At this point, since your new cells are already beginning to “peek” above the nominal 3.2V-3.3V per cell (they are around 95% full I would guess), and since it is stable with nothing jumping out… I think it might be interesting to leave it alone and see what the balancer in the BMS can actually do, and how long it takes.
That information would be more interesting to me than the actual process of getting it there faster.
As Andy saw over the years, it still takes a very long time, the 5a, with large aH cells as I have. His 5a balancer in one video, where he realised 10a at 48v going in, the balancer trying to balance at 5a, it is “overpowered”. Once the cells are closely matched, THEN it works quite nicely, over a few days, week.
I could do that yes … but as mentioned, these big aH cells, at 260mA balancing, will take a very VERY long time … read, I cannot comfortably leave the house until I KNOW the bank is settled …because of previous horrible experiences.
More importantly, I need to know how to do it quickly and efficiently. I’m basically doing what a brand name supplier will do on a bank still under warranty, kinda.
I going where no Lifepo4 owner has ever been. Nowhere have I seen similar done, bar here, because of the huge aH involved.
Cell 17 is being charged under Lipo setting:
The rest is “slowed down” balancing whilst discharging, LS and clouds.
GOTCHA!!!
NOW I’m getting somewhere fast! LiLo setting is working much better!
Bank stops charging, 16-18 far behind, then shoots to 62.85v whilst balancing.
Definitely not a good idea to push those three cells hard for weeks whilst the BMS tries to cope.
Victron is now balancing the bank to 3.45v, flot volts.
Cell 17 is being charged with the EV Peak. Matters none that it is “balancing” as the EV Peak is connected to the battery direct.
Why is cel 17 still charging?
EV Peak maxes out at 250000mAH. The moment it gets close to that, it stops.
Seldom charges above 3a
LiLo setting, now it charges at 6a at ± 3.78v.
If I have to leave, I simply switch off the EV Peak, as one MUST be around to monitor it.
Will go to 3.50v max.
Question for @TheTerribleTriplet You are quiet. Either it’s a good news “quiet” or the new pack is 18s GIGA 48v is keeping you VERY busy!
Bah … what is this GIGA thing you speak of?
Do you mean this?
And that after I said something about FOCUSSING … ???
Ok ok mense, Relax! … 4.2v is where the cell goes BOOM … and I can say:
"I let the smoke out!! - of a Lifepo batt.
"I let the smoke out!!
"I let the smoke out!!
Na na Nana na na, yeah yeah I’m a w … ok, enough of that.
I’ve got this ok!!! …
Why did that cell shoot out?
EV peak counts the amps going in right? The max is 250ah, I need 280ah.
Sooooo, what if one was to set it to LiLo, at 180ah at 3.85v, you can get mos close to 280ah right (?) Ah x volts thing.
… but ONLY if the aH counter is not resetted … I forgot that, must stop at 220ah not start when the bank is full and then the aH count starts at 0, a teeny weeny peeny little detail see … do NOT do this at home unsupervised. Just a Tittibitty that … and IF you do … Do. NOT. WALK. AWAY. or read a forum … nope, don’t
The big drama:
LS + bad weather + ones BMS keeps on balancing when discharging, the correct technical term … is a “cluster f…”.
The whole week was like:
Set Balancer manually On, Balancer Off, On, Of …
Having to keep a damn close eye on schedules, or balancer will mess it all up, by setting starting volts higher when discharging and lowering it before the sun takes over to get a longer “balancing session”.
PITA I tell you. PITA … if I could have spoken to the tea girl I could have asked for some tea … 
.
Got BMS fixed AT LAST this afternoon. 
When the supplier said: My engineer says it works … yeah right, let me speak to the tea girl then please, she knows more …
Ok, now that that is out of the way …made LS work for me:
Started to drain at 4kw (have to wangle loads with leads) during the LS … did you know our dishwasher, induction plate, air fryer, MW can all work nicely, one at a time, sometimes 2 at time on a 4kw inverter? 
Recharge at max after LS is over … over and over … still a hell of a lot faster than parallelling.
Ideal settings are:
10a at 3.35v - discharge as low as I can … 17kWh is a LOT to discharge.
Recharge at 10a at 3.38v.
etc.
Ps. It is better to balance cells with a constant current, Eskom. With panels, it takes longer.
And I’m getting there … faster and faster.
I’m up to 3.4v tonight.
So, now to see how the "fixed’ BMS behaves … for now.
Also found this gem … whilst watching the BMS all on my own, alone.
So be patient young one (ito dabbling with this stuff) … I should not have said: Here, hold my beer … it is warm now, isn’t it?
That 3.80V column looks very fishy?
Now with the BMS fixed, I can relax … actually, in true TTT style, I got one teeny logic wrong.
Note: I asked Overkill Solar + another SA importer of these BMS’es, both told me they have seen this behavior now and then with no idea wh, nor dit the App developer for Apple version know why.
I now know why … will inform them.
Running on Keep Charged all the time, to make LS useful.
To test at this high SOC, asked Shelly to please switch on the 3kw geyser for us, to see if the balancers are off.
Charging at max, balancers start near instant.
So, now that my biggest pain has been solved, and because I don’t have a proper charger, @plonkster, let’s see how long the BMS now takes, as you asked before.
We start here:
Ps. Some 10-100w is still going into the batts, way to low for the BMS to register.
