Help balancing 16 cells

My goal is to top balance qty 16 304Ah CATL prismatic cells that are currently all sitting at 2.90v except 1 cell that measures 2.88v.

My question is how to achieve this efficiently?

The 16 cells will eventually be wired in groups of 4 series connected cells and each of the 4 "12v" packs with it's own JK balancing BMS

The resulting "12v" packs will be connected in parallel, and feed a VIctron Multiplus II, 3kw inverter.


Is it best to top balance all 16 cells at once in parallel with a 3.6v high current lab / bench power supply? This will require many days / weeks /months

Or, can I wire my 12v packs (4s), connect each pack to BMS, then connect the 12v packs in parallel to the Victron Multiplus. and charge at 12volts, letting the BMS do the balancing?

JK BMS is limited to 2A of balancing current and since it doesn't simultaneously draw from the high cells and transfer to the low cells, its only operating at a 50% duty cycle. Realistically you would need 4 high current balancers, like in the 10 to 15A range to balance the fully assembled battery in a reasonable time.

Probably easier and much less cost to connect cells all in parallel and charge at 3.5V first.

With four BMS with 2a active balancers I would assemble the batteries and place in service. No specific top balance in advance.
Possibly need to limit the charging voltage while the balancing completes. Cells will always be protected with the BMS.

Wire batteries in series. Charge at 55.0v until it reaches 55.0v. Repeat at 0.25v increments (55.25, 55.5, etc.) until you reach 57.6v).

The first time the BMS cuts out (cell over voltage):

Option 1:

Disassemble and measure the voltage of each cell. Charge the lowest to the next lowest's voltage. Wire in parallel. Rinse and repeat until they are all connected and charged to 3.60v.

Option 2:

Set inverter to charge rate of 2 amps. On BMS, set high cell over voltage to 3.48 v, resume charging at 3.47v.
Set charger to just under cell over voltage. Let BMS balance. When High cell goes to 3.630 volts or lower, Raise charge voltage by 0.10v. Rinse and repeat. High cell should range between 3.630v to 3.645v during the process (under the 3.65v limit for LiFePO4).

Thanks so much for the help!

15 of 16 cells measure 2.90V and one measures 2.88V. If I look at volt / current graph typical for this chemistry, it looks like my cells are only charged to about 10% capacity.

My cells are 304 Ah, so am I correct in estimating each cell needs about 270 amps to fully charge?

I have the EBC-A40L 40 amp charger @ 3.45v. 270A x qty 16 cells in parallel = 4320 amps, so would take 180 days? Can that be right?

I also have the bench supply below which is supposed to do 60A @ up to 15 volts. If I charge in groups of 4 in series @ 13.xx volts, I'm a little confused with the math; does each "12v" pack need about 270A to fully charge @ 13.xx volts?

Also, would it be better to charge all 4s "12v" packs at the same time connected in parallel with the 60a charger set @ 13.xx volts?

304AH * 16 = 4,864 AH (@ 3.2v nominal)
4,864AH / 40A = 121.6H (5 days)
The fastest way to charge them would be to connect them in their final configuration (4s4p) and charge them at your desired voltage (i.e. 3.45v = 13.8v)
304AH * 4 = 1,216 AH (@ 12.8v nominal)
1,216AH / 60A = 20.3H or about 5 hours per pack. (Note you are charging 4x faster by putting them in series, if you charge at the same rate (i.e. 40a) then it would take 1/4 of the time. 30hrs vs 120hrs. Since your 12v charger is also 50% faster, you get 20hr instead of 30hrs)

As mentioned above, the easiest/fastest way to top balance them is to connect them to the BMS and charge them. Because of the flat charge curve of LifePo4, you can charge to 80 or 90% before the voltage starts to diverge. So just charge up one pack to say 13.2v at full speed and see how close the cells are once you hit 13.2 (you don't need to let them soak at any voltage during this process). Then increment up to 13.4v, etc. Once you start to see the cells diverge, you can either slow your charge rate to 2amps or slower so the bms can balance and charge up to a final voltage of 14.4, or disassemble your pack and top balance the cells in parallel at 3.6v (at that point they will be 90% full and won't take days to reach 3.6v). If you do this make sure to monitor frequently so you don't accidentally go over 3.65v in case the bms is misconfigured etc.

Over the years I've built a few batteries.
I take them in sets of 4, set them in parallel and charge the cells starting @ 40A to a max of 3.65V and allow the CC/CV Charger (bench unit) to fully saturate the cells down to the point where tehyare taking 3A only. I set those aside, leaving them connected in parallel while I charge the next set up. One my sets are done, I link the "packs" worth of cells together in parallel again and let them get back to 3.650 BUT stop at EndAmps/Tailcurrent and then I take it all apart and put into case/box and setup the rest. Typically by the time I am done, the cells are usually sitting around 3.5V and ready to go. LFP ALWAYS SETTLES and so this is perfectly normal & acceptable.

