Moonlight long duration testing

[WalkIntoTheLight]

Has anyone ever done any tests on how long moonlight modes really last? I'm interested to know if they really last as long as advertized. For example, my XPG2-based Quark says 720 hours. My SC52 says from 3 weeks to 3 months. Are those real performance numbers, or just marketing guesses?

I tried to do an accelerated test using a AAAA battery in a SC52, but even that lasts too long for my patience.

I tried measuring current draw in moonlight mode, but I don't think my DMM is accurate enough because it kept bouncing around.

The best I could get out of my testing with depleted batteries is that the SC52 acts strangely when the battery gets very low. The L1 (2.7 lumens) mode continued to shine properly, but the L2a (0.34 lumens) mode would flash instead of shine continually.

None of that tells me what I really wanted to know... how long would it really last? Has anyone done long-duration tests on similar 1xAA or 2xAA based lights?

 

[kj2]

http://www.candlepowerforums.com/vb/showthread.php?361620-The-ArmyTek-120-day-Challenge!!!

 

[Hondo]

Here are a couple of good ones from a quick search - bottom line on the Quark is no, it lasts longer than claimed:

http://www.candlepowerforums.com/vb...ight-Runtime-Test&highlight=moonlight+runtime

http://www.candlepowerforums.com/vb...ight-Runtime-Test&highlight=moonlight+runtime

I have heard about the Zebras getting a bit wonky at the end of a battery. I think Quarks, and the OLD HDS lights, and the Novatacs that followed them, are some of the best loooong runners out there. Many of the lights nowadays that have very low modes also have so much overhead current draw from the circuit that they don't really run as long as you would expect for such a low level. Control ring lights are that way.

The one that blows my mind is my old CPF special edition VB16 from Wits' End. Not only does it have modes that go all the way down to nothing (32 total, actually), but I've been using a fairly significant sub-lumen level (0.1 - 0.2-ish lumens, I would guess) for a nightlight, most of the night long. I had not charged the old laptop 18650 in about 3 or 4 months, so I took it out and checked it. Still something like 4.05 volts. Now that is a battery sipping light.

 

[mcnair55]

I see no point in the maker advertising a claimed moon light run of so many hours then for it not to reach its specified time.Why not do an experiment yourself and then you can compare notes with the maker.

 

[reppans]

Yes, I have - also on AAAAs from a 9V, and I use a lightbox, but have no data logger. Runtime is one thing, but efficiency in lumen-hours is another and takes into account the light's output. I'm a low lumen efficiency fanatic and prefer the "brighter" moonlight modes in the 0.3 lm range, since they're bright enough for most close task work with dark-adapted eyes, yet will run 200+ hrs per AA - it's my most often used mode along with 3 lumens. Here's a pic from somewhere in the middle of my test run:

There's lots of caveats as some lights ML mode greatly varies with temperature, voltage, and QTC settling, but the lights you mention have rock solid current regulation, so they're easy.

Light / AAAA hrs / AA Alk equiv. (x 4.62) / Lumens / Efficiency (Lumen*Hrs)

QAA-XML2 / 61 / 281 / 0.3 / 84
QAA-XPG2 / 75 / 346 / 0.2 / 69
SC52 CW / 92 / 424 / 0.07 / 30 (< set on 0.34 L2a mode)

The SC52 did run the longest, but it is so dim, it turned in the lowest efficiency number (interestingly, it claims the highest lm-hr efficiency spec by 2-3x ). The surprise light was the Malkoff MDC AA which, with a PWM driver, pulled 155 lumen/hours efficiency (75 hrs, avg 0.45 lumens) but the caveat is that this moonlight is highly temperature sensitive - @ 60 degrees or below, it's ~0.6 lumens, and ~80 or above (hand warmed), it's ~0.15 lumens. I tested @ 60 (winter attic)... Don't know if testing in 80 degree temps, dropping output to a quarter, will increase runtime by 4x, but I doubt it.

 

[WalkIntoTheLight]

Very interesting, thank you.

I'm surprised the XPG2 Quark only ran a little longer than the XML2 Quark. I have both, and the XPG2 Quark moonlight mode lists it as running twice as long as the XML2. Do I understand correctly that I should double your AA Alk numbers to estimate the light running on 2xAA cells? If So, the XML2 runs much longer than 4sevens states.

The SC52 surprises me. I know its moonlight isn't as bright as the Quarks, but I didn't think it was that dim! It doesn't look that much dimmer to my eyes, but it's really hard to estimate. Its beam is more floody than either Quark, so I thought it just looked dimmer because it was more spread out. Your estimate would put it at 2.5 weeks runtime, a little less than spec. I wonder if the L2b and L2c modes run anywhere near 2 and 3 months? The L2c is extremely dim, so it wouldn't surprise me.

I tried measuring current draw on the SC52's L2a mode, and it came out to 80mA. So I'm obviously doing something wrong. Edit: Tried again. This time, I'm getting around 5.4mA on L2a, 22mA on L1, 0.6A on H2b, and 2.5A on H1. I think those numbers sound a bit more reasonable. Given that, I should get around 370 hours on L2a moonlight from a 2000mAh Eneloop. That seems to fit your results on a (higher capacity) alkaline.

