Ebike Battery pack got unbalanced. Reasons?

[eikido]

I got an ebike battery with brand cells (MJ1) and after about 1000km of usage, the pack is heavily unbalanced.
How can a battery pack all of a sudden get this unbalanced? I know it's a difficult question, but it could give me some hints

The cell groups measure voltage:


3.13 (+)
3.12
3.06
3.4
3.6
4.09
3.98
3.92
3.98
4.07
4.08
4.08
4.08
4.10

[mercly]

Do you have a bms?

[eikido]

Do you have a bms?

Yes it has a BMS. But i don't think it's a balancing BMS.

[mercly]

So that could be the problem. Every battery needs balancing.

[pwd]

The source of the imbalance could be a few things:

mismatches in cell group internal resistance
uneven current share on the series connections
faulty BMS
frequently draining the cells down too low

[999zip999]

did you balance all the cells before you hooked up the BMS

[docw009]

I've have two packs get badly unbalanced this year, In each case, one row would hit the 3,0LVC when the rest of the cells were close to 4.0 volts. My theory included (a) faulty BMS (non balance) that was leaking down that row of cells or (b) running the battery well past the C-rating of the cells. In one pack. I had run 13A per cell thru some 21700's only rated for about 5A, In the other, it was a steady 6A per cell on generic 18650's.

In both cases, I replaced the BMS with a balance BMS and was able to use the bike packs, I noticed that the unbalanced row on both backs would hit about 4.0V on charging, but never got that extra .2V to match the other cells. Also noted that my capacity was down. Eventually, I replaced the rows on both batteries with new cells. This was hard work, by the way, and required me to buy a spot welder,

The mystery continued with my purchase of a Littokala cell tester. It can discharge up to four cells at a time cell to LVC, fill it back up, and tell you how many maH were used and the final voltage, I expected to test the cells and find at least one from eacb row that had diminished capacity. I did not. All tested to the rated capacity marked on the cells, or even higher, and reached a charge of 4.2V The explanation is that the Littokala discharges at 250ma or 500ma for its load test. So it isn't comparable to the 2A-4A per cell often seen on an ebike,

I conclude that put of the 70 cells in the above two batteries, at leas two of them are unable to recharge or sustain current at their rated C-rates, but I'll need a better tester to find them,

[eikido]

Thanks all.

I'm suspecting frequent full discharges or a bad BMS as this type of battery pack is quiet common, so the design shouldn't be a problem.
Just wondering, how can a broken BMS without balancing function unbalance the pack?
I'm getting my Active balancer this week, a year later....

[Chalo]

Just wondering, how can a broken BMS without balancing function unbalance the pack?
I'm getting my Active balancer this week, a year later....

Yes, some BMSes draw their power from just one or two cell groups rather than the whole pack, so over time those groups get drained down if the battery isn't periodically cycled and balanced. A malfunctioning BMS can leave the bleed resistor on for any (or all) of the cell groups and drive them to 0V. I have run across both these things several times when diagnosing misbehaving packs.

People tend to leave a bad pack to sit for a long time even when they suspect a BMS problem. This is exactly the wrong thing to do. My advice if you suspect the BMS isn't working right, or you're going to let the pack sit for a prolonged period, is unplug the multi-pin connector from the BMS so it can't drain any cells. You can leave the heavy wires attached because they only contact the battery negative terminal, and there's no way to close a circuit. At that point you can let it sit, but maybe make a note of what you know about it and keep it with the battery for later reference.

[eMark]

My advice if you suspect the BMS isn't working right, or you're going to let the pack sit for a prolonged period, is unplug the multi-pin connector from the BMS so it can't drain any cells.

Does anyone really know what is "working right" when it comes to relatively inexpensive BMS discharge balancer or even an expensive so-called 'smart' BMS. Sometimes the BMS is OK, and the fault is with the cells in the pack.

Yes, it's very possible that these Lithium ion BMS' will continue to slowly discharge balance over time. Possibly as low as 3.2v as is the case with the RC Lipo "Blinky" [BMS} discharge balancer (cells within 25-30mV) ... https://www.astroflight.com/106-123-ast ... tails.html

Whether using a RC Li-po "Blinky" or inexpensive Li-ion BMS neither is a cure for 2nd rate cells and/or pack abuse. What i mean is that with 2nd rate cells or an abused pack it isn't possible to balance the p-group cells within 25 to 30mV of each other no matter how long (i.e. weeks--months) is the BMS discharge balancing. Not sure about "active" balancing, but probably the same.

Protection is just as important a factor as balancing with any BMS. When a pack, even a good pack, is not in use for weeks--months always follow Chalo's advice ...

unplug the multi-pin connector from the BMS so it can't drain any cells.

