[sdiy] David's sortof biquad filter

[Tim Stinchcombe]

> I didn't demand that he put in a more complicated expo control source 
> since he wound up spending a huge amount of time fighting the 
> effects of 
> DC offsets and I took pity on him. He'd change his control 
> current, and 
> the DC of the output would swing all over the place.

Looks to me like most of the DC offset problem will arise from the 1meg at
the second integrator: assume no AC input to the OTA, so the feedback cap is
o/c, then the magnitude of the output will be 19.2*Iabc*Vin*10^6, so
assuming a max Iabc of approx 350uA and max output swing of 13V, then
Vin=13/(19.2*350*10^-6*10^6)=2mV of DC at OTA input for the op amp to
saturate. If the input bias current is of the order of 2uA, then if the
positive OTA terminal was merely grounded, you'd be halfway there with the
drop across the 470 at the negative input...

Doesn't look like that resistor needs to be there anyway (and I'm also
assuming the sketch circuit erroneously shows the OTA out to the positive of
the op amp, rather than to the negative with the pos to gnd). Have you
worked out the transfer function at all? Sound samples?

Tim
__________________________________________________________
Tim Stinchcombe 

Cheltenham, Glos, UK
email: tim102 at tstinchcombe.freeserve.co.uk