Calculating actual runtimes for emergency use

LED Flashlights – General Info
1

I figure its a good idea to know how long each battery would last in case of power failure, but its not easy to make a list for a few lights of how long they will last

Starting with 105C/Qlite driver, which all of us will have at least one of


I have 2600mAh batteries, so i just assumed i would get 1 hour of runtime as the battery falls out of regulation and power consumption decreases (2.8A driver), and in an emergency you won’t use high for very long if at all since you want to conserve battery power





HKJ:Due to the fast pwm, the driver does not turn fully on. This makes the driver drop out of regulation at a fairly high voltage.
So how can i calculate actual runtime for a theoretical 1000mAh battery, for easy conversion ability i can multiply by 2.6 for 2600mAH or whatever battteries i have available or can’t i (experts please chime in)?
Also i have found medium does not last the over 4 hours that is theorized, i seem to get about 1.5 hours on sanyo 2200mAH, but i have never measured an exact time

This may also explain the small difference we seem to notice between medium and high for most of our lights, the consumption is above what its proclaimed to be giving us more lumens then we are expecting, and starts falling in current before high does, so medium to high is never an actual percent, it always changes spending on battery voltage even if high is running at full power (battery is over 3.6V under load) this seems a real mess to me





HKJ In low mode the light is never in regulation.
This would be the most used mode, the only possibility i can think of is to do a field test, or can someone come up with a calculation?






I have been using a Crelant V21A to get a feel for it before i leave it in the car for emergency usage, it is supposed to be putting out 420 lumens for 106 minutes, but i get an estimated 250-300 lumens for about 45 minutes on fully charged eneloops, again official numbers are not useful to have on hand because they are misleading. When i post my review for this light i will post some links with actual lumen tests that show its nowhere in the ballpark of 450 lumens on high

I also have no reliable numbers for an Olight S20, or Klarus P1A so it turns out i have no real numbers to work with at all.
Your thoughts and experiences?

2

I’ve been trying to test / estimate runtimes for my lights and can’t seem to get repeatable numbers. Hopefully we can get others to chime in.

3

I think the best solution is to just choose your preferred emergency light and perform a runtime test when there is no actual emergency going on.

Document your findings and store them in your emergeny kit with the lights.

Then publish your findings on BLF!

4

I object to your last step, why should i publish findings to a forum that has given me so much :stuck_out_tongue:

5

Just imagine if everyone did it! Would be a bit rough and ready due to variations in battery health, testing methodologies, etc but could really build into a useful database over time.

6

that was my thought as well, you would need battery capacity to determine an actual consumption average and tests on various capacities to determine what is causing the variation and to be able to extrapolate or have correct data for different batteries that are commonly used by us

7

I've found it tracks pretty well to the amps you measure at the tail in general. I got about 25 mins runtime from a 2.05A tail measurment on a SANYO 840 mAh cell for example. Works out about exact what you would calculate (25 mins at 2.05A = 833 mAh). There's a few factors effecting amp draw, though - varies by light/driver, etc.

8

These 3 Amp 7135 regulators are not so well behaved at the medium and low modes.
The 1 to 1.4 Amp max 7135 driver works better, but for maximum runtime and better regulation on medium and low, something like this is much better:

Shiningbeam used to sell these without the flashy modes as the ‘Perfect Regulator’.
Not much interest around here with the 1 Amp max but perfect for my long run time requirements. XM-L2 t6 still puts out 430+ lumen at 1 amp. Medium and low are well placed too.

9

If that were redesigned for 2.5 to 3 amps it would sell like hotcakes
i remember reading through that review and i liked the driver, but even 2 amps would be very nice, its regulation is very good

10

Then carry spares in a protective plastic box in your pocket

11

I favor lights with moonlights modes or pretty low modes since some light is better than no light in an emergency and you really don’t need that much light if it’s pitch black.

My 2 lumen Klarus ST20 can light up trees in a well lit street and it lasts far.

For the high modes you can do real tests to simulate whatever use you’ll be giving the lights, the low or moonlight modes are the ones that require theoretical calculations.

12

I have runtime graphs of quite a lot of lights to 50% (On rechargeables) and to shutoff on primary cells.

It should be possible to dig out that data and bung it on the spreadsheet linked in my sig.

Just don't hold your breath waiting for it - it will be a lot of work to add this.

13

For Klarus P1A (1XAA):

Nitecore MT1A (1XAA)

Fenix E11 (1XAA)

Rofis ER12 (1XAA)

Sunwayman C15A (1XAA)

Thrunite Neutron 1A (1XAA)

Xtar WK41 (1XAA)

Nitecore SRT3 (1XAA / 1X cr123a)

Sunwayman S10R (1X cr123a)

Sunwayman D20A (2XAA)

Solarforce Z2 (1XAA)

Spark SF5 NW (1XAA)

Foursevens ATOM AL (1X cr123a)

XTAR WK50 (1XAA)

LUMAPOWER LM21 XP-G2 (1X AAA)

14

I understand nothing of the voltage/amp graph that is always used… And I think it is infact meaningless because its actually more complicated.
I would like a very simple graph for the most popular lights using popular battery. Example :
XinTD C8 V4 : Runtime with Panasonic 3400 protected : Moonlight X hours : / low: X hours : / medium: X hours / high : X mins

Anyway for emergency use I think you should not be worry of the runtime because low mode can last for 2 days in many lights.

EDIT : I love the sleek simple graphs from Trevilux !

15

It used to be that Google docs were viewable just from the link. Google has obviously changed their policy since it became a sub-functon of Google Drive. Typical Google, put out something useful then change it to allow greater data grabbing. If you have some sort of Google account, you should be able to access it, if not I can bung it up on Dropbox or Copy or whatever.

At the moment it contains no runtime data but it is no big deal to add it if I've kept the original data logs - if not, I'll have to extrapolate from the jpg charts I still have. See here for an example (first one I found). https://budgetlightforum.com/t/-/189

16

The volt/amp chart shows how the led/driver works, together with discharge charts from a battery it is possible to calculate a runtime chart. I do not have time to play with it, but if anybody wish to try, they can send a email and ask for some driver/battery datasets (They are in CSV format).

What is missing from the volt/amp chart is heating effects, that depends on what light you month the driver in, how you hold the light, the temperature and how windy it is. All factors that are impossible to incorporate in a driver test.

17

Its a great idea, i wish the drivers had a constant draw then it would be an easy calculation because you will get different numbers based on battery charge level unfortunately :frowning:
If i get a light meter in future i will calculate runtimes with what i have, the popular 105c and qlite, i suppose it is possible to take HKJs driver and battery data and design an app that will calculate runtimes if you pick the driver and battery, but thats way beyond my skills.

If someone wants to help they could take a capacity tested or relatively new sanyo ur18650fm and panasonic 3400 and use a qlite or 105c light to measure runtime with a light meter on medium, low and moonlight, low and moonlight would have minimal heat output so the results would be pretty universal. That is a lot to ask so if anyone steps up to the plate who has these items and the time and inclination that would be extremely awesome!