Juno-106 Restoration story/Power Supply woes

[Mooger5]

What about this? In Play Mode, not Test Mode, BTW.

One way to find out: at 50% duty cycle it must measure +6VDC at the junction of C63 and R119. While playing ANY note. It must also read the same on pin 12 of every waveshaper.

27% should read around 3V at those locations everytime you hear it.

And also in Play Mode, with Poly1 selected, play one different note at the time (ex. press C2, lift, press D2, lift, and so on) until you hear the change from 50% to 25% duty cycle. Then keep playing just that note and identify the voice with the audio probe.

[Mooger5]

I think i can see the problem. But it could only happen to the same voice. I cant access the sm right now but there are two small value capacotors near each wVeshapwer. One for each boice. In the sm are marked g and integrator. Theyre responsible for the eveness of the amplitude of thesaw wave across the kbd. Without it the saw wave decreases in amplitude as freq rises, or at higher notes. Since the comparator from which the squ is derived needs to see both 6v and saw at full amp to output 50 percent dutycycle, failing one of such conditions is enougj for less than 50 duty cyclr.

[Mooger5]

The above is ofc assuming the dco cv is correctly calibrated or faultless. Selecting just saw wave should confirm lower amplitude in the higjer registets.

[Mooger5]

In case its the voltage thay goes lower than 6v and it does happen to any of the voices, then it must come from dac process starting with the demul6tiplexor. Dont replace just uet. Measure instead the voltage T the putput or monitor the signal. Doesnt matter what appears as long as its consistent when 50 percent duty cucle. And something different is read that coincides with the offending event.

[The LT]

Thanks Mooger5, I'll get to the testing when I get to the workshop.

[The LT]

C63/R119 measures 6.01V with PW slider all the way down and around 0.54 when it's up. Switching to saw gives the same peak-to-peak voltage on the TP outputs across the keyboard.

The unit is otherwise fully calibrated. The 6V at R119 never changes and stays at 6V when the pulse width jump occurs. So the problem must somewhere else down the line, I guess. I, frankly, don't understand where exactly. I feel like a dumbass because from the schematic, the 6V goes straight into those MC5534s and measures at exactly 6.01V on them when I hear the wrong PWM note playing. How come?

It happens on all voices. I am not really sure how the voice assigner works in Poly-1 but once that PWM jump happens, all the voices exhibit the same behaviour on the higher notes.

[Mooger5]

What if it´s not exactly a PWM problem? Doesn´t it sound like two voices an octave apart instead? In such a case the probable culprit(s), could be the 4052 ICs, ic10, ic2 and ic6. But it´s a longshot. I´m actually clueless now.

About Poly1: when a key is pressed, it will be assigned to one voice, and always the same voice as long as no more than 6 keys are played and the polyphony isn´t exausted.

Ex: Press A3 - keep pressing or lift the key and play repeatedly , doesn´t matter - voice 1 will sound.
press B3 - voice 2 will sound.
press A3 - voice 1 will sound again.
press c3 -voice 3
press d3 voice 4
e3 voice 5
f3 voice 6
now press G3 - voice 1 will sound.
Press B3 - voice 2 will sound.
Press A3 - voice 3 will sound. Because G3 is now linked to voice 1.

Poly 2 is different:
Play any single key -voice 1 will always sound.
Play two keys at the same time -voice 1 and 2 will sound.
play thre keys, voices 1, 2 and 3
And so on.

Playing staccato always the same key under Play2, if the issue does affect all the voices, then voice 1 will be affected sooner or later, and pobably every six times...

[The LT]

Yeah, I know Poly-2 assigner logic, just wasn't sure about Poly-1. Actually, you confirmed it. I could not have put it into words better myself.

Well, it's actually almost unnoticeable when playing the synth musically. The synth plays like it should and is almost indistinguishable from the first 106, as long as you play higher notes, the problem does not manifest itself. It doesn't sound like the two voices are an octave apart. It's only the duty cycle that sometimes runs away on C4 and higher and that's it. No other problems whatsoever.

Duh... Not sure what to do. Swap IC10, 2 and 6 just for the heck of it?

