Tuning Telewave 1444C duplexer for 2m ham with IFR1500

[WX5812]

I am not experienced with the IFR1500 as of yet (hope to be one day). Is there anywhere online or can be explained the exact settings and steps to take to tune this duplexer with this analyzer?
The specs on the duplexers show they can handle .600khz of spacing.
Any guidance would be appreciated

 

[mmckenna]

The Youtubes are your best bet.

I sat down with an experienced tech and did one. A few years later I had to do a couple of UHF duplexers and found a Youtube video to refresh my memory. Once I started it made sense.

 

[AK9R]

You'll find lots of good information on the Repeater Builder website including duplexer tuning instructions from Telewave:

http://www.repeater-builder.com/antenna/telewave/telewave-duplexer-tuning.pdf

Amongst the YouTubers, there's a guy named James Eagleson who I think is a retired Telewave engineer. He has some useful videos on duplexer tuning.

 

[prcguy]

What frequency was the duplexer originally delivered on? If it was a very low 150Mhz or close to the bottom edge of 2m it should tune up fine, if it was in the high 150-160MHz range you might have to replace the interconnect cables. Things can go smoother on many duplexers if you remove one can at a time and tune it individually for pass/reject then do another, etc. When the cables are all hooked back up it will be very close and sometimes right on.

 

[cmdrwill]

Exactly what
prcguy said.

 

[wa8pyr]

Exactly what
prcguy said.

With one exception. . . once you get them all connected back together there can be some detuning effects, so you'll want to run through a quick alignment with everything connected together to make sure everything is OK.

 

[prcguy]

Yes, of course, but some duplexers like the UHF PD/Celwave 526 series will usually meet full specs after carefully tuning each individual cavity and just connecting them all back up. Another thing on the Telewave 1444, its only got about 70dB isolation when everything is perfect and that's not going to be adequate for some repeaters like a couple of wide band mobiles cobbled together as a repeater with a 150 watt amplifier attached. A good repeater with sharp tunable front end and 50 watts should be ok.

It will also be a struggle balancing insertion loss and notch depth at 600KHz with that one and an IFR1500 might get you there or it might not. Its nice to have a good spectrum analyzer with tracking generator and return loss bridge or a scalar network analyzer to get the most out of tuning. A return loss bridge will let you know if the cable lengths are correct and if not you can see it with dual return loss humps that you can't tune out instead of one deep one at the pass frequencies. Its also a good idea to calibrate your analyzer with 6dB attenuator pads at the ends of both test cables so the attenuator is what interfaces with the duplexer. That will insure a good 50 ohm match right at the duplexer so the tuning is less cable dependent.

With one exception. . . once you get them all connected back together there can be some detuning effects, so you'll want to run through a quick alignment with everything connected together to make sure everything is OK.

 

[vagrant]

Is there a particular reason for 6dB instead of say 1, 3, or 10 dB pads?

Its also a good idea to calibrate your analyzer with 6dB attenuator pads at the ends of both test cables so the attenuator is what interfaces with the duplexer. That will insure a good 50 ohm match right at the duplexer so the tuning is less cable dependent.

 

[prcguy]

6dB is a good compromise giving an instant 12dB return loss on top of whatever the cables and equipment already give. Its also 12dB insertion loss for two cables and attenuators which is easily made up for using a spec analyzer w/tracking gen or scalar analyzer but it could be too much for some simpler setups.

10dB pads would be a better match but your generator and measurement range will be reduced by that much. 3dB is better than nothing but it only gives you 6dB return loss if your cables connected to your equipment are questionable. 6dB pads on each test cable was what I was taught when entering the aerospace industry as a microwave tech a long time ago and it seems to be what everyone in the know uses.

Other "in the know" people have 1/2 wavelength test cables adjusted with velocity factor made up for various bands and that's an excellent alternative.

Is there a particular reason for 6dB instead of say 1, 3, or 10 dB pads?

 

[prcguy]

As a follow up comment, I've used an IFR1500 for tuning cavity filters and other things but its been too long ago and I can't remember the exact setup. Once you find it I believe it will involve sweeping out of the transceiver or duplex port and measuring into the antenna port.

I don't remember any specific amplitude calibration routine on the 1500 and its a good idea to connect your two test cables together while sweeping and adjust the horizontal position knob so the flat sweep is at 0dB or the top reference line on the analyzer screen, then use 1dB/div when measuring the duplexer or single cavity insertion loss. At 10dB/div and a sweep range of maybe 1MHz you will see the pass hump and notch but it will be hard to read the exact notch depth on the tiny screen and this is where a dedicated spectrum analyzer or scalar analyzer is your best friend.

