[sdiy] (More) Really Low, LFO

[Kyle Stephens]

I just finished a term paper this morning that had interrupted my LFO research. Glad that's over for the time being.

Anyway, at any rate lemme try and answer what questions people had about the LFO, and many, many thanks for all the replies - it's more than I could have hoped for.

My primary application for this is in a guitar effect, an FX loop panner: input, buffer, 2x send to pair of FX/whole loops, receiving end has two variable resistance elements controlled by the LFO with one running antiphase to the other, ergo panning. Biggest issue here is that I have 9V supply to play with.

Using the basic design, I wana make a module for synth use too though. I've a lot of filters on my bench, and when they're all done I don't think I could pick just one, so to pan through them super slow would let me make balanced use of them. It'll be subtle, but I like subtle/slow effects, and I could really rip through them at higher speeds just as well too. I recall someone saying a while back that the maximum practical modulation rate was ~500Hz, so call that the top speed.

Micros as a solution have come up, and that's something I'd love to get into, but I'm still getting the swing of this whole analog thing, and I jump from ideas a lot as is, so trying to keep a lock on the present mission.

An integrator with a big cap, feeding into an op amp Schmitt trigger, then into an OTA and back into the integrator, seems like a decent bet so far. 

The top of page 11 here gives you an idea:

http://www.intersil.com/data/fn/fn957.pdf

An SSM2164 looks very nice too. One cell for one LFO, maybe use another to sweep that one, and the remaining two as the variable resistance elements.

As for stability, in the stomp box form at least doesn't need to be rock steady - it's a glorified trem pedal after all. The ability to compose a song around a relatively predictable sweep time was what I had in mind. The module should if it could be more consistent, though I'll worry about that more when I come to that point.

~+/- 5 seconds off is reasonable at longer sweeps (or is it? from a design/music standpoint?). Someone mentioned getting at long as 4 hours with the CA3140 design - even half that (~length of a concert) would be grand.

More questions/suggestions, just ask again, and hopefully I can be more expedient in responding. Hanging around here is more fun than term papers but due dates there aren't as flexible as my projects unfortunately (least it means one thing gets done sooner than later...!).


_Kyle

(Loved the photocell/clockwork idea, Paul!)