Lumen-hour capacity of flashlights? FL1 and beyond!

(I appologize for a nerdy thread, but it's a flashlight forum so what would you expect? )

Wouldn't the light capacity in lumen-hours be an interesting parameter when comparing different flashlights? Formally I guess it would be defined as the integral of the light output in lumens wrt time and could be measured in a lab, but one could also do a rough estimate using the FL1 specs:

If T is the runtime when 10% of the flux remains (I believe that's the FL1 definition), then assuming the light output decreases linearly (which it doesn't..), the integral would be 0,55*P*T, where P is the maximum flux. A maglite LED 3D would then have approximately 6,000 lumen-hours worth of light in it . I believe a Surefire R1 Lawman would contain around 1/10 of that, for comparison. Given the bulk of the mag that's not a surprise, though..

Thoughts? Be nice, it's my first post

baxtrom said:

(I appologize for a nerdy thread, but it's a flashlight forum so what would you expect? )

Wouldn't the light capacity in lumen-hours be an interesting parameter when comparing different flashlights? Formally I guess it would be defined as the integral of the light output in lumens wrt time and could be measured in a lab, but one could also do a rough estimate using the FL1 specs:

If T is the runtime when 10% of the flux remains (I believe that's the FL1 definition), then assuming the light output decreases linearly (which it doesn't..), the integral would be 0,55*P*T, where P is the maximum flux. A maglite LED 3D would then have approximately 6,000 lumen-hours worth of light in it . I believe a Surefire R1 Lawman would contain around 1/10 of that, for comparison. Given the bulk of the mag that's not a surprise, though..

Thoughts? Be nice, it's my first post

Click to expand...

I don't think I'd like that because most lights have different light levels. And those levels are going to be different efficiencies. So you'd have to say the lumen hours for each light level, when you could just say the run time, which is a real-world thing. The major thing i really don't like about this is it doesn't tell you how many lumens it gives during the specific amount of lumen-hours. Sure you, could say one light is more efficient on one 18650 versus another light on one 18650. but, maybe in the future, this will catch on.

I understood this right? By thinking lumen hours as "mah". Like that both "mah" and "lh(lumenhours), are both ways of measuring capacity, just different units.


EDIT: I think though that this may be useful to people needing long run time lights. Being as most led's are at least 50 lumens on high, which is quite usable for night tasks, the knowing about at what light level the lumen hours is at would be negligible. So, I guess, yes; it would be a good idea.

Sintro said:

I understood this right? By thinking lumen hours as "mah". Like that both "mah" and "lh(lumenhours), are both ways of measuring capacity, just different units.

Click to expand...

yes, the analogy with ampere hour capacity for batteries is a good one. The lumen hour capacity of a flashlight would then be a function of the capacity of the battery, the control electronics of the light (micro controllers, current reducers) and the LED diode itself. Two lights running on the same type of battery pack and having the same nominal lumen output could have different lumen hour capacities depending on the quality of the electronics.

I thought of making an excel table with lumen-hour capacities for different modes of certain flashlights. It's a lot of work and is only interesting for those looking for maximum efficiency; i.e. those looking to:
— maximize light output for a given runtime;
— maximize runtime for a given light output;
— decrease the total amount of cells they carry/cycle/use in total.

Obviously, none of this matters unless you impose some significant limitations. For instance, I don't want to carry a box of cells on me when I'm backpacking, especially if they can't easily be disposed. Stuff like that. In normal everyday use or some short outdoor activity it's not something worth pursuing other than on principle.

Well, it actually IS interesting. Many people make runtime graphes for lights with different modes, but I have not seen anybody acutally integrate over the graph and make a comparison. It would be very enlightening - are lower modes more efficient because of lower led current or less due to higher driver losses?

IMSabbel said:

Well, it actually IS interesting. Many people make runtime graphes for lights with different modes, but I have not seen anybody acutally integrate over the graph and make a comparison. It would be very enlightening - are lower modes more efficient because of lower led current or less due to higher driver losses?

Click to expand...

Just saw a thread over at blf yesterday where a guy took like 10 emitters of 6 or 7 types of of cree leds. and xml's gained lumens linearly compared to mahs applied up until like 2200 or so mah where it slowed down (the lumen gains for mah).


Sorry this was a little choppy, have to go to a brewers game now...

IMSabbel said:

Well, it actually IS interesting

Click to expand...

I thought so too!

Sintro said:

Just saw a thread over at blf yesterday where a guy took like 10 emitters of 6 or 7 types of of cree leds. and xml's gained lumens linearly compared to mahs applied up until like 2200 or so mah where it slowed down (the lumen gains for mah)

Click to expand...

I found this graph of lumens (in %) vs mas for a Cree XM-L. It's also in the XM-L datasheet. Still, correct me if I'm wrong - that's an indication of the efficiency of the led alone, not including the driver. :thinking:

baxtrom said:

I thought so too!


I found this graph of lumens (in %) vs mas for a Cree XM-L. It's also in the XM-L datasheet. Still, correct me if I'm wrong - that's an indication of the efficiency of the led alone, not including the driver. :thinking:

Click to expand...

Oh yeah, never thought of that part. The guy who made the efficiency graphs did it directly with a power supply on his workbench. I suppose you'd have to factor in a driver, which will complicate things greatly, but once you boil it down to lumenhours, not so much.

Sintro said:

Oh yeah, never thought of that part. The guy who made the efficiency graphs did it directly with a power supply on his workbench. I suppose you'd have to factor in a driver, which will complicate things greatly, but once you boil it down to lumenhours, not so much.

Click to expand...

If you want those xml and sst graphs of efficiency, go over to node 2603. Also, I made a post about this topic over at blf and heres the responses I got:


Interesting concept I think. So how then do you account for the variation in cell capacity? That will completely vary from cell brand to brand. Even the same exact brand of cell will have differing capacities from unit to unit, batch to batch.
For me, Lumens per watt efficiency is a more useful way to evaluate each light. (Lumens OTF divided by tailcap power consumed). The variables still remain: OTF Lumens will drop with temperature, and voltage sag. Wattage consumed will drop (with Vin drop), unless its a theoretically perfect boost regulated circuit.

While interesting it makes it so manufactures could have a high lh rating by driving a XM-L with low currents. I think a runtime graph is much more useful.

Now I look at it, lumens per watt still is the best way to show each light because of factors that would affect lumenhours:
-cell capacity
-variations
-voltage sag
-temperature sag

I agree regarding the cell capacity variation. However, the runtime specified in the FL1 standard also depends on cell capacity. I assume each manufacturer would pick the best cells available when measuring runtime and lumen-hour capacity (as well as the FL1 lumen rating which I believe is measured after the flashlight has been turned on for some time). The lumens per watt measure is a great indicator of the efficiency of the light, however, a total light capacity indicator might be of interest to the layman user.

It's getting late over here in scandinavia. Good night and good luck! :sleepy:

Yes, very useful for looking for long running, low lumen lights.
Also very useful for looking at published lumens and runtimes to determine a light's most efficient levels.