I received some Samsung 21700 40T cells from 18650batterystore and they arrived reading 3.43v. I charged one and it took 3600ma before reading full.
I don’t have a setup to perform high current discharge tests, but it doesn’t seem to perform better than generic rewraps that come with the lights. After a few cycles on turbo it was down 1600ma (this could very well be legit).
What do you all think? Should I call 18650batterystore, or are these cells fine? 3.43v seems very low, and 3600ma for a cell rated at a nominal 4000mah also seems extremely low to me, but I have no experience with these particular cells, so maybe this is fine/normal?
Thanks, I’m used to getting cells at closer to 3.6v and 50% capacity for storage. My understanding was that cells should be kept between 20-80% capacity for maximum life, and ~400ma is closer to 10% capacity (assuming the cell isn’t exactly 4000mah), so I was a bit concerned. I’ll keep an eye on them and see how they hold up.
That would entirely depend on the driver used in each flashlight. If it’s FET driven, then the output is limited by the cell, but if it’s limited by the current, then you may not see any increase at all.
To test the capacity you require an analyzing function on a charger or use a CBA. This needs to discharge down to 2.5v (most don’t) with a load of roughly 0.2C, the C being the capacity (0.2C x 4000 = 800ma discharge current)
I will run it through my analyzing charger and if it’s at full capacity I will assume it’s ok. The light I tested was a FET driven SBT90.2, so I would expect the 40T to dump a lot more current than a the Wurkkos branded cell that came with it.
Lower storage voltage is better for the cell all the way down to 2.8V. Pretty much the only reason batteries are not stored at this voltage is that self discharge can bring it to below safe levels.
Not necessarily, sometimes cheap FETs perform terribly and don’t offer that much of an increase. You should also bypass the springs.
Download a light/lux meter app on your phone. In a dark room point the flashlight at the ceiling and sit your phone on the desk beside it, try each cell to see if there’s an increase. For something to look twice as bright it needs to be 3-4x the lumens, so it maybe brighter, but without a second flashlight to directly compare you may just not notice.
I was trying to capture the difference using an Opple, I know we perceive brightness logarithmically. Unfortunately the Opple has a very slow slow update time and the light heats up very quickly. I have considered bypassing the spring, currently it is double sprung, but a bypass may still be in the cards.
I am currently laid out with a bad back, and just got a new puppy so I probably won’t be able to test for a few days, but next time I will try a ceiling bounce app. Good advice though. The light in question is a Wurkkos TS30S, I have been chatting with another user here who was seeing roughly 15% uplift with the 40T, so assuming they components are consistent, I don’t think the quality of the FET is a limiting factor.
Ultimately the brightness gain is not a practical issue due to how quickly the light heats up and how much current a DD SBT90.2 can pull. I did want to make sure that the cells weren’t an issue because there is a limited window I could deal with that, and a. Cell with poor health can cause other problems (like not being safe).