So, my modular synth might be done now. Or it might not…
The A-110-4 Quadrature Thru-Zero VCO arrived a couple of days ago, and the Mazzatron Quad Audo AC Coupler a couple of days before that. In general, the QZVCO sounds good… but not entirely mind-blowing I would say. Perhaps the folks that were blown away by it are the type who avoid digital modules, so linear TZFM is novel to them?
The idea here is, you can use this as a carrier for FM, and feed it a separate oscillator as the modulator. You can use exponential FM like most analog oscillators, or linear FM and it will flip its phase to go “thru zero” for a deep modulation range. Its design is unusual though, with the LFrq knob in essence just being a selector between three states (no TZFM at +/- 5, “pretend you’re a waveshaper” at 0, and TZFM anywhere around the 1-4 or -1 to -4 ranges, with choice of positive or negative having no consequence).
The problem my unit has is that at any LFrq setting (except 0, and possibly even there), increasing the modulation amplitude will shift the VCO’s frequency out of tune. At some LFrq settings, it goes flat. At other LFrq settings, it goes sharp. And in between, it goes sharp and THEN it goes flat as the amplitude increases. There’s no neutral setting where it stays in tune.
It’s far enough out of tune that, with a base frequency of 300Hz, it might drop all the way to 200Hz. That’s not just a little bit that I can live with. It’s way out of tune and whole new FM sidebands appear that shouldn’t be there with my chosen ratio.
Normally, I would expect this to be caused by a DC offset in the modulation signal — and I would expect it to travel in a consistent direction, always sharp if the DC offset is positive, always flat if negative.
I know there is no DC offset because I’m using the QAACC. I also tested highpass filtering with Blades, Angle Grinder, and VCFQ. I confirmed lack of a DC offset with the O’Tool+. I tried adding my own DC offset with Blinds. None of that helped.
The upshot of this is, I don’t get to use dynamic FM amounts — which is the major advantage that linear TZFM has over the much more common exponential FM. Angle Grinder can already do expo FM well and has not just the 0 and 90 degree outputs, but 180 and 270, and the comparator waveshaping stuff too. And Blades can also do expo FM well and crossfade from 0 to 90 to 180 degrees.
There are some calibration trimmers on the circuit board that might or might not be relevant, but these are supposed to be set at the factory and not need adjusting when new. I’ve written to Doepfer’s hardware support email for advice, but it turns out they’re closed until January 10.
I’ll decide what to do once they write back. I can see scenarios where I return or resell it and don’t count it against my 2021 gear plan. But I also could see keeping it even if dynamic FM isn’t going to work well, because with LFrq at or near 0 interesting things do happen. There might be some benefit in keeping it, flaws and all, and learning what it can do.