[sdiy] stringfilter

[Dino Leone]

Magnus and Juergen,

Thanks so much for pointing me in this direction 
and for Juergen's explanations!

I would love to try to build it using the 
textbook GICs. But there's still a lot that 
confuses me... first, why 20 bands per channel? I 
assume each of the band is a bandpass filter with 
its own resonance frequency. But I'm only aware 
of the 3 main resonances for each string 
instrument. So what would be the values of the 
other filters?
And, would one wire the 20 bands in parallel and sum them together?

Furthermore, (and this leads back to the initial 
question), assuming a completed filterbank 
providing the desired resonances, what else would 
be necessary to get the "stringy" character that 
Juergen demonstrates on his webpage?

And I agree very much, it would be *so* 
interesting so hear Ken Elhardt giving us a few 
hints how he generated his string sounds!!

Many thanks and best regards,

Dino





>Well, to start off, the General Impedance Converter (GIC) is very simple.
>You just put the resistors and one of the condensators in a string like this:
>
>                    |\
>   +----------------|+\___
>   |              +-|-/   |
>   |              | |/    |
>   |  ___     ||  |  ___  |  ___     ___  |
>o-o-|___|-o--||--o-|___|-o-|___|-o-|___|-||.
>           |  ||  |               |       |
>      R1   |  C1  |  R2      R3   |  R4
>           |   /| |               |
>           |__/-|-+               |
>              \+|-----------------+
>               \|
>
>The inductance becomes
>
>     R1*R2*R4*C1
>L = -----------
>         R3
>
>Now, just toss a second capacitor on the left terminal either in series or in
>parallel with the above circuit depending on your needs and you got yourself
>the high-Q resonator using the GIC Jürgen is talking about.
>
>To interprent what he is saying, that the 
>resistors and capacitors are equal it
>becomes
>
>R1 = R2 = R3 = R4 = R and C1 = C2 = C
>
>we end up with
>
>      2
>L = R C
>
>           1                 1              1
>f = -------------- = --------------- = --------
>     2*pi*sqrt(L*C)   2*pi*sqrt(R"C")   2*pi*R*C
>
>So, calculating R and C for a given frequency is really trivial under these
>conditions. Similar benefits could be done by choosing R3=R4 to be some other
>value common for all resonators (say 100k). The only R1=R2=R.
>
>It you need lower Q, just add a resistor in parallel with the capacitor.
>
>Now, with this at hand, you realize that it is fairly trivial to make a bunch
>of these on a board and you only need a handfull of op-amps on the input and
>output of this to make it alive and kicking. Actually, now you get more out of
>looking at Jürgen photos. Each IC is a dual op-amp and they are neatly put in
>two 4x5 banks. Then you have 3 extra ICs per bank.
>
>Cheers,
>Magnus