We have had a number of enquiries to look at Li-ion batteries that radically have lost capacity or indeed that have stopped working altogether. We are never that optimistic at the start, as there are many reasons for a battery to stop working, not least because it is genuinely worn out. That being said, it is unusual for a noticeable decline in performance over a short period of time so we’ve had a look and been encouraged by the results.

One particular annoyance is the use of LED indicators on batteries. They are great for gauges as they need only a few milliamps of current when lit; it’s a problem when they are designed to be on whilst the battery is in use, and worst of all when there is a manual switch to turn off the battery and the LED. Those few milliamps of current are 100 to 1000 times the normal, no-load operating current of the battery. The fact that it is a fraction of the load current is immaterial. So, what happens is that the battery is used until it runs out. At some time later it is re-charged or maybe it is put on a shelf or left in the boot of your car. It is left switched on and the LEDs are quietly running the battery down further.

The last battery we took in was a case such as this. A nominal 36V battery was run down to 1V. Usually this is game over, but the cells were from a reputable source and so we gave it a go, slowly charging the cells with a few milliamps over a long period before putting it back on normal charge. We were surprised that the battery actually recovered and it had most of its normal capacity given it was a few years old.

Another one had a slightly different issue. All the cells were fully charged when we got it, except for one bank. There was a cell balance circuit in the battery so this should not happen. Often, the balance current is quite low in comparison to the charge current and so there can still be cases where the cells drift off. We charged this one bank, and the battery worked again. Simple thing really: the charged cells would not allow the battery to charge because they were at maximum voltage and the discharged cells would not allow the battery to discharge because they were at minimum voltage.

A third case took a bit more working out. Much the same as the previous example, we quickly fixed the battery, but were tasked with finding out the root cause of the failure. Everything looked to be sound and we didn’t initially discover the problem, which was that the battery would lose effective capacity after a relatively short time of use. As a side issue we noticed that one cell alway seemed to discharge before the rest. We had seen quite a few batteries of the same type from different manufacturers but it was always the same cell. This had to be an uneven load current on the cells, but as it was not the main issue we decided to park this in favour of finding the main problem.

It was LEDs again; well sort of. The product had a few that lit up when you switched it on. They used a low current until the product was in use, at which point the load current increased to several amps. When not in use the load current dropped again to a few tens of milliamps. Then as the battery discharged the LEDs went out. This was not due to the battery however. The LED drive circuit shut down before the battery, but the load current was still there, albeit reduced to 20mA or so. The battery protector was also unusual in that it required a residual load current to be flowing in order to trigger the protections. 20mA was under the radar so the battery continued to discharge. Overnight, this could take the battery well below normal operating voltage.
The protection circuit was updated by the manufacturer and this problem went away. So the moral is, don’t give up on your dead batteries; it might not be as terminal as you think (no pun intended).