> As another note on the tx-rx isolation curves, a synthesized
> transmitter is
> not always worse than a crystal controlled transmitter as far
> as broad band
> noise output. If I remember right There are somewhere some
> curves showing
> that a certain model GE synthesized transmitter has less
> broad band noise
> than a typical crystal controlled transmitter therefore requiring less
> isolation in a duplex operation.
I was told that one of the design critieria for GE Delta-S/SX series mobiles
was that the transmitter noise was required to be LESS than that of the
Mastr II series PLL exciter. Some circuitry in Mastr III series equipment
is similiar to that found in the Delta-S, so maybe that's what you're
remembering.
> This leads to another subject, duplexer isolation. Lots of
> people are under
> the impression that if a duplexer that is adequate with 70 db
> of isolation
> and provides no desense then using one with 100 db of
> isolation will be much
> better. In this case the 100 db duplexer is only better at
> emptying your
> wallet. Once you have reached the needed isolation, more does
> nothing for
> you.
Except you have to take into account external influences that may affect Z,
phase, noise, etc., including those caused by changes in temperature and
humidity. Just squeaking by with 70 dB under perfect conditions doesn't
mean that you'll still have no desense when there's ice on the antenna or
the building tempature spikes at 120 when the HVAC fails.
> It is all about how
> much total power
> a mixer can handle before being overloaded. The narrower the front end
> filter the less total signal power reaches the mixer. All
> signals received
> add together in the mixer and are a determining factor in
> overload not just
> the strongest signals.
Well, sure, but if you take 99% of the out-of-band signals out of the mix
via tight filtering, you've eliminated the potential for that many mixes
from occuring.
> This is particularly true of analyzers that are
> incorporated within a service monitor. Most suffer badly from lack of
> dynamic range.
Yes. Even good SM's like 8920's are easily overloaded, often making
harmonics appear grossly exaggerated when transmitting into the RF in/out
port. If you can reduce the input level by adding X dB of attenuation, and
you see the harmonic drop by greater than X dB, you know that you're
overloading the instrument...
> Using a sharp notch tuned to the strong signal
> will often
> allow you to look much further down to reliably see other
> noise and spurs.
Not really practical when using the RF in/out port on a service monitor.
You'd need to use a lightly-coupled sampler or directional coupler, with the
transmitter operating into the antenna or dummy load, otherwise the
transmitter would be looking into a very bad (and potentially damaging) load
Z.
--- Jeff
