Memorymoog "DEAD OSC" on all voice cards OSC3

[whinylittlerunt]

Here's hoping someone could help me pinpoint this...

Started having issues with my Memorymoog (non-plus) tuning voices, so I accepted that it was time to start some major overhaul on it. I did the following:

*Re-capped all the boards minus the pot boards (I really don't want to remove those)
*Replaced the CEM sockets with machined-type on all voice cards and the one on the common analog board.
*Replaced the bus IC sockets on the digital board, analog board, voice cards.
*Performed the service bulletin pre-reqs and then did the auto-tune upgrade. Etched the buffer board from a file and burned my own eproms for the 2.4 software.
*Replaced the fan with a DC fan.
*Calibration up to the end of the voice cards.

I'm now having an issue where all voice cards OSC3 are coming back as "DEAD OSC" in the C7 routine. I was looking at the schematics and trying to understand the OSC3 summing circuit because I'm also seeing that on OSC3, the mixer volume is very low and only comes alive when KB TRACK 2/3 is switched on, and then the mixer volume just doesn't work unless KB TRACK is turned back off.

I'm probing around on the common analog board, but man I'd be super grateful if anyone has ever run into similar on theirs and could shoot off some suggestions....

WLR

[Markyboard]

Hi Mike-

Now you've gone and done it - upsetting the memorymoog Gods, shame on you!

I'm suspicious that the other osc are really being updated. Maybe start by clearing out the auto tune values (press C5 iirc). When you hit C7 are you seeing the osc 1 and 2 values change? Maybe tweek a trimmer pot or 2 to verify they're really being read.

If osc 1 and 2 are working as expected in the C7 routine have you tried tweaking one of the OSC 3 range trimmers to see if you can bring it out of dead osc? I'm not sure what the revised auto tune software does but maybe it skewed these values.

Its also possible there's a timing issue, maybe on the DB affecting the timing signals such that they aren't sampling osc 3.

[Markyboard]

Scratch what I wrote previously. If there’s a problem with the control signal for Osc 3 Amount it could affect the C7 read out. The firmware enables and sets each Osc amount to max (I believe) to ensure a proper read out.

The Osc Amount doesn’t come from the CA board. The CA board sums all of the mods and frequency components. Instead check the connection from the DMUX board to the voice cards. Also make sure the decoder output (U12, U27) is updating for Osc 3 amount

[whinylittlerunt]

I could also be dealing with an issue in the ribbon cable for all I know. I hate how the cable is shared across all 6 voices; that’s so gross. I did replace all the ribbon connector sockets on the voice boards since they were on the bench anyway.

Thanks for clarifying the difference between what happens on the DMUX as opposed to the common analog board. Since I put the DMUx board back in with some components missing the first time due to my sheer stupidity, it also wouldn’t surprise me if I fried something out. I’ll keep posted. Thank you for the assistance!

[Markyboard]

Components just missing i.e. left off won't fry anything.

[whinylittlerunt]

Well, for what it's worth I found a little something...

When I lift the ribbon connector off S12 on voice card 5, OSC 3 immediately gets louder (obviously a note is missing) and seems to go back into 'range' if that makes sense. So there is something going funky on that voice card maybe? Only happens when I lift S12 - if I life the S56 off the DMUX board and just briefly touch the contacts it gets louder too. Sure seems like a short or connection issue with something on Voice 5 bringing the rest of them down....

[Markyboard]

S56? Did you mean S54? S54 carries the 3 Osc amounts so I can see this affecting the Osc level/volume.

Word of caution- lifting those connectors while power is on and making contact is extremely risky. I strongly recommend you not do this.

[whinylittlerunt]

I did mean S54, sorry. And taking your advice; I realize they carry control voltages. I won't lift them with the power on. I should probably check my soldering on that card since I did replace the socket. I will say, if for nothing else, I'm very good at soldering and have good tools but if I was capable of leaving out components on job, who knows what else I fell asleep at the wheel on...

EDIT: why does S54 have 3 different sockets all routing the same?

[Markyboard]

Excellent. You’ll figure this out, of that I’m certain.

Regarding the 3 parallel socket locations for S54 I can only guess that when they built the DMUX they had a different routing scheme in mind for those signals.

[whinylittlerunt]

In theory, I should be able to completely disconnect all the voice cards on the ribbon cables and just connect 1 card, right? Just to isolate? I want to see if it's the card or the actual connector on the ribbon just in that particular spot bringing everything else down...

[Markyboard]

I assume you mean removing S54. Yes that should tell you if the problem is at the signal source or if one of the VCs is pulling the Osc 3 amount down. Or maybe something else altogether. You can also defeat all but the one voice that’s still connected using C4.

[whinylittlerunt]

Well I disconnected the entire ribbon assembly of S54 and S53 and removed all the cards. I labeled the cards and then did some tracing on VC E which is the one that was showing me different activity with the ribbon S12 was lifted off. Everything traces out fine from that socket and compared to another card. So I connected S54 and S53 to only 1 voice card and I get nothing. Unless something else is wrong, I would have assumed you could have a 1 voice Memorymoog if you only had 1 card installed?

[Markyboard]

Ah...I believe removing the cards is different than just removing those ribbon cables. Only I don’t recall what it is. There maybe something funny with the power distribution to the boards or routing the audio outputs back to the CA board.

Are you working with a scope?

[whinylittlerunt]

Yeah...
Here are a couple videos to show what I'm dealing with.

https://www.youtube.com/watch?v=R_jZAdS30j4

https://www.youtube.com/watch?v=COWooH6rLtg

[Markyboard]

Oh man, I wish we started with those videos. I thought Osc 3 had no sound or very limited output. This is a Osc 3 pitch/ frequency issue which does point to a voltage input common to all VCs. Now we’re back to the CA board.

I would look at the OSC 3 Sum signal coming into any VC as you play a note.