Battery protection circuit - why is this T400 battery dead?

[Zak]

The battery in my girlfriend's T400 diminished in capacity over time and eventually went completely dead. It provided zero runtime, and I could not measure any voltage with my multimeter. It's my understanding that the non-resettable protection circuit trips in the event one or more cells over-discharges, as over-discharged cells have a high probability of developing internal shorts, and continuing to use them can lead to fire or explosion.

I disassembled the battery and removed all the cells. Every cell was between 3.8 and 3.8 volts. It's my understanding that a fairly conservative lower limit for discharging lithium-ion cells is 3.2 volts.

I suspected the cells might simply have a very low energy capacity. I charged a pair to 4.2 volts in an external charger and put one in my flashlight. While this isn't a scientifically instrumented test, the flashlight's manual rates runtime on high at 1 hour, 45 minutes on a 2600 mAh cell. The cells in a T400 battery should be 2600 mAh. I noticed the flashlight's brightness decreasing gradually after an hour, and stopped it at 1:45. The cell read just over 3.2 volts.

I'm not sure I'll repeat the capacity test with all the cells, but my initial test here suggests they're fine. They weren't over-discharged, and they still hold their rated capacity, which is actually much better than I expected. Given that, why did this battery go (over the course of a couple years) from lasting hours to lasting minutes and finally not working at all?

[rkawakami]

It's possible that one of the other cells you have not checked yet is at the end of its life despite the voltage measurements (the real test is how much current it can supply over time). It's also possible that the firmware inside the battery is set to a conservative level in order to prevent "bad things" from happening. It's also possible that the factory does this to sell more batteries .

[Zak]

I suppose I could stick every cell in my flashlight and see if one of them fails to last at least 90 minutes. If they all pass, I might try to jump start the protection circuit, though I have no idea whether that works on batteries other than the T4x battery shown in the video.

[Zak]

I tested all of the cells. The worst of them got to 4.1 volts when fully charged. All read right around 3.6 after 90 minutes in the flashlight. A known-good ~3000 mAh protected 18650 read a little over 3.6 volts after 115 minutes, so this seems pretty much in line with 2400-2600 mAh from the Thinkpad cells. None of them got more than warm to the touch during charging.

[Zak]

One of the cells got uncomfortably hot to the touch the second time I charged it. If anything like that happened inside the battery, it seems reasonable that a protection circuit would conclude the cell is no longer safe.

[rkawakami]

There should be a thermistor-type device that's laid between/atop the cells inside the pack. If it detects an over-temperature condition it is supposed to notify the intelligent controller of that fact. That could have been the cause of the pack rapidly going "dead".