<<< see embedded responses below >>> --- In [email protected], "Gary Schafer" <[EMAIL PROTECTED]> wrote: > > Duplexers and cavities have been tuned for many years > without a tracking generator or spectrum analyzer. It > is a little easier with a tracking generator or even a > spectrum analyzer and a signal generator. But you can > do the job with a signal generator that you can keep > on frequency and a receiver.
W6NCT: This sounds encouraging. > It helps to have some kind of signal level meter on the > receiver but you can also just tune for quieting by riding > the generator level so that the receiver is always noisy. > That way you can hear if the level goes up or down. W6NCT: I think that understand the notion of using a receiver with a quantifiable indicator of receiver signal strength; but I don't understand what you are saying in your statement "tune for quieting by riding the generator level so that the receiver is always noisy". Could you explain this a bit more? > > Using a sweep generator works fine for tuning pass band > cavities or even the pass band part of a pass/notch type > duplexer but you can't tune the notches with a sweep > generator. The reason being is that the detector is not > sensitive enough for the low signal levels that you need > to work with. A typical diode detector will only work at > about -20 to -30 dbm. Even if your sweep generator is able > to put out +20 dbm that only lets you look down -40 to > -50 db and the duplexer notch is going to be (hopefully) > in excess of -70 db. You will never find the peak of the > notch. > > So just a signal generator preferably a synthesized unit > so you know what frequency it is on. You could use a tunable > generator and a frequency counter but lots of generators > tend to shift frequency slightly when you crank up the output > level (high enough for the counter to see) and you will be > working with the signal attenuated down to -80 to -100 dbm > most of the time while tuning the duplexer. You can not afford > to have the signal shift around or drift while you are trying > to tune things as it doesn't take much of a frequency shift > to mess up your tuning. W6NCT: Can I use the sweep generator to tune the band-pass; then just the generator in the IFR-1100 to generate the driving signal into the cavities to tune the notch? The IFR is synthesized and I'm guesing that it might be able to put out more than the sweep generator without becoming unstable. > > With a good generator tune the pass bands at the proper > frequencies first. Then tune the notches at the opposite > frequencies. On the transmit cavities you would tune the > pass to the transmit frequency and then tune their notches > to the receive frequency (etc.). Then tune the pass tuning > at the proper frequency again now that you have everything > in the ball park. Then tune the notches again at the proper > frequencies. > > ALWAYS tune the notches last. Do NOT touch the pass tuning > again when you are done with the notch tuning. Tuning the > pass just slightly will throw the notch way off. W6NCT: Understood. Tx cavities are to band-pass Tx freq and notch Rx freq; similarly, Rx cavities are to band-pass Rx freq and notch Tx freq. Right? > > Be sure and have a dummy load on the port of the duplexer > you are not working with and preferably 6 to 10 db pads on > each of the ports that you are working with or you may not > have the same tuning when you hook it up to your radio. W6NCT: This confuses me a little. In addition to the X and Y outputs to the oscilloscope, the Wavetek 1080 has an "RF output", "demod In", and a "marker In". Not having a manual to tell me the proper setup, I was guessing that the sweep generator's "RF output" would connect to the input port of the cavity, and that maybe I should connect the output of the cavity to the "demod in" (leaving the marker input on the sweep generator unconnected). I wasn't sure if I should use some sort of an external circuit between the cavity output and the 1080's demod in. Do you know? Alternatively, I thought that maybe I should put a T connector on the output of the cavity with one side connected to a (50 ohm) dummy load, and the other side connected to an RF-detector that feeds into a second channel of the oscilloscope (leaving both the demod and marker inputs on the sweep generator unconnected). I tried this setup, since I didn't know what the "demod In" expected. Am I off-track with these possible setups? Either way, I don't understand what the pads are for. Please explain? > > 73 > Gary K4FMX
