It is possible it could regulate the entire range but it would be very close due to the low Vf pushing the heat dissipation right to the limit.
Although since we do not have a viable buck or boost driver it is the best option.
Thanx! In the JM07, the LED centering piece is the stock XM size, but the XPL2 is sitting nice in the middle. Should file down the top - might pick up a few lumens. It usually works, not always. Could also try filing down the backside to get it more into the reflector. The reflector is nice for wiring - no flat bottom, like an EE X6.
I do have qty 4 of the XPL2 V6 3B's on order from KD since Dec 24th. They show stock but my order still hasn't shipped, so suspect they really don't have them on hand. They are displaying the V6 3B CREE label now.
In output, figure the V5's are about equal to XM-L2 U2's, and the V6's are equal to U3's, while W2's are same as U4's, but of course you get a whole lot more amps out of the XPL2's and they won't fry, plus they seem to run cooler and lower overall watts because of the lower Vf.
I just ordered a few different XP-L2’s from arrow, gonna give them a go to see how they look and see which version looks the best.
I am also going to try a taller centering ring that might block/diffuse the tint shift around the edges, no idea if it will work but even if it looses a few lumens if it can reduce or eliminate the tint shift it would be worth it.
It is a good thing these latest gen LED’s can handle more power, otherwise FET drivers would not longer be an option.
This does leave us in an interesting place for future options though as the Vf will keep dropping as time goes on and sooner or later the LED will not be able to handle the current it can draw. The 219C already is past that point but luckily it survives the abuse.
It does mean that we should be able to maintain regulation much longer though.
We would have to be looking for efficient buck drivers too.
Otherwise you just burn off up to 25% of the voltage.
On 2 Amperes this would be around 2 Watts turned into useless heat.
Maybe LiFePo will be interesting in the future, or even 3x NiMH.
(Would be nice to have 18200 NiMH cells)
I had someone request that I build them some XP-L2 Convoy S2’s recently and thus ordered some more of the V5 E3 LED’s from Arrow.
These seem a lot nicer then the first one I had, maybe the stress test caused a tint shift? This one while still cooler then the claimed 5000k IMO was not nearly as bad, just a hint of blue in the beam.
I also tried printing off some super tall centering rings that went to the top of the dome and the tint shift was not improved, although it did make the beam even colder. It makes sense as the tint shift comes at the edge of the spill which is the light that doesn’t hit the reflector. This can be seen visually when I was playing with the centering ring. I honestly thing the tint shift comes from the yellow tinted dome they are using now.
It is hard to explain but the angles I was seeing the tint shift were where the die itself could not be directly seen and only the dome. Has anyone sliced the dome on one of these to see what difference in makes?
I also noticed that a deeper reflector makes a BIG difference int he tint shift. The S2+ reflector has a very noticeable “ring” of tint shift but the S2 reflector is just a hint of tint shift around the edge for the most part. The hotspot itself also looks better and closer to the claimed 5000k as well.
My guess is the rated tint is the average of all light produced, so the larger the reflector and the more light mixed into the beam, the better it looks.
Overall the S2 OP reflector produces an acceptable beam with minimal tint shift although still a lot more then older Cree’s. I am curious what a larger reflector would do, I might toss one in a C8 I have laying around.
I also ordered a 4000k XP-L2 80 CRI, I just have not decided what to put it in yet.
Oh, the S2’s both made a bit over 2000 lumens with 20AWG, NXP FET’s and a 30Q. They also get hot.
In other news, thanks to the donation of CRX I have a few more tests in the works, including an XP-L V5 for comparison to the XP-L2 V5.
Amazing work!! Watching this thread closely. I am finding the tint shift aspects of the XP-G3 (5000k) work very well outdoors. The colder hot spot in the middle seems to aid throw distance. The warm-yellow side flood beam helps with near field. So I can light the trail down range and see whats out there, yet still pleasantly illuminate the ground I walk on.
If the XP-L2 V5 is like a beefed up XP-G3… I could definitely make that work for my intended use. Even if what I am seeing is an optical illusion resulting from color tint perception that’s alright by me!!!
Yeah, the tint shifts is very close to the XP-G3, I don’t mind it much myself. Although I do prefer the 219C over the G3 due to the better CRI and overall tint.
The XP-L2 could make a good output monster though, I am out of drivers at the moment so waiting till I build some more to build a C8 with one.
Hhm, waiting for XP-L2 triples for a proper Convoy S2(+) build!! :smiling_imp: …
After trying them as a single in a S2+ I can say that you will have only mere seconds to use such a light. Even a single XP-L2 gets silly hot.
Although I do plan to try a triple in an X6, still won’t have amazing runtime but should at least last a minute or 2.
All of that said, you should be able to crack 4000 lumens in a triple if you tried I figure.
You can run it without overheating until the battery depleted……underwater…. :cowboy_hat_face:
I actually did exactly that with testing the waterproofing on my S2+ EDC with TIR lens. I managed to noticeably raise the temperature of 1 gallon of water while doing it. It also lasted a surprising ~1.5 hours before shutting down (output dropping the whole time naturally).
Try high speed diving in an ice cold lake :laughing:
That should reduce the output drops considerably
The angle-specific tint changes are one of the traits I like least about Cree emitters. It pretty much goes away if you use a frosted lens or similar, but I’d rather not have to do that to get a reasonably consistent color throughout the beam.
