I found these:
This one has the highest thermal conductivity…but is corrosive to copper…umm not going to be so good on a device with a copper inductor 
Well, I know this one works. Had it in there previously, but only 2.8 amps. Guess what? Its going back in...
However, that potion is working very very well, strongly recommended. JB weld and some of that pixie dust I mentioned above. And of course, you need a reliable driver. There aren't any unfortunately >3.5amps on 2 - 3 cells.
With silicone thermal cubes I have very little issue with the 5 amp buck driver from Mountain besides the high pitch whine.
Did you verify that your voltage reducer is actually bringing the voltage all the way down to 12 volts? I’m not sure how picky the driver is but if you measure over 12.5 volts maybe that is the issue.
What about 5amps? RMM has been testing one for awhile. He released the parts list so you can build your own, or wait till he starts selling them pre-built. (that could be awhile)
No comment......
Does this help at all?
I'll give em a try, thanks guys.
If you have a 12v + ish power supply and want multiple XM-L emitters, would it be possible to run x number of XHP50’s (in the 12v flavor) ?
Heck maybe heat sink the hell out of them and run them direct drive?!!
Then again I usually have no idea what I am talking about 
Check my math
12v at 5 amps = 60 watts is equal to
6v at 10 amps = 60 watts or
3v at 20 amps = 60 watts
Sheesh no wonder things are frying (Or maybe I skipped somthing)
Have a read through this thread. There are several proven options. Unfortunately, some of the links may be outdated by now.
This one was used by quite a few members, & is easily tweaked by changing the resistors.
Hmmmm... makes me doubt my chances...
I've got a STL-V2 clone type light with one of the FT 3A drivers linked above. I've been waiting for Richard to get the U4-1A bin XM-L2s in stock so I could dedome, mount on a Copper DTP board, and combine with the 5A driver to driver this thrower a bit harder. Now you make me think this might be a waste of a good emitter.
Everybody who has used these drivers said they worked fine with thermal cubes. Which is why I tried them. The cubes seemed like a weak solution to a good mix of potting compound. I learned on the first 5 amp driver. I ran it for 1 minute 2 times back to back. That let the magic smoke out. Then I bought a 3 amp version from FT, that one ran slightly cooler but I wanted to make sure it was heatsinked very well before lighting it up for anything more than 10 seconds.. So, I potted it very well in an established potting compound and within 2 days a lightweight use it let the magic smoke out. I should also mention that before these 2 drivers failed they took out(killed) emitters first. So now I'm using a different driver all together.
Was not using the dropper for the initial build, I was using the first three LiPo cells I had to hand. They were not even fully charged, seem to remember 11.6 volts
But either way, a device for three cells should be good for 12.6 at least, and I;d say it should be a bit more too, …running any device that close to it maximum rating would not be good.
15 volts is what I read for that driver.
I hope you have a better day tomorrow.
I would give you a big ole hug, but ……well ya know…
Thats all I came here to say. Oh…and WE LOVE YOU!
Are you saying you measured 15v or you thought it could handle 15v? Mountain Electronics lists it as a 5.5-12v driver. I’m fairly confident 15v could kill it being as touchy as it is.
Its the same driver only Rich or someone else added some R020 resistors to make it 5amps....
I suspect he(Rich) does the mod though as I have not seen this driver offered anywhere else at 5 amps.
I'm really happy to hear that these drivers are actually working for people. I've tried 2, the 5 amp and the 3 amp, heatsinked well and they both crapped out of me.
I had a couple crap out before I started using thermal cubes. They’ve worked fine since then.
How are you using yours? Or, what did you install it in?
All in Uniquefire 1405’s so far. Most with aluminum pills, one with brass. I guess I should mention that I also put a thick blob of thermal paste between the contact board and main board, maybe that’s what you’re missing?
I’ll turn on the brass build for 4-5 minutes straight and let you know if it dies. It’s unique in that it’s the first one I’ve done using thermal cubes instead of the thermal paste between the boards. Don’t know if that’s a good or bad thing though comparatively, I only did it because I had more space to fill on this one.
Could be a change to a lower quality component on the board? It was not just me who had one burn for no good reason. 2 actually. I can't trust them anymore. Some others have suggested some good replacements. Maybe I'll try one at some point.