Top balance struggle

[Gould]

Appreciate there are dozens of top balance threads going, but I'm struggling with an issue now for a couple of years and hoping for some guidance. I'm happy to provide my detailed approach but to summarize, I'm using the tried and tested method that most are using; cells in parallel taken to 0 amps. I've just completed my third attempt on a 12V bank and just can't seem to get all cells to the same level. Everything goes well, with the the cells getting to about 3.64V and .30 amps on the power supply. When I put the bank back into series and connect to my charging system, inevitably 1 cell is slightly higher causing the BMS to go into overvoltage protection.

I've ruled out any 1 bad cell because each time I go through the top balance process it's a different cell causing the overvoltage. I've even attempted to top up just those cells that are lagging compared to the others but no luck. I'm looking for some guidance here. Everything is working fine and the overprotection on the BMS is doing it's job but I'd rather not have my last line of defense being used on the daily.

Is there a different approach you would recommend? Is there a smarter way of top balancing, I'm not apposed to buying something to save my sanity!

 

[Bob B]

Appreciate there are dozens of top balance threads going, but I'm struggling with an issue now for a couple of years and hoping for some guidance. I'm happy to provide my detailed approach but to summarize, I'm using the tried and tested method that most are using; cells in parallel taken to 0 amps. I've just completed my third attempt on a 12V bank and just can't seem to get all cells to the same level. Everything goes well, with the the cells getting to about 3.64V and .30 amps on the power supply. When I put the bank back into series and connect to my charging system, inevitably 1 cell is slightly higher causing the BMS to go into overvoltage protection.

I've ruled out any 1 bad cell because each time I go through the top balance process it's a different cell causing the overvoltage. I've even attempted to top up just those cells that are lagging compared to the others but no luck. I'm looking for some guidance here. Everything is working fine and the overprotection on the BMS is doing it's job but I'd rather not have my last line of defense being used on the daily.

Is there a different approach you would recommend? Is there a smarter way of top balancing, I'm not apposed to buying something to save my sanity!

How much higher is slightly higher .... What BMS are you using? can you post a screen shot of the cell voltage readings?

 

[Gould]

When I caught the overvoltage it was at 0.100 on the single cell. Screen shot shows a delta of .049 after some balancing. I'm using the OverKill BMS.

 

[Bob B]

It seems like they are settling at a different rate .... How long did you let them sit together in parallel after coming off charge?
What are your balance settings for the BMS?
Did you put the cells in a different order the 2nd time?

I personally wouldn't worry about that degree of imbalance .... they will probably come more into balance over time unless the balance settings prevent that.

 

[Gould]

I moved them back into series pretty much right after top balance. Cells are in the same order as before. I've added a couple of screen shots showing the BMS configuration.

I'll run them this weekend and see how they make out. The only issue is it's going into overvoltage with an absorption set down to 14.2V.

 

[RCinFLA]

If that is open circuit rested cell voltage after a several minutes with no load, all cells are fully charged and balanced.

Common misperception is seeing delta drop off from 3.65v top balancing and thinking they are not balanced. This is just cells dissipating its surface charge which has no significant capacity. This bleeding rate of surface charge does not relate to cell balance.

As long as cells stay above 3.43 vdc after a few hours of rest time they are fully charged and balanced. Balanced cells does not mean they are matched cells.

 

[sunshine_eggo]

Appreciate there are dozens of top balance threads going, but I'm struggling with an issue now for a couple of years and hoping for some guidance. I'm happy to provide my detailed approach but to summarize, I'm using the tried and tested method that most are using; cells in parallel taken to 0 amps. I've just completed my third attempt on a 12V bank and just can't seem to get all cells to the same level. Everything goes well, with the the cells getting to about 3.64V and .30 amps on the power supply. When I put the bank back into series and connect to my charging system, inevitably 1 cell is slightly higher causing the BMS to go into overvoltage protection.

I've ruled out any 1 bad cell because each time I go through the top balance process it's a different cell causing the overvoltage. I've even attempted to top up just those cells that are lagging compared to the others but no luck. I'm looking for some guidance here. Everything is working fine and the overprotection on the BMS is doing it's job but I'd rather not have my last line of defense being used on the daily.

Is there a different approach you would recommend? Is there a smarter way of top balancing, I'm not apposed to buying something to save my sanity!

