Guitar Steiner filter

[tca]

Hi, here is a guitar Steiner filter which is not voltage controlled! It uses two log 1M pots to do the frequency sweeping, it has, as the original Steiner filter, a HP, LP and BP inputs. Here is the schem



Some considerations about it and the questions later on. I'm powering it with two 9V batteries, still waiting for the voltage pump IC. I've never build a filter like this before, it is very illuminating to ear plain music with it, you can really dissect the audio spectrum. Using a distortion pedal before it you can really, really, change the finger print of you sound!

I'm still wrestling the calculations to get a complete understanding of the filter, but mean while I did some simulations. You will see in the picture above a table of R (which controls the frequency of the the filter) as a function of the frequency. It is a non-linear relation, after a least square approximation it turns out the data is well fitted by

R=exp(10)/f.

Any clues about this relation?

So a 1M log double pot would balance this non-linearity and gives a good sweep as the cursor changes position. Another thing is that you probably could put it inside a wah-wah foot pedal (I never liked the automatic sweep of a synth when used with a guitar).

Have fun, cheers.

P.S. (edit)
Here is a sound sample with distortion from a pair of diodes in the feedback loop of the first opamp in parallel with P1.

[WhenBoredomPeaks]

Is that seriously a steiner?

[tca]

Is that seriously a steiner?

What do you mean by that?

[slacker]

Looks like a Steiner to me, the original uses transistors but there's plenty of opamp implementations of it.

[chptunes]

Interesting schematic tca.  Is "Steiner" the name of a classic product.. or a filter type.. or both?

[tca]

Interesting schematic tca.  Is "Steiner" the name of a classic product.. or a filter type.. or both?

Here you go:
- http://yusynth.net/Modular/EN/STEINERVCF/index.html
- http://yusynth.net/archives/ElectronicDesign/N-Steiner-VCF-1974.pdf

[tca]

Here is another sample, it is a typical sound of a filter but I can get enough of hearing it. It can go to really high/low frequencies and very selective. You should use headphones to listen to this, it starts with LP, then HP and finally BP (you can hear the self resonance).

Cheers.

[John Lyons]

Interesting. I'll have to try this.

[slacker]

Sounds very nice, this is my favourite filter I've got two in my synth. Have you tried it with diodes or LEDs as clippers in parallel with P2, I have those in mine and at certain settings you can get interesting non linear things happening.

[gcme93]

Just to check with your relationship, do you mean R = (e^10) / f ?

Cool circuit anyway!

[tca]

> Have you tried it with diodes or LEDs as clippers in parallel with P2, I have those in mine and at certain settings you can get interesting non linear things happening.

Yes, two 1N4148 in parallel with P2. But I prefer the sound without them.

> Just to check with your relationship, do you mean R = (e^10) / f ?

Yes, e≈2.71828 (Nepper number), i.e.,  exp(1)=e.

Thanks for your comments.

[slacker]

I think the original is supposed to turn a linear control voltage into an exponential frequency response, does the way it does that using transistors help to explain what you're seeing?

[tca]

I think the original is supposed to turn a linear control voltage into an exponential frequency response, does the way it does that using transistors help to explain what you're seeing?

I don't think there is a linear control voltage in any case. The answer to the non-linear relation between R and f comes from the transfer function, if I could calculate it, I could have a straight answer. A typical low-pass filter has the same non-linear relation between fc and R (on-line ref: Analysis of the Sallen-Key Architecture - Application Report TI), something like

fc = 1/(2*pi *sqrt(R1*R2*C1*C2))  

The coefficient is, for now, undetermined, have to make the calculations. Any ref.?

Cheers.