Nichia 219 emitters still change tint a little per angle, but not nearly as much. And Cree’s HI emitters are quite a bit better about it too. But with a Cree dome, it’s pretty noticeable.
Broke 2,000 lumens with a XPL2 V6 2B from KD, in a Warsun X60 FET DD: lumens: 2,090 @start, 2,000 @ 30 secs.
Throw was measured at 5m for 66 kcd. Definitely lose some throw with the XPL2's. Previously I had a XML2 T6 4D in this light and got 1,610 lumens @30 secs, and 67 kcd. Tail amps measured with a clamp meter was 7.18A before the mod, and after 8.45A. The old T6's really could crank amps.
I just built an MT-12 with an XP-L2 V3 4000k 80CRI.
While not a nichia quality tint by any means it is FAR improved from the 5000k in an S2+ reflector I was using before.
First off the tint shift around the edge of the beam is completely gone, I think my theory of the deeper reflector fixing that was correct. The hotspot is very nice, good tint and above average CRI.
The spill does taper in tint getting cooler as it goes out. This is only really noticeable on a white wall, outside I didn’t notice it but the sun was already coming up so I will see better tonight.
Still making ~1900 lumens even with the lower bin though which is nice. I have to say, I really like this MT-12 light. The threads need serious work and the shelf should be way thicker but overall it is a nice light and the finished product is very nice. Kinda like what I am guessing the X7 is like, only cheaper and way better looking.
I am really liking these XP-L2’s the more I use them.
I have some more 4000k V3 80CRI XP-L2’s I will be building a triple with but they are a different part number then this one, they are binned differently then the easywhite system. I am curious both how the triple goes and how the tint looks on these other LED’s.
Is an MT-12 UltraFire branded and looks like an F13, this MT-12 here at FT? I assume you mean the Eagle Eye X7, not the TrustFire X7, or definitely not the Olight X7. These model names is get'n very confusing....
The Warsun X60 is basically a C8 deep reflector, SMO. I've found the V6 2B in this light to have 3 clearly different tints: the spill, corona around the hot spot, and the hot spot itself. I think some/many may find it not so pleasing. The V5 4000K XP-L2 in my ThorFire JM07 with OP reflector looks much better.
That's quite good to get 1,900 out of a V3.
Yep, that is the one I got, they had a sale on it for 15% off awhile back so I got 2 of them. Although I really want to get my hands on the F13, it looks like an updated version of the MT-12. It might have fixed the threads (way too small of threads and too many and not even remotely square) and possibly the shelf is thicker.
Yes, the Eagle Eye X7. I thought about getting one but it is just too dang ugly and I had to pass in hopes that someone releases a similar model that looks better. Like that X6/X7 combo render someone made.
Yeah, the tint could be separated into 3 areas like you mentioned although the deeper C8 style reflector could be making it more distinct then it is in this light as the reflector is not all that deep. I plan to build a C8 with a 5000k as well to compare so that will be an interesting datapoint.
Overall the LED is good if you want output IMO but it is not a tint queen by any means. But then I have 2 basic modes, max output or high CRI tint. For the first the XP-L2 is proving to be a good option since it is 3V and easy to use in most lights. For the latter, one word. Nichia.
I just built my other MT-12 with a 219C 4000k 90+ CRI.
So it is 4000k vs 4000k and 80cri vs 90CRI.
The interesting part is how close they are, don’t get me wrong they are still quite different but way closer then I expected them to be.
The hotspot tints are quite close, the 219C is a bit rosier but that could easily just be from the 90CRI. Same for how it looks, about what you would expect for 80 CRI vs 90CRI. The XP-L2 just seems to have a bit less red/more blue overall.
The corona on both has a different tint then the hot spot. It is quite close to each other in size and shape of this corona. The nichia doesn’t shift the tint as much as the XP-L2 and remains more even overall. The XP-L2 seems to actually get warmer in the corona with a bit of yellow.
The real noticeable difference starts at the spill.
The nichia gets cooler as it moves into the spill but it is much more mild then the XP-L2 and it simply changes color temperature, not the tint.
The Cree on the other hand gets significantly cooler in the spill. The offset from the spill to the hotspot is what causes the beam to not look good on a whitewall.
I have not tried the nichia outside yet but I tried the XP-L2 last night and outside it looked good, you don’t notice the tint shift much outside. I look forward to the 219C. I really like these hosts.
Oh and some quick data on the 219C as a comparison. It is making about 850lumens @ 3A vs 1250 for the XP-L2 and in “turbo” (have to limit the duty to about 75% to keep it from overloading the 219C) it makes around 1100.
The hotspot is about 65-75% of the XP-L2 size.
Some quick throw measurements.
219C Turbo - ~1100 lumens - 27Kcd - 330m
219C @ 3.15A - ~850 lumens - 22Kcd - 300m
XP-L2 Turbo - ~1900 lumens - 30Kcd - 350m
XP-L2 @ 3.15A - ~1250 lumens - 18Kcd - 270m
Seems like the 319A Ra8000 will probably hit in the middle between 219C and XPL2, closer to XPL2 I guess.
I kind of doubt that the 319A will reach XP-L2 levels but it should improve on the 219C.
The weak link in the nichia setup seems to be the thermal pad setup, it is much smaller then the Cree version and as such it seems to top out at around ~20W vs the 40W that Cree seems to reach before reaching max lumens.