Broken cell? Voltage shooting up when charging at .2C

[MatthiasU]

Hello,

so I have a new 16-cell bank (EVE 280), not top-balanced yet, with voltages between 3.3 and 3.37V. Charging at 60A causes one of these cells, previously seen with unremarkable 3.33V, to shoot up to 3.7 within less than a minute – apparently too fast for the change controller to react, as the overvoltage caused the BMS to trip the relay, which can't be good for any of the solar hardware around here, but I digress.

Would you agree that this cell is bad?

 

[houseofancients]

so this is typical behavior when you don't top balance

as the SOC reaches 100% the voltage will spike up
so no the cell isnt nessecarily bad, you just skipped a pretty important step in you battery build

i highly suggest you do a top balance, and do so before anything else

 

[DThames]

If you can connect a resistor across the high cell, you can drain it down a bit. Try to get its voltage below the next highest cell. You need to get the others to catch up. Also, that is a high charge current if any of the cells are nearly charged.

 

[MatthiasU]

@houseofancients Top balancing is what I am trying to do … I already reduced the charge current to ~3A (which is what the balancer can eat ). my point is that when I have 15 cells that are (by now) mostly in the 3.4…3.5 range and don't move much when fed, and this one cell shoots up from 3.3, I can't help thinking that said cell is almost entirely unlike like any of the other cells, and hence might need replacing.

@DThames I already have a resistor, it's in the BMS. Yes I know that the charge current is high if any cell is nearly charged; however this one cell was among the lower half of the bunch and suddenly its voltage shoots up way above the others. Ugh. I really don't want to depend on capacity monitoring, esp. since I have no idea yet how efficient a charge/discharge cycle is going to be.

 

[rhino]

How do you know that 1 cell isn't already near 100% SOC?

 

[ArkRob]

"voltages between 3.3 and 3.37V"

I went through this too.

The thing I had to learn coming from lead acid batteries, is that relatively minor differences in voltage - particularly in the range you have described - can mean that the cells are in very different states of charge.

If you are currently top balancing, put a load on that single cell - there are instructions on this site for using several methods. I understand that the cell voltage differences seem minor, and the charging current may seem small (again, I went through this too) but you may have a 10%-30% - or maybe even more - difference in state of charge even with all that said, and the BMS resistor by itself isn't going to be able to overcome that in any kind of reasonable time frame.

If this cell were to discharge faster than the rest also, then I would be more likely to agree that it was a bad cell. Everything you have described so far though indicates that that cell was at a higher state of charge to start with.

 

[Solar Guppy]

Would you agree that this cell is bad?

No, you skipped the required step of top balancing ( all cells in parallel ), and got expected results of the highest SOC cell getting fully charged first. Reducing the charge current to the active balance ability of your BMS would be the next step OR do an actual top balance.

How do I top balance the Lithium cells?

There are a couple of easy ways to charge and balance a number of cells: Just putting them together in parallel and leaving them does NOT do much. 1. If possible - top each of the cells up to the voltage of the highest cell prior to wiring them in parallel. 2. Wire the cells in...

diysolarforum.com

 

[HighTechLab]

I have 15 cells that are (by now) mostly in the 3.4…3.5 range and don't move much when fed, and this one cell shoots up from 3.3

From this statement, I would agree something might be wrong with the cell. If you have 16 cells resting and the 1 lowest cell is around 3.3 and the rest 3.4-3.5, then that would be the most discharged cell. If you put on charge and this one spikes and hits COVP first, then it seems to be the cell with the lowest capacity.

I would suggest loading the battery down and see if the voltage drops down extremely low. If the cell has a defect, it likely will drop like a stone in the lake.

 

[houseofancients]

@houseofancients Top balancing is what I am trying to do … I already reduced the charge current to ~3A (which is what the balancer can eat ). my point is that when I have 15 cells that are (by now) mostly in the 3.4…3.5 range and don't move much when fed, and this one cell shoots up from 3.3, I can't help thinking that said cell is almost entirely unlike like any of the other cells, and hence might need replacing.

@DThames I already have a resistor, it's in the BMS. Yes I know that the charge current is high if any cell is nearly charged; however this one cell was among the lower half of the bunch and suddenly its voltage shoots up way above the others. Ugh. I really don't want to depend on capacity monitoring, esp. since I have no idea yet how efficient a charge/discharge cycle is going to be.

did you follow the top balancing guide ?
top balancing should be done by connecting cells in parallel after your bms stops for overvoltage protection...
you kinda missed the most important bit

 

[Solar Guppy]

If you have 16 cells resting and the 1 lowest cell is around 3.3 and the rest 3.4-3.5, then that would be the most discharged cell.

Yes, and all have to be brought up to the 3.65 volts to know they all have a full charge ( not necessarily same capacity ), this can only be done by one of the follow before putting into a string configuration:

Top charge, all cells in parallel OR Individually charge batteries to 3.65V

If one does ignores the above, then in the string configuration, you have to use an active balancing feature of the BMS and do not exceed the current that the balancer is capable of to bring all cells to 3.65V

The OP didn't do all three options described above, and was charging at a current much higher than the BMS can active balance at.

