I recently purchased a later model, non MIDI Korg Poly 61 from eBay, advertised as not working. It was in a pretty bad way.
Here I detail some of the issues I had with the synth and how I attempted to resolved them.
First of all there was no power. This was due to a bad capacitor an the power supply. I checked and adjusted power rails, -15,5v etc
Of course the big issue with the Poly 61 is battery leakage. I noticed that the battery had been replaced at some point by a previous owner. There was PCB damage but it was not that bad, at least at first glance. Components tested ok around the battery and track continuity was ok with the multi meter.
I now had some lights and the display sat at program 11. I loaded in the factory presets by taking a line out from my mixer into the tape input of the synth.
There was no audio on the output just some very low audio on the headphone output. This turned out to be a bad opamp. I used a DIY audio signal tracer to probe pin 1 of each voice opamps (IC 109,209 etc), then discovered that the summing opamp, IC6 on the voice board was faulty.
5 out of 6 input opamps were also fried! Even one bad opamp here can cause other, non related voice to misbehave. So it's worth checking them all. IC101 > 106. These are matched JRC072's so worth maybe replacing them all.
There was also a large trace missing from the audio jack board and the voice board! Looks like a short with the power supply or something nasty was plugged into the audio out jack.
I was now getting sound although sustaining and distorted.
I set about replacing all the logic 40 IC's on the CPU board KLM 509A. I have a little 'ebay special' IC tester which was handy to identify faulty chips. I think it's wise to replace 40 logic as many of these chips can suffer from metal migration and fail. Modern equivalents are far superior. The 74 series IC's are generally more reliable.
There were several broken molex connectors. These were affected by corrosion from the battery leak. They can be seen by looking down the molex plug. The solution is to pull the bad ones and replace the pin. New pins can be bought. Code - WMLX-106
Another issue can be corrosion on the legs of components
Oscillator 2 was also missing on voice 2. This turned out to be a bad 8253 timer. One of 4. The others seemed ok. I discovered this after checking all the outputs of the pins and found that IC5, pin13 was sitting at 0v rather than around 2.9. The OSC and filter inputs to the voice board seemed fine.
At this point I still have two outstanding issues to solve.
Issue 1: Voice 1 has a low end distortion on it. I checked that it was not specifically voice related by swapping round chips from the other working voices. Voice 1 was still faulty. I suspected the problem was not with the voice board but rather with the CPU board.
There is a multiplexer IC2 a 4051. I check the waveforms coming from the outputs and found a small visual discrepancy in the signal. This was not present on the other voices. I changed the 4051 but the problem was still there on pin 13. This IC seems to take control voltages (CV's) and send them to the voices. I made up a bodged 4051 with pin 13 nd 14 swapped. The issue now moved to voice 2, the issue seemed to be with the input rather than the voice.
I have another Poly61, an older model with the separate Anti-log board. I bodged the KLM508A voice board into this chassis and powered up. The distortion was also present, confirming that the problem was on the voice board after all. To be really sure I placed the newer CPU board into my known good, older chassis and the voice board worked fine.
I will now check every resistor and capacitor on voice 1. Tedious but hopefully I will find the problem...
Bad trace ---------------------------------------------------------Good trace
Issue 2: All notes seem to sustain forever when EG mode is set to envelope. The release controls seem to have no effect. The only way to stop a voice is to reduce the sustain to 0. These signals are output from IC28 a 8255 Programmable Peripheral Interface chip. I used a logic probe to show the corresponding pins going low and high when the release controls were used. This suggested that the 8255 was ok. The associated opamp, IC29, output pin 7 also showed voltage change. This path was traced all the way to the SSM2056 ADSR chip. I noticed that the sustain voltage was higher than the other voltages for release at around 5v.
Well it turned out voltages are ok. ALL the SSM2056 chips are toast! I had a known good one and it restored the ADSR on one voice. So this is an expensive repair as these IC's typically sell for over £25 each.
I have just bought 2 more 61's that are either not working or partially working.
One had a very bad mains plug. I always check these before power up. You can see why...
One was a MKI with the brown voice PCB - KLM475 and separate anti-log board - KLM478A
Both had bad keyboard contacts. I cleaned the contacts with a pencil eraser and reassembled.
The other was a later model with it's battery completely missing. So a new unit was fitted. Logic checked out ok on this board so I didn't bother changing it all. I did add sockets though incase of future problems.
It had a bad 47k resistor in the battery area. It read 78k. I therefore replaced all the resistors around the battery. Some disintegrated during removal due to acid eating through the legs and weakening them.
Most Poly61's will have intermittent tactile button switches. It makes changing patch and editing a chore. I got some new ones on ebay. Manufacturer: Omron B3F-5050
After some work, I now have 2 working 61's out of 3. The third still has the bad voice 1 issue, which I just can't seem to track down.
One of the 61's I got had really yellowed UV damaged buttons. So I thought I would try a trick that worked before on my TR505 drum machine. I got some peroxide from a hair salon supplier, 10% seems to be about the strongest in the UK at least. I removed all the switches and put them in a Tupperware container. I then applied the peroxide neat to the buttons. I covered the top of the container with cling film to stop the liquid evaporating. Then I left them in the sun for a few days, reapplying every 7 or 8 hours. At night I used a UV disco light I had. The results were excellent and totally rejuvenated the buttons and knobs.
Below a comparison
I installed a MIDI IN kit to provide use as an expander