In my 38+ years of working with amateur repeater stations, it has become obvious to me that tube-type transmitters were easier to make duplex. The first 2 meter repeater I helped build was for the local amateur radio club and the radio equipment used was GE MASTR Pro. We had a new Sinclair Hybrid-Ring duplexer. This duplexer worked fine on the tube MASTR Pro, however when I built a brand-new Hamtronics repeater some years later, I had trouble. The transmitter output power was the same, the only difference was the all-solid-state exciter and Amplifier.
The 'problem' was not the Hamtronics repeater; one of the cans of the duplexer was not tuned properly (wasn't even close) and this prevented the solid-state repeater from working, even though the tube-type repeater worked fine. But WHY?
One reason was receiver sensitivity. The Hamtronics receiver was quite a bit more sensitive (8 or 9 dB) so the duplexer had to make up for this. Also, solid-state amplifier stages that are not tuned are therefore broadband, amplifying everything within a reasonable bandwidth the same amount. The stages in the Hamtronics exciter are tuned, but noise at 600 kHz still gets through, plus the 15 Watt amp is somewhat broadband.
Tube-type exciters and amplifiers use tuned input and output stages that are "High-Q" meaning their operating bandwidth is VERY limited without re-tuning. This is a definite plus in repeater service because they amplify the repeater's desired signal but not the exciter side-band noise. Since most repeaters don't QSY often, limited tuning is not a limitation, but rather a desired effect.
What am I getting at?
It has been my experience, on 2 meters with a 600 kHz. transmit to receive split, that you can run up to 100 watts of transistor power using a good duplexer and a single duplex antenna with good results. Over 100 watts of transistor power, with even the best duplexer and a single antenna, may not give enough isolation for NO desense. On the contrary, tube power can exceed several hundred watts with excellent results.
My 145.270 KQ3M repeater in Hays Mill PA has a GE MASTR Pro/II 4CX250R tube-type amplifier model 4EF5A1 capable of 330 Watts. This amplifier has been driven with solid-state Hamtronics and Motorola exciters to this power level with a 4-cavity WACOM duplexer (93 dB isolation) into a single duplex antenna with NO desense. The receive sensitivity on this repeater is under 0.125uV for 12 dB SINAD (yes, receiver sensitivity plays a part in desense). I have tried to run transistor PAs at one-half of the tube level and, all else being the same, with unacceptable results. Unacceptable to me is the existence of any receiver desense. The difference between the solid state and tube power amplifiers ability to duplex is simple. The "Q" of the tube PA is very high beacuse its circuitry needs to be tuned on frequency. This makes it difficult for the tube amp to amplify signals (or noise) off frequency. In fact, the tube PA can actually 'clean-up' a noisy multiplier exciter because of its inability to amplify signals which are off of frequency to which it is tuned. Solid State PA's are broadband and noisy in nature - - nothing new here.
Always strive for NO desense if at all possible. The use of a simple toggle switch in the PTT line, and listening to the receiver on a local speaker, is the only good way to test for desense. Use a weak signal into the Antenna System, not by inserting it into the receiver or duplexer line with a "tee" connector or some other unacceptable manner, (everyone else has to go through the antenna). Don't be afraid of tweaking the last drop from the equipment, especially the duplexer. Duplexers that are "factory tuned" are done so with laboratory test equipment with near perfect 50 ohm impedance. Is your antenna, transmitter and receiver a perfect 50 ohms? I doubt it. The impedance presented by the transmitter and receiver can affect duplexer tuning. This is no big deal in tube type equipment because you can tune out reactances with transmitter output and receiver input tuning controls, but with no-tune transistor equipment you are stuck with whatever the impedance is. This may mean using a PI network or stubs cut and placed properly to improve return loss (match). Like Jeff DePolo - WN3A says, "you shouldn't use your duplexer as an antenna tuner".
Here's something else that commonly is misunderstood:
> I've Googled and Googled and can't find the answer I'm
> looking for so I'll try here. I'm trying to find out the
> correlation between the rated power in "watts" and amount of
> isolation "dB" of a duplexer. Example" a duplexer has 70db of
> isolation and has a power rating of 50 watts. How do I figure
> out the amount of power a particular duplexer can handle at a
> given amount of isolation ?
The two aren't directly related.
The power rating is the maximum power that the duplexer can handle before
failure caused by heating or voltage breakdown. If the duplexer is rated at
50 watts, that's the most it can take with a flat VSWR. A prudent user
wouldn't run a duplexer at its maximum - you should always have some
headroom to account for antenna system anomalies, whether constant (poor
antenna VSWR) or occasional (antenna iced up).
The isolation spec doesn't have any relationship to the maximum
power-handling of the duplexer. How much isolation you need depends on a
multitude of factors, transmitter power output being just one of them. The transmitters purity
and the amplifiers noise figure are two more, then there's the receivers actual sensitivity. All
of these factors and more go into how much isolation is needed to duplex.
--- Jeff WN3A
So now we understand why a duplexer rated for 350 Watts might not be adequate to make a particular set of equipment operate without desense at a TX power level far removed from the duplexer's power rating.
Dec 19 1996 Kevin K. Custer W3KKC
Updated October 2016 due to aging text (first line).
Last updated: October 2016 - - W3KKC