On Sat, 30 Aug 2003 11:58:55 -0500, W5DXP
wrote:

Richard Clark wrote:
Simply going to the
bench would be adequate for amateurs, guessing would eclipse the
efforts of everyone else. ;-)

I'm ready to go to the bench but I don't know how to separate the
incident reflected power that is re-reflected from the source from
the generated power. That's what I am asking. How does one separate
those two coherent superposed forward waves?

It should seem obvious that the power reflected at the source
interface is governed by the standard mechanics no one wishes to
impart to the source.

I agree that the source obeys the wave reflection model rules. The
fly in the ointment is the unknowable source impedance encountered
by the incident reflected waves.

Worry about it if the example of the challenge presents more than 3+
dB loss. If not, there's hardly any point is there? We can work out
the differences then, but only if you care about accuracy and the lack
of confirming this exotic characteristic. I can appreciate that this
is not everyone's interest.


In other words the power returning to the
source will be reflected through the same mechanics it met at the
mismatched load, with the "exact" value varying only by the reflection
coefficient of the source mismatch and the power incident upon it. As
the line presents a media of 50 Ohms, and the source presents a 100
Ohm discontinuity, ...

I must have missed how you know the source presents a 100 ohm impedance
to incident reflected waves. Do you have a 100 ohm pad between the
transmitter output and the transmission line?

Hi Cecil,

I didn't say I did the example, it is drawn from a reference. We've
already been through the mechanics of how to do it employing a
variable transmission line. Consult our correspondence for specific
details, you already offered that your equipment could tolerate that
mismatch, however THAT discussion is separate and distinct from the
example of the challenge. The challenge merely offers another
approach. You can add a 50 Ohm Dummy Load in series to the output of
your rig (this, of course presumes it presents 50 Ohms characteristic,
but as many declaim that specification perhaps they could offer
another value - eh, unlikely).

Another issue of adding a series resistor is one of shielding and
common mode issues. You asked me in that earlier correspondence if I
considered this, and yes I did. That is why I used massive parallel
loads that insured an entirely shielded system. This is one of those
methods that one correspondent pondered:
There is no institutionalized ignorance, just a
lot of skepticism regarding the reliability of the
analysis methods and the measurement methods.
which is understandable from those not trained in the art of designing
measurement scenarios. I am trained but those still caught in the
quandary are immobilized by rejecting every method (institutionalized
ignorance). The measure of RF power is not simple by any means so it
is best left to those who are serious about accuracy - hardly an
amateur pursuit, and it hardly matters anyway as it is not a problem
with modern equipment.

I offered that you already had the tools to perform a simple first
pass approximation, you really need to consult that thread of
correspondence again. None of this with forced mismatches is all that
hard in the first place. The greater the mismatch at each end, the
more compelling the evidence. The simple fact of the matter is that
most rigs conform to 50 Ohm source Z and do not exhibit this issue.
If those who held to their cherished fantasy of source Z being other
than 50 Ohms, then they should be able to ace this test from the
beginning (the lack of their correspondence reveals the invalidity of
their claims).

I'm working on two repeater systems today (10M and GMRS), so enjoy.

73's
Richard Clark, KB7QHC