[sdiy] 2164 Korgasmatron VCF

[David G. Dixon]

> LOL...  After reading your description, I happened across an old
> electronic workbench file where I had the same circuit (with three
> antiparallel
> diodes btw).  I see how that would work, the big decision is where
> to locate the circuit.

Mines in the resonance amp feedback loop.  Since I'm using VCAs and
integrators (rather than OTAs and buffers), my resonance amp is inverting.

> I was thinking you put the variable resistance in series with the zeners
> (which would not work correctly as the zener clipping would get
> progressively
> more 'squishy').

That doesn't happen with mine.  Really, mine is (almost) straight out of
Jung's cookbook.

> I've tried a few different methods with the EFM version, including making
> the
> input resistor into the resonance OTA into two series resistors (like 10K
> and
> 50K) and using single antiparallel diodes. The 10K gives a series
> impedance
> that the diodes can work against (without loading the opamp LPF output)

My Korgasmatron (like the original) does not have VC resonance -- just a
knob (actually, two knobs per filter) to crank -- so it's just an opamp
inverter.  Perhaps that's the main difference...?

> Now how are you getting the notch response ???  (I'm getting a bandpass
> :^)

No, it's definitely a notch.  Only, not much of one.  It's basically a flat
response with a dip down to -6dB at the cutoff frequency.  It's just the
filter with both HP and LP input options engaged simultaneously (just like
the MOTM 420, and a couple others out there, with similar results).

> I have the two EFM filters set up with a Moog-like coupler, one CV moves
> both
> filters in the same direction, the other sets the difference between the
> cutoffs.
> Its an interesting module.

That would be cool.  It might be worth consisering a little 1U "Korgasmatron
helper" (perhaps I could call it the "sex toy") for processing CV signals
prior to feeding them to the two filters.

> The neatest thing (imho) about the MS-20 filter is that the resonance
> takes place
> at the zero cross of the input waveform, not at the peaks like most
> filters...

Yes.  I wasn't aware of that, but I have noticed that in my simulations.  It
makes a resonant filtered saw, for example, look very different than it
would through, say, a COTA filter, with much higher resonance peaks right at
the sharp edges.