And for anyone doubting this dual definition of SWR, just take a look at
Agilent, Narda, or similar web sites, and notice that their high priced
and precision terminations all come with an SWR specification. You'll
also find SWR specs for test equipment inputs and many other devices not
containing transmission lines. These are clear examples of Definition 2.

Roy Lewallen, W7EL

Ian White, G3SEK wrote:
Reg Edwards wrote:


The only "inconsistency" is that an SWR meter is obviously NOT a "TLI",
or transmitter loading indicator, as you've been so kind as to
graphically point out.

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

It would be more true if you had said the SWR meter is not being used
as an
SWR meter because there's no SWR for it to measure.

There's an inconsistency because the line between transmitter and
meter is
not 50 ohms. The Tx is incorrectly loaded but neither meter scale
gives any
indication of it to the user. It could be a serious matter but
there's no
warning. The operator is allowed to believe he has set up the equipment
correctly.

In that case, calling the meter a "Transmitter Loading Indicator" will
only *add* to the confusion!

The rho/SWR meter has only ever been claimed to measure conditions on
the *load* side of itself. It knows nothing about conditions on the
transmitter side; and until Reg came along, nobody ever claimed that it
could.

Now here comes Reg, saying "Let's all call it a Transmitter Loading
Indicator!" Later, he realises that his "TLI" will only work correctly
if the connecting line between the transmitter and the meter is (a) of
impedance Zo, or (b) vanishingly short.

And his reaction to that: blame the meter! Me, I'd go looking for that
bloke who invented "TLI"...


A TLI suffers from the same disadvantage as an SWR meter - it gives the
correct answers only when making measurements on 50-ohm lines. This
should
not be surprising. They have identical circuits.

But when there is no line of any impedance, just a few inches of wire,
the
TLI indicates correctly. Whereas the SWR meter requires at least a
1/4-wavelength of 50-ohm line before it stops being dishonest. And it
doesn't stop telling white lies even on longer lengths.



What Reg refuses to accept is that there are two alternative definitions
and usages of the term "SWR":

Definition 1. Given an impedance Z at any single point in any circuit,
and a system reference impedance Zo, then SWR is a *mathematical*
function of Z and Zo that tells you how close together the two
impedances are.



Definition 2. Given a transmission line that is terminated in an
impedance Z that is different from the line's own characteristic
impedance Z(line), then a standing wave of varying voltage and current
will appear on the line. If the line is long enough to identify a
voltage maximum and a voltage minimum (an electrical quarter-wavelength
away) then SWR is defined as the ratio Vmax/Vmin.

That was the original definition of SWR, but it has severe limitations.
For a lossless line with a characteristic impedance the same as Zo, the
two definitions of SWR are related by a simple, fixed mathematical
formula. But that's a special case; in all other cases, definition 2 has
to be applied with care (or you may even judge that it isn't valid at all).


Definition 1 is by far the most common definition and usage of "SWR" -
it's just one of several alternative measures of impedance match (others
including rho, return loss, S11 etc). All these different alternatives
are inter-related by defined conversion formulae. RF/microwave engineers
move freely between all of them, using whichever term is most convenient
at the time, or more normal in that particular area of electronics.

It's no big deal. Life is full of examples where the same word is used
to mean many subtly different things. The whole of the
electronics/RF/microwave engineering profession is very comfortable with
"SWR" meaning more than one thing. Most amateurs never need to think
about that difference; but if they do, they don't find it difficult.

AFAIK, Reg is the only person on the planet who understands all of this
perfectly, but refuses point-blank to accept it.