[sdiy] Multiplier built from op-amp

[jh.]

>     But, No, there is no way to make a multiplier out of a regular
opamp...which
> is a shame when you think about it.

This is true (of course), but it reminds me of a very strange *VCF* that was
built around a quad opamp. I'm not sure about the part number from memory
(LM349 ? LM346? - a quad opamp similar to the LM324, but with a
bias current that could be set on two external pins. It was intended for
use with a resistor to set the bias current in the 100uA or 1mA range
(or something like that) - optimizing for speed or low power operation.

Farfisa used (abused) this for a VCF in one instrument at least
(Synthorchestra
probably). Connected a expo current source to the programming pins.
So the differential input stage of the opamp would act like an OTA,
the (common emitter) gain stage with (internal) phase compensation
capacitor would act like an integrator (yes, even picofarads work
when the current is accordingly low !), and the output stage - unity
gain buffer - is familiar from ordinary OTA VCF designs as well.
Feddback applied with high resistor ratio, and dividing down the
input voltage to match.
It's surprising that it works at all. I tried it last year, and while it's
ok
for cutoff frequencies of a few 100 Hz or more, getting a deep bass
sound was impossible. Current too low, parasitics would start oscillation
at higher frequency, stuff like that. But for a slight filtering of the
top end of a sound it was surely ok. Things like that always amaze me.
Brilliant idea, even if the sound would not be comparable to a "good"
VCF, and even when that opamp costs as much as ten 3080's *today*.

JH.