In a like-for-like comparison, of course there's no big jump in output. The main benefit is more headroom, I guess you could say - they can be driven a lot harder and keep going up in output past the point where the original XML begins to fall off a cliff. What led you to believe there'd be tons more light with no other changes?
There is a thing that not many people address: the majority of the chinese flashlights -according to specs- are fitted with XM-L T6 leds, and for whatever reason we all truly believe that the bin that is used is indeed T6 .With all variability that is going on in a flashlight it is notoriously difficult to verify that bin, so there is no punishment whatsoever for the factories for using a T5 or even T4 (T4 is the lowest bin that ever was available for cool white I believe), they sure have that sitting on the shelve (as I understand leds are bought in reels with various bins) . It could just be that the reason that upgrading a T6 to a U2/3 makes such a difference is that you are not upgrading a T6, but perhaps a T4 or T5. And the difference between upgrafing from XM-L to XM-L2 will in comparison not be impressive.
>>>>What led you to believe there’d be tons more light with no other changes?
One BLFer said that he got a 20% gain over a T6. But I just wanted to try it …. The fasttech U3s really did work well, an average very noticeable 15% on quite a few lights. Figured it would be great if the L2 would add another 10% on that. You’d think if Cree was making a big hoopah about the XML2 that it would beat the U3. U3 looks brighter to me with existing drivers anyway. Someone here was talking about how they need 2 cells? Need to look at voltage-output chart I guess.
I have one more 21mm L2 lck-ed star left. Was going to try it in a Sky-Ray clone … A something I think A0? Can’t remember. Anyway, bought another one of those to try a U3 or L2 in it.
I’m keeping many of my T6s stock though. Warm lights out here don’t work well. Adding a brownish cast to a brown landscape makes everything this really unappetizing yellowish brown. The warm LEDs may work great when used on green foliage, but in the desert, the lowered contrast and kinda ugly hue make for really reduced range, reduced acuity and extremely muted color rendition. IMHO anyway.
Gotta get some sleep. Haven’t seen 2 a.m. in three years. 
Were you swapping T6’s for L2’s or U3’s for L2’s?
Are U3’s as cool as U2’s? If so it would make sense that a swap from a very cool U3 to a kinda warm L2 might seem dimmer. But I don’t see how a L2 wouldn’t be brighter than a T6 which isn’t usually all that white.
Also you mentioned the moon, with a full moon everything seems much dimmer always. Did you put a stock ZY up against a L2 ZY at the same time or were you just remembering how it was the last time? Hopefully that is what the big difference is.
Well anyway I hope it works out one way or another.
Ps. Let me know if you want to trade those T6’s I need some for a project.
1) state the exact names, bins, tints, othewise it’s not comparable
2) XM-L U3 = XM-L2 T6 in output, so yes they look the same in brightness
3) XM-L2 T6 is brighter than XM-L T6, see my pictures in another comparison thread
4) provide pictures and measurement numbers or leave your rants for rant thread
I put an XML2 T6 (on sinkpad) into a light that was doing 475 lumens with a (supposed) XML T6 on aluminum. Output went to 650 lumens. It is being driven at 3A, so the copper sinkpad is not too important there.
some valid points Jack, but not an especially helpful tone.
Ubehebe, tints and bins of old vs. new would be helpful. A like to like swap (T6 3c to T6 3C) should net a small but noticeable difference at high drive current (3A) and a just about noticeable difference at 1.5-2A (in my experience). Going from U3 CW to T6 NW won’t be any difference in brightness. Plus there’s the possibility that you were replacing a genuine U3 with something that wasn’t as advertised.
Absolutely on the color. I spent several weeks in CA last summer - after growing up in Bay area and S Cal had forgotten just how brown and dull the colors are once things go ‘golden’ for the rest of the year. Additionally, we went into Yosmite NP for a few days and all that rock, sure does reflect light and colors different from what I am used to back east. I needed something bigger to throw on my nightly wanderings… Thought two of my favorite lights were going south, nope, just the notable difference in the native colors.
Told my wife we have to go back again this year so I can decide which I like better - plus I have more lights to play with - bigger and better.
Don't look at it as time wasted. You now have brand new LEDs in those lights that aren't going to burn out on you any time soon.
TGIF!!!!!!!!
Lets skip the modded aurabuy C8 for now. That adds confusion, and I don’t yet have another light to compare it with, one on one. So this comment pertains ONLY to the modded Small Sun ZY-T08 and the stock SS.
>>>>Were you swapping T6’s for L2’s or U3’s for L2’s?
I swapped a XM-L2 for a existing T6 in a Small Sun ZY-T08.
>>>>Are U3’s as cool as U2’s?
