Teaser XML-3 Coming 5A rated

djozz · 2020-06-15T22:28:47.000Z

Cree has provided a design example (a torch ) for the XM-L3 that seems to show not too much tintshift over the beam, but a pic of a headlamp does not look good qua beam.
https://www.cree.com/led-components/media/documents/PortableLighting_XM-L3-TorchRerefenceDesign.pdf

hodor · 2020-06-15T22:56:22.000Z

8300K

JaredM · 2020-06-15T23:00:39.000Z

So if I’m interpreting this right, then what they showed in that mod/retrofit example is very misleading. Efficacy is barely improved and the xml3 used was COLD white. It seems to just show that if you run the emmiter at 5A, you get about 50% more output. Go figure…

BurningPlayd0h · 2020-06-15T23:03:45.000Z

vestureofblood:

Th558:

Hopefully it’s not a disappointment. What is the actual difference between XML and XPL. Is it just the footprint? or the vf aswell.

If I had to wager a guess I would say that because of the XML foot print this emitter will continue to show an increase in output at higher drive currents than the XPL. That yellow dome does make me a bit apprehensive about how the color quality will be. So far all of the yellow dome versions of crees LEDs have been a down grade in color blending IMO. Getting the super low VF like in XHP50.2, and XHP70.2 has sure been nice though.

To be honest I am surprised they made this at all. I was assuming XHP50.2 was their answer to 3V XML footprint. I will be trying some of these bad boys for sure

I would bet these will be significantly cheaper than the quad-die emitters. Basically a 5mm footprint equivalent to what they designed the XP-G3 for - absolute highest efficiency from a cheap emitter for the lucrative street/area lighting industry.

Sirstinky · 2020-06-15T23:04:12.000Z

djozz:

Cree has provided a design example (a torch ) for the XM-L3 that seems to show not too much tintshift over the beam, but a pic of a headlamp does not look good qua beam.
Document Library - Cree LED

A performance boost at 5A is nice. I’d like to see the voltage curves change over current input though. I wonder how they get 8300k though. The sample looks g0od, bit I’m wondering how they don’t have corona? This is definitely not a replacement for the xhp50.2 3 volt though, amd we’ll see how 9t compares to the sst40!

Th558 · 2020-06-16T00:59:57.000Z

I’m also wondering why at over 8000K the efficiency is almost the same. Maybe the production version will be better?

Sirstinky · 2020-06-16T01:48:32.000Z

No clue on bins either? How will it compare to the 7 year old XM-l2? I was thinking the xhp50.2 3v was going to replace the xm-l2, XM-l2, but obviously the cost difference is an issue for industrial applications. At $10 for a 50.2 vs $6 for a sst40 or $7 for an xp-l2 it makes sense for a new more efficient emitter.

Hikelite · 2020-06-16T06:24:17.000Z

djozz:

Cree has provided a design example (a torch ) for the XM-L3 that seems to show not too much tintshift over the beam, but a pic of a headlamp does not look good qua beam.
Document Library - Cree LED

It’s NiteCore TM06S which they make out to be way more nice than it really is with the XM-L2.

Actual beam of the flashlight with XM-L2.

What they show XM-L2 looks like in the TM06S, basically a crop of the hotspot only.

Considering that most of those type of reflectors always have all kind of bad artifacts, seems like a poor choice of testing their own LEDs on how they perfom, even if that was is not what they were trying to point out.

Agro · 2020-06-16T06:39:07.000Z

Marginally lower Vf than XM-L2.
At least now we know that it’s really not a flip chip.
Dome appears slightly smaller.
It’s slightly throwier, delivering 8.3 cd/lm rather than 8.0. The die might be smaller.
Huge CCT shift with current.

Agro · 2020-06-16T08:28:03.000Z

There’s a product page with a datasheet already:

Cree LED

XLamp® XM LEDs Archives - Cree LED

The first generation XLamp XM-L® LED was the industry’s first single-die LED to deliver over 1,000 lumens of light output. Newer generations of XM-L expand upon the performance advantages of the XM platform with more light output, better efficacy and...

Agro · 2020-06-16T09:58:55.000Z

Comparing datasheets:
table(table#posts).
|LED|Thermal resistance (K/W)|Max current (A)|Vf at 3A|CRI|CCT (K)|
|XM-L2|2.5|3|~3.17|unrated to 90|2600-8300|
|XM-L3|2.2|5|~3.14|unrated|5000-6500|
|SST-40|1.2|6 (8 pulse)|~3.01|70|4500-7500|

welight · 2020-06-18T22:39:21.000Z

chip looks good

its bright but no testing done yet, I got some U4 1A and 1B bins
Cheers

JaredM · 2020-06-18T22:41:50.000Z

interested in results!

ch1ir · 2020-06-18T22:50:38.000Z

Great news

Yokiamy · 2020-07-15T18:08:48.000Z

Kaidomain sells them as well

http://kaidomain.com/Cree-XM-L3-U4-1B-White-5000K-LED-Emitter

HONDARIDER · 2020-07-15T18:53:07.000Z

Looks like they only have the 6500K version at this time. I more fond of the 4000K tints personally.

BurningPlayd0h · 2020-07-16T01:27:11.000Z

Since there are both visible bond wires (so not flip-chip) and clearly phosphor over the entire pad, could this and the recent “XP-G2 HE” be using the same design and production process?

djozz · 2020-07-16T08:18:10.000Z

BurningPlayd0h:

Since there are both visible bond wires (so not flip-chip) and clearly phosphor over the entire pad, could this and the recent “XP-G2 HE” be using the same design and production process?

Could be, I do not have high expectations. But I will do a quick test when I get a couple in from KD.

Rolz · 2020-07-16T08:24:36.000Z

It will be really interesting to see how they perform. Great to see XM-L3 come out.

Rolz · 2020-07-16T08:43:42.000Z

Just saw a flashlight on KD which uses it,

http://kaidomain.com/KDLITKER-C21-Cree-XM-L3-1800-Lumens-5-Mode-LED-Flashlight?search=KDLITKER%20C21

The listing has errors but assuming it’s a L3 and the driver produces 5A, the 1800 Lumens is probably accurate