As I use JKBMS' I set Active Balancing to start @ 3.42V and allowing only a 0.010 deviation and this works well consistently.
The charge profile I use charges Bulk/Absorb to 3.450VPC and transitions to FLOAT when EndAmps/Tailcurrent is triggered and continues at 3.437VPC.

VPC = Volts Per Cell.
EndAmps/TailCurrent calculation examples: 100AH X 0.05 = 5A or 280AH X 0.05 = 14A
* NOTE * When multiple packs are in Parallel, use the EndAmps for the Largest Capacity Pack within the bank. This can be "Tweaked Down" a bit to accommodate deeper charge levels if you have a few "quirky" packs. NEVER EVER go below the EndAmp value of the lowest AH pack in the bank. A potential side-effect is larger packs get over saturated resulting in some cells "running" and causing HVD (High Volt Disconnects).

Lasertrapper said:

Thanks so much for the help!

15 of 16 cells measure 2.90V and one measures 2.88V. If I look at volt / current graph typical for this chemistry, it looks like my cells are only charged to about 10% capacity.

My cells are 304 Ah, so am I correct in estimating each cell needs about 270 amps to fully charge?

I have the EBC-A40L 40 amp charger @ 3.45v. 270A x qty 16 cells in parallel = 4320 amps, so would take 180 days? Can that be right?

I also have the bench supply below which is supposed to do 60A @ up to 15 volts. If I charge in groups of 4 in series @ 13.xx volts, I'm a little confused with the math; does each "12v" pack need about 270A to fully charge @ 13.xx volts?

Also, would it be better to charge all 4s "12v" packs at the same time connected in parallel with the 60a charger set @ 13.xx volts?

https://www.amazon.com/dp/B0C6JV1VYK/ref=pe_386300_440135490_TE_simp_item_image?th=1

Click to expand...

40 amps into 4320 = 108 hours or 5 days.

Charge all four assembled batteries at the same time if they are ready. Although if you assemble one battery every day or three it is perfectly fine to charge them individually as you go.

The bottom balance measured here gives no indication of how top balanced they are or are not already.

Thank you, thank you, thank you!!!!

I'm hearing that most practical / efficient way forward is to assemble my 4s packs with BMS connected.

Then I can charge each pack individually, or go for broke and connect the 4 packs in parallel, then charge.

I can charge either single packs, or all 4 in parallel with either my bench power supply, or my Victron Multiplus II.

Would there be an advantage to one over the other?

When you buy cells, they've been shipped to vendors after the factories perform their validation tests and put them into storage, usually at 3.2V. These can sit in warehouses for months, even years before they get to sale. We are still seeing people get shipped cells made in 2019. Think B & Bulk grades. Nothing wrong with that if properly stored. Manufacturers therefore generally recommend that "New" cells or those that have been in storage for > 6 months be top charged to ensure the electrolytes are fully activated.

There has been a lot of discussion around this and numerous other points related to Top Charging & Top Balancing - a few Massive & Excrutiating threads on it (tylenol required LOL). I have personally tried a few ways to decide for myself & test it out, and guess what, KISS Applied is best. I outlined how I do my packs, I have quite a few out there (other people using them) and 99.5% of the time no issues. BMS' can poop or even a cell after a few months. BTW ! being able to replace a cell "if" ever required is also a thought to consider when building your pack. Let's just say that my first set of packs, I built the boxes and did it in a way which did not account for "polite disassembly" and it wasn't easy.

Lasertrapper said:

Would there be an advantage to one over the other?

Click to expand...

Time and cost. Inverter charging is usually faster than bench powersupply.
You can do it with BMS and Inverter, and avoid the bench powersupply, but that takes a lot of time (holding at top and waiting for BMS to balance).

Thanks!

So I should charge my 4 packs (4s) through the BMS's with my bench power supply?

DIYrich said:

Option 2:

Set inverter to charge rate of 2 amps. On BMS, set high cell over voltage to 3.48 v, resume charging at 3.47v.
Set charger to just under cell over voltage. Let BMS balance. When High cell goes to 3.630 volts or lower, Raise charge voltage by 0.10v. Rinse and repeat. High cell should range between 3.630v to 3.645v during the process (under the 3.65v limit for LiFePO4).

Click to expand...

This is the most reasonable answer to your problem. OP

Lasertrapper said:

Thanks!

So I should charge my 4 packs (4s) through the BMS's with my bench power supply?

Click to expand...

Bench supply is for parallel charging

My 60amp bench supply is adjustable 0-15vdc so I could charge 4s packs, no?

Lasertrapper said:

My 60amp bench supply is adjustable 0-15vdc so I could charge 4s packs, no?

Click to expand...

Only with a bms.

Yes, with BMS, but is it better / easier / safer to top balance 4s BMS packs with my bench power supply or Victron Multiplus?

Inverter is more complicated....

Lasertrapper said:

Yes, with BMS, but is it better / easier / safer to top balance 4s BMS packs with my bench power supply or Victron Multiplus?

Inverter is more complicated....

Click to expand...

Bench takes more physical work.
Inverter takes more time, but you build the battery once.