I've got a Celwave 6-cavity Bp/Br duplexer, equivalent to the Phelps Dodge PD526. This one actually has a Motorola part number on it: 0185417U05 and it's the same one they use with Quantar and MTR2000 systems. It uses RG400-size coax and looks to be in very good condition; very few scratches. I've got three of the PD units running on UHF ham repeaters (with RG214 coax) and they work great: over 100dB notch depth and about 1.5dB passband loss.
This one has me stumped. I've disconnected all the jumpers and tried tuning each section manually. With the notch tuning screws at the top (when looking into the business end), the LO side is on the left and the HI side is on the right. This is how they're marked and the manuals and Quantar documentation agree with this.
I used a return loss bridge to tune each of the bandpasses properly. For kicks, I'm tuning for 452.000 and 457.000 MHz. The three sections on the LO (left) side tune right up on 452 MHz, and the three on the HI (right) side tune right up on 457 MHz. The tuning screws are fairly close in length, almost 2 inches sticking out.
I then connected the spectrum analyzer and tracking generator (through a pair of 6dB 50 ohm coaxial attenuators) to the leftmost section and tried to tune the reject tuning to 457 MHz, but it just won't go there. I get a nice peak 5 MHz below the passband frequency, but when I turn the reject tuning screw inward, I can see the notch move down and another notch appears above the passband freq, but as soon as it gets within 10 MHz, the passband loss increases to over 10dB because the skirt of the notch is too wide. By this point, the other notch is about 40 MHz below where it was. All three of the LO side cavities do this.
OK, maybe the unit is labeled wrong, so I left the LO side as it was and moved to the HI side.
On the HI side, the pass is 457 and the notch I need is at 452 MHz, and all three tune just fine there, like they should. I did try to tune one with a notch 5 MHz higher than the pass freq and it wouldn't go there either. This leads me to believe that the duplexer is probably marked correctly.
Unfortunately what I seem to have is six cavities set up for notch below passband. Kinda hard to make a duplexer work that way. If this was some sort of multiple frequency filter, it sure doesn't have any markings saying it's special. No "SPnn" in the model number either.
Am I missing something? I've tuned these before and usually the notch can be moved symmetrically above or below the passband frequency on either side; it really doesn't matter how you want to arrange it. This one just won't give me a proper notch above the passband unless I want it 30 MHz above; if the two freqs are that far apart I probably don't need a duplexer. All six sections act the same: passband tuning is fine, reject tuning is only good below the passband frequency and any attempt to get a notch above also kills the passband.
Bob M.
UHF Duplexer Tuning Weirdness
Moderator: Queue Moderator
Re: UHF Duplexer Tuning Weirdness
I'm interested in knowing the answer to this as well. I had a similar issue with the very same duplexer out in the field. I monkeyed with it for two straight days in a row before finally saying that it was indeed "bad", and replacing it. It's the only duplexer I've ever encountered where I think something just plain went wrong.
Re: UHF Duplexer Tuning Weirdness
Could be birdies in your spectrum analyser rcvr at 452.000. I've had a number of service monitors choke on .000 freqs. Take the 6db pads out and try again. Try a different analyser if you have one.
Re: UHF Duplexer Tuning Weirdness
I tried it without the pads. I also tried it with a lower tracking generator output level. No change. Birdies at BOTH 452 and 457? Highly unlikely. I also tried tuning the passband into the 445-450 range; again I could only get a decent notch at a lower frequency. The notch tuning works, it moves the two notches all over the place, but the best notch is below the passband freq, and at 5 MHz lower I get a very nice low-loss passband and about a 35db deep notch. If I start turning the notch tuning inward, the notch moves lower in freq and the upper notch starts moving down towards the passband freq, but the notch width of the upper notch is much too wide, and by the time it comes down to 5 MHz above the passband freq, the passband loss is 10dB and in fact there's less loss at a much lower frequency, but it's about 15 MHz away from the notch I'm trying to achieve.
