Request: Use Optocoupler to Replace Dual-Gang Pot

[thehallofshields]

I have a couple designs that use dual-ganged pots, and I'd like to streamline the design using some sort of photocells.
I understand I probably won't be able to get a linear sweep, but I'm interested nonetheless.

I see some 8 pin optocouplers out there, but I have no idea how to turn a transistor into a variable resistance.
If anyone can find some beginners material on the topic I'd be very grateful.

[R.G.]

I see some 8 pin optocouplers out there, but I have no idea how to turn a transistor into a variable resistance.

If it was easy, someone would already have done it; and if it was easy, cheap and broadly applicable, no one would put up with photocells - right?

Bipolar transistors are very hard to turn into a variable resistance. It can be done, but at the cost of some sincere side effects. The biggest two side effects are offset voltage and signal size limitations. You have to put current into the base of a bipolar to change what the collector-to-emitter path is doing. This necessarily offsets the internal voltages a bit and this offset with the control current is hard to get rid of, mixed with the audio that's being resisted as it is. The signal size is limited because the "linear resistor" region of a bipolar is quite small, much smaller than a JFET's equivalent region. So signals bigger than 25 to 50 mV are significantly distorted.

I have used centertapped photocells as dual pots before, notably in the speed control for later versions of the NeoVibe phaser. That does work, but it requires that the two pots have at least one shared terminal, and that both be used as variable resistors, not pots, and that you can put up with the variations in the resistor values as the photocells light-adapt and dark adapt.

Without knowing what your couple of designs need, that's about all I can say.

A better choice for most analog designs is to go to some form of duty-cycle averaged resistance or switched-capacitor setup. This will probably need LED to FET-type optos to implement, as the switching has to be much faster than LED-photocell can do, and much more noise-immune than LED-BJT optos can manage.

[thehallofshields]

Excellent response. You've saved me a mountain of wasted time.

What I'm trying to do control the Gain of 2 Inverting Opamp's simultaneously and inversely.

https://upload.wikimedia.org/wikipedia/en/e/e6/Opamp-inverting.PNG

Okay to be a little more forthcoming, I'm doing a CMOS distortion and the character is completely different when an opamp is doing the bulk of the amplifying as opposed to when the CD4049 is doing the leg-work. With a dual-ganged pot I can get a sweep at a pretty consistent volume/clipping level, but I don't like dual-gang pots and I'd like to be able to direct-mount the pots in the finished design.