[sdiy] Shruthi 4PM (was re something else...)

[Neil Johnson]

Tom Wiltshire <tom at electricdruid.net> wrote:
> I don't know whether it's "weird in theory", but there are clear differences in practice.

Putting diodes parallel to one of the integrating capacitors is not a
sensible route, since the leakage current through the diodes will ruin
low frequency performance.  Even zener diodes.

In my 3-pole 2164-based VCF I was able to get it to self-oscillate
from a few Hertz up to about 16kHz with visually nice sine wave at
constant amplitude (maybe 1% distortion, haven't measured yet).  But
not with diodes across capacitors.

My next project after the current one (the list is long) is to develop
the filter further to get the self-oscillation performance an order of
magnitude better (I'd like to get to below 0.01% distortion).  Diodes
just won't cut it.

Neil
--
http://www.njohnson.co.uk