Pardon the extensive pasting, but it will allow a clearer post, and save a
lot of click time. / RF

"Ian White, G3SEK" wrote :
The meter measures nothing that involves the source, except
the level of RF that it supplies. It does not respond in any way
whatever to the source impedance.

Richard Fry wrote:
Not that I said it did in my part of the thread, but nevertheless the
above statement is not strictly true. In the case where the source Z
of the tx PA does not match its load Z (which is typical), power
reflected from the load mismatch will at least partly be re-reflected
from the PA -- which then contributes to the power sensed by a
"wattmeter" in the output path.

"Ian White, G3SEK" wrote :
Sorry, that statement cannot be correct. It would mean that the
impedance you measure at the near end of a transmission line
terminated by some arbitrary load at the far end) would depend
on the internal impedance of the device that's doing the
measuring - and that is not true, either in transmission-line
theory or in the real world. It is a function only of the line and
the load. etc

Richard Fry wrote:
How, then, do you explain the "ghost image" that can occur* in
analog(ue) TV transmission systems arising from reflections
at/near the antenna end of the station's transmission line?

*with sufficient round-trip propagation time in the transmission line

Ian White wrote:
Yes, that is a true observation, just as true as the one I made... so
now you have *two* different things to explain!

The so-called SWR meter is a steady-state instrument, so it always makes
sense to use that quicker, easier way of thinking. Since you're the one
who chooses to think of this particular situation in terms of multiple
reflections, any difficulties you encounter are entirely yours.

This reads to me as though you know they are there, but choose
to ignore them...?

If you ever see a conflict between two different theories that explain
the same observed facts, then there's an error somewhere.

We agree on the subject of conflict resolution, but apparently not on
the location of the error.

If the multiple-reflection theory is extrapolated to infinite time, so
that it calculates results for the steady state, it *must* give identical
results to the steady-state theory. But whenever the steady-state
theory can be used, it will always get you there much more quickly.

This is true only to the extent that all the power ever generated by the
transmitter eventually either is radiated by the antenna or is dissipated by
losses somewhere.

For simplicity, let's assume a tx with a source impedance of zero ohms feeds
a lossless transmission line of uniform impedance throughout its length to a
mismatch at the far end. The mismatch reflects a percentage of the incident
power back down the line to the tx, and continues to do so as long as the
transmitter generates power. The tx will re-reflect the reflected power
back to the far end -- in this case all of the reflected power it ever sees,
in fact. To this easily-seen, real-world reality you agreed above ("Yes,
that is a true observation, ...").

The re-reflections combine with the power generated by the tx at that
instant to create a vector sum at the sample point used by the meter. The
typical tx meter is a frequency-domain device, and cannot by itself separate
the RF output of the transmitter from re-/reflections of it. That requires
a time-domain device. So the magnitude of the transmission line samples
driving the tx RF metering circuits during normal operation under these
conditions become a function of both the source impedance and the
load impedance.

The defense rests.

RF

Richard Fry wrote:
This is true only to the extent that all the power ever generated by the
transmitter eventually either is radiated by the antenna or is dissipated by
losses somewhere.

The fly in the ointment is a definition.

If a signal generator is sourcing 100 watts and 20 watts of reflected
power is being dissipated in a circulator load resistor, we say the
source is sourcing 100 watts and 20 watts of reflected power is being
dissipated in the circulator load resistor.

If the identical thing happens in a ham transmitter, we say that the
source is sourcing 80 watts, BY DEFINITION. What's wrong with this
picture? Ham transmitters NEVER re-reflect anything, by definition.

The reason that the source impedance doesn't enter into the forward/reflected
power values is that it has been defined out of any relationship to them. By
definition, there is zero power re-reflected from a ham transmitter NO MATTER
WHAT THE IMPEDANCE OF THE HAM TRANSMITTER MIGHT BE. Never mind that we can see
those reflections with our own eyes in TV ghosting. We must be crazy because
they have been defined out of existence. How dare we have the gall to observe
them!
--
73, Cecil http://www.qsl.net/w5dxp


-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 100,000 Newsgroups - 19 Different Servers! =-----

Richard Fry wrote:

Ian White wrote:
The so-called SWR meter is a steady-state instrument, so it always makes
sense to use that quicker, easier way of thinking. Since you're the one
who chooses to think of this particular situation in terms of multiple
reflections, any difficulties you encounter are entirely yours.

This reads to me as though you know they are there, but choose to
ignore them...?

Oh no, quite the opposite - but since these difficulties are entirely of
your own making, you get to do the work :-)

If you ever see a conflict between two different theories that explain
the same observed facts, then there's an error somewhere.

We agree on the subject of conflict resolution, but apparently not on
the location of the error.

Thank you for the more detailed explanation below... which, sure enough,
revealed where the error is.

If the multiple-reflection theory is extrapolated to infinite time, so
that it calculates results for the steady state, it *must* give identical
results to the steady-state theory. But whenever the steady-state
theory can be used, it will always get you there much more quickly.

This is true only to the extent that all the power ever generated by
the transmitter eventually either is radiated by the antenna or is
dissipated by losses somewhere.

That is exactly true in the steady state.

For simplicity, let's assume a tx with a source impedance of zero ohms
feeds a lossless transmission line of uniform impedance throughout its
length to a mismatch at the far end. The mismatch reflects a
percentage of the incident power back down the line to the tx, and
continues to do so as long as the transmitter generates power. The tx
will re-reflect the reflected power back to the far end -- in this case
all of the reflected power it ever sees, in fact. To this easily-seen,
real-world reality you agreed above ("Yes, that is a true observation, ...").

The re-reflections combine with the power generated by the tx at that
instant to create a vector sum at the sample point used by the meter.

There's the error: you can't "combine... power" in that way. You can
only create vector sums of voltage; and separately, vector sums of
current.

To make the multiple-reflection theory work correctly, you have to do
two separate vector sums at output port of the transmitter. First you
add all the voltage vectors: the 1st (original) forward, the 1st
reflected, the 2nd forward (re-reflected), 2nd reflected... and so on,
summed to infinity to give the correct result for the steady state.
Then you do the exactly same for all the forward and reflected current
vectors.

In order to account for reflection from the transmitter, you have to
assume some value of source impedance. Any value will do, for reasons
we'll see in a moment.

Now you can calculate two things: the vector ratio, which is the complex
impedance that the transmitter sees as a load; and the scalar product,
which is the power the transmitter can deliver into that load.

If you vary the source impedance of the transmitter, it will change all
the summed voltage vectors and all the summed current vectors - but each
voltage term in the sum will be changed by exactly the same factor as
its corresponding current term. Certainly the product (the output
power) will change, but the ratio (the load impedance) will not.

So, when correctly worked out, the load impedance is *not* a function of
the transmitter output impedance or the output power. Likewise, the
indication of the SWR meter is not a function of either the transmitter
output impedance or the output power - this last one being a well-known
fact.



--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek