When this Minimoog Model D arrived a couple of years ago, it was probably the absolute most thoroughly wrecked synth I had ever agreed to work on. I made a deal to do it for a flat fee, even though I knew it wouldn’t be technically “profitable,” under the condition that I could take my time on it, kind of as an absurd challenge to myself and because I knew it would be really satisfying when I finally finished it. Since it is finally done and restoring it was such an insane gauntlet of tasks, I thought it might be worth reviving the long-neglected Shop Blog with a post about it!
When the synth arrived, it did not work at all. I was told it hadn’t worked in about 30 years. It had been painted black with oil-based enamel paint, had parts of it including an entire circuit board removed and lost, been sat on, taped together with duct tape, gotten wet, and then sat untouched since the late 80s (when it stopped working after being sat on). It was also very, very dirty inside and out.
I knew it had been modified, but I did not realize just how much. The entire “keyboard circuit” part of one of the original PCBs had been sawed off and replaced with a sloppy wire-wrap circuit meant to change the glide from linear to exponential.
One of the card edge connectors had been removed and the wires soldered directly to the gold fingers, but had later torn off. Elsewhere, wires had broken and been haphazardly spliced.
The keyboard / left hand controller connector had been removed and the wires twist-spliced together and taped. About 60 of the gold key contact springs were broken. The oscillator board had been badly damaged by a previous owner’s messy desoldering while replacing the wire-wound trimmers with multi-turn ones, and showed a lot of corrosion and damage to some of its components’ leads. (Unfortunately, the photos I took of these last few things were lost upon the untimely death of a previous phone.)
The first step was to create a new keyboard circuit PCB. I etched one by hand using the same type of fiberglass PCB stock originally used by Moog, and a PCB layout image that was kindly shared by Bob Grieb at tauntek.com.
Though factory PCB manufacturing is fast and affordable and I design and order factory-made boards often, I opted to use the old method in this case for a few reasons: 1) it seemed like it would be kind of fun 2) it seemed like it would be easier since I didn’t want to have to figure out how to get “gold fingers” done and space them out perfectly in Eagle 3) the new board was guaranteed to be the correct thickness for the card edge connectors because it was the same material, and 4) it would look more visually consistent with the others.
The next step was simply to populate it according to the Moog schematics, including selecting hand-matched transistors to ensure the correct behavior of the keyboard current source.
(This “Peak atlas DCA” transistor tester is great for both Hfe and VBE matching transistors, verifying pinouts and helping to select transistors for different applications. It’s a big improvement over grappling with trying to use the “official Moog transistor matching circuit” on a breadboard.)
I then had to repair all of the damaged wiring and wire up a new card edge connector for the one that was missing, the “A” connector of the Oscillator board. All 22 of the wires that had been originally connected to the Oscillator board were so worn and brittle from being jerked around and flexed that I ended up replacing all of them all the way back to their origin points with new wire, before pinning them up into the new card edge connector.
I sanded the solder off the gold fingers on the edge of the PCB, which ended up removing much of the gold, and then replated them with nickel and then gold using a simple “brush electroplating” set up I have for such occasions. I gold-plated the copper traces for the edge connector fingers of the new board I had made from scratch as well (above).
While I was wiring the new card edge connector and interacting with the panel wiring a lot, I decided to tackle figuring out what was going on with the wiring and all the modifications in the panel controls. There were many mods involving extra switches and pots that intercepted the normal signals, and that had been soldered poorly, causing a lot of the wiring to break off. I ended up opting to undo almost all of them, which meant untangling a lot of crazy reroutings in the wiring. This entailed looking at the schematics and meticulously checking, wire by wire, where each signal was going and correcting any that had been rerouted.
These photos show the contrast between the wiring in a “normal” Minimoog (already pretty wild) and the wiring in this one, during the time that I was correcting it and repinning the card edge connector.( Or maybe they don’t, maybe they both just look crazy.)
There was one mod that I decided to leave in, though I cleaned up its execution: the output of the synth being normaled to the “external input” jack to make it easy to get wild feedback effects! This is a popular and not abnormal modification for the Minimoog.
I rebuilt the keyboard’s Pratt-Read action with new bushings, removed and polished the bus bars, and replaced nearly all of the gold spring key contacts. Unfortunately the photos of these steps are all lost! I also had to redo all the wiring for the keyboard. Whoever had made the new “keyboard circuit” had removed the keyboard connector entirely and a lot of the wiring/signal paths were very different from what it should have been, not to mention very messy.
(Sorry about the gross blood blister on my finger in the photo. If you’ve ever worked on a Minimoog you can probably guess how I got it.)
I also went through and did all of the normal servicing that we would do for any Minimoog: rebuilding the power supply, replacing all of the polarized (electrolytic and tantalum) capacitors and Mullard film capacitors, replacing all of the op amps* and their terrible sockets, cleaning all of the controls, replacing the multi-turn trimmers that a previous owner had replaced the wire wound trimmers with with new ones and cleaning up the damage that he had caused to the PCB, replacing the oscillator CV summing resistors with re-matched new ones, and building and installing the octave buffer board per Moog’s service change. The tempcos on the oscillator board were “relocated” and thermally bonded to the tops of the oscillators’ 3046 transistor arrays, which I also always replace before doing this.
*We always replace at the very least the op amps on the oscillator boards, because the linearity/offsets of old originals is significantly worse than that of new ones, allowing the oscillators to be much better tuned once they are replaced. In this synth, I replaced all op amps everywhere.
Meanwhile, while all of the boards were out I also had been working on restoring the wooden cabinet.
I used Citristrip to strip off all of the oil-based enamel paint and duct tape residue. It took multiple cycles of applying the stripping gel, wrapping the cabinet in plastic wrap to let the stripper work, and then scraping off as much as I could. After all of that, it still needed to be sanded quite a lot to get it completely down to bare wood.
I originally had aspired to give it a light-colored finish, but the wood turned out to have a lot of irregularity and greenishness to it, even after all of these years. I guess it was probably poplar? I ultimately gave it a very deep, red-mahogany inspired finish by layering several different colors of stains followed by several coats of satin polyurethane.
The other seriously nasty part of its appearance was the lifting lamination layer on the front panel. Minimoogs often have this but this one was the worst I’d ever seen, and it was extraordinarily dirty as well.
I decided I was going to have to peel it off even though the idea scared me, but I also knew it would be pretty fun, so I decided that I would do it as a “reward” to celebrate when I put all the boards back in and was ready to test it.
The results were better than anyone had allowed me to believe they likely would be!
When I finally put it all back together and tested it, I was relieved and surprised to find that it worked pretty well, but the tuning was atrocious.
I was disappointed, and over the months that followed I procrastinated about dealing with this last issue.
I eventually figured out a lot of the issues were being caused by the resistors that I had noticed corrosion on before.
I ultimately ended up making the difficult decision to replace all of the resistors in the oscillator board, because so many were corrosion damaged and out of spec. With those, hand-matched in the CV summing section, new op amps and sockets, and thermally-bonded tempcos, the oscillators now track extremely well.
I tested it over and over again for weeks before I felt ready to declare it finished, and perform the final steps of reassembling it and putting on the new knobs I had gotten to replace its mismatched knobs and its new rubber feet.
Darian said I had been hired to “turn a pile of garbage into a Minimoog” and I think that’s a pretty good way of describing it. I guess you could say it was an “enriching” experience, but I will probably never agree to work on something like this again!