I asked that question for most of the people on the list that have not tuned a duplexer themselves or don't have the proper test equip to do it right and have done it with 2 radios or signal gen and a receiver. These work but it is so much easier with a tracking generator once you learn how to do it.
That's the way I've always done it with the 6 or 10 db load/attenuators on the ports being used at the time and loads on the other. Some people tune duplexers w/out the loads on unused ports and that throws everything off in the tuning. I have never seen it done but there should be a way to "see" a particular cable's resonance on a tracking generator, but don't know how to do it as in shorted or open stubs. Does it take a network analyzer to do this? I have not tried what you suggest to tune the cans in the duplexer seperate and then put the cables in and see how much the tuning changes, but will try that and see how much it changes. Thanks, Mike ----- Original Message ----- From: Nate Duehr To: [email protected] Sent: Thursday, July 24, 2008 7:07 PM Subject: Re: [Repeater-Builder] Duplexer cables Mike Dietrich wrote: > What is the best way(hookup) to "test" cables like these to see if they > are resonant on a particular freq? > With or without the coupling loops attached etc. for duplexers say The technique I've seen used and participated in is to test the entire section of the duplexer as a unit (with appropriate 50 ohm loads on the other side) using a tracking generator/spectrum analyzer, after getting the individual cans in tune. If connecting multiple cans and testing through them "pulls" the tuning of the cans (there are multiple "humps" in the image shown), the cable lengths are wrong. When the lengths are correct, two cans "add" to make a steeper/taller pass... for example. (It's easier to see this than to describe it in e-mail. Maybe someone will send some screen-shots of what it looks like someday... I've tried taking photos of my IFR 1500 with my cheap digital camera (4MP point-and-shoot Nikon) and it just doesn't look right. Someone with an HP service monitor would have a bigger screen and along with a good camera could probably capture what it looks like better. This technique is a "wholistic" approach to it, I suppose you could call it. You're shooting for the length that ends up looking the best after tuning the individual cans, and locking them down. As long as everything else has proper 50 ohm loads on everything that's not under test. Also always a good idea to use a 6dB 50 ohm pads between the signal generator and the device under test... of course. Normal practice... just in case. (Depending on the test gear and the device being tested, things can get really squirrelly if the test gear is expecting a 50 ohm load and doesn't have one...) It includes the loops this way, of course... because the end-result you're shooting for is that particular duplexer to work at your desired frequency. I hear this process is even nicer to do with a Vector Network Analyzer, but I've never even seen such a beast, let alone used one. Nate WY0X