 

[HighTechLab]

Yes, and all have to be brought up to the 3.65 volts to know they all have a full charge ( not necessarily same capacity ), this can only be done by one of the follow before putting into a string configuration:

Top charge, all cells in parallel OR Individually charge batteries to 3.65V

If one does ignores the above, then in the string configuration, you have to use an active balancing feature of the BMS and do not exceed the current that the balancer is capable of to bring all cells to 3.65V

The OP didn't do all three options described above, and was charging at a current much higher than the BMS can active balance at.

Are you reading this thread or just commenting?

OP said

I already reduced the charge current to ~3A (which is what the balancer can eat )

The member may be new, but I think some of you guys are assuming he made the same mistake as the last guy and are ignoring the actual issue he is describing.

 

[HRTKD]

@MatthiasU , check your connections. A bad connection can manifest as a voltage issue.

Take that one cell out of the pack and charge it separately at ~3.6 volts using the power supply that you're going to use to do the top balance. If the amps goes to almost zero right away then the cell is charged.

 

[Solar Guppy]

Are you reading this thread or just commenting?

Yes, I read the posts.

@houseofancients I already reduced the charge current to ~3A (which is what the balancer can eat )

If indeed the BMS could shunt the full charge current, then the cell voltage would not rise.

There are few BMS's the can "eat" 3 amps, most are in the 100 - 200 milliamp range. The OP could help by listing what BMS is being used.

Top Balance the cells, as the OP proved the BMS can't do the job, that the only option.

 

[MatthiasU]

My BMS has large-ish 1-Ohm resistors and a cooling fan. Hence 3A, assuming I'm not balancing 15 of 16 cells.

Yes I know of the need to top balance, that's what I was trying to do. The problem is that I don't have some beefy lab power supply that can feed 100A at 3V to 16 cells in parallel, in order to get them balanced in some reasonable time, so what I was trying to do was to get one cell to 3.4 or so (they're supposed to have been delivered with matching charges …) and then reduce the current to balancer-compatible levels.

Thus my surprise when I came back to see a tripped BMS relay, and a voltage log that showed the shoot-up I described.

 

[HRTKD]

Yes I know of the need to top balance, that's what I was trying to do. The problem is that I don't have some beefy lab power supply that can feed 100A at 3V to 16 cells in parallel, in order to get them balanced in some reasonable time

Most of us don't have that either. What we do use is a less than $100 power supply that can put out 10 amps. That and a good amount of patience. The urge to get it done quickly is going to have a less than desirable result.

 

[houseofancients]

My BMS has large-ish 1-Ohm resistors and a cooling fan. Hence 3A, assuming I'm not balancing 15 of 16 cells.

Yes I know of the need to top balance, that's what I was trying to do. The problem is that I don't have some beefy lab power supply that can feed 100A at 3V to 16 cells in parallel, in order to get them balanced in some reasonable time, so what I was trying to do was to get one cell to 3.4 or so (they're supposed to have been delivered with matching charges …) and then reduce the current to balancer-compatible levels.

Thus my surprise when I came back to see a tripped BMS relay, and a voltage log that showed the shoot-up I described.

you wont need such a powerhouse.
a simple 5-10 amp bench powersupply should be plenty, given you have already bulk charged your cells in series.

 

[MatthiasU]

you wont need such a powerhouse.
a simple 5-10 amp bench powersupply should be plenty, given you have already bulk charged your cells in series.

Yeah, well, I was trying to avoid overcharging … and frankly it didn't take a whole lot of bulk charging to run into my little problem.
Oh well. Live and learn. I'll have to see what the actual usable capacity of the battery with the funky cell will end up at.

 

[houseofancients]

Yeah, well, I was trying to avoid overcharging … and frankly it didn't take a whole lot of bulk charging to run into my little problem.
Oh well. Live and learn. I'll have to see what the actual usable capacity of the battery with the funky cell will end up at.

another thing to check is your balancer.
at what point does it start and stop ?
is it just a passive balancer ?
what bms are you using ?

 

[740GLE]

Did you ever verify the high voltage cell with a multimeter? Or just go off what the BMS saw? Could be a bad sensing wire. Always trust but verify ?

Also if you’re resting voltage of the other cells resting voltage is 3.5v that’s very very close to 100% SOC.

Also what brand model BMS? You said fans, so is it a Daly?

 

[MatthiasU]

The balancer isn't the problem; the problem is the control loop behind it. There's a balance[sic] to be found between "reacts too slowly" and "is too jittery". I simply did not expect to have to go from "plenty of room in the battery, according to those pretty charge curves" (not to mention some past experience) and "oh no I'm overcharging this thing NO STOP PANIC" in half a minute or so.

Yes I have verified the BMS voltage calibration. It's based on diybms. Passive but not too pasive..