Circuit Overview:
Early MIDI synth with some scaled back aspects but a lot of typical subtractive analog things going on of course. Timer chips used for the DCO design as was typical in the first generation of those units. Analog VCF and VCA chips but digitally generated envelope controls and oscillators. BBD based chorus. 8253 chips are used as w/many dco synths to create digital sqare waveforms. S+H cells via mux chips store voltages, including for ramp rates in saw converters. (in mks-30, IC3, IC4 in each of 6 voices). TR5, TR6 are clocked to give proper saw frequency for DCO1 and DCO2, and S+H voltages from IC23 and IC24 mux's set ramp rate. IC2's do the voice mixing w/ 'bal' lines setting mix through TR3/TR4. And this signal to pin 1 (vcf in) on mks-30's IC1 modules (80017 as in juno 106, mks-7 etc.) Pin 3 is out to gain adjust to pin 9 vca in. And pin 10 vca out to the resistor for that voice whose other end is the summing node on ic60b's input. The summed out is applied to HP Filter then to Chorus.
Service Tips:
The MKS-30 has the failure prone filters of course. I've had good results with soaking them in acetone for several days then peeling off the resin coating. Crystals tend to form that are conductive under that jacket and it just all washes away when you perform this procedure. Otherwise I've only done key cleanings and pot replacements and so on with JX-3P that I can recall. Here's Florian's FAQ. I have one finally on bench with all lights on. Corrosion on eprom socket pins and eprom reads/won't verify. Burn new one...lights on a droning sound still. Tim at custom shop pdx has had cpu issues but....nope not this one. TURNS OUT the CE line to SRAM was open! Apparently the code insists on reading valid data there before allowing ic41 to start enabling data to turn off all led's etc.! NOTE that trace from IC42 pin 6 goes through hole to top side...over a ~2CM then down again for absolutely no reason. ALSO I recommend adding .22uf to 2.2uf cap from ground to cathode of diode to non-volatile voltage supply. It's noisy and theoretically some HF spike could foul memory I suppose.
i had a phantom sub harmonic it sounded like but it was a flux induced leak on an S+H cell for saw ramp rath.. so a saw modulation on saw ramp rate..
Parts:
I have a variety of parts that might help you on these as it appears Roland used many of the same parts those years. And also there are some discrete components that I usually stock.
Contact Sound Doctorin'
GOOD NEWS for MKS-30 users!!!! I've purchased and tested one of the 40 euro module redesigns. GREAT WORK! This should drive the price of the used modules down since these are way more reliable and sound identical! Here is the Module Retro page. This device must be hard wired which takes a few minutes but well worth the effort to get something in using common 13700 ota's.