Emitter Test Results pt4: XML2, XP-E2, XP-G2, and Nichia219! (along with sinkpad tests)

Bob - you should clarify this - where's your question coming from?

+1 with Dale, but I'm not gonna get upset bout it - Match's test results represented in this thread and all the others is the "de-facto standard" of our hobby as tech ref data is concerned and nothing has, or probably will, change that.

What do you mean?
Big Success? Budgetlightforum Standard?
Do you have any idea it take how much effort to provide this data?
Will you provide some data of testing result using different method?

Thanks, Didn’t mean to stir it up. I was just reading some conflicting results and did not mean to offend anyone

No prob - thinking was just a mis-understanding. Match's results may be questioned at times, but it's pretty darn good for what he or anyone had/has to work with. I myself see some things I don't understand, but I'll spectulate it's probably my testing method or result being off, more so than Match's.

One detail lacking in his original posts has been the exact bin/tint of the LED's tested, but since then. that's been clarified in later posts. I believe the XM-L2 was a U2 (Post #40), so for me, I'd be interested to see it tested the same way for a 1C vs. a 1A, for example, to know if they have identical output or not -- always has been a question in my mind... For Match, he's posted he's got to give this testing effort up - too much time and expense, and I can't blame him for that, for sure... The U2 bin has a range of output level spec'd by CREE, so I always wondered if a specific tint would be in the high range or low range, but that's getting pretty nit-picky.

Hi Match,

I was just wandering if you would be interested in running one of your tests on the new generation SinkPad II so we can get some independent data on these? I know a guy who could send you some :)

Just let me know...

Match, thank you for this awesome data :slight_smile:

Every aluminum star I have tested has been terribly unflat. So I have been sanding them flat and polishing to 1000grit US based on Crees recommendations in the CXA product design guide. Do you think this test would be closer with an aluminum star that has been flattened? (I have noticed that copper stars are always much closer to flat)

Hi supraspl, match has not been seen for some time on the forum, so if you don't mind I try to answer your question.

I think that match, as he was always busy improving the thermal paths in his flashlight mods, would also have taken care of good mounting of the aluminium boards, he probably would have noticed if they were not flat and been sanding them.

Regardless that, my and other people's experience is that you might see a small difference in output if the board is not mounted perfectly (because it is not flat, or there's a hole under it and board only rests on the side), as long as there is good metal-metal contact at least partly under the board, the big difference in output at high currents is almost exclusively caused by if or not there is a 'dielectric layer' (non-conductive, non-metal layer) present between the center thermal pad of the led and the metal core of the ledboard.

hi, i am new here. i am not new to the topic. but also am no expert. I saw this post and was hoping someone could help me out. I have a nitecore srt6 that was modded by vinh from cpf, he did a fantastic job. the mod was dedoming a xm=l2 u2 and putting a boost on the light taking the max output and making it 4 amps instead of 3. according to him initially it should have boosted my lumens by 200 or 300 otf and a 10% loss from dedoming. with all of that said as far as i can figure that should have given me approximately 1100 otf lumens or a little better on max. but i sent it to be tested and found out that it was only putting out 830 lumens. so someone in cpf said that it was because the 18650’ s even the samsung or panasonic high drain (10 amp +) batteries can’t maintain a high enough voltage for the xm-l2 u2 at 4 amps. they suggested getting lg 4.3v batteries which i did and that seemed to help some. they said it was because of the higher vf of the xm-l2 u2 at 4 amps. which was what brought me here. is there better 18650’s to use? or are there better 18350’s than efest imr or aw imr’s? or if i had someone mod it to 3.5 amp boost instead would that allow for a greater otf lumen? also is there anyone here that is suggested as a good modder of this type? and would it be possible to mod the srt6 to flat regulation at max with a buck boost circuit then have it step down to a specified amp or output? and i probably would not just buy a different light. however that is not out of the question. i just like the srt6 and would prefer to mod it. sorry for so may questions and thank you for whoever responds.

my apologies i just realized from reading that the author is a modder. but i am interested in whoever else might be.

The main problem asking questions in this thread is that match hasn't been posting for quite a while now.

But concerning the lack of output of your modded light: every flashlight running at high current on a single li-ion will face the same problem: not enough voltage overhead over the led-voltage. The XM-L2 at 4A needs 3.5V. The Efest IMR battery at 4A has a voltage of 3.6V. Consider a few tenth's of a volt loss in the flashlight (even while vinh usually tries to minimise it) and your flashlight can't maintain the 4A.

But, this said, you should still see more than 830 lumen OTF, more like 1000 or so. How did you have it measured? I ask this because getting an accurate light flux measurement is not obvious/cheap.

thank you for responding, i appreciate it.
i sent my light to jmpaul320 who as far as i can tell has a pretty good set up based on his results of other lights and other peoples post about it. if you know more please let me know. you can pm me if you want.

If I ran an XM-L2 or XP-L off of 4S lithium AA’s, would the emitter and driver survive? (Qlite)

First, AA lithium primaries do not like currents above 1A very much, the capacity suffers and at 3A the voltage is not really stable (I look at HKJ's review of the Energizer Ultimate Lithium AA: https://budgetlightforum.com/t/-/20614 ).

That said, at 3A the voltage is between 1.2 and 1.3V, so four in series would be max. 5.2V theoretically, but there's losses from springs (in a 4x-carrier there's at least four of them), switch etc. The Qlite MCU can have 5.5V so you will be fine in that aspect.

This would be for an emergency light in my vehicle. It would be 4S and 2or3P, so the current would be split up, and I’d probably only do 2amps to the emitter as runtime is key. So the XM-L2 can handle it without issue?

I got somе XPG2 S3 2A samples this friday, and today i decided to test them with A17DD V024 wights driver, expecting 4amps or so, but to my surprice they are driven with a fresh battery( 25r) at 5.5amps…

5.5amps…will the emitter last( in Convoy C8 lets say) at that current? Cause i like the result alot

PMed djozz ofc :bigsmile:

Mitko is this a tail cap reading or at the emitter?

Tail, sorry for not mentioning it at first place

The current at which the XP-G2 blows thusfar seems to be the same current at which the XM-L2 blows: the old batches at about 9A, newer batches around 7 (from what I hear, I have not tested a newer XP-G2). That is not surprising because they seem to have the same size bondwires. The difference is that the XM-L2 increases output up to when the bond wires fry, but the XP-G2 is over the maximum output and the output already is less than max and turns blue when the wires go.

The XP-G2 is about at its max at 5.5A, which is nice because during battery drainage the current will drop but the output will be quite steady for a while :-)

10x, i knew youll help - thats C8 setup is actualy a very good choice for light FL setup when you hunt- the strange part is that i`ve bulded several XPG2 S2-3 C8s, but with that driver i usualy get 4….4.5amps max, it seems those samples have lower Vf