Need input. Running XML2 LEDs on a 12v Power Supply?

I was thinking of a "house lighting" project using a 12v 300mA Power Supply to power the XML2 emitter [thinking of going with 3 of them, . . . 4 volts each, running them in series]. Would this power supply be inadequate to power the LEDs to any significant brightness?

I plan on putting them on a computer heatsink with fan [fan to be hooked up to the same power supply]. I may not necessarily use the fan and may have a switch on that part. The heatsink is pretty large, actually, with fins that are about roughly 1 1/4 inch long, with surface that is 2 1/8" x 3 1/2".

What do you think about this? Be as thorough as you can about this project and its viability. Thank you!

That's only a 3.6 watt power supply, and you want to run two or three 10W LEDs with it? You certainly don't need a fan, or even a particularly large heatsink. What you need is a good reason.

If the supply is current limited, then you don't need a driver. Just direct drive them. If it's not, you can use a resistor to limit the current.

More of an experiment than anything. I figured using three of them because of the 12 volt output of the power supply. The XM-L2 emitters are only up to 4.2 volts.

Here's what it is.
Vector AC/DC Adaptor
Model: LF1230D-41
Input: AC 120V 60Hz
Output: CD 12V 300mA

Is it the 300mA figure that's the part that won't drive them very hard?

Whatever help anyone can offer would be great. Thanks!

Orion said:

More of an experiment than anything. I figured using three of them because of the 12 volt output of the power supply. The XM-L2 emitters are only up to 4.2 volts.

Here's what it is.
Vector AC/DC Adaptor
Model: LF1230D-41
Input: AC 120V 60Hz
Output: CD 12V 300mA

Is it the 300mA figure that's the part that won't drive them very hard?

Click to expand...

Yes. Also, a wall-wart (AC-powered adapter) is not an ideal voltage source. "Ideal" in the strict sense. It will not provide 12.00V at any load between 0 and 300mA and then gracefully decline in voltage to create constant power as current increases (12.00V at 100mA, 12.00V at 300mA, 11.00V at 327mA, 8.00V at 450mA, etc). They actually have a mushy voltage-current curve, but generally work well enough for LEDs.

I have a wake-up light on a timer. It is nothing but 4 XM-Ls, a 2A 12V wall wart, and a computer heatsink with a fan attached. At 5:30, the outlet clicks on and the LEDs light up brightly. The LEDs are rated for more current than the wall wart, so they will not fail. The wall wart does get mighty warm, though. It will be your likely point of failure.

At 3.6W, you could use a CPU Heatsink (Avoid the ones with heat pipes) without a fan. For temperature protection, you could add a snap switch in series with your LEDs that goes open-circuit at 80, 100, or 120C. Good gracious you'd need some poor airflow to reach that temperature with 3.6W. Mount the snap switch securely to the heatsink, close to the LEDs.

For output, the XM-L2s are about 120 lumens per watt, so look for 400 lumens or so. You could double or triple the power supply and add a fan for almost double or triple output... With fan noise.

The heat sink I have has fins, but does have two copper tubes that are in a 180 degree arc, but there are still fins. It's a "squirrel cage" type and the thing puts out a ton of air, . . . and sound.

I may be out of my league here. Mostly, I'm just interested in a project light. Probably running 4 XML emitters, but now not sure how to do that with 12 volts, unless it does drop down to 8 volts. Then I could just do two parrallel in series. The XML emitters I'm getting are fairly cheap [from china], so again, it's more of an experiment than anything else. As far as house lights that use a wall socket for power, I'm not sure of any other way, other than using the "wall wart" method.

Orion said:

The heat sink I have has fins, but does have two copper tubes that are in a 180 degree arc, but there are still fins. It's a "squirrel cage" type and the thing puts out a ton of air, . . . and sound.

I may be out of my league here. Mostly, I'm just interested in a project light. Probably running 4 XML emitters, but now not sure how to do that with 12 volts, unless it does drop down to 8 volts. Then I could just do two parrallel in series. The XML emitters I'm getting are fairly cheap [from china], so again, it's more of an experiment than anything else. As far as house lights that use a wall socket for power, I'm not sure of any other way, other than using the "wall wart" method.

Click to expand...

Ah, it's easy, and I don't mean to scare you off.


