Nice! That's great with a NW XP-L.
It’s nearly the same as a dedomed 219c at 6A, and almost exactly 3/4 the candela of a dedomed XP-G2 S4 2B at the same current. (all in 1504’s or 1405’s)
djozz, I would like to echo the others by thanking you for your efforts. Your data is awesome! When I still made flashlights I referred to it many, many times.
If I ever come out of (flashlight making) retirement, you are on my list to receive (and keep as yours) a sample light for an honest evaluation.
wow, thanks 
There are many sound reasons to stop this flashlight madness, but if you ever decide to dive in again I think many of us will be very pleased to see you back here, if only for your builds, they are beautiful!
Thank you.
I fear it is not a matter of if, but when I will dive in again. I still have a lot of things on my plate that take precedence, but the list is getting smaller.
I ordered a couple of XPL V6 0D's from FastTech(fasttech.com-xp-l-v6-0d, with the deal: https://budgetlightforum.com/t/-/35653). In theory, should be excellent for de-doming, specially with the test results posted here. Iv'e dedomed XM-L2 0D's and they came out pretty nice.
Let us know how they turn out. The XP-G2 0D is lovely dedomed.
Ok - of course FT is not, what I'd call, a trusted source. You, Hank, Cutter are trustable - is that a word? 
Guess I'll see when they arrive - not in stock, they say to be shipped Monday, Nov 2nd.
Recently, I bought a few LEDs from FT (XP-G2 R4/XP-L V2), LED-DNA (ebay, XM-L2 U3) and lck_led_com (ebay, XP-G2 S2). I tested them on my homemade IS. It is uncalibrated so I can only compare the relative results. I would estimate the lux number should be divided by 25 or 30 to get the Lumen number. The results on XP-G2s are all pretty consistent. However, the results on the ones with higher power rating are not. At 3A using a KD C8 with 8x7138 as the power source, reading on the XM-L2 U3 is about 24500 lux. So it is about 800 to 1000 Lumen and pretty good, I think. However, the 16 mm XP-L V5 from FT only gives me about 21000 lux under the same conditions. The reading from 20mm XP-L V5 is only 18000 lux. I also tested the XP-L Hi V2 come with the KD-C8. the reading is about 22000 lux. For the tests, I had the LED pulled out as to eliminate flashlight loss and I only turn the lights on for a few seconds to avoid burning the LEDs. With the reflectors on, the readings are about 10% less.
My questions are:
- I thought XP-L should provide higher output then XM-L2. However, it is not the case from my tests. Is it normal?
- The 20 mm XP-L V5 has such a low output. I am wondering if it is not really V5 as it is claimed to be.
- Comparing XP-L Hi V2 and XP-L V6, which one should have higher output?
Hope some more experienced member can make sense out of my results and let me know. Thanks.
I am waiting for a XP-L V6 from lck_led_com. I will know how it performs after I receive it and tested it out.
May be FT is my problem.
For C8, the best results I got is XM-L2 U3. XP-L V5 is not as bright. I don’t know if it consist with your results. All of them are 1As. (Is your U2 XM-L2 U2 or XP-L U2?)
Well, can't answer it all, but the 1. statement above just cannot be generalized that way. An XM-L2 U3 should be better than a XPL V5 based on djozz's results in the OP (an XP-L V6 beats an XM-L2 T6, but not by much). You are testing a XM-L2 2 bins higher, and a XP-L 1 bin lower - the XM-L2 should beat the XP-L. Looks like both your XP-L V5's from FT are not as listed. FT has done this before. some of their listings are obviously wrong, some not so obvious and no telling til you get them and compare against something from a more reliable source.
It's sounding like my FT XP-L V6 0D's will be over-rated in bins 
. They are just not a reliable source for LED's, never were. You always take your chances.
Thanks Tom. That was I thought. Like other people say, I am a person always do things by rules. Maybe not everybody is like that.
As for FT, my feeling is that they are not really very technically savvy as they should be. They just rely on whatever others told them.
I just made up my xp-l v6 3D. from intl-outdoor
In short, with my jbweld+zinc powder, aluminum reflector, ar glass, 2 al heatsinks (one finned cpu sink) glued together: I am measuring 374 lumens in a relative fully diffused lux comparison to my 277 hp11 fenix light—at 700 ma to led 640 ma draw on fresh cells (320ma per 2x18650, with buckpuck). This is dead on the pct.cree.com tool.
All my xml2 t6 builds came in around 280 lumens in this test. My xml2 U2 builds would come in around 305 lumens.
My complaint is that the smallish pcb board limited my choice in optics, else I was bumping my solder job on the star with wider based optics. I may need to learn to reflow solder to get a wider 26 mm star. I see how they do it, but don’t understand how it works.
Still, 374 warmish/neutral lumens with 10 hour runtime minimum, is nothing I can complain about. Also the next batch of xpl’s should be %9 brighter, meaning 400lumens/200 lpw at 700ma, with this homemade setup of mine (that seems to run way cooler than anything previously, I have seen).
Just did a xp-l hi V2 build, compared it to the Fenix hp11. In relative diffused lux ratio test, the V2 came in consistently at 309.6 lumens at 700ma to led. About 1.5 lumens off the pct.cree.com tool. Also, the candela is roughly 12.5k in a 35x26mm smo reflector, at 700 ma.
Visually, it is not only throwier than the xml2 t6 in the 42mm smo, but brighter in a corner of a hall. Though, only 30 lumens more: 12.5k candela v. 8k, plus brighter coronua.
So, I can’t complain with two 4200 Kelvin lights at 374 and 309 lumens, running at only 700 ma, constant brightness, for a full workday. Yes, only a slight step up in usefulness from the xml2, but a step in the right direction toward the goal of 8k candela over 15 degree hotspot, which will take way more than 400 lumens. But, I am not budging from the limit of 700 ma, until liion tech gets much better.
Anyone else noticed that Cree changed their datasheets on many LEDs for the reflow characteristics
including XPL, XML, XPG2, XPG3 and XHP35
before they were 3K/s now they are 1.2K/s like XHP50/XHP70 were in the past
No, I did not notice that. Would that matter if my reflows are more like 30K/s? :innocent:
what is K/s? I search in google but nothing…
oC/second ,or how fast temperature rises.
the more thermal stress a reflow does the higher is that you reduce lifetime or even destroy the diode
XHP70.2 with 30K/s would destroy it