Chip Surface Intensity: Cree XM-L2, and Luminus

[matt304]

Assume in this discussion the two LEDs being compared have adequate heat sinks.

I believe it is well known around here that the XM-L2 reaches the highest phosphor surface candela (lm/mm²) over XP-G2/XP-E2 when speaking of a chip being fully driven at maximum spec or above spec amperage levels.

This generally defaults in an XM-L2 choice for high Kcd large throwers (aspheric for example) as it will cover a square of more area, but at higher lumen per area. I believe this is why the German-built light that reached somewhere around 2.2 million Kcd (200mm lens) at a competition they have, used an XM-L2 in his light.

So in your opinions, is there any other LED out there that theoretically could reach higher light density at the die level, over the Cree XM-L2?

I noticed that Luminus LEDs of course are more wattage hungry, but they have much lower die-to-solder point C/W thermal resistance. I wonder if they could be pushed harder for that reason if the amperage/heatsink was there to do it.

Comments?

 

[eVoX]

Have a look at this:

http://www.taschenlampen-forum.de/lux-messungen/23324-leuchtdichten.html

 

[DIWdiver]

I was going to say I don't think it's well accepted that the XM-L2 has the highest surface brightness, then eVoX showed us the data. Of course that data showing the 'G2 with much higher brightness also doesn't show the peak of the 'L2. The brightness is still on an upward slope at 7.5A while the 'G2 peaks around 5.13A. There's no telling what would happen if you pushed the 'L2 to around 10A.

The data is inconclusive, but the 'G2's smaller size suggests it should be able to be pushed to higher current density, which suggests higher brightness. I'm not aware anyone has proved this to be true, but I doubt very much that anyone has proved it false.

The Luminus devices, having a larger die, must have lower C/W ratings to achieve the same power density. You should be looking at C/W * W, or just temp rise at max power. Keep in mind that 'power' means thermal power, not electrical power. Thermal power is lower by the amount of power radiated as light. Thus a higher radiometric efficiency (not the same as luminous efficacy, but related) will result in lower thermal power.

The larger sizes of Luminus devices mean the heatsink effectiveness will be lower (though perhaps only slightly). That coupled with lower device efficacy to start with, means it will not be possible to achieve higher surface brightness than with the Crees unless substantially higher temperatures can be sustained, or there is less loss of intensity at higher temperatures. I don't think either of those is true, but on that I could be wrong.