Why should I buy XP-L? Need some basic pointers.

I'm thinking this is why on djozz's chart (https://budgetlightforum.com/t/-/27500), the XP-L peaks at bout 7.5A while the XM-L2 is still climbing, though slowly. Also maybe this explains why the Vf is higher on the XP-L?

Actually leaving aside lumen density of the XP-L, the LED is claimed to have a certain brightness in the Product Characterization Tool, the same or better than XM-L2 (Vf aside). CREE does not have a track-line of lying regarding LED brightness. Why would they start now? Just one test from one seller of one LED is enough?

I did a XM-L U2 mod to a Tank007 E09 once. >> XM-L U2 Tank007 E09 mod !
the XP-L should be easier with the smaller substrate footprint.

XPG2 or XPL both dedomed. Which has more throw on the same reflector/driver/etc? I am guessing XPL because it has the same footprint, but is able to push out more lumens. Is this true?

The XP-G2 de-domed has a more concentrated point of focus, hence further throw given the power level. The only way to really achieve throw with the bigger die face is through use of a massive reflector (barring aspherics) such as the 96mm TK61.

Granted, the biggest throw is now coming from the honking big lights with XM-L2s de-domed, so it’s probably arguable.

The only 'breakthrough' with XP-L is that it is by far the brightest led with a xp-footprint. That is more significant than most flashoholics think, because this so-called 3535-footprint has more and more become an industry standard (LuxeonQ, Nichia219, numerous chinese leds), while the xm-footprint is not.

Djozz do you belive 1 XP-L is enough to be tested to judge this LED, lumens and Vf? (maybe you tested more than 1 and was not specified?) As I said earlier CREE has had a good track-line of specs so far. Now they changed?

No, I tested only 1, and this means that you have to be beware that there is variation among equally spec'ed leds in Vf and lumen output, I might have measured an extreme one. On the other hand, I would not expect another specimen to behave much different in curve-shape and the current at which the max output is reached, because that I think is design related.

I did promise somewhere else (, man, they gave me a hard time over there, demanding hard scientific proof of every detail, lots of (ir)relevant theories, but in the meantime not really interested in practical information) to repeat the test with my second XP-L, it is the same bin/colour (in fact they were next to each other in the reel). It would give some combined insight in the variation among equal leds, and of course the repeatability of my measurements. I do not promise anything, but will try to do that test one of these days.

That seems to be it in a nutshell.

djozz, do you think the XP-L will be a good performer in the 3A range? I’m looking at reduced heat over a 3A driven XP-G2, what do you think? Just ordered a couple of reflectors for my little custom Ti light (my avatar) and will have to modify the reflector to fit the XM sized die, but I’m getting 552 OTF lumens at 3.1A to a de-domed XP-G2 now, from an Efest IMR10440 cell. So I think the lumens will go up significantly, the tint should be better in the de-domed XP-L V5 2A and quite possibly the heat will be less.

Please share your thoughts on this.


Each new chip seems to be about a 5% increase in lumens, so not a lot, but after a few generations it will add up a lot more lm/w

Excellent idea to look at that tool! The flux-bins in the datasheets are not so easy to compare due to different currents. I guess you should compare XM-L2 U3 (currently highest bin and available) and XP-L V5 (2nd highest bin, already announcing the next higher earlier than for XM-L2 for marketing reasons),

and voila: almost identical!

2nd observation: an interesting official statement here : "As a “successor” product to the XLamp XM-L2, lighting manufacturers seeking ENERGY STAR® qualification can use just 3,000 hours of LM-80 data, potentially saving up to four months in the approval process." Do I get it right that existent XM-L2 evaluation data can be used, which means they need to be sufficiently similar?

3rd observation: the center bar in the solder profile of the XP-L is 1.3mm wide (matching the XP-G*). But the XM-L* die is 2mm wide. This might be an interesting issue (possible disadvantage?) about heat management when overdriving the XP-L to the limit.

I think you may be missing the point Slim Pickens was trying to make. Cree is saying that this package represents an advantage based on physical size alone - not due to die / binning / efficiency / etc improvements. The 200L/W efficiency stuff is just for press releases and headlines, it has little to do with how they are actually marketing this product to engineers. Cree is terming this OCF, or “Optical Control Factor”. Take a look here for more info: www.cree.com/ocf

Fair enough, but if it is a smaller chip and competes or even gets a few more lumens then the XM-L2 then why not try to market both.

They do market both.

EDIT: I wanted to say this in a non-provocative way, but couldn’t think of a good way to state it. Basically any claims of improved performance from the XP-L over XM-L are extremely questionable. The die appears to be a standard “Gen2” style die, so it’s most likely that Cree is doing exactly what they did with the transition from the large XR-E package to the smaller XP-E package: putting higher binned dies into the new package and using lower bins for the old package. For XR-E it got to the point where there apparently weren’t enough low-binned dies so Cree started using a smaller die in order to keep form releasing XR-E LEDs with a higher brightness bin!

I played around with the characterization tool (which i now love for all the data available) and it seems the XP-L is a bit better then the XM-L2, but not revolutionary, but then again we had many threads a while back about why upgrade our lights to XM-L2 from XM-L the difference is not there, and i guess thats what i meant by 5% or so improvement in each generation, one or two generations doesn’t make much difference so no point upgrading all our lights each time a new chip comes out just for more lumens, wait a few chips then upgrade to see a noticeable change, but if there are other reasons to favour the new chips then they can be worth it, if one has an XP-G light upgrading it to XP-L2 and one can get much higher currents running through it, essentially getting XM-L2 output from the smaller chip in some lights that can’t fit XM-L(2)



any Carlco triples for the XPL yet ?

No, and I don't think any will be coming. The dome size is huge in relation to the tiny little 10mm lenses in the 20mm 3-up TIRs. If there were practical applications for it we would have already seen parts for a 20mm triple XML2 a long time ago, and the XP-L will have the same beam characteristics as a plain XML2 since they have exactly the same size die. The only difference between them is the size of the square base.