On Feb 1, 6:24*pm, "Antonio Vernucci" wrote:
As long as it is ONLY 2 ports, has ONLY passive linear components, AND
is small enough to be considered a lumped element, then you can always
reverse the terminals and not know the difference. *The most
frequently violated condition is the last one, put in a piece of coax
with that is long enough to be measured at the highest frequency you
will use and all bets are off.

Could you please explain me the meaning of your sentence:

"put in a piece of coax with that is long enough to be measured at the highest
frequency you will use and all bets are off"

as I have difficulties to interpret it due to my non-mother tongue english.

73

Tony I0JX
Rome, Italy

the general case is, if any part of the circuit is more than a small
fraction of a wavelength in size you may be able to detect the
difference between the ports. one common way to do that is to try to
measure a circuit that has a long piece of coax in it, the results may
be very different when you reverse the terminals.