Easywhite XLamp MP-L LED driver for headlamp?

[purduephotog]

I just discovered Tyco makes a solder-less socket for the XLamp MP-L LED (http://www.newark.com/cree/mplezw-a1-r100-0000c040f/easywhite-leds/dp/63R1258) - which means I can drive one of these guys without having to get all super fancy soldering skills.

The socket (TYCO ELECTRONICS 2106946-2) has the requisite pins to carry the low current higher voltage and has a standard AMP mini connector- making it pretty portable. Solder the MP-L to a chunk of copper and either epoxy or screw the copper down to the aluminum and you're good to go.

So... what drivers are out there that can do 25V/150ma/250ma? I believe the chemistry I would drive is LiFePO4 (safe) 2S.

I really do love the boostpuck (4015) but it only goes as low as 350ma (max)- the smooth dimming with the external pot is something I suppose I could wire up to limit it- but that is one more failure mode.

 

[qwertyydude]

2s of LiFePO4 batteries only yield 6.6v. Not enough to drive the MP-L led. And efficiency will go way down if you want to try a 4x voltage boosting circuit. Plus this led was made for lighting class devices, so beam control would be difficult without large reflectors unless you only wanted pure flood. A buck driver with 8s batteries would work but you're talking one huge battery with it's requisite charging difficulties.

If you're trying to rig up a headlamp I think you're better off either learning how to solder or just sticking with a tried and true technology like the magicshine headlamp. It's about as powerful as the MP-L headlamp at the nominal currents you want. And if you're going out of your way to cobble together just to avoid soldering, the magicshine is already available easily. The mp-l led is kinda tough to find retail.

 

[purduephotog]

2s of LiFePO4 batteries only yield 6.6v. Not enough to drive the MP-L led. And efficiency will go way down if you want to try a 4x voltage boosting circuit. Plus this led was made for lighting class devices, so beam control would be difficult without large reflectors unless you only wanted pure flood. A buck driver with 8s batteries would work but you're talking one huge battery with it's requisite charging difficulties.

If you're trying to rig up a headlamp I think you're better off either learning how to solder or just sticking with a tried and true technology like the magicshine headlamp. It's about as powerful as the MP-L headlamp at the nominal currents you want. And if you're going out of your way to cobble together just to avoid soldering, the magicshine is already available easily. The mp-l led is kinda tough to find retail.

Thanks for the information.

I was going out of my way to build something- the Tyra Diffused Spot (http://www.ledil.com/datasheets/DataSheet_Tyra.pdf) shows both excellent spill and enough spot that I could use it for quite a few applications. That was the intended reflector.

I was worried about the efficiency issues on a boost regulator- the draw would still be small and at the lower current density I'd get much less heat- although still compared to 3x XPG running at 1A each...

I just chose 2x LiFePO4 because it was easy. 4x LiFePO4 is just as easy from a pack to build.

It is starting to look like a custom driver is going to be the only solution at this point- I don't see anything else out there at the voltages I need to do so.

 

[qwertyydude]

The P7 led in a mid size reflector like the magicshine has a nice smooth beam, a tighter hotspot but easily remedied with a light dusting of clearcoat on the reflector. This increases the orange peel effect without much impact on total output.

 

[purduephotog]

I really do appreciate everyone's feedback.

I purchased and have been running the MP-L Cree light from HD. The thermal design is quite simple- I haven't cracked certain pieces yet (I like the light too much) but everything is very cool to the touch- not hot at all.

Which gets me back to my original point- all those dies run at a low current produce alot less heat. I understand that I can use an MCE or a triplet XPG at 1.5A to get more light- no worries there- but I also get more heat.

And heat is not what I want. Thus I'm willing to accept lower light output for lower heat output/higher efficiency.

Thanks for the feedback and if you do come across this type of boost topology driver with a good 150ma/250ma drive current please let me know.

 

[MikeAusC]

. . . without having to get all super fancy soldering skills.

. . . Solder the MP-L to a chunk of copper .

I see a major problem here.

 

[purduephotog]

I see a major problem here.

I don't understand how. Sweat soldering isn't hard, and soldering anything to copper is fairly easy. We're not talking tiny traces on the bottom of XPGs- we're talking about a large die on a package that is designed to be soldered.

Tin the surface, add some flux, heat up on hot plate, add pad, begin cooldown.

Pretty simple.

 

[MikeAusC]

we're talking about a large die on a package that is designed to be soldered..

Yes - a die - coverd with phosphor, several optical compounds on a metal substrate - all with different co-efficients of expansion. Some of the components absorb moisture once removed from their hermetic packaging.

If you're interested in reliability of your light, I suggest you read the soldering temperature profile in the data sheet.