Note: Boosted cells 16-18 a lot manually, they used to be sitting at ±3.34v a week ago, at 100% SOC.
But it is not really 100% SOC in this case, the cells were not top balanced at the correct aH and volts, the EV peak limits.
Man, learned a bit here to this point.
Update, sorry @plonkster the wait it out will take “months”.
Been upping the bank volts, starting with 3.3v then 3…35v etc, using Eskom.
Currently on 3.45v, think I must drop it to 3.42v, as this is the first time on 3.45v.
It always gets here quite nicely - read “bottom balanced”.
The moment we get here, it derails properly! - Ignore the BMS SOC, it is WRONG! BMV says 100%. I trust that.
Then, “The Bulbs” are applied, batches of 4 cells, 1-4 currently … the PITA process … to help the BMS.
Think I should invest in a Neely.
Two things are a fact:
These big cells need big wattages to charge.
These big cells need big balancers to balance.
These big cells need do not like small wattages.
See, it works pretty fast. - The Bulbs
And if I need to “o sheeet” moment … switch on the oven. 
Again:
It works well, The Bulbs … but I cannot leave the system with so much power available…
I’m getting a Neey … this is why:
But it will take time …
Mooi - basically the same thing but with lots of options for you the “test”!
Watch from here on the Heltec … I NEARLY bought on earlier:
@Phil.g00 , can you watch this video and explain it to me, I recall you once explained that one must always only balance when charging if my memory serves me correctly.
No BMS that I’m aware of, can see ±<100w charging still happening when the batt is full. Can see that on Victron that it is still being charged … ALWAYS a problem at the end.
Now I know why.
@Louisvdw watch this
It’s a custom BMS from the Auz supplier. Maybe its available direct from China but it looks great!!! Big busbar’s etc.
Without watching the video, I will give you an answer that hopefully helps. Here it comes…
“It depends”
With a passive balancer, you can only balance while charging. That’s because the balancer is merely little bypass resistors that sends some current past a high cell, allowing the low ones to be pulled up. They only work while current is actually flowing.
With an active balancer, you have more options. In an active balancer you have a circuit that literally moves charge between cells. It connects to a high cell, slurps off some charge, then connects to a low cell and dumps it. These BMSes are more expensive.
On an active balancer, you can balance at any time, although it is typically still unwise to attempt balancing during hard discharge. An active balancer will typically be active while the battery is at a higher voltage, moving charge around, and it can do so without needing a charge current.
There are two kinds of active balancers. The one uses a large bank of capacitors to transfer the charge. That balancer linked earlier looks a bit like that. The other kind uses an isolated DC/DC converter, and an inductor for interim storage. For those BMSes you will see a big old coil somewhere on the board, and a row of FETs to connect to each cell, but you won’t see the string of balance resistors you see in passive balancers.
The additional cost of an active balancer, over the life of a battery… unclear if it is worth the bother, honestly.
Now, on that topic: If you have a balancer that relies on seeing a charge current with its own current sensor, and that sensor fails to see a current below 100mA (as is often the case)… that sounds like somewhat of a disaster.
That Change Balance setting on the JBD is a bit confusing, but it is actually a great tool.
In normal opperation you will only want to balance while charging. (Off setting) This will keep your bank balanced as you would expect. JBD is a pasive balancer, and like plonkster explained it can’t move energy, only remove energy from cells.
Now when you set up a new battery you need to balance the cells. This is where the Charge Balance setting can come in handy as it will balance when not charging (only for your first set up of the battery). You can charge the battery as much as possible and then stop the charge and use the Charge Balance option to balance the cells. After a while you charge the battery again and then again stop and leave the BMS to balance. Repeat until all is balanced.
This is the same as using a resistor or light bulb to reduce the high cells.
If I recall correctly, you balance whilst charging simply because 100% charged is the finish line that each cell should cross simultaneously.
With a balancer that can only do milliamps, if you try to keep thing perfectly level on the discharge to are on the plateau of the curve, where a few mV can be 10’s of % difference in SOC.
Without the balancer interfering, it is unlikely that the SOC would drift like this during a normal discharge. A normal charge would see everything cross the finish line together.
If you try to balance during discharge, then the balancer has to undo everything it has done on the way down when the battery recharges.
Correct, they are now “planning” to get a 48v one made IF there is enough orders.
I have followed that titbit to the letter. See below the PROBLEM.
He tested 2 batts, one ON, one OFF, I see EXACTLY that on my BMS, when ON.
You say exactly the same as Andy and many many others.
Has anyone tested that? I have never seen the results over the years in real life.
So Andy did the test for all of us over 3 weeks or whatnot comparing the 2 batts next to each other.
Go and watch the vid. 
I’m now testing Andy’s test.
The PROBLEM is every single JBD BMS I have used, 12v and 48v, after setting the settings a few times, the software stops working as advertised.
You can set the Charge Balance On or Off … it keeps Charge Balance ON in the BMS, and the screen shows OFF.
To “fix” that, you push the balancing start volts to 3.45v, to NOT balance when discharging … or switch the Balancing Off when there is LS … and remember to switch it back on.
Spoke to the supplier, pictures and all … everything was lost in translation … just a “My engineer says it is working.”
Bottom line, where I’m heading … using my JBD BMS with a “Neeeeeeeey”.
You then have a Big BMS with a really cool feature added, Neey, the 4th version, for the manufacturer listened to all of the communities’ comments and replies, adding the features they all needed.
