One cell ahead of others

[hackenstein]

Hi,
I have one new package of 8 of 3,2 280ah cells. When I charge them one cell runs ahead always and in few minutes it is 0,020v difference from the others. And when started charging it was 0,005v difference in whole bank. I have changed the position several times and also busbars. The same thing happens at each new position. It starts drifting at 3,300v already. What can I do about this? It is active balancer JK BMS 2A active balancing.

JK app does say that the resistance of each lead is roughly the same. All of the cells show ~0. 038-0,039. Also when I change the position of the cell.

 

[BentleyJ]

Its a good idea to top balance all the cells first while connected in parallel. 3.5V with a bench top power supply until the current falls off on ALL the cells indicating they are full.
Alternatively the the 2A active balancer will eventually get the job done if the cells are already connected together in series. The battery pack should be float charged at 28.0V for 24 hrs then check the Bluetooth app and verify all cells are within 10mv or less. Keep in mind the accuracy of the the voltage monitoring IC is + or - about 3mV. This is a total error range of 6mv. Therefore I usually set the balancer for 10 to 15mV cell to cell deviation. Anything lower is really not going to get better results.

LFP cells have a very flat voltage curve across a major portion of the usable SoC so any analysis of voltage deviations between cells has to done at 3.4V or higher (alternatively bottom balance can also be done at 3.0V and lower but not recommended.)

 

[hackenstein]

Its a good idea to top balance all the cells first while connected in parallel. 3.5V with a bench top power supply until the current falls off on ALL the cells indicating they are full.
Alternatively the the 2A active balancer will eventually get the job done if the cells are already connected together in series. The battery pack should be float charged at 28.0V for 24 hrs then check the Bluetooth app and verify all cells are within 10mv or less. Keep in mind the accuracy of the the voltage monitoring IC is + or - about 3mV. This is a total error range of 6mv. Therefore I usually set the balancer for 10 to 15mV cell to cell deviation. Anything lower is really not going to get better results.

LFP cells have a very flat voltage curve across a major portion of the usable SoC so any analysis of voltage deviations between cells has to done at 3.4V or higher (alternatively bottom balance can also be done at 3.0V and lower but not recommended.)

Yeah thats what I want to do. I want to charge them up till someone hit 3,65. And then put in parallel. But Im getting kind of worried this cell is bad because after just 5 min it differs so much even 0,030v or 30mV compared to the other cells. And this is with an active balancer of 2A.

Should I keep it in series and charge til one hit 3,65 and then parallel? I have never had a problem before with cells so I notice this is odd.

 

[BentleyJ]

If they are already connected in series I wouldn't take everything apart, at least not for a 30mV deviation, its not that much. Let the pack float for 24hrs at 28.0v and let the balancer work. A 280Ah cell may take some time to rebalance at only 2A. Keep in mind, its not 2A of continuous current. The JK balancer works on a 50% duty cycle. Half the time its extracting from the high cell(s) the other half delivering to the low cell. As cell voltages begin to even out you will notice the balancer starts skipping around to several different cells. The process takes time.

 

[hackenstein]

If they are already connected in series I wouldn't take everything apart, at least not for a 30mV deviation, its not that much. Let the pack float for 24hrs at 28.0v and let the balancer work. A 280Ah cell may take some time to rebalance at only 2A. Keep in mind, its not 2A of continuous current. The JK balancer works on a 50% duty cycle. Half the time its extracting from the high cell(s) the other half delivering to the low cell. As cell voltages begin to even out you will notice the balancer starts skipping around to several different cells. The process takes time.

Thanks buddy. Will try this tomorrow and give an update!

 

[hackenstein]

This is is after 1h of charge. From 1mv difference to 39mv. Im currently charging it till one hit 3,65v.

 

[timselectric]

Let it charge and do its thing.
You're not in the upper knee yet. That's when the delta will grow.
Its not even balancing yet.

 

[hackenstein]

Let it charge and do its thing.
You're not in the upper knee yet. That's when the delta will grow.
Its not even balancing yet.

I it is balancing. I just wanted to check for a min what happened without balancing so I turned it off. I will let it run now.

 

[timselectric]

I wouldn't have it balancing that low.
It just makes it worse for balancing in the high knee.
Balancing should not start until you reach 3.4v or higher. (Mine starts at 3.45v)
Running the balancer lower than 3.4v is actually unbalancing the cells.

 

[hackenstein]

I wouldn't have it balancing that low.
It just makes it worse for balancing in the high knee.
Balancing should not start until you reach 3.4v or higher. (Mine starts at 3.45v)
Running the balancer lower than 3.4v is actually unbalancing the cells.

Okay will turn it off! Do you mean it should start when one cell is at 3,4 or all of them above?

 

[timselectric]

It starts balancing when the first cell hits the start voltage.

 

[hackenstein]

Hi, the cell has now reached 3,6 and you can see the difference. It drops fast when charging is off and in 60 min it has dropped to same level as the other cells. 3,363V. See the attachment.

Should I put all in parallel now and balance it to 3,65?

 

[timselectric]

Let the BMS balance them.
You may need to limit the amount of charging amps to avoid cell over voltage protection, until they are balanced the first couple of cycles.

 

[hackenstein]

Let the BMS balance them.
You may need to limit the amount of charging amps to avoid cell over voltage protection, until they are balanced the first couple of cycles.

It is already finnished balancing after 1h. Cell diff is 3mv. Check the attachment. And when I turn on the chargin the cell starts running away fast from the others.

 

[timselectric]

It is already finnished balancing after 1h. Cell diff is 3mv. Check the attachment. And when I turn on the chargin the cell starts running away fast from the others.

Check the BMS connection to that cell.
Or, more balancing is required.

 

[BentleyJ]

Reduce the charging voltage to 27 then increment it up 0.1V, wait until balancing completes at every step up. Repeat until you get to 28V and all the cells are balanced.

 

[hackenstein]

Check the BMS connection to that cell.
Or, more balancing is required.

Nothing wrong with the connection. I have tried several different position. Same behaviour.
Okay I will balance it more.

 

[hackenstein]

Reduce the charging voltage to 27 then increment it up 0.1V, wait until balancing completes at every step up. Repeat until you get to 28V and all the cells are balanced.

Okay will try this, thanks.

 

[hackenstein]

Reduce the charging voltage to 27 then increment it up 0.1V, wait until balancing completes at every step up. Repeat until you get to 28V and all the cells are balanced.

Should I stop the charging between every step and let the active balancer balance it? The cell gets fully charged pretty fast. While the others are around 3,84.

 

[BentleyJ]

Should I stop the charging between every step and let the active balancer balance it?

If you have to turn off charging to prevent the BMS from shutting off due to over current protection, your charging voltage is too high. What is it now, Screen shot??

The cell gets fully charged pretty fast. While the others are around 3,84.

Not completely sure what behavior you are describing. The goal is to charge with Constant Voltage of 28.0V (3.5V per cell) until all the cells are balanced within 10mV.