Trimless VCA

uniqview

I have updated my electronic music blog with a post about a new VCA I recently designed.

I wanted to see if it was possible to achieve the performance goals, including very small control voltage feedthrough, yet require zero trimmers. The topology I developed does this, and has excellent performance in most respects.

The circuit design would admit alternative component choices for the active devices, which are a precision matched transistor array, and a precision op amp.

To acheive the specific goals, I made certain engineering tradeoffs with passive components that other SDIY'ers might or might not find favorable to their needs or aesthetics. Much of the design relies on components matching tightly, so for that reason there are several ±0.1% resistors.

The VCA works really well in my superstrip breadboard test. It sounds great, and it's simple. It only uses 24 parts, with only 2 ICs. I didn't post scope fotos in my blog, because I wanted to test on a PCB implementation, which I've not yet done.

There's more details in my blog post, but to summarize some basic performance the VCA is designed for ±2.5V audio with DC signal path, and 0-5V linear-responding control voltage. It has excellent waveform fidelity at 20 KHz, and full power bandwidth to at least 50 KHz. I don't have an Audio Precision tester, but with FFT on my Tek scope, the main harmonic contribution is a 3rd-order spur at -45 dBc, everything else is down in the mud. I'm reasonably sure this VCA is not a high-fidelity type, and it likely doesn't have 100 dB dynamic range. But I also think the requirements for a synthesizer VCA are also more relaxable.

When controlled to be off, the VCA output measures "quiet" on a scope when viewed at 20 mV/div, with very apparent little noise. It also sounds quiet when I hook up a sound amplifier. The main reason I wanted a trimless design was control voltage feedthrough that worked -- many VCAs I've studied or built over the years have interacting trims, and control voltage feedthrough is a tradeoff. So I made a VCA topology that appears to remove audible "clicks" that come from fast control signal changes, or even the control voltage as an audio signal, when the input audio is at 0V.

blue hell · Posted: Mon May 31, 2021 1:53 pm Post subject:

Looks like a nice design .. was wondering .. you state +/- 2.5V for the signals .. that seems a bit low for synth levels?

uniqview · Posted: Mon May 31, 2021 2:40 pm Post subject:

blue hell · Posted: Mon May 31, 2021 3:34 pm Post subject:

Ah ok, I understand, thanks!