Blades v/oct circuit design - not understanding it!

Hi - I’m working on a stereo multi-mode filter eurorack module (similar features to the Elektron Analog Heat) and using the 2164.
I’ve been using the Ripples schematics as inspiration as well as the ssi2164 datasheet, the SSI AN701 app note and @pichenettes papers - the smr analysis, pole mixing and the 2164 svf one. I have a design all sorted out but then looked at Blades’ design to see how Blades worked and got very confused.

My understanding of the transfer function is that the frequency will double (exponential response) with each approx -200mv change in control voltage which is consistent with my understanding of ripples (1v input is converted to about -200mv by an inverting opamp and trimmer in the loop which is then fed to the control port of the v2164). The initial frequency is determined by the RC time-constant (typically about 30-33k and 220p).

I then looked at Blades and am really confused - the inverting opamp converts 1v cv input to -300mv. There is then a series trimmer (2k) into the control port of the v2164 (input impedance of 5k in the ssm2164 and v2164) so the trimmer and control port act as a voltage divider but the control voltage seen by the 2164 would be between 300mv and 215mv ish depending on trimmer value… (unless the effective resistance to ground in the v2164 is closer to 4k) I’m just really confused why the trimming isnt around ~200mv and how this then gives a nice v/oct response - I’m obviously either totally misunderstading the maths of the filter response or how the Blades circuit works (or just being very dumb which is most likely!) I’d understand if the feedback resistors were 20k but am obviously totally missing something important !

Anyone looked at this and can help me out with an explanation ???

The trimmer is hooked to two V2164 inputs, so the input impedance is approx 2.5k. Assuming the trimmer is in the middle (resistance = 1k): 300mV * 2.5 / (1 + 2.5) = 214mV, close to the 200mV you mentioned.


Ah, of course !!! Makes total sense - I really was being very dumb !!!
Thanks a lot for the amazingly quick reply !!!

And if you’re wondering why this funny business with the trimmer not being in the op-amp feedback loop (or before), this is for the cutoff coupling feature: F_CV1_PRE must not have any trimming for the coupling to work!


Yes I’d wondered! The thought had crossed my mind :slight_smile:
Thanks again - it’s just amazingly kind of you to reply to these sorts of questions !