Putting filters into feedback loops

[WhenBoredomPeaks]

I checked out some state variable filter designs and it seems like that most of the time the resonance control is achieved by feeding back some portion of the band pass output into the input of the filter section. (i've seen this in the mutron iii/neutron/meat sphere, mxr envelope filter, polivoks filter and in some other filters)

Then it seems logical to me that if i put a bandpass filter which does not seem enough resonant to me into a feedback loop that the filter will be able to go into self oscillation. (which means maximum resonance i think) Because essentially i just feedbacked it's bandpass output into the filters input.

Continuing that thought, could i make the Meat Sphere more resonant by increasing or removing that 1k5 resistor which bleeds some of the feedbacked bandpass signal to VREF? (which would be ground/0V in the synth world that is why i think that 1k5 resistor probably decrease the amount of resonance)


(the leftmost 10k resistor is where the unfiltered signal enters the filter)
(by the way there is a small mistake in the pdf, that vertical line between the opamps seventh pin and SW4BP is connected to the horizontal line which it crosses, there should be a dot there.)

The image is from this pdf:
http://musicpcb.com/wp-content/uploads/2011/07/Meat-Sphere-build-PDF.pdf
Which is hosted here:
http://musicpcb.com/pcbs/meat-sphere/

[slacker]

The colour control is negative feedback the more signal you feed back the less resonant the circuit is. So if you make the 1k5 resistor smaller that will give more resonance, if you make it small enough or just replace it with a jumper it might even self oscillate.

[WhenBoredomPeaks]

The colour control is negative feedback the more signal you feed back the less resonant the circuit is. So if you make the 1k5 resistor smaller that will give more resonance, if you make it small enough or just replace it with a jumper it might even self oscillate.

I jumpered it and it makes the resonance noticeably higher, i'd say it was quite close to self-oscillation, on the other hand this mod makes the resonance pot (i think it is called color originally) unusable, so this mod should be made switchable.

[CynicalMan]

How is it unusable? Except for the high-resonance end, it should act the same way as before.

[PRR]

[WhenBoredomPeaks]

How is it unusable? Except for the high-resonance end, it should act the same way as before.

About 95% of the pot's turn have no inpact on the resonance, then in the last 5% the filter goes silent. Btw i did the resonance pot mod a while ago which means that the lug that is connected to pin 3 of the opamp and the middle lug is connected together.

@PRR

It seems to me that decreasing that resistor would mean that more of my unfiltered signal would "skip" the filter, what exactly that resistor does?

edit: or i should look at that resistor from the other way and it actually feeds back some signal to the input of the filter so when i decrease it i get more feedbacked signal at the input?

[PRR]

> unfiltered signal would "skip" the filter

Output of U1C is "zero" impedance. A 10K resistor can't influence it directly.

> it actually feeds back some signal to the input of the filter

Going that way it mixes with several other 10K resistors. It can influence that.

But why think? Isn't it quicker to tack another 10K across that one (5K) than to type about it?