D119: It's nice to know someone else has seen this. I've tuned probably half a dozen with no problems. This one is just weird.
Bob M.
D119: It's nice to know someone else has seen this. I've tuned probably half a dozen with no problems. This one is just weird.
Bob M.
Re: UHF Duplexer Tuning Weirdness
For S&G's, have you tried tuning it for .025, or back to the original freqs?
Re: UHF Duplexer Tuning Weirdness
Yep. I know how frustrated you are. I tried tuning it in the field with an R2600 w/ tracking generator to no avail, just screwball madness the entire way - extremely lossy passband and miserable notches. I then pulled the thing down to the shop and put it on the bench R2670 which is our "de-facto-standard", and to my surprise, same thing.kcbooboo wrote:I tried it without the pads. I also tried it with a lower tracking generator output level. No change. Birdies at BOTH 452 and 457? Highly unlikely. I also tried tuning the passband into the 445-450 range; again I could only get a decent notch at a lower frequency. The notch tuning works, it moves the two notches all over the place, but the best notch is below the passband freq, and at 5 MHz lower I get a very nice low-loss passband and about a 35db deep notch. If I start turning the notch tuning inward, the notch moves lower in freq and the upper notch starts moving down towards the passband freq, but the notch width of the upper notch is much too wide, and by the time it comes down to 5 MHz above the passband freq, the passband loss is 10dB and in fact there's less loss at a much lower frequency, but it's about 15 MHz away from the notch I'm trying to achieve.
D119: It's nice to know someone else has seen this. I've tuned probably half a dozen with no problems. This one is just weird.
Bob M.
After at least ten hours of screwing with it, and the fact we were a sub for another company back East on this particular job, I told them to send us a replacement (exact same thing arrived), sent them the screwball one back, and checked the tuning on the new one and had no issues with it.
I've tuned all kinds of stuff on VHF, UHF, 800 and 900 and I've never seen anything like this. Even with improper harnessing I've never seen stuff this stupid. All I can chock it up to is a plain old failed duplexer - something went wrong internally.
I too have probably tuned up 4 to 6 of these things in my time and all of them except this one were effortless. We've gotten them to do a few strange TX/RX spacings that were wider than normal without issue.
Good luck! Don't pull your hair out over it - cans are cheap.
Re: UHF Duplexer Tuning Weirdness
The original freqs are unknown. When I got it, I did hook it up to the SA/TG; the LO side had a passband and two notches that were reasonably close (it was originally connected as two sections on the LO side and four sections on the HI side); the HI side had four passbands and eight notches from 300-500 MHz. That's when I disconnected all the jumper cables and looked at each section individually.
While in the process of tuning it (some of the passband tuning slugs had to come out 1/2 inch; that's a lot of turns with a wrench when cables are attached) I could watch the passband tuning move up the screen, and the notch tuning went right along with it. I roughly tuned each section with just the SA/TG, so I wasn't exactly on 452/457, but the signal peak was within 100 kHz. I just wanted to see if the unit worked. I then adjusted the reject tuning and that's where I encountered this weirdness. The notch tuning screws were also all over the place; some had perhaps 4 threads showing, otheres had an inch showing above the jamb nut.
Yes, cans are cheap, but I don't look a gift duplexer in the mouth. This one was as cheap as they come. I don't have the option of calling someone and asking for a replacement. I bet I could get $100 for it on FleaBay. Hey, it says UNTUNED (and it will be again) and it's in great condition. A little wire brush action on the male N connectors, a dip in Tarn-X, and it'll be almost as good as new. Except...
Oh, by the way, these units are NOT water-tight. The rivets have holes in them and some of the corners are not sealed. It should have dried out by now; I'll see if anything has improved in a day or two.
Bob M.