Given the numbers, I wonder if you've set your OVP level too low.

Cell OVP should be set to trigger at 3.65V. your charger should be set to charge to 3.55V or lower. The BMS is a safety feature that should only be used when needed, not as a part of routine operation.

 

[RCinFLA]

Given the numbers, I wonder if you've set your OVP level too low.

Cell OVP should be set to trigger at 3.65V. your charger should be set to charge to 3.55V or lower. The BMS is a safety feature that should only be used when needed, not as a part of routine operation.

Except for the somewhat misguided, but common practice of many BMS's firmware requiring a cell overvoltage trip to reset SoC Coulomb counter to 100% fresh reference point.

 

[Gould]

If that is open circuit rested cell voltage after a several minutes with no load, all cells are fully charged and balanced.

Common misperception is seeing delta drop off from 3.65v top balancing and thinking they are not balanced. This is just cells dissipating its surface charge which has no significant capacity. This bleeding rate of surface charge does not relate to cell balance.

As long as cells stay above 3.43 vdc after a few hours of rest time they are fully charged and balanced. Balanced cells does not mean they are matched cells.

Apologies, daft me kicking in - what voltage are you referring to "if that is open circuit rested cell voltage"?

Regarding charge settings, the BMS is set to 3.65 and the charger set to 14.2 or 3.55.

 

[RCinFLA]

First picture you posted show all cells fully charged and fairly well balanced.

Showing 13.88v with cell voltages shown, if you applied 14.2v absorb charge, BMS should not generate a cell overvoltage trigger.

Only a poor cell terminal/bus bar connection with very high charging current or bad cell with high impedance could cause a cell overvoltage at the 13.88v and shown cell voltages starting state.

 

[Warpspeed]

Apologies, daft me kicking in - what voltage are you referring to "if that is open circuit rested cell voltage"?

Regarding charge settings, the BMS is set to 3.65 and the charger set to 14.2 or 3.55.

That is fine.
What I suggest you do is let it fully charge up to 14.2, and while its at 14.2 connect a discharge load across the highest voltage cell.
Not a lot of load. You do not say how large your cells are, but try say a one ohm discharge resistor (about 3.6 amps).
Bring down the highest voltage cell to 3.55v and stop the discharge there.

Let the whole battery go through another normal discharge / charge cycle. That high cell will still be a bit high, but not as high as previously.
Do another discharge down to 3.55v and slowly the high voltage cell will end up matching the others.
Do not try to hurry the process by discharging too fast, or over discharging the high cell.
Always discharge just one cell at a time, the highest voltage cell.

The same cure works for a low voltage cell. Do the same thing, but slowly charge it up to 3.55v at maybe three amps.
Only do one cell each time, and let the whole battery settle down and go through another full nights discharge, and a full days charging before correcting another cell or doing anything else.
Its a long slow process if you have a lot of cells (I have thirty). But once done, the cell voltages track wonderfully well, and they should stay balanced for a long time (several years) without any further attention.
The secret is go at this very gently, and be very patient, and it will only need doing once.

If you try being too clever and aggressive by speeding up the process, what was the highest voltage cell may then be a bit low, and you end up chasing your tail around and around. When you think its right, after a few days it will be unbalanced again and you get to start all over again.

 

[Gould]

Thank you for this, very much appreciated.

When discharging do I leave the pack together, still in series and BMS connected? Disconnecting other charging sources of course. Also, just bridge the resistor between positive/negative in the one cell?

FYI I have 2 - 4 cell packs, 270ah each.

 

[Warpspeed]

No need to change anything.
Let the battery fully charge so its at 14.2v and charging current has fallen to zero.
Just bridge the discharge resistor directly across the particular cell. A pair of clip leads will be fine.
Measure the voltage directly across the cell terminals.

Discharge the voltage down to the calculated average, and then disconnect the resistor.
The voltage will bounce back up a bit when the discharge load is removed, that is to be expected.
Then just let everything run normally for a day and give it time to settle down.
The improvement the first time will not be a lot, but it will definitely be there.
After a few days of doing this the voltage of the high cell will gradually come down to match the others.

 

[Gould]

Closing this out. Thank you for the great information and @Warpspeed for your practical assistance. Worked like a charm only need to discharge slightly once. Both packs are now absorbing to 14.2v with no over voltage alarms.

 

[Warpspeed]

Great news !!