LED Flashlights: Output power vs. Run Time.

I know a lot of us here love to modify lights to get as much photons out of them that is possible,
and i have some 150 lights in my growing collection, and many of them modded, some under construction, and some to be modded.
i noticed however that with many of your modded lights and the ones i have modded, for the quest to squeeze as many lumens as possible from the lights the run time becomes ridiculously low, along with the problem if we run some of these lights for longer than 5 minutes continuously we can cook a meal on them faster than a Coleman Stove or risk a fusion nuclear melt-down & third-degree burns.
What about those of us who use lights for Search & Rescue, Firefighting, night fishing, long multi-night camping trips, Canoe trips, night Geocachers, Expeditions, night security, and so on who need a good balance between output and a long run time.
I have searched for, and even modded some lights for extended run-time just for these uses.

What is your favorite stock light with the best run time, and still have a good output enough for actual use outdoors at night ?

For me i have several stock & modded lights i use a lot for these purposes listed above, some of my “used often” lights are:

For stock lights: - a Nitecore P25, Olight S20, ( both have Moonlight modes) in the single 18650 class, a stock SRK Kung on Low mode, a TK35 Fenix in the multi-18650 class, a Thrunite Ti2 for AAA.

For my modded lights: - a two-18650 in parallel top-handled light with a XM-L, N/W swap, a Trustfire Zoomie with XM-L & Qlite/Revision nanjg 105C driver and single 18650, a modded Roche F12 with the Qlite-Revision nanjg 105C driver & modded to use a protected 18650 cell, a modded Sipik SK68 (original) modded with the 20-mode 7880 driver & XM-L N/W for AA class, a 2-AA side-switch custom SK68 with N/W XR-E, and various other small lights for ambient lighting that have a running time of over a week continuously on a set of cells.

It’s a good question. I have several “prepper” friends that have asked me to mod lights for longer run time. The Q-lite driver with the 5% is a good one for that. They usually want small bug out lights that don’t take up a lot of space or weight. I have done several single 26650. This is a good one for that and I kept one and use it all the time. In a pinch you can use an adapter for AAA.

http://www.aliexpress.com/item/Super-High-Power-LED-Flashlight-Cree-XM-L-U2-1800-Lumens-5-Mode-Portable-LED-Torch/1607885240.html

This is my next project for myself more of a lark really but fun. It has a USB interface 26650. It’s getting a XP-G2 on copper and It’s going to have an Electronic Cigarette attachment. So it will charge a cell phone, run as a decent flashlight and work as an E-Cig too. Most likely it will also get a reflector swap and a Q-lite driver. It also will run AAA

http://www.aliexpress.com/item/Q5-LED-300-Lms-Flashlight-Torch-USB-charge-26650-18650-AAA/1472152241.html

As far as stock goes I’m not sure I know what a stock light is anymore. :wink:

I am assuming that when a modder increases wire size from the emitter to the driver or solders copper strapping to the springs he is looking to get more current flow. Does this decrease run time as well as increase the brightness of the light?

Yep but a lot depends on the modes and the driver. But increased current will sap your batteries faster. Upgrading to better more efficient emitters will get you more brightness without sapping your batteries or allow you to remain the same with longer run times depending on the driver.

The best lights are the one that can do it all. Just because many make hot-rods does not mean that runtime have to suck. There are other modes than high. Its all about having the right driver circuit with the proper modes.

In a light with say 2 batteries in series and a buck driver, copper braiding the springs will minimize voltage sag. Higher voltage to a buck driver means it needs less amps in order to deliver the correct emitter current. So it will actually improve runtime in certain lights.

In typical single emitter/single cell lights with a regulated linear driver it will just make sure you stay within regulation longer. Which to most people are desirable..

In certain simple (usually crappy) "direct drive lights" it will increase the output on all modes. This is also typical with resistor mods. If the output on the lower levels become higher, you will naturally get less runtime.

With a decent driver, copper braided springs and improving wires does not have any downsides in terms of runtime assuming you use the correct lower modes.

For people who want runtime, get a good driver with suitable modes and good efficiency. Get the most efficient emitter, mounted on a proper copper mcpcb. AR lens, well centered emitter, copper braided springs, etc.. The usual stuff most modders does are beneficial to getting the highest amount of lumen out of a flashlight, in many cases, even in the lower modes.

The simplest thing is just to bring a spare battery (or set of batteries) and stick with good high capacity cells.

im assuming your referring to the same “Qlite-Revision” version of the Nanjg 105c , (the one with the 4-modes, Moonlight, Low,medium,High) which is the one i have modded in so many of my lights because of the greart run times it has on the Moonlight,Low, and even medium modes. Also if im correct it has the 3.0 volt shut down or warning protection for Li-ions built in ?
i am planing on building a maximum-runtime light, using possibly the four-18650-in-parallel body of another SRK, a nanjg/QliteRevision driver, and a single XM-L2 NW head for a extended runtime light. (after i finish the 4-DST/SRK super thrower)

a good efficient driver is a reason why i like my P25, it has great run time for a single-18650 light, and with the XM-L2 U2 NW mod its still plenty powerful on medium and low mode, with the moonlight that will run continously for weeks.

