diy solar

diy solar

A quicker way to Top-Balance? Is this idea an option?

Watzi

New Member
Joined
Jan 6, 2022
Messages
14
Hello,

I have been researching the best way to top-balance my cells (4x 280A) and came to the conclusion that most of the people connect the cells in parallel and charge them up to 3.65V with a regulated PSU.

Story why it would take me so long:
I have tried it with my 5A (set at 4.5A) PSU and it obviously will take ages. In additon it will take even longer, since I do not want to leave the PSU + Cells unsupervised (Fan already turned off at one time and the PSU nearly overheated). I have already charged one cell for around 15h from 3.28V up to 3.31V which is roughly 65-75% Soc and although the charging curve will become steeper, charging all four cells under supervision will take me ages because of my limited free-time. Money is short - buying a reasonable bigger PSU just for this purpose makes no sense for me.

Alternative idea to top-balance after some research:
I have read about another option to top-balance in a detailed article on a marine-website which takes way less time but is more labor extensive and requires full-supervision.
(I am not sure if I am allowed to post external links - please advise and then I include the link here)
It basically consists of charging up the batteries up to ~3.6V as a pack (12V bank) and then disconnect the cells and charge them up to 3.65V individually.
In this article the person directily connected it to the boat's alternator (supplying ~14,4V) and monitored each cell regulary and turning the system of if one cell hits 3.6V.

My adaption of this way:
I already have my Victron Multiplus 12/800 which can act as a battery-charger ("charger-only" mode) supplying up to 35A to a battery-bank.
One cell sits now at 3.31 (obviously will be full at first) and the rest sit at 3.28V.
My steps would be:
1) Connect the cells in series (4s - 12V )
2) Charge the battery with the Multiplus up to 14,4V (cut off voltage) and monitor each cell constantly with a multimeter
3) Turn off if one cell hits 3.6 or lets say 3.55V
4) Individually charge the cells up to 3.65V with a regulated PSU

My questions:
Q1: What do you think about this alternative methode?
Q2: Is it even safer if I connect the BMS to the battery and let it do the cut-off if one cell reaches 3.6V?

Thank you in advance.
 
Last edited:
First, this method has been mentioned/recommended previously.

You should really use a BMS when charging as a pack.
But yes, it's alot faster.

Yep:

 
How many hours did you charge that individual cell at 4.5 amps?
 
Yep:

Okay, thank you for the reference. I tried to find similar topics but did not look well enough I guess.
I will read up on this tutorial - thank you again!

Consider the topic closed. (At least for me)
 
I did not count but i would say at least 15h.

My best guess is the cells were delivered at "shipping voltage" of ~3.295 volts.
Lets say shipping voltage equates to 30% state of charge.
So each cell contained roughly 280 amp hours * 30% = 84 amp hours.
You put at least 4.5 amps * 15 hours = 67.5 amp hours into that one cell.
So that one cell now contains ~151.5 amp hours and the rest contain ~84 amp hours.
Charging them serially with a bms is going to result in the high cell tripping on over-voltage before the other cells get anywhere near full.
 
My best guess is the cells were delivered at "shipping voltage" of ~3.295 volts.
Lets say shipping voltage equates to 30% state of charge.
So each cell contained roughly 280 amp hours * 30% = 84 amp hours.
You put at least 4.5 amps * 15 hours = 67.5 amp hours into that one cell.
So that one cell now contains ~151.5 amp hours and the rest contain ~84 amp hours.
Charging them serially with a bms is going to result in the high cell tripping on over-voltage before the other cells get anywhere near full.

Good point. Probably worth paralleling them and putting them on charge for ~4 hours. That might help normalize the SoC by taking charge from the higher SoC cell and shifting it to the lower SoC cells during a charge.

'course it might make more sense to roll this into the parallel charge after the series charge. Too tired to run the numbers. :)
 
Put a 50W 12V incandescent bulb across the cell you started charging while you're doing the initial series charge. If you're watching it anyway that will help burn off some of the power, or at least reduce the amount the cell takes up, bringing them closer to equal when it hits full charge. If you've got a decent amount of current feeding them, put two bulbs in parallel.
 
If the cells have the same voltage when you first get them and you do not want to verify the capacity (by doing a full charge/discharge cycle), you can let the BMS do the top balancing. It might take a few months though.
 
My best guess is the cells were delivered at "shipping voltage" of ~3.295 volts.
Lets say shipping voltage equates to 30% state of charge.
So each cell contained roughly 280 amp hours * 30% = 84 amp hours.
You put at least 4.5 amps * 15 hours = 67.5 amp hours into that one cell.
So that one cell now contains ~151.5 amp hours and the rest contain ~84 amp hours.
Charging them serially with a bms is going to result in the high cell tripping on over-voltage before the other cells get anywhere near full.
That seems about right. I did not calculate it as you did but I estimated arround 20-30% difference in SOC witch results in 55-82 amp hours. I was aware that i would not be able to put that much energy in the rest of the cells.
I was still making up my mind if i should:
1) put them in parallel and let them equalize (like @sunshine_eggo mentioned)
2) Charge them up until that one cells tops out and connect my 5A active balancer (if the cells are in the steeper "part" of the charge curve)

I will probably go for Option 1.
Still have to read up a bit how big the equalizing current/inrush current will be and whats the best option to go to equalize the cells the best way.
 
That seems about right. I did not calculate it as you did but I estimated arround 20-30% difference in SOC witch results in 55-82 amp hours. I was aware that i would not be able to put that much energy in the rest of the cells.
I was still making up my mind if i should:
1) put them in parallel and let them equalize (like @sunshine_eggo mentioned)
2) Charge them up until that one cells tops out and connect my 5A active balancer (if the cells are in the steeper "part" of the charge curve)

I will probably go for Option 1.
Still have to read up a bit how big the equalizing current/inrush current will be and whats the best option to go to equalize the cells the best way.

Andy did this for you:


Given the tiny voltage difference, you won't see anything near these numbers. I wouldn't hesitate to connect them in parallel knowing there's only a ~70Ah different between them.
 
Back
Top