Crowd sourced run time testing

[3_gun]

So I got bored today & did a RT test on my Sofirn IF25a/4K light. That bring me to this .. most run times are BS marketing propped up by an ANSI standard that most don't understand. It is ..
"The duration of time from the initial light output value (that's 30 seconds after the light is turned on with fresh batteries) until the light output drops to 10% of the initial value".

Not really useful in the real world if you buy a light needing 2000L for 2hrs but end up getting 2000L for 1min and end up at 200L at the 2hr mark. That's what a F1 test allows to be reported as run time

So I picked a level that I use often for normal GP flashlight use .. 125L. I used a fully, freshly charged quality battery & just let it run until the FIRST step down, making notes of the starting & ending volts.

Light .. Sofirn IF25a @ 4000K
Battery .. Samsung 30T 3000mah 21700
Level .. 125L
Time .. 8 1/2hrs
Starting V .. 4.2 Ending V .. 3.0

Try and test a light level that is a normal set level for that light. Sofirn reports the IF25a starts at 125L after a battery change so I set that as memory for testing. I don't know that it really is 125L but every IF25a should be close to that doing the same thing. I tested using stepped ramping so the step down would be clear; some lights might be harder to tell but do your best

So lets get some REAL numbers to compare. Use whatever light you want just be sure to stick as close as you can to the outline.

Light .. Fenix E35 v3
Batteries .. Fenix arb-l21 5000mah 21700
Level .. 450L
Time .. 5 1/4hrs
Start V .. 4.2 End 3.1

Who's next?

 

[3_gun]

Light .. Acebeam E70-AL
Battery .. Sofirn 5000mah 21700
Level .. 180L (Mid-1)
Time .. 14 3/4hrs
Start V .. 4.2 End 3.2
The light stepped down to low. I ended the run there w/o timing how long it might have run at that level. Run time was a bit better than claimed for Mid-1 [12.75hrs] but 3rd party testing had Mid-1 closer to 140L than the 180L claimed, was that the reason for the longer run? I may recharge & try Mid-2 next

 

[3_gun]

Light(s) .. Sofirn S11c
Battery(ies) .. Sofirn 3000mah 18650
Level .. 300L
Time .. 3 1/4hrs
Start V .. 4.2 End 3.6

Ran both of my lights side by side for this one. Used 300L middle for testing as it's the level I usually have set as my memory level. Ended the run before an actual step down as I could see a drop in brightness when tested in my "black hole" hallway. Rough guess I was still getting at least a 200L level of light. All levels, ML to turbo would still activate when cycled. Wasn't really surprised by that when I found the battery level was still over 3.5v. The lights are not identical but are within what I'd imagine is a reasonable tolerance. Claimed run time of 4hrs seems possible is you waited for the light to actually drop to low(30L) mode. Solid useable lights for a price before tax of under $10.

 

[3_gun]

Light .. Sofirn SP32a V2
Battery .. Sofirn 3000mah 18650
Level .. 160L (med)
Time .. 10 1/4hrs
Start V .. 4.2 End 3.2

Light stepped down (low 50L) & LED indicator turned red pretty much at the same time. Didn't run the light on low or wait for LED to start blinking, I'd guess at 2.9 or 3.0v. Did try & run a cycle; high would only blink so I didn't even try for turbo. Really not much to dislike about the light. OK UI, good spacing of the light levels(plus there is ramping mode), beam is a nice mix favoring flood but still having reasonable throw. Might try a run on high (500L) but might run into temp issues, 10 or so minutes gets warm but I've never run it enough to get hot. For $20 (w/battery+charger&shipping) it's hard to go wrong with this one

 

[fuyume]

No one really buys a handheld flashlight expecting 2000 lumens for 2 hours, do they?

The latest models, like the Fenix PD36 TAC I just got, will put out 350 lumens for over 10 hours with a flat brightness curve on a fully charged 5000 mAh 21700. None of the modes above that will produce maximum brightness for the entire duration of the battery's charge, and I don't think anyone is really expecting any single-cell light to do otherwise.

 

[3_gun]

No one really buys a handheld flashlight expecting 2000 lumens for 2 hours, do they?

If I only kept a record of how many "bubbles have been burst" on guys who've come to a low/no light gun class because that's just what they thought. Most had gotten gun show specials thinking they bought a light for 1/3 the cost of a Surefire that was going to work better.

I'd bet 1/2 of this board hasn't read or fully understands the ANSI F1 standards. Sadly for many of my classmates their "GSS" didn't even meet/have F1 testing.

So yes there are a lot of people who see the marketing without understanding the truth behind it. And just like an Apps user agreement they don't read the small print about the light either

The latest models, like the Fenix PD36 TAC I just got, will put out 350 lumens for over 10 hours with a flat brightness curve on a fully charged 5000 mAh 21700.

