Hi everyone, this is my first post here, hope to find someone who can clear things up for me 
I’m building a modified version of Moritz Klein’s VCO (take the one appearing at this timestamp on his video).
Now this would be the circuit I’ve come up with after adding my modifications:
The circuit is very similar, I just added some more op amps to get a triangle wave too, changed some resistor values, etc. The way it works essentially remains the same. Now, I’ve wired it up on some breadboards and it works just fine, all the waves look clean, the frequency range is ideal to me and the exponential tracking is accurate enough, however when I tried to test if the CV inputs work, or rather, how the VCO behaves when you plug in a CV signal, I found the output gets heavily distorted.
I don’t have breadboards left and the ones I’m using are already quite full, so I can only build one VCO at a time. Because of that the only way I have to test this is by making the VCO modulate itself: I plugged one of the outputs (that would be C6, C5 or C8) to a CV input such as R3 or R10. That way I would be self-modulating the VCO with its own oscillation. The result was the output signal level getting severely distorted. Let’s say I plug C6 (SAW OUT) to CV IN (R3), I would then measure with the scope at C6 and find a very distorted low level signal which doesn’t even look like a sawtooth wave anymore.
I wanted to ask if this is normal when you try to make a VCO self-oscillate. To my logic it isn’t, I can’t find a reason for it to be happening “just because”. I also don’t think it is due to load/impedance issues since the op amp has a very low output impedance, plus, the way I plugged it to the CV input, the op amp is attacking a much larger resistance, meaning little current should be flowing, and so the op amp is not forced to work too much and the voltage level should remain steady.
I thought maybe the issue is the fact that the input stage is a passive resistor mixer, and maybe replacing it with an op amp active mixer would fix it, but darn it, Moritz Klein’s VCO is even being comercialized in collaboration with Erika Synths, which means this input stage design has to work just fine, right…? I also have the theory that these issues appear because of the capacitors I’m adding in series at the outputs. Could it be that, after plugging them to the CV inputs (which have resistors to ground afterwards) I might be creating some kind of filter that distorts things out?
If anyone’s feeling generous today, I’d be grateful if you could give me some guidance on what’s happening exactly. Thanks a lot beforehand!
Kind regards,
Enf.
PS: I do have another question. In Klein’s design, the capacitor he uses to make the Schmitt trigger oscillate has a value of 2.2nF, which he obtains a good range of frequencies with. For a reason that I don’t understand I needed a much larger capacitance to get a usable range, since in my experiments a 2.2nF cap made it go from a few thousand kHz with 0V to over 100kHz with larger input voltages. Although on the modified schematic that I linked I placed a 100nF capacitor (C1), this is still quite high to me. I’ve put together two caps with a total value of 320nF, and that gives me a 30Hz-6kHz range roughly speaking, which is where I’m standing with now. Any idea why? Thanks again!
