# Resisting current to fan

Discussion in 'The Projects Forum' started by GRNDPNDR, Apr 2, 2014.

1. ### GRNDPNDR Thread Starter Member

Mar 1, 2012
449
8
Back to my 30W LED project, I have a small northbridge heatsink with a fan on it that I've mounted to the back of my LED.

I'd like to poer it from the LED driver, but it's constant current and pumps out 1A, whereas the fan is only 50mA.

Would there be a way to restrict the current to 50mA for the fan without a drop in voltage? it doesn't move a lot of air so I don't want to slow it down, but it seems to move enough to keep the LED cool.

Last edited: Apr 2, 2014
2. ### AnalogKid AAC Fanatic!

Aug 1, 2013
5,590
1,586
Since the fan likes to see a voltage source, let's start with what we got. Wwhat would be the range of voltages the fan would see when the LED is bright and dim?

ak

3. ### GRNDPNDR Thread Starter Member

Mar 1, 2012
449
8
Near as I could tell it would be about 11-13V at a rough guess, based on the current consumption of the LED.

I measured the driver s putting out 1A @ 12V, but it would obviously flucuate to keep the current at 1A.

The LED only draws 1A without a heatsink, and with the heatsink mounted it is now only pulling 0.94A @ 12V on my CV bench supply.

4. ### GRNDPNDR Thread Starter Member

Mar 1, 2012
449
8
I'm really tired so I could be wrong, but could I use a 1W, 240 Ohm resistor to keep the current at 50mA with 12V?

5. ### MrChips Moderator

Oct 2, 2009
14,300
4,199
You need to put a resistor across the fan in order to divert 1000 - 50mA, about 12Ω.

6. ### THE_RB AAC Fanatic!

Feb 11, 2008
5,435
1,309
Then the fan would have to be in series with the LED? The LED already draws 30v and 1A, so putting another 12 load in series with it requires a PSU of 42v + headroom.

7. ### GRNDPNDR Thread Starter Member

Mar 1, 2012
449
8
The fan would be in parallel, the driver is only 12V.....where did you get 30?

8. ### mcgyvr AAC Fanatic!

Oct 15, 2009
5,065
1,099
Where did you get 12V?

30W/1A=30V.. Most of the 30W LEDs are rated 32-34Vf @900mA or similar..

9. ### GRNDPNDR Thread Starter Member

Mar 1, 2012
449
8
This LED is 12V, 30W.

Unloaded, the driver measures 52V. Once the LED is connected it drops to 12V running at 1A.

10. ### MrChips Moderator

Oct 2, 2009
14,300
4,199
Sorry, I obviously don't get the situation.
If the LED is powered at 12V then put the fan in parallel with the LED.

11. ### GRNDPNDR Thread Starter Member

Mar 1, 2012
449
8
That's the plan, but the driver is a constant current source, so it's going to feed that little fan 1A (or .5A if it's to be divided equally) which is far too much for the fan.

I need to make sure the fan runs at 12V and only ever sees 50mA

I'm really just trying to avoid having a second circuit just to power the fan when the driver is already 12V and plenty of current.

ALTHOUGH, the ebay page says the output of the driver is 32V, my measurements show otherwise when loaded.

12. ### MrChips Moderator

Oct 2, 2009
14,300
4,199
If the 12V fan takes 50mA, it is going to take 50mA as long as the 12V LED is in parallel and not blown.
If the LED blows the fan is in trouble.

13. ### GRNDPNDR Thread Starter Member

Mar 1, 2012
449
8
so even though it's a constant current source, the fan will only take what it needs still?

unless the LED dies, then it's going to get the full fury of the driver.

Would there be a way to protect the fan from such a thing happening, should the LED fail one day?

14. ### MrChips Moderator

Oct 2, 2009
14,300
4,199
You could put a 12V zener diode across the fan but the zener would have to be rated at more than 12W. And if the zener goes so does the fan. This is getting messy.

The simplest solution I can think of is to power the fan from a separate 12V supply.

15. ### GRNDPNDR Thread Starter Member

Mar 1, 2012
449
8
that would be a big zener.

I'm not terribly concerned about the fan, just exploring options really.

Given nothing strange happens, the LED would probably last for years so the fan would probably quit before the LED anyway lol. It's already an old used northbridge heatsink.

Not only that I doubt I will be running the LED at full brightness all the time, so that would further extend it's life.

I'll mess about with the fan a bit today I think, take the chance and hook it up in parallel and see what happens.

16. ### Alec_t AAC Fanatic!

Sep 17, 2013
7,042
1,459
If the LED is drawing 1A and has 12V across it then it can't be a 30W LED (unless the Chinese have their own version of Watts) .

17. ### GRNDPNDR Thread Starter Member

Mar 1, 2012
449
8

Ya, understandable but the LED itself was bought at a local shop, only the driver came from china.

I wondered a bit about this myself because the rated output doesn't match what I physically measured.

The current matches, but the voltage doesn't agree.

18. ### GRNDPNDR Thread Starter Member

Mar 1, 2012
449
8
just tested it, and looks like it works fine with the fan connected in parallel to the driver. But my meter may have been acting up when I took measurements because now I'm reading 14.6V @ 1.06A with or without the fan in parallel.

19. ### GRNDPNDR Thread Starter Member

Mar 1, 2012
449
8
Just letting it run for a bit to see what happens, and so far it's at 1.08A @ 15.1V

This may cause concerns for the voltage being applied to the fan.

20. ### AnalogKid AAC Fanatic!

Aug 1, 2013
5,590
1,586
A little late to the party, but think of this.

If you have a voltage source that might go as high as 52 V, and a load that wants to see no more than 12V, then the answer is a voltage regulator for the fan. Voltae regulator and fan are in series, and that string is in parallel with the LED.

If you don't need full speed out of the fan, it can be a 317HV and the fan will see 10 or 11 V in normal operation (normal being 12V across the LED). If you need full speed then go with an LDO regulator. Either way the regulator should be able to handle at least 60 V in to run reliably in a 52 V world.

OR, since you don't need tight regulation and we're talking only 50 mA, this can be as simple as a 20K resistor, 12V zener, and TIP29B or C pass transistor. Worst case transistor power dissipation is 2W, but the load is after all a fan so heat sinking should be minimal.

ak