UHF MSF won't transmit

[motorola_otaku]

I have a UHF station, C74CLB that I upgraded to CXB with a SSCB and interconnect board robbed from an 800 station (pictures in the last post of this thread.) Everything powers up fine, both VCOs lock and it's not throwing any error codes at me, but I cannot get the transmitter to key, even with the power control pot cranked all the way back. I have swapped the TX VCO, the UHF uniboard, the IPA, and the PA with good spares, and still cannot get transmit. I have the cans to convert it to internal duplex, but it came to me as an externally-duplexed station with the IPA output run straight through to the PA. I tested it both in that configuration with the PA output run straight into a service monitor and with the transmit pre and postfilters installed and tuned to the operating frequency; it would not transmit either way. And yes, the SSCB was programmed with the CONV.DEF file to make it a UHF station and the entire station was fully aligned by the book before I started testing transmit.

Well, today I tried keying it with no PA attached, using the method outlined in this thread but with a 10K pot in place of the 2.2K resistor. With 1.8~V fed to pin 4/the green wire and the power control pot cranked all the way back, I can get the station to key and all the lights on front to light up normally... but I am not getting any power out of the IPA. None. At all. Furthermore, after several key-up attempts C451 on the IPA popped. I popped the lid off and saw that I had forgotten to connect the RCA-terminated cable coming off the uniboard to the IPA. D'oh. Luckily I had another spare IPA, but even with it installed and connected properly I still am not seeing any power out of the IPA. As far as I know the station was operating properly in its factory CLB configuration, but it came to me without the CLB controller so I couldn't test it before starting the conversion. But like I previously mentioned, everything in the transmit "power train" has been swapped all to no avail.

I should also mention that I am operating it with no TTRC and a secure board installed. All the jumpers have been properly installed, no errors are present, and it takes programming just fine. SSCB firmware is 5.42.

I'm at the end of my rope here, guys, and about ready to throw the $%^& thing out the window and plug the GR500 back in.

[Doug]

otaku...
The first thing that comes to mind is check your fuses on the power supply distribution board. I had one do the same thing to me and it turned out being a 4 amp fuse on the board. There's 3 or 4 of them up there behind the PA. Let me know if my hunch was right
Doug

[motorola_otaku]

No such luck.. all 4 fuses on the power supply are good. Guess I'll keep plugging away at it and hope my last IPA doesn't pop.

[motorola_otaku]

Okay, I swapped the CXB interconnect board with a CLB interconnect, and modified the SSCB to supply 5V to pin 9 of J801. Now I am able to vary the input voltage to pin 4 of the PA control lead between 1.5 and 2.5V and subsequently vary the setting of the station power control pot while being able to key the station and hold it keyed.. but I am still seeing not so much as a milliwatt out of the IPA. I almost wonder if there's an issue with this particular SSCB since it's been the only constant so far.

[kcbooboo]

The Uniboard does all the work of controlling the output power. It receives a signal from the control tray (SCB or SSCB) that basically tells it when to transmit. Signals from the power control circuit goes back to the control tray so it can illuminate its front panel LEDs.

The power control circuit does need feedback from the PA telling it that RF power is being made. This comes in via the 6-pin Molex connector. On some stations this cable also plugs into the interconnect board under the RF tray; on other stations it's hard-wired to the interconnect board. It goes through some resistor network there before making its way back up to the Uniboard. This feedback is balanced by the power adjust pot to ultimately set the output power of the station. It responds very fast.

If there's no feedback, the IPA/PA might key up for a few milliseconds before shutting down. You might see the PA KEY and PA ON leds blink on briefly.

