> An isolator wont cause intermod, but it may cause harmonics.
How so? What phenomenon occurs in a ferrite circulator that creates
harmonics but won't generate IMD?
> Commercial installations usually use either a harmonic filter
> and 3db hybrid coupler, or a special type of band pass cavity
> to couple the output from the isolator to the feed line.
> (This info from an RX TX application note on transmitter combiners)
Well, that's just two out of a myriad of ways of combining, duplexing,
multicoupling, etc. Hybrid-ferrite and TX-RX's T-pass are by no means the
only two ways of combining transmitters.
> I'm proposing a novel application of the circulator (an
> isolator without the dummy load on one port)..
Yes, I know what you're trying to do, and it's nothing new. UHF Micor
mobiles have an isolator in the antenna network that routes received RF
through the isolator to the receiver when the transmitter is not keyed. I
forget how the full-duplex Micor med radios were set up; I believe the
antenna network in those radios was different, maybe somebody else
remembers. Circulators are used in the manner you're contemplating in
duplex microwave systems all the time, but with the addition of filters in
both paths.
> the
> transmit chain will still use more or less conventional
> combining techniques to merge the signals from the 3
> transmitters. the output from the transmitter combiner goes
> to the input of an additional circulator, the circulator
> output goes to the antenna as you would normally expect for
> an isolator, and the "load" port goes to the receive chain
> instead of a dummy load
In your example, you will have three transmitter carriers going through a
common isolator. That's where I think you may have spectral problems show
up due to IM amidst the high-level carriers.
> Since the path from the transmit
> chain port to the receive chain port is "reversed" compared
> to the normal signal flow in a circulator, it will incur
> 20-30db of loss, depending on the circulator specs.
It's not only a function of the circulator's inherent directivity spec; the
usable isolation is likely going to be dominated by the mismatch at the
antenna port as I said before. Even if you match well when you install it,
feedline phase change with temperature, feedpoint Z varying with weather
conditions, and all of the other uncertainties and uncontrollables in the
antenna system are only going to degrade the match (and therefore isolation)
over time.
--- Jeff
