MS20 Filter CV input level

DIY STUFF
1

Hi folks,

I’m back into building and modding sections of my synth after a big contract job ate my life for a few weeks and I’ve suddenly realised (I don’t know why it took this long) that my MS20 filter (similar to the LMNC filter, based on the Eddy Bergman design) doesn’t fully open the cutoff frequency with eurorack standard 5v CV levels.

I’d imagine this is because the Korg MS20 operated on a 10v or more CV level, but being that we’re building in KOSMO which is electrically eurorack underneath, I’m surprised so many seem to still be using this design. Is it perhaps because many use the Beatsteps which have very hot 12v CV levels so they don’t notice?

Here’s LMNC’s design:

I’m wondering (apart from adding another op-amp in line with the CV input) what the best scenario might be in terms of allowing this 5v input to feed the filter properly? If I lower the input resistor (100k - feeding the BC558), that would take us out of standard best practice input impedences. I also could increase the resistance pulling us down to ground (currently 1.8k) and add resistance to the 12v feeding the cutoff potentiometer to essentially make it a 5v source, but I guess this could also have some unintended consequences…

Any hints or thoughts? It seems like this would be something pretty important so we have full range at Eurorack CV voltages.

EDIT:
Actually, partially answering myself here as the coffee kicks in:

Using a voltage divider calculator, it looks like if we dial the cutoff all the way up with the 12v-sourced potentiometer so it’s feeding out directly 12v, once that passes through the 100k resistor, and is pulled down by the 1.8k resistor, we’re feeding roughly 0.21v into the BC558 transistor.

So if we want this to act the same way (at least with voltage), we can put a roughly 140k resistor between 12v and the cutoff potentiometer (where directly it was sourcing straight from 12v) which should feed this potentiometer ~5v, which then makes it act like the 5V CV input.

After that, we can mod the pulldown resistor (heading to ground) on the transistor to be higher so our 5v has more effect. It looks like 4.5k would be good to get us sitting at the 0.21v that the transistor base again.

Any thoughts? Am I completely incorrect here?

4 Likes
2

Is there such a standard? I was under the impression Eurorack designers notoriously fail to agree on things like CV levels.

Here it says, for Kosmo:

Control Voltages (CV)

  • -10V to +10V typical
  • 0V to 10V for “normal” envelopes
  • 0V to -10V for inverted envelopes
  • 1V/oct for pitch/frequency control.
3 Likes
3

Hmmm as far as I read, mostly we’re looking at 5V these days, but looks like that’s not the case. The problem is then that I have a MIDI to CV converter that assumes 5V is our CV/gate voltage…

1 Like
4

Well, gates are another matter: from the Kosmo spec

  • Gate Signals
    • Off: 0V
    • On: minimum 4V, typical 5V?, maximum ???

(Gates are gates and envelopes are envelopes, you probably shouldn’t try to use one in place of the other.)

Anyway, to take the first three examples I found:

  • Befaco VC ADSR output is 0 to 10 V
  • Erica Synths DIY EG is 0 to 10 V
  • AI Synthesis Looping ADSR is 0 to 7.5 V
4 Likes
5

Hmmm… but an envelope generator is fed often by a gate coming from your keyboard interface to modify it, which to my mind would mean that envelopes should be interchangeable with gates (at least logically) as a gate is just an envelope that is just a square? So I’m not sure why we should jump to a different level once we’ve created an envelope from a gate?

6

I’ve always considered a gate to be an on/high or off/low state (however you define those). From your keyboard (or midi2cv), it just indicates that a key is pressed.

1 Like
7

Right, it’s a signal to turn something on or off. The level is to be ignored, unlike an envelope. Gates and envelopes have entirely different purposes — a gate can start and stop an envelope, but it isn’t itself an envelope.

2 Likes
8

When I was playing with my synth, I was wondering the same thing. I think I could get much cooler sounds out of it if my MS-20 filter would open up more. I built my LFO’s and EG’s to output 8V, which gives pretty reasonable results with my filter, but I would like more.

I’m considering to just replace the 100k resistor by 68k or something. I don’t think bad things will happen if the input impedance is less than 100k.

2 Likes
9

I have similar problems with the B2600. So I built a go between Two-and-a-Half-o-Matic (times 2 and divide by 2) attenuverter to interface with it.

4 Likes
10

In which circuit? Depending on the configuration, changing the resistor value might not change anything except the input impedance.

2 Likes
11

My MS-20 is built using this diagram. Would swapping the resistor work, you think?

12

Assuming your control signal is coming in on the pitch, cv1 or cv2 inputs, those 100k resistors form voltage dividers with the 1.8k resistor, so, yeah, reducing any of them to 68k should increase the sensitivity on that input.

I don’t know how well that circuit tracks to 1V/octave, but changing the 100k on the pitch input will definitely break that 1V/oct tracking on that input.
The cv1 and cv2 inputs are already not tuned as they can be modified by their pots.

3 Likes
13

So I’m coming back to this problem now I have some time, and I just realised that the AD/AR I’m using (LMNC’s simple envelope generator with the big button: https://www.lookmumnocomputer.com/projects#/simple-envelope-generator is only outputting about 5.8v or so fully open. But when I go through the schematic for it, as far as I can see, the basic comparator feeding an amplifier should be trying to push near the 12v rail (probably more like 10v or so) but isn’t.

Has anyone had the same problem here? Until now I’ve mainly been only using this envelope generator to drive my VCA so it was fine, but seems this might be the root cause of my problems with my MS20 filter not opening far enough (though I still want to mod it to have a more sensitive input).

14

In the output there’s a diode, which will drop the voltage some, and then there’s a resistor and LED after the 1k series resistor, which will act as a voltage divider. Depends on the LED but a quick LTSpice check gives me 5.99 V at the output with 10 V going into the diode, so what you’re seeing sounds roughly right.

That’s indeed a rather low amplitude. I think most envelopes are 8 to 10 V at maximum. Removing the LED or moving it (and its resistor) before the output series resistor would give you something around 9 V, probably.

1 Like
15

Ah yep, that’s true about 0.7v from memory, but yes you’re right I hadn’t been thinking about that indicator LED. I’ll just cut it from the circuit for now, but I guess I could implement one of the CGS indicator circuits in its place, but I guess the nice thing was that this LED faded nicely to show you the envelope, whereas it probably won’t be quite so clear with a transistor switching the LED.

16

That did it! 10v out now, but still the MS20 filter isn’t fully opening, so now to increase the sensitivity on the CV input.

As an aside, when I hear the MS20 filter opening on a slow envelope, I can hear a little “bzzzt” at the start of the envelop where I can hear some almost bypass sound happening with a little sizzle - is this standard for this filter?

EDIT: I put 83k of resistance on the MS20 input and now it mostly opens, I might drop it a little further (maybe 68k or something like that) so it can handle slightly lower than 10v for other possible modules in future, but for now it works. Still that little bypass type sound bleed happening, and if I roll off the frequency to the bottom, there’s a slight pop each time an envelope starts. I can raise it to a point where it’s just silent, but this doesn’t solve the slightly bypass sound as the filter begins to roll up which is strange as I don’t seem to hear it when I manually just roll the frequency potentiometer up…

1 Like
17

I have a slight pop too with my envelope generator, its one of sams from the shop. From what ive heard on here somtimes you get a bit of noise going back into the power strip or whatever you are using and it can bleed into other modules slightly.

(I can hear the triggers?)

Just as guess though.

Hope you get it sorted :slight_smile: