Low Temperature Runtime Comparison

saildude · Dec 18, 2007

Here is a Low Temp vs. Room Temp Runtime

We have had several threads about what type of batteries to use for low temperature operations. I did a system comparison light and batteries.

Streamlight 4AA Luxon 2000 MHr Eneloop Cells

Room Temp about 70 Degree F
Freezer Temp about 4 Degrees F
Meter Extech 401036 - Sample interval 10 min.
Light box milk carton similar to that at Flashlightreviews.com

I did one run at room temp. I then put 4 charged Eneloop cells and the Streamlight 4AA Luxon in the freezer ( the cells were separate from the light in a plastic bag to let them cool a bit faster ) initial soak time was a bit over 4 hours - I then put the cells in the light and and let it soak for another hour+++. The light box and light sensor were also soaked in the cold temp for about 4 hours. Light meter display was outside the freezer.

Results:

Freezer Run Time about 17% shorter
Freezer Light output about 15% higher

I hope this helps you with your battery & light selection.

mb

Streamlight%204AA%20LUX%20-%20Temperature%20Variations%20-%20700.png

soffiler · Dec 18, 2007

saildude, is the Extech 401036 sensor rated down to 3 deg F? Clearly it still *worked* but was it accurate?

With that said, it certainly makes sense to have more output because the freezer is keeping the LED very cool. And shorter runtime makes sense because the chemical reaction in the NiMH is less efficient when cold.

Size15's · Dec 18, 2007

Interesting... Are you able to measure the 'areas' under each line?
I did a rough'n'ready measurement and discovered that there's only 1% difference between the area (the freezer being 1% less)

Based on this I don't see the freezer making the flashlight any more/less efficient - it comes down to whether you'd want higher output or longer runtime.
The trade-off for higher output is the time and energy taken to freeze the flashlight.

Or I may be way off and happy to be stood corrected?

Kiessling · Dec 18, 2007

McGizmo did some testing about this some time ago and found out that intermittent or cooled runtime curves sported significantly less runtime than "hot" ones, contrary to popular belief.
We thought it might have something to do with the battery chemistry working better once warm and that this eff4ect overshadowed the diode being less efficient when warm.

If it is of interest I can search for the thread.

bernie

meuge · Dec 18, 2007

I wonder though, whether the draw on the batteries would cause them to heat up.

That's why the diode efficiency wouldn't suffer - I doubt it's at ambient.

The Streamlight is a plastic light, so it wouldn't act as a very good heatsink. If the batteries inside heated up to even 0C, they'd still be well within their operating window.

abvidledUK · Dec 18, 2007

In true Blue Peter style, here's one I did earlier.

https://www.candlepowerforums.com/threads/111218&highlight=nimh+temperature

What surprised me most, with my 2,000mah nimh's, was that Frozen torch and battery (-15ºc) gave virtually identical results, with the Room temperature (20ºc) Control.

Also, the fact that at fridge temperature 4ºc, the capacity appeared to be less.

saildude · Dec 18, 2007

No I don't have a good way ( my math is real rusty ) to do the area under the curves. The Data Log has been added below.

ExTech rates the system to 32 degrees F - so their may be something about the sensor.

I did this because while I live in a mild climate area I do travel east of the mountains from time to time where it is cold and was wondering if my light would work or if I needed to have extra Lithium Cells for the cold weather. This looks to me like I am good to go at least down to 0 Degree F.

Range:20000 Unit:LUX
10 Min Sample Rate

- -- Room - Freezer

1 -- 3770 - 4320
2 -- 3240 - 3600
3 -- 3080 - 3440
4 -- 3000 - 3390
5 -- 2950 - 3370
6 -- 2960 - 3370
7 -- 2950 - 3360
9 -- 2920 - 3330
10 - 2910 - 3300
11 - 2920 - 3330
12 - 2930 - 3350
13 - 2910 - 3350
14 - 2910 - 3310
15 - 2910 - 3280
16 - 2930 - 3310
17 - 2930 - 3330
18 - 2910 - 3320
19 - 2910 - 3290
20 - 2900 - 3290
21 - 2930 - 3310
22 - 2910 - 3300
23 - 2890 - 2730
24 - 2890 - 1330
25 - 2890 - 820
26 - 2900 - 550
27 - 2880 - 390
28 - 2850 - 290
29 - 1240 - 240
30 - 790
31 - 600
32 - 480
33 - 400
34 - 350
35 - 310
36 - 290
37 - 260
38 - 250

abvidledUK · Dec 18, 2007

Size15's said:
Interesting... Are you able to measure the 'areas' under each line?
I did a rough'n'ready measurement and discovered that there's only 1% difference between the area (the freezer being 1% less)

