To add to what Bob said; Band pass/reject type duplexers are basically notch type duplexers with the loop tuned to form a little bit of band pass effect. The cavity itself tunes the notch and the tuned loop forms the pass band part. Most of the pass band rejection will fall between the two frequencies of the duplexer. Part of the pass band rejection comes from the wide skirt of the notch of each set of cavities i.e. the overlap of the high and low notch skirts. Above and below those frequencies there is little rejection, sometimes only 10 or 15 db. Think of the pass band above or below the notch frequencies as what you would get by using a tuned coax stub for a pass band filter, it will be pretty broad.
On UHF with 5 MHz spacing you can make a duplexer using all pass cavities, although it will take more cavities than using notch cavities, but it will give you a superb duplexer with a nice narrow receive window and you will get very little response above and below the duplex frequencies. Pass cavities do have some spurious responses out away from their pass frequency but they are usually pretty far down. A pass cavity duplexer will have a little more insertion loss than a notch type duplexer too. You can learn a lot about how much rejection and how wide a pass or notch really is by looking at the curves of cavities or duplexers on the web sites of the various manufacturers. Look at the frequency verses loss and you will see just how much, or how little, the rejection is as you move away from the tuned frequency. With a pass cavity they will usually show the curve with 3 different insertion losses set so you can see the difference. Some will also show the result of 2 and 3 cavities in series. 73 Gary K4FMX > -----Original Message----- > From: [email protected] [mailto:Repeater- > [EMAIL PROTECTED] On Behalf Of Bob M. > Sent: Thursday, May 31, 2007 8:13 PM > To: [email protected] > Subject: Re: [Repeater-Builder] Duplexer tuning - sanity check > > Some duplexers pass the desired frequency, lose > something on each side, but then pass stuff from DC to > over 1GHz as if it's not even in the circuit. My > Celwave 526 operates that way. The bandpass tuning is > rather broad and serves to match the impedance more > than effect the bandpass of the signal. The notch > tuning is very deep (40dB per section). It will keep > the transmit signal out of the receiver very nicely. > > If you take an isolated bandpass can and sweep it, > you'll probably see low loss at one frequency and high > loss everywhere else. Duplexers don't always respond > this way. Your mileage will vary depending on the > design. > > The duplexers on UHF are typically set for 5 MHz > spacing, although closer spacing is often available. > The purpose of the duplexer is to allow the receiver > and transmitter to share one antenna. They offer very > little filtering of other nearby signals. You need a > good, selective receiver front end to handle that. The > transmitter can often be cleaned up by adding an > isolator to the output; some units may already have > that feature, especially commercial base stations. > > You always tune your duplexer to your two frequencies; > other nearby repeaters are not a concern when tuning > the duplexer. It's just there to allow your TX and RX > signals to share the coax/antenna. If you have nearby > stations that you need to filter out, you must add > your own very narrow filters external to the duplexer. > Usually these will be notch filters or even shorted > stubs. A bandpass filter is rarely narrow enough to > reduce interference from signals less than several > hundred kiloHertz away. It will have no effect on the > signals from a repeater 16kHz away. The tuning of the > duplexer just isn't that tight, and in fact even the > best duplexer will allow a lot more leeway in the TX > and RX frequencies without requiring retuning. > > Bob M. > ====== > --- w6nct <[EMAIL PROTECTED]> wrote: > > > I'm tuning up a cavity-type duplexer for a 70cm > > repeater, and in the > > process, I showed my results to a fellow ham. He > > asked a couple > > questions that cause me to think; so I decided to > > ask the experts... > > > > FIRST QUESTION TOPIC: > > > > For reference, my primary cavities were made by > > Tx-Rx Systems, and > > have both bandpass and band-reject tuning controls > > on each cavity. > > When tuning them (either singularly or as pairs > > in-series), I see and > > can adjust the band-pass and notch for the desired > > Tx and Rx frequencies. > > > > However, in the broad-band sweep, I can also see a > > bunch of other > > signals passing through the cavities; all of which > > are well away from > > my Tx/Rx frequencies. I suspect that these are > > normal, and are a side > > effect of how the can-type resonant cavities work. > > Am I correct in > > this assumption? > > > > As a sanity check, I combined this set of cavities > > with another > > (Phelps-Dodge) set I have, and tuned these > > supplemental cavities for > > band-pass only (one cavity for my Tx frequency, and > > one cavity for my > > Rx frequency). Sure enough when I put these in > > series with my primary > > set, I can eliminate nearly all of the "other > > signals" from the > > broad-band sweep. This observation seems to > > reinforce my initial > > assumption about the Tx-Rx cavities. Do you agree? > > > > > > SECOND QUESTION TOPIC: > > > > The other ham thought that I should end up with a > > band-pass that is > > narrow enough to eliminate adjacent repeaters (at > > 16kHz spacing, as > > per the current SCRRBA band-plan separation). I > > tried but I cannot > > get either set of cavities to have that narrow of a > > band-pass; at > > least not without sacraficing most of the signal in > > the process. I > > suspect that the receiver and transmitter need to > > actually inforce > > these much narrower bandwidth requirements within > > the broader > > protection provided by the cavity-duplexer. I > > suspect that I should > > focus my duplexer tuning on passing the desired > > frequency, notching > > the alternate repeater frequency, and trying to do > > so with the least > > amount of signal attenuation. Am I correct in these > > understandings? > > > > For both of these question topics, feel free to > > point out anything > > that I might be missing or misunderstanding. I'm by > > no means > > sensative about this stuff, and still consider > > myself on the learning > > curve about duplexers and repeaters in general. > > > > Thank-you (in advance) for your time, thought, and > > opinions. > > > > <<< vern >>> > > > > __________________________________________________________________________ > __________ > Pinpoint customers who are looking for what you sell. > http://searchmarketing.yahoo.com/ > > > > > > Yahoo! Groups Links > > >
