Flashlight with extreme battery endurance??

37 posts / 0 new
Last post
aswang
Offline
Last seen: 1 day 22 hours ago
Joined: 07/17/2018 - 07:53
Posts: 280
Location: Philippines

Under $50 I think skilhunt H03/H04 or the Sofirn SP32A are one of the more pound for pound efficient ones. I also find multiple low mode options being more important than efficiency alone when trying to extend battery life. The Skilhunt’s 0.5, 5, 20 and 70 lumen modes and the SP32A’s ramping has that covered. The skilhunt has the advantage since it can double as a headlamp.

Multiple emitter Anduril lights that have a channel that can run on a single 7135 chip (e.g D4, D4v2, Fw3A, MF01 mini or maybe the new HL3A) can also be very efficient if you run them on that single 7135 channel. Usually on default (7) stepped mode they’re the first 4 modes (up to 120 or so lumens). You can set the ramping ceiling to this level and not use turbo and a MF01 mini could run a very very long time on a single 26650.

Also, consider a thrower if you need long distance lighting since they should be way more efficient for this than any super efficient floody light.

ggf31416
Offline
Last seen: 10 hours 47 min ago
Joined: 02/25/2016 - 17:38
Posts: 339
Location: Uruguay

XP-L2 or SST-20 have a Vf of 2.7-2.8V at around 100lm, so a linear driver in the best case only has 75% average efficiency. If PMW is used the LED is actually run at a higher current, which it’s less efficient.

aswang
Offline
Last seen: 1 day 22 hours ago
Joined: 07/17/2018 - 07:53
Posts: 280
Location: Philippines
ggf31416 wrote:
XP-L2 or SST-20 have a Vf of 2.7-2.8V at around 100lm, so a linear driver in the best case only has 75% average efficiency. If PMW is used the LED is actually run at a higher current, which it’s less efficient.

If you look at numbers, yes they don’t have the best efficiency. But in real life, I find that having the ability to adjust brightness to what is needed (especially moonlight and good low mode spacing) is much more important than just sheer % efficiency numbers when looking at runtimes.

As an example, I have a nitecore hc30 which uses a relatively efficient buck boost driver vs my modded H03 with Lexel’s FET+7+1 TA driver. If you look at numbers alone yes the nitecore is more efficient. But during camp time, for instance, I actually only need 3-20 lumens most of the time. But the HC30 only has 1 and 70 lumen mode so you end up using the 70 lumen mode all the time. For my H03 I just ramp to the minimum light I need. With the same led, 20 lumens at 70% efficiency eats way less power than 70 lumens at 90% efficiency. When going to the loo or needing light with night adjusted eyes and you only need 0.1 lumens (or even lower), you have to use 1 lumen on the nitecore. Etc, etc. So I’d argue that my skilhunt H03 is way more efficient than my hc30 for my use case.

I think the main disadvantage of having FET driven light is its very difficult to resist the urge to go Turbo for no good reason.

Lightbringer
Lightbringer's picture
Offline
Last seen: 3 hours 9 min ago
Joined: 08/30/2016 - 14:12
Posts: 10982
Location: nyc
BrightLight wrote:
A brand that lies with their specs is for instance Black Diamond with their Icon headlamp claiming 500 lumens measured with ANSI method for 70 hours on 4 x AA batteries.

ANSI spex calculate runtime to the point where light output is down to I think 10% of its initial value or value at 1min operation, something like that. It’s geared towards hotwire bulbs, when that’d likely be considered “dead”. (Ie, burn off the initial voltage peak and run down ‘til “dead”.)

So a 500lm light that runs ‘til it’s down to 50lm is the case in question. So some mfrs of LED lights will have a drop after a few minutes, then just languish at about 11% output forever, ‘til the cell can’t even maintain that, then finally drop to 10%. Call it the Volkswagen method of runtime calculation.

So no way will it sustain 500lm for 70hrs or anything close to that.

09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0

Lightbringer
Lightbringer's picture
Offline
Last seen: 3 hours 9 min ago
Joined: 08/30/2016 - 14:12
Posts: 10982
Location: nyc

Whups, just saw Pete’s post.

Okay, so 30sec, not 1min. Close enough. LOL

09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0

Agro
Agro's picture
Online
Last seen: 13 min 34 sec ago
Joined: 05/14/2017 - 11:16
Posts: 5944
Location: Ślōnsk
aswang wrote:
ggf31416 wrote:
XP-L2 or SST-20 have a Vf of 2.7-2.8V at around 100lm, so a linear driver in the best case only has 75% average efficiency. If PMW is used the LED is actually run at a higher current, which it’s less efficient.

If you look at numbers, yes they don’t have the best efficiency. But in real life, I find that having the ability to adjust brightness to what is needed (especially moonlight and good low mode spacing) is much more important than just sheer % efficiency numbers when looking at runtimes.

As an example, I have a nitecore hc30 which uses a relatively efficient buck boost driver vs my modded H03 with Lexel’s FET+7+1 TA driver. If you look at numbers alone yes the nitecore is more efficient. But during camp time, for instance, I actually only need 3-20 lumens most of the time. But the HC30 only has 1 and 70 lumen mode so you end up using the 70 lumen mode all the time. For my H03 I just ramp to the minimum light I need. With the same led, 20 lumens at 70% efficiency eats way less power than 70 lumens at 90% efficiency. When going to the loo or needing light with night adjusted eyes and you only need 0.1 lumens (or even lower), you have to use 1 lumen on the nitecore. Etc, etc. So I’d argue that my skilhunt H03 is way more efficient than my hc30 for my use case.

I think the main disadvantage of having FET driven light is its very difficult to resist the urge to go Turbo for no good reason.


I second that. I got really a lot of walktime from my Peak Eiger AAA – precisely because the ramping UI allowed me to adjust the output to the min. I used more at dusk and ramped down as the night set in.
I so wish there were efficient buck or boost drivers with BLF firmware available…

Also, when it comes to runtime-to-weight, 18650 with spare cells is going to win over 21700, 26650 or multi-cell lights. Though 21700 will not be much heavier and will save you some cell swaps.

Pages