"Cecil Moore" wrote in message
m...
Reg Edwards wrote:
But it appears to me that ALL SWR meters work on the principle of a
common, identical, take off point for both voltage and current
samples.

There are a number of different designs. One very popular class of
designs samples the voltage and current essentially at a point. If
the device is in a 50 ohm environment and the two samples are equal
in magnitude and phase, there is no reflected power. If the device
is in a non-50 ohm environment and the two samples are equal in
magnitude and phase, reflected power exists at the sample point even
though the device reports none.

Therefore, anything inside the meter which could constitute a
transmission line, no matter how long or short, if it separates the
two effective take-off points, it is a small source of error and its
actual length otherwise plays no part in the measuring process.

( I mean its actual length is not included in the basic design
calculations, except perhaps in rare instances to minimise the error.)

( The error arises from a phase-shift error in tapping off the current
sample. There may be no problem with defining the location of the
voltage sample. All depends on meter construction and the tapping
methods which vary but do not affect the basic design principle of a
common tapping point.)

I am not trolling. Please read very carefully. Try to limit yourself
to HF and VHF. Do you agree? Yes or No?

All of the above is why this class of wattmeters, including the Bird,
does not establish a 50 ohm environment merely by being inserted in
a non-50 ohm system. The Bird samples voltage and current essentially
at a point, assumes it is in a 50 ohm environment, and reports the
power reading. If it is not in a 50 ohm environment, the power reading
that it reports will be wrong as Owen's experiment proved.

ah, but if the bird does establish a 50 ohm 'environment' within its 40mm
housing then the reading that was actually obtained is indeed correct. so
now, how else could you read zero reflected power at that point if it didn't
establish that environment? what other evidence to you have that 40mm is
too short? if 40mm is too short then why do we worry at all about impedance
bumps in connectors that are 20mm or so long??? if 20mm is enough to see a
bump then that is obviously affecting the 'environment' on the line.