[Chalo]

Whether using a RC Li-po "Blinky" or inexpensive Li-ion BMS neither is a cure for 2nd rate cells and/or pack abuse. What i mean is that with 2nd rate cells or an abused pack it isn't possible to balance the p-group cells within 25 to 30mV of each other no matter how long (i.e. weeks--months) is the BMS discharge balancing. Not sure about "active" balancing, but probably the same.

Active balancer boards provide full time cell voltage detection, plus balancing when variation exceeds 30mV, at any state of charge. They balance at a current of up to 1.2A (or more for some boards), though surely this value is much lower for small voltage differences. They go back to standby when the cells come within a 10mV range. They are rated to consume 25 μA or less while standing by.

So cells would have to be very shabby, very dissimilar, or seriously damaged to outrun the active balancer's ability to bring them to the same voltage, even in real time discharging. Then of course while the battery is charging or idle, the active balancer will keep moving charge around until the pack is balanced again. Maybe "leveled out" is a better description than "balanced" in this case, because a pack that requires much intervention to bring to a uniform cell voltage is inherently imbalanced.

I think the main disadvantage of an active balancer is that it can hide symptoms of rapid self-discharge in individual cells or parallel groups. It would be good due diligence to unplug the thing once in a while when the battery is not being used, and then check cell voltages before plugging it back in.

[spinningmagnets]

Its been commonly posted that "BMS" stands for battery management system, but...ebike mechanics and design engineers have told me it actually stands for Battery Murder Suspect.

Do with this knowledge whatever you wish...

[Chalo]

Its been commonly posted that "BMS" stands for battery management system, but...ebike mechanics and design engineers have told me it actually stands for Battery Murder Suspect.

Do with this knowledge whatever you wish...

Ah, but when you eliminate that suspect... then you become the person of interest.

BMSes are not so much for the benefit of the battery, as they are for somewhat protecting the battery from the user.

[spinningmagnets]

BMSes are not so much for the benefit of the battery, as they are for somewhat protecting the battery from the user

"I have met the enemy, and he is us" - pogo

[eikido]

Well, turn out i was wrong.
My BMS does balance so it actually might be a faulty BMS.
I got two suspects now

My balancer is working hard now..
haha

[docw009]

I have the suspect now for one of the two unbalanced batteries. Faulty cell. I had a 13S-2P 21700 pack with one bad row. I replaced the cells. Battery now works fine, As for the two cells I took out, a month later, one was completely discharged to zero volts. sitting unconnected with no load. The other seemed OK. So it was a bad cell with a slow self discharge.

The other unbalanced pack remains unsolved. I changed to a balance BMS and also replaced the bad row.The five cells I took out are in safe storage and appear to be healthy.

[999zip999]

What motor and what app controller are you using plus this pack is only 2p right ? Maybe you're stressing the pack too hard ?

[docw009]

What motor and what app controller are you using plus this pack is only 2p right ? Maybe you're stressing the pack too hard ?

If you're asking me.

Yes, I knowingly overstressed it. 13S-2P with 5AH 21700 cells. I had it on my BBS02, which draws 26A with throttle. (I normally run it on a 14S-4P from luna, that uses 30Q's.) No voltmeter so couldn't see the voltage sag, but it never stalled. A couple of no drama rides down to LVC. Recharges from there needed 8AH, so I was happy with capacity.. On the 3rd/4th recharge, one row hit LVC early. I recall the rest of the cells were around 3.6V.

Repaired and now used on a bike ridden by a grandma at 12 mph. That was where it was originally intended.

[leffex]

I hate it.

I mean I got new packs that has unbalanced cells and if you read one group they can be 3.0v when the rest are sitting at 4.0v. BMS cuts charge at different voltages for a 36v battery; 37,5v. 38,5v 40,5v . Total capacity measured before BMS cuts again at discharge can be from 3,5% to 15% capacity.

My hypothesis is that if we go back some years, recycled cells became Chinese cheap cells (not taken and converted but made out of recycled material from these). In the end recycling centres start recycling more and more low cost and low quality cells that already may have been recycled once. One cell that I would avoid totally is LG M26. Totally crap. It's like you are selling sewage water for consumer use.

So if the BMS is not a balancing one or not. How do the cells end up unbalanced to hell. I mean if we take it up a notch if kind of all cells are all over the place. What can you imagine or theorize it to be so that it became so bad in the first place?

Another weird issue is in transit regarding battery packs that have lights if the stay lit in transit. Weird thing is still only one cell group is low in these packs even though the connection for the light goes to the whole voltage of the pack.

I cut my losses and just scrap the intestines for new modules ready to go in. That saves some time down to 1-2 hours at most to handle a pack any way you decide to fix it. Still an hour to much I think.