Cold weather finally hit Finland last week and today we had cold temperatures down to -26 °C (~ -14 °F). As it hasn't been snowing I saw my opportunity to go ice skating on a lake and to test my flashlights in the cold. I packed the lights in a pouch outside my backpack to make sure they'd be truly cold when I would turn them on. I skated for an hour and sat on a pier to test the lights. None of the batteries were recently recharged to simulate actual circumstances. All of my li-ions are protected. Here are my notions after quick testing:
18650:
Convoy S2+ (7135*3) with Keeppower 3200 mAh, 4.13 V: No notable change in use. The driver doesn't draw much current even on high mode so the result wasn't a surprise.
Oligh M22 with Panasonic NCR18650B 3400 mAh, 4.00 V: M22 lost its high mode (no notable difference to mid). The twist-the-head-back-and-forth UI didn't work well in the cold as the lube and O-rings started to freeze. I've noticed this problem last year too while using it when bicycling in cold but it has been able to reach its high mode in warmer temperatures.
Thrunite TN12 (CW) with NCR18650 (grey). 4.11 V: TN12 with a battery from an old laptop in it had only slighty notable difference between mid and high. The difference between high and turbo was not notable. The result isn't surprising as the battery is one of the weakest I own and is meant for backup purpose only.
Thrunite TN12 (NW) with Panasonic NCR18650B 3400 mAh, 3.75 V: I carried this light in my pocket so it was a little bit warmer than the others. I forgot to test it on the lake but it hasn't had problems maintining its med or med-high mode while cycling in the cold.
Armytek Predator Pro v2.5 (warm) with Keeppower 3200 mAh, 4.13 V: No notable change in use. The twisty head was working better than in M22.
Armytek Wizard Pro v1.5 with Keeppower 3400 mAh, 4.01 V: Wizard had no problem maintaining it's low, med or high modes while I was skating, but I noticed the turbo mode had become actually lower than the high mode after using it a while on low in the beginning.
14500 / AA:
Eagtac GX25A3 with 3 x Sanyo Eneloop 1900 mAh, 1.29V: The high mode seemed to become lower than in room temperature. I guess it's direct drive so the result shouldn't be a surprise. No problems with the twisty UI.
Armytek Tiara A1 Pro v1 with Keeppower 840 mAh, 4.07 V: No notable change in use. The headlamp was able to maintain its med and high modes while I was skating. Turbo was notably brighter compared to high.
with Sanyo Eneloop 1900 mAh, 1.30 V: No notable change in use. The headlamp was able to maintain its med and high modes while I was skating. Turbo seemed to be even brighter than with the li-ion.
Other:
Nitecore Tube with its li-ion: I attached Tube to my backpack and turned it on low when I started skating. I had used it on low for 2 hours in cold earlier and wasn't sure if it would even work. Surprisingly the light worked the whole time I was out.
Motorola Moto G with its li-ion: I carried my phone in my pocket with TN12. It felt really cold in the hand and the buttons on the rubber protection cover had freezed, but the touch screen worked fine and I was able to update my location to my roommates.
Summary:
Edit:
I measured the voltage on batteries once they were warmed up again and added the voltage readings on this post. I think it might be more interesting to test the readings in cold, but my hands were really freezing. Maybe next time.
18650:
Convoy S2+ (7135*3) with Keeppower 3200 mAh, 4.13 V: No notable change in use. The driver doesn't draw much current even on high mode so the result wasn't a surprise.
Oligh M22 with Panasonic NCR18650B 3400 mAh, 4.00 V: M22 lost its high mode (no notable difference to mid). The twist-the-head-back-and-forth UI didn't work well in the cold as the lube and O-rings started to freeze. I've noticed this problem last year too while using it when bicycling in cold but it has been able to reach its high mode in warmer temperatures.
Thrunite TN12 (CW) with NCR18650 (grey). 4.11 V: TN12 with a battery from an old laptop in it had only slighty notable difference between mid and high. The difference between high and turbo was not notable. The result isn't surprising as the battery is one of the weakest I own and is meant for backup purpose only.
Thrunite TN12 (NW) with Panasonic NCR18650B 3400 mAh, 3.75 V: I carried this light in my pocket so it was a little bit warmer than the others. I forgot to test it on the lake but it hasn't had problems maintining its med or med-high mode while cycling in the cold.
Armytek Predator Pro v2.5 (warm) with Keeppower 3200 mAh, 4.13 V: No notable change in use. The twisty head was working better than in M22.
Armytek Wizard Pro v1.5 with Keeppower 3400 mAh, 4.01 V: Wizard had no problem maintaining it's low, med or high modes while I was skating, but I noticed the turbo mode had become actually lower than the high mode after using it a while on low in the beginning.
14500 / AA:
Eagtac GX25A3 with 3 x Sanyo Eneloop 1900 mAh, 1.29V: The high mode seemed to become lower than in room temperature. I guess it's direct drive so the result shouldn't be a surprise. No problems with the twisty UI.
Armytek Tiara A1 Pro v1 with Keeppower 840 mAh, 4.07 V: No notable change in use. The headlamp was able to maintain its med and high modes while I was skating. Turbo was notably brighter compared to high.
with Sanyo Eneloop 1900 mAh, 1.30 V: No notable change in use. The headlamp was able to maintain its med and high modes while I was skating. Turbo seemed to be even brighter than with the li-ion.
Other:
Nitecore Tube with its li-ion: I attached Tube to my backpack and turned it on low when I started skating. I had used it on low for 2 hours in cold earlier and wasn't sure if it would even work. Surprisingly the light worked the whole time I was out.
Motorola Moto G with its li-ion: I carried my phone in my pocket with TN12. It felt really cold in the hand and the buttons on the rubber protection cover had freezed, but the touch screen worked fine and I was able to update my location to my roommates.
Summary:
- Note that this test is more qualitative than quantitative
- All of the lights turned on and could reach their med mode
- I didn't notice any flickering or unexpected behaviour (except Wizard's lower turbo mode)
- None of the protections circuits on the batteries cut the current due to low voltage even though the lights were used on their med/high modes
- I'm not sure if 1 hour is enough to freeze the batteries, but my Keeppowers, Panasonics and Eneloops listed above seem to be able to work in cold
- Except for Tiara and Wizard, the lights were on max. 1-2 minutes. Longer runtime tests may be required to see the difference in battery behaviour.
Edit:
I measured the voltage on batteries once they were warmed up again and added the voltage readings on this post. I think it might be more interesting to test the readings in cold, but my hands were really freezing. Maybe next time.
Last edited: