>
>
> Hi Tony,
>
> to answer your question if it will work, the answer is yes it
> will and i am doing it on a site the same way. You will
> however need two dual isolators to add to each transmitter.
> You will tune the reject the same way as you would normally
> for the specific frequencies.
>
> Good Luck,
>
> Mike K7PFJ
Dual-stage isolators on top of a good pass/reject duplexer seems like
overkill to me. Between the duplexer and the isolators you're talking about
130+ dB or more of isolation. 130 dB of isolation is about what you'd have
if the two transmitters were 100 miles apart!
50 dB of isolation is a *very* coarse rule-of-thumb target to use for
combining isolation requirements. Sometimes you need more, sometimes you
can get away with less, particularly if you're using a transmitter/PA that
has either a built-in isolator or an output combining network that affords
high turnaround loss (such as a hybrid). As always, you won't know for
certain whether or not you have enough isolation until you *measure* the IM
products, unless you know the turnaround losses of the transmitters ahead of
time.
To wit, I'm combining a UHF Mastr II and a Glenayre PA using a hybrid with a
single-stage isolator w/low-pass filter on each transmitter (run-of-the-mill
Celwave hybrid combiner panel), about 55 dB transmitter-to-transmitter
isolation. IM products are better than -80 dBc, which is about how far down
I can measure due to test equipment dynamic range limitations (because the
two frequencies are very close, filtering out the fundamentals to increase
the usable dynamic range isn't practical when looking for the third-order
mix products). In other words, if there are mix products, I can't see them.
>From a practical standpoint, I can drive around in the vicinity of the tower
(like within a few hundred yards) and tune the radio in the truck to the
third-order mix frequencies and hear nothing.
Aside from measuring the relative (in terms of dBc) amplitude of the mix
products to gauge the merit of the system, what's really important is the
absolute amplitude (in dBm or whatever you want to measure it in) that makes
it to the antenna, and what they could potentially affect. If one of the
mix products falls on a repeater receive frequency, and that repeater
happens to be co-located, chances are you're going to be up the proverbial
creek. Even if the potentially-affected receiver isn't co-located, if the
absolute level is strong enough, you may still have problems. Nothing new
or exciting here, isolation is isolation whether it's achieved spatially or
through filtering, there's a good chance you'd have the same problem even if
you were running separate antennas on the same tower. Free space path loss
buys you a lot of isolation over a short range initially - at UHF, about 70
dB for the first tenth of a mile. To pick up another 20 dB, you have to go
out 1 mile. To get another 20 dB, you need to go 10 miles. So the goal is
to keep the radiated energy low enough that it falls into the noise before
it "gets out" too far to cause trouble...
Sorry, I drifted a bit off track. Bottom line, while I always advocate the
use of an isolator in a repeater installation, I think spec'ing a dual-stage
isolator on top of a good pass/reject duplexer is overkill in this case if
the sole reason for adding the isolators is prevent IM.
--- Jeff WN3A
