Well done Bort. That is great info. So true, but I would not have realized it on my own. Thank you 
Thanks for the reference to the Convoy M1, Bort. I didn’t know about this light and the choice of emitters has me intrigued. 8)
Good to see that you weren’t willing to give up on this project…! And out there who would know?!
Thanks for the pics. Used to drive around there back in HS days w dad in the family wagon, had the old canvas water bag on the front of the radiator of the car, canvas tarp & tent poles to create shade mid-day and massive Cibie and Marchal driving lights that just disappeared into the inky darkness for night crawls on some knarly roads…thanks for making me think of those trips.
Stock, my experience says U3 over L2 everytime, w budget lights that I have multiples of. Same w some P60’s as well.
>>>>>>>Stock, my experience says U3 over L2 everytime,
>>>>>>>w budget lights that I have multiples of. Same w some P60’s as well.
I ordered the kai domain C8 U3 recently and regretted it the next day. WTH did I order that for? I have like 10 C8s at this point.
But the regret vaporized when it arrived two weeks later last Friday. That light beats just about every C8-C2-C12 clone that I have, including the unbeatable famous xin and a couple u3 self-modded lights.
The thing I love about the light is that it never seems to dim. Even at 3.8 volts, it STILL beats every other light I have. That thing is bRIGHT. How bright? Well, I just ordered a lux meter because my other light meters really aren’t suitable for this kind of measurement, so I’ll know soon enough.
Further, last night, I was bored. So against my better judgement, I put a fasttech U3 into a Uniquefire C108, a nice light with meh performance with its stock T6.
All I can say is wow!!! WOW WOW WOW.
Now here is a recommended simple mod. Not as bright as my new kaidomain C8, but darn close. What a thrower!!! With the existing T6 in the C108, no fourth phone pole in my testing range was visible. With the new U3 in it, that fourth DARK pole against a dark background lights up like no tomorrow. (I can’t do pictures of the pole for another couple weeks. There has been a HUGE moth hatch out here and I am absolutely COVERED in moths — including bird-size sphinx moths — within 10 seconds of turning on any light.)
But I haven’t given up on the xm-L2. I also ordered the famous light malls C8 XM-L2, and will put that up against the other C8s.
Finally, I REALLY REALLY hope that the people who have had great luck with L2 upgrades will get some beam shots up on BLF showing how it does in real world situations.
Tried my aurabuy C8 XM-L2 AGAIN last night (enduring MAJOR moth attacks). Confirming again, when the battery reaches about 3.8-4.0 volts, the bottom drops out of that L2. Really huge loss of light output. Like I’m suddenly on the second highest mode instead of the highest. This must be the higher voltage requirements of the L2 that everyone is talking about. (The modded Small Sun T08 does the same thing but takes longer for it to happen because of the two 18650s in parallel.)
I’m sure that I will be bombarded with how I measured the voltage wrong or whatever, but to me another reason that the L2 is not a great choice for a swap-out for an existing T6: The light REALLY dims as the voltage drops in this really cheap aurabuy C8 host. It also dims in the L2-modded SS T08.
It looks like the L2 really needs a dedicated driver that will optimize its higher voltage requirements. Yes, as BLFers have been saying, it looks like L2 will really need 2 cells, probably in parallel, to keep feeding its high-juice requirements.
Again, these observations are about MY experiences with these two lights in a XM-L2 upgrade. Your results may be different with different lights.
Presently, 2*series lithium-ion buck drivers are the way to go in order to drive an XM-L2 hard. An XM-L2 is more difficult to drive hard with 1*series lithium-ion cell than XM-L due to higher Vf. Bill Zeng and his engineers at LightMalls have been working with several members here on a single cell, 17mm high-current boost driver to hopefully rectify that. Check out the last few posts at the bottom of the page of this thread.