Some of the fasttech U3s are cool; some are warm. They are very inconsistent. Blfers suggested that was normal in another thread:
>>>>>Also you mentioned the moon, with a full moon everything seems much dimmer always.
Absolutely. The full moon is a killer out here. You could read by it. But as I said, I was testing against phone poles receding into the distance. Judging against the light hitting the fourth one in particular. I hesitate to estimate how far that fourth pole is because I might be off a few yards, but I’d guesstimate that it’s 1/4 mile away. (I very roughly measured it with my car odometer driving on a road parallell to the phone poles, so I could be SEVERELY off). I will measure with my GPS later when I get a chance later today.
>>>>>Did you put a stock ZY up against a L2 ZY at the same
>>>>>>time or were you just remembering how it was the last time?
Nope, no memory involved. Two Small Suns side by side. One with the stock emitter. One with a lck-led XM-L2.
>>>>Let me know if you want to trade those T6’s I need some for a project.
I don’t think you can use the SS emiitter for anything. it is mounted on a humomongous star.
>>>>>It could just be that the reason that upgrading a T6 to a U2/3
>>>>>makes such a difference is that you are not upgrading a T6, but perhaps a T4 or T5.
Absolutely. Who knows what was originally was in the SS. But since it was one of my brightest lights, I assume it was a T6. I will say 100% for sure that the two SS flashlights produced the exact same output before the mod, as far as I could see. I kept one stock as i said because I already know that the XM-L2s i have from lck-led are warmish. As I said, they don’t work well in the desert.
>>>>> so the copper sinkpad is not too important there.
No, I know. I just decided to use the copper ones because he had them.
>>>>> put an XML2 T6 (on sinkpad) into a light that was doing 475
>>>>>>lumens with a (supposed) XML T6 on aluminum. Output went to 650 lumens.
Thanks TP for chiming in. I think you volunteered that before and perhaps was one of the reasons I decided to try the XM-l2. I’m not getting an increase that anywhere approaches that.
In my original experiment, I used tenergy cells that were 4.17-4.20 in both lights. Right before I went to bed, i tried once more with four fully charged tenergy cells, all exactly at 4.20. This time, I maybe noticed that the XM-L2 light was a hair brighter. But just a tiny tiny hair.
I will post some beam shots later. I should have done that before, but as I said it was 2 a.m. when I finished. I should have a gossen meter here also that measures lux and anything else. I will try and find that for measurements.
You guys are all completely right in criticizing the lack of documentation. I should provide more documentation. I was just so ticked after the money and the work to have no real gain that I posted without it. Fasttech U3s DID produce gains of 5-20% in all the lights there were installed in, over the last moth or so.
I have two more lights to try. Both are EXACTLY the same. I will do beam shots at least before I change the emitter. If I can find my light meter I will also measure it. I will be using two of these lights. One will get the L2; one will remain stock.
Or I HOPE to mod one of these. Like I said, I had minor issues fitting a 21mm star in the SS because of the massive stock star. I beleve this light also has a giganto-normous star. If it looks like i will run into the same problem, I might nix the third light, but we’ll see.
Thanks for all your comments. I have a ton of work, but i will try and get some beam shots up over the next couple days. (as well as shots to prove that two SS flashlights were used).
And tahnks for suggestions in another thread about using flux to get the solder to stick on the stars. Some fasttech U3s took the solder. Other times, the solder would just pop off, no matter how high i heated the star. The lck-led L2s DEFINITELY require flux to get the solder to adhere. It just pops off the star without flux.
This doesn’t seem like “rant” thread to me. Ubehebe is providing his real experience and it should be respected. Members should not be criticized for not providing “enough documentation”. It is nice when it is provided and I think it is proper to ask for politely.
Back on topic. Thanks for the post Ubehebe. I think this is a worth while discussion because many folks are debating whether to upgrade to XML2. It doesn’t jive with my experience, but I’m biased because I’m so partial to NW tints.
Also, please try a xm-l2 upgrade yourself in whatever light you have. I would love to see more comparisons.
I would be happy to find out that these L2 emitters are somehow dimmer than usual. I REALLY wanted to upgrade some other lights, using the XM-l2 instead of the U3s. So don’t believe me, and do your own tests. Jut please post your results here on BLF.
A 10-15% increase in brightness is all but impossible to discern by the naked eye.
I believe a difference in tint is likely the key factor here. A lux meter, along with some current measurements, would likely suss out any other contributing issues.
Real-world feedback is always important, folks; please don't be so quick to dismiss the practical for the theoretical.
>>>>>I believe a difference in tint is likely the key factor here.
Yes, I think so too.
If I try the aurabuy C8 (modded with the xml-2) against all comers on a white wall, it SLAUGHTERS ’em. Outside the same lights slaughter the aurabuy c8. To me, that means tint is playinga huge role.
