Surefire L2 with Cree MC-E

[rayman]

Nice mod :thumbsup:.

Can I ask you what reflector you used for your mod?

thanks
rayman

 

[Energie]

I´ve used the stock reflector.
Some additional details here.

 

[LLCoolBeans]

:thanks: keep us updated about your mod. :thumbsup:

New emitters just arrived today. We'll see if I can't destroy these too.

Going to rework the heat sink again to make soldering easier. I tried just making two new small holes for the jumper wires, but that required me to solder wires to the emitter that were already bent at a 90 degree angle, so that they would then be in position to fit through the tiny holes. That make it a tight squeeze to get the soldering iron in place and that may be why I ended up overheating the emitter. This time I'm going to drill big fat holes for both wires on either side to pass through, like on Energie's custom sink. This will give me more of a fudge factor and the position of the wires won't need to be so precise.

With my method I am seating the emitter and riser in place in the reflector, then inserting the heat sink in behind, rather than attaching the emitter to the heat sink then soldering to it. This makes it much easier to get the emitter centered, but I can't use one single jumper wire. So, what I did before and what I plan on doing again, is to hook splice the two jumper wires together after the heat sink has been screwed into place, then shrink wrapping over the splice. Is this an acceptable method? Seems OK to me, but please let me know if this is bad form. And yes, I am soldering the hook splice together.

 

[LLCoolBeans]

Finished reworking the heat sink.

I should have done this to begin with. Now I've got plenty of room.

 

[mudman cj]

It sounds like you soldered the wires to the emitter before it was heatsinked. If that is true, then that means the slug, and through it the die, had nowhere to dissipate heat as quickly as it came in the leads from your soldering iron. The die will not get as hot if the slug is attached to the heatsink before you solder.

Here is how I did it. First I applied an even and thin layer of thermal epoxy to the heat sink using a razor blade. Then I mounted the emitter to the heatsink/riser. I carefully pressed down on the sides of the emitter to thin the epoxy layer. While the epoxy is still wet, I carefully inserted the heatsink into the reflector to check if the LED was centered. If not, then adjust and repeat until it is just right. Don't try this step until you have made clearance on the back of the reflector for the larger MC-E, of course.

Let the epoxy cure for a whole day (if you can possibly wait that long :nana:), and then feed your wires through the heatsink, solder them onto the emitter leads, reinsert the heatsink into the back of the reflector, and use the screws to secure them. You need the wires to be about 24 gauge to have enough flexibility to leave extra length for attaching the screws while still being able to then get the circuit board to seat against the heat sink.

 

[LLCoolBeans]

It sounds like you soldered the wires to the emitter before it was heatsinked. If that is true, then that means the slug, and through it the die, had nowhere to dissipate heat as quickly as it came in the leads from your soldering iron. The die will not get as hot if the slug is attached to the heatsink before you solder.

Yes, that's what I did. I realized that is probably what caused the problem. I'm going to do it differently this time. I was hoping to avoid using the thermal epoxy on both sides of the riser, but I guess it doesn't really matter.

 

[mudman cj]

Well, your instincts are right. The extra layer of epoxy does add thermal resistance, but the only way around it is to build a new heatsink out of a solid piece. It just so happens that AaronM has some of them for sale. But if you are driving the LED with the stock driver at a maximum of 5W, then I agree that it probably matters very little. If you were to use a driver that pushes it to 10W then you would have more to lose from the extra layer. I used one of Aaron's heatsinks in my KL4 and I am very pleased with it. He has the focal distance spot on to eliminate the dark cross artifact. The beam is quite the wall of light!

 

[LLCoolBeans]

The thermal epoxy between the riser and emitter is curing now.

 

[Energie]

...But if you are driving the LED with the stock driver at a maximum of 5W, then I agree that it probably matters very little....

You are right.
After 10 minutes runtime the head-temperature of the mod. L2 is approx. 42 °C, a bit lower than the stock L2.