I've had problems with the interconnect male-to-male filter assemblies that bring signals between the Uniboard and the interconnect board under the RF tray. Crud builds up around the base of the pin and provides a low resistance path to ground. This will totally screw things up and make the station do weird things. One of my 900 MHz MSFs had this and it would only make half power when it would stay keyed. Everything seemed to be fine with an ohm-meter until I pulled the filter assembly out and actually measured the pin to ground; 10k, but it was obviously lower when more voltage was across it. The crud acted like a semi-conductor rather than a fixed resistor. I had to scrape the stuff off the pin and get the resistance up to infinity. Everything worked fine after that. The filters are just feedthrough capacitors with long pins on them, and they solder the pins to both ends of the feedthrough caps. There's more info about this problem on Repeater-Builder, in the Motorola section MSF&PURC area, in the "900 MHz Experience" article, but no one reads stuff anyway, so it's a moot point to even mention it.

Bob M.

[motorola_otaku]

..but no one reads stuff anyway, so it's a moot point to even mention it.

Oh, I wouldn't say that. I read the MSR pages and linked conversion articles repeatedly for another project I have going on one of those guys, but this seemed like such an oddball beat-your-head-against-the-wall problem that I didn't think it was common enough to have been documented. Chalk it up to the naivety of youth if you want. ;p

In any event, I pulled the interconnect-uniboard pass-throughs, and what I'm seeing is a black substance around the base of each pin that won't scrape off with a fingernail. Is this what you're referring to? I had noticed it before and written it off to a soldering or coating byproduct.

[kcbooboo]

That's the stuff. Measure the resistance of each pin to the mounting plate with a good digital meter. You should get many megohms or infinity from each pin to ground. Anything that you can measure repeatedly means the gunk is conducting. I used a small pen-knife to scrape it off, but I suspect some other solvent will also work. I think this is some kind of growth on the flux that wasn't properly washed away during manufacture. Clean the soldered joints (where the pins meet the feedthru caps) with flux remover or alcohol and a toothbrush, or even a soft brass brush. Check the affected pins with the meter and put it back in once they all read 20 megohms or higher.

It would also be useful to note which pins you find with low resistance, and trace them back to the signals they're interfering with. That might give you a clue to the problems and at least a good explanation if it helps the situation after being cleaned.

Bob M.

[motorola_otaku]

Well, I did that, and verified infinite resistance from each pin to ground.. and still have no transmit or power out of the IPA. I'm beginning to think this station is cursed.

[George]

[quote="motorola_otaku"]Okay, I swapped the CXB interconnect board with a CLB interconnect, and modified the SSCB to supply 5V to pin 9 of J801. Now I am able to vary the input voltage to pin 4 of the PA control lead between 1.5 and 2.5V and subsequently vary the setting of the station power control pot while being able to key the station and hold it keyed.. but I am still seeing not so much as a milliwatt out of the IPA. I almost wonder if there's an issue with this particular SSCB since it's been the only constant so far.[/quote]

Have you tried putting it into test mode and seeing if you get anything at all?

[kcbooboo]

You may just have to troubleshoot it step-by-step.

Make sure there's output from the TX VCO right at the RCA connector (actually at the spot that it connects to the Uniboard).

Make sure there's RF at the input to the IPA on its RCA plug. I'm pretty sure this should be active all the time.

Key the station using the front panel Xmit switch. Check the voltages on the IPA. There should be two or three pins (if yours has the RF module) that have 6-8VDC on them, or more. Don't put a meter on either the input or output RF pin.

There are a few signals that feed the IPA: one is a TX Inhibit signal that shuts down the first couple of stages in the IPA to keep the VCO signal from leaking out to the PA. A second signal is the control signal that comes from the power control circuit on the Uniboard. I think the Uniboard pulls the base of a PNP transistor towards ground through a resistor; the further it goes towards ground, the more voltage/current is applied to the IPA and the more output power you should have. Then there's probably 9.6V that comes directly from the 9.6V regulator on the same board. There are several bypass caps on the power supply lines; if one of those popped, you might have burned a foil on the IPA/REG or even the Uniboard. Someone contacted me with such a problem; a surface-mount tantalum on the IPA went up in flame and the nearest low-current trace to burn out was on the Uniboard between a connector and the cable that feeds the regulator.