Haven’t I read somewhere here on BLF that warm emitters perceptually seem to the eye as 15-20% (or so) dimmer over the more bluish emitters? If that’s the case, the warm emitter would be perceived as less bright and that would pretty much cancel out the XM-L2 gain in output when compared to a bright white T6.
Someone on BLF (I think) said that they meaured a 27% lumen increase with a XM-L2 over a T6? But the wamer tint of the emitter makes the XM-L2 seem 15-20% dimmer. So, when all is said and done, I am only seeing a 7% increase with a L2 over a T6 — and that seems about what I’m seeing in the aurabuy c8 with the L2 mod.
Gotta get some work done. I’ll try to post some beam shots later.
Yes, Chicago X, I think you nailed it. I think the XM-L2 is probably theoretically brighter in — say — an integrated sphere, but the warm tint makes it seem dimmer TO THE EYE than its “real” output and — therefore, it would SEEM (to the eye) to be only very marginally brighter than a bright white t6. 
Compounding this perceptually dimmer warm tint is that I am looking at a brownish-tinted light on a brownish tinted landscape.
Yes, the more I think about it — it’s the tint. Makes perfect sense.
I’m not sure about CW LEDs being perceptibly brighter then NW LEDs of the same bin, especially as this will depend hugely on the environment (and colour) they’re being used in. I do think that NW tints provide better colour contrast in the wooded/ leafy environments I use my lights in, to the point where I’d rather use lights that are NW and half as bright over CW lights.
I seem to remember reading that eyes don’t respond linearly to increases in light (damn those adjustable pupils!) and that you need something like 4x the emitted light for it to appear twice as bright. Hence a lot of people like Mid modes that are 25% of the max drive current.
At the end of the day, as long as you’re happy with your lights, who gives a monkeys?
I’ve only built up one xm-l2 light, a c8 upgrade host with op reflector, ios 4a ramping driver and nw emitter.
It’s definitely not super bright compared with other lights but it does have a nice beam so I am happy with it, it did show me that I’m still happy with the output of my xm-l 4c emitters so dont need to panic about buying loads of xm-l2’s.
I think you’d see a bigger difference in fitting an xm-l2 u2 in place of an xm-l t6, the nw definitely softens the “wow factor”.
I also work on the same basis - if I cant “see” a benifit by eye, its not really worth it, I dont see with a lux meter, I see with my eyes. If they cant see the difference then I’m happy with lower bin emitters. In my case, nw is more usefull, but I generally use my lights in shiny reflective environments, nw lessens the glare so lets me better see what I’m doing, I can see where your coming from, in brown environments a cw tint would probably be more usefull. Like everything else, modding is usefull in tailoring things to suit your specific requirements, which wont be the same as mine.
i had some lights upgraded, one is keygos M10 using U3 and the other is ultraOK using XM-L2
TBH i think the keygos M10 is slightly brighter… they both pushing around 4 Amp.
slightly disappointed as i thought the UltraOK will throw farther since it has a bigger head.
i’ll say XM-L2 is over rated compare to U3 , unless it’s XM-L2 U2
When looking for maxim brightness from a single cell the difference between XM-L U3 and XM-L2 (even U2) will not be that much, in most circumstances, no more than 10%.
Difference is mostly because of the XM-L U3 BIN. Which is not available for XM-L2.
When looking at all tints below 6000K the advantage of XM-L2 is higher compared to XM-L. More like 20%.
I would say tint can easily change the perception of light output too.
20% increase in brightness is not much to the eyes. 10% is not much at all… So it all depends on the person.
To OP, comparing tailcap readings is a mistake when looking for small differences in output. For all you know, the drivers can have different efficiency. Drivers can easily be between 60-95% efficient. The chance of having 10-20% difference in output when comparing random chinese light with similar tail cap reading is not what I would consider low. They change around on driver circuits all the time, even for similar lights. Also, several of the lights you mention dont excactly come from stores/manufactures that are considered trustworthy (in terms of bin/tint)
Since LCK-LED dont sell XM-L U2 emitters, its a T6 (at best). Which are equevilent to XM-L U3 I dont see how you cold expect to see a difference. The perceived difference you see is most likely a combination of tint and how many amps the emitter are getting (not the driver).
All in all. Your results are not surprising at all.
As for the copper MCPCBs, im pretty sure the ones from LCK-LED are with insulation layer…
Being a flashoholic I will certainly bother upgrading from XM-L to XM-L2. The difference is defiantly there. Is it big? no… (but that is subjective)
I would only care for XM-L2 U2 if looking for CW though, not T6 (or possibly lower bins) that most Chinese lights will use.
hmm… maybe if i upgrade all the LEDs on my Tr-J18 mod’ed with KD driver to XM-L2s, would that show much improvement? since they use 3 26650 batteries
+1