Hi Alberto- I don’t think to many vendor’s will take responsibility for it unless proof get’s shoved in there face.
The three xm-l2 u3 leds that I bought at mtnelectronic’s clearly visibly were worse then the xm-l2 u2 version’s I own installed in the same light as seen and judged by my friends and I but all I get from mtnelectronic.com is there should only be a 7% difference which he say’s is not visible……
To make matter’s worse 2 of the 3 xm-l2 u3 led’s I purchase all the sudden just died when my friends and I were testing them again tonight so IMO these led’s have issue’s but the problem remains’ is that proof is what the seller’s will ask for….
Sure I could paypal dispute and win but I’m not going to worry about it as I know mtnelectronic’s is a US company trying to make it out there so I will just take it as a lesson learned hope your issue work’s out well for you my friend.
Alberto- Here are some here mounted on copper—> XM-L2 I like the U2 1D that’s a great price going right now to they also ship realy fast. Hope this help’s.
I had a U3 XM-L2 emitter from intl-outdoor, but it was the 2A, not the 1A that most people complain about. I thought it was interesting enough to test it. And I'm glad I did because I had plans for a 5+amp direct drive mod for this led, and I would have been disappointed!!! First, the voltage was way higher than the other two XM-L2's that I tested, up to 0.5V at 6A! Second, between 6 and 6.5A the bond-wires fried. Third, the output hardly exceeds that of the T6 3C led. This could have been a freak-led of course, a bad apple, or is this just how bad other people found that the U3's perform?? BTW, I have killed many of the newer Cree leds lately (XP-L's and now an XM-L2) , but you can not blame Cree, up to 3A they perform fine of course.
Apart from the XM-L2 U3 2A, in the graph are a XM-L2 T6 3C (test posted before) and a XM-L2 T5 5A2 (test done a few weeks ago but never posted), both leds (and specs ...) were from Fasttech. (my cool white XM-L2 'crash' testfrom last year was not done with the same set-up but with the ceiling bounce method, so I do not want -can not- to compare them here)
All measurements were done with the exact same set-up described here, before each test the output calibration of the integrating sphere was checked with a constant output reference light. Remind that the lumens are 'djozz-lumens', they compare well to my other tests but may differ from other people's calibrations.
I got some bare U3’s in today, so right away I set about testing against the ones I got from Richard earlier, already mounted on 20mm Noctigons (the new ones)
I found that the pre-mounted U3 1A on the new Noctigon, (Ver 1 to keep it straight) pulls 3.87Vf and won’t go higher than 5.43A with my BLF17DD FET driver.
The new bare U3 1A mounted on an older style Noctigon (Ver 2) pulls 3.841A and won’t pull over 5.32A. This is using a freshly charged Efest 35A Purple cell at 4.22V right off the charger. Freshly charged each time .
The Ver 1 made 1794 lumens at start, 1718 at 30 seconds (from 5.43A) [New Noctigon pad, reflowed by Intl-Outdoor]
The Ver 2 made 1777 lumens at start, 1694 at 30 seconds (from 5.32A) [Old style Noctigon, bare led from Intl-Outdoor reflowed by me]
This is with each one put in the Eagle Eye X6 with a BLF17DD Ver 1.0 driver. Everything the same, but changing out the Noctigon/U3 1A.
I put the V1 back in the X6 and with a freshly charged Sony VT C5 it makes 1842 lumens at start, 1742 at 30 seconds. The wider spaced contact pads are very helpful in the X6, I had issue with the older style pad and had to keep the solder/wires to the very edge and use an insulation sheet to prevent shorts. Easy with the new board.
So, output is decent if not amazing while the current is lower than the U2 would do. Works for me!
FWIW here is my testing. Identical setups including all wiring and DMMs. Testing was carried out ten minutes apart of my bench power supply directly hooked up to the led. The XML-2 U3 1A was from IOS on a 26mm copper star. The XML-2 U2 1A was from CNQG and on a 20mm copper star. An Intel cpu cooler was used on both leds with the fan running. Conditions were as identical as possible to minimise errors. Maximum current tested to was 5 amps as some of you here will know why. All measurements were at the wiring into the led.
XML-2 U2 1A XML-2 U3 1A
Voltage Amps Voltage
3.11 1 3.1
3.40 2 3.38
3.64 3 3.69
3.86 4 3.87
4.11 5 4.08
If I can jury rig something up to measure light output I'll post it up at a later date.
I just bought a UltraFire C8 with XM-L2 U3 from AliExpress. Looks like it could be worse than a U2 if even real U3 at all? Well… guess i won’t notice anyway since i’m stepping up from a xm-l(1) t6.
Nice… I should get something similar then- mounted on a 20mm noctigon from RMM and it’s going on a copper spacer. Just waiting for the coolabs tim to get here and finish it up. I got the same battery too.
The voltage Mr has up looks like it’s a bit more efficient too. Hope mine is good it’s the 1A
Did you measure the voltage direct on the led wires, or more 'upstream' towards the power supply? These voltages are as high as what I measured above, also the U2, and way higher than the two other XM-L2's in my graph.
The leds had about 100mm of 18 Awg wire soldered to them with banana plugs on them. this is where the voltage was measured. I normally test with a switch in line but found that it added to much resistance for accurate measurements. Is it possible for you to measure an XML-2 U2 and add it in if you have one?
18AWG wire sounds solid enough even with 2x100mm of that, a bananaplug can have quite some resistance (voltage drop) but I would expect the voltage drop still to be way under 0.1V.
I'm sorry, I do not have U2 XM-L2's. I'm a neutral to warm high CRI adept so I never bought them.
They refund me $3 for the one working U3 I have that I kept as I gave the flashlight with that led in it to one of my friend’s who did not have a good flashlight and also they sent shipping label to me for the two led’s that are dead as Richard want’s to test them……to me this is amazing and is just another reason I will remain a customer of there’s.
I want to point out I never meant to say anything negative about this company in particular as I never felt this issue was any retailer’s fault.
I just think there are some more then usual random bad emitter’s from the XM-L2 U3 series floating around out there right now that hopefully get’s sorted out.
Thankyou to all of you putting in the extra time to run the test’s.
, but you can not blame Cree, up to 3A they perform fine of course.
