The XP-L2 has a lower Vf than the XM-L2, so it might even give you just a bit more run time to go with your higher output! As for the Loctite, I’d say a little heat would be the perfect prescription for that. Red is supposed to be permanent to 260C, but maybe with a good grip, you can get it loose at a lower temperature. A quick search just now found gasoline and acetone separately recommended as well.
I think there are new and old xpl V6’s . I got some from KD (V6 1A) that they don’t dedome easily , like 0D .
I would appreciate if anyone knows where to order old V6’s from .
Got the TM16 apart, found the wires so twisted up that the small thermal sensor wires were actually broken! I replaced all the wires, went with 18ga throughout on power wires. Will be putting XP-L W2’s on 26mm MaxToch mcpcbs and fitting them to the proprietary plastic reflector. In the meantime, I’ll reassemble it and see if the larger wires made any difference. 
Did you do anything special to break it loose? Or just use the lathe?
TM16 doing 4122 lumens, so it’s about the same as what I measured when I got it yesterday. Should have better efficiency with the larger wires…
If you converted this to a 4s setup you would get a lot more current to each emitter. I don’t know if that is feasible in this particular light, though.
With 4 18650s in parallel directly connected to 12 XPLs in parallel, you are getting ~29A, which is only ~2.4A to each emitter. But with 4 18650 in series connected to 4s3p emitters, you would send ~14A, which is ~4.7A to each emitter. With the all-parallel setup, the higher current causes a larger voltage drop across the circuit resistances, which kills the resulting current.
These numbers were calculated using the method I describe here. It assumes each 18650 cell has 0.03 Ohms, the other circuit resistances (wires, traces, springs, FET, etc.) total 0.025 Ohms, and the LEDs used are XPL V6 1A.
This method graphs the voltage at the LEDs as a function of the current, and the forward voltage of the LEDs as a function of current. The intersection of the two curves is the current (x-axis) that will flow.
All parallel:
Googleand(.19**(x%2F12)%2B2.74)
4s:
Google(4.15*4)-x**+and+4**(.19*(x%2F3)%2B2.74)
All well and good, and completely irrelevant. The Noctigon Meteor is intensely compact, there simply isn’t space to convert the cell layout.
Not sure about all that math, never have seen theoretical numbers actually depict the amperage a given cell supplies to a given driver/emitter. Interesting though. I have a light with 4 Nichia 219C in parallel, run from a single 18650. One particular cell can deliver 21.1A to this quad, which makes it difficult for me to believe that 4 LG HE-4 would only be capable of 29A, as this is the cell that delivers 21.1A all by itself in my quad.
And then, of course, there’s the wrench in the gearbox, moving to the new XP-L W2 emitter. It obviously has a much lower Vf, having run at 6.6A in a C8 where the XP-L V6 1A you mention does not. At any rate, I’m not in it for the mathematical exercise, but merely to build a light.
Maybe I’ll get an amperage reading on this one, before and after, see what the reality is.
xp-l2 is coming 
must have ultra low Vf
Too bad you can’t make it 4s; almost twice the current to each emitter is a nice boost.
The analysis does work. I used it to predict the current in a single XPL DD, a triple XPL, and my 2s 7xXPL convoy L2 mod and they all were very close to reality (within ~5%). Of course, for different lights the circuit resistances might be different (tailcap/switch resistances, etc), and if you bypass the tailcap while measuring with a clamp meter the current will be higher.
Except for the fact that the analysis doesn’t work.
4 LG HE2 cells, button tops, give me 33.70A at the tail as it’s set up right now. Will test HE-4’s directly.
As far as bypassing the “tailcap” there isnt’ one on the Noctigon. There is a copper plate with beryllium springs, which I removed, then I used a strip of copper contacting the tail end of all 4 cells to get my clamp meter reading from. Virtually no difference in the copper adapter for the read as compared to the copper plate with copper springs. (FWIW, I bypassed the beryllium springs anyway when I built it up some time back.)
It does work. Since I don’t know the details of your light and measuring setup, it will only give an estimate, and I’d say given that, the fact that it’s within ~10% is pretty good. If you don’t want to take advantage of the tool I won’t make you.
FWIW, LG HE4 button tops just showed me 11,040 lumens.
Thank you, but instead of doing calculations I’ll stick with taking actual measurements. I only really take the measurements to see what a modification gains me, or loses. So it’s relevant in perspective anyway and not necessarily real world as our eyes don’t see and calculate lumens any more than we can feel amperage numbers coursing through the light.
I mod for stress relief, for me trying to keep up with all the number crunching would be stress inducing. Wasn’t always that way, but I’m getting old.
Sometimes there are little things that happen in a build that I can’t figure out, like in this one if full contact of the cells was made then the emitters would glow ever so slightly, even off. So current was bleeding in somewhere, but I couldn’t find it. After removing the copper contact plate in the tail and cleaning it up to re-install it, the light no longer has this glow. So that in and of itself made the current reading worthwhile. 
Of course I don’t want to influence anyone into making the hobby less enjoyable for them.
In my view, we are all doing calculations/predictions of some sort, even if they are based on the intuition that one has gained through experience.
Anybody find a reliable source for these W2’s yet?
Bump
I don’t know? Pays yo money, takes yo chances, I guess!
I’ve got a second order of 10 on the way, first 10 were correct, we’ll see if they’re consistent.
That ali seller has also xpl hi W2 so I would not expect to get W2 HD there
I got 10 XP-L W2 1D, have another 10 on the way.
How could you tell if an HI variant was W2 or V2? Output would be the only indicator that I’m aware of, haven’t seen any W2 HI listed nor have I seen em listed by Cree, haven’t checked lately though.
And there has only been very few XP-L HI V3’s out that I have seen too!