On Fri, 25 Jan 2008 04:13:56 GMT, Gene Fuller
wrote:

The irradiance equations
work fine for detailing the external effects, but they don't give any
hint of what happens inside the interface. Think Thevenin.

Hi Gene,

Cecil isn't going to rise far enough to catch a breath of air on this
one. You may as well drop the shoe for lurkers (and me).

I don't see the connection to Thevenin (specifically); but, for me,
inside the interface we can draw the correlation of TIR failure (the
chapter that Cecil hasn't drug across the Xerox yet) to evanescent
waves to near fields to how antennas work.

73's
Richard Clark, KB7QHC

Richard Clark wrote:
inside the interface we can draw the correlation of TIR failure (the
chapter that Cecil hasn't drug across the Xerox yet) to evanescent
waves ...

In the Ramo & Whinnery discussion of the distributed
network wave reflection model, I don't recall them
mentioning evanescent waves. Perhaps you could point
me to the correct page.
--
73, Cecil http://www.w5dxp.com

Richard Clark wrote:
On Fri, 25 Jan 2008 04:13:56 GMT, Gene Fuller
wrote:

The irradiance equations
work fine for detailing the external effects, but they don't give any
hint of what happens inside the interface. Think Thevenin.

Hi Gene,

Cecil isn't going to rise far enough to catch a breath of air on this
one. You may as well drop the shoe for lurkers (and me).

I don't see the connection to Thevenin (specifically); but, for me,
inside the interface we can draw the correlation of TIR failure (the
chapter that Cecil hasn't drug across the Xerox yet) to evanescent
waves to near fields to how antennas work.

73's
Richard Clark, KB7QHC

Richard,

Perhaps I have misread the message traffic for the past 5 years or so,
but it appears that most of the heat over wave reflections is about what
happens during the reflections, including detailed concern about energy
and momentum. The issue is never (or at least rarely) about what one
would find in external measurements on the transmission line (or free
space, as the case may be). This goes back at least to the battles
between Steve Best and Walt Maxwell. (I pointed out that both models
were correct, although they arrived at the proper conclusions in quite
different manners.)

More recently, we have great battles over interferometers and cute Java
demonstrations. Absolutely nobody would question the overall wave
superposition principles shown in the FSU Magnet Lab Java applet. Most
people know how to add sine waves, or at least they can look up the
technique. The entire debate usually comes down to arguments about how
that addition actually takes place physically. This is what I am calling
the Thevenin equivalent. In particular, the external observations are
unambiguous and non-controversial. The standard models and equations
give all the correct answers for the observables. At the same time those
models and equations say nothing about how the two FSU input waves
suddenly jump together to superpose and interfere. Do the waves
"cancel"? Is there a requirement for auxiliary waves that are created
and immediately destroyed? How does one account for missing momentum?

That attempted under-the-hood analysis with tools suitable only for
external description drives my suggestion of the parallel to the
Thevenin model.


73,
Gene
W4SZ

"Gene Fuller" wrote in message
...

That attempted under-the-hood analysis with tools suitable only for
external description drives my suggestion of the parallel to the Thevenin
model.

the Thevenin (or Norton) equivalent circuits i believe are rarely used in
analysing transmission lines and antennas, they are more commonly used for
breaking down lumped circuits and networks. In fact Jackson doesn't even
have them in the index of Classical Electrodynamics 2nd ed... It is
mentioned in Ramo,Whinnery, and VanDuzer Fields and Waves in Communications
Electronics in the index, but in scanning the 2 chapters it is listed for i
don't see an actual reference to Thevenin... one of those chapters deals
with microwave networks and components and does break them down into lumped
equivalents so that is probably where the reference belongs, the other
chapter wouldn't seem to be related so may be a typo.

The problem with the use of the Thevenin or Norton equivalents is that you
have to exactly respect the limitations in order to use them properly...
that is the part of the circuit being replaced with the 'black box'
equivelent must be linear and time invariant, and the analysis is only valid
for sinusoidal steady state. This last one is what gets everyone, it
eliminates all the transients and makes it impossible to use to figure out
what happens when that first reflection physically happens... you have to
ignore all that stuff and only consider the steady state solution.

We have seen that some posters on here don't want to accept those
limitations as they try to figure out 'what is in' the black box
equivalent... this is of course a non-sequitar as the whole purpose of
replacing a part of a circuit with the Thevenin equivalent is to simplify
the problem so you don't have to know what is inside and can focus on the
rest of the problem. So any attempts to measure the length of the line in
the black box, or figure out if it is a lumped circuit break the rules for
using it in the first place.

Dave wrote:
So any attempts to measure the length of the line in
the black box, or figure out if it is a lumped circuit break the rules for
using it in the first place.

Ramo and Whinnery warn against even trying to calculate
the power dissipation inside the black box - something
that is regularly attempted on this newsgroup. The
calculated power dissipation inside a Thevenin equivalent
and a Norton equivalent can result in an infinite difference
even though they are both "equivalent".
--
73, Cecil http://www.w5dxp.com

Dave wrote:
. . .
The problem with the use of the Thevenin or Norton equivalents is that you
have to exactly respect the limitations in order to use them properly...
that is the part of the circuit being replaced with the 'black box'
equivelent must be linear and time invariant, and the analysis is only valid
for sinusoidal steady state. This last one is what gets everyone, it
eliminates all the transients and makes it impossible to use to figure out
what happens when that first reflection physically happens... you have to
ignore all that stuff and only consider the steady state solution.
. . .

That's not true. A Thevenin or Norton equivalent generator can produce a
voltage or current, respectively, which is any function of time. This of
course includes single pulses and pulsed sinusoids, as well as an
infinite number of others.

Roy Lewallen, W7EL

One more note about Thevenin and Norton equivalent circuits:

In spite of the frequency with which this topic appears on this
newsgroup, I don't recall seeing anyone actually use a Thevenin or
Norton equivalent circuit in illustrating a point on this newsgroup. I
don't believe I ever have.

Roy Lewallen, W7EL

Roy Lewallen wrote:
I don't recall seeing anyone actually use a Thevenin or
Norton equivalent circuit in illustrating a point on this newsgroup. I
don't believe I ever have.

Some of your example sources walk and quack like
a Thevenin equivalent circuit. :-)
--
73, Cecil http://www.w5dxp.com