Bipolar optocouplers good for anything?

[Jonas H]

I managed to save a board with 7 quadrupel PC817 optocouplers (in sockets, mind you!) which means a total of 28 optocouplers. These are of bipolar type. Does anyone know something I could use them for. According to a post by R.G. recently bipolar optocouplers are no good for remote switching. Pity!

[Nasse]

:?:  :idea:  :?: The type does not ring a bell, maybe if I look after it in some data book or page I will know more. I dont see why they can not be used in remote controlled applications. Maybe R.G. means there are simpler and cheaper (optocouplers are not always cheap) alternatives for it, he is clever and cost concious genleman what I believe. But if uou have optocouplers  free... A light organ for your band or home disco?

I once breadboarded and even made a pcb and built an unit (found schematic in old elektor magazine) What I was trying to do was a audio signal isolator (audio going trough led / transistor combination inside a common and cheap optoisolator. It worked, but I had some level matching problems and level loss was too much for what I was looking after ( I worked at "112" ("911" in other side of Atlantic) and I wanted it be totally isolated of some life saving equipment. anyone messed with these?

[Paul Perry (Frostwave)]

Seeing that some wah circuits use a transistor as the 'variable' leg of the twin-T filter (in place of a pot) maybe you could make an autowah, by driving the LED section & thus wobbling the impedance of the transistor section?
Maybe a 'three peak' unit, seeing you have multiple units in there?

[R.G.]

Let me clarify my comments.

LED/bipolar optos are designed for fast switching. They have very limited linear signal range. I've been disappointed in them every time I tried to couple a linear signal through them.

If you keep the signal level low - like under 50mv, maybe 25mv, you could use the variable resistance of the collector region for a variable resistance, as in the twin T circuit, as Paul notes. However, the response curve of LED current (which is what you can control) to collector resistance will probably not be either linear or exponential, which is the two response curves you'd like to have. I haven't run down what the response curve is.

It's always possible that the signal distortion through an LED/bipolar would not be something you like. Odd distortion mechanisms are part of our stock in trade. It's worth listening to what happens that way before you throw them out. I didn't like it when I did that experiment, but my taste in distortion is not the same as everyone elses necessarily.

They might be good for shunt switching, where what you want is to short a signal to ground to switch it off. That's something that bipolars are good at.

If the opto is a high speed TTL one, it has an internal integrated circuit on the output side that fully switches the output on or off, no intermediate. These will always have a power and ground-connected pin as well as the output pin.

Overall - I messed with them, didn't like them. That doesn't say there's nothing there; I just didn't think it was enough to justify my time messing with it.

R.G.

[Paul Perry (Frostwave)]

Right you are RG! If they are optimised for switching, maybe they would be good in soe ind of pulse width modulated phaser? (like the Elektor one) where you have resistors being switched in and out at an ultrasonic rate.