Does this mean I have a bad cell in my 48V system?

[jdy2kgt]

The battery is a 48V using 16x 100AH EVE cells and a JK BMS. I built the batter back in March and it's worked great all year till now. Yesterday I noticied that anytime I put a load of 2,000W or more on the battery cell number 7 goes from the same as the rest to a lot lower and fast. The wire resistance is the same as the rest of the cells and all the terminals are torqued to the same measurement. I checked the wires and nothing is hot or lose. I looked at the cell without taking the entire battery apart and it doesn't look bloated (see pictures). Here's two videos showing when I put just 1,000 W on it and then one showing over 2,000W. The 2,000W+ one you will see the "Protection (Cell Under Voltage)" error at 1:25 seconds into the video.

1000W video1000W video

2000W video2000W video -

 

[42OhmsPA]

The battery is a 48V using 16x 100AH EVE cells and a JK BMS. I built the batter back in March and it's worked great all year till now. Yesterday I noticied that anytime I put a load of 2,000W or more on the battery cell number 7 goes from the same as the rest to a lot lower and fast. The wire resistance is the same as the rest of the cells and all the terminals are torqued to the same measurement. I checked the wires and nothing is hot or lose. I looked at the cell without taking the entire battery apart and it doesn't look bloated (see pictures). Here's two videos showing when I put just 1,000 W on it and then one showing over 2,000W. The 2,000W+ one you will see the "Protection (Cell Under Voltage)" error at 1:25 seconds into the video.

1000W video1000W video

2000W video2000W video -

Nice build!
Could you swap cell 7 with another position and see if it behaves the same?
Did you measure the voltage of the terminals on cell 7 to make sure it's the same as the BMS is seeing?

 

[Alkaline]

Check wiring make sure it is not loose, yes swapping position will also help.

 

[SeaGal]

Does this happen at much higher SOC?

Am wondering if the cells are just not balanced well and #7 is the lowest one. The BMS figures are all lowish - cell #11 is only 2.8V, so you're into the low-end knee already.

And you have a 369mV differential between cell#1 and cell#11 anyway.

Suggest a top balance and retest at higher SOC as well as the suggestions above.

 

[jdy2kgt]

Does this happen at much higher SOC?

Am wondering if the cells are just not balanced well and #7 is the lowest one. The BMS figures are all lowish - cell #11 is only 2.8V, so you're into the low-end knee already.

And you have a 369mV differential between cell#1 and cell#11 anyway.

Suggest a top balance and retest at higher SOC as well as the suggestions above.

I can certainly do a top balance but I did that when I built the battery and thought the JK with active balancing would make it so I never had to again. But they were at 100% SOC for these test.

 

[Partimewages]

How are you charging the bank? What are your settings?

 

[SeaGal]

But they were at 100% SOC for these test

Sorry, I'd only looked at the pic - have looked at video now.

That's odd - they all drop so much... even cell #11 goes from 3.3 to 2.9V, that 400mV drop on 2000W. Just looked at my voltage graphs from the dishwasher and kettle this evening and I'm seeing only a 30mV drop at 2000W and about 50mV drop at 3000V on cells that are 3.25V - ish.

Something not right, but not just cell #7 IMHO - it's across all cells.

 

[Nobodybusiness]

The battery is a 48V using 16x 100AH EVE cells and a JK BMS. I built the batter back in March and it's worked great all year till now. Yesterday I noticied that anytime I put a load of 2,000W or more on the battery cell number 7 goes from the same as the rest to a lot lower and fast. The wire resistance is the same as the rest of the cells and all the terminals are torqued to the same measurement. I checked the wires and nothing is hot or lose. I looked at the cell without taking the entire battery apart and it doesn't look bloated (see pictures). Here's two videos showing when I put just 1,000 W on it and then one showing over 2,000W. The 2,000W+ one you will see the "Protection (Cell Under Voltage)" error at 1:25 seconds into the video.

1000W video1000W video

2000W video2000W video -

Is that the only battery you have for your system or others in parallel?

Seems like a lot of Voltage sag for .5C.

The suggestions given such as checking the connections on that cell and swapping it to see if the problem moves or stays with the slot are good ones.

Once you do that, let us know the results.

 

[SeaGal]

The suggestions given such as checking the connections on that cell and swapping it to see if the problem moves or stays with the slot are good ones

I thought that at first.... but on the video, all cells drop a huge amount.

 

[Nobodybusiness]

I thought that at first.... but on the video, they all drop a huge amount.

I agree it’s a huge voltage sag for .5 C.

Very strange.

#7 just happens to be the worst..

I would still swap it just to see what happens.

