Using bounce with light meter I am getting a lux reading of 6.5 with the 60LL with one 17670, 8.5 with 2 CR123's and 10 with two RCR123's. Looking at Big Waffles M60LL runtimes, he was getting perfect flat runtime with two CR123's, and two RCR123's, though higher output with the RCR123's. Looks like the 17670 is not allowing regulation. mA draw with two RCR123's is 140mA's, 150mA's with two CR123's, and 150mA's with the one 17670. I have not tried it with an 18650 yet.
Bill, that is excellent data, both the lux info as well as the mA draw on various voltages. :thumbsup: This info belongs in gsnorm's definitive M60/M30 thread. I did find the info from BW that the M60LL on 2xLiIon / 3xCR123 has essentially the same output as the M60L on 2xCR123 to be very interesting as well.
Edit: This is even more interesting, thinking about it some more. Seeing this constant-current regulation for the M60LL begins to make more sense, the M60LL has always behaved somewhat unusually IMO - remember that GreyShark got adequate performance from the M60LL on 2xAA, much better than the M60 on the same cells? I wonder if the M60L behaves like your M60LL, or more like a constant 140 lumens across various voltages, similar to M60 behavior (otherwise the M60 would overdrive itself @ 7.4v, and I doubt Malkoff would set it up this way). It does sound like the M60 is driven somewhat differently. :shrug:
Malkoff website, M60L:
The input voltage is 3.8 - 9 volts. Below 3.8 volts it will drop out of regulation and run direct drive. The output is 140+/- lumens. The current draw is approximately 350ma at 6 volts.
So the way I understand what Malkoff is writing here is that the M60L will draw more ma as the voltage drops, maintaining ~140 lm until the cell can't deliver 3.8v, which a single large LiIon under load may or may not be able to do. This is most definitely different than your (BB's) M60LL data, which is drawing the same ~150 ma no matter what the voltage is. This perfectly explains BW's odd data with the M60LL on 2xLiIon (7.4v).
M60L:
- 2.1 watts drain (350ma @ 6v benchtop power supply - I am assuming from how I am reading the Malkoff text on his site)
- 2.1 watts @ 3.8 volts would require ~550 ma for the same (regulated) output
- Looking at the Silverfox data for an 18650 for a 0.5 amp load, the voltage of the cell should be at or above 3.8v for 1-2 hours, at least this is the best estimate I can come up with looking at what data there is.
So the fact that large-capacity 18650's suffer such minimal voltage drops at 0.5 amp drain rates should permit the M60L to be fully regulated in this situation.
However, I am making a number of assumptions here, the biggest one is that I am assuming that I am modeling the regulation circuit in the M60L correctly. We've already seen that the M30 behaves somewhat strangely with regards to its output within its voltage range (per Justin Case).