Yesterday, I noticed that one of the two oscillators stopped working again. I decided to open up the Polivoks once more to figure out where the issue was coming from this time.
The last time I worked on the Polivoks, I had just built an audio probe. It worked fairly well, but I wanted to go a step further, so in the meantime, I bought an oscilloscope. I ended up getting the Rigol DHO804, a 4-channel scope that should be more than enough for this kind of work. I immediately put it to use to look for a waveform on the oscillator boards.
As soon as I started measuring, it became clear that there was no signal on one of the boards. Fortunately, it wasn’t the board I had worked on before. Could it be the same issue?
In any case, there was no waveform on pin 6 of IC A7. That meant something went wrong before that point. I measured the transistors, and they were all still fine, as were the resistors and capacitors. Luckily, I had the second board that was working properly, so I could use it for comparison. After some measurements, it became clear that there were several points showing much higher voltages (12V). See below. I decided to measure the resistors, and I noticed that R23 had a much higher resistance than specified. It measured 2.9kΩ instead of 100Ω. I then swapped the 100Ω resistor from the good board to see if the waveform would return, but unfortunately, that didn’t fix it. The voltages were still high in some places, although pin 5 of A5 now showed much better values! I then measured some values on A4, and on one pin, I measured 12V, whereas on the good PCB, that wasn’t the case. Internally, the IC likely shorted, causing a higher current through resistor R23. It then overheated and likely caused the resistance to permanently increase. Fortunately, I had bought a spare KR198NT1A from Erica Synths, which I could now use.
After soldering, I measured again, and yes, the waveform was back! The only thing that stood out was that the PCB was drawing more current than the good board. It was now drawing 0.035A instead of 0.015A. Even though the oscillator seemed to be back to normal, something still wasn’t quite right. I decided to measure and compare a bit more around the components I had replaced. It quickly became apparent that pin 10 of IC A5 had too low of a voltage. I tested the capacitor C4, which sits between pins 9 and 10, and it was fine. I then swapped IC A5, and that turned out to be the issue. This is a transistor array, and likely, one of the transistors inside had failed, causing pin 10 to show the wrong value. When I powered the PCB with the new IC, the current draw returned to 0.015A. So, I unfortunately need a new IC. It’s the KM551UD1B, which is quite rare, and there isn’t even a datasheet available for it. Eventually, I managed to buy a couple from a Latvian store (http://evita.lt) for 2 euros each.
So, the cause was a damaged resistor and two faulty ICs. One was probably the cause of the other. Fortunately, I still had one spare left. We have a working Polivoks again!