How to deal with a battery with mismatched cells?

[Haysdb]

I'm repeating myself but here is a chart of 16 cell voltages showing how the high IR cells go high at 100% and then immediately go below the low IR cells as soon as charging stops. The second chart is the cell voltage delta between the high and low cells. There is a peak at 3.55v and then a second peak as charging stops and the voltage of the high cells drops below the low cells.

I tried balancing tighter, to 30mv, but this just made the delta greater during discharge. The tighter I top balance the more SOC is shifted from the high IR module to the low IR module, which if taken far enough would reduce the dischargeable capacity of the battery. In order to maximize capacity I'm balancing only enough so that the high cells stay below 3.6v (an arbitrary threshold).

If I could stop charging at 3.5v the delta would be much lower, and if I could stop at 3.45 it would be such a non-issue that I would probably just top balance to 10 mv like I do my other batteries. There would still be the issue of current sharing but that's a separate thing unrelated to balancing. But Seplos in their infinite wisdom has essentially hard-coded the charge voltage to 3.55v. They told me that if they allowed customers to charge to a lower voltage they would have customers complaining they aren't getting full capacity from their batteries. :-|

 

[Haysdb]

Seplos sent me this "equalizer" (balancer) but I don't understand why. I told them not to but they were insistent that it was the right tool for the job. But when I've asked how it will address a difference in internal resistance, they simply don't answer. I received it quite some time ago but I just unboxed it today. They have instructed me to discharge the battery until the BMS shuts off and then connect this balancer and let it run until it finishes. This would be a bottom balance. But I'm going to top balance the battery after it's charged, so what is the point of a bottom balance? It makes no sense to me. I can give it a try, but I really think it's a complete waste of time. Should I bother? Can it help? If so, how?

 

[Haysdb]

Still no answer from Seplos in answer to my question of how this balancer will help with a battery containing cells with mismatched internal resistance.

I have somehow managed to completely bork this battery. Two days ago it suddenly started acting like I had completely top balanced it. I didn't, or at least I didn't MEAN to. Now what I see is the voltages converge to near zero at 3.55 and then spread apart to 85mv at 3.4v. What I'll do when I get back home in a couple of weeks is to charge it to full then take it offline and let it sit for at least an hour to reach resting voltage (3.4v or below) and then balance it down from 85mv to maybe 60mv as a first pass. Then charge it back to 3.55 and see what the voltage delta is. I _want_ the voltage delta at 3.55v to be somewhere around 80mv. Balancing to a lower delta just shifts more (too much) charge from the high IR cells to the low IR cells, which will bite me in the ass at low SOC.

The chart shows the battery being charged to 100% around 1 PM and then the voltages mostly just setting toward resting voltage for the rest of the afternoon. The voltage spread at 3.375 is not IR but rather actually charge having been shifted from one group of cells to the other via too-aggressive top balancing. Managing this battery is going to one more complicated than I thought. It's going to need some gentle top balancing with maybe an occasional maintenance balance at 3.375 to 3.4v. Actually, after reducing the 3.4v delta down to 60mv or 50mv I need to discharge the battery until it shuts off to see what's going on at the bottom. I want to (1) protect the cells from daily high voltage above 3.6v and (2) retain full capacity if it's possible.