I have also run 11A+, but these buck drivers seem to be able to give a very large spike when pushed too hard in buck mode. With 2S cells I think it's just about impossible to kill an MT-G2, but with 3S not bucking correctly or giving a large current spike apparently it is.
Wow that pill is ridiculous.
Yeah i literally killed half of it :_(
Nice build and report. Thanks for taking one for the team on the 3S/Dry/MT-G2. Could you check to see if there is any kind of gap between the emitter and the thermal pad? A real close macro shot may help. Also, it you have the emitter base (Assuming Notigon) well connected to the pill when the failure occurred?
Just wondering it was the Dry allowing too much voltage or a thermal connection issue.
Measuring 7 amps of current draw from 3 cells, would probably mean that the led was pushing maybe over 15-16 amps across that MT-G2. 
I was just beginning to think the MT-G2 was invincible. Now this, I was really expecting someone to kill one by hooking it straight to a fresh car battery or plugging it directly in the wall receptacle. :bigsmile:
It does look like there is a gap and not enough solder, could this of did it. The LED is a little messy because after I killed half of it, it went in the dead diode pile.
This is the side the damage is on……
This is the opposite side that is still working……
Nice effort. Thanks for taking one for the team. Hows the beam? I have a couple of those reflectors I will one day build into a MTG-2 light and was hoping it would tighten the hot spot up compared to other reflectors that size.
I am happy with it beam wise. It isn’t as tight as the Torchlite 2x26650 which is now discontinued. The torchlight reflector is deeper compared to it. If there was a way to have the Torchlite reflector with this new host it would be perfect. IMHO the much beefier pill and the ability to perfectly fit the 32mm dry driver makes it a better host then the Torchlite. We are getting some crappy freezing rain weather right now so it’s not a good time for night shots lol I’ll post some tomorrow night most likely.
I tried to get some pictures but this is all I could do tonight
New 2x26650 Host @ 5A
These next two pictures are using the same setting and you can see the difference between the reflectors
This is the Torchlight MT-G2 @4A
I have Torchlite host, still deciding on what to put in there… do you think the new DRY driver fit in Torchlite host?
No the dry driver won’t it is a weird size pill, it is meant for a 23mm driver or something weird like that. I had to solder a 20mm contact board to the inside or the driver retainer ring and use a qlite with the mt-g2 mod. If you ever want to sell it let me know 
I think that host would make a thrower with a de-domed XM-L2 at 5A+ but there is no sense in ripping apart the MT-G2 since it is running just fine.
Thanks for the pictures. The spill is definitely larger on this light but the hotspot appears about the same size or is the picture deceiving?
The bottom of the reflector looks pretty wide. Did you have any trouble with clearance between the bottom of the reflector and the wiring soldered to the LED board?
Hotspot is definitely bigger it was hard to get a picture of last night but you can see from the picture the flood is definitely different.
I had absolutely no issues with the reflector hitting, I did make sure that my solder joints weren’t huge though. If anything it hits the screws used to hold the board down. I also use one of the 3m gaskets from mountain instead of the paper sticker that is supplied.
I was also concerned about the solder joints hitting. If you get them pretty flat it isn't an issue (but should be insulated) because the reflector does seem to hit the mounting screws before hitting the joints. As another benefit it seems to help protect the dome of the MT-G2 which can be easily damaged by a reflector tightening onto it (again, don't ask me how I know 
)
sorry… I was trying to buy more of this host… and regret I didn’t buy more 
but the new 26650 host from Mountain electronic looks similar enough 
it’s a beautiful host and plenty heat sink, the only issue I have with this light is the pill are for LED is too deep.
the LED couldn’t get close enough to the reflector, heck the reflector didn’t even touch the MCPB.
I was dry testing it by putting the noctigon MCPB on top of 2 copper pennies to make it a perfect fit.
how did you put yours?
Hmmm, there is an indication of some type of void in there. Question is, how large is it. We may be be seeing the only solder there is or there may be solder under most of the emitter.
That's one thing I don't like about the Notigons. The copper board is not pressed up like in the Sinkpads. I quite buying MT-G2's premounted on boards for that reason. I rather reflow myself and get it right.
The void may have cause the failure or it may have only contributed to it. At that current, any gap probably is fatal. I guess the Dry would need to be tested for output using 3S cells and the voltage sensor bank shorted to know if excess voltage is delivered to the emitter. I don't currently have the knowledge or equipment to do that.
The old Noctigon boards like that do not have the thermal pads up. Get some new ones like the 3XP and you will see all are up (got a lot of them to know). Ask Hank for the newer batch that have the thermal pad at the same level.
it does look like a gap under the led, who knows how much of it is actually soldered to the mcpcb, and how much of it is hanging in the air.
i used sinkpad2 for mine mtg2, it has heatpad and contacts on the same hight, no gaps.
Been testing my Noctigon mounted MT-G2 at 10A and 11A for a few minutes at a time. I know that DBCStm has been also. I don't think that the issue here was the amount of solder under the emitter but that it simply received too much current.