Put this in .basics and didn't get much. Trying here. More on topic than trolls who can't build a computer or grasp the difference between electromagnetic and acoustic waves, I'm thinking...
APC...Automatic Power Control
Lots of complex hoo-ha out there for this. I'm trying to simplify, using the LM317 since I like it and have them on hand. Costs a few cents more than a transistor solution, but low component count and built-in protection, as well as supply voltage range are nice. Plus I already have one in use for the dumb (constant current, non-APC) version of this circuit.
Laser Diode has a photo-diode in the same package. Point of APC circuits is to pay attention to the photo-diode output and adjust laser current to maintain consistent photo-diode output. Common positive for the parts I'm playing with.
L In---V+ M
3 Out-------- 1 + 7 Adj |---|---| | _P V L | ^D - D | | | | R6000 RX |----| | ----- --- - My (too) quick think through this is that if RX is set to provide adequate reverse bias to the PD (30 max, 5 suggested) in the realm of operating current (80-100 ma), the 6000 ohm should result controlling at a 0.2mA photo-diode current. ...But that ignores voltage across the photo-diode, and the fact that the only voltage the photo-diode will see is the 1.2V from Out to Adj. Crap. That won't work.OK, turn it around and use the LM337. Adj terminal only sees the voltage across the 6000 ohm resistor now. RX wastes power (under half a watt?) to get photodiode reverse voltage up, at least I think that works. 47 ohms should about work, given a mid-range current of 90 mA, laser diode forward voltage of about 2V, and drop across R6000 of 1.2V.
|-------------| + ----- |---|---| --- _P V L - ^D - D +--+ | | R6000 R47 | |-------| | | | | L Adj | M |
3 In---V- | 3 | 7 Out-----------|Please poke holes in this, if there are holes to be poked. I'll try building it shortly if it's not hopelessly flawed.
One hole I've noticed since putting it up in basics is that the 50-100uA at the adj terminal is significant (I had recalled when cooking up the design to begin with the PD output as being 1-3 mA and put in a 600 ohm, then revisited the data sheet and upped it to 6000 ohm, since it is actually 0.1-0.3 mA, or 100-300 uA - how much of an issue that is would probably depend on how consistent the adj current "error" is.
I may yet just throw together one of the transistor solutions, but want to dissect this idea first.
...and yes, I know it's much cheaper/simpler to just buy a built diode and APC circuit. That's not the point here. The intellectual exercise with a semi-practical result is the point.