[sdiy] question about a Mankato filter

David G. Dixon dixon at interchange.ubc.ca
Fri Jun 4 06:25:05 CEST 2010

> I didn't see a response to this yet.  I used a 1 KHz square wave and
> adjusted the output (input to Mankato) to 10.0 V pk-pk.
> Input:   10.0V pk-pk
>    6 dB:   5.0V pk-pk
> 12 dB:   6.9V pk-pk
> 18 dB:   9.9V pk-pk
> 24 dB: 13.7V pk-pk
> Yes, your assumptions are correct.

Thanks, David.  David Ingebretsen also did the experiment, and replied to me
offline.  I guess I should have mentioned it before, but it doesn't hurt to
have two confirmatory responses.

Obviously, I'm not too enamored with the design of this filter.  The Mankato
seems to be optimized for use as a quadrature oscillator, but the filter
outputs shouldn't be giving these disparate levels.  This is the result of
using dividers and inverters at various gain levels to equalize the sine
waves during self-oscillation.  Distributing the resonance gain across the
four stages overcomes this problem, but does require another quad VCA, and
hence, increases the cost of the filter.  I guess it's a tradeoff.
Personally, I'd rather pay the extra $3.50 or whatever to have a unit which
is optimized as both a filter and a quadrature oscillator.

BTW, I really like your scope graphic.  What are you using?

More information about the Synth-diy mailing list