Constant Current Source question

[Jazznoise]

So I'm building an opto coupler based AM circuit which I want to be able to control with 0-5 volt signals, partly because my onboard Osc. is digital but also for other cool stuff like CV.

Now in order to get some sort of sensible responsible from my LED I've been told I need a constant current source to drive the LED, in my scenario I'd need to move it between about 0.1 to 1mA and maybe above. The opto couplers I'm using are the Solinex 32SR and they provide example circuits, but they don't seem to work.

Any help is appreciated!

Regards,
Jazz

[PRR]

> I've been told I need a constant current source to drive the LED

Not essential. A large voltage waveform and a series resistor is constant-enough current for many chores.

> Solinex 32SR

Silonex NSL-32SR

> they provide example circuits

Where?

> they don't seem to work.

"Doesn't Work" can mean many many things. Fire, smoke. No-signal. Signal, but not what you expected.

What have you done? Have you comfirmed that current flows, right direction, roughly right amount? What is this "digital oscillator"?

[Jazznoise]

Hey PRR, thanks for the reply!

> Not essential. A large voltage waveform and a series resistor is constant-enough current for many chores.

But would this allow me to set a specific range of resistances? Realistictly I want to keep the resistance between 2K - 200 ohm range so I need to be able to keep my current between 0.1 to 1.1-1.2 mA. I figured since I'd be driving it with a few different sources - the PWM out of my 18F2455 and possibly an external audio source - that a constant current source would the stablest way to do it. However, writing this I'm starting to wonder if I just need a buffer before the opto-coupler just so whatever is driving it is of a fixed impedance.

> Silonex NSL-32SR

Sorry, that's the one, though it's the SR3.

> Where?

http://www.silonex.com/audiohm/levelcontrol.html

I was attempting to build figure 5

>"Doesn't Work" can mean many many things. Fire, smoke. No-signal. Signal, but not what you expected.

Well with the Op amp driving the input (I used a TL084 instead of the 074, just what I had) of the transistor I found the signal was sometimes oscillating when I put a simple sine tone through (making a square wave whose duty cycle varied with the level of signal put in)
and that it wouldn't "drive" the transistor stage in any functional sense - any LED in the signal patch stayed very much dead, with no signal on the collector side and the Emitter displaying a sort of badly rectified signal.

Yet I rebuilt the transistor stage to isolate it and it does indeed seem to be able to drive an LED and I could get usable resistances across my opto-coupler. However since my signal generator decided batteries are delicious, I've yet to retest this with a signal driving the base. I will update as soon as I get to!

>What have you done? Have you comfirmed that current flows, right direction, roughly right amount? What is this "digital oscillator"?

I think I've answered this to an extent above. I'm using an 18F2455 to read a table of a few waveforms - Saxaphone, Sine and generating a few others such as a PW Square wave, Saw and Triangle and using the PWM output to drive the opto-coupler.