crosstalk, are
> > > > easy examples.) It isn't always convenient to handle SI that way, though. Let me
> > > > explain.
> > > >
> > > > Most engineers aren't specialists in field issues although they may know something
> > > > about them. Since most of the issues SI folks deal with appear in the time domain,
> > > > many times it is easier to stay in that domain (i.e. keep the problem as simple as
> > > > possible and still get the job done!) Also, SI folks need to speak the "language"
> > > > of the design engineer who is working in the time domain.
> > > >
> > > > The field simulators that we use for SI are designed to relieve us from having to
> > > > directly deal with field theory and make our jobs much easier (thank goodness!).
> > > > The "bad news" is that we might begin to lose touch with our field theory skills
> > > > because the simulator is supposed to do that for us (and that can create even
> > > > larger issues.) The "good news" is, in the near term, we become more effective
> > > > in producing meaningful results which, of course, results in products entering the
> > > > market which then translates into a steady paycheck for the SI engineer!
> > > >
> > > > Michael E. Vrbanac
> > > >
> > > > Sainath Nimmagadda wrote:
> > > >
> > > > > Interesting! What SI issues have to do with fields anyway.
> > > > >
> > > > > Sainath
> > > > >
> > > > > Michael Vrbanac wrote:
> > > > >
> > > > > > D.C.
> > > > > >
> > > > > > re: change in orientation of the fringing fields
> > > > > > Yes. That's a very good point.
> > > > > >
> > > > > > re: Vss plane being cut back instead
> > > > > > In the original application of the 20H rule, it probably wouldn't have mattered.
> > > > > > In later applications (and alterations) other than the original, it might have mattered.
> > > > > >
> > > > > > As as side note:
> > > > > > The original 20H rule came out of a "field" problem being solved. From
> > > > > > that standpoint, it might have actually been true to some extent to have
> > > > > > called it a "screwy SI rule" because it really wasn't an SI problem to begin
> > > > > > with. Knowledgeable folks dealing with SI issues would never have
> > > > > > classified the original use as an SI issue. Subsequent uses of the 20H rule
> > > > > > may indeed have blurred the lines a bit.
> > > > > >
> > > > > > Michael E. Vrbanac
> > > > > >
> > > > > > "D. C. Sessions" wrote:
> > > > > >
> > > > > > > Keeping in mind that There Is No Ground, what happens if
> > > > > > > the Vss plane is the one cut back?
> > > > > > >
> > > > > > > As an alternate consideration, note that the stepback would
> > > > > > > change the orientation of the fringing fields, which would
> > > > > > > interact with the enclosure to potentially limit net radiation.
> > > > > > >
> > > > > > > --
> > > > > > > D. C. Sessions
> > > > > > > [email protected]
> > > > > > >
> > > > > > > **** 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
> > > > > > > ****
> > > > > >
> > > > > > **** 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
> > > > > > ****
> > > > >
> > > > > **** 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
> > > > > ****
> > > >
> > > > **** 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
> > > > ****
> > >
> > > **** 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
> > > ****
> >
> > **** 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
> > ****
>
> **** 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
> ****

**** 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
****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Thu Apr 20 2000 - 11:34:38 PDT