Filter Question (was LM13700 Filter noisy)

[grapefruit]

The filter I'm building doesn't need to be voltage controlled, I was only using the LM13700 so I could use a single gang pot for easy control of frequency.

I'm now planning on building an op amp based 2 pole state variable filter. From my calculations this (as any filter where f=1/2*pi*R*C) would really need an antilog pot to get a linear feel to the frequency adjustment. Am I correct in assuming this?

I'm thinking I may have trouble finding an antilog dual gang pot...

At least the filter will be quiet with low noise op amps.

Thanks,
Stew.

[Transmogrifox]

For a music related filter, linear sweep may not be what you're looking for.  Something that feels "linear" would be an exponential sweep in frequency.

Your guitar's low E is tuned to 83 Hz.  The next octave up has a fundamental frequency of ~166Hz, the next octave up is now at 332 Hz or so.  You see the 2^n power forming here.  If you wanted to drive a filter with a linear control (linear pot, for example), you would want an accelerated frequency change per increase in frequency so that you give about the same amount of time to accent each musical interval throughout the sweep.  

This 1/x curve done backward (starting at large R and decreasing) produces about the right sweep to my ears.  The problem I have with my LED/LDR modulated pedal is that the exponential curve on the LED kinda fights the filter's "accelerated' frequency change per resistance change.

You won't have this problem with a linear pot.   You can also try a modified version of the state variable biquad filter that relies on one resistor for center frequency--the others define Q and gain and such.  This is the constant bandwidth filter.  I haven't tried it yet, so I don't know what it's like for guitar.

[Transmogrifox]

For a music related filter, linear sweep may not be what you're looking for.  Something that feels "linear" would be an exponential sweep in frequency.

Your guitar's low E is tuned to 83 Hz.  The next octave up has a fundamental frequency of ~166Hz, the next octave up is now at 332 Hz or so.  You see the 2^n power forming here.  If you wanted to drive a filter with a linear control (linear pot, for example), you would want an accelerated frequency change per increase in frequency so that you give about the same amount of time to accent each musical interval throughout the sweep.  

This 1/x curve done backward (starting at large R and decreasing) produces about the right sweep to my ears.  The problem I have with my LED/LDR modulated pedal is that the exponential curve on the LED kinda fights the filter's "accelerated' frequency change per resistance change.

You won't have this problem with a linear pot.   You can also try a modified version of the state variable biquad filter that relies on one resistor for center frequency--the others define Q and gain and such.  This is the constant bandwidth filter.  I haven't tried it yet, so I don't know what it's like for guitar.

[grapefruit]

I'll try it with a linear pot when I get home and see how it goes. I need a highpass and a lowpass filter. I was going to use sallen key filters but I want to be able to control freq and Q indepependantly so I thought a state variable design would be the easiest. If anyone has any other ideas I'd be glad to hear them.

Stew.