Charging a couple low cells in a pack that can not get to balance.

[Jmhai3]

This is part of an on going issue I’m trying to solve from the Eco-worthy 48v server rack battery thread among a few things.

How would I go about charging a couple cells (top balance) to get all the cells above 3.4 volts so they will balance by the BMS.
I watched a couple videos but they all were not specific to a 48v server rack battery.
Will I need to take it apart or can I just use my 60v Drok power supply and start at one cell and charge to ???
What would be a good volts to level them all at so I can see if the issue I’m seeing is a BMS problem or bad cell.

The issue is that no matter how much I cycle the battery one cell (12) will take off and another will lag behind (cell 5) so bad that the fast one hits the covp before 4 or 5 other cells get to 3.4 volts (best I could do). The slowest one will be 250 to 280mV part from the high one which will be atleast 100mV above all of them when it hits covp.

Thanks for the help.

 

[Rednecktek]

Either charge the low cells to 3.4, drain the high cells with an old car headlight or the like, or set the "start balance" voltage lower, to like 3.0 and see if it can balance itself out. Make sure if there's a setting like "Balance on Charge" or the like that it's set to OFF.

 

[wpns]

Did you ever try holding the battery voltage at a point just below where the high cell trips out on sell over voltage? For like a week?

 

[Jmhai3]

Either charge the low cells to 3.4, drain the high cells with an old car headlight or the like, or set the "start balance" voltage lower, to like 3.0 and see if it can balance itself out. Make sure if there's a setting like "Balance on Charge" or the like that it's set to OFF.

Waiting on the password for the software so I can make changes. That and the sleep time and function I want to see I can set to off. Don't want it to sleep shut-off if not in use as a UPS.

 

[Jmhai3]

Did you ever try holding the battery voltage at a point just below where the high cell trips out on sell over voltage? For like a week?

Yes, several times I brought it up slow and try what you suggested by increasing 1000mV every time it settled down.

It will balance the cells within a few minutes of running under load (discharging) down to 4-5mV difference. So BMS is working it appears.

The issue is once it gets close to 53v the difference starts to increase exponentially and the run away cell increases how fast it charges while the lager (5) stays put. Even if I keep it low with very small increase the high cell (12) will not stop no matter how low I set the incoming charge. It is like it has no control and incoming charge that should be split between the cells is being gobbled (Happy Thanksgiving joke) up by the one cell and leaving every one else behind.

The BMS looks like it is doing the opposite it should be, by slowing down the charge of the lowest cell and increasing the charge of the highest.

Maybe it is a bad cell or two in the pack, OR they built it with batteries no where near the same voltage to start, (5) and a few others were never cycled correctly from the get go so they don't know they are not maxed out. I hope if I top balance the low cells they will learn a new individual SOC and align with the other cells.

If they will get to top balance but after a couple cycles are way out again even adding a active balance like BaracudaBob suggested could not be out of the question if EW wants to throw the cost of it back to me for my troubles.

Until I get a return shipping label, I am going to try what I can to correct my issues.

 

[wpns]

Yes, several times I brought it up slow and try what you suggested by increasing 1000mV every time it settled down.

Not 1000 mV (one volt), 100mv (one tenth of a volt)

It will balance the cells within a few minutes of running under load (discharging) down to 4-5mV difference. So BMS is working it appears.

The fact that the cells come into line while discharging is not a surprise. They all do that. Has nothing to do with the BMS balancing function. Discharging is not balancing.

What voltage can you hold it at in order for the one cell to not trip out? 40 V? 45 V? What is the number?

 

[Jmhai3]

Not 1000 mV (one volt), 100mv (one tenth of a volt)

The fact that the cells come into line while discharging is not a surprise. They all do that. Has nothing to do with the BMS balancing function. Discharging is not balancing.

What voltage can you hold it at in order for the one cell to not trip out? 40 V? 45 V? What is the number?

Yes, you are correct, typo.
Sorry for the late reply… Friday and night was gallbladder attack (stone) fun. Still recouping from the pain and no sleep.

Left it on a charge set at 54.4v @ 3a and right up to 90% SOC it maintained a 4-5mV difference. Once it got close to finishing it started to show the mV separation from there and went to idle @ 54.4v and a 130-140 mV difference.
Cell (5) 3.368, Cell (12) at 3.498 this sat at this for 10 hours while I enjoyed the evening.

I gave it another 100mV on the Drok and it charged a few more minutes and cell (5) never gained and all the charge went to cell (12) to finish at 3.520

This is the same result I get every time. I usually end up with 12 at 3.52-3.54 half the cells within 100mV of it and the rest 200mV with cell 5 the lowest every time.

 

[wpns]

Yes, you are correct, typo.
Sorry for the late reply… Friday and night was gallbladder attack (stone) fun. Still recouping from the pain and no sleep.

Left it on a charge set at 54.4v @ 3a and right up to 90% SOC it maintained a 4-5mV difference. Once it got close to finishing it started to show the mV separation from there and went to idle @ 54.4v and a 130-140 mV difference.
Cell (5) 3.368, Cell (12) at 3.498 this sat at this for 10 hours while I enjoyed the evening.

I gave it another 100mV on the Drok and it charged a few more minutes and cell (5) never gained and all the charge went to cell (12) to finish at 3.520

This is the same result I get every time. I usually end up with 12 at 3.52-3.54 half the cells within 100mV of it and the rest 200mV with cell 5 the lowest every time.

OK, so what’s the highest pack voltage you can set on the DROK power supply that keeps the highest cell under the COVP point (of I assume 3.65V)?

 

[Jmhai3]

OK, so what’s the highest pack voltage you can set on the DROK power supply that keeps the highest cell under the COVP point (of I assume 3.65V)?

54.7v if I go another 100mV it will push cell 12 covp and it will keep going and go into protection.

 

[wpns]

54.7v if I go another 100mV it will push cell 12 covp and it will keep going and go into protection.

Ok, and if you hold at 54.7V does anything change in 24 hours? 48 hours? 72 hours?

 

[Jmhai3]

Ok, and if you hold at 54.7V does anything change in 24 hours? 48 hours? 72 hours?

Cell 12 will slowly creep up higher every time it goes from idle to charge till it hits covp and protection.
I think tonight sometime it will do it. Will know in the morning, if it does not shutdown for no reason which it has done before too.