My first Moog modular restoration. This beloved American classic is certainly the most terrifyingly valuable synth I’ve ever worked on, worth several times as much as my car, more than the equity we own in our house after paying our mortgage for two years, and more money than I’ll probably earn this entire year. So I’m very relieved to have completed the restoration without setting it on fire. I also realize it’s really time for me to form an LLC.
It’s a very unique and specific kind of restoration experience, which I’ll try to show mainly in photos because it’s less boring for you. It will be like a vacation slide show with a fun anecdote to accompany each picture.
Speaking of not setting it on fire, the first thing I did was replace the frayed power cable with its giant circular connector and the giant inlet for it on the back of the synth with a regular IEC socket. The big Amphenol connectors are for connecting keyboards and other “Controllers” and for connecting other cabinets of Moog modules. The male end of the connector has a metal sleeve that screws on to this end like a garden hose. I also had to rewire the male end of the 952 keyboard for this one, which was damaged.
The “Console Panel,” the front bottom panel which holds all of the connectors for accessing the signals brought into those Amphenol connectors on the rear, what’s supposed to be the main output mixer, and some other basic things opens up via a hinge on the bottom. Behind you can see parts of the power supply. Notice the capacitor mounted on stripboard. Sometimes you see a huge can-sized electrolytic capacitor in a power supply with many wires soldered to some very large solder lugs that it has for that purpose. My solution for replacing those is often to mount a normal electrolytic on a piece of stripboard, which allows for both easy connection of several wires to each lead and easy mounting of the whole thing by screws. Of course, I replaced all of the other electrolytics in the power supply as well.
Each module has a tag showing when it was built and inspected and some other info. These were all built in 1973 and 1974.
Each module can easily be slid out to be worked on, but the ones that are double-width have two PCBs inside, mounted back-to-back, that have to be removed which is a bit of a pain. Above, here’s a look inside the 921 Voltage Controlled Oscillator. The density of the components on the board at right, as low as it seems, is actually atypical among these modules. Most of them are simpler and have even more spread-out layouts, like the one below which is one of the two parallel boards of the 904A Voltage Controlled Low Pass Filter. Get a load of those polypropylene capacitors! They’re the biggest ones I’ve ever seen. Also notice at bottom left, the transistors of the famous transistor ladder that makes the Moog VCF special.
All of the modules needed a lot of work and had what I think of as a “high casualty rate.” I started off by replacing all of the electrolytic capacitors in all of them.
All the pots in all of the modules are these big Allen Bradley pots which are unfortunately sealed and therefore cannot be cleaned. I guess it wasn’t unfortunate for the first couple decades, but by now they were all very noisy. Sometimes cranking them back and forth a couple dozen times will clear up the noise, but for most of these it was not the case, so I ended up replacing something like 19 of them.
Before I went back through for the in-depth troubleshooting I needed a reliable source of scaled CV, so I did the keyboard next. This included a full Pratt-Read action rebuild which I did not document (this is generally one of my least favorite tasks, that has to be done in many Moogs and most ARPs), cleaning the key contacts and working on the keyboard PCB. The keyboard PCB had a fun surprise for me.
The keyboard was acting strange and there were all kinds of cut traces and wiring jumps on the board. What I finally figured out was that at some point a transistor had failed in the Trigger 1 circuit and was actually causing the fuse to blow. Someone had then come up with the “solution” of cutting traces to cut the Trigger 1 circuit out entirely and adding jumpers to tie CV 1 to Trigger 2. This meant that you could not use the keyboard unless you were playing two notes to activate the second (and now only) trigger circuit. If you patched something up with CV1 and Trigger 1 and played one note you would get nothing. I can’t believe they figured out that stupid workaround but could not figure out to just replace the transistor. Anyway, I undid all the “edits” but it still wasn’t working. Guess why.
One of the rework corrections (shown above after I put it back) was supposed to be there. I eventually figured it out by painstakingly cross-checking the PCB with the schematic. There is no documentation of the change because apparently it was a PCB layout error originally made by Moog and probably corrected in every unit even before they left the factory. The schematic only matches the board after the above change is made.
The keyboard trigger design, by the way, is fairly ridiculous and convoluted due to their insistence on using a keyboard with only one buss bar. Probably not what a modern designer would consider a solid design. There are situations here where an op amp comparator is precariously balancing in its high or low state by a margin of something like 16 mA. This is risky and not remotely the “best practice” way to set up a comparator and indeed, occasionally when I played two notes a semitone apart the comparator would just fail to flip and the second trigger wouldn’t work. I actually changed two resistors to adjust the references for the trigger comparators to give them a little more leeway.
Troubleshooting everything else was… a whole other thing. Most modules pretty much worked but almost all of them didn’t quite work! For example, here are all the things I ended up having to replace to get everything working right in the 951 VCO. Notice that I replaced 5 cermet timmers. Later, as I calibrated the other modules, I ended up replacing probably over half of the cermet trimmers overall as many of them were so bad they were impossible to set.
The VCF had fallen victim to maybe the same person who messed up the keyboard. Two transistors had been replaced but put in backwards because they didn’t realize the ones they swapped in had a different pinout than the originals.
The last step was calibrating everything which takes long enough that I actually don’t think it can fairly be described as “the last step.” It’s more like somewhere between 50 and 100 steps.
One thing that is kind of funny is that I failed to order the correct replacement pilot lamp multiple times. At one point I actually texted the Bulbtown.com “Miniature Bulb Identification Hotline.” Can you believe that’s a thing? Well, they didn’t text me back but I eventually figured it out.
There it is again. I was pretty underwhelmed by the Moog Model 15 as a synth but I guess that’s not the point. It’s history and stuff. But gosh, I really wish there was more than one mixer.