I'm interested in this one. I really like my KD C8, but at the same time I'm noticing that at least the reflector in the Brinyte version starts further up the parabola. I wonder if the reflector is also more shallow or what. And most importantly... several beam shot comparisons is what I wanna see. :)
What do you guys think about the difference in reflectors?
Sorry to hijack your thread. But there is alot of buzz around XM-L. AFAIK, they generate half the heat and twice the output of XPG? Also better than MC-E?
Definitely. More output for the same current. It is a single die so no cruciform hole in the beam which is there in all but the best-sorted MC-E lights. It is a smaller die than the four that make up the MC-E so higher surface brightness and more throw. My best MC-E does a little under 600 lumens, my worst XM-L (of two) does pretty much the same. My better one (of the two P60 XM-L dropins I own) does a lot better.
Half the heat, but at the full power of XP-G. It heats no different than an XP-G if driven at full power, 3A.
For the OP,
I built my own, smooth reflector C8 XML and saw that it isn't a thrower anymore. Then I took a TR1200 body, put in a 52mm reflector which is much shallower than a C8 reflector but wider. And that XML threw well. So, deep reflectors designed as XR-E throwers don't work the same for a wide emitting LED. Instead, you'll need a wide reflector to catch and throw the sidebeams. Yes, C2 type wide reflector may work better. I ordered a C2 DIY kit to see that.
An XPG @1.5A produces about as much (a bit more) heat as an XML @1.5A,
an XML @3A (a bit more than) twice as much.
An XML T6 @1.5A gives about 17% more light than an XPG R5 @1.5A,
two XPGs @1.5A each give slightly more light than one XML T6 @3A though.
All calculated from the data sheets, and at 25 degC, so no thermal sag is taken into account (of which the XML will probably suffer more, as it produces more heat @3A).
An XP-G at 1,5A mean 5,3W that from 3,5V Vf .1,5 A, in an hypothetical body with a total thermal resistance Junction to ambient of 15 C/W the temp of the junction reach around 105 C.
5,3W.15C/W + Ta 25 C =105 C = 80% flux 139 x 325% = 451 x 80%=360 Lumen
In the same body an XM-L perform way better since Vf 3,1V x 1,5A =4,65W.11,5C/W =53C + Ta 25C = 78 C 280x200%=560 x 90%=504 Lumen
The driver heat is no counted however with full cells is around or a bit more than 1W with XML and a bit less with XPG
Ah, I didn't look at the thermal resistance, 6K/W (XP-G) vs. 2.5K/W (XM-L).
That would indeed be 30% more light for XM-L. But much less throw though.
It's 40% more light, the XM-L at 4,6W is in the most efficient part of the curve the X-P-G it´s not, to put it simple, 2 XM-L at 4,65 W total 9 W give roughly the same light as 3 XP-G at 5,3W total 16W , that's a big chunk.
All with a lot less heat generated...
I used different values:
XM-L: 280lm * 1.95 (rel flux) * 0.90 (sag) = 491lm
XP-G: 139lm * 3.33 (rel flux) * 0.82 (sag) = 380lm
(carefully read from the rel flux graphs)
So I got 29%.
The 'most efficient part' is at a few mA and efficacy goes down with increasing current.
2*XP-G S2 are about as efficient as 1*XM-L (they eat a bit more power and give a bit more light).
Those numbers are correct still the gap between emitters is huge. XP-G working above 100 C . XM-L below 80 C ,380 Lumen=71 L/W, 491 Lumen =105 L/W
The hypothetical body mentioned above could be the UF P10 which modded with a XM-L and an AK105 at 2,8A 1A 40ma is an amazing light now
Glad the thermal sag is accounted otherwise the graphs are meaningless since we are talking of power.