Top Balance - JKBMS - drift voltage between pair and odd cells

[Andre Magro]

Hello,

I'm building a 16 cell (48V) 272Ah battery bank, and yesterday I started to top balancing the first 8 cells.
I'm using JKBMS 24S 2A 200A BT and two benchtop power supplies (30V10A anda 60V5A) with 15A total.
I put the cells in series and the top balance procedure started, btw the "Balance" feature of the BMS was turned on during the top balance.
After aprox 14h the BMS was triggered since "Cell Over Voltage" 3.6V was reached.

And this is what I got:

Cell	initial state (V)	4h later (V) (charging)	14h later (V) (stop charge)	17h later (V) (w/Balance:On)
1	3.295		3.555	3.419
2	3.296		3.381	3.354
3	3.295		3.549	3.419
4	3.292		3.366	3.350
5	3.293	3.375	3.551	3.419
6	3.295		3.380	3.354
7	3.295		3.544	3.419
8	3.290	3.392	3.371	3.348
Delta Cell	0.006	0.017	0.189	0.071

As we can see 14h later there is a big diference between the cells voltage, but curiously the cells 1,3,5,7 are very close, and cells 2,4,6,8 also.
(the column "4h later" shows the lowest and highest voltage cells)

So, what could happen? Different SOC cells, but why exactly between pair and odd cells?
Should I leave the cells in series so the JKBMS could balance them? Should I connect them in parallel and let them balance?

Edit: add "17h later" column (the cells are settling with BMS "Balance" mode on)


Thanks

 

[740GLE]

If the JK is like JBD (except higher current), only the odd or even cells get balanced when balancing. So it looks like it was keeping odds higher balanced than evens and as time wore on it did it’s job.

 

[RCinFLA]

You are still out of balance. Make sure you do not have terminal connection problems that will affect what the BMS thinks the cell voltage is. Resistance in terminal connections results in voltage drop of connections with current. A good terminal connection is 0.05-0.07 milliohms. A poor connection can be 0.2 milliohms. A bad connection can be over 0.5 milliohms.

Aluminum cell terminals builds aluminum oxide which needs to be cleaned off prior to making bus bar/terminal connections. Some white vinegar mixed with water, 50/50% mix, cleaning followed by 91% isopropyl alcohol cleaning will help to clean off aluminum oxide. Don't touch cleaned terminal surfaces and connect soon after terminals dried off. Aluminum oxide regrows in short amount of time.

It is best not to balance below 3.4v cell voltage as there is little difference between cell voltages and balancer may make wrong decisions on balancing direction.

When first cell gets above 3.45v you can reduce charging current down to have a higher percentage of balancing to charge current ratio. This will allow more time before highest state of charge cell causes a BMS cell overvoltage shutdown. If you get an overvoltage shutdown just let BMS continue to balance. When BMS resets it will start charging again but it may get a few more shutdowns before cells finally get balanced enough to allow achieving 3.55v-3.60v on all cells. Charge current should be allowed to taper off at absorb voltage of 3.5-3.6v per cell.

 

[740GLE]

I believe JK has 2A balance current, correct? That’d make short order of some imbalance. I think if he keeps the charger with cells above 3.5 and low current.

Report back after 20-24 hrs that cell difference should be much better.

 

[Andre Magro]

Hello.
Thanks for your reply's.

After some more hours with cells settling and being balanced by BMS, they have now very similar voltages.

Cell	23h later (V) (w/Balance:On)
1	3.345
2	3.340
3	3.344
4	3.339
5	3.345
6	3.340
7	3.345
8	3.340
Delta	0.006

Should I assume that the top balance procedure was well succeeded?
I'm not comfortable with this situation since "pair" cells (2,4,6,8) did not pass the 3.3...V during the top balance procedure.

Should I connect them in parallel and let them reach 3.6V?

Thanks

 

[John Frum]

Hello.
Thanks for your reply's.

After some more hours with cells settling and being balanced by BMS, they have now very similar voltages.

Cell	23h later (V) (w/Balance:On)
1	3.345
2	3.340
3	3.344
4	3.339
5	3.345
6	3.340
7	3.345
8	3.340
Delta	0.006

Should I assume that the top balance procedure was well succeeded?
I'm not comfortable with this situation since "pair" cells (2,4,6,8) did not pass the 3.3...V during the top balance procedure.

Should I connect them in parallel and let them reach 3.6V?

Thanks

3.34 volts is in the flat part of the charge/discharge curve so the state of the top balance can't be determined at this voltage.
Charge battery into the high knee to see what the cell delta is at 56.8 volts.
If the lowest cell is higher than 3.45 and the highest cell is lower than 3.65 its workable.
The tighter the better though.

You want your bms to only balance above 3.45 and my preference is to only balance during charging.

 

[Andre Magro]

Hello @smoothJoey

When the BMS shouted down the charger, the delta was 0.189V (Cell1: 3.555V and Cell4: 3.366v), should I re-charged them again in series?
Maybe with active balance feature disconnect and at lower amperage?

Tks

 

[John Frum]

Hello @smoothJoey

When the BMS shouted down the charger, the delta was 0.189V (Cell1: 3.555V and Cell4: 3.366v), should I re-charged them again in series?