Eventually you should get some RF out of the IPA. I'd terminate it with a 50 ohm load just to be safe. Some service manuals have the schematic of the IPA and regulator board, but there were two versions: the older one used discrete parts in the IPA, the newer one uses a 5-pin RF brick that's about 1 inch by 2 inches. If you give me the TLE number on the IPA/REG board, I'll see if I have any documentation on it.

Most MSF5000 problems can be fixed if you attack them logically and methodically.

Bob M.

[motorola_otaku]

The TX VCO is active. I am measuring -17dBm on frequency at the RCA input to the IPA. I also checked the 14.4 MHz reference and am measuring -14dBm at its output.

There are 4 lines going to the IPA. Looking at the front of the station, from left to right these are yellow, orange, red, and brown. Yellow is +9.6V and does not fluctuate when the station is keyed, orange is +15V and dips slightly when keyed, red shows 68mV idle and spikes briefly when keyed, and brown is +15V when idle and drops to 0 when keyed.

The part # on the ceramic substrate that holds the actual transistors is TLE5332A. This is an older, 1985-vintage station, so it doesn't have any other orange stickers on it.

Now get this: I have another, 100% functional station (the 40W CXB on 462.700 referenced in my user profile. ;p) When I cleaned the pass-through "filters" from this one, I tested them in that station and it ceased to transmit. When I replaced the 4-pin filter that supplies DC voltage to the IPA with the original, it started transmitting again. I guess these filters are like Spectra capacitors in that they can only go so long before they're un-repairable. Anyway, I also have a complete set of near-pristine filters from the 800 MHz drawer that supplied my CXB SSCB for this station, but they have completely different part numbers. I tried them in this station, but it didn't solve the problem.

[kcbooboo]

The VCO output should be around +10dBm or 10mW. Similarly, the signal feeding the IPA should also be around 10mW.

The A+ and 9.6V signals on the IPA connector are fine. The TX Enable line should go high during transmit to allow RF to proceed through the IPA. When low, it turns on a transistor that swamps the IPA's incoming RF in an attempt to stop the flow of RF through the IPA.

The control signal should run around 8-13VDC. The lower it goes, the more power the IPA should put out. This feeds a PNP transistor base through a 100 ohm resistor, so it's the current in this circuit that controls the output power.

First verify the RF levels, then it's time to drag out the schematic and find out whether you're getting voltage to the actual IPA hybrid module. That TLE number you gave is for the 435-475 MHz IPA hybrid module; the whole IPA/REG assembly has another number, but at least I found the IPA number in my manual that has full schematics and troubleshooting procedures for the IPA. I could scan some pages and post them.

I doubt the feedthrough capacitor / filter / connector assemblies are really different. If the number of pins and their orientation is the same, they should be interchangeable. They use simple ceramic feedthru capacitors; I don't think they would suffer the Spectra Time-Bomb capacitor syndrome since there's no electrolyte to leak. If you could find some dual-row pins that were long enough, you could probably use those in place of the filter assemblies that go through the RF tray casting.

Bob M.

[motorola_otaku]

Well, I finally got it transmitting. It ended up being a combination of problems:
-My supposed "good" spare IPA apparently wasn't. I took the ceramic IPA hybrid off the assembly that I burnt the capacitor on and dropped in onto the IPA assembly that came with the 800 drawer.. that had a SP model # and no hybrid of its own.
-Bad interconnect passthroughs. I went with the clean ones from the 800 station.
-Bad PA. After verifying that the IPA was, in fact, cranking out power I dropped my known good 40W PA on it and got rated power out easy.

All in all, I am not looking forward to repeating the process with this station's 3 brothers that all came to me from the same source and in the same outwardly-good-appearing condition, but at least this one works now.

[kcbooboo]

Remember those immortal words spoken by Spencer Tracy in Desk Set: "Never Assume!"

Murphy's Laws are always with us. The supposedly good IPA wasn't. The feedthrough filter assemblies didn't. Won't get into the PA.

Let's hope the other stations are in better shape.

Bob M.