[sdiy] Filter Musings

[JH.]

>> Excepting the trivial case of two identical filters, you will get a more complex behaviour than with a single filter.
>> Two 4-pole LPFs in parallel form an 8-pole filter indeed. But you also get zeroes, so there will be no 48dB/Oct slope.
>>
>> Serial connection of two 4pole LPFs without zeroes means *multiplication* of the transfer functions -> poles only -> 48dB/Oct 
>> slope
>>
>> Parallel connection of two 4pole LPFs without zeroes means *addition* of the transfer functions -> finding the common denominator 
>> to
>> perform the addition means you get "jw"-terms into the numerator -> zeroes are created.
>>
>> JH.
>
>So say I take a transistor ladder core and put it in parallel with a diode ladder core... is that enough of a difference so that 
>they're not identical? Or should it >be, for example, two TB-303 cores but with differing values of caps?

Two identical filters, set to different cutoff frequencies, are "different enough" already.
Whether such a setup is the most usefull or not, is a different question. Moog Voyager has it, and I have it in my JH-5. But a 
combination of two entirely different filters certainly has more flexibilty.

JH.