A few comments to your mail, Bruce as well as to George's original mail:
>Synth: Korg PS-3100
>Filter: three 6db/oct BP
>MOTM: 410
One MOTM-420 (per voice), for the VCF(s) in the PS-3100, plus one
MOTM-410 (for all voices in common).
>Synth: Oberheim Ob-Xa
>Filter: multimode 12/24db/oct LP w/res
>MOTM: is this simply a subset of the Matrix-12?
Just use the standard MOTM filters without overdrive:
MOTM-440 (bass enhance off) for 4-pole and the soon to come
SEM-style VCF for 2-pole. (While nothing in the world will replace
anything else to 100.0%, this will be very close. Oberheim tried
to emulate their own SEM VCF with the OB-X 2pole mode, after all.)
>Synth: Roland Jupiter 6
>Filter: multimode 24db/oct LP/HP or 12db/oct BP, w/res
>MOTM: future?
This is from memory (must check the schemos), but I think the JP-6
has two state variable filters in series. They don't have individual
control,
you just select the configuration and the (common) cutoff frequency.
So I'd say you can emulate it quite well with a pair of the announced
SEM-style filters.
Which brings me to the following (long quote included):
> OSCar was designed by Chris Hugget, the person who was behind the EDP
> synths before. The EDP Wasp used a pair of OTAs to produce a multi
> mode 2 pole filter. This filter exists as a module for some german
> company today, as well as Jurgen's schematic as well.
>
> The OSCar filter put two of these filters in series to create a 4 pole
> multi mode filter, with a couple of interesting twists.
>
> The filters funtioned as 2 12 db sections, and each section could have
> both low pass and high pass output. Chris could have designed BP and
> NF in there as Jurgen did, but wait for the interesting trick.
>
> Because the whole thing was under CPU control, there were some
> interesting manipulations that could happen in the circut. First of
> which is a knob called "Separation". The 2 12db stages could have
> their cutoff frequencies controlled independantly. By moving this
> knob the FC would be moved up or down (depending on stage and setting)
> to widen out the resonance peak. At high Q levels this was (and still
> is for those of us with OSCars) lots of fun.
>
> MODE STAGE 1 STAGE 2 SEPARATION
> Highpass HP HP +Stage1 / - Stage2
> Lowpass LP LP +Stage1 / - Stage2
> Bandpass LP HP -Stage1 / + Stage2
>
> Now, looking at the circuit, there are a lot more things that could
> have been done. Each stage is really 2 6db OTA based things, so it
> should be possible to use 3 of the 4 to get an 18db response, or just
> 2 of them to do 12db. Again because the CPU is supplying the CV for
> cutoff and resonance, it should be possible to have nice separation
> effects in each mode.
No need for a CPU here. A pair of SEM-style VCFs and some (analogue)
processing of the CVs will do. What you need is to send one CV to
both cutoff modulation inputs, and a second CV to one input directly,
and to the other inverted. You have all degrees of freedom here - that's
part of what Modular is all about. BTW, that famous Serge filter with two
resonant peaks does nothing else (to my knowledge, at least).
>Synth: Yamaha CS-80
>Filter: 12db/oct LP w/res and 12db/oct HP w/res in series
>MOTM: is this akin to the Korg MS and therefore covered by the 420?
Not exactly. A better approximation will (surprise !) be done with the
SEM-style VCF, I mean with a pair of these. I have done some research on
the CS-80 VCF some months ago, and it's really just a pair of state
variable filters, ∗but∗ with both, cutoff and resonance ranges limited
to rather conservative values. So, don't turn the resonance all the way
up, and limit the cutoff range, and you should get a nice CS-80 VCF
emulation with two SEM-style VCFs in series.
JH.