XP-L vs XM-L2 OTF Tests

I know a lot of you are interested in the XP-L as a replacement for XP-G in multiple emitter setups, but I tested the XP-L against an XM-L2 in a P60 setup to compare the OTF performance.

The flashlight setup:

XM-L2 U2-1D with P60 XML smooth reflector
XP-L V5-2A with P60 XPG smooth reflector

16mm noctigon
3.8A FM driver
p60 drop in -potted and copper wrapped
LG E1 4.35v battery
Solarforce L2P host

Two of each lights were built and all test results were averaged.
The overall output tests were powered by a lab power supply set for 4.2v at the driver and the lights were fan cooled.
The throw tests were performed using an LG E1 battery.


Surprisingly, it appears the XP-L is slightly better thermally…maybe due to it’s slightly better efficiency. (the XP-L had a output drop of 60 lumens vs the XM-L2 drop of 80 lumens)
As expected, the throw from both emitters was nearly identical -even though the XP-L had greater output.
I like the XP-L for the 2A color bin, but I’ll take the extra 60 lumens.

Thanks for the test!

Good to know. Thanks for sharing that info! :)

Would you mind repeating throw using the XM-L reflector on the XP-L pill? Many who do this will have XM-L reflectors since they are far more common and easier to get.

Great! Yes, please do the XP-L with the XM-L reflector, as many of us will not use the XP-G reflector, especially in larger reflector lights. Everything I have done with XP-L, tells me I like it better than XM-L for a thrower, even though there should be no difference, there is.

I copper wrap and glue my reflectors on, so it will be easier to build another to test the xm-l reflector -give me an hour or so.

Way cool tests. I always look forward to your finds. TY.

But I think there is something I am confused about.

You say that "I like the XP-L for the 2A color bin, but I’ll take the extra 60 lumens."

So you like the XPL. But you will take the extra lumens. But if I am reading your graph right then the XPL HAS the 60 extra lumens... And the least thermal drop @ 30s.

So you will take the XPL... right?

Sorry, I worded it wierdly:
The xpl has the tint I want and I also get 60 more lumens with it.

Clairification: ‘thermal lumen drop’ was the amount of light decline from 30 seconds to 5 minutes

Here’s data for the XP-L with an XML reflector:
|xpg reflector|248m|
|xml reflector|241m|

The hot spot from each is almost identical in size…the XPG reflector gives the hot spot a hard, well defined edge. With the XML reflector, the hot spot has a softer edge. Both hot spots are good, but the XPG reflector has a little more throw than the XML reflector.

Thanks for the tests, are XM-L2’s being outdated with the advent of the XP-L?

I did not know there was a reflector difference, what is different about them (other then the opening on the bottom)

My guess is: if they can fit an xml emitter on a 3.45x3.45mm xpg platform by clipping the dome edges, then they can put something larger on the xml’s 5x5mm platform and keep the xml footprint. I think the XP-L will kill the xpg since they share the same footprint…remember, the vast majority of the led market is not looking for throw like the xpg delivers…they’re looking for output. The differences between the xpl and the xml aren’t drastic.

I never thought there was a difference either. When I built the xpl, I just naturally use the xpg reflector because it fits the isolation pad I use for centering. I really didn’t expect a difference when I put on the xml reflector. Keep in mind, the difference isn’t huge…just noticeable.
When the xml smooth reflectors were backordered from all my suppliers about a year ago, I drilled out some xpg smooth reflectors an they looked good on the xml.
The real difference may be who they were bought from and what day of the week they were made.

The difference, is of course, in the binning. Like a U2 is brighter than a T6, the V series binning is the next step up from U. These new XP-L’s should be in the 18-20% brighter range, as we’ve seen before with a new bin.

It will be a sad day when they have a ‘Z’ bin…if they use up all the letters…they can’t get any brighter…

Why I’m impressed with the XPL V5…they managed to generate a 2A tint…not a 1A or 0D like you would expect.

When I had the light in my Avatar built, it came to me making 135 lumens. Now, with a de-domed XP-L and a driver of my own build (BLFTiny10) it’s capable of 921 lumens. :slight_smile: Radical difference, I’d say! The bigger XM-L2 die wouldn’t fit in this light without major alterations. The die size of the XP-L did indeed still require opening up the emitter hole in the reflector.

I also like these XP-L for a triple configuration. Big step up over the previous XP-G2.

I’m finding the tint shift upon de-dome to be much less and of a better quality overall. Seems that the dome is concave on the die side, not actually in contact with the die face or phosphor so there is less effect removing the dome from the adhesive strip around the outer edge. Win win.

okay now it makes sense, made me thought the graph legend was wrong

btw its smaller and still better cooled? weird

Milan, the die is the same size it’s just put on the substrate of an XP footprint.

I think Cree’s intention was to allow the higher powered XM class die to fit together tighter in arrays for street lighting and such, so using the existing compact XP size base and cramming the big XM-L2 die on it made a lot of sense. Notice that the dome has flat sides to allow these to nest together without gaps, this is ideal for array use where multiple dies fit into one light.

Sure we use em in our lights, but I doubt Cree gave us much thought when designing these.

This is interesting, different results from other tests done and posted here, that showed the output about the same (think myself included Smile). XM-L2 U2 1D's I've used tested a little lower than XM-L2 U2 1A's. Also the XP-L is not the highest bin available - the V6 is available from Illum.com.

Did you use copper DTP stars on both?

Also the interesting result of the XPG reflector throwing better than the XML reflector is also true for C8 reflectors.

Thanx for doing these tests!!

After I test more lights, I will be more comfortable with the accuracy of the outcome…this test was merely 2 of each…but I think it’s a good start.
I’m not sure the xpg reflector has a different geometry than the xml -the contact ring on the bottom of the reflector -the part that presses against the isolation pad, it holds the xml reflector .46mm higher -this could change the focus enough to affect the throw 8 meters. Also, my xpg reflectors are .5mm larger at the top…so there could be a difference. For the little difference in throw, I will use the reflector that gives the best beam pattern.
One thing to note: the OP xpg reflector looked terrible on the XPL.

I think your tests are better/more thorough than my own for sure... Using qty 2 of each is one more each than most of us use Smile. Maybe those V5 2A's are pretty good. Think I and djozz used V6 2C's. I truly believed the cooler tints in the same bin will perform better, but after djozz's light meter tests, I'm not so sure now. His test results indicate cheaper light meters (maybe Extech included) measure cooler tints higher than neutral and warm tints.

djozz's here: https://budgetlightforum.com/t/-/27500

Actually in looking over the results, they are actually pretty similar to yours. His lumens are higher but he's not using any reflector or glass, just bare emitter I think.