That's highly dependent upon the state of boost drivers in the future. Right now getting a boost driver to operate efficiently while drawing more than about 3 amps from a AA cell is problematic given the necessarily small size of the driver board.
But let's say 3 amps and 1.2 volts. That's 3.6 watts into the convertor. Assume 85% efficiency (about the best you can do nowadays for such a setup), and you have 3.06 watts to the LED. Maximum theoretical efficiency for blue plus phosphor whites is around 250 lm/W, so that's about 765 bulb lumens. If we can perfect RGB whites by developing efficient red and green emitters we can potentially get up to 400 lm/W, or roughly 1225 bulb lumens. Of course, LEDs will never reach 100% efficiency. Nothing in this world ever does. At best I'll assume 80%, so the figures are ~600 and ~1000 lumens, respectively. Ironically, in both of these 80% efficiency cases nearly half the waste heat will come from the convertor.
With better battery and convertor technology we can of course greatly exceed these figures.
You cant use 80% efficient because in the real world it just doesn't work that way. You have to consider heatsinking and cell capabilities at high current.
Ex: MrGman testing!!!!!!!!!!!!!!!!!!!!!!
Testing w/ P60 P7 direct drive in a 6P w/ IMR 18650.
1. w/IMR 18650 = 350 Lumens
2. w/ 2 primaries = 450 lumens
3. w/ 3 "C" NiMH cells in Fivemega body = 550 ish Lumens.
Then the same bin P7 driven the same way and made by the same person "Nailbender", but in a 1D Mag. Only improvement was the heatsink.
1. w/IMR 18650 = 704 Lumens
In this test you can see that in the past we used a 6P hosts and using the same cell we got only 350 Lumens, but by just increasing the LED cooling factors using an Electrolumens heatsink for Maglights we more than doubled our lumens.
Next. Using the same bin P7 and the Mag made by Nailbender the same exact way as the 1D Mag P7 w/ IMR 18650 and the P60 P7 drop-in.
This time he built it around a 2D Maglight running "3" NiMH Tenergy 5000mAh.
1. w/ 3 "C" cells= 870 lumens
In this final test we increased both the cells (bigger cells, more ideal for high current) and the heatsink and we were rewarded with lumens almost as the specsheet dictates. 3A of current at the tail with this set-up.
The single cell lumens for the future will rely on the cooling factors needed for high current LED's. Without vast improvements in cooling I don't see how much improvements could be made. My Single Cell P7 and MC-E P60 drop-ins running on a single IMR 16340 were 320~345 Lumens for a split second before dropping. This includes the Malkoff, DX MC-E, and Nailbender made P60 drop-ins.
bigchelis
