On Tue, 10 May 2011 19:35:55 -0700 (PDT), walt wrote:

To Richard C:

Richard, I must totally disagree that the same degree of mismatch
occurs at the tx output as at the dissipative load, the antenna. As I
have said many times, with the non-dissipative output source impedance
of the RF amp (tubes) the reflection at the junction of the line input
and the pi-network output is total. In other words, all of the
reflected power incident on the output of the pi-network is TOTALLY re-
reflected. Can't be anything less!

Hi Walt,

There's an active load behind that 2:1 mismatch seen by the reflected
power returning from the load. That is my part of the discussion.

However, if you are going to neglect that interpretation, then there
must be some explanation for TOTAL reflection. Fortunately there is,
it is called an open, or a short.

Now, we come to a philosophical quandry of the "Conjugate Match." If
we tune to a load, and we guarantee that the tune is the conjugate of
the load, then at any point along the transmission line in between
there exists looking at either end a conjugate of the opposite end.

In other words, putting a network analyzer into the line looking at
the load it sees R +jX, and looking back at the source it sees R -jX.
This is conjugation. Moving the analyzer along the length of the line
will transform R and X, but they will always track to a solution of
the conjugate.

You've been arguing this for years.

And, as the load is neither a short, nor an open (which is a special
solution of conjugation); then it stands to reason it is something
other which necessarily demands a solution in the form of:
R ±jX
from the source, where R has some non-zero, and non-infinite value.
For trivial antennas, that value of R falls somewhere between the 10s
of Ohms and the 100s of Ohms. For trivial, retail transmitters
available to Hams, similar values of R abound but most often fall
closest to 50 Ohms, or they would have a hard time getting optimal
power into the 50 Ohm pipe. Vendors are not going to shoot themselves
in the foot before that race.

A TOTAL reflection is available courtesy of the "active" part of this
special load. The "active" load (the transmitter sinking the
reflected power from the dissipative load) can offer one of two
mechanisms to provide the necessary short or open. When the source
wave coming out meets the returning wave coming in, and they are phase
coherent at 0 degree difference, then there is no potential difference
through the connector - no current flow. No current flow = open
circuit. Open circuit satisfies TOTAL reflection mechanics (but
violates conjugation). If they are phase coherent, but 180 degrees
different, then there is maximum current = short circuit. Short
circuit satisfies TOTAL reflection mechanics (but violates
conjugation).

However, in either of these scenarios, both powers (actually energies)
must be equal. Obviously that isn't the case with a nominal 100W
source looking at a 10W reflection.

Hence the analysis draws out slightly more along the lines I've
already posted and won't repeat here.

However, for the sake of argument, the condition of tune (conjugation)
and the state of coherency (a slippery slope, but here we are),
combine through a ringing in the line to create a virtual TOTAL
reflection at the connector of the Tuner/tuned amplifier. So, Walt, I
am not denying your case, but in the earlier posting I do preserve
conjugation in the explanation.

No transmitter, even with the best of tuning attempting to match will
forgive a short or open as load however. And, yes Richard, I have
experienced such conditions with 10s of W to 100s of kW and not all of
them accidental.

Another issue of moving sensors along the transmission line between
high mismatches is that there will be considerable error in the
determination of power - I presume this is where Jim's bookkeeping
explanation will lead, but that is still hanging on the line (no pun
there). I have experience the difficulty of those conditions too when
making precision power measurements.

73's
Richard Clark, KB7QHC