Xiaoxiang BMS Charge Balance

[DubP]

Hi Folks,

I put together a battery with a Xiaoxiang BMS, was wondering what people's experiences are with it's charge/balance operation. My pack was admittedly pretty far out of balance when I connected it all, but it's taking FOREVER to get to a balanced state. Like 2 weeks. And more concerning, I set it to charge and balance and it seems like it's going to blow right past 4.2V on the higher voltage packs. Where does it turn off at when you leave it on a charger?

 

[amberwolf]

I put together a battery with a Xiaoxiang BMS, was wondering what people's experiences are with it's charge/balance operation. My pack was admittedly pretty far out of balance when I connected it all, but it's taking FOREVER to get to a balanced state. Like 2 weeks.

Most of these have 50mA or less balancing current ability. Meaning, if it's a high capacity pack, with large imbalance, it could take a very long time to fix.

For instance, if just one cell group is out of balance by 2Ah, then if the BMS just sat there and drained all other groups down (simultaneously) to the level of that group (which isn't how they work), it would take 1.6 days (40 hours) at 50mA to do. If it was out by 20Ah, it would take ten times as long to do. If it's only a 10mA balancing currnet, it would take fives times that long....

Since the way it really works is that the BMS first stops charge when any group exceeds the cell-level HVC, then it drains down all groups that are above the balance-start-HVC until they are at the balance-stop level, then it restarts charge (which will be at a low current by the time the pack is far enough charged to start balancing), and repeats this process over and over until all cells are below the balance-start-HVC point, but above the cell-full point, it can take a lot longer to actualy do the balancing.

Add to that that many BMS don't power themselves from the entire pack, and may be drawing power from only the first group or first few (at the most negative end of the pack), it may take even longer if the BMS is drawing any significant power during this process, and any of the low groups are within the groups powering the BMS.


If a cell group has a problem (leaky cells, etc) then it may never finish balancing, as the leakage might be close to or greater than the fill rate.

 

[DubP]

Thanks so much for the detailed response! I think the biggest difference in pack voltages when I connected the BMS was about 0.4V (4.1-3.7), and my pack capacity is about 24Ah. Assuming those Ah work linearly across a 1V span (like 4.2 to 3.2), I think that it means it was imbalanced by 40% of the capacity, so 9.6Ah. does that make sense? Probably it was out more because I think the discharge is not linear and most of the capacity is at the higher end of the voltage span.

OK, so yeah, 9.6Ah/.05A gets me, wow, 192 hours to balance. And you say it doesn't necessarily balance between all packs at the same time, so now I see that it could take a lot longer.

I guess I'll have to cycle this battery a few times to see how out of balance it gets during use. It seems like a BMS really needs to keep up with balancing overnight if it's for a battery that gets cycles daily. I'm starting to see why a single weak cell or two could bring down a whole battery.

 

[amberwolf]

Thanks so much for the detailed response! I think the biggest difference in pack voltages when I connected the BMS was about 0.4V (4.1-3.7), and my pack capacity is about 24Ah. Assuming those Ah work linearly across a 1V span (like 4.2 to 3.2), I think that it means it was imbalanced by 40% of the capacity, so 9.6Ah. does that make sense? Probably it was out more because I think the discharge is not linear and most of the capacity is at the higher end of the voltage span.

It's not really linear; and the actual curve depends on both the specific cell model and the load on the cell vs it's capabilities. Example:
https://lygte-info.dk/review/batteries2012/Samsung%20INR18650-25R%202500mAh%20%28Green%29%20UK.html

I guess I'll have to cycle this battery a few times to see how out of balance it gets during use. It seems like a BMS really needs to keep up with balancing overnight if it's for a battery that gets cycles daily. I'm starting to see why a single weak cell or two could bring down a whole battery.

Improperly connected groups do it too--a broken connection to one or few cells in a group means that group doesn't ahve hte same resistance or load-bearing ability or capacity as the others. It's a pretty common problem, as is poorly-matched cells, so that different groups have different properties even when all the cells are well-connected.

A well-made pack of well-matched cells doesnt' even *need* to be balanced, especially if it is not run hard (meaning, not near it's LVC or near it's cells' max current limits). I don't even use a BMS at all on my decade-old EIG NMC 20Ah cells; they were designed to be used at high currents (100A each), in large-EV use, and I'm only using them at a fifth of that for a few seconds at a time, probably a tenth of that most of the time, and I rarely drain them nearly empty, so they stay in nearly perfect balance year after year. Occasionally I do testing of higher power systems with them that make them actually work for a living, but they dont' have to do that very often.

So for me, the cells are already pretty closely matched in characteristics (or were originally, at least), and easily capable of what I use them for without breaking a sweat.

For most packs, people choose something that can just barely do what they want it to, and *then* also go the cheapest possible route for it, which means it is not really capable of what they want, without fairly severe stress to it. So their packs get out of balance easily, and even worse as they age, and the BMS systems used are typically barely capable of dealing with the "normal" stress, much less what they actually put them thru unknowingly.

It gets worse from there.