 

[reppans]

Keep in mind that sample variation can be a big factor in my results, and moonlight mode may possibility be the most volatile of any mode. Yes, you should double, and perhaps even add a bit more, for 2xAAs since the higher voltage will require less boost to attain Vf of the LED, and therefore be more efficient. I know 47 specs say the XPG is more efficient on the lower modes, but I've run a few of these tests (and w/ different lights), and on other low-range modes and find them in the same ball park on a lumen-hr basis some better, or worse, depending upon the mode - in any case, the results were close enough that efficiency is not a factor between which light I grab.

The human eye has a high dynamic range and natural auto exposure/white balance that tends to minimize differences, IMHO, or perhaps the camera just exaggerates tint and lumen differences. I don't know if my SC52 is abnormally dim, but the few photo comparisons I've shared with others seem to indicate it is normal. Since we share the same three lights, I'd appreciate it if you could post 2 similar side-by-side photos to help verify. One photo in the dark to compare spill, and one with enough ambient light to compare hotspots so they are not badly overexposed (underexposing a manual camera a few stops would do as well).

Keep in mind, since these lights are well regulated, the current draw should increase as voltage drops, but your readings certainly sound close enough for government work .

 

[WalkIntoTheLight]

Sure, I'll take some pictures to show moonlight levels. But I'll wait until I get the SC52w-L2 I ordered, so I can include it too.

 

[Wiggle]

I know someone has tested Armytek lights and they seemed to meet or exceed the very long moonlight runtimes they promise.

 

[subwoofer]

I know someone has tested Armytek lights and they seemed to meet or exceed the very long moonlight runtimes they promise.

:wave:

http://www.candlepowerforums.com/vb/showthread.php?361620-The-ArmyTek-120-day-Challenge!!!

And yes that would be me with my Armytek 120 day challenge. I didn't bring it up originally as the OP seemed to be asking about AA powered lights.

 

[think2x]

My son did his own moonlight challenge with his older XP-E Quark AA2, it lasted every bit of the 30 days they claimed from a pair of Duraloops.

 

[WalkIntoTheLight]

I don't know if my SC52 is abnormally dim, but the few photo comparisons I've shared with others seem to indicate it is normal. Since we share the same three lights, I'd appreciate it if you could post 2 similar side-by-side photos to help verify.

I posted this moonlight comparison in another thread, http://www.candlepowerforums.com/vb/showthread.php?387195-Zebralight-SC52w-L2-comparison-to-SC52, but I guess it also belongs here.

It compares the L3 Illuminations L10 moonlight (on left, with Nichia 219A), followed by the Zebralight SC52w-L2 (XML2 neutral), then SC52 (XML cool), then the 4sevens Quark QP2A-X (with XML2), and lastly the Quark QT2A (with XPG2).

The Zebralight's L2a moonlights seem fairly consistent in brightness. The XML2 based Quark has a much brighter moonlight. The XPG2 based Quark is also brighter than the Zebralights, but quite a bit dimmer than the QP2A-X. Finally, the L10 is way dimmer than all of them. I think it's probably closer to the L2b moonlight of the Zebralights.

I also did some more accurate measurements of current consumption on the various modes of the SC52w-L2 (after I figured out how to use my DMM!). It was done with a partially depleted Eneloop battery (2000 mAh).

Here they are:

High 1 : 2.6 A (~ 45 min)
High 2b : 522 mA (~ 3.5 hours)
Medium 1: 290 mA (6.5 hours)
Medium 2a: 189 mA (11 hours)
Medium 2b : 86 mA (23 hours)
Low 1 : 29 mA (3 days)
Low 2a : 5.6 mA (15 days)
Low 2c : 2.1 mA (40 days)
Off : 20 uA (11 years)

It's interesting to note that the very lowest moonlight mode still consumes more than a third of the power that the brightest moonlight mode does, even thought it's WAY dimmer than a third. I think Zebralight is overestimating the run-time of the moonlight modes, especially the dimmest ones.

 

[reppans]

Thanks for following up with the ML pics. Yeah looks pretty consistent with my lights too.

I appreciate everyone likes different ML levels - I like "brighter" ones which tend to be my general purpose low modes used ~40% of the time. For those limited times I need dimmer than 0.3 lms (middle of the night bathroom runs) I bury the bezel in my fist and use my pinky/ring fingers as an aperture control. The SC52 went on the shelf for me - wish they were more truthful/accurate . Also seems silly to offer 3 ML modes will all being under 0.10 firefly, speaking of which, here's an old pix next to "0.09 Fireflys" of a Thrunite T10 (I measure 0.10) and the L10 N219 (I measure 0.02) - CLICKY.

Not sure how trustworthy tailcap current readings are - for example, the 3.5 hours for H2b is more like 2.5 hrs which Selfbuilt, THIS review, and I got. (and it's interesting to note that ZL specs ~70% greater efficiency in lumen-hours than Eagletac for what is virtually the same performance - 108x3 vs 75x2.5)

The SC52 surprises me. I know its moonlight isn't as bright as the Quarks, but I didn't think it was that dim! It doesn't look that much dimmer to my eyes, but it's really hard to estimate.

FWIW you can use your DSLR as lumen meter in very much the same way as you described in the "Lux Meter" thread..... check out my reply there if interested.

Thanks again!