The L4 is different, because of the driver in the head.
After 10 minutes runtime the head-temperature is approx. 51 °C.


Surefire L2 stock version (Lux5)
Surefire L2 modified (MC-E)
Surefire L4 modified (MC-E)

Runtime 10 minutes,
Power from the battery approx. 6,2 W (all lights)

´
18 °C temperature of the background
29 °C min. temperature (L2 MC-E tailcap)
51 °C max. temperature (L4 MC-E head)

 

[Illum]

Thats it! I'm getting me a L2 MC-E light instead...51C is too darn dangerous for my liking:green:

 

[toby_pra]

Real craftmanship Energie!

 

[LLCoolBeans]

Emitter and riser are now attached to the heat sink...

Now, for the tricky part.

 

[mudman cj]

Thats it! I'm getting me a L2 MC-E light instead...51C is too darn dangerous for my liking:green:

How is 51C dangerous? At a 5W drive level (stock circuit), that equates to a 15C rise (3 degrees C/W for MC-E) for the die over the temperature of the slug, which in this case is somewhat higher than 51C. If we assume the slug is at 60C, then that puts the die at 75C. From the datasheet, the MC-E would be in the region of 85% luminous flux output due to temperature derating. The L2 LED is running at about the 90% level in terms of temperature derating for comparison (assuming 5W).

I am surprised to see the difference in operating temperature between the L2 and L4. I thought they were driving the LED at the same power level and with about the same heatsink effectiveness, but this makes it seem as though the L2 is better at conducting heat from the head to the body. I think the L2 and L4 drivers are driving the LED at the same power level. Your battery draw measurements confirm this unless one circuit has a much lower efficiency than the other.

Notice how the L4 head seems to be thermally disconnected from the body compared to the L2 as evidenced by the sharp transition in temperature. This makes me think that the L4 is not a good host for the MC-E driven to 10W after all.

Thanks to Energie for posting those thermal imaging pics. Very nice. :twothumbs

 

[LLCoolBeans]

Thanks to Energie for posting those thermal imaging pics. Very nice. :twothumbs

No, kidding. It's great you have access to thermal imaging equipment, what a terrific resource. :thumbsup:

Looks like your job is a bit more interesting than mine, if you have FLIR equipment just laying around the office. Sweet!

 

[Energie]

A thermal imaging camera is a nice toy for flashaholics.
Slightly OT:
Mod. Surefire E1e with a good heatsink design.

 

[LLCoolBeans]

Energie, I'd like to request that you start a thermal imaging thread. How fun! :popcorn:


My jumper wire is ready for soldering. Now it's time to eat some lunch before I get started. :sweat:

 

[LLCoolBeans]

Connections soldered...

Insert heat sink...

Solder to terminals...

Done.

Well, either I over heated it again (not likely, I was very careful this time) or there's something else I'm doing wrong or my expectations were just sky high.

Light functions fine but not much brighter than the old Lux5.

Did I wire it incorrectly? :sigh:

 

[LLCoolBeans]

Just a note here on tint. The 5A tint is worlds nicer than the WD. Much easier on the eyes. It'll be dark in about 3 hours, looking forward to trying it in the dark.

 

[mudman cj]

It looks like you are bypassing the circuit and running direct drive with the emitter wired 2s2p. For direct drive, you need to wire the LED full parallel. And you need to measure the current with different Li-ion cells before using them because larger cells (and especially an IMR18650) will overdrive a parallel MC-E. If you want to use the emitter 2s2p then you want to use the circuit.

 

[LLCoolBeans]

It looks like you are bypassing the circuit and running direct drive with the emitter wired 2s2p. For direct drive, you need to wire the LED full parallel. And you need to measure the current with different Li-ion cells before using them because larger cells (and especially an IMR18650) will overdrive a parallel MC-E. If you want to use the emitter 2s2p then you want to use the circuit.

Bypassing what circuit? The driver is in the tailcap. I thought I was supposed to wire it 2S2P?