In this ongoing experiment which started back in late September here, I’ve been diligently putting a powerbank through controlled charge-discharge cycling while measuring the effective capacity based on a methodology I developed to measure the real capacity of power banks. All the care wherever possible was put in to try and ensure the experiment would be able to show the “rate” of degradation of the unit over a decent number of cycles, given that this was all done by hand, it’s a bit of an endurance test for myself as well.
The testing was put on hold while I was overseas in Hong Kong and China over summer for a conference, and resumed as soon as I got home. I’m glad to report that we have now reached 250 cycles in the test regime, for a total of 256 cycles from new. This represents a lifetime cycle consumption of about 51 to 85% based on the often quoted 300-500 cycle lifetime for a lithium-polymer based cell.
Due to a loss of power, data for experimental cycle 233 was lost, making the amount of lost and anomalous data being 3 in 250 (1.2%).
How did it fare?
Surprisingly well. It seems the rate of degradation seems to be marginally higher, resuming the earlier linear trend somewhat. Based on the linear model, the cell would be good for 1379 cycles, with the polynomial model predicting a more reasonable 865 cycles. The last measured capacity was 3705mAh, which is a reduction of 169mAh from the fresh measurement – this is still not much and is only marginally bigger than what I estimate would be the margin of error.
Rather surprisingly, this unit doesn’t perform badly at all and will probably outlast my expectations. Testing continues, with the next report to be given when we reach 300 cycles in the experiment (or 316 cycles total from new).