Making a Roland MC-8 cable

Today my Roland MC-8 arrived. Unfortunately, it arrived without any cable to connect the main brain to the CV/gate box. I knew and expected this. The guy I bought it from mentioned that details for building your own cable were floating around on the internet.

Apparently, information about how to build the cable for the MC-8 is a tightly-guarded national secret. About the most I could find was a short messageboard thread where a couple people discussed which possible parts to order. I scoured page after page for any more details but came up empty-handed. So I took the scant information from that thread and started to order the parts, only to discover that one of them was now out of stock. An hour later, I determined that pretty much nobody had it in stock.

That’s when I decided to roll up my sleeves and try to research this cable as best as I could. I knew that the cable was likely a 60-pin cable with connectors made by a company called Hirose. The model number of the connector seemed to be Sumicon 1600. However, a check on Mouser showed that there were dozens of different parts in this series. There were rarely any pictures, and it wasn’t obvious from the description which needed to be ordered. To make matters worse, the connector pieces were kind of pricey, so I didn’t want to just order a ton of stuff I didn’t need.

After going through a surprisingly difficult registration process, I was able to get the original spec sheets from Hirose. It’s rather old, but it explained the different types of connectors in the series, and how to decode their part numbers. Apparently there are a couple different shells, a couple different connection styles (crimped, soldered, wound), and of course, the male and female versions of everything. Eventually, I ordered the following parts from Mouser:

P-1660A-CA(50): this is the black hard plastic shell. It’s pricey, but I must admit that it’s a pretty sturdy thing. I ordered two of these.

P-1660BA(09): this is the actual block of pins. This is the solder-style block. I didn’t feel like figuring out their crimping tool or trying to wind wires. Of course, two of these also.

Incidentally, I also ordered the P-1660A-STA(51) but later realized that it wasn’t needed at all. This part appears to actually be something on the female side of the connector.

Those appear to be the only two parts you absolutely need.

Next I needed some wire. I looked briefly into trying to find a 60-wire multiconnector cable, but this seemed pretty far-fetched and expensive. I stopped at a thrift store and picked up an old parallel cable, which has about 25 wires inside. But I didn’t like the idea of investing hours of time soldering a bunch of wires of unknown quality. So I finally just decided to use single solid wires and band then together somehow.

This afternoon the MC-8 arrived in two large boxes. Everyone remarks about how much bigger it is than they expect. Indeed, it’s a large piece of equipment. I think somebody joked that it looked like it was ready to launch some ICBMs. If somebody else didn’t joke that, then I’ll claim that as my own. That’s a pretty good one.
Roland MC-8
I was waiting for the actual sequencer to arrive so that I could check one final thing on the cable: did I really have to connect all 60 pins? I was hoping that maybe all 60 pins weren’t used. So I opened up the bottom of the CV box and had a look.

MC-8 bottomThis Sumicon 60-pin connector actually attached to a 50-pin ribbon cable inside. Without access to any schematics or anything, I couldn’t easily tell whether all 50 pins were needed. But I found that there were a few that I could scratch off my list. Pins 52 though 60 and pin 21 were all grounded together, so I figured that I only needed to run one wire for all of these.
MC-8 ribbon connectionI also wanted to know which pins to wire together on the two connectors. Wiring the same numbers together made sense, but I’m pretty new to this. For all I knew, you had to wire pin 1 to pin 60, pin 2 to 59, etc. So I opened the main unit and checked it out. The grounded pins that were all soldered together was a tell-tale sign. After studying it over, I was confident that pin 1 was wired to pin 1 on the other connector, etc.

So then I spent the next couple hours cutting up wire and soldering them. Actually it was a little faster than that. At one point my mom stopped by and helped me cut wire. Yep. My mom. We chatted about our respective Christmases and Roger Miller. In no time, the wires were all soldered.

Soldered MC-8 cableI wanted to test it out before I clamped the plastic shells on it. I plugged the cable carefully into the two boxes, stared at it for a long time and then panicked. Somehow I had reversed the orientation on at least one of the plugs. The wire on pin 60 was actually supposed to be plugged into pin 1. I couldn’t figure out how it had gone wrong

I certainly didn’t want to just power on the sequencer to see what happens. With the wrong voltages going who knows where, I could see something getting fried easily.

I started thinking of having to unsolder each of the 50 pins. Then an even worse thought occurred: maybe I had actually reversed both connectors. These Sumicon connectors can only be plugged in one way. They can’t be flipped, so it’s possible that I had actually screwed up the orientation on both plugs.

I opened the main unit first to check the orientation of pins 52-60. Since these were all soldered together, they made a convenient marker. To my relief, the pins on the main unit looked good after all. Then my eyes drifted back over to the CV box and I started to think about it again. I had the box flipped upside down when I was checking the pins. Suddenly everything made sense again. The pins that appeared to be on the top were now actually on the bottom. I double-checked the photos I had taken. Slowly I started to convince myself that I had wired it up correctly after all.

So with a little fear and trembling, I powered it up, and it actually works! I should temper that: CV1 appears to work, albeit in the mysterious way that the MC-8 works. I haven’t read the manual yet, but I did manage to get it to spit out a string of notes to my Minimoog. Tomorrow I’m going to give it a more thorough test and then finish assembling the cable. Anyway, I thought maybe somebody, someday, might want a little info about building a new cable for their MC-8.