Could you share the data for all 4 samples instead of just the emitter averages? It would be interesting to see the spread within the emitters as a backdrop to the spread between them. Especially given all the variances just in the data sheets.
I thought the TangsFire C8 8W XM-L2 U3 1300LM Aluminum Alloy Flashlight Black from here TangsFire-C8-8W-XM-L-U3-1300LM already had a XM-L2 U3?
Sure - TangsFire has a U3. It's a Chinese U3 
. The only real U3 in a light I'm aware of was the ArmyTek, but they ran out pretty quick. Since there was a real CREE batch of U3's released a while back, all the marketing hyped Chinese sources claimed they had them. I finally had enough of these false claims and proved it by comparing one source's claimed U3 to T6's and found the T6's equal or brighter, and U2's definitely brighter. If they claim U3, like TansgFire, chances it's a T6 at best.
Yeah, my $12 Ultrafire C12 I bought last year also came with an "XM-L2 U3", it even says it on the side! 
iOS has them and I will have them soon.
- now this I'm feel'n!! "Trust in Hank" - the only trustyworthy source on that side of the planet. Illumn, IOS, MtnE, Cutter - only ones I know of to trust bins and tints from.
Thanks Richard, you work very hard for us and it’s greatly appreciated!
Thes are the XP-L dyes - interesting because the early availability of "2A" match's well to the XP-L's.
I bought 10 XML-2 U3 2C tints from Cutter a while ago. I went back 2 weeks later and bought the last of their stock which was about 4. As far as I know they have not had any since. I have in the one light, its a triple, two of the XML-2 U3 2C and one XML-2 U2 2C and I think I can tell the leds apart shining the light on a wall up close with brightness but probably kidding myself. Heres three of the most scientific pictures l've ever taken with the light rotated each picture.
Looking at these shots you can pick it.
U2 is on top in the middle picture? (you should consider a carreer in science Steve, looking good! :bigsmile: )
how about mouser?
If you could get specific tint info from Mouser, yes. I know in the past all they had was the range grouping. DigiKey of course, as well - not sure if they carry much though.
The problem with Mouser & Digikey is that they're expensive compared to the other options. Even buying 500 LEDs they are still a little bit more expensive than what I charge for the bare LEDs! The only way to come out ahead is to buy 1,000.
Crash testing an xp-l will give us more information about their limits. Can it handle the current better than xm-l2?
No, they max out at about 7A, while the XM-L2 still gains output up to the point that they crash at ~8.5A
:_(
XM-L2's peak output is a little higher but at much higher current. What's with the fascination of pushing the maximum current possible? I'd rather have equivalent output at less current, thankyouverymuch.
How much more efficient is xp-l at 3+ curent?
Djozz's thread shows a pretty underwhelming XP-L V6 compared to what CREE says. Basically only 50lumens difference from XP-L V6 to XM-L2 T6 at 3A.
On PCT.CREE.COM they show a 177 lumen difference between XP-L V6 and XM-L2 T6 at 3A
Thinking of Djozz tests I would think theoretically V6 would only be identical to U2, but when I look at Pflexpro's test in this thread I see a V5 that is brighter than U2, more like in the line of what CREE says.
+1 HikeLite - these result are quite different. Not really sure what to make of it, but maybe the poor accuracy for cheap light meters to measure neutral light is an influence. djozz's were done on neutrals, while here (pflexpro) were done with cool whites. Also, djozz did the tests on the bare emitter while here they are done in an assembled light (more real world maybe?).
Also I've seen the U2 1D's under perform compared to U2 1A's, but who knows - could be because of the tint's effect with the cheap light meter I use as well. I betcha though the poor performance I've seen with 1D's had an influence on these test results.