On Jun 13, 6:09*pm, Richard Fry wrote:
On Jun 13, 4:33*pm, walt wrote:

The point I'm emphasizing
is that because the output resistance of these amps is non-
dissipative, no reflected power from a mismatched load enters the amp,
but is totally re-reflected in the forward direction.

Walt, wouldn't that non-dissipative output resistance also reflect
reverse power on other close frequencies, such as those coupled from a
co-located transmission system within the bandwidth of the tx output
network?

If so, such external signals would never enter the transmitter, and
could not cause r-f intermodulation by mixing with the main signal in
the (non-linear) PA plate circuit of that transmitter to produce a 3rd
order product *at *2F1 *- F2, and other combinations.

But such r-f intermods are a fact of life, as shown in the Mendenhall
paper, and in the existence of elaborate and expensive r-f filtering
hardware needed to enable several stations to operate from one
location while meeting a radiated r-f intermod spec of -80 dBc -- even
from one antenna!

RF

Richard, I'm not qualified to answer your question about intermods, so
I won't even try. However, interference to the amp from radiation from
an antenna driven by another amp will not be phase or frequency
coherent with the tx receiving the interference. Perhaps that's the
reason the interfering signal gets through while the waves reflected
from a mismatch don't? In other words, could the coherent relationship
between the forward and reflected waves result in the total re-
reflection at the non-dissipative output resistance of waves returning
from the mismatched load?

Walt