Did a timing test on my XM-L2 U2 1C/SinkPAD, Nanjg 3.85A C8 light a couple of days ago - used a Samsung 20R (2000 mAh) battery - starts out at 1,250 lumens or so, after 28 minutes of runtime, batt voltage is at 3.6v and lumens output is about 800. So, yes it's a significant drop but still brighter than it was stock. In another test, compared a U3 1C/SinkPAD 3.85A light to a XM-L2 U2 1C/SinkPAD 3.85A light, and after 15 minutes of runtime, the U3 caught up and passed the XM-L2. Forgot details of the batts, etc., got it written down at home.
Alll depends on your thinking, usage preference, opinion, but for me, I'd prefer the first 10-15 mins of super bright (wow) effect, after that, well guess you don't have to worry too much of over-heating because it really cools down at the lower output levels, and the further you go down on the battery, the more battery runtime you got left. 
I agree with that point of view. Given the choice of a light that starts at 1250 lumens vs a light that starts at 1170 lumens, I will always pick the higher output light. Especially if the 1275 light can hold that lead in power for 15min…
Plus I’ll bet if tested with a bigger capacity Panasonic 3100-3400 that lead might go up to 25min or so before the voltage drops enough to give the edge to the U3.
On my tests of an XML2 with a 3.0A Nanjg driver, the driver went out of regulation when the cell voltage dropped below 3.60V
I’m working on that. We’ve got a BTU, T08, C8, & 2ea TN31’s that have all had XM-L2’s swapped into them. Everyone of them is VERY impressive to my and my brothers eyes. And according to out meter and lightbox.
But we will be getting beam shots done ASAP. Tried the last two weekends, but complications arose.
Just out of curiosity I decided to try this with my Small Sun ZY-T08. Well sort of I swapped the original led for a XM-L2 T6 1-C 6500.
First of all it was a pain in the back side trying to get it to focus properly. But I did manage it, but I was also disappointed. There really was no real noticeable change. Maybe just a tad bit brighter but not enough to make it worth the hassle. However, what I did next made a huge difference.
I de-domed the L2 using a gasoline soak for about 6 hours and filed the reflector hole until I got the hotspot perfect, it pretty much took the very bottom of the reflector out. Now it’s no longer a cool white but more neutral without being yellow. The hot spot is smaller and much more intense. It throws farther now by far.
For me I am very happy with it. I like the tint better and the extra throw is super. It doesn’t have as much usable spill but that’s ok with me because it really is a thrower. If I wan’t flood I have other lights for that.
But family like the urban. Lifestyle only downside is WATER flashflood and first responder response time.
It depends on how good is your constant current circuit. And temperature of your LED are.
Forward voltage varies depending on temperature.
XML : –2.1 mV/degree
XML2 : –1.6 mV/degree
At 25 degree, XML at 3A has spec of 3.35V.
At 85 degree, XML2 at 3A has spec of 3.3V.
With 60 degree difference in temperature, if calculated back to 25 degree, XML2 at 3A should have spec of 3.3+60*0.0016 = 3.396V
This shows that at 25degree, XML2 is harder to drive as it needs 3.396V to have current flow of 3A while XML needs only 3.35V.
Looking this in power watt term, it is 3.396*3 vs 3.35*3 = 10.188W vs 10.05W, a difference of 0.138W
Let’s calculate XML at 85 degree, XML at 3A should have spec of 3.35V-60*0.0021 = 3.224V
This also shows that at 85 degree, XML2 is harder to drive as it needs 3.3V to have current flow of 3A while XML needs only 3.224V.
Looking this in power watt term, it is 3.3*3 vs 3.224*3 = 9.9W vs 9.672W, a difference of 0.228W
Summary of above:
1.At same temperature, XML2 is harder to drive compare to XML.
2.The higher the temperature, the harder to drive XML2 if compare to XML.
Why XML2 then?
Because XML2 has better intensity at higher temperature.
