Idea for optical compressor or expander

earthtonesaudio · Post by **earthtonesaudio** » 06 Nov 2008, 17:22

Simple enough idea, using a CMOS hex inverter chip. U1 is the compressor or expander depending on where VR1 is set. U2 makes up gain and buffers U1.

Not shown are all the other details to make it work: the envelope and LED driver part, and input and output circuitry. But I thought I'd just throw this out there to see what people think. I am thinking of using something like this as the input stage to an auto-wah, to have lots of control over the envelope.

darkglass · Post by **darkglass** » 03 Dec 2008, 15:38

Is it me, or I'm missing an input coupling capacitor? No signal will flow into the inverter since you're shortcutting the power supply...
Good idea though! You could develop it more, try by simulating your snippet with Ltspice, or some similart software, it really really helps!

Greetings
D.

earthtonesaudio · Post by **earthtonesaudio** » 03 Dec 2008, 16:18

darkglass wrote:Is it me, or I'm missing an input coupling capacitor? No signal will flow into the inverter since you're shortcutting the power supply...

Nope, it's not just you!

earthtonesaudio wrote:Not shown are all the other details to make it work: the envelope and LED driver part, and input and output circuitry.

It seems to work fine on paper using "ideal" values. The most dramatic effects are if VR1 is large and the fixed resistors are small (and if the photocell can go from 0 to infinity). With an ideal photocell, "R"=10k, and VR1=1M, I think the gains worked out to something like this:
Max Compression: gain=101
Min Compression: gain=1
Max Expansion: gain=.01
Min Expansion: gain=1
VR1 set to midpoint: gain=1 regardless of photocell resistance, i.e. no change in dynamics.

Basically you could turn the knob one way for compression, the other way for expansion, and in the middle for unity gain buffering.

I have MacSpice on my computer, but haven't had time to play with it yet! Hopefully this snippet works in the real world.

briggs · Post by **briggs** » 03 Dec 2008, 19:27

That is a very, very interesting snippet indeed. Especially goes along with my long term design project... It could be the inspiration I have been looking for! A simple envelope follower would be pretty easy to get out of a decent cmos chip too. I'll have a fiddle later

earthtonesaudio · Post by **earthtonesaudio** » 03 Dec 2008, 22:13

I'd like to see your ideas about using some CMOS inverters for generating an envelope voltage. I got stuck there because I figured it the inverters would not work well with DC signals close to ground... but maybe I should think about that some more.

briggs · Post by **briggs** » 04 Dec 2008, 00:28

http://www.generalguitargadgets.com/pdf ... filter.pdf

It can be done, that's just one example. I'm sure with a bit of fiddling it wouldn't be too tough to whip up an envelope follower with cmos. Just using them as gain stages should be enough? I'll try it! AG probably knows, hopefully he'll read and post some insight

The project I'm working on is an overdrive/limiter mutant pedal. Pretty sure there is nothing out there that works quite like it at the moment.... I have been using op amps, but I need a quad opamp plus another semi conductor so it's taking too much space. The cmos option would clear that problem for me - it all depends on working up a decent envelope follower though!

earthtonesaudio · Post by **earthtonesaudio** » 04 Dec 2008, 13:39

Oh yeah, I keep forgetting about the MXR envelope filter. Pulse width modulation is my favorite! I see they just full-wave rectify the signal with diodes and use that voltage to control the pulse width. I guess my worrying about inverters for DC was unwarranted.

Thanks for the reminder!

briggs · Post by **briggs** » 04 Dec 2008, 17:40

You could probably employ something like the old DOD 440 follower, just using an invertor as the gain stage. http://www.generalguitargadgets.com/pdf ... 440_sc.pdf