> My question has to do with how much TX/RX isolation is required for
> different frequency spacings. I'm assuming that as the TX/RX
> frequency spacing increases, the amount of isolation that the cavities
> needs to provide decreases. My question is how do I calculate this?
There are two things to consider when determining how much isolation is
necessary to avoid desense:
1. Carrier supression. This is how far down the transmitter's carrier
needs to be attenuated to avoid overloading the receiver.
2. Noise supression. This is how much attenuation is required of the
transmitter's noise energy that falls on your receive frequency.
How much isolation you need to satisfy #1 is a function of your RECEIVER's
performance. A receiver with a tight front end that is able to provide some
attenuation of the transmit frequency will require external less carrier
supression filtering than one without. On 2m, this is typically not
relevent, because most all receiver front ends will pass a signal only 600
kHz away with little attenuation. On other bands, it may play a bigger
role, but it depends on the receiver. If you take a typical ham rig, or a
Maxtrac, or something else with a barn door front end, you're not going to
get any additional carrier supression from the receiver, unlike a radio with
a tight front end. There are other factors beyond just front end filtering
that determine the usable dynamic range of a receiver at a given frequency
offset; I'm just using that as the most obvious and easiest to discuss in a
reasonable amount of typing.
For #2, how much isolation you need is a function of your TRANSMITTER's
performance. Noisier transmitters will require more filtering, plain and
simple. Modulation products ("sidebands") are typically not a major
contributor to noise falling on the receive frequency in a
properly-operating transmitter at the offsets we typically deal with;
broadband noise is the enemy. The better/cleaner the transmitter, the less
noise supression is needed at the receive frequency.
Take a look at http://www.repeater-builder.com/pdf/GE_Isolation_Curves.pdf.
You'll see isolation curves for highband Mastr II stations, comparing a PLL
exciter versus a PM exciter. You'll notice that the amount of carrier
supression is the same for both stations, but that the station with the PLL
exciter requires substantially less noise supression.
GE was probably one of the best when it came to publishing isolation curves
for their repeaters. Based on those curves, you had a pretty good idea how
much isolation you were going to need. Of course, you always want to have
excess isolation to account for changes in temperature, variations from unit
to unit, etc.
Dave KA9FUR has scanned all of the GE curves he was able to get his hands on
and posted them on his web site:
http://www.ka9fur.net/geduplex/duplex.html
These can be used as a starting point, but you may need more or less
additional isolation than what the curves predict.
--- Jeff
--------------------------------------------
Jeff DePolo WN3A - [EMAIL PROTECTED]
Broadcast and Communications Consultant