Yes.
cell 4 is well below 3.45.
Cell 1 could have been charge as high as 3.65 volts.
I suspect you have modified your bms settings to trip at 3.6 volts, confirm?

 

[Andre Magro]

Yes.
cell 4 is well below 3.45.
Cell 1 could have been charge as high as 3.65 volts.
I suspect you have modified your bms settings to trip at 3.6 volts, confirm?

The BMS settings are by default:

 

[John Frum]

cell ovp(v) should be 3.65
cell uvp(v) should be 2.5
start balance should be 3.425

max charge current should be ~140 amps


Try to charge to 56.8 volts and see if the high cell stays below 3.65 and the low cell gets above 3.45.

Because we set the bms to start balancing at 3.425 volts, your float voltage must be lower than 54.8 volts.
I like a float of 53.6 volts.

 

[Andre Magro]

cell ovp(v) should be 3.65
cell uvp(v) should be 2.5
start balance should be 3.425

max charge current should be ~140 amps


Try to charge to 56.8 volts and see if the high cell stays below 3.65 and the low cell gets above 3.45.

Because we set the bms to start balancing at 3.425 volts, your float voltage must be lower than 54.8 volts.
I like a float of 53.6 volts.

Those charge voltages that you mentioned is for 16s cells right? I'm using 8 cells so I can simply divide by 2.
Since cells was charging all night long, do you recommend me to wait some time before charging them again?

 

[John Frum]

Those charge voltages that you mentioned is for 16s cells right? I'm using 8 cells so I can simply divide by 2.

yes.

Since cells was charging all night long, do you recommend me to wait some time before charging them again?

Why?

 

[John Frum]

To be clear I want you to charge with a charge voltage of 28.4 volts
As long as the highest cell doesn't trip the bms by exceeding 3.65 volts and the lowest cell is above 3.45 I expect the bms can manage the pack.
You charger's float voltage must be lower than 27.4 volts and I suggest 26.8 volts.

 

[Andre Magro]

To be clear I want you to charge with a charge voltage of 28.4 volts
As long as the highest cell doesn't trip the bms by exceeding 3.65 volts and the lowest cell is above 3.45 I expect the bms can manage the pack.
You charger's float voltage must be lower than 27.4 volts and I suggest 26.8 volts.

I will charge the cells based on your suggestions, and them I will give feedback.

Thanks @smoothJoey

 

[RCinFLA]

Hello.
Thanks for your reply's.

After some more hours with cells settling and being balanced by BMS, they have now very similar voltages.

Cell	23h later (V) (w/Balance:On)
1	3.345
2	3.340
3	3.344
4	3.339
5	3.345
6	3.340
7	3.345
8	3.340
Delta	0.006

Should I assume that the top balance procedure was well succeeded?
I'm not comfortable with this situation since "pair" cells (2,4,6,8) did not pass the 3.3...V during the top balance procedure.

Should I connect them in parallel and let them reach 3.6V?

Thanks

If this is taken with no load on battery and voltage is below 3.45v, the cells are not fully charged and still not balanced.
Below 3.45v the cells should be closely matched in voltage with no load.

During charge absorb voltage level (3.50-3.60vdc per cell), you should hold this absorb voltage until charging current tapers down to low level to ensure all cells are balanced.

Don't be concerned after absorb charging is completed if the cells drop down to 3.45v at differing rates with no load on battery. This is just the surface charge bleeding off at different rates and does not mean they are unbalanced. If you discharge battery with an amp or two for 60 seconds it will dissipate the surface charge and they should all come down to near 3.45 vdc after 3-5 minutes of resting with no load current. That is fully charged and balanced.

 

[Andre Magro]

During charge absorb voltage level (3.50-3.60vdc per cell), you should hold this absorb voltage until charging current tapers down to low level to ensure all cells are balanced.

How this is actually made, how could I "hold" the absorb voltage?
Tks

 

[John Frum]

How this is actually made, how could I "hold" the absorb voltage?
Tks

Hope @RCinFLA doesn't mind, I will answer this one because I suspect the answer is gating your progress.

Usually a proper charger has an absorption timer and/or a tail current parameter.
Either method tells the charger when to terminate absorption.
If your charger does not support either function you can just set the float voltage to the same level as absorption.
Don't leave it this way for normal operation though.
Do this only as part of the commissioning procedure to get the cells top balanced.

If you just using a desktop power supply that doesn't have charge termination logic just leave it running after it gets to the desired voltage.
Current flow will taper off but hopefully the bms will continue to balance.
All this depends on the bms not tripping on one of the cells exceeding 3.65 volts.

 

[740GLE]

What charger are you using again?

And was this charger on all night?
Do you know what voltage was applied to the terminals? My guess is it went to float too quickly.

 

[Andre Magro]

Hi guys,

After aprox 1h of charging I started to got the same behavior from the beginning, pair cells with aprox voltages, and odd cell also with aprox voltages between them.
Why this pair/odd behavior, could the BMS be faulty?

 

[Andre Magro]

What charger are you using again?

And was this charger on all night?
Do you know what voltage was applied to the terminals? My guess is it went to float too quickly.

For this top balance procedure I'm using this benchtop power supply - 60V5A