HPF + soft clipper

[fryingpan]

Some bassist friends of mine inquired about a HPF (which makes more sense than you might think on bass guitar). I got inspired by Rod Elliott's Project 197 - HPF with adjustable Q and since I had an extra opamp, I added in a mild boost + bypassable soft clipper that would only basically engage on peaks for some compression (since the filter can get somewhat resonant, for safety).

This is my design:



It's a 12dB/oct sweepable filter, plus a single-pole high pass at 30Hz or so, to limit subsonic frequencies even better.

This is the LTSpice project: https://drive.google.com/open?id=1FOj8PN_x17MP5WEfggKKYmmmbZewvRzm&usp=drive_fs (the only external symbol is a pot, which you can get here: https://drive.google.com/file/d/1FeGO5jI2rWfO3Zp77TjDU1wYfJcvG34H/view?usp=sharing ).

Do you have any suggestions?

As opamps, I was thinking an OPA1678 (overkill? Yes. Blameless? Also yes) for the first two stages and a 4580 or similar for the last two stages (or some bipolar input opamp that can reliably work at 9V supply). The pedal should be able to operate up to 24V (again, overkill, 18V is plenty). My only doubts are:

- should I place a resistor (like, 1K or so) in the buffer's feedback path? Apparently they can help against instability.
- should I add input resistors to the opamps (again, 1K or so)?
- the cap in the third opamp's feedback path is dimensioned for an f3 at about 20kHz at maximum boost, but I'm thinking I can get away with half that.

[fryingpan]

Also, the third opamp has its input bias path into the second opamp (via R4), I'm assuming it's fine. There could be minimal offset, but nothing significant.

[antonis]

Delete C11 & R12 and connect R3 to Vref..
(I usually avoid putting garbage into Vref unless it saves items..)

[fryingpan]

With those in place the noise analysis is somewhat better (and I don't know why ).

Also, would it make sense to convert those diodes at the input to Schottky ones? The absolute maximum ratings for the opamp I'm looking at state no more than 0.5V over or under supply. Schottky diodes leak, but a quick simulation with small signal BAT diodes show that the current through one of those diodes is about 100nA. Together with the other diode + the 1Meg resistor, it should still be something like 800kohm, right?

[Mark Hammer]

I made a unit with 2-pole variable highpass and variable lowpass, each with a few Q settings, for my modular system.  The objective was to accomplish what you describe, except with no particular drive in mind (PLENTY to choose from).

[fryingpan]

I made a unit with 2-pole variable highpass and variable lowpass, each with a few Q settings, for my modular system.  The objective was to accomplish what you describe, except with no particular drive in mind (PLENTY to choose from).

It's not really designed to be a drive effect, this is more of a "safety clipper" of sorts to bring down that low frequency boost if you set the HPF to be resonant. And a very mild one at that. It's basically up to 3:1 compression, at full boost, and the body of the bass or guitar tone would essentially be unaffected. (A 200mVpp signal would not be clipped at all, the simulator calculates 0.5%THD at that level, at full boost).
I also added it in because I had an extra opamp (it was either that or a discrete buffer). Using one half of the opamp to set and filter Vref felt like a waste, with the low gain the circuit has.