[sdiy] VCR - Voltage Controlled Resistor

HOSHUYAMA houshu at muj.biglobe.ne.jp
Mon Jul 19 19:03:12 CEST 2004


Hello Scott and List,

Maybe I don't follow all the discussion in the weekend.
But, I think I can contribute something.

Scott Gravenhorst wrote:
> harrybissell <harrybissell at prodigy.net> wrote:
>>I suggest you hie hither to Osamu Hoshuyama's site. He has some ideas there on
>>FET linearization iirc.  One of them is truly a new idea that has not been
>>commonly used !
> 
> I've been there, he has an SVF using 4069UB integrators, but with an M5222 dual
> expo OTA (unobtainium), essentially as WASP type filter.  He also has a 4069UB
> integrator SVF using MOSFET OTAs built from more 4069UB (but says they are tricky
> to make them work).  He doesn't seem to use FET VCRs in any of his filters. 
> Perhaps I know why now.

Actually, at the beginning of my CMOS VCF project, I tried to
design some VCF structures using 4049 as shunt VCRs, but
failed in those days.

The reason I used inverters as OTAs in my CMOS VCF is that I was
inspired by Pearl phaser PH-03, where an MC14069U is used as VCRs
based on an OTA technique.  I thought the structure is the only
way to obtain "smooth" and "wide-range" sweep.

  http://www5b.biglobe.ne.jp/~houshu/synth/VcfMos0206.GIF
  http://www5b.biglobe.ne.jp/~houshu/synth/CmosOta.GIF


I've seen a schematic that uses 4007 as VCRs in the SVF.  It's an
envelope-controlled filter of Electroharmonics, which may be
useful, though I don't think it has sufficiently wide sweep-range
for a synth VCF.

If I have "matched" and "floating" FETs (MOSFETs), I'd try a 4-pole
VCF based on the LPF structure in the following schematic replacing
the OTA by a FET, and the opamp by a CMOS inverter.

  http://www5b.biglobe.ne.jp/~houshu/synth/Vcf0210.GIF

In the structure, the signal level across the FET is small
thanks to negative feedback, which leads to less distortion
in pass band.  My linearizing technique can also be applied.

  http://www5b.biglobe.ne.jp/~houshu/synth/PhaseFet0205.GIF

Linearizing the relationship between CV and the resistance of
the FET is possible if master-slave technique is applied.

I'm busy these days.  So, don't expect quick response please.
Writing English is not easy job for me  ;-)

Sam Hoshuyama
Saitama, Japan




More information about the Synth-diy mailing list