@MTB2019 I’m assuming from the the latest map video that you got the pack balanced. Nice.
I was having a similar problem recently and wanted to drop some data here so that others looking for balancing info could find some hard numbers.
- My pack is relatively new, uses (48) Samsung 30Q cells in a 12s4p configuration.
- I use a Neptune 15 BMS which handles the charging and balancing. Conveniently is has an Android app for monitoring.
- The BMS is wired as charge only since I pull more peak amps than the BMS can handle
- The BMS settings before the event:
5 Hour charge and balance time (before a timeout)
50mV delta for all P-groups (it will continue to balance until it hits this delta)
On August 27th, at about 4PM I noticed that my battery had a P-Group delta of ~350mV (yipes):
I had no idea how the pack had become so unbalanced. I searched the forum and found this thread, which helped give some perspective on the lengthy times required to balance a pack (without charging a specific P-group using a standalone charger)
Taking this into account, I changed the BMS to:
- Use the maximum balance timeout (18 hours)
- Shrank the p-group delta to 30mV (to help keep the pack more aligned)
- Started the process of balancing the pack
I may have missed a few hours here and there (as I restarted the balancing after hitting a timeout), but between 2020-08-27 ~8PM and 2020-09-04 ~8PM the pack was actively balancing. Here’s the results from after the balance completed:
This means the balancing process required ~192 hours to move from a ~350mV delta to a ~30mV delta (~320mV change). This means that to move the needle 10mV required about 6 hours.
After the pack balanced I charged up to full (which for me tops out at 4.1V/group) and went for a long ride. After the ride I checked the resulting pack balance:
The pack remained balanced throughout the discharge (which was a relief). I have since run through a couple charge / discharge cycles and the pack has remained in a balanced state.
Interesting thoughts on the data
Note that the ~350mV delta was the largest difference, but half of my p-Groups were unbalanced. This appears to essentially be a worst case scenario where half the pack needed to be brought in line with the other half. Based on this I would say the balance time I note here is probably the longest expected for this configuration. I imagine that larger deltas, or higher parallel cell count would result in even higher balance times.
For me, the lingering question remained, how did this happen if my pack is healthy and I have a good BMS?
From the results, I would conclude that the 5 hour charge and balance timeout stopped the balance cycle before the pack was done balancing. After several charge / discharge cycles this resulted in an ever increasing delta between the groups, until the pack arrived at the 350mV difference. Increasing the timeout seems to have addressed this issue, which I think bolsters this conclusion.