Korg Poly 61 pitch issue consistently up 5 semitones.

[capacitorless]

@Midphase that's really narrowed down, good work I would first suspect the pot itself (since it's mechanical) but you've probably already tested it for 5k on each side when centered. Thinking open wiper here; that might be making the output of IC8/pin 7 go to the maximum or minimum voltage, eg +4v or -4v, that this part of the circuit provides.

Is 4 steps the maximum amount of modulation that normally happens, when everything is working ok? I played a Poly61 in a previous life and dimly remember the range being something like a perfect 4th.

Also, briefly grounding the right side of R34 might be a quick way to check to see if that theory holds water. The way I'm reading the schematic is that when IC8/pin 7 is not modulating anything it's at 0v, then swings up to 4v in either direction, controlled by one side/axis of the pot.

 

[Script]

As a test, I swapped out the joystick with one from my Korg DSS-1 which is close enough to the Poly-61 one, and lo and behold the tuning is suddenly where it's supposed to be.

Doesn't that point to a faulty joystick indeed?

 

[capacitorless]

+1^

I think this is how the mod circuit does it's thing: The first half of IC7 generates the mod waveform; the frequency of the mod waveform is controlled by the speed pot. This waveform is then fed to the 2nd half of IC7 to be amplified; pin 7 is then always swinging that waveform at +-10v, and is is fed through a 100k resistor to the wiper of Y (pin 4).

When Y is at the center position, the center tap (pin 2) grounds the mod waveform. When Y moves to either side, it provides some of the mod signal to one or the other of the two op amp inputs of IC8. This is sent to either the master clock or to the common VCF cutoff depending upon which side of IC8 is in use

In other words, the Y axis provides a variable voltage divider for the mod waveform, and (via either side of IC8) it is sent to the master clock or to the filter cutoff.

Does the filter mod (the other half of Y) work? It could be forcing a higher cutoff or lower cutoff like the pitch side of the circuit; it wouldn't be as noticeable as the pitch always being up 4 steps, of course.

Edit: easiest way to test would be to ground pin 1 of the Y axis pot and see if the pitch returns to normal.

 

[Midphase]

IC6 is a Quad OpAmp. I checked the voltages on IC 7 and 8 and they read correctly.

Quite honestly, the circuit isn't terribly complicated, with relatively few likely faults. Since I already replaced IC 7 and 8 with no difference, I now have to assume that perhaps one of the PNP transistors might be acting up. It's all quite unlikely but at this point I'm not sure what else could be the culprit.

 

[Midphase]

Doesn't that point to a faulty joystick indeed?

That is what I thought initially -- however swapping out both pots from the other joystick didn't solve the issue, so even though the two pots from the other joystick were marginally more to spec (10k vs 8.6k-ish), the pitch problem persisted regardless of which set of joystick pots were connected.

Ultimately the only difference where I am able to get the correct tuning on the keyboard is when I disconnect what to me looks like pin 3 (out of 4) on the Y axis potentiometer which, according to the schematic, connects to what is labeled as Pin 1 (which I tested with my continuity meter).

So -- assuming that the pots are not the issue, I also tested the Tune and Tune Trim pots which both checked out, the dual 4558 (IC 7 and 8) opamps are not the issue, and the resistors are not the issue -- that would really leave either IC 6 which is a quad opamp, and Q7 or Q2 which I believe are both PNP transistors.

Unless I'm missing on something else that is obvious?

 

[Midphase]

@Midphase that's really narrowed down, good work I would first suspect the pot itself (since it's mechanical) but you've probably already tested it for 5k on each side when centered. Thinking open wiper here; that might be making the output of IC8/pin 7 go to the maximum or minimum voltage, eg +4v or -4v, that this part of the circuit provides.

Is 4 steps the maximum amount of modulation that normally happens, when everything is working ok? I played a Poly61 in a previous life and dimly remember the range being something like a perfect 4th.

Also, briefly grounding the right side of R34 might be a quick way to check to see if that theory holds water. The way I'm reading the schematic is that when IC8/pin 7 is not modulating anything it's at 0v, then swings up to 4v in either direction, controlled by one side/axis of the pot.

Each side gives me around 4.5K..not exactly up to spec, but if the pot is merely acting as a voltage divider, it shouldn't make much of a difference provided both sides are the same, which they are. Having said that, the service manual states a maximum pitch bend of +/- a perfect 5th and not a 4th when the Bend knob is set to 10. Perhaps that's due to the less than 10K range of the pots? There is also a pitch bend range trim that I have not yet adjusted which might be a way to compensate for the less than ideal pot values.

I wish I could just substitute a regular 10Kb pot to test, but unfortunately Korg uses these weird 4-pin pots, and I can't find exactly what that fourth pin is used for.

BTW, I also used some contact cleaner on both pots just to make sure, although it didn't really have any effect.

I will try grounding R34 and see what happens. Pin 7 reading basically 0v would explain my readings. I will also test with the joystick at the extreme positions which, if I understand it correctly, should give me a solid + or - 4V.

 

[Script]

Don't know then. I'd check the original pot again. Maybe your alligators connected wrong.

The fourth pin is called a 'centre tap'. It taps at mid travel of the pot.

Bourns have some, I think. AliExpress have duals, but only 50K and 100k.

 

[capacitorless]

Yeah that 4th pin centre/center tap is there to provide a voltage reference (or ground, in the case of the LFO Y axis) so that there's the max voltage (or nothing) on the wiper when the pot is centered.

Edit: in the case of the LFO pot, an AC voltage is fed into the wiper, and as the wiper moves away from the center position, that varies the AC voltage appearing at either end of the pot. Those voltages make their way to the master clock or common filter cutoff sections elsewhere on the synth.

I'm really curious what grounding pin 1 of the Y axis pot does

The way I'm reading the schematic, +-4v is an AC waveform, as is the +-10v a little earlier on. (edit: I added that to the 2nd paragraph but will retract it if we find out that the LFO circuit does something different)

 

[Script]

Only an example of a centre-tapped pot

 

[capacitorless]

It almost looks like someone could DIY their own, with something like conductive paint.