[Mooger5]

I don´t really think all three chips would fail at the same time. Could be something common to all three though.
What about Key Transpose and octave shift? When transposing by say a fifth, does it affect anything? Does it matter if 16, 8 or 4 is selected?
I´d look at the processor, ROM and latch chips joints. The problem seems to come from them, if not because they were recently installed.
Did you press the MIDI Channel prior to DCO CV adjust? Try pressing C2 to purposely add a bias and only then adjust for zero volts if possible. Who knows...
The SM includes the Parts Modification page. There are three versions of the MC5534 and at least two versions of the Module Board. There is also the masked vs. External ROM thing.

For "remote troubleshooting" that´s all I can do.

[The LT]

I don´t really think all three chips would fail at the same time. Could be something common to all three though.
What about Key Transpose and octave shift? When transposing by say a fifth, does it affect anything? Does it matter if 16, 8 or 4 is selected?
I´d look at the processor, ROM and latch chips joints. The problem seems to come from them, if not because they were recently installed.
Did you press the MIDI Channel prior to DCO CV adjust? Try pressing C2 to purposely add a bias and only then adjust for zero volts if possible. Who knows...
The SM includes the Parts Modification page. There are three versions of the MC5534 and at least two versions of the Module Board. There is also the masked vs. External ROM thing.

For "remote troubleshooting" that´s all I can do.

Yeah, I did adjust the DCO CV properly. (Pressing MIDI CH and adjusting for 0.0V). What latch chips do you have in mind? Not sure what you mean. I will reseat the CPU and EPROM. If you remember, the module board had to be rebuilt for external ROM due to a dead slave CPU.

I've tried a different MC5534 set, all 3. Made no difference.

I will add sockets for IC 10, 2 and 6 and see how things go.

[Mooger5]

Adress Latch IC39.
There is also IC23, 24 and 26 that the SM strictly says must not have the TC prefix. Just in case. The common CD4051 will do.

Now, I was here trying to show how the DCO CV matters in the square wave generation process, but it´s complicated (to explain).
Just try a little experiment: readjust the DCO CV trimmer by ear. Forget about the zero volts DAC offset for now.
Rotate the trimmer one way or the other while playing the octave above C4 and see if the PW issue disappears. The idea is to add a bias to make the DAC see the last key in the keyboard, the C5 key, as C4, so there won´t be any keys above. DCO CV controls amplitude, not pitch of the waveform, so it won´t be critical, and easily set right later.

It may or may not work. Fingers crossed.

[The LT]

Socketed IC 10, 2 and 6, tried brand new TI CD4052BE in them. Same problem. Returned the original 4052s in the sockets.

IC39 was indeed added when switching to external mode. I used an original NOS part. I have a spare. I will try to replace it tomorrow.
I think I understood the DCO CV idea. I've tried it by setting VR33 to lowest and highest settings. No change whatsoever.

I will try replacing IC23, 24 and 26 with CD4051BE tomorrow.

This is getting interesting day by day, although it seems we are getting to the end of the line. We have changed almost every logic component on this module board. Thanks for bearing with me, Mooger5!

[Mooger5]

Don´t replace or socket any more chips. It won´t fix nor improve anything.

Yeah, vr33 was a shot in the dark.

I made the same old mistake of not verifying where C4 actually is in the keyboard.
It´s shown in page 19 of the service manual. Please confirm it´s the exact same key that was used for PWM adjust.

I keep thinking there was some fault in the calibration. Sorry for insisting. Do you realize that it must be done in Test Mode and tat Poly1 must be selected and also Bank5?

[Mooger5]

I mean, it´s not actually necessary to calibrate in Test Mode, if you know what you´re doing. It just simplifies things so it´s best to advice such good practice.

[The LT]

Don´t replace or socket any more chips. It won´t fix nor improve anything.

Yeah, vr33 was a shot in the dark.

I made the same old mistake of not verifying where C4 actually is in the keyboard.
It´s shown in page 19 of the service manual. Please confirm it´s the exact same key that was used for PWM adjust.

I keep thinking there was some fault in the calibration. Sorry for insisting. Do you realize that it must be done in Test Mode and tat Poly1 must be selected and also Bank5?

I have recalibrated this board around 10 times already. And yes, I am using the correct C4 key, I've seen the keyboard picture in the Service Notes. Yes, I am calibrating PWM in test mode bank 5.