4PM filter tuning, some experimenting

I had mentioned in another thread that I was going to try tuning a 4PM filter board that I’ve used normal capacitors on. I fired it up and got some interesting sounds but not what I expected. Before I start fooling with it, I thought I would ask, probably Olivier because he has fooled with the cap values, if this had happened to him.

I set up a patch as recommended in the manual, but the sound I’m getting is actually just noise, with a very faint sine wave under the noise. Sounds like pink noise.

I’m assuming the standard answer is, “put the recommended caps in” but I’m curious to know, if possible, where the noise is coming from, and if it is caused by the non-standard caps or something else I screwed up.


The noise is very faint compared to the normal signal you send by playing an oscillator through the filter, right? In this case what happens is that the filter is not self-oscillating - which could be due to the wrong cap tolerances but also to other assembly errors…

Thanks Olivier and yes it is very faint. The normal patches are fairly distorted. Since I have gone through this with you before I thought I would check the values of everything else just in case and found two potentially incorrect resistors. One is marked on my PCB as supposed to be 16.5k but on the build page it is10k. It sits between two 1.0k resistors near the LM13700. Which value should I use?

All of the other values look okay. The other incorrect resistor was the last one before the two 220 resistors for the eyes.

Interested to know about this 16.5k resistor, since I just soldered whatever the board was asking for, and have some issues too, maybe I solve them through this swap to 10k?

I’m lost.

BOM, PCB silkscreen and schematics say that R30 is 10k. Where does it say it is 16.5k?

There is one single 16.5k resistor - R57 - it’s in the pole-mixing circuit for the all-pass mode.

I´ll check and report, I don´t have my 4pm here right now.

Maybe Piscione does?

Maybe I’m going nuts or just not paying attention or looking at the wrong photo of the board or who knows what else?

It looks like you have an older version of the board. No slot for the dog eye, vertically mounted resistors for the to LEDs, and this resistor value change. Weird. How/when did you get this board? It doesn’t look like it uses the chips I packed with the “semi-kits” I used to sell.

Use 10k in place of the 16.5k resistor left empty.

Thanks Olivier. Someone gave me this board when I bought another Shruthi from them.

I might have sold or given away the 2 or 3 pieces of the prototypes I had - with a note about the silkscreen/design changes.

Too bad I didn’t get the note too! I think I’m going to change the caps to your recommendations.

I put the recommended caps in and the 10k resistor. I am getting sounds and the patches sound close to what they are supposed to sound like but I cannot get the filter to self-resonate as suggested in the setup procedure. I also probed the four points as suggested in the setup and get very low voltage readings that do not stop fluctuating. These are between 0.01 and 0.07 volts.

I’ve already checked all of the resistor and capactor values and made sure the electrolytic caps are properly oriented, guess I’ll double-check everything.

This may be off topic but the confusion of how/when of this board kind of fits.

For self manufactured boards, are we allowed to put a personal Hallmark on it? An example would be a filter board with a few modifications. Printing a Hallmark would identify the board as not manufactured by MI and it would be accompanied by a modification version number or other significant information.

Its Open Source but what are the limitations on this?

The four points to probe are supposed to carry audio, so a multimeter won’t say anything useful.

@kreeper_6: cc-by-sa

Olivier, that makes sense. I figured that but the way I read the instructions seemed to hint at voltages, “A quick check if something goes wrong: Probe the signal at those points – their voltage is the signal 1, 2, 3, and 4-pole filtered.”

This is interesting 'though. If I just play a patch and fool with the cutoff and resonance, it seems to sound fine. The resonance acts as I would expect, but if I turn the oscillators off and just play a note to get the filter to sound on its own, all I get is a thump, and only if I turn the volume up on the mixer quite a bit.

> their voltage

It’s voltage varying at audio frequency, that needs to be probed by a scope or to be sent to your mixer/audio interface/amp. Maybe you have swapped the diodes?

Diodes are oriented correctly. Two are larger than needed because that was all I had around but they are the correct values. I could change those if you think it would help.

Changed one of the over-sized diodes, no difference. Olivier, can you tell me if there are any specific components I should check that would influence the ability of the filter to self-oscillate?


IC2A and all parts around it generate the current that gets into the resonance VCA.

IC3OTA1 and all parts around it are the resonance VCA (OTA).

D2, D3, D4, D5 and the 4053N (IC10) that switches them on/off are also responsible for shaping the resonance.