 

[jdy2kgt]

How are you charging the bank? What are your settings?

Charging through a Victron 100/20 Solar Controller with 440W of solar - attached are my settings.

 

[Surferdudemi]

Given this has been in service for 8 months without problems until now, yeah, looks like you've got 1 weak cell and 2 more than are on their way. Given the flat curve of LiFePO4, 11 looks weak as well. I'd consider replacing both of them. How many AH did you draw to get to that point? If they are top-balanced, then once you discharge them, one of the cells has to be the lowest. If you've drawn less than 70% of the capacity at that point, just replace the two low cells. The remaining capacity shows 0, but I'm guessing that's based on the voltage the BMS sees, not the cumulative charge.

 

[Partimewages]

Those are BMS settings. What are your SCC settings?

 

[Nobodybusiness]

The battery is a 48V using 16x 100AH EVE cells and a JK BMS. I built the batter back in March and it's worked great all year till now. Yesterday I noticied that anytime I put a load of 2,000W or more on the battery cell number 7 goes from the same as the rest to a lot lower and fast. The wire resistance is the same as the rest of the cells and all the terminals are torqued to the same measurement. I checked the wires and nothing is hot or lose. I looked at the cell without taking the entire battery apart and it doesn't look bloated (see pictures). Here's two videos showing when I put just 1,000 W on it and then one showing over 2,000W. The 2,000W+ one you will see the "Protection (Cell Under Voltage)" error at 1:25 seconds into the video.

1000W video1000W video

2000W video2000W video -

When watching the 2000watt video.
It’s blurry but when you hit 32 seconds the battery capacity goes from 94% to 0 and stays there.

Something is definitely not right.

Voltage sag shouldn’t be that big on a .5c pull.

You definitely have a runner with #7 but your other cells are pulled down to 2.8-3.1.
That shouldn’t happen.

Short of putting a shunt or VOM on the negative side to see if you are pulling 50 amps or it’s actually more?

Checking all the connections to make sure tight.

Swap in a new cell for #7.

Swap BMS for spare?

If you can take it offline measure individual cell voltages.

These are the things I would start with.

It’s definitely strange..

 

[jdy2kgt]

Those are BMS settings. What are your SCC settings?

Here's my SCC settings..

 

[#40Fan]

Do you have a multimeter to confirm voltage readings instead of relying on BMS?

 

[jdy2kgt]

Do you have a multimeter to confirm voltage readings instead of relying on BMS?

I also have the Victron BMS on their so I can do another test tomorrow and compare the two (shunt vs bms). What else should I check with the multimeter?

 

[#40Fan]

The shunt measures overall battery voltage (as well as other things, but not individual cell voltages). The BMS measures individual cell voltages (as well as other things).

Using a multimeter to check individual cells and comparing them to what the BMS is telling you, would be a good way to diagnose the issue you're having.

For instance, the BMS is saying #7 cell is low. Check the voltages with a multimeter to confirm the BMS is correct.

 

[SeaGal]

That's odd - they all drop so much... even cell #11 goes from 3.3 to 2.9V, that 400mV drop on 2000W. Just looked at my voltage graphs from the dishwasher and kettle this evening and I'm seeing only a 30mV drop at 2000W and about 50mV drop at 3000V on cells that are 3.25V - ish.

Something not right, but not just cell #7 IMHO - it's across all cells.

Something is definitely not right.
Voltage sag shouldn’t be that big on a .5c pull.
You definitely have a runner with #7 but your other cells are pulled down to 2.8-3.1.
That shouldn’t happen.

Using a multimeter to check individual cells and comparing them to what the BMS is telling you, would be a good way to diagnose the issue you're having.

This has been bugging me in the middle of the night ?

With a 2000W load all cells are dropping hugely (like 300mV+) - e.g. from 3.3V to 3.0V. When the BMS cuts out due to Cell#7 undervoltage, they ping back up to 3.3V. So it can't be the cells are discharged or they wouldn't go back to 3.3V.

To me the big drop implies too much resistance in a cell or connection, but why across all cells, roughly evenly?

Def need to check with multi-meter as it happens.

I'm thinking either faulty BMS or what if there is a bad connection on the -ve... I can see how the BMS would report voltage drop across the pack in that case, but still can't get my round why every cell is being reported as low.

 

[400bird]

In the 1000 watt video, cell 7 was the highest voltage at the start, but still dropped lowest under load. It's either a bad connection (remove bus bars, clean, reassemble), bad cell -looks like high internal resistance to me(but they all drop too much under load), or a voltage measurement problem (this seams unlikely, but would be easy to test with a volt meter)