4AA vs. 2x2AA with regulation?

Intuitively it makes sense that a regulated light -- let's use the EverLED bulb as an example -- will run longer using more cells, such as 4AA vs. 2AA, possibly because the higher voltage available will cause the regulator to demand less current from the batteries. It also makes sense that two parallel strings of 2AA (2x2AA for short) will run longer than 2AA since the parallel strings effectively double the power capacity of the light.

But which is the better of the two strategies to lengthen the runtime of a regulated light, 4AA or 2x2AA?

post deleted, stupid post.

I think they are about the same.

shiftd-You having a bad day ? Take an aspirin and call me in the morning.

lol JohnK.
not really. It is just that my left hemishpere takes over today, so i am in the feeling of babbler-lover and my frontal lobe got something going on in it. ups, now i am babbling again. I reallly need the aspirin john /ubbthreads/images/graemlins/wink.gif

This is a difficult question to answer in general. Too many variables. But in the case of the EverLED, it appears intuitive that the 4AA would run longer than the 2x2AA.

My reasoning is as follows:
The EverLED draws less current/is more efficient with higher input voltage (up to 9V IIRC). From someone else's experiments with input voltage and current versus output (I can't seem to find that post at the moment), it would appear that 3V is sort of the beginning of the end, ie the input current starts to increase dramatically for every slightest drop in input voltage. While the 2x2AA setup might be able to handle the amp requirements at this stage better than the 4AA setup, it is already paying the price for the reduced efficiency, whereas the 4AA setup has already expended half its capacity while the circuit was performing more efficiently.

To summarise, this conclusion is specific to the EverLED and that the 4AA setup will provide longer runtime because it is the efficiency of the circuit at the higher input voltages that will make the difference.

The parallel system could handle twice the current since it would be shared across both sides of the parallel system. Or to put it another way, if the current draw in boost mode in the 2-3 volt range is double the current draw in buck mode in the 4-6 volt range, then it would be a toss up. That chart mentioned above could answer the question. My guess is that series would last longer than parallel. It's a good question.

I'm actually running an EverLED in a 2x2AA arrangement right now; maybe I ought to move it to a 4AA light instead. /ubbthreads/images/graemlins/ooo.gif

Actually, I used the EverLED as a specific example, but asked the question in order to gain general knowledge. It's a design choice that's always been troubling because I couldn't see a clear answer. Thanks for the insights, folks! /ubbthreads/images/graemlins/grin.gif /ubbthreads/images/graemlins/thumbsup.gif

milkyspit: Don't take what we say as the final answer. It is merely an exercise in logical argument from figures that aren't even readily available (still can't find it /ubbthreads/images/graemlins/smile.gif )

Since you have the EverLED and the 2x2AA setup, perhaps you could device a way to do a runtime test? Brightness vs Time and input current (and voltage if possible).

Steelwolf, it's not my highest priority (the runtime test, that is), but I'll give it some thought. It would be nice to get some experimental data on this.