R = (VS - VL) / I

[petemoore]

  2v - 2v divided by 40 = 0
  2v being the CV output set at max as supply voltage,
  -2v being the 'LED voltage',
  40 being 40ma from data sheet, "test conditions [2]"...whatever that means.
  But it doesn't 'ballpark', I think I need a current limiting resistor on there.
  Is what I come up with, maybe I'll try another full equation.
   http://www.kpsec.freeuk.com/components/led.htm
  just less than half way down, methods for calculating Current Limiting Resistor for LED's is shown.
  The other way I figured out to adjust the current limiting resistor for an NSL-32:
http://www.alliedelec.com/Images/Products/Datasheets/BM/SILONEX_INC/Silonex-Inc_Actives-and-Passives_6995014.pdf
   ...is use a pot as VR for choosing the value using the actual CV output, and measure with a DMM the resistance on the photocell side, pushing current through the LED to get just under 1k, which will work, but I wanted to finish some mathwork since I did some on it...I hardly ever finish the math as much as I should.

[R.G.]

I don't know what the preamble to this is, but:

An LED has a fairly constant forward voltage when it's conducting. A typical green LED might need the 2V you talk about.

You cannot successfully stabilize the current in an LED with a resistor if you don't have a minimum of about 2.5-4 times the LED forward voltage to play with. That is, you cannot reasonably expect to power a 2V LED from a 2.5V source and also have some control of the current with only a resistor to limit the current.

It's only when the voltage across the current limiting resistor becomes equal to the LED forward voltage or bigger that it can control the LED by making variations in the LED current insignificant compared to the voltage across the resistor.

[Faber]

If I may comment, looking at the datasheet, it says that the Vf is 2.0V at 16mA, not at the 40mA in the previous row.  (couldn't they just make these datasheets easier?)
There's no voltage to current graph for this one either...

Sorry I don't have a real answer for you, but I hope this helps in some way.

[R.G.]

OK, I found the equation.

They're telling  you very much what I was telling you - it is not practical to run an LED from a supply voltage which is very near the voltage the LED needs to turn on if you rely on a resistor to limit the current. You must either use a significantly larger voltage than the LED forward voltage so there's some voltage dropped across the resistor, or use something like an active current source so the voltage is always ...exactly... what the LED voltage is and the current source circuit makes up any differences.

It's all consistent.

[puretube]

Go PWM ?

[R.G.]

I don't think so. The LED starts to conduct at (to make up an example) 1.90v. It hits perhaps 20ma at 2.00V and 40ma at 2.05V. At 2.1V it may be 100ma. Those are not exact numbers, but illustrative. The conduction ramp of an LED is as steep as an ordinary diode.

PWM works great if you have something to average the on/off times. An LED is not an averager, except perhaps for the few 100 pF of the junction. PWM works great for cases with series inductance or resistance + shunt capacitance to do the averaging.

It's not that the light output wouldn't be modulated, you just have no way to keep the peak current from burning out the LED without a peak current limiter in the circuit somewhere.

[puretube]

Ooops, maybe I possibly misunderstood/-interpreted petemoore`s problem?

[R.G.]

It's always possible that I did as well. I'm just going from what I read on that first link in his post.