Yep that’s a good driver for run time and yep it has low voltage warning. You would sure have all kinds of run-time with the setup you are thinking about.

Are the newer XM-L2s brighter and more efficient than the XM-L?

I have an old short “lantern” style 4D light that I converted first to krypton incandescent and then to led with drop in bulbs, as well as adding glow in the dark paint in the grooves. I remember days long power outages in Mountain View, CA and have read that the power companies are operating the grid for economy rather than reliability.
Dividing 12 AH from Wikipedia by 0.23 A that I measure across the switch, I get 52 hours. Running it a third of the time, it should be good for almost a week.

Yes but it’s kind of an either or. If you put the same amps to it it’s brighter. But you can put less amps to it and get the same brightness. If you put the same amps to it you are still using the same amount of battery but you don’t have to put as much so it’s more efficient.

Interesting discussion. Thanks.

that reminds me of a 4-D lanter/flashlight made by Dorcy/Rayovac, like this one: >> http://www.dorcy.com/images/product/large/109.jpg
i had one of these for 15 years or more, and modded it too with a LED a fiew years ago, ( salvaging a Luxeon Star & resistor from a Garrity LED spotlight to swap into it) and now on four D Alcalines in will run for over a week continously drawing only 0.15 amps, and is twice as bright as the orginal old krypton incan bulb.

One I’m coming to appreciate more as I use it is a Convoy L1.

As much as I like the Convoy ‘S’ lights, this one has more of everything - more throw, very good spill, and plenty of run time. For some reason it seems to be discontinued; I can’t imagine why - 2 x 18650s behind an LD29 driver and a T6-3c emitter works really well.

To be picky, I think the “efficiency” is defined as the lumens per power, so the small difference in forward voltage also comes in. Is that right?

Yes I think so anyway. Someone who understands it better might be able to chime in and explain that part better.

As long as an XM-L / XM-L2 is being regulated through a linear driver, which is the most common setup. The extra voltage that the XM-L does not "use" compared to XM-L2 is just burnt off by the driver. Runtime is calculated based on mAh on the battery and amps to the emitter. Only difference is that XM-L2 will more easily drop out of regulation due to slighly higher Vf demand when battery voltage declines. This will usually mean that you get better runtime and similar output as XM-L when the XM-L2 is out of regulation, and similar runtime and more light when they are regulated at the same amps.

In a light with a regulated buck driver the XM-L2 will require slightly more energy/power at similar amps. (Im not sure if there is much difference at lower amps btw) So runtime should become marginally lower compared to XM-L. But no matter how you look at it, XM-L2 is more energy efficient.

Here is an example of the balance of long run time and a usable amount of light for camping, close-proximity use, etc.

This is one of those small Dorcy 4-AA lights,
(The adjustable from beam to lantern mode model.)
It formerly had a Krypton Incandescent bulb.
with the incan it ran roughly 3 to 4 hours on a new set of Energizers to about 50 % flat, and in 6 hours it was dimmer than a birthday cake candle burning out and cell voltage was around .6 volts.
So I used a 1 watt warm white ( large-surface emitting area) generic LED, replaced the filament in the incan bulb with the LED, (with no heatsink what so ever, no star, just the bare LED !)
I did some testing with many different resistors, (using no driver circuit.)
and eventually worked it out to a balance, & added two 1/4 watt, 56 ohm resistors in series in the circuit to regulate the LED to a brightness level just under where it generates almost no heat, and can run with no heatsink or no star suspended in air just by the incan filament wires, but still seem brighter than the original Krypton bulb was with new Energizer AAs, …and measured a load amps of only 0.05 amps with the four new AA alcalines, then after a fiew seconds that dropped to 0.03 avarage.
I ran this light over a full WEEK continously, and at the end of the week it was still as bright as the orginal Incan was on new cells, and i measured each Energizer voltage to be still at 1.39 volts each.

There seems to be fine line between run-time efficency in most regular Light Emitting Diodes, and when that efficency quickly drops drastically for that given extra light output when its pushed to the level and beyond, when it begins to produce alot of heat that it requires a heatsink or it will burn out.
I have many various mini-lanterns/lights with this modification set-up, that we would turn on at campsites for ambient lighting outside at the start of 3 – 4 day camping trips, and they never get turned off for the entire trip, not even in the day.
(some of them have been used for several camping trips during the summer and on the same set of alcaline cells all summer.

This photo below is of the Dorcy light with over 40 hours running time on the set of Alcalines in it. ( its still brighter than my stock Tank007 E09 with a diffuser-cap on high.)

I got a Spark SX5 Headlamp for that purpose

It uses about every battery out there, so I can switch out sells from any other device and use them on it.