Fixing a broken JX-3P

[prinsen]

Hi. I'm trying to troubleshoot and fix a non-working JX-3P for a colleague. Right now I'm at the point where I'm moving into guesswork territory, so I'd be very thankful for some input before I start randomly replacing transistors and ICs. Any input. Qualified or not.

Here's what I've found so far:

Errors:

• Voices: Only voice channels 2 and 3 work - the rest are either silent or make rumbling sounds.
• Saving patches: Saving patches to user memory (Patch banks C and D) doesn't work. Patches stored in factory preset banks (A and B) seem to be OK.
• LED's: "8", "16", "D" and "Tape Memory" indicators are constantly blinking ("Tape Memory" at a higher rate than the rest). These LED's are on the same row in the LED matrix, sharing the D7-signal. I've chased the activity for the D7-signal on IC44 pin32, which on my 10MHz scope doesn't seem different from the D0-D6 signals.
• Switches: Nothing happens when the "8", "16", "D" and "Tape Memory" switches are pressed. These switches are on the same row in the Switch Matrix and share the D7 line with the above mentioned LED's.
• Crashing: The entire synth crashes from time to time. I've had it on for hours without any crashes, and then at other times all it takes for the synth to crash is starting the internal sequencer or pressing a few keys simulteneously.

Tested. OK (it seems):

• Power supply: All supply voltages are OK, and present as needed on each individual boards.
• Battery: OK @appr. 3V.
• Internal sequencer: Sequencer plays, records and saves. The recorded sequence is retained in memory, even after having the synth turned off.

[rhino]

Just finished repairs on a trashed JX3. One thing I found that might help you is this: Check the control panel board for cracks near the mounting screws - I had to jumper 14 traces. Check very carefuly around the pitch bend range switch.
Befor you change the expensive filter chips (IR3109 iirc), check all the cutoff CV inputs for errors caused by failed 4051 DeMux chips. Note: member Synthparts has some IR3109s for sale if needed.
The buttons on this one were fine, but some owners have reported buttons sticking "on".

[prinsen]

Thanks, Rhino.

I spent an hour googling JX-3P issues earlier today. It seems cracked traces on the panel board is a relatively common problem. There's a couple of small dents on the panel/case of this JX-3P, so I'll definitely do a closer inspection of the panel board.

[prinsen]

Did a thorough inspection of the panel board today and didn't find any broken traces. The board is clean and appears to be in very, very good condition. I'm thinking that whatever caused the minor dents in the case isn't the reason why this JX-3P is broken.

I've concluded for the moment that the there's at least two independent problems with this synth:
- One is affecting the D7-digital signal and f**k up everything on that line.
- The other causes four out of six synthesizer voice modules to malfunction.

I've focused on finding out what's going on with the digital signal, and haven't really looked at the voice modules yet. But I'll definitely look into the 4051-mux/demux'ers as rhino suggested.

[rhino]

Well, Damn! I had hoped my sage advice would carry the day. (John Belushi voice: "Buuuut NNNNOOOOOOOooooo!")

Could be IC 46 - a TC40H245 tri-state buffer. This is one crazy little synth: much more going on inside than most people would think.

[Don T]

Did a thorough inspection of the panel board today and didn't find any broken traces. The board is clean and appears to be in very, very good condition. I'm thinking that whatever caused the minor dents in the case isn't the reason why this JX-3P is broken.

I've concluded for the moment that the there's at least two independent problems with this synth:
- One is affecting the D7-digital signal and f**k up everything on that line.
- The other causes four out of six synthesizer voice modules to malfunction.

I've focused on finding out what's going on with the digital signal, and haven't really looked at the voice modules yet. But I'll definitely look into the 4051-mux/demux'ers as rhino suggested.

Just curious, did you take the board out and hold it up to a strong, bright light? Sometimes cracks aren't easy to spot with the board still in place. Since the synth did take a hit or two, checking for bad solder joints might be a good idea, especially if it plays for awhile, then seems to misbehave after it gets warm.

Along that line, when it crashes, hitting some of the chips (especially the ones rhino mentions) with a quick blast of freeze spray, and restarting the synth immediately after may help isolate a dying part.

[prinsen]

Thanks a lot for chiming in mr. T!

Yes, I took the panel board out and checked it under a magnifying lamp. Even did a bit of continuity testing. I'm fairly confident that the board, its copper traces and solder joints are fine.

I'm more focused on the main board which is where both the digital circuitry AND the voice modules are located. I've checked all cables and connectors between main and panel boards.

There doesn't seem to be any relation between crashes and temperature of the components. I believe the malfunction is the same whether its just been turned on or its been on hours, but I'll need to double check to make sure. I'll also try to see if I can provoke crashes, look for patterns there and do the freeze spray-trick where it makes sense.

Actually the JX-3P doesn't run hot at all. I checked it with the thermal cam we have at work. The synth is very well designed in that sense, I think.

[synthparts]

You might want to install sockets for the IR3109s so you can swap them around to see if any are bad...