While in the process of tuning it (some of the passband tuning slugs had to come out 1/2 inch; that's a lot of turns with a wrench when cables are attached) I could watch the passband tuning move up the screen, and the notch tuning went right along with it. I roughly tuned each section with just the SA/TG, so I wasn't exactly on 452/457, but the signal peak was within 100 kHz. I just wanted to see if the unit worked. I then adjusted the reject tuning and that's where I encountered this weirdness. The notch tuning screws were also all over the place; some had perhaps 4 threads showing, otheres had an inch showing above the jamb nut.
Yes, cans are cheap, but I don't look a gift duplexer in the mouth. This one was as cheap as they come. I don't have the option of calling someone and asking for a replacement. I bet I could get $100 for it on FleaBay. Hey, it says UNTUNED (and it will be again) and it's in great condition. A little wire brush action on the male N connectors, a dip in Tarn-X, and it'll be almost as good as new. Except...
Oh, by the way, these units are NOT water-tight. The rivets have holes in them and some of the corners are not sealed. It should have dried out by now; I'll see if anything has improved in a day or two.
Bob M.
Re: UHF Duplexer Tuning Weirdness
I'll give you one guess as to why it was probably freekcbooboo wrote:Yes, cans are cheap, but I don't look a gift duplexer in the mouth. This one was as cheap as they come.
I jest. So it had water in it?
Re: UHF Duplexer Tuning Weirdness
No, it was apparently in service until something happened to the site. You don't want to experience what the rest of the equipment looked like or smelled like. The mice called it home for about a year and they had no maid service.
The water was from me hosing it off, then washing it. I didn't submerge it but I definitely had water on every surface as I scrubbed it to get rid of the mouse urine smell. When I picked it up to shake off the water, I heard water sloshing around inside so I inverted the unit and saw a small stream come out of one corner and several rivets.
We're having a warm spell up here this week so whatever water is left inside should evaporate and dry out.
If nothing else, it will make a fancy receiver filter with notches for six transmitters. It just won't act like a duplexer.
Bob M.
The water was from me hosing it off, then washing it. I didn't submerge it but I definitely had water on every surface as I scrubbed it to get rid of the mouse urine smell. When I picked it up to shake off the water, I heard water sloshing around inside so I inverted the unit and saw a small stream come out of one corner and several rivets.
We're having a warm spell up here this week so whatever water is left inside should evaporate and dry out.
If nothing else, it will make a fancy receiver filter with notches for six transmitters. It just won't act like a duplexer.
Bob M.
-
grandnational
- New User
- Posts: 7
- Joined: Sat Feb 20, 2010 2:24 am
- What radios do you own?: Motrac, Mastr Pro, Kokusai, NC
Re: UHF Duplexer Tuning Weirdness
It would seem to me, just from skimming this half-awake, that the loops on one side may either be wrong or damaged somehow. Although their being wrong is unlikely if it saw prior successful service. You may also have a corrosion problem with the notch adjustments and/or the loops affecting continuity in the unit, hence poor Insertion Losses & other tuning anomalies. I had a tuning anomaly earlier this evening with a DB-4076, but tuning the two HI PASS cans individually first allowed me to remedy it. 526s can be a rain pain in the arse. Personally, I think they're great for notch function, but useless as a meaningful bandpass device. In-band response is OK, but out of band is nil. Might as well be a Tee. Still, the "right price" even convinces me! Good luck. I'd try cleaning up any contact surfaces for the jam nuts, notch screws & pass rods just on GP. (The pass rods are Invar & rust very easily)
The 526s are fun to service with their big soldered box design, too. Ugh!
The 526s are fun to service with their big soldered box design, too. Ugh!
Re: UHF Duplexer Tuning Weirdness
I agree that the passband response out-of-band is about as good as a TEE connector. I've never tried to open one up. This one looks great from the outside. I have removed two of the notch tuning slugs and they're pristine inside and out. It's hard to do anything to brass. The nuts can easily be tightened and they can be adjusted throughout the entire tuning slug length. If something's broken inside, I don't know how it got that way unless the unit fell against something pretty hard, but there's no indication of that from the outside; not even a dented corner.
