Just playing with some of Match’s figures after extrapolating some rough numbers from these graphs in post 1: Emitter Test Results pt4: XML2, XP-E2, XP-G2, and Nichia219! (along with sinkpad tests)
XM-L2 on copper mcpcb at room temp mounted on massive al sink:
2.7A vF 3.38V = 1050 lumens x 40 = 42000 lumens (Base configuration) Power requirement: .2V (driver) .68V (wire losses) + 13.52V (4 emitters) = 14.40V (2 drivers x 13.5A) 27A x 14.40V = 388.80W
3.0A vF 3.44V = 1125 lumens x 40 = 45000 lumens (7.143 % increase) Power requirement: .2V (driver) .68V (wire losses) + 13.76V (4 emitters) = 14.64V – just for comparison @3.0A at the emitters…
3.4A vF 3.51V = 1240 lumens x 40 = 49600 lumens, (18.095 % increase) Power requirement: .2V (driver) .68V (wire losses) + 14.04V (4 emitters) = 14.92V (2 drivers x 16.9A) 33.8A x 14.92V = 504.296W
In going from 2.7A to 3.4A, there is a difference of 7600 lumens (18.095 % increase) and a 29.9 % increase in watts. It is said that the human eyes need a 20% increase in lumens to notice a change in perceived brightness. It is also noted that emitter surface brightness increases throw, which is often why we overdrive XM-L2 on copper up to 6.5A in thrower flashlight applications. But with the C8 sized reflectors that I will be using, the increase wont be dramatic… only noticeable.
To run at 49600 lumens means that Id have to rework my alternator regulator to put out 14.92V nominal. Thats getting up just a bit too high to risk frying other electrical components on the truck. At 42000 lumens, I can run the stock alternator voltage regulator at 14.4V nominal… so this looks like it will probably be a 42000 lumen light.