diy solar

diy solar

Lifepo4 balancing question

meetyg

Solar Enthusiast
Joined
Jun 4, 2021
Messages
1,072
Hi there.
I am planning to build an 8S 24v 40ah Lifepo4 pack for a mobility scooter.

I am planning to use only 80% capacity as a charge-discharge cycle, for extended pack life.
I intend to do this by charging only up to 90% SOC, and discharging to 10%.
My question is: Will cell balancing work on this setup?
From my understanding, balancing usually occurs when cell voltage is in the upper knee (above 3.4v or so...).
I am planning to use a JBD BMS, so that I will be able to tweak the parameters.
But will the BMS be able to balance (during charging) on the flat curve? Or will the voltage delta be so small, that it wouldn't see the difference?

Of course I will properly top balance my cells before first use.
Will the cells get out of balance in the 80% DOD at all (assuming that they are of moderate quality)?

Thanks.
 
Cells only require balancing if one cell reaches its cutoff voltage before the pack capacity is reached.

Once in service, the differences in self discharge will determine how often the cells need balancing. Generally a passive balancing BMS will be greater than this self discharge difference and will keep an individual cell from reaching cutoff voltage before pack capacity is reached.

Don’t parallel top balance new cells, it’s a waste of time. Assemble your pack and if one cell reaches high voltage cutoff before the others, remove pack load and individually charge the remaining cells (without dismantling the pack) to the cutoff voltage of the BMS.

If the pack reaches your target voltage without an individual cell reaching cutoff voltage you don’t need to do anything further.
 
Thanks.
So from what you are saying, it would be OK to set a start-balance voltage at around 3.25v and cell over voltage to 3.3-3.35v ?

I actually ordered 12 cells for my 8S setup, just in case one is damaged, or another has a lower capacity. If they are all OK, I will keep the 4 extras as spare, just in case a cell goes bad in the future. I will be testing all 12 cells, and I will choose the 8 ones with best capacity.
I'm still not sure how I'm gonna too balance all of them. Doing so in parallel might take ages, depending on their initial SOC when they arrive.
But on the other hand I will be using an 8S bms, so just building a 24v pack won't work.
Maybe I will do a 6s2p build and charge it up like that, assuming the JBD 8S BMS can be configured that way.
 
If you are going to capacity test each cell, it is better to get a capacity tester and run each cell through a capacity cycle. You will know the capacity, and all cells will be left at the same SOC.
 
I think this is bad advice.
IMHO the keyword here is parallel and I agree with @toms, especially if you receive decent matchrd and batched cells.
Even for the crappy mystery meat cells I would charge them as a pack first and see what the differential is at the top before I decided the next step. If the cells are close enough at the top and the pack pulls its capacity and the cells are close enough at the bottom its becoming increasingly popular to forgo any top balance at all.
Also topping off the cells individually while they are assembled in the pack is much quicker than a parallel top balance.
Cell capacity has increased and 16s packs are becoming more and more prevalent.
Bench PSU's are not keeping up their end of the arms race.
 
Last edited:
IMHO the keyword here is parallel and I agree with @toms, especially if you receive decent matchrd and batched cells.
Even for the crappy mystery meat cells I would charge them as a pack first and see what the differential is at the top before I decided the next step. If the cells are close enough at the top and the pack pulls its capacity and the cells are close enough at the bottom its becoming increasingly popular to forgo any top balance at all.
Also topping off the cells individually while they are assembled in the pack is much quicker than a parallel top balance.
Cell capacity has increased and 16s packs are becoming more and more prevalent.
Bench PSU's are not keeping up their end of the arms race.
Yeah, that's my problem. My power supply is rated only up to 10 amps.

I will take your advice and first try to charge as a pack (again, probably as 6s2p just to cover all 12 cells). Then I will see if additional top-balance in parallel is needed.

I am really unsure what I'm getting in regards to quality.
The seller (on Aliexpress) told me that they are Grade A+, but I guess he would say just about anything to get me to buy.
These are the cells I ordered:

I did contact Amy, but unfortunately they aren't able to ship to my region currently (Middle-East).

These sized 40ah prismatic cells are pretty rare. I have seen some Littokala ones, but they look different. So I am hoping that there is less of a chance of grade B cells.
These cells are the largest capacity that I could find that would still fit in the fairly small battery compartment of the mobility scooter in an 8S configuration.
Even the Headway cells are too long at this amperage. I don't have room to parallel Headways either, so I went with these prismatic.

I'm planning to do a full capacity test for each cell using a DL24 capacity meter (similar to the cheap one Will uses, but a newer and improved generation).
It looks like this:
20211105_125427.jpg

Anyways, do you guys know if it's possible to do 6S pack with an 8S JBD BMS?
 
Last edited:
I will take your advice and first try to charge as a pack (again, probably as 6s2p just to cover all 12 cells).
6s2p is two discrete batteries with a nominal voltage of 19.2 volts.
I think you have the nomenclature a bit mixed up.
If you are going to have cells in parallel within a battery then you really should top balance.
Then I will see if additional top-balance in parallel is needed.

I am really unsure what I'm getting in regards to quality.
The seller (on Aliexpress) told me that they are Grade A+, but I guess he would say just about anything to get me to buy.
Grade A+ is what all the vendors say.
A+ means it was factory tested including a ~30 day test to measure self discharge and the cells are matched by capacity and internal resistance.
That costs premium cash.
These sized 40ah prismatic cells are pretty rare. I have seen some Littokala ones, but they look different. So I am hoping that there is less of a chance of grade B cells.
I've heard that name before and it does not fill me with confidence.
Anyways, do you guys know if it's possible to do 6S pack with an 8S JBD BMS?
6s is not a valid configuration with a typical 8s bms.
 
Yeah, I heard and seen some bad reviews about Littokala cells, not to mention some shady sellers.

That's why I got mine from the Mr.Li store.
They claim they are the cell factory, but again, words these days don't mean much...

They have overall good reviews, but these specific cells weren't ordered much, so they have few (but still good) reviews.

Anyways, I was thinking of the 6s2p setup just for the initial charging phase.
In the end my pack will be an 8s 24v, with 4 cells to spare/as backup.
 
Back
Top