I haven't been following this thread lately, but happened on this
particular posting...

Seems to me you want to say "Linear TIME INVARIANT system" to get to
not generating new frequencies. Certainly I can build a system that is
linear but not time-invariant and generate new frequencies with that
system. A simple one is a signal going into a potentiometer, coming
out the wiper, in which the wiper is rotated continuously. It's
linear, but not time invariant, and obviously any input will be
amplitued modulated at the rate of the time variation.

Is a double-balanced mixer with LO a linear system (for input signals
in the intended amplitued range)? Increasing the input amplitude by
1dB causes the output amplitude to increase by 1dB, though the output
is not at the same frequency as the input. If the response of the
DBM/LO system to input x1 is y1, and to x2 is y2, then is (y1+y2) the
response to input (x1+x2)? Is a DBM/LO system time-invariant: if I
apply stimulus x1 at time t1 do I get the same response as if I apply
it at time t2 (where the response is also shifted by t2-t1)?

Perhaps this will be useful food for thought...

Cheers,
Tom

Steve N. wrote:
wrote in message
oups.com...
Thanks, Steve,

Good post.

Thanks.

You do need to read a little mo by mentioning Doppler
shifts, I was illustrating that a linear system can generate new
frequencies without violating the law of superposition.

Hi Glenn,
I got that. Perhaps my use of "relativitistic effects" was inappropriate.
I still maintain that Doppler shift is not an example of (at least what I
would call) "creating new frequencies" because an electronis system did not
cause said shift. A system is linear or non-linear, but a system can't
"make Doppler happen". It is a frequency change, yes, but an "electronic
system" can't cause it. It occurs for reasons other than "system
characteristics". Yes, a space craft can be considered part of a system in
teh general sense, but not as I believe an "electronic system" should be
thought of when discussing such things. Perhaps symmantics in your view,
but not mine. I think it is a fundamental, but, perhaps can't explain
adequately why.
What a thread. usually when a thread gets this long it digresses far
outside the original intent.


...Earlier posts were
confusing as the term 'linearity' ...electrical superposition sense
... and ... current distribution along the antenna element.

So you *were* questioning the linearity of antennas as we understand the
term?

73, Steve


73,
Glenn AC7ZN



Steve N. wrote:
wrote in message
oups.com...
... By the way, generating new frequencies is not necessarily a
violation
of superposition (though it usually is). Consider a system undergoing
a constant Doppler shift.

73,
Glenn AC7ZN


Glenn,
boy! I can't read all these posts, so I was trying to see where you
were
going and asking by just skimming, primarily your posts.

The above caught my eye. Doppler is not "generating new frequencies" as
a
non linearity does. A non-linear system will produce harmonics with one
exitation frequency and produce the common mixing / IM with multiple
exitation (superimposed) frequencies. I think trying to mix
relativistic
effects in with the stationary world is an unnecessary complication of a
linearity discussion.
If I have time I'll try to follow the thread to see what you're really
after...but. If it takes mega watts to see some non linearity in an
antenna, who cares? and more importantly how will you know whree it is
occuring since things like the junction of two connectorc can produce
enough
IM to mask other, smaller sources. If you tried an experiment
looking
for IM / Mixing you might try to use a receiver becaue a receiver could
be a
very sensitive detector...but you'd have to have a pretty good receiver.
Something like a kW LO and mixer to have a really good intercept.
I'm not sure of the point here... Do antennas cause IM?
Sounds like a deadend arena to me.

73, Steve, K9DCI