Your wall wart may die in a few hours, or it'll last for over a month*. Thermal-adhesive the LEDs to a chunk of metal (The cooling pipes are mostly just decorative in your application). Solder them in series (3 or 4 of them won't matter at this power level - The forward voltage of the XM-L2 is mighty close to 3v at 300mA). Attach the power supply to the LEDs (I like barrel plugs, 2.5 x .5mm size or so) and go.

*Wall warts quit when they want to. Unless they overheat, it's usually a boring death.

Is there any other better way for a household application?

I know that there are LED bulbs with the traditional house lighting screws to replace incandescent lights, but they must have some sort of major resistence to counter the 120 volts. The "wall wart" was mostly an "I'm not sure how else to do it" kind of thing. But I really am out of my league if I had to get much more involved electronically. Anyway, I'll try the LEDs in series and see what happens. I may try to get a variable output switch for the fan, though. The thing is really loud and pushes far more air, and really I probably don't need the fan on, but in a way it may make the light more interesting then.

It's fun to play with this stuff, though.


Thanks for your input, sir!

Do it! It'll work fine. The low-power wall wart is like training wheels.

Next, you'll want to take that 'milli' off your amps. It's fun.

With any kind of heatsink, you won't need a fan. An 1/8" aluminum plate 3" square (flat, no fins) would be enough in room temperature air. The fan would just draw off power that could otherwise go to the LEDs.

I'll let you all know how it goes. It will be awhile before the emitters get to me. Thanks!

Guys, just wanting to make sure this is "in series".

That's a parallel setup.

So that 12v wall wart is going to destroy those LEDs. Can someone show me how I would set them up in series so the 12v is split up between the three? Would it be the wall wart positive to the positive on the first, and the negative of the first wired to the positive on the second, with second negative wired to the third, . . . and third negative connecting to the negative on the wall wart?

Or should I just get a battery holder for an 18650 and make a flood style bicycle light with it wired like that? *contemplating*

here is a picture of a series connection:
http://www.google.de/imgres?imgurl=...r=5730&page=1&start=0&ndsp=26&ved=0CC4QrQMwBA

btw don't forget to add a resistor or driver otherwise you might blow your power supply

Here is an LED on a star:

(-O+)

You can see my lovely terminals on it, Negative on the left, Positive on the right. In your application, you'll need to pull up all but one wire to start cleanly.

My power supply is connected to my supply wires (======). These wires bend wherever I need them to. Wire your LEDs like this:

==========(-O+)======(-O+)=====(-O+)=========

Thanks guys. When it comes to homemade stuff, I'm a lightweight/newbie. I'll rewire it. I plugged it in several times for a relatively short amount of time and nothing popped on me, so I should be good to go after the rewiring.

BTW, if I DID run it direct [as it currently is] to an 18650, would that work or would it be a bad thing? I don't have any drivers and was really just experimenting on something cheap. Maybe there IS no "cheap" when it comes to modern LEDs.

Cheap puts your LEDs at risk. Make sure your heatsink attachments are good.

A parallel array like that will have a 'rated' forward voltage of about 3.4v. Check the Cree datasheet to see what the forward voltage is.

However, an 18650 is probably underpowering it. Unless you use a super IMR-cell, rated for a few dozen amps, the battery's output current will top out at about 6 amps, with a voltage of around 3.3v. This is hard, hard, duty for the cell, and isn't healthy. If your heatsinking is good, if your LEDs are bonded to the heatsink well, you'll only hurt the battery.

I have 4 XM-Ls on a computer heatsink, with a 12v computer fan blowing, run on a 2A 12V wall wart. It works nicely and is cheap - But if the fan fails, the heat will ruin the LEDs. I should* add a thermal snap switch to open the circuit if the heatsink heats to over (say) 80C.


*should: This would help, but I haven't done it yet.

The heatsink isn't bad. I used Artic Silver and the binding compound mixed. I'll just rewire it for parallel and run the wall wart. Else, if someone is interested in these LEDs on this heatsink, I could send it to them [at a fair price] so they can operate it more appropriately.

Orion said:

I'll just rewire it for parallel and run the wall wart.

Click to expand...

Series! You connect the (-) of one to the (+) of the next.

Also, if you want a neato "emergency lamp" brick, you could put 2x NiMHs in series to run this in direct-drive through a (say) 1 kOhm resistor. The runtime would be somewhere north of 500 hours on AAAs at a drive current of under 1 mA. That's enough to qualify for the 'super long runtime' projects we enjoy on CPF.