Issue with constant current source for LED

[Aluz]

I've been trying to build a constant current led control circuit, but I'm not sure why it's not working. I'm using the arduino output to control I_SP. Output voltage for LED is extremely low for some reason. Should be around 7-8v but for 12v supply is around 2-3v.

Any help would be much appreciated.

[PCB.Wiz]

It's more common to put the LED in the FET Drain, as that gives lowest voltage loss.
As you have it now, the VGS of the fet subtracts from the VCC range.
Moving the LED eliminates that Vgs dependance.

[Aluz]

Oh I see, should the wire between R3 and Imeasure exist?

[PCB.Wiz]

Oh I see, should the wire between R3 and Imeasure exist?

Yes, that's needed for feedback.
You move the UV LED up on the SCH to connect the + to VCC and - to MOSFET Drain

[magic]

Measure voltages at all inputs and output of U1 and Q1 and see what isn't behaving the way it should.

If you have a scope, check if there is no oscillation.

Not sure what's the point of R3? Maybe you copied this schematic from somewhere and R3 was meant to go to U1 IN- rather than VCC?

[DavidAlfa]

Don't use a N-Fet like that.

The gate threshold is usually around 2.5-3V, that voltage Will be dropped bt the fet.
Then the led drops another 2-3 volts, so 5Vcc won't work.

Place the led at the drain!
Also lower power loses by replace the sensing resistor with 1ohm, and dividing the output of the RC filter by 10.

[SiliconWizard]

Yes to the above.

Use a current sink topology with a NMOS (source to GND) rather than a current source one.
If you insist on a current source topology for some reason (which doesn't seem justified here), in the same context, use a PMOS instead (the current will in this case be proportional to the difference between VCC and the input voltage).

[Zero999]

Put the LED in the drain.

[Aluz]

Ok awesome, I will test this out. Thank you.