[sdiy] Filters

[Grant Richter]

> Now I don't KNOW how the diode acts as a resistor... but Ohms law assures
> us that it
> does....

Well, it turns out that ANY non-linear conductor will change it's AC
reactance with a change in DC current flow. I'm not entirely clear on this
either, but you can use it to advantage with diode, transistors and... MOVs.

See
http://www.musicsynthesizer.com/Waveshapers/movs.html

for a filter example.

I'm prototyping a transformer coupled MOV ladder filter. Just because it can
be done. I call it the "Anti-Filter" because it violates every engineering
precept of sensible design practice.

Diodes and transistors have a very short non-linear curve, where MOVs have a
long one. I wonder if that will have any audible artifact, what the heck.
Lack of matching can be compensated by adding a nulling signal from the CV
control into the audio path, rather than relying completely on balanced
nulling.

This is why the Stiener design is so complicated, with no isolation between
DC and AC in a non-linear transconductor, the CV signal has to be made
common mode to cancel it from the AC path.