Another Oberheim OB-X saved from harm

While Clusterchord has been burning midnight oil lately in his studio I’ve been working on his OB-X. This is the third time I have been working on this machine. First time was some minor service when PWM wasn’t working right. Turned out to be a bad IC in the PWM section of the digital board. At that point I’ve noticed unit has a lot of old CMOS chips, so i replaced them with new ones.

Fast forward two years later, Clusterchord contacted me to do a complete refurbish of the boards inside. No problemo. I replaced VCF trimmers on the voice boards with multi turn. Replaced all the caps with new ones (also found one voice board was dead). Not a big deal it was just one dead TL op amp. Fixed it. So i call Clusterchord to pickup his unit. He takes it away, gets home, and the unit no longer functions correctly. Autotune no longer works. Horrid s/r ratio of the unit with some constant noise going on in the background.

So he returns it back. I check all the chips, everything ok. No clue what the heck is the problem. At that point I got a bit mad because 24 hours before, the unit worked flawlessly. So Ii’ve removed all boards out, resoldered all voice board connectors, all male molex connectors on the motherboard and digital board. Assemble it all back, same problem. This was really starting to become frustrating. (!)

While messing with molex connectors I’ve noticed that at one point the noise got low a bit. So i figured out the source of the problem. Female molex connectors were oxidized, old, and wires were lousy, two of them partly broken (which mean they would conduct while synth is in one position and not in other). So…. I disassembled all the plastic molex connectors, bought new pins, new wire and started a nice whole day work of manually rewiring the whole thing by building new connections. To further prevent oxidation and enhance the connection of the wires on the place where they were being crimped i simply soldered them. This took a lot of hours but it was worth it.

Red wires = something like 160 new connections that I had to build

Power up the unit and the Autotune now works! Great! Noise was … well almost gone. I’ve realized there were two type of noises actually. One was digital like, and this one was now gone. Another one was hiss and after some research turned out to be one of the output op amps. Replaced it, no more noise. Great!

Other source of the noise was one bad IC in this area

At this point the unit worked perfectly! So I called Clusterchord, who at that time was burning some midnight oil to come and pick up the unit. New day he arrives to officially pick up the unit. But after some close inspection we figured out the Decay of the filter doesn’t work right. So he leaves the unit for another day for me to fix it. I start the oscilloscope, start digging through the signal path of the VCF Decay, I get close to one CMOS, and out of nowhere the unit goes totally crazy! I mean totally. It started producing a wild sounding drone, no controls worked right, it was almost frozen with this loud unison tone going out, no matter you press the key or not.

At this point i had no idea what the hell is wrong. Because at that point ALL of the molex connectors were brand new. All of the capacitors on digital board were new. All of the ICs (LOGIC and op amps) were brand new. What the hell is going on!!! In fact I was so mad I returned the unit and said, “dude you get someone else to fix this one, it just gets worse and worse the more we work on it.”.

Fast forward 1 year. While Clusterchord has been burning some midnight oil I sent him an email saying that perhaps I should check the OB-X because I have some new ideas about downgrading his unit to non-Encore and then doing simple side by side test against my OB-X. Mind you, my unit is original and it’s easy to follow the schematics. He had Encore modified unit, and it was a bit harder without schematic. Plus someone did some extra mods on his CPU board. Not sure why. So he brings the unit and I start some more work. At that point the machine is still behaving totally crazy, releasing only digital drones, almost as if it was a PPG. You just don’t know where to start with the service when in such condition! Really frustrating.

New trimmers and battery in the house

But i took it calmly. First I’ve replaced old trimmers on the digital board. And I found three tantalum caps that weren’t replaced on the panel board (thanks to The Real MC for info about what they do). Also the battery was dead, so I’ve installed a new one. Then I took out Encore, installed it into my unit to verify that it work properly and is not the source of the “unit gone crazy” problem. Encore turned out working perfectly. So I’ve installed it back into Clusterchord’s unit. I knew at this point that I will fix his unit. I just knew it. It was only a matter of inspiration.

And now comes the turning point – has to do with servicing Moog Voyagers which i do from time to time. One thing I’ve noticed (and i need to thank Rudi Linhard for info) is that so many Voyagers have problems and it always turns out to be the Texas Instruments chip. So I’ve came with idea – hey, what if that was the case with this OB-X. Four years ago I’ve replaced all of the chips, what if some of those has failed in the meantime.

<dramatic pause> You bet it was! Turned out exactly that. The reason why the machine was going crazy was because 2 of the brand new ICs that i have installed 4 years ago were dead. Can you guess who was the manufacturer? TI – which stands for Texas Instruments. Damn!!! Of course i never expected this outcome. To make things worse, each time i would painstakingly trace things with an oscilloscope looking for the source of the problem i totally ignored these chips, yet the search would always bring me to them! But i still ignored them knowing they are just 4 years old, they can’t be the source of the problem. I was wrong! Thanks again to Rudi Linhard for inspiration!

Goodbye TI chips!

I’ve replaced all 4 chips in that area that were TI, and now the unit finally works!!! Yesterday I’ve calibrated digital and bender boards to factory specs and as of this moment, the unit is finally FULLY functional. This is OB-X was bumping on and off the service bench for the past 4-5 years or so, always something a problem, never working right. And now these days are finally over!!!

Now Clusterchord can take it and burn some midnight oil!

Bonus images:

The search would always lead me to de-mux chip (left of the screws) but since chip was new I’ve ignored it.

I was surprised by the completely dead battery. So I’ve installed a new one. Measured the current draw, and according to calculation this one should last 13.7 years:

OB-X is massive !!!! I’m glad that it is completely fixed now.

Leave a Reply

Your email address will not be published. Required fields are marked *