This is one of my ongoing experiments to see how a rather no name power bank ages with charge/discharge cycles, and it’s been quite surprising all the way. This is the 18th installment, meaning that we have reached a total of 900 cycles since the commencement of the experiment, or 916 cycles since the power bank left the factory. It’s been quite an effort to keep taking care of it and analyzing the data, but now that we’ve gotten so far, it would be a waste to just stop it.
The latest 50 cycles seems to show a further downward trending of available capacity, which is consistent with the trend established from cycle ~775 to 825. The reversal at that point was not prolonged, and may just be a sign of variation due to temperature. There appears to be some indication that the degradation rate might be accelerating at this point, but it’s not clear. Capacity low is just above 2650mAh, so we are about 67% of original capacity.
At this point, the power bank still delivers more than you would obtain from a “lipstick-style” single-18650-cell type power bank, so it is still functional, but it’s likely that a user might notice the degradation in capacity. Surprisingly, while the cell holds up rather well, I anticipate that the connectors (especially the microUSB-B charging port) may not. There is some increased “slop” in the connector fit, probably due to wear from plug/unplug cycles on both the power bank and the cable I’m using with it. While it has not impeded charging or discharging at the present time, they could potentially fail at any time into the future.
One interesting observation seems to suggest the capacity degradation is internal-resistance related, as noted by some other sources. When the power bank has finished its discharge cycle, connecting it to the charger presents a “voltage based” charge status estimation. Previously, the power bank would spend a short time at 1/4-LEDs indication before rising up to 2/4-LEDs. Now, when plugged in to charge, it immediately goes to 2/4-LEDs, suggesting the voltage of the cell rises significantly upon application of charge current.
At this point, the unit has reached three-times the 300-cycle “low point” estimated life and has about 67% of the initial capacity. The unit has quite good endurance, it can be said, which was a big surprise. The connectors may prove to be another point of failure of the whole assembly, although this remains to be seen. The experiment will continue for another 50-cycles.