Hi Jeff, If your duplexer presents 50+j0 at its input at the operating frequency and you are using 50 ohm cable to connect it to the transmitter then the transmitter is always going to see 50+j0 at the operating frequency no matter what the cable length is. But at some off frequency that is not 50+j0 that impedance is going to get transformed into something yet again by the time the cable reaches the transmitter. And depending on how long or short you make that cable will depend on what transformation it will make to that off frequency impedance all the while still delivering 50+j0 at the operating frequency. So cable length can be used to change unwanted reactance of off frequency things.
Even if the duplexer is not a perfect 50 ohms, changing cable lengths can have enough of an effect on off frequency things to make the system work sometimes. Transistors are strange animals. The do not equate to tubes in there operation especially when broad band combining devices are used at the output of the transistor amplifiers as most do. The devices themselves do strange things in the presents of reactance. That's mostly why many amplifiers have swr shutdown circuits or power roll back. The problems are generally not one of excess current because of reflected power but the excess current the devices draw is from the unwanted reactance that they see. Sometimes a duplexer just doesn't present a perfect 50 ohms to the transmitter either. Some transmitters will start cutting back the power with little reflected power. Others may not be able to put out all the power it is supposed to if it does not see a perfect 50 ohms. These are some of the reasons most duplexer manufacturers recommend trying different cable lengths between the duplexer and transmitter. All transmitters are not created equal. I do agree with you that gross problems that show up are signs of problems that need to be fixed other than changing cables. 73 Gary K4FMX > -----Original Message----- > From: [email protected] [mailto:Repeater- > [EMAIL PROTECTED] On Behalf Of Jeff DePolo > Sent: Sunday, July 01, 2007 11:36 AM > To: [email protected] > Subject: RE: [Repeater-Builder] Cable formula > > > The other ways to correct the problem, other than using a different > > transmitter that is not bothered by reactive loads as much, > > is to use a Z > > match or try different length cables that make the > > transmitter happier. > > But if the transmitter is bothered by the bad Z at frequencies outside of > the pass, any matching device you put in line is only going to throw off > the > Z at the pass frequency. You can't have a duplexer that presents 50+j0 at > the pass frequency, and then add a matching device between it and the PA > and > still have 50+j0 at the transmitter at the pass frequency. Sure, you can > give the PA 50+j0 at some other off-channel frequency by doing this, but > at > the expense of messing up the pass performance, this seems like it's only > creating new problems. > > > Both > > of these transform the impedance/reactance presented by the > > duplexer to > > something more palatable to the transmitter and allow it to > > produce the > > power intended. > > If you have a PA that is happy with a good load at the carrier frequency, > but isn't happy with the Z at other frequencies where it shouldn't be > making > power, I'd have to say you have a problem with the amplifier that needs to > be fixed. The antenna system shouldn't have to be the cure for the > amplifier's ills. I've never had a good PA (emphasis on good) that got > unstable if it was looking into a matched load at the carrier frequency > regardless of what was going on off-channel. > > > There is no sin involved in using different length cables to > > make the system > > work properly. It is not a band-aid approach to mask other > > problems. > > Well, I guess we'll have to agree to disagree. > > > The > > real problem is that some transmitters, because of the way they are > > designed, do not like reactive loads. > > They should like a good load at the carrier frequency, and be tolerant of > strange load Z's off-channel. A PA that has a tendency to "run away" just > because the load Z at frequencies well removed from carrier isn't perfect > is > an accident waiting to happen. Even if you mess with cable lengths and > Z-matchers and walk away from the site with the PA running clean, there's > a > good chance the next time the antenna ices up or the HVAC fails that > you'll > be getting calls about spurs coming from your box. > > If I have a PA go spurious, for ANY reason, whether during commissioning > or > sometime during operation, it comes out of service until it can be fixed. > That's just me. > > > Close spaced duplexers will be the worst with > > off frequency > > reactance as the impedance has to change quickly as you move > > away from the > > wanted frequency in order for the duplexer to do its job. > > Mostly agree, except for notch-only duplexers with tend to have a fairly > wide range of tolerable load Z outside the notch, and notch duplexers are > fairly common in close-spaced situations. > > --- Jeff > > > > > > > Yahoo! Groups Links > > >
