Reg Edwards wrote:
"Ian White, G3SEK" wrote -
No - I meant exactly what I said. The meter can only indicate the
rho/SWR of whatever is connected downstream (load side) of the meter
itself.

==================================

We're agreed, down to and including point 5.

(1) I'm sure we are agreed our meters will correctly indicate Rho and SWR
only on 50-ohm lines.

(2) Insofar as the meter is concerned the transmitter's load impedance is
the input impedance of the transmission line between the meter and the
antenna. There may or may not be an intervening Z-match network.

(3) Insofar as the transmitter is concerned the line between meter and
antenna can be ANY impedance. It is desirable only that line length with its
Zo transform the antenna input impedance to somewhere near to 50 ohms (Like
a G5RV on 14.15 MHz). If things become difficult then a Z-match can be
inserted. Once we have selected Zo for this line we are no longer
interested in its SWR. And if we WERE interested a 50-ohm SWR meter would
be incapable of correctly measuring it.

(4) Note that when a Z-match is located at the transmitter end of this
feedline, and varied, the actual SWR on this line cannot change - yet the
SWR meter responds readily to the Z-match settings.

(5) The only line which, ideally, MUST be 50 ohms coax and have a small
SWR, certainly if it is of appreciable length, is that between the meter and
the transmitter. Otherwise the load directly presented to the transmitter
would not be 50 ohms. Any other impedance would transform the 50 ohms seen
immediately on the antenna side of the meter to some other value.

.... so we're OK as far as here.

The reasons you give why the impedance of that connecting line between
the transmitter and the meter must be 50 ohms are all correct. But
there's another reason - or at least, another way of looking at the same
situation: because the SWR meter has been calibrated for a 50 ohm system
reference impedance, it will not correctly indicate a 50 ohm load to the
transmitter unless the connecting line is also of 50 ohms impedance (or
is too short to matter).

(6) It is the SWR on this line which the meter indicates.

No, it isn't! The value of rho or SWR obtained from the forward and
reflected readings of the meter is the rho/SWR of the *load* connected
to the meter's *output* terminals.

Where the confusion arises is because the SWR on that 50 ohm connecting
line will be numerically equal to the SWR indicated by the meter - but
that's only because they are both 50-ohm devices. It is *not* because
the meter is reacting to the standing waves on its input side.

(If you change the impedance of that connecting line, then as you say in
(5) above, the load impedance presented to the transmitter will change,
so the power output will change. However, the forward and reflected
signals from the meter will both be affected in the same proportion, so
the rho/SWR result will not change.)

Or look at it another way: think of the meter as a bridge, fed by that
same connecting line - after all, some SWR meters are literally
three-resistor bridges with the unknown load impedance forming the
fourth arm. The bridge balance is not affected in any way by the length
or impedance of the line that is merely energizing the bridge. Only the
load impedance affects the bridge balance.


(If this line is
NOT 50-ohms then the meter incorrectly indicates the standing waves on it.
Which is what I said before and a lot of people disagreed.

And they were right, because the rho/SWR reading will *not* change.

What's incorrect is to believe that the rho/SWR meter is measuring
"standing waves" OR ANYTHING ELSE to do with its input line. It ain't.


Not that a false
indication is of great consequence when it is the incorrect choice of line
Zo where the problem arises.)

That contradicts what you said in (5) above. If the meter falsely
indicates correct transmitter loading, that certainly could be of
consequence.


(7) In practice at HF the length of this 50-ohm coax is often negligible.
The meter is often inside the transmitter box. As the misleading idea of
standingwaves on this short, even zero-length coax is nonsense

Reg, you must be about the only person in the world who worries about
that. The term "SWR" has almost completely freed itself from its
original literal associations with standing waves. Everyone else accepts
SWR as just one among the many mathematically interchangeable ways of
expressing the quality of an impedance match or mismatch. None of them
is either more or less valid than any of the others.

the name of
the instrument should be changed to TLI. (Transmitter Loading Indicator).
Which is all that it is!

It is indeed, so the very best of luck with your campaign to change its
name... ;-)


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
Editor, 'The VHF/UHF DX Book'
http://www.ifwtech.co.uk/g3sek