The new XP-G: another big announcement from CREE

Candle Power Flashlight Forum

Help Support CPF:

Press releases are a marketing tool to "stir up" the waters, make people talking. For this reason performances of the product are a lot or slightly exaggerated.

As someone else said above: just wait for mass production and tests.
 
An easy way to see if they did anything about current droop would be to test an XP-G, then plot the relative output versus current, normalized to 750 mA. Compare this to a graph of the XP-E or XR-E with the relative output normalized to 350 mA. This procedure factors out the current density differences. I plan to do this whenever I can get an XP-G.

Regardless of how they did it, this is a very exciting development! When I originally read the LED technology roadmaps several years ago, 150 lm/W in production wasn't forecast until 2012. We're practically there in 2009, assuming this isn't vaporware (I doubt that because Cree never makes announcements unless the product is well into the pipeline). In fact, my guess is they are already making these in production quantities, and holding out for the official release until they build up a reasonable stock.

Interestingly, I'm not expecting any dramatic increases in efficiency at very low currents. I've noticed a trend where as LEDs get more efficient the relative efficiency increase at 20 mA compared to 350 mA has gone down. For example, with a P4 bin XR-E the ratio was 1.507. For the R2 bin it was down to 1.422. For the new XP-G I'll hazard a guess of 1.3, which translates to about 172 lm/W. BTW, highest lab efficiency I've ever heard of so far was for a superflux type LED by Nichia. It achieved 169 lm/W at 20 mA and IIRC 191 lm/W at 3 or 4 mA. See this thread.
 
Last edited:
spencer said:
I think I remember someone saying it had a 2^2mm die. Thats 4 times larger than the current die.
Click to expand...

The xp-g has a 2 square mm die (1.41mmx1.41mm if it is truly square). It is double a normal 1x1mm die, it is half the 2x2 mc-e.

How small is the xr-c, xp-c?
 
I am dissapointed at the cree announcement - if you double the chip area, then 139 lumens / 2 = 69.5 lumens per mm2 at 350mA. Compare this to 114 lumens per mm2 at 350mA for an R2. Thus, the surace brightness per unit area at the same overall current is way down, and as such the spot will be much dimmer using the same reflectors (unless of course they can make the apparent size of the die smaller using a cunning built in dome - please correct me if I am wrong - which I would like to be as i look forward to these 'advancments'). What does everyone else think???

HL
 
Aaaaaaar! Reading back through the posts it looks like my point above has been mooted before. Sorry - I am currently in Korea on business from the UK and have missed a few days posting!!!!!

HL
 
Just drive it with a higher current. You'd lose any efficiency advantage, but your spot would be as bright as the XP-E.
 
Your lights will be antiques quicker.:twothumbs

If the xp-g does 345 lumens at 1 amp.
What would a hypothetical xr-g do, 500 lumens at 1.5 amps?

and a theoretical mc-g, 1000 lumens at 700mA per die?
 
Probably adding red. Pitty they can't get the same effect by adding some red phosphor to the mix (typical Cree CRI being 75, and 80 for warm white).
 
but it would have lower efficiency than just adding LEDs that make the right color from the get-go
Click to expand...

That's kind of a loaded statement with a lot of complexities to argue on both sides.

If it's true, then we might as well just build monochrome LEDs and give up on multi-spectrum phosphor mixes because the former will always be more efficient.

If it's false, then we have to contend with electro-optical laws that dictate that higher CRI (more spectral gaps to fill) means lower efficiency given current technology.

Also, while not entirely familiar with the details, it's my understanding that an additional reason Cree uses multi color emitters in their high end fixtures is they are dynamically/automatically adjusted for precise color mixing throughout the life of the fixture. My own opinion on this is it's just another widget to break, but also feel it's because Cree isn't currently as yet focused on pushing high CRI because they want to funnel R&D in other directions. So, the multi LED approach is just a stop gap.
 
blasterman said:
That's kind of a loaded statement with a lot of complexities to argue on both sides.

If it's true, then we might as well just build monochrome LEDs and give up on multi-spectrum phosphor mixes because the former will always be more efficient.

If it's false, then we have to contend with electro-optical laws that dictate that higher CRI (more spectral gaps to fill) means lower efficiency given current technology.

Also, while not entirely familiar with the details, it's my understanding that an additional reason Cree uses multi color emitters in their high end fixtures is they are dynamically/automatically adjusted for precise color mixing throughout the life of the fixture. My own opinion on this is it's just another widget to break, but also feel it's because Cree isn't currently as yet focused on pushing high CRI because they want to funnel R&D in other directions. So, the multi LED approach is just a stop gap.
Click to expand...
Anything can be argued but that does not make it valid.😉
 
I'd like to contest that point 😀

...But it is true that Cree uses more than one emitter type in their fixtures for reasons other than elevating CRI.
 
They use yellow and red LEDs. I don't know if they use white though... There is a datasheet somewhere that states this. Just look up on the cree site the model of that "bulb" and you will find a pdf.
 
You must log in or register to reply here.

Latest posts

Back
Top