Reconfiguring batteries to 4 series 4 parallel of 3.2v and one battery cell sparks

[Jades]

Hi all,

I have two twelve volt LiFePo batteries that I'm reconfiguring to be 4 series of 3.2v (see image below), because I could only find a BMS that would accommodate a 12v 4s (using same BMS used in video link below). In doing so, I connected one set of batteries just fine, but when connecting the 2nd 12v battery, one of the cells would spark when connecting. I did a little research and found that this was caused by a difference in voltage between battery cells. Sure enough, I used a multimeter and while the three other battery cells were at 3.35, the fourth was at 10. I then tried to connect it again (to the one battery connected in series to the other two connected in parallel). Still sparked. I measured the voltage again...it seemed to go up to 16...then it went to zero.

I thought that I would just configure a 12v battery and use the BMS on that twelve volt battery to balance all of the cells per Will Prowse's instructions here:

. However, the BMS has been on the battery now for over 24 hours and nothing has changed. Any ideas?

Best,

James

 

[Samsonite801]

Something doesn't sound right with that description. Are you saying a single cell with no bus bars attached is reporting 10v? That can't be possible. Please check again the points where you're taking voltmeter measurements from just to be sure what you're seeing there.

Start from scratch, remove all bus bars from all cells, check individual volts on each cell, and keep track of cell numbers, record them on a piece of paper. Then arrange your bus bars for an ALL parallel configuration so you can top balance them first (prior to switching to a parallel / series configuration).

To do this, you would hook all the + together with bus bars, and all the - together with bus bars. Then using a power supply (set to constant voltage 3.65v prior to attaching charge leads). Then disconnect power supply when it reaches 3.65v and 0 amps.

So do you only have 1 of the 4s BMS? Or two of the 4s BMS?

Based on the picture you provided above, assuming only one 4s BMS (2p4s), then you have to arrange the batteries in parallel groups of 2 (so you have 4 groups total), and then each group of 2 parallel cells wires in series ( + to - on each cell group), so you wind up with 4 series groups...

 

[Jades]

Thanks for the response.

So I checked again, and currently the voltage of the battery is 0.

I only have one of the BMS, hence that configuration.

I've arranged everything in parallel as shown except for the final battery, which I can't, because of the sparks/heat being generated. How do I know if it's a bad cell? Should I buy a desktop charger and try charging the battery before connecting it?

 

[Samsonite801]

Start from scratch again, remove all bus bars from all cells, check individual volts on each cell, and keep track of cell numbers, record them on a piece of paper.

Do that step first, confirm if any odd cells exist. Report back the independent voltages here if you can, let's see what we got on that level.

Don't forget to check polarity. If one or some cells are backwards, you might get backwards push and 0v... Check voltage and mark all cells with correct + or - in case they labeled cells wrong with + or - , or something silly like that...

Voltmeter reads -xx.xx volts if it is measuring reverse voltage, and reads just the number like xx.xx volts if red is positive and black is negative, that's how to confirm polarity.

Also make sure the voltmeter is on DC scale and has good batteries in it (that's the only thing I can think of that could cause a 0v reading, unless the cells really are all bad)...

 

[Jades]

Okay, I just went out and took off the busbars and reread for each battery. They were all at -3.35 except the 8th one, which was at -0. I had the setting on the first setting to the right, on the DC (10 megohm):

 

[snoobler]

That's an auto-ranging multimeter, it should adjust to whatever voltage range is needed. I have a few. Sounds like a dead cell. The sparkes are the big 7 cell battery trying to dump their charge into the 0V cell.

Terminology-wise, a battery is a collection of cells connected in series and/or parallel. A cell is a single 3.2V unit in this case.

 

[Samsonite801]

Okay, I just went out and took off the busbars and reread for each battery. They were all at -3.35 except the 8th one, which was at -0. I had the setting on the first setting to the right, on the DC (10 megohm):
View attachment 29158

When you get the -3.35v , you are seeing the - before the number? Is the Red lead of the voltmeter going to the Red pad on the cell? If Red is really the Red pad on the cell, and your voltmeter is reading -3.35v and not 3.35v then that indicates all the cells are marked backwards. Now if you have the Black lead of the voltmeter on the Red pad of the cell, and getting the -3.35v then you just had the voltmeter leads on backwards.

Also check if the Black cable of the voltmeter is plugged into the COM of the voltmeter, and the Red cable of the voltmeter goes to the DC volts port...

 

[snoobler]

If the red cable was in the 10A port, the meter would be fried, and it never would have displayed anything resembling a voltage.

You can see from the picture the black and red wires are reversed, so red to red/black to black will read (-).

 

[Jades]

Yes, black was not in com; so everything should be positive. What should I do for next steps?

 

[snoobler]

Replace dead cell.

 

[Jades]

Are we sure that it's dead and can't be recharged? If it is truly dead, and I'm connecting 2cells in parallel, can I just forego one of those pairs and just use one cell, such that I will have seven cells, 3 of which are paired and one of which is not?

 

[Samsonite801]

I suppose you could try to pump some juice to it to see if it could take a charge (I've heard there are some tricks for reviving certain lucky dead lithium cells), but exactly 0v is a bit suspicious to me as if the circuit went open internally. I wonder what's the current internal resistance of the cell now, measured at the terminals. Maybe it's open.. If it's open then no amount of tricks will revive it.

 

[snoobler]

Are we sure that it's dead and can't be recharged? If it is truly dead, and I'm connecting 2cells in parallel, can I just forego one of those pairs and just use one cell, such that I will have seven cells, 3 of which are paired and one of which is not?

You've already dumped massive current into it with each attempt to connect. It's still reading 0V. It's shorted. It's done.

You absolutely, positively can't have an odd number, 2P3S+1S as you describe. You might as well just use 4 cells as you would literally be working the single cell to death with the other 6 getting half a tickle.

Replace cell or only use 4.

 

[Samsonite801]

You've already dumped massive current into it with each attempt to connect. It's still reading 0V. It's shorted. It's done.

That's true, it didn't come up when he had it paralleled, and still back to 0v when pulling it back off, it's dead (and the spark meant shorted, not open), I'm agreeing with you there now