It's hard to explain what the spectrum analyzer trace looks like when adjusting the notch tuning. I'll have to hook it all back up and take several snapshots, then post them somewhere for viewers to see the progression. I do get two notches: one above, one below the passband. With the lower notch at 5 MHz below the passband, the upper is probably 30 MHz above and the passband loss is under 1dB. As I tune the notch slug inward, both notches start moving down in frequency. As this happens, the passband starts rolling off and gets much lossier as the upper notch gets close to it. By the time the upper notch is 5 MHz above the passband, there's 10dB passband loss and the lower notch is 30 MHz below. The notch depth is fine. Above the upper notch the passband loss is back around 1dB. It's as if I'm looking at the lower notch but the passband tuning moved up several MHz, and I know that isn't happening.
I've got all the junction coax cables removed and I'm tuning each section separately. This eliminates any interaction.
The unit has been sitting in my hot porch for the past two weeks so any moisture in it should have dissipated by now.
Bob M.
It's hard to explain what the spectrum analyzer trace looks like when adjusting the notch tuning. I'll have to hook it all back up and take several snapshots, then post them somewhere for viewers to see the progression. I do get two notches: one above, one below the passband. With the lower notch at 5 MHz below the passband, the upper is probably 30 MHz above and the passband loss is under 1dB. As I tune the notch slug inward, both notches start moving down in frequency. As this happens, the passband starts rolling off and gets much lossier as the upper notch gets close to it. By the time the upper notch is 5 MHz above the passband, there's 10dB passband loss and the lower notch is 30 MHz below. The notch depth is fine. Above the upper notch the passband loss is back around 1dB. It's as if I'm looking at the lower notch but the passband tuning moved up several MHz, and I know that isn't happening.
I've got all the junction coax cables removed and I'm tuning each section separately. This eliminates any interaction.
The unit has been sitting in my hot porch for the past two weeks so any moisture in it should have dissipated by now.
Bob M.
Re: UHF Duplexer Tuning Weirdness
Forgive a silly question but: have you tried bouncing between tuning the notch and the pass? Specifically, deliberately displacing both notch and pass way off where you want them, and then walking in? I've had a couple of cases where a cavity was tuned to what mathematicians call a local minimum - a place where it sort-of-works, but not the optimum location. As you move away, things get worse, but they aren't best, because you aren't at the global minimum. When you completely move off, and re-tune, you sometimes find the global minimum and get the correct response.
This is my opinion, not Aeroflex's.
I WILL NOT give you proprietary information. I make too much money to jeopardize my job.
I AM NOT the Service department: You want official info, manuals, service info, parts, calibration, etc., contact Aeroflex directly, please.
I WILL NOT give you proprietary information. I make too much money to jeopardize my job.
I AM NOT the Service department: You want official info, manuals, service info, parts, calibration, etc., contact Aeroflex directly, please.
Re: UHF Duplexer Tuning Weirdness
My first "rough" tune used just the SA and TG and I tuned the passband for a signal peak. Then I tuned the notch for 5 MHz above (I was doing the low side first) and that didn't go well, so I did the two other sections on the low side; same results. I found the notch would only give me a decent (under 1dB) passband loss when 5 MHz below the passband freq. Then I tried the high side; same results. I can get two notches but only the one on the low side gives me proper passband loss.
Then I took out my Eagle-1 return loss bridge and tuned the passband tuning for the lowest indication on the SA. The RLB was connected to one port and a known good 50 ohm load was attached to the other port of the section I was tuning. I found that the sections were rough-tuned within 0.5 MHz of optimum (best return loss, lowest SA indication). There was only one dip (that I saw) in each section, but I will take a wider look incase there are two resonant points, something I've never seen when adjusting other Celwave 526 duplexers.
Bob M.
