I've had two XM-L's desolder from their boards

Two different lights.

1. Surefire C2 mod with a P60 drop-in. I turned a piece of copper and AA'd it into the recess of the brass heatsink to add mass. 2.8 amp driver (Shiningbeam)

2. Maglite mod with H-22A heatsink that I cut the tower down flat to accommodate the 20mm star. Star glued on with Arctic Silver ($$$). Direct drive off 3 alkaline D's. 3.36 amps at tailcap.

No blue light seen on either light. I'm wondering if the melting point of the solder some guys are using is too low? With good heatsinking the XM-L should be good for 3 amps. Am I correct?

Thanks,

Greg

certainly seems things are heating up

how hot did the hosts get? and how long did they run befor de-soldering?

first thing to suspect would be a bad thermal interface from the star to the heatsink, did you get a really thin layer of AS?

it could be a low melting point solder being used, were the board from the same supplier?

Umm... wow. Typical 60/40 Lead solder melts at 188C. The max Tj for the XM-L is 150C, so I'm extremely shocked the LED didn't fry.

Also 3A is too much for a P60 drop-in for long duration use...

Did you compress the adhesive as it cured? In order to desolder like this the LED would have to heat rapidly, as stated, the max temp for operation is 150C, and running above that for prolonged time would fry the LED.

May be star board has bad thermal conductivity.. I saw some cheap star boards (MCPCB) has low thermal conductivity. even if you screw it to larger heat sink, no success.

The one that disassembled itself last night was the Mag mod. Nice massive heatsink. Expensive AA Silver thermal epoxy. Everything was nice and flat. No high spots to hold the star up off the sink. I weigh 250 pounds and use my hands for a living. figure I can apply good pressure to the star while the glue is curing?

I popped that star base out and reflowed the emitter onto it with my heat gun. Works like a champ again.

I looked at the epoxy film and it was pretty darn thin.

I put a different emitter in it and it's running. the sink is working. The light gets warm.

I don't know.

jasonck08 said:

Umm... wow. Typical 60/40 Lead solder melts at 188C. The max Tj for the XM-L is 150C, so I'm extremely shocked the LED didn't fry.

Also 3A is too much for a P60 drop-in for long duration use...

Click to expand...

I agree. I now make a chunky solid aluminum sink in my lathe that presses into the light body for the single 18650 P60 style lights. The reflector sits on top of the light engine.

I just ordered two lights with anXM-L led I will post if I have any issues.

Greg G said:

I agree. I now make a chunky solid aluminum sink in my lathe that presses into the light body for the single 18650 P60 style lights. The reflector sits on top of the light engine.

Click to expand...

+1, 10W is too much for a p60 sized light when its on constantly, but if his thermal connections were solid, his hand should be nicely burned well before the LED desoldering on its own. Even at 10C/W the light body would have been at 80C, a bit too toasty to be held.

My 6P with single XM-L generic module P60 gets hot quite quickly on high. I am using the aluminum foil heatsinking method, but took great care to use flat pieces of foil, no crinkles. I manged to wrap the drop-in 3x more wraps than with using hand-flattened pieces. My logic tells me that means there is less air between the drop-in and the host wall.

Haven't run it through a full cycle or anything, but I've run it 10 minutes straight with no visible negative effects.

Edit: the real reason I posted here is to talk about the smaller MCPCBs. The MCPCB core is punched out from an aluminum sheet. Therefore, the bottom of the star isn't flat, but rather it starts rounding off towards the edges. I know the PC overclockers go ape-sh about polishing their heatsinks to a mirror finish, but that's because the chip surface is also a mirror finish. I wonder if sanding our stars would have any appreciable positive effect on thermal transfer.

I've always taken it for granted that the bottom of the star, round base, etc. that I was using was flat. I started on a 3 XP-G light engine for a mag mod tonight.

I touched up the bottom of the stars on a piece of 120 grip wet and dry. They weren't flat. Two were low in the middle, and if applied to the sink without sanding, the epoxy thickness would have been thicker there. Not good.

Thanks for that food for thought.

CKOD said:

+1, 10W is too much for a p60 sized light when its on constantly, but if his thermal connections were solid, his hand should be nicely burned well before the LED desoldering on its own. Even at 10C/W the light body would have been at 80C, a bit too toasty to be held.

Click to expand...

The problem is......they will turn the light on then set it down and forget it about while they're working on the car, etc. I tell them the high mode is like turbo, it's for short use only. They don't listen. in the future Im stepping them down from the 2.8 amp board and using the 1.4 amp board made for the XP-G with the XM-l emitter. It'll be safer for the unwashed masses.

I would say that they do benefit from being sanded flat. Every pcb I have ever seen were not flat. I always sand to a good flat finish. The less gap the better.

try some lead free solder

kuksul08 said:

try some lead free solder

Click to expand...

I avoid lead free like the plague. Fixed too many broken joints because of that garbage.