Single Cell AA light shootout

MDXL cheapie 15 lux / 3 lumens

Aurora SH-035 12 lux / 2 lumens

Zebralight H50 11 lux / 2 lumens but because of its design the lightbox may not be capturing all the light it is putting out.

Ultrafire C3SS 4 lux / 1 lumen

Powerlight <1 lux / not a lot

I can only say that the Akoray K-106 ran for more than 8 hours, likewise the F25 (All the remaining lights were lit when I checked at 2am). Given that the BLF DeLight and the F22 were both flashing I've allocated them a >600 minute runtime.

I do not have the facilities to do this for 20 lights simultaneously.

13hr exactly, the C3 stainless started to flash at me. Voltage 0.85V.

Hmmm, interesting. I should test my Ultrafire C3 Stainless again. I would have expected at least 20 hours. And it appears than the F25 is significantly brighter than the C3 Stainless on low? I wonder if my 3/5 star rating on the 14 hour runtime of the Ultrafire M2 is a bit too low after all?

I got 12.5 hours on an alkaline to 50% on the C3. I'll need to stick the F25 on the box and get some better numbers. Thus far it hasn't even been out of the house, let alone exposed to outdoors darkness. At some point I need to gather together all the runtimes and put them on a spreadsheet. Problem is, runtimes are the slowest of all to do - must play with the Linux box which is steadfastly refusing to update itself and see if I can persuade it to do two simultaneous sets of measurements, the Windows box will only see one of the meters at a time - and it isn't always clear which.

I suppose a full data set would be something like

Outputs at all (non-flashing) levels from all cell types. Since I'm not interested in the flashing levels and runtimes are likely to be long, it would be tedious to test those - let alone I have no way of measuring with the necessary time resolution.

Current draw at all levels from all cell types

Runtimes at all levels from all cell types (This gets expensive and is unlikely to happen)

Temperatures reached (This is not very easy to get reliable results from)

Weights

Sizes

LED type

Construction quality

Beamshots indoors and out

Presence or absence of PWM.

Other stuff

Complete data sets are an ideal rather than something I ever expect to achieve. Comparisons do interest me though. I'll get numbers for the F25 soon.

Thanks Don for your hard work! Fascinating stuff. Is there a way to make the Google spreadsheet temporarily sortable by column perchance?

I don't think so, but it will be easy enough to put in more tabs sorted by each heading - problem is that sorting it comes under editing it which is likely to lead to vandalism sooner or later unless I were to specifically authorise each person using it which would be tedious in the extreme for everyone. I can easily enough copy and paste into new tabs and sort them by each heading.

Might get around to graphing it some day.

You only have to spotcheck each at your convenience. Batteries also conserve their charge. :)

The Powerlight has now given up - some time in the last 4 hours. I didn't expect it to make it till lunchtime - runtime over 14.5 hours though. 0.81V

The MDXL is very, very dim so I'm calling it 18.5 hours - 0.79V which is lower than I'm keen for an NiMH to go

SH-035 still going strong, as is the Zebralight. Both showing 7 lux / around 1 lumen. Now 18hr 50min

But it does considerably increase my investment of time. Remember, I'm doing this for fun, not as a job.

I did not have the time last night as I had stuff to prepare.

You want better methodology, feel free.

Aurora and Zebralight still going strong at 23hr.

The Powerlight will make light from an alkaline for longer than with an NiMH it seems, the current draw is low enough that you actually get the benefit of the higher draw at lower current. Likewise the MDXL both of which should be capable of over a day on an alkaline.

Aurora SH-035 and Zebralight H50 still going. I wasn't that impressed with the Aurora as I was expecting a better Ultrafire C3 Stainless Steel Q5. I got a dimmer light that is severely overdriven (To the point of the LED turning bluish) with a 14500 - the XP-E in it can't tolerate the over-current that an XR-E can. Its niche would appear to be incredibly long runtime on low. It is still putting out 12 lux to the Zebralight's 10, both around the 2 lumen mark. One lux is about the minimum level a human eye finds useful without severe dark adaptation, i.e., the level at which colour vision is still possible. You can see at lower levels, at least for sufficiently low values of see.

Aurora SH-035 drops out, Zebralight still going. Cell at 0.845V

Maybe it's just a matter of perspective, but I'm a curious person by nature and it would bug me to not know why something occured the the way it did. It's not a matter of "methodology", but I wouldn't feel like I understood the runtime number otherwise.

For example, I don't care much about absolute runtimes per se, but at least initial draw rates (corresponding to "lightbox" numbers) would be very telling about the efficiency of different lights, especially if they are related to lumens. For the all the talk about bins and whether a light is Q3 or Q5 or whatever, not many people ever bother to try and figure out if a light is actually efficient. In a sense a simple draw power-->output calculation is much easier and more telling than the emitter bin. It would substantively qualify what your lightbox number means. Again, not methodology for the sake of it, just quite valuable info. Better would be this value integrated over time (ie this test), but just one simple measure initially would carry a lot of weight.

You know I would do it if I had the lights. :)

Sorry - bad day at work.

However, I genuinely couldn't have spent the time doing it last night - I am firmly of the opinion that the more measurements the merrier. I must investigate the Linux datalogging stuff so I can run two meters at once. Then unattended measurements which are rather less time-consuming and can log a lot more information can be done. Like you say current vs. output is useful. For instance the new Q5 Ultrafire C3 on 2AA NiMH compared to the old (Also 2AA) makes more light while drawing an amp less. Which is just one of the reasons why it is enormously better than the old one.

The Zebralight is still going. OK this guy isn't a budget light, but is seriously good and can be found for around $40 sometimes. They ship from Shenzen which can be slow - 4 weeks EMS.

Still at 12 lux/1 lumen. Current and voltage: somewhere between 112 and 35mA, 1.26V. Looks like it has some time to run.

And no, I have no idea why the variability in current is happening - I suspect mode switching.

It cannot be 112mA as 112mA for 28 hours = 3,136mAh and Eneloops just can't hold that much.

Yes, I've noticed the same thing. The Powerlight is good for almost 24 hours on a junky alkaline, considerably less with an Eneloop.

Thanks again for these very useful tests Don!

The adjudicator gave up before the Aurora and Zebralight - she went off to her bed an hour or so ago. :)

The Zebralight is stil at 10 lux on the box. Around 1 lumen. Current measurements seem to measure what it'll do on high after a day. Not exactly meaningful. Will leave it till it gives up.

hey don, did you try the headlamp with a 14500? if so, does it take apart easily for troubleshooting?

Zebralight still hanging in there. Still giving its 12 lux. If the manufacturer is correct it ought to run for 53 hours.

I did try the headlamp with a 14500, no luck and I've not got around to disassembly yet mainly because I've not figured out how to.