Then I took out my Eagle-1 return loss bridge and tuned the passband tuning for the lowest indication on the SA. The RLB was connected to one port and a known good 50 ohm load was attached to the other port of the section I was tuning. I found that the sections were rough-tuned within 0.5 MHz of optimum (best return loss, lowest SA indication). There was only one dip (that I saw) in each section, but I will take a wider look incase there are two resonant points, something I've never seen when adjusting other Celwave 526 duplexers.
Bob M.
Re: UHF Duplexer Tuning Weirdness
For the record, I have also seen notch/pass filters with two resonant points...it acted just as you describe and had me chasing my tail for a while. Ultimately I did just what Wowbagger suggested, starting way out & working back in a bit at a time before finally getting it right. It was years ago & I don't recall what make/model it was.
No trees were harmed in the posting of this message...however an extraordinarily large number of electrons were horribly inconvenienced.
Welcome to the /\/\achine.
Welcome to the /\/\achine.
Re: UHF Duplexer Tuning Weirdness
Since the Celwave passband rolloff is rather gradual, it is possible that there's another peak with two notches. I probably had the SA/TG set for a rather narrow width, perhaps 10 MHz, while tuning the notches. It would have been wider while tuning the passband but not much more.
I'll take a look at the passband over the range of 0 to 1000 MHz and see how many peaks there are, but the passband tuning certainly seemed to be working properly and tuned smoothly.
Bob M.
I'll take a look at the passband over the range of 0 to 1000 MHz and see how many peaks there are, but the passband tuning certainly seemed to be working properly and tuned smoothly.
Bob M.
Re: UHF Duplexer Tuning Weirdness
I connected the SA/TG to the duplexer's first LOW section and swept it from 0 to 1000 MHz. There was just one peak near 455 MHz and two notches, but the upper one was a lot closer to the passband. I zoomed in and found 0.5dB passband loss and a notch at around 460 MHz. I retuned the passband for 452 and the notch was almost 40dB at 457, right where it should be. It was also symmetrical, in other words I could get a nice notch at 447 and still have 0.5dB passband loss. I retuned the remaining LOW sections and they all worked just the way I expected them to. The HIGH sections were already working properly (notch on the low side) but I retuned them to pass 457 and notch 452. I did not bother tuning the notch to 462 and checking for acceptable passband loss; I suspect it would have.
I then retuned the passband tuning using the return loss bridge. They were all within 1 MHz but I got them dead on. I then repeated the notch tuning and got those dead on.
Final results with everything cabled together and the unused port terminated:
LO passband loss: 1.5dB
LO reject loss: 110dB
HI passband loss: 1.6dB
HI reject loss: 114dB
When I checked from LO to HI with the ANT port terminated, I get 104dB of notch depth and it's nice and wide (as in more than one hundred kHz).
So to end this long saga, apparently moisture/water inside the sections can do weird things to the notch tuning. The corners are NOT waterproof and the pop-rivets used to assemble the connectors and notch assembly are hollow, thus letting water in. After two weeks sitting in my hot porch, everything evaporated and it's working just fine. I will probably remove the jumper cables and soak the connectors in Tarn-X and use a wire brush on them to clean them up a bit.
Thanks to all who cared to respond or comment.
Bob M.
I then retuned the passband tuning using the return loss bridge. They were all within 1 MHz but I got them dead on. I then repeated the notch tuning and got those dead on.
Final results with everything cabled together and the unused port terminated:
LO passband loss: 1.5dB
LO reject loss: 110dB
HI passband loss: 1.6dB
HI reject loss: 114dB
When I checked from LO to HI with the ANT port terminated, I get 104dB of notch depth and it's nice and wide (as in more than one hundred kHz).
So to end this long saga, apparently moisture/water inside the sections can do weird things to the notch tuning. The corners are NOT waterproof and the pop-rivets used to assemble the connectors and notch assembly are hollow, thus letting water in. After two weeks sitting in my hot porch, everything evaporated and it's working just fine. I will probably remove the jumper cables and soak the connectors in Tarn-X and use a wire brush on them to clean them up a bit.
