This blog post continues with the disassembly and reverse documentation of my 1982 hybrid synthesizer design, based on 8 RSF-KOBOL Expander I cards, interfaced to an Apple II motherboard via a custom design 128-channel, 12 bit digital to analog convertor system (DAC). Only two of these were built.
I started this development in July 1981, worked on it for the better part of 1982, and demonstrated this on the Omegalfa booth of André Stordeur (1941-2020) and Paul Snaps (Les Cahiers de l'ACME) at the March 1984 Frankfurt Music Messe. Peter Beyls was also in attendance, he had organized my participation with André and Paul. I knew André from previous events and meetings.
Yesterday I pulled out the Apple II expansion board containing my MIDI interface. This prompted some history checking.
Sometime in mid 1982, I had received an early document detailing the early specifications for MIDI, revision 0.7, the now ubiquitous Musical Instrument Data Interface. This came to me through Peter Beyls, my conspirator and instigator in many things electronic and experimental music since 1978. This early MIDI document was signed-off by Dave Smith and Tom Oberheim, respectively Sequential Circuits and Oberheim Electronics. Japanese instrument maker Roland was also represented, I believe it was Kakehashi. As far as I remember, this MIDI 0.7 revision was the consensus at the time reached after the Jan 1982 NAMM meeting in Los Angeles.
I immediately understand the importance and the future impact of this instrument-to-instrument and instrument-to-computer connectivity and in my book, I needed to implement MIDI from the start, even when this was not an official standard yet. and there was no other music equipment available for testing or to connect to.

The MIDI board is pure simplicity. The large chip is the serial communication interface I plugged into the Apple II expansion slot. The other chips are glue and support chips. I couldn't get the original recommended SHARP PC900 optoelectronic isolator, so I used the 6N135 from HP. Turned out that this would become the chip of choice later for so many MIDI peripheral designers.
The metal can is the crystal oscillator. I had a 4 Mhz unit in the box, so I had to divide by 4 with the 74LS74 dual flip-flop to get the the required 1 Mhz per spec for the UART chip 68B50. (UART stands for Universal Asynchronous Receiver Transmitter.
The signals for MIDI in, out and through were routed to the 26 pin header, and a flat-cable routed this to a DB25 connector on the back of an instrument. I used a plastic box with three 5-pin DIN connectors to finally connect to the external world, except, there was no one to talk to (yet).
This didn't stop me from writing the software to record and playback MIDI command signals, and it is the latter feature that got a lot of attention during multiple visits from Kraftwerk's Florian Schneider and a Yamaha VP at the 1984 Music Messe in Frankfurt. (Yamaha had not yet launched the CX5M at the time but I did get a preview.)
This actual 1983 MIDI implementation was ahead of prime time, especially developing it in an isolated Belgium bubble. There was no internet and I was not on first name basis with any of the developers, but by fall 1983, I had a fully functioning MIDI recorder and sequencer in three versions, one and other based on the memory model. Apple II memory was restricted to 16kB, but I had written a version for 64kB and 129kB memory expansion cards. (By Januari 1984 I was at software revision 3.3 !)
Online MIDI history will teach you that in 1981 Dave Smith and Chet Wood of Sequential Circuits proposed a "Universal Synthesizer Interface" (USI) at the Audio Engineering Society (AES) show. Both worked with Ikutaro Kakehashi (Roland) with input from Yamaha, Korg and Kawai. to refine the proposals.
The first public MIDI demonstration took place on the January 1983 NAMM show in Anaheim, California, this between a Sequential Circuits Prophet-600 and a Roland Jupiter-6. The connection worked perfectly, marking the birth of the industry-wide standard. The Yamaha DX7, would bring MIDI to the mass market, the latter officially launched in June 1983 at the Summer NAMM Show in Chicago, and commercially available in late 1983.
I was introduced to the DX-1 and DX-7 lineup at the Frankfurt Music Messe by private invitation of the Yamaha VP to a dealer-only event. He had visited my exhibit many times and indicated that what I had done wax extraordinary for a single man operation, but the future would be digital and not analog, and I should be aware of what lies ahead and what I was up against.
Back to 2026 disassembly reality.
Below are the three power supplies in use for the synthesizer and the digital to analog convertor. These were all linear power supplies for minimal noise. The Apple II had its own factory standard switching PSU in a separate compartment.

All power supplies followed the simple layout of bridge rectifier, electrolytic smoothing capacitor, fuse and linear regulator. In the case of the 5 volt line, I used a hefty 3 Amp device. Overkill for the application but I wanted the headroom for future expansion.

The analog circuitry was fed by a plus and minus 12 Vdc supply, seen above. same concept, the power draw was only few hundred milli-Amps, so no need for large heatsinks.
The addition of the 24 Vdc section was somewhat random. The transformer had an extra 24 volt winding, so I put it to use, just in case i needed to drive relays or actuators for a future project. I don't recall ever needing it.
The RSF-KOBOL cards had their own plus and minus 18 Vdc power supply immediately mounted behind that section, bolted to the side rail of the rack. No PCB was designed for it, as this was a one-off design. So prototype board with traces and direct point-to-point wiring and soldering at the bottom. This power supply was adjustable with the venerable LM317 design.

I needed two multi-winding iron-core transformers for all of this and those make up the bulk of the weight. I verified that all these power supplies are still operational under no-load conditions, 45 years after being built and some 30 years after their last power-up in Los Angeles, 1995. (These boards will not be put to use in any future design as I don't want to recap them.)
Cheers !