Love it! having no current spreaders will look nicer in aspheric.Cree is notorious for releasing higher bins later down the road,so I might wait for the better binned stuff to surface.Other companies are really floundering wirh their new stuff,especially SSC with their z series...fail.I'd like a few to play around with,even if they are lower bin first thing I would do is throw some in my outside security light.
i think it’s 20% more from about 20% efficiency —> 5% more energy into light
since every new generation of cree leds seems to have a lower Uf this could be an interesting idea.
I’m going to read up on buck converters, i never liked the idea of burning off excess voltage in linear regulators.
Been lazy about ordering a U3 on copper to pop in my HD2010, and it pays off!
Will be looking for the XML2 in high bin on copper now! Direct drive off a freshly charged king kong with a big jump in surface brightness, this should throw nicely.
Is this a “real-world’ 20% increase in light output with an optimized reflector and driver? Or will it be realistically more or less?
For example, with the correct driver, let’s say a Fancy Fandy hoo-fire ke-hd-2010-5-897 now puts out 875 lumens with a T6. With an optimized driver and reflector, the comparable light with an xm-l2 will now put out 1050 lumens?
Where do you see lower Vf?
Maybe I’m wrong, but when I calculate with the old xm-l’s “temperature coefficient of voltage” which is –2.1mV/°C , I get 3.224V Vf at 3A and 85°C.
3.35V + (–2.1mV/°C * 60°C)= 3.224V
this value for the new xm-l2 is 3.35V
And because of the lower “temperature coefficient of voltage” of the new xm-l2, which is –1.6mV/°C, higher core temp means automatically higher Vf.
Although I believe that lumen ratings are weighted according to the human eyes’ response to the wavelength/power spectrum of the source, so taking into account that these LEDs are white they should really be compared to the theoretical maximum for white light (around 250lm/w) instead of the theoretical maximum for monochromatic 555nm green (683lm/w), which gives a value of about 40% luminous efficacy, meaning nearly 10% less heat.
Anyway, I wanted to put a positive spin on it
To be honest I don’t think the increase in heat from a linear driver will be more than that amount anyway. It’s only a small drop in vf, so it should about balance out heat wise.
“Up to 20” is what Cree says. I believe I saw someone say the 20 takes into account the variability in bins where you could have one LED at the bottom of the bin and another in the same bin, but at the top of its range. That gives you 7, but that’s true of any two LED’s in the same bin. On CPF, Selfbuilt compared Foursevens XP-G lights against the newer XP-G2 lights where 20 is also thrown out and he said he wasn’t getting that much.
The Cree datasheet for the XM-L2 goes up to a U2 bin, but I don’t know if that will really be available. Because the LED’s are binned differently, what do you compare against? The two top bins: XM-L U3 vs. XM-L2 U2? Right now it seems like T6 and U2 are the most available XM-L bins. Once we know real-world bins for the XM-L2 and people start taking measurements, it will be easier to do comparisons.