I usually generate into the antenna port of the duplexer and measure out of the transmit and receive ports and don't forget to put a good 50 ohm load on the unused port while measuring.

 

[wa8pyr]

Yes, of course, but some duplexers like the UHF PD/Celwave 526 series will usually meet full specs after carefully tuning each individual cavity and just connecting them all back up.

I've never been lucky enough to get a set of cans to cooperate that well, on any duplexer (including a 526). No matter how carefully I tune each can individually, there's always some improvement to be found once you get all the cans connected together. A lot of duplexers I've worked with worked fine doing it the way you describe (especially if I was out at a site and had to tune "by ear" in a pinch), but I'm kind of picky about duplexer tuning.

Another thing on the Telewave 1444, its only got about 70dB isolation when everything is perfect and that's not going to be adequate for some repeaters like a couple of wide band mobiles cobbled together as a repeater with a 150 watt amplifier attached. A good repeater with sharp tunable front end and 50 watts should be ok.

Concur on the limited isolation. The FD I used to work for had an MSR2000, with input and output frequencies only about 1.5 MHz apart, and the PA running full-bore 120 watts. Duplexer was rated for 1 MHz separation but only about 70dB isolation, and we had awful problems with desense; when the incoming carrier dropped the repeater stayed keyed for several seconds making the most obnoxious crackling noise due to the TX getting into the RX. Drove me nuts, as I knew exactly what was going on. We had to put up with it until I was finally able to convince the nitwits from the shop we did business with to crank the power down by about 30 watts (fire chief wouldn't let me do it). We noticed no reduction in coverage and the problem went away.

I usually generate into the antenna port of the duplexer and measure out of the transmit and receive ports and don't forget to put a good 50 ohm load on the unused port while measuring.

Ditto, especially on the 50 ohm load. I forgot that once and the wonky readings I was getting drove me batty until I took another look at the connections, and there it was right in front of my face: the TX port just sitting there all nice and empty. D'oh!

 

[prcguy]

On your next 526 try the attenuators on the end of the test cables and tune each can individually. I've been pleasantly surprised the last few times I've tuned one and I couldn't improve it any further tuning as a complete unit. In fact that's the procedure called out in the Motorola instructions for a 526 series on a Quantar repeater. BTW on duplexers capable of 120dB isolation its a real pain to see the notches and I often have to use an amplifier at the output of the sweep gen or sometimes add a preamp to the measure input just for the notch adjustments because they can be buried in the analyzer noise floor.

I've never been lucky enough to get a set of cans to cooperate that well, on any duplexer (including a 526). No matter how carefully I tune each can individually, there's always some improvement to be found once you get all the cans connected together. A lot of duplexers I've worked with worked fine doing it the way you describe (especially if I was out at a site and had to tune "by ear" in a pinch), but I'm kind of picky about duplexer tuning.

 

[AK9R]

On your next 526 try the attenuators on the end of the test cables and tune each can individually.

Which end? SA/TG end or the duplexer end of the cables?

BTW on duplexers capable of 120dB isolation its a real pain to see the notches and I often have to use an amplifier at the output of the sweep gen or sometimes add a preamp to the measure input just for the notch adjustments because they can be buried in the analyzer noise floor.

Isn't that where the return loss bridge comes in so you can, in essence, "flip" the notch and see it better?

 

[prcguy]

On the duplexer end so the cavity sees a good 50 ohm match. A notch will be a high return loss when viewed with a return loss bridge but I usually tune the notches in the through loss mode. A return loss bridge is good for making sure the whole assembly with cables plays nice and that the cables are the right length. Not a problem when tuning close to the factory made frequency but absolutely necessary when using a surplus commercial freq unit on amateur. Some of those will not tune up because the cables are the wrong length and the resulting return loss is high and the insertion loss will be higher and the notches get funky, etc.

Which end? SA/TG end or the duplexer end of the cables?

Isn't that where the return loss bridge comes in so you can, in essence, "flip" the notch and see it better?

 

[WX5812]

Ok I believe I’m making progress. This is the tx side of the duplexer at 145.250. Does it look as it should?

 

[prcguy]

Sort of. If the IFR has different generator output levels you want the highest one possible to get the best dynamic range. If that's the band pass in the middle you don't have the analyzer calibrated to anything and you can't tell how much insertion loss there is. If that's the notch for the receive frequency off to the left see how it flattens out into the noise floor? You need to be able to see the ultimate notch depth to place it accurately on the receive frequency and you will need more generator power or less attenuation on the receiver side if there is an option for that. I see about 65dB of on screen dynamic range and you need more. More more more!!!

Ok I believe I’m making progress. This is the tx side of the duplexer at 145.250. Does it look as it should?

 

[WX5812]

Ok thanks! I will do some more trial and error today and see what I can come up with.