Would be nice if we all have similar cells, similar equipment, to agree on the best settings that we can all use and see how the banks perform over time, adjusting as needed.

Balancer: Start Voltage: 3.200v Delta to Balance: 0.010v Balance Only when Charging: Yes

BMV: … if one has one … Charge Efficiency: 99% Peukerts Exponent: 1.05 Charged Voltage: 0.3% below Absorption Tail Current: 4% (~2 to 4%) Discharge Floor: 20% Charged Detection Time: 3 (~2 to 5 min.)

5kva MPII: State of Charge when Bulk is finished: 95% Charge Efficiency: 0.99 DC input low shut-down: 2.9v - 46.4v DC input low restart: 3v - 48v DC input low-pre alarm: 3.1v - 49.6v

Absorption: 3.55v per sel - 56.8v Float: 3.4v - 54.4v Charge: with 280ah, max inverter can do, 70amps Repeat absorption: 1.00 Repeat absorption interval: 7 Absorption time: 1

With the 280ah EVE, I want to get it right … so with 17/18 cells, the above will have to be adjusted, as Victron can handle 38–66V, so methinks for 18 cel bank to go for 3.4v x 18 = 61.2v

@Louisvdw is incorporating the BMV midpoint measurement, so an even number of cells will be better.
Also, in terms of physically arranging the cells, an even number is neater and more flexible.

In terms of tweaks, (if my understanding of @Louisvdw’s script is correct) is that at 98% of capacity the charging rate is hard-coded limited to 1 Amp. I think that should be equal to the balancing capability of the BMS. (Which I think is 160mA for the JBD BMS).
Otherwise, the highest cell voltage will still climb and result in a shut-off.

FWIW, I am starting to understand why brand names differ so much in their settings. It depends on what cells they are using and how many.

From that they work out exactly what the settings must be, to ensure longevity for “those” cells they opted to use.

So the ideal is to get to a point where if one uses like say 280ah EVE cells, with XYZ BMS, then set the system and BMS settings to so, and one should be sorted for a very long time.

We don’t have a clever highly trained department for that, but man, we do have a lot of experience here on this forum, to figure it out.

You can spesify a very small charge amount but that does not mean the charger can give you such a small limit. Initially when I did test I found that smaller that 1A makes no difference.

@Phil.g00 you comment about the midpoint graph does not make sense to me. With lithiums we have the values per cell. Any midpoint is just a fun graph for a lithium battery. You already have much better in the min/max cell graph.
Physical arrangement is easier with even numbers, but it is also doable with odds. 15 cells goes in nicely 3x5 (but nothing can help with 17 cells - you will have something off somewhere)

As you state this is probably redundant for most people, but it seems that you have encountered the request and are incorporating it, so I assume somebody somewhere wants it.

I was really addressing @TheTerribleTriplet suggestion of 17 vs 18 cells.
However:
The busbar arrangement gets messy adding the third row, two rows allows for neat +ve to -ve busbars. The first row of cells can turn back on themselves, but turning back for a third row the positive and negative terminals of the batteries are no longer adjacent.
Doable, but not as neat.

And we “bought it”, hook line and sinker, as it makes 100% sense, large ah, running it “lightly”.

Brandnames with 15 x 100ah cells in, I think we are putting together the mother of all banks.

Another “joke” we make is if one has:
18 cells, and for whatever reason, one goes, take it out. 17 cells left, so another one goes as you now realize “o my, what did I do”.
16 cells left … "why me!?
… can take one more out for a 15 cell 48v bank.

Technically I’m a Virgo too (if you believe that stuff), and I kinda like odd numbers. What I like more than odd numbers are prime numbers. I like them because I know other people tend to avoid them, so they make numbers that tend to be unique. Especially useful when you’re choosing subnets for your internal networks, with the future potential for VPNs and bridging such networks, I’m pretty sure there are comparatively few people out there running 172.16.13.0/24, because 1) for some reason 10.0.0.0/8 and 192.168.0.0/16 are more popular, and 2) people tend to not like 13…

Palindromes! I spot palindromes on car number plates while driving.

For other things to keep you awake at night, look at the Collatz Conjecture, or the Hilbert Hotel.

But back when I picked a subnet for my home, I somehow picked 192.168.8.0/24, because I thought it would be faster to type (you already have one finger on the 8), completely forgetting that 10.0.0.x is even faster…

So yeah… different day, different rules. Spotty consistency.