At 25 degree, minimum luminous flux of cool white U2 Bin at 700mA, XML2 calculated score 340lm while XML score 300lm.
I believe most of the improvement seen from modding to XML2 is due to better heatsink using copper plate. I mean if someone were to mod two TN31 with XML and XML2 under same good copper heatsink, difference will be small if TN31 circuit favour XML more when it comes to above 15W or so? Higher wattage means higher temperature, higher temperature means XML2 more difficult to drive. And XML2 relative flux drop more when temperature increase.
Stock XML TN31: 1169-1125 lumens (@start-@30 secs), 128 kcd
Stock XM-L2 TN31: 1326-1281 lumens, 156 kcd (13.9% gain in lumens, 21.9% gain in kcd)
Same driver, same batteries doing the test, same star. Didn't do amp measurements but I believe they are the same - the driver has the same model and date stamped on it.
Of course with 3 batteries in series using a buck driver, there's a big advantage to the XM-L2 in this light to maintain volts/amps and resulting output.
Tom E, did you check the resistor value? Are they same?
Because Thrunite doesn’t mention it is XML2-U2, I think it might be XML2-T6, while XML is stated to be XML-U2.
See this huge difference you posted make me suspect Thrunite increase the amp with lower resistor value for XML2 model.
I'm confused - spasmod in his review of the TN30 and TN31 lists them as XM-L2 U2, and -JOE- is his sales thread lists it as XM-L2 U2, and he's got a pic of a label listing it like that, though can't be sure it's an official ThruNite label. I see the ThruNite site doesn't list it, so I just emailed them asking what the bin is. Have you done that yet? I didn't notice you questioning spasmod or -JOE- about it either in their threads.
I'm thinking the XML U2 to XM-L2 U2 upgrade they did corresponds well to the percentage increases I'm seeing - it's not the "20% brighter" number ThruNite quotes either, and their throw/kcd calculations are wayyyy off, always have been. I wouldn't call 13.9% hugh...
I'll check the resistor values when I get a chance, but I'm thinking I confirmed they are the same already because I looked for that, though the old XML version I have has a resistor laying in top of one of the existing ones (
). There is a pic around of the resistor in the XML version though, so it could be confirmed.
I wasn’t paying attention to this XML2 thingy in the past, and I didn’t read every post. I can understand the improvement from XML-U2 to XML2-U2.
I am guessing some brand can’t get hold on quantity of XML2-U2, so they have to put other bin in their product, that is why they don’t put XML2-U2 in formal advertisement. If even worse, to match the lumens output, they might increase the current a little bit.
That is the risk of using latest higher bin of LED, quantity is limited.
Maybe so, I was thinking along those lines as well. ThruNite is a weird one - China company (not Hong Kong), supposed to be high end, why they quote 90000 cd and 840 meters boggles my mind, and they have had some real suspicious QC problems, well published here, but seem to support/service their products well. The kapton tape left on the underside of the emitter on apparently many pieces is another mind blower...
Here's ThruNite's response and my Q:
---------------------------------------------------------------
THanks Tom:
It is U2 1C bin. we always choose the best !
Regards
David
------------------ Original ------------------
From: "Tom E";
Date: 2013年6月11日(星期二) 上午9:54
To: "sales";
Subject: TN31 XM-L2 model - what exact bin is it?
Hi!
Your website lists many specs. on the TN31 but fails to mention what bin the CREE XM-L2 actually is. So, is it a T6 or a U2? And if you can answer this question, why not post it so we know what we are buying? Please explain.
Thanks,
Tom E
I have many multi-emitter flash lights light X100, TK70, Darth, S6330, RC40, 2 x Terminator and TN30. Only TN30 has different color in one of the emitter, looks like 2 T6 and 1 U2, damn ugly beams it puts out especially on white wall. I saw one post complaining about this before so it is not unique case. Adding flickering issue and I had clean the reflector due to finger prints, I told myself I won’t buy this brand anymore.