Thanks to all who cared to respond or comment.
Bob M.
-
chartofmaryland
- Batboard $upporter
- Posts: 411
- Joined: Sat Dec 28, 2002 11:25 pm
- What radios do you own?: Alot
Re: UHF Duplexer Tuning Weirdness
FYI
DO NOT WASH A DUPLEXER, EVER, UNLESS YOU DO NOT WANT IT TO WORK AGAIN
Also, getting any characteristic 50 ohm port wet will change the 50 ohms, and may cause it to not return should the dielectric be compromised.
Had a chief buggy with a mobile duplexer in the middle seat get hosed out after a nasty fire.
The duplexer and cabling did not survive.
Most decend quality duplexers and the like are silver plated internally and will corrode should they get wet.
The most I have put a piece of equipment through after a mouse piss-a-thon was Super Clean on a Rag followed by a diluted bleach soaked rag.
You guys make me laugh sometimes, glad to hear the duplexer is working again . . . .
CoM
DO NOT WASH A DUPLEXER, EVER, UNLESS YOU DO NOT WANT IT TO WORK AGAIN
Also, getting any characteristic 50 ohm port wet will change the 50 ohms, and may cause it to not return should the dielectric be compromised.
Had a chief buggy with a mobile duplexer in the middle seat get hosed out after a nasty fire.
The duplexer and cabling did not survive.
Most decend quality duplexers and the like are silver plated internally and will corrode should they get wet.
The most I have put a piece of equipment through after a mouse piss-a-thon was Super Clean on a Rag followed by a diluted bleach soaked rag.
You guys make me laugh sometimes, glad to hear the duplexer is working again . . . .
CoM
If the lights are out when you leave the station and then come on the second you key up, you know you have enough power.
Re: UHF Duplexer Tuning Weirdness
And you made me laugh pretty hard with that one. So hard in fact, you made my sig line along with wazzzzup.chartofmaryland wrote:FYI
The most I have put a piece of equipment through after a mouse piss-a-thon was Super Clean on a Rag followed by a diluted bleach soaked rag.
You guys make me laugh sometimes, glad to hear the duplexer is working again . . . .
CoM
Re: UHF Duplexer Tuning Weirdness
None of the other pieces of equipment around the duplexer suffered from their hot water bath. Silver will tarnish just sitting in air; water won't accelerate the process unless it sits there for a long time. Thorough drying is the key.
With all the mouse excrement, nest material, fur, etc. IN and ON this equipment, I didn't even touch it until the outdoor hose sprayed water on everything. I then took each piece under the sink and scrubbed them. I can still smell the urine almost 6 weeks later.
An Argus battery charger did not survive, nor did the power supply in the station that was connected to the Argus. One rather wide foil must have been bubbling under the urine and now does not make contact with the battery terminals. Other than completely disassembling it and pulling nest material out from between every heatsink fin and inside every toroid, the unit has not been plugged in or tested. If I could get a schematic and board X-ray view I might try to fix it, but I think this particular unit really needs to be thrown away. If I had unlimited funds I could send it to Argus and they would probably condemn it for me.
Bob M.
With all the mouse excrement, nest material, fur, etc. IN and ON this equipment, I didn't even touch it until the outdoor hose sprayed water on everything. I then took each piece under the sink and scrubbed them. I can still smell the urine almost 6 weeks later.
An Argus battery charger did not survive, nor did the power supply in the station that was connected to the Argus. One rather wide foil must have been bubbling under the urine and now does not make contact with the battery terminals. Other than completely disassembling it and pulling nest material out from between every heatsink fin and inside every toroid, the unit has not been plugged in or tested. If I could get a schematic and board X-ray view I might try to fix it, but I think this particular unit really needs to be thrown away. If I had unlimited funds I could send it to Argus and they would probably condemn it for me.
Bob M.