Yet another MSF issue - no RX

[motorola_otaku]

Well on the last one I worked on it was no TX, this time it's no RX.

Station is an externally-duplexed 75W CLB that was upgraded to CXB and is operating on a 441 MHz pair with standard TX low/RX high spacing. The problem I'm having is that the receive is non-existent.. like, handheld right next to it won't key it nonexistent. The only way I have been able to get it to pass a signal at all is to remove the lid from the RF tray, remove the cable from the RF uniboard to the injection amp, and key a handheld right next to it. More telling is that there is zero meter reading on position 3 of the test set, and pumping hundreds of microvolts into J1 on the preselector results in no meter movement on position 2.

So far I have swapped the injection amp, the mixer/injector assembly, the front end preamp/1st mixer board, the entire preselector can, and the RF uniboard all to no avail. I have literally run out of options - what else is left? I have verified many times that programming is correct. Also, TX is spot-on, is cranking out full rated, and both VCOs tuned up by the book. Oh, and I replaced all of the RF tray pass-through filters with known good parts from a parts-donor 800 station.

So what say ye?

[kcbooboo]

Assuming the Receive VCO is locked, you should be able to pick up the injection signal coming out of the mixer amp. It's at a level of 1 watt there (+30dBm) and should be 10.7 MHz lower than the programmed RX frequency. Feed it into a spectrum analyzer, possibly with a 20-30dB attenuator ahead of it to protect the instrument.

I had one with the same symptoms. Found an open circuit inside the mixer coil assembly going to one of the RCA connectors. A signal tap, soldered to a coil, had broken off, leaving a fillet of solder where the wire belonged. Took a couple of seconds to heat it up and solder it properly. But since you've already swapped everything and still aren't getting any receive, this probably isn't your problem.

Are you getting any errors on the CXB's display? Are both VCOs locking? Did you start with a default code plug and build it up specifically for that station?

Bob M.

[motorola_otaku]

No errors, and yes both VCOs are locked with an even 38mA on the test panel. I'll check for RX +10.7 and report back. Codeplug was built from the RSS-included default.

[motorola_otaku]

Okay, I have an update. I took another 100% functional station from the same lot and vintage that was programmed and tuned for GMRS, and loaded the 441/446MHz codeplug into it.. and it did the exact same thing! Both VCOs would lock but no RX whatsoever, no meter reading on position 3, and won't tune up. TX died too. I reloaded the original codeplug and reset the VCOs and it came back to life.

The R2 working range for a MSF is 435-470 MHz, right? There's no Range 3 for T-band or something that won't work past 461 MHz? This whole lot of stations I have came out of an old LTR system that operated in the 461-462 MHz range.

[kcbooboo]

There are only two ranges for UHF MSFs: 400-435 (roughly) and 435-470 (roughly).

Make sure the VCOs are both for range 2. The last digit on the part number stickers should be "2". The transmit VCO has components for modulation; the receive VCO does not. They can be swapped and will still oscillate, but you won't have any modulation with the receive VCO in the transmit position.

Since the VCO are locking up, they're obviously oscillating on some frequency. Put a freq counter probe near the injection amp and see what frequency it's running at. That might provide the answer. Once you know that, add the IF freq and see if the receiver hears a signal at the frequency IT wants to pick up, rather than the one YOU want it to pick up.

MSFs have two different synthesizer chips. The software usually selects the proper one based on the operating band. One is called Mosaic, the other is Non-Mosaic. I don't know what the difference is, nor what the meaning of Mosaic is, but perhaps the divide ratios are slightly different between the two chips, so you have to get it right. Maybe your old CLB Uniboard has it one way and the code plug has defaulted it the other way.

Bob M.

[Doug]

I have a sneaking hunch its in your injection filter tunning, moving a station so far away in frequency has a natural effect on meter position 3 as in 0 uA. I'm going to guess that L7,8,9 are roughly flush with the cover. Its a time consuming operation and its going to make you go cross eyed but if you move L7 down slowly(possibly a 1/4 turn at a time) at some point you will see a very slight increase in uA. At that point move to L8 and go through the same procedure at some point every thing should start coming alive. Move to L9 and do the same thing at somepoint in there you should start seeing a noticable change. Once you start seeing it then start peaking for max on 3. Should roughly end up around 30uA. Give or take keep us posted.
Doug
By the way its good to see you still lurking about Bob

[motorola_otaku]

I have a sneaking hunch its in your injection filter tunning, moving a station so far away in frequency has a natural effect on meter position 3 as in 0 uA. I'm going to guess that L7,8,9 are roughly flush with the cover. Its a time consuming operation and its going to make you go cross eyed but if you move L7 down slowly(possibly a 1/4 turn at a time) at some point you will see a very slight increase in uA. At that point move to L8 and go through the same procedure at some point every thing should start coming alive. Move to L9 and do the same thing at somepoint in there you should start seeing a noticable change. Once you start seeing it then start peaking for max on 3. Should roughly end up around 30uA. Give or take keep us posted.

