TFW you spend the last hour trying to figure out why your integrator isn't exporting a triangle from a square, playing with the RC time constants to try and figure out what makes it look like a triangle, only to realize that it is impossible to use an integrator in this circuit because inherently the frequency is changing and as such you can't use a fixed time constant and would have to find a way to sync it with the frequency of the square wave ;-;
Ah well it doesn't matter. I'll just keep this oscillator bank as a pure square, and use THIS one on the output of a *saw* bank (I've built this successfully in the past, LMNC has a great tutorial on this).
Oh and with every knob comes a way to LFO it, I need to read into the theory of LFOs and figure out how hook up each of these pots (and in the filter bank, caps) to a LFO.
Furthermore there's the complexity of the LFO itself (i.e. defining LFO shape) which is in and of itself a bit wild to do.
HOWEVER once I get the circuit design done in KiCAD once, then I can order like 10 of them for like $30.
BUT the circuit as it currently stands is designed to work with a ribbon controller, not a keyboard.
Getting it to work with a keyboard controller first and *then* allowing for ribbon control is more sensible. So I need to look into how to add keyboard control.
Oh and then I have to add ASDR and mixer panels. Mixer panel should be easy as fuck. ASDR is complex cap stuff but that I can use an existing circuit without modification.
One thing I want to figure out how to do is the "prism" effect in Harmor, but in analog synth terms. I need to do some digging to see what exactly it's doing (my guess is that it's keeping the mid tones, but making high tones higher and low tones lower, which I assume is splitting the signal into high pass, band pass, and low pass sections, subsequently increasing or decreasing the frequencies, and then mixing them back, need to figure out how to increase the frequency of a given input signal)
I need to do more research into what actually went into Moog banks and try and recreate it as much as possible.
I could just straight up use the Moog VOC circuit and throw that onto a PCB and get that made but where's the fun in that? That's not something I can show off to maker faire. I might as well just copy and paste code off of Stack overflow and call it my own.
Anyway so basically the oscillator thing is not going to be a 40 voice thing. It's still going to be max 10 voice, and each one has a switching mechanism to switch to which waveform you want to use. Then there'd be sections for square, saw, triangle, and sine controls on each board.
I'd design it in such a way that it would be completely modular and you can attach as many voices as you want.
Now the *filter* banks are going to be fun because they're both pretty straight forwards and pretty complex. Same with the LFO thing.
That's a *lot* of things to get done by Maker Faire (mid November) but it's really only 6 circuits, everything else is just ordering $300 worth of PCBs to build a synth worth $5000.
I think I'd just straight up buy my own sequencers and stuff, I'm not gonna bother trying to build that.
I mean yes I'm... doing this the hard way. I could just use an existing schematic, throw together a PCB, and order it. But again there's no fun in that, so I'm taking my time to enjoy this project.
