All good points... however, there is perhaps something to add here ...
It is not appropriate to consider planes as "virtual" structures with
no electrical distance associated with them. Also, it is very easy to
find high frequency noise potentials across planes in spite of their
low impedance due to surface transfer impedance and Ohm's law. This is
very easy to measure and I have done this quite often. It is also
true that "good decoupling practice" may mean a wide range of things
to a group of engineers. If I were to take your statement and go with
it to the best possible scenario, then I might be inclined to agree
with the "good practice" idea. However, (according to my definition
of "good power decoupling practice") this is usually not done very
often and often cannot be done for various reasons.
A question or two posed to address the "low impedance of the planes"...
What happens when you have a low impedance driver driving a low
impedance transmission line? How low is too low? Is there not energy
transmitted along the line? Also, where do the "resonant peaks" come
from which you have astutely observed? Can there be radiation from an
improperly terminated lossy transmission line? Coming to the correct
answers for these questions should bring one to at least an intuitive
understanding of what is going on here.
re: field measurement... radiation off the edge of a plane
If you are inclined to really do the measurements of these effects,
great care must be taken as there is frequently other fields
produced on the board which can mask the problem. This means that
some very clever instrumenting must be done to single it out.
re: impedance will be under an ohm
A question... what does the power plane impedance look like to the
residual energy at the plane's edge? Is it "under an ohm"? I think
not. At the board edge, there is a non-uniform field distribution
compared to the area found much further in. This is due to the
dielectric constant changing from, say 4 something (FR-4), to 1 (air)
and/or the loss of the capacitance from the conductors (i.e. they
end). This demands that the impedance be higher at that point than
the rest of the plane for the same reason that microstriplines with
comparable dimensions/materials to striplines are higher impedance.
What happens at such a boundary? The high frequency voltage will be
at a relative maxima at the plane edge compared to the rest of the
board where the field is more uniformly distributed in a "uniform"
dielectric... unless, of course, something is done about it.
re: reality
Again, this is a matter of degree... Just what is the noise immunity
of the system that you are trying to protect? If the noise immunity
is higher than the field induced energy that is obviously (by
Maxwell's equations) going to be there (and if the physical associations
are not appropriate to couple the energy to the subject circuit), then it
is a moot point. If not, it is a problem. I think that there should
be no question of whether there is a field or not. This is basic
stuff. There will be. The question is if it is large enough to be a
problem for your design. I specifically know of at least one case in
point in my experience where such a situation would have been
intolerable for proper operation.... and this was not a guess but reality.
At this point, I want to stress that I have built many systems with
and without using the 20H rule and this was due to various reasons
that "came with the design/requirements". My point here is that there
should be no question that if there is any potential difference at the
edge of a plane and any nearby conductor, then fringing E-fields will
form. Surely a conductor placed close by the time-varying fields
will, by Faraday's/Lenz's Law, have a current induced into it. If the
potential is large enough to overcome the losses (it frequently can be
in digital systems) and the electrical dimensions are right, then an
EM wave will be launched (and it doesn't have to be particularly
strong). Will any particular situation be a problem... always? No,
but potential for a problem does exist if the system is sensitive to
it or can radiate the energy at high enough levels to become a
compliance problem.
So it is a matter of responsibility to know whether the effects of the
field will be a problem for the particular design in question. It is
usually not be good thing to be adversely surprised by such a basic
phenomenon. If you need to do something about it, then the 20H rule
is a good place to start. If you don't know, perhaps it might be a good
idea to include where you can it until you can find out.
I hope that helps....
Regards,
Michael E. Vrbanac
>
> I agree with Larry- also, if your power planes have a high-frequency potential with respect to ground, then they will radiate; but I would want to see proof first that the power plane is radiating. With a goodly amount of
> interplane capacitance, and good decoupling practice to minimize the impedance between the power/ground planes, there should be no difference in how the edge of the board is treated. I've measured the impedance of one of our
> boards recently, and although it had a number of resonant peaks, the impedance was always well under an Ohm. The impedance of the plane will be significantly less than this, because of its low inductance; but I would want field
> measurements taken in a screen cage to prove that there is radiation off the edge of a good power plane design, and that there could be any improvement with this 20-H rule.
> --
> Regards,
> Gary L. Sanders, Staff Analog Eng., [email protected]
> L3 Communications, Inc. Celerity Systems www.csidaq.com
> Cupertino, CA USA dir. (408) 861-7325 fx (408) 873-1397
> Ultra Fast Acquisition & Data Generation Systems
>
> Larry Smith wrote:
>
> > I don't believe in the "20-H Rule". Suppose the power plane was at
> > 3.3V and the ground plane was at 0V. It would be easy to reconfigure
> > the system so that the "power" plane is at 0 volts and the"ground"
> > plane is at -3.3V. Does this mean that the power plane should now be
> > bigger than the ground plane?
> >
> > The only difference between the power and ground plane is that one is a
> > 0V and the other 3.3V WRT (...thats with respect to, lest I start
> > another discussion...) earth ground. But even this is not true in a
> > battery operated system. In any modern digital system, the impedance
> > between the power and ground plane is much less than 1 ohm well into
> > the EMI frequencies.
> >
> > The ground plane probably has a path out to frame ground and eventually
> > earth ground somewhere. But if that path is more than an inch long, it
> > is going to be well over 10 nH. Ten nH is 1 Ohm at 15 MHz (Z=jwL) and
> > higher impedance at higher frequencies. So, above 15 MHz, the voltage
> > between the power and ground planes is insignificant compared to the
> > voltage across the earth ground connection.
> >
> > The power and ground planes should be exactly the same size. To make
> > one larger than the other will simply have the effect of turning nice
> > diffential currents into common mode current and common mode
> > radiation.
> >
> > regards,
> > Larry Smith
> > Sun Microsystems
>
>
> **** 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/si-list ****