Korg MS-20 mods/schematics

okay, so i got hold of the schematics and board layouts here.
if you want to reproduce some of these mods, you'll need them, as i already screwed the bugger back together, so no handy step-by-step pix.


1. patching out the VCO's.
this is easy - just tap into the right points on the waveform selector switches.

as you can see, the white noise on VCO1 is the same white noise source that's on the patch bay, so to me there's no real use in having that twice.

VCO2 is a little more difficult to read, but ok.
the bottom throw i highlighted is the straight saw signal. easy.
next: IC6 turns that saw into a square or pulse wave, depending on the state of its non-inverting input.
as you can see,  the second pole of the selector switch connects that input either to ground or not. when connected, you get a square wave - if not, pulse. that's why the square and pulse waveforms are on the same point, aka the output of the opamp.
therefore it is impossible to tap into a square and a pulse wave at the same time here - the same opamp is responsible for both, and it can only do one or the other. it's basically a hardwired version of the PW pot on VCO1.

the four logic gates on the bottom here are your ring modulator (there's an error in this factory schematic!), and the output goes from R65/66/67 straight to the top throw of the selector switch.

notice that the square/pulse from VCO2 gets used to drive the ring modulator through D8 -
the second pole turns the pulse wave into a square wave when on the Ring setting.
however, the ring gets fed the pulse wave (same opamp output) when in pulse or saw mode as mentioned in my previous post. so out the jack comes.


listening to the waveforms on the new jacks, i noticed the volume drops if the same waveform is selected as the VCO out to run through the internal filters. (listening to saw out patchpoint with saw selected on VCO, for example.)
i might want to buffer them, but for now, it doesn't really bother me.
might depend on the circuit they get patched into, though.

2. sync switch
hard sync is not all that hard to implement.
you need to reset one VCO (slave, preferably the one with the higher pitch, so VCO2) with every cycle of the other VCO (master, here VCO1).

the trick is around D4 and D7.
in this case, you want to pretend that D4 (VCO1) is attached to Q10 and Q11 (VCO2), overriding D7.
you cant't just make that connection, because D7 will then also override D4, and that's no good.
so we throw in a fake D4 substitute.
a wise man on a nice forum mentioned that it would be best to use shielded wires and connect the shields to ground to avoid the oscillators locking when their frequencies are close. you run the risk of them syncing through stray wire capacitance even when the switch is not engaged.
i took his advice - never even tried otherwise, but it works beautifully.

3. EG2 to VCA switch
this, to me, is a big one. i feel it expanded the possibilities of the synth a lot.

slightly edited schematic:

there's a length of wire conveniently running across the board, connecting the EG2 output to the VCA in.
i believe it's labeled 'No.4', but don't quote me on that - anyway, it connects to R125.
if you snip that wire and put in a switch, you can now defeat EG2 control of the VCA.
i also put an extra jack on the side of the EG, as an EG2 0-5v output.

of course, you need to plug some other voltage source into the 'initial gain' to open the VCA, or else it'll be dead silent. but at least now you can choose whether or not you want EG2 in the CV mix!

4. VCA floor pot
since i felt it was a bit of a waste to always plug something (mostly the mod wheel) in the initial gain jack, i decided to hardwire a mod wheel duplicate into an initial gain duplicate, freeing both the input and the mod wheel up for other uses.
it was almost as easy as copy/pasting, except that i didn't need the negative voltages of the mod wheel.

i kind of hustled the schematic around a little, to show what i copied.
i left out the new EG2 jack for clarity.
everything still connects to the same spot, the CV mix node.
i copied the initial gain input (diode + 1M pulldown), and one half of the mod wheel.

this way, the initial gain jack, EG2 and the floor pot all play together nicely.
you can modulate through the initial gain jack, on top of a floor pot drone that'll last until the wall runs out of juice.

if you ever think of offering this mod to people, some credit would be nice.


Dylan said...

This is the most useful and descriptive MS20 mod info I've found online.

Thanks for your help.

Remörk said...

why thank you sir.
it's mostly gobbled together from different sources on the net, but i'm glad it was helpful to have them sort of centralised..

Lars Christian said...

Hey! This is amazing. Im going to do this mod for christmas! I was wondering, what expanded possibilities could you get from the Eg2 to VCA switch?
Also do you know of any way of getting eg2(or eg1 for that matter) to modulate osc2 without taking out the cv to osc2?


Remörk said...

cheers! well, expanded possibilities.. basically it allows you to turn a hardwired functionality OFF. so you can choose whether or not EG2 controls the VCA.
thing is, with the initial gain jack, you will add whatever you put in there to EG2's envelope - so you always end up mixing your external envelope with the internal one.
now i just like being able to control the VCA externally, without EG2 interfering with every trigger it gets. very nice for sequences etc.
so, technically, it's less functionality - but to me, in a good way.

and as far as modulating only OSC2 goes - i would try patching kbd cv out > VCA > VCO 2 cv in, with the envelope controlling the VCA. that would make the note you play on the keyboard the highest note your envelope will reach. not ideal (you'll never go higher than that note), but it's the only option i can think of without resorting to external CV mixers.