Thanks to everyone for their responses. All have been helpful, but the one Allan mentions, measuring the XL of the loops and using that to calculate the amount of coax it "replaces" is the one that best fits the sort of answer I was looking for.
While the empirical "trial and error" method is sure to get results - given enough time and materials - it's great to augment it with an understanding of the theory behind the results. SO what I'll do is use the VNA to measure the loop's XL, calculate a starting point for coax lengths based upon it, and make a few other cables which are slightly different in length. Then I'll see what works best. Thanks guys! I needed a little coaching on the theory end. FYI the cavities are older, inexpensive DB4001s that I converted to Bp-Br using a coil and capacitor between the two connectors. I thank those who have contributed to the knowledge base on the Repeater Builder web site for the information describing cavity duplexers. Once I get the cabling perfected I'm confident that they will work very well. Even now, with the cables a little off, they're working well enough (with three cans per side) that the RX desense is quite small, only a dB or so. But, as I've said, that's after carefully retweaking the cans as a set, even though they had been optimized individually on the VNA before connecting them together. Once I get the cables just right and I squeeze out the last few dB of isolation I think the old MastrPro receiver will be completely free of desense even when running the transmitter at its rated 80W output. When I get it finished I'll try to report back on what I found. 73 Brad KB9BPF --- In [email protected], allan crites <[EMAIL PROTECTED]> wrote: > > Or one could measure the XL of the loop assembly ( off the resonant frequency of the tuned cavity), plot the measured results on a Smith Chart, determine the WL towards .25 WTG on the circumference circle, and subtract that length in the coax dilectric, from the 1/4 WL coax cable. > > Allan Crites WA9ZZU > > --- On Tue, 9/16/08, Jeff DePolo <[EMAIL PROTECTED]> wrote: > > From: Jeff DePolo <[EMAIL PROTECTED]> > Subject: RE: [Repeater-Builder] Re: Cavity interconnection cable length > To: [email protected] > Date: Tuesday, September 16, 2008, 11:06 AM > > > > > > > > By actual length do you mean the length of the entire conductor, the > > circumference of the loop? That could be a couple inches on VHF > > cavities. > > I think some of the comments made in this thread are misleading. > > The coupling element (loop, probe, whatever) adds *some* electrical length, > but it's not correct to say that the actual physical length of the element > is what you should use to determine how much to subtract off a true > quarter-wave (or half-wave or whatever) length of cable. How much should > you subtract? Very difficult to say. The amount may even change depending > on the coupling angle (i.e. loop orientation) . It is for this reason that > manufacturers, and guys like me out in the field, have an assortment of test > cables that we use to find the ideal length. We also use line stretchers > and "connector savors" (short male to female adapters) to find the ideal > length before cutting the final cable. > > > My supply of RG-400 and silver-plated UHF connectors (my old DB cans > > use UHF connectors) is running a little thin and I don't have enough > > to make a lot of trials-and-errors. > > You can cut the cable a little long, put a little dab of solder the center > pin, and just slide the ferrule over the braid without crimping. That way > you can re-use the connector until you find the right length, then do the > final crimp on the center and ferrule. You can do the same thing with > old-fashioned PL-259's and UG-176 reducers too, just soldering the center. > > > Although we don't have a UHF-connector calibration standard and I > > haven't seen any available even if I could convince the boss they're > > needed, what I did was I made a set of N-to-N cables, which I can > > calibrate, and a set of N-to-PL259 cables the exact same length and > > cable material which I use to set the cans. I'm counting on the > > small differences caused by impedance bumps and change of velocity > > factor in the PL259 won't be critical on VHF. See any big flaws in > > this approach? > > Good adapters with Teflon dielectric will have the same velocity factor as > RG400, so the resulting cable length should come out the same. > > There will be some trial and error involved unless you have a line > stretcher, there's no easy way around it. One technique, if you don't have > a line stretcher, or a whole lot of test cables cut to incremental lengths, > is to start with a piece of cable that's a little short and then add > high-quality (very important) connector savers or elbows until you find the > right length. If you have a VNA, measure the electrical length (delay) of > the resulting cable (with the adapters) and then cut a new cable of the > right length. > > --- Jeff WN3A >