Will it? My Fenix E35v3 got sorta, kinda close to the claimed run time of 7hrs. Close enough (5 1/4hrs) I ain't going to *****. But I only know this because I charged the battery, I set the light level & I timed the run.

Some lights like the IF25a don't make run time claims. At least now (if they see this thread) they have something to go by. Will their light match mine? I doubt it given the unknown tolerances built into the lights & batteries. But hopefully they'll be close & if enough people post the same lights often enough we'll see a trend

 

[3_gun]

Light .. Sofirn IF25a @ 6500K
Battery .. Samsung 30T 3000mah 21700
Level .. 125L
Time .. 8hrs
Starting V .. 4.2 Ending V .. 2.9

I was curious if light temp made any difference in run time. The minor change in the figures are well within what I might expect if I'd retested the same light again, so no not really. Might a five test average show a trend one way or another? Or better yet testing the lights side by side to reduce variables help? Maybe, maybe not but since the usefulness of knowing that % is kind of pointless I'll leave it to someone with OCD to work out

 

[vicv]

Are you using a lux meter to insure output isn't dropping? It can be hard to tell

 

[aznsx]

I do fully understand and fully support the FL1 standard as it currently exists, and it does as intended; to allow me to compare one light to another using the same 'yardstick'. When I want to know (subjectively) if a light will meet my personal expectations, as I've said before, nothing can replace an output / time graph. Most of the lights I buy are from manufacturers who provide both of those pieces of information. I won't speak for the average consumer, because I'm not in that market segment.

 

[3_gun]

Are you using a lux meter to insure output isn't dropping? It can be hard to tell

I try to stay with lights that have published testing done by the Mfg (Like Fenix,Acebeam) or by 3rd party(Flashaholic,1lumen) testing or both with flat graphing. Failing that I try doing levels that match a known & trusted light in my collection. Last (& biggest PITA) I take a pic every 1/2hr or so at a fixed iso, shutter speed & iris and trust my eyes. I'm just trying to get close.
I try to avoid levels that could or should have heat issues in longer run times. Plus those (lower) levels are where I commonly use my light(s) 90%+ of the time and heat is seldom an issue there.

I do fully understand and fully support the FL1 standard as it currently exists, and it does as intended; to allow me to compare one light to another using the same 'yardstick'.

That is if the mfg follows it. F1 states that run time levels as claimed are to be at 30 seconds after start up on a fully charged cell until a 10% brightness level is reached. Yet how many of the tested lights market the start up levels not the 30sec out put?

The F1 would be more useful if run time was until the loss of 35-50% of the light at start up. The numbers wouldn't be as marketable but they would be closer to the reality and much more understandable to the "norms" if they bothered to read it.

I didn't set out to do a scientific paper, just establish numbers for comparison that anyone could do & report. I'm trusting every IF25a (or any other light) when tested at known level will be the same.

 

[vicv]

No cool idea I'm not disparaging it. Just asking your method. Thank you for putting in the work. If the eye can't tell if the brightness has dropped, it doesn't matter anyway

 

[3_gun]

Light .. Sofirn IF25a @ 6500K
Battery .. Westingtonhouse 2000mah 18650
Level .. 125L
Time .. 4 1/4hrs
Starting V .. 4.1 Ending V .. 3.0

No shock here, a smaller battery gets you less run time. Expected a bit more run time but not enough under that I'm alarmed. Other battery slightly over 3k/mah; this one slightly under 2k/mah? My guess both are within spec since none of the time are far from expected.

 

[bykfixer]

When Maglite first introduced the 3 cell LED version of the ML25 it claimed 17 hours to 10% yet the 2 cell version claimed only 2 hours.
The 2 cell version was rock solid same output for the 2 hours and quickly sagged after that. Well, in about 10 minutes it went from full to dim. So that was easy to see.

The 3 cell version had a slow taper so a lux meter was needed to see how it faded. Now it did meet and exceed the 10% after 17 hours but it did it so slowly my eyes could not tell. Did I stare at it for 17 hours? No way. My life aint that dull. But checking on it from time to time I could not really see a difference until it had reduced to around 50% or less.

I was impressed with the results. Yet these were not lights touting massive outputs so heat was not a factor as much as the source of power to light them was.

I think that too is where "the public" is misinformed due to many rechargeables (called re-wraps) claiming this or that and cannot deliver the required supply.

 

[3_gun]

Light .. Klarus EP10
Battery .. Klarus 2000mah 18650
Level .. 200L
Time .. 7 3/4hrs
Starting V .. 4.2 Ending V .. 2.8

On board battery level still shows a 25% or less light & if restarted 200L is reached .. for about 20seconds. I'd guess it was still at 100-125L at the end. Still useful but noticeable it had dimmed. It had not stepped down to low (30L) when I ended the run. Really don't think I'd like to get to or below below 2.8v on a regular basis. Run time well short of the 11hrs claimed by Klarus & a 2000mah battery is what they say was used for testing. I'm not buying it. Still for the cost of the light & how well it does work, it may be worth while up grading to a better battery when this one kicks it.

