Output level limiting for simple OpAmp mixer?

Hi there,
The next Kosmo module in my pipeline is a 4 channel mixer for both CV and Audio, which can also accept Line level instruments like external synths:

I found a couple of examples on the interwebs for basic Eurorack mixers (https://www.youtube.com/watch?v=g1eAsHxHG6g&t=285s, https://www.youtube.com/watch?v=q8tmUgaXrEQ), but I noticed the following: they don’t seem to limit the output to the Eurorack convention of 10 Vpp (+5 to -5 V) for Audio and CV signals.
In my experiment with the Moritz Klein one, I realized that if you mix in 2 oscillators and you tune them very close together, the mixed output will produce peaks of 20 Vpp (+10 / -10), of course, since the oscillators output 10Vpp each. Isn’t this a problem for other modules down the line if they are expecting a “Eurorack like” signal in their inputs?

I’m thinking about adding a couple Zener diodes facing each other between GND and the output of the OpAmp, so that the summed output will never go beyond 10Vpp. Or will I regret it?

(I’m currently out of zener diodes so I can’t experiment with my ears yet :stuck_out_tongue: )

There isn’t really such a convention, at least none that’s very widely adhered to. A good number of envelope generators, for instance, will go something like 0 to +8 V. Raw outputs from the AS3340 VCO chip are typically 0 to 4 V for triangle wave, 0 to 8 for ramp, and 0 to 10.5 V for pulse; some designs normalize these to ±5 V, but not all. (LMNC 1222, for instance — Kosmo version at least, but Eurorack version too, I believe.)

It’s very unlikely you’d damage anything with a ±10 V signal. Any well designed module containing something that can’t handle ±10 V inputs should have protection circuitry already.

Limiting outputs to 10 Vpp using Zeners will clip and distort anything that tries to go higher, but it’s up to you. (Also, nitpicking slightly, you’d be limiting to ±5 V which is not the same thing as 10 Vpp — a 0 to 10 V signal is 10 Vpp, and with AC coupling can be brought to the ±5 V range, but your Zeners would truncate it to 0 to 5 V.)

What I mean with 10Vpp is really just +/-5V.

I wasn’t really thinking about damage to modules down the line. Was thinking more in terms of “compatibility”, since I was under the (perhaps silly) assumption that most Eurorack modules would be designed to work best with +/-5V at their inputs. As I moved on from VCO to Mixer design, I realized it was probably a silly assumption.

Anyway, the idea I have now with the Zeners is to use 3.3V ones in series but facing each other, which means the first OpAmp stage would clip its output to +/-4V (considering the 0,7V forward bias of one of the diodes) even when a single chanell gain knob is fully clockwise. The second OpAmp stage can make up in gain for the remaining +/-1V. I think this is an interesting way of using the mixer to “dial in” some color to the sound, and I could still avoid that coloring by simply dialing back the gain knobs of each channel without losing peak volume (+/-/5V) in the output.

I still need to receive my zeners to check if that would actually sound nice :stuck_out_tongue:

I think you’d probably end up wanting a separate mixer for CV, then. Or make the Zeners switchable.