I suspected this as well, but running L7, 8, and 9 their entire length net me nada on meter 3.

Thanks for the input, guys. I have to sideline this project to get another repeater re-tuned (that is completely functional, thank God) but will pick it back up in the near-future and post my results. Seems like every time I get into one of these things it ends up being a learning experience.

[Doug]

You may have to go back and forth L7,8,9 several times over before seeing anything.....I did one this AM and it took me about 15 minutes on that stage before I saw the end result.
Doug

[kcbooboo]

He did indicate that he swapped all of the major assemblies. I will assume that he tuned the mixer coil assy, or at least that the one he swapped in was working before he swapped it. If the station is being programmed for the same freqs, then swapping the mixer coils should still result in some indication on M3.

Now this could be a bad assumption on my part, or a bad part on my assumption, or an assumption of a bad part.

Bob M.

[motorola_otaku]

Given my past experience with stations form this lot, it wouldn't surprise me if the installed parts and the spares were both defective.

How do you get the pressed-in shields out of the injector/image filter? I can get the two out that cover the coils themselves, but not the one covering the 3 RCA jacks.

[kcbooboo]

They are just press-fit into the housing. I inserted the tip of a black dry-wall screw (it was the smallest, strongest thing I could lay my hands on) into the tiny stress-relief hole at one corner and managed to lift the cover just a bit. After that pry it off with anything you have handy. There are no nearby parts you can damage unless you go in 1/2 inch. I don't think the orientation is critical, so just put it back so it's flush with the rest of the casting.

I had one station with exactly the same symptoms: no M3 and all the coil turning in the world didn't yield a thing. I pulled the coil assy out and connected it to external test equipment and still saw nothing. Upon opening it up, I discovered that one wire (a tap) had been side-soldered to one coil and one RCA connector, and the wire had pulled out of the solder blob at the coil. I could see a nice concave indentation in the solder where the wire belonged, but as it was several thousandths of an inch away, it didn't provide any signal path. I heated the joint up, pushed the wire back where it belonged, and instantly had signal on my test equipment. Problem solved.

Bob M.

[Bill_G]

Those coils are known for growing hairs and broken solder joints. As kc said the bottom covers are pressed in. Insert something small and strong like an Xacto blade tip through one of the small ditches in the side the raise the cover out enough to grab with a pair of cutters.

[motorola_otaku]

Ooookay, here's where I'm at now.

I have two sets of receiver hardware: preselector, preamp/mixer, and mixer/injection filter (both disassembled and found to be clean and intact.) I have two RF uniboards. I have two injection preamps. I have a MSF station that is 100% functional and the "specimen" that refuses to receive properly.

I tested both receiver hardware sets in the functional station and they both work as advertised and tuned up quite nicely at 446 MHz. I tested both RF uniboards in the functional station with the same result. Ditto for the injection filters. But when I assemble the receiver hardware, RF uniboard, and injection preamp together in the specimen station... yep, you guessed it, no receive. So what does that leave? Interconnect board, pass-through filters, and possibly the RX VCO which is tuning and locking properly. I tried swapping the interconnect board (which, by the way, is the one with an onboard 5V regulator for use with a CXB controller deck) with a known good specimen which didn't solve the problem. Next up is swapping the pass-through filters with known good parts, and then I guess swapping the RX VCO with a known good part. Will advise...

[d119]

Let me know if you end up needing any parts. I've got a few junker MSF's around the shop.

[kcbooboo]

I'd like to know what frequency the receive VCO is actually operating on. A freq counter with even something like a scope probe on its input should pick up the output from the injection amp, so that's where I'd try to measure it. Report back with the programmed RX frequency as well as the RX VCO frequency; they should be off by exactly the IF frequency: 10.7 MHz.

If you have a spectrum analyzer, connect it to the output of the injection amp; there should be almost +30dBm there.

Bob M.