I got around 3.2% difference.

soffiler · Dec 18, 2007

Size15's said:
Interesting... Are you able to measure the 'areas' under each line?
I did a rough'n'ready measurement and discovered that there's only 1% difference between the area (the freezer being 1% less)

Based on this I don't see the freezer making the flashlight any more/less efficient...

The freezer increases the LED efficacy - lumens per watt. Cooler LED puts out more lumens for the same electrical input. At the same time, the cooler temperature is decreasing the efficiency of the chemical reaction that makes power in a NiMH. It appears from your rough'n'ready measurement that these two effects are coming very close to cancelling each other out.

saildude · Dec 18, 2007

Data Points added above

Illum · Dec 18, 2007

Kiessling said:
McGizmo did some testing about this some time ago and found out that intermittent or cooled runtime curves sported significantly less runtime than "hot" ones, contrary to popular belief.
We thought it might have something to do with the battery chemistry working better once warm and that this eff4ect overshadowed the diode being less efficient when warm.

If it is of interest I can search for the thread.

bernie

Gizmo also noted that the batteries desired warmer climates to operate efficiently while the LED/driver assembly would prefer much cooler environments....nice graph you got there!
as for the link to this concept: A paradox to thermal relief??!?!?
here ya go :popcorn:

soffiler · Dec 18, 2007

saildude said:
No I don't have a good way ( my math is real rusty ) to do the area under the curves. The Data Log has been added below.

ExTech rates the system to 32 degrees F - so their may be something about the sensor.

I did this because while I live in a mild climate area I do travel east of the mountains from time to time where it is cold and was wondering if my light would work or if I needed to have extra Lithium Cells for the cold weather. This looks to me like I am good to go at least down to 0 Degree F.

Range:20000 Unit:LUX
10 Min Sample Rate

- -- Room - Freezer

1 -- 3770 - 4320
2 -- 3240 - 3600
3 -- 3080 - 3440
4 -- 3000 - 3390
5 -- 2950 - 3370
6 -- 2960 - 3370
7 -- 2950 - 3360
9 -- 2920 - 3330
10 - 2910 - 3300
11 - 2920 - 3330
12 - 2930 - 3350
13 - 2910 - 3350
14 - 2910 - 3310
15 - 2910 - 3280
16 - 2930 - 3310
17 - 2930 - 3330
18 - 2910 - 3320
19 - 2910 - 3290
20 - 2900 - 3290
21 - 2930 - 3310
22 - 2910 - 3300
23 - 2890 - 2730
24 - 2890 - 1330
25 - 2890 - 820
26 - 2900 - 550
27 - 2880 - 390
28 - 2850 - 290
29 - 1240 - 240
30 - 790
31 - 600
32 - 480
33 - 400
34 - 350
35 - 310
36 - 290
37 - 260
38 - 250

By piecewise integration (with the help of Excel) the area under the two curves is equal at about 255 minutes. The freezer curve wins prior to this. At 290 minutes, the room-temp curve has 23% more area and at 380 minutes it has 36% more area.

Size15's · Dec 18, 2007

That's quite a large difference between mine (1%) & abvidledUK's (3%) [similar in my book] compared with soffiler's 36%.

Low Temperature Runtime Comparison

saildude

Newly Enlightened

soffiler

Enlightened

Size15's

Flashaholic

Kiessling

Flashaholic

meuge

Enlightened

abvidledUK

Flashlight Enthusiast

saildude

Newly Enlightened

abvidledUK

Flashlight Enthusiast

soffiler

Enlightened

saildude

Newly Enlightened

Illum

Flashaholic

soffiler

Enlightened

Size15's

Flashaholic

Similar threads

Latest posts