Re: [SI-LIST] : Trace Impedance Selection

About this list Date view Thread view Subject view Author view

From: Neven Pischl (npischl@cisco.com)
Date: Mon May 08 2000 - 16:52:22 PDT


Vinu,

free space can be modeled as a transmission line only to take advantage of
the mathematical analogy which makes possible to describe E and H field
propagation using the same expression as for V and I in the transmission
line. And in cases such as plane-wave propagation through free space, it is
only ONE t-line that can mathematically describe it, not a network of
multiple t-lines.

But it is not possible to mix/combine these two, i.e. connecting a "current
and voltage" carrying t-line such as a stripline in parallel to such an
imaginary t-line such as free-space. It is same as the case of sound
propagation through a medium, where again the propagation can be described
using the same expressions, but you can't add such a "sound-carrying
transmission line" in parallel to a stripline.

I think that much of the confusion (my perception) comes from the original
question, which was which characteristic impedance would work best for EMI.
The characteristic impedance of a trace is a result of the geometry, and the
radiation is also a function of the geometry. But the radiation is not
directly related to the characteristic impedance, which Mary pointed. In
addition, I guess that the far-field radiation of a single trace was meant
in the original question. In that case, it does not matter whether the
source is more of "low-impedance" meaning "mostly" current in the traces, or
high-impedance, meaning mostly voltage. It is only the radiated power that
matters, and the same far-field levels can be obtained with either current
or voltage drives. What I mean is that you can have the same far field
levels originating from a loop or from a dipole or monopole (even if they
are electrically short).

That is different from crosstalk on the PCBs, when in the most practical
cases due to proximity and parallelism of the traces, we have mostly
inductive coupling caused by the currents. If we look at the indirect
radiation caused by coupling on the board first and then radiating, perhaps
by some more efficient radiator (e.g. longer trace), than the current-caused
inductive coupling may play a significant role, and it is the current that
matters most, not the voltage, and as well-known, reducing the current will
reduce EMI. It is not the same as saying that increasing the characteristic
impedance will reduce EMI. You may increase the impedance by rising a trace
higher, which in turn may cause more EMI if you happened to build a more
efficient antenna. As long as the size of such a trace is electrically
small, it is guaranteed that you will increase radiation by doing so, by
both direct radiation and indirect radiation due to coupling. I believe that
at higher frequencies, above the first resonance, anything can happen, and
you may actually also decrease EMI (which by no means should be hoped for
:).

What really matters in reducing EMI is to build very inefficient antennas,
by keeping them (the traces) close to the reference planes - thus reducing
the loop area and crosstalk as well, by assuring that the return currents
flow adjacent to the traces, by keeping them short .... For a given drive
voltage, the power available to drive the antenna (trace) and cause EMI and
crosstalk should be minimized, which leads to reducing the current. Any
answer to the question about which characteristic impedance is better for
EMI must be viewed from such a larger angle, because otherwise it can not be
answered.

Sorry for taking up much space, I actually only wanted to answer to your
question but I got carried away :)

Neven

----- Original Message -----
From: Vinu Arumugham <vinu@cisco.com>
To: <si-list@silab.eng.sun.com>
Sent: Monday, May 08, 2000 10:24 AM
Subject: Re: [SI-LIST] : Trace Impedance Selection

> Can free space be modeled as a network of transmission lines, arranged in
such
> a manner that the impedance seen looking into any two points in space, is
377
> ohm?
>
> Thanks,
> Vinu
>
>
> Neven Pischl wrote:
>
> > Here's my $0.02.
> >
> > The scenario doesn't work because it roughly assumes that if you have a
377
> > Ohm source, and a 377-Ohm resistor somewhere in free space, that you
will
> > have max power transfer (even without any other connection). That's what
you
> > are saying when you say that half power would go through the space (at
least
> > it seams to me so).
> >
> > It mixes the wave-guiding concept in which there is at least
two-conductor
> > structure that guides EM-waves, and it is characterized by its
> > characteristic impedance, with the free-space EM-wave propagation in
which
> > the intrinsic impedance of the medium (not characteristic impedance of a
> > waveguide) is 377 ohm. There is an analogy between these two ways of
> > propagation, in terms of mathematical description, but the terms do not
have
> > the same meaning.
> >
> > Same as a 377 ohm line will not radiate half power to the space, when
> > immersed into the EM field, it will not couple half power of the field
> > (which should happen if the concept is right).
> >
> > It can be seen that the concept does not work also if you examine a
50-Ohm
> > air- microstrip. If you assume the concept of splitting power between
the
> > line and the air, 50/377 of the total power which is about 0.13 (or 13%)
> > would always be lost to the space, even in a perfectly matched 50-Ohm
> > system. We know that it does not happen when you connect a matched load,
> > source and a line.
> >
> > Neven
> >
> > ----- Original Message -----
> > From: Vinu Arumugham <vinu@cisco.com>
> > To: <si-list@silab.eng.sun.com>
> > Sent: Friday, May 05, 2000 5:55 PM
> > Subject: Re: [SI-LIST] : Trace Impedance Selection
> >
> > > If you were able to connect a transmitter to a receiver using a 377
ohm
> > > transmission line, this line would be in parallel to the "transmission
> > > line" between the two formed by free space. Therefore, one half the
> > > transmitted power would go through free space and the other half
through
> > > the line. As the line impedance is lowered, more power would be
> > > transmitted through the line and less through space.
> > >
> > > What's wrong with this scenario?
> > >
> > > Thanks,
> > > Vinu
> > >
> > > Mary wrote:
> > >
> > > > Somone recently claimed that higher impedance transmission lines
> > > > radiate more because their impedance is closer to the 377-ohm
> > > > impedance of free space. This is not true. It is not possible
> > > > to judge anything about the radiation from a transmission line
> > > > based on the value of its characteristic impedance.
> > > >
> > > > Characteristic impedance is the ratio of voltage to current in a
> > > > forward traveling wave. The ratio of electric to magnetic field
> > > > strength in a free-space transmission line is approximately
> > > > 377 ohms regardless of what the characteristic impedance is.
> > > > Even if you were to build a transmission line with a 377-ohm
> > > > characteristic impedance, there is no reason to believe it would
> > > > radiate any better or worse than a 300-ohm or a 400-ohm line.
> > > >
> > > > Mary
> > > >
> > > > **** To unsubscribe from si-list or si-list-digest: send e-mail to
> > > > majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
> > > > si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
> > > > si-list archives are accessible at http://www.qsl.net/wb6tpu
> > > > ****
> > >
> > >
> > > **** To unsubscribe from si-list or si-list-digest: send e-mail to
> > > majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
> > > si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
> > > si-list archives are accessible at http://www.qsl.net/wb6tpu
> > > ****
> > >
> > >
> > >
> >
> > **** To unsubscribe from si-list or si-list-digest: send e-mail to
> > majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
> > si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
> > si-list archives are accessible at http://www.qsl.net/wb6tpu
> > ****
>
>
> **** To unsubscribe from si-list or si-list-digest: send e-mail to
> majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
> si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
> si-list archives are accessible at http://www.qsl.net/wb6tpu
> ****
>
>
>

**** To unsubscribe from si-list or si-list-digest: send e-mail to
majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE
si-list or UNSUBSCRIBE si-list-digest, 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 : Wed Nov 22 2000 - 10:50:14 PST