Hey Synth-builders,
I’ve built two MS-20 filters and I noticed that the sweep range for the CV input is a bit limited compared to the actual cutoff knob. This makes a lot of sense if you look at the schematic
The cutoff input voltage varies from -12 to +12V (I’m not doing the 15V like suggested by the schematic) while I don’t think that any LFO’s or ADSR’s cover that range (nor are they supposed to). My LFO’s and ADSR’s go between 0 and 8 volts.
I could tweak the circuit by replacing the 100k CV resistors by 47k to make the circuit more sensitive to CV. That way the CV attenuator will also have more impact. My understanding is that the filter cutoff is ultimately determined by the sum of all currents flowing into the first transistor. Is that right?
I was wondering whether any of you made these kind of tweaks and how you get the most out of your filter musically? A bigger range allows you to use your filter as a gate since the filter can close at low input voltages. That can be extremely useful since you could potentially do without a VCA in your synth voice.
1 Like
Typically a filter’s CV is calibrated to at least roughly 1 V/oct with the attenuator open. That means if you use the same keyboard or sequencer or whatever to control both a VCO and its associated VCF, you’re getting about the same waveform out of the VCO regardless of pitch. (In the absence of the keyboard CV going to the filter, in low pass mode it’d filter hard on high notes while leaving low notes more or less unfiltered.) It also means if you turn the resonance up to self oscillation you can control that oscillation with a keyboard and it’ll be more or less in tune.
If you change the sensitivity to CV, then the cutoff can’t track keyboard CV to give you consistent filtering of a VCO or to make it usable as an oscillator. Of course if you don’t care about either then it’s not an issue.
If you have the front panel space you could have two (or more) CV inputs, one 1V/oct (maybe even with no attenuator) and one more sensitive (with an attenuator). Just like what’s shown in the schematic but with one CV’s 100k reduced, as you say, to for instance 47k and the other left at 100k.
Another option is to leave it at 1 V/oct and use a boost module to increase the range of the CV as needed.
3 Likes
Interesting. But 1V/oct does not seem supported because otherwise there would probably be a trimpod somewhere in the circuit. Does anyone know if tracking can be supported by the MS-20? If not, I might as well lower that resistor to get some punchier filter curves out of that thing 
I also found this topic discussing the same issue. MS20 Filter CV input level - #16 by jessecakeindustries
I did say “at least roughly”. There’s also little to no provision in the schematic for temperature compensation. The assumption is that you want approximate 1 V/oct response but that it isn’t worth trying to make it as precise as is needed for a VCO; if you want your self oscillating filter to be used as a really accurate VCO you’d have to make those modifications.
2 Likes