[sdiy] Pole Dancing

[Richie Burnett]

> You bring up an interesting point.  What's the effect of the Moog poles 
> being mistuned a bit?  The answer is, not much.  Worst case, you may need 
> a little more feedback to get the job done.

Exactly.  The effect is much more pronounced in the TB-303 filter though. 
Here there is no buffering between stages in the ladder and the open-loop 
poles are spread over more than a decade in frequency!  That's why the 
TB-303 filter requires so much gain in the feedback path to get it anywhere 
near to self-oscillation.

> Even if the tuning is substantially off, the two right poles will pair up, 
> and the two left poles will pair up, and they'll split off into a very 
> similar X pattern.

Many people have pondered the effect of the bottom capacitor in the TB-303 
ladder filter being a different value from the others.  As you say, it has 
little consequence other than that you need a bit more feedback gain to make 
it howl.

> Also note, a Moog Ladder pole pattern is nothing like a Butterworth 
> filter, where all the poles contribute roughly equally.

Yes, from a technical point of view once closed-loop feedback is applied the 
response doesn't resemble any of the classic textbook filter responses.

-Richie,