/diy/ - Do It Yourself

>>1418451
I was going to try to make a gilbert cell to use as a phase detector, but then I realized it's pretty much pointless or too much effort to do with discretes, and I discovered that a SA612 can sort of work, even though it's meant to have an oscillator circuit placed on one of the inputs, as opposed to another "input" signal from an op amp; but THEN i realized that a multiplier phase detector is pointless for my intended application because the output is 90 degrees out of phase with the inputs (holy run-on, batman!). My two inputs are (nominally) in-phase.
The idea is something i've been tossing around in my head for literally years without being able to come up with a realistic solution until now: an equalizer-type circuit where the filter bands follow the input frequency. Instead of having fixed bands spanning the entire audio spectrum, specific harmonics are adjusted. As you can imagine, simultaneously tuning 5+ voltage controlled bandpass filters to accurately follow a noisy signalfrom a guitar pickup over several octaves is a crapshoot, but a paper just published in 2016 shows a solution that might just fit inside of a large stompbox: https://www.worldscientific.com/doi/abs/10.1142/S0218126616500456
The bandpass filter for the fundamental frequency of the input is used INSIDE a phase-locked loop in place of the VCO, with the phase offset adjusting the control voltage. So instead of trying to independently tune an array of VCFs to ridiculous precision, the only requirement is to have each successive filter be tuned to a multiple of the loop filter's resonant frequency, which is easily accomplished with some precision capacitors. As for OTAs, LM13700s aren't prohibitively expensive.
requirement for through-hole components is just for ease of breadboard prototyping. I know this is a few lines of code with a microcontroller, but that's no fucking fun. Currently looking at phase-frequency detectors as the final puzzle piece, as they have 0 phase offset.