 

[3_gun]

When Maglite first introduced the 3 cell LED version of the ML25 it claimed 17 hours to 10% yet the 2 cell version claimed only 2 hours.
The 2 cell version was rock solid same output for the 2 hours and quickly sagged after that. Well, in about 10 minutes it went from full to dim. So that was easy to see.

The 3 cell version had a slow taper so a lux meter was needed to see how it faded. Now it did meet and exceed the 10% after 17 hours but it did it so slowly my eyes could not tell. Did I stare at it for 17 hours? No way. My life aint that dull. But checking on it from time to time I could not really see a difference until it had reduced to around 50% or less.

I was impressed with the results. Yet these were not lights touting massive outputs so heat was not a factor as much as the source of power to light them was.

I think that too is where "the public" is misinformed due to many rechargeables (called re-wraps) claiming this or that and cannot deliver the required supply.

Batteries are key to light performance, no doubt.

I try not to stare to much or often. Having a known background helps in noticing dimming. And a light near the test level to use as a reference is huge. Short of that I try to avoid doing a picture routine with manual settings. Mk2 eyeballs work well with side x side pics but it's a pita. This isn't a science project so just starting with known specs & paying attention is good enough to identify trends.

Old D & C cell lights were horrible. I'm sure many replaced batteries WELL BEFORE light levels dropped to 10%. Better batteries came along before LEDs but they were a minor improvement compared to 123a & xenon bulbs later. Even then it wasn't unusual to use a set of 123a a night (or more) if you wanted to keep the light at full power. Most of the guys I knew back then replaced the light assembly once a month, needed or not. Surefires gave little to no warning before they died. It wasn't odd to see time notes for the light within field notes. Things have gotten so much better

 

[bykfixer]

Back when PK was with SureFire he said they were using Duracell batteries until they developed their own in house in an effort to supply a more steady performance.
We tend to take things for granted now-days although the physics of todays massive current draw in some flashlights still taxes the laws of physics.

I like the plain-Jane aspect of runtimes based on old school approaches versus the computer generated charts and graphs, which add value in my view but are just a guideline to me. I think ANSI is still relevant but like you said, what a product does at say 35% tells me how useable in real world situations the product would be, then perhaps as a caveat the 10% could be noted as well.

 

[Stefano]

Lately I don't even trust the graphs that are published.
I have several lanterns of a famous brand, at the level of 25 lumens they (according to the graph) should have constant brightness, I found that after exactly two hours it starts to decrease and after 30 minutes the brightness is exactly half of the initial value.
At first I thought of a faulty specimen but I have three pieces bought at different times, all of them do this.
As someone wrote in another discussion in the past there was the lumens war, now there is the runtime war.

 

[3_gun]

Lately I don't even trust the graphs that are published.
I have several lanterns of a famous brand, at the level of 25 lumens they (according to the graph) should have constant brightness, I found that after exactly two hours it starts to decrease and after 30 minutes the brightness is exactly half of the initial value.
At first I thought of a faulty specimen but I have three pieces bought at different times, all of them do this.
As someone wrote in another discussion in the past there was the lumens war, now there is the runtime war.

Feel free to do a post .. name names. You may save one of us from the same fate. Try to follow the guidelines set in the first post & include as much extra info you think we might need

 

[vicv]

Lately I don't even trust the graphs that are published.
I have several lanterns of a famous brand, at the level of 25 lumens they (according to the graph) should have constant brightness, I found that after exactly two hours it starts to decrease and after 30 minutes the brightness is exactly half of the initial value.
At first I thought of a faulty specimen but I have three pieces bought at different times, all of them do this.
As someone wrote in another discussion in the past there was the lumens war, now there is the runtime war.

Do you have a lightbox? I'm curious how you're getting the exact numbers. If so that'll be a great resource. Thanks

 

[Stefano]

No, I don't have a lightbox (wish I had)
But I noticed with my eyes a decrease in the lantern, after several times that I saw it happen I turned on the lux meter and took note of the initial value and I checked it periodically.
It keeps the brightness constant but after exactly two hours a slow decline begins, after 30 minutes from the beginning of the decline the lux is exactly half of the initial value.
Test done several times with two different lanterns (one red and one green) both with original battery and with others battery.
I was worried it might be a flaw but I don't have it, it's just a way to save the battery when the eyes get used to it.
If someone thinks that it depends on the voltage drop of the battery, I mean that if after the drop I turn the lantern off and on again I have constant brightness for exactly two hours.
This lantern also exists in AA version and the behavior is identical (but it happens more quickly, I have not tested exactly how much but it happens before two hours)