Last winter, my friends, Ryan and Lisa came for a visit and brought along a Roland TR-606 and a Korg MS-10, both of which were not working. The 606 was easy to fix since the lead to the battery compartment had come loose. The MS-10, however, was a completely different story and required a lot more research since I’ve never undertaken a repair like this before. I’ve managed to get the MS-10 back up and running, and I want to share some of this information, just in case someone out there runs into a similar situation.
The Korg MS-10 was purchased in Europe but came with an unpolarized NEMA AC connector that looks like it would work in Japan or North America. When Lisa and Ryan used this in Europe, they had it connected to a power adapter which bucked the voltage down from 220V. After moving to the US, they (Ryan) plugged it into a regular 120v socket, and it started smoking. Since then it has been dead. Whenever I would visit their home, i would see the dead little MS-10 sitting as a bit of decoration on a table.
About a month ago, this project seemed like it was ready to be tackled. Fortunately, the service manual is available from several sources online. Synthfool is not only a great synth tech, but provides the community with the archive of these materials. After studying the schematics I started to open up the unit and noticed something odd. The badge says it’s requires 15 volts ac. This was a head scratcher, and I thought perhaps it was a misprint and should have been stamped 115 rather than 15. It simply makes no sense especially since it has a NEMA plug.
So after opening up the unit, I discovered that the transformer is basically used as a coil. The incoming voltage from the AC line runs into one leg of the primary then passes through the other leg of the primary right into the motherboard. The other leg of the AC line passes into a fuse (which was blown) and the power switch then into the motherboard. I was expecting to find a normal AC connection where the incoming power connects through the fuse and switch to the primary of the transformer, and the secondary windings of the transformer passing to the motherboard. However the way this was wired, the badge was indeed correct, and it was intended to run off of a 15 VAC power source… not 220/240 or 110/120vac. When the MS-10 was plugged into a socket, 120 volts was sent straight into the motherboard and did some serious damage.
At first, i thought maybe I can get lucky and just replace the rectifier diodes and the 7815 and 7915 voltage regulators and it will be done. I pulled the diodes and regulators and found that the diodes were OK, but the regulators were shot. Fortunately a these components are easy to get, so I popped in the new ones and started to see some signs of life. The LFO light started blinking, and I could hear some noise. The keys wouldn’t play, but after patching the LFO into the VCA, I could hear some activity and I could adjust the filter into self-oscillation. While far from being perfect, there were some signs of life, and I figured, why not push forward to see how far I could get.
According to the schematic, the MS-10 oscillator section takes power right off the transformer… however in this case, 120VAC was injected into this part of the circuit. This would be good place to start looking. Sure enough some of these FETs were damaged and replacing them got a 60Hz hum being shaped into a sawtooth and pulse, however the envelope generator and keyboard were still not working. It dawned on me that half the AC leg might have still been active after the fuse blew, so there was still a lot of voltage passing through the circuits and to ground through the audio out socket. It made perfect sense that the CMOS components were damaged and many of the op-amps (TL0s, etc). After replacing many of these more sections started to function. The envelope generator, the modulation routings, etc. Sure enough, after replacing all of these components everything came back to life!
Thinking that a critical component might have been damaged, i also sourced a special NPN transistor pair (2SC1583) at the heart of the oscillator. This part is also used in the TB-303 and x0xb0x and has become rather difficult to find. To make matters worse, according to a thread on the adafruit forum, counterfeit versions are floating about, so i avoided the foreign eBay suppliers. As it turns out, the original transistor was not damaged. I popped in the new transistor, and I was surprised that there was a bit of a difference in sound. The original transistor has a bit more character and bite, so i put it back in. I’ll keep the spare in case one of my old synths needs work.
I thought about properly wiring the transformer back into the circuit, but instead opted to leave the power section untouched. This particular MS-10 still runs off 15 volts AC, but the original power cord has been removed and replaced with a 2.1mm power jack and a wall wart. Hopefully this mistake will not be made again in the future. The moral of the story is this: Yes, apparently Korg made some MS-10s and MS-20s that were designed to work on 15VAC, but have a North America AC plug. If you plug one of these 15V units into a 100/110/120 socket, it will definitely fry. If you’re in Europe and plug it into a 220/240 to 120 converter, it will fry.
I tracked some raw samples of the unit. You can download the Reason session here:
and a short video: