The cells we bought, Batched Matched … to do that, the seller must have a factory, the relevant charge/discharge equipment, and software systems.
Not to mention how labor-intensive it can be, unpack the cells, mount, connect, run the tests, label, disconnect, pack all again, wrap it all … and don’t mess up your record system … send the incorrectly labeled boxes to the wrong customer.
Pylons don’t have adequate protection, though…
Yeah but from your inverter/MPPT params should be correct, why send the battery something it was not designed for? What happens if the battery is so flat that the BMS is off?
There is nothing the inverter can do to help with the balancing issue.
Could just be a faulty cell/s … nothing to do with the balancer?
Cause from my experience, no balancer can handle cells with issues … Andy from “down under” has done copious tests on that. In his Frankenstein Battery, I did a very similar one (12 x 100ah 2nd Life and 4 x 280ah new), the balancer will not cope if a cell/s is/are too far off kilter.
Send it in.
Complete noob question. What (if any) makes the E-tower’s protection better? Just want to educate myself somewhat when it comes to justifying the price tag. 
Is it the breakers they have?
Could be, but I could almost bet my balls that if you drop that battery down to 10% to get bottom level balance then the battery would come right.
It’s not that bad. Your cells all reach 3.45V to 3.5V at about the same time. After that, one of them jumps out, which is normal. Then it sends a CCL=0 to the inverter, it stops charging, and that causes the lower cells to drop back like that. I see it with almost all batteries.
Would be interesting to see what it does if you set the DVCC charge voltage to 55.2V. 3.45V per cell.
No, the BMS should protect the battery from a failure or user error upstream - that is it’s most important task, second to that (and related) is banancing.
From your figures I see no issue, the BMS is working as expected. You don’t have to be too pedantic about those values, the BMS really does protect the battery. In fact you’ll probably find that about 99% of new batteries behave exactly like that where one cell causes overvolt disconnect on charge for the first week or two. After that it will skirt the limit for a while and eventually settle.
Pylontech seem to have somewhat proprietary cells, which behave slightly different from other LiFePO4 I have dealt with. In the sense that they have a much lower internal resistance, though that could by due to physical design and using multiple smaller pouches. But they seem to also be much less tolerant of high voltages. Normal LiFePO4 cells can take 3.9V occasionally without missing a beat. On Pylontech hitting 3.6V regularly will cause bulging in some models. So taking that into consideration as well as the fact that they don’t disconnect on cell overvolt (on some of the BMSs) it explains why they need higher balance current.
The fact that it actually disconnects on battery overvoltage and cell overvoltage. At least some Pylontech models don’t (I don’t know if that is a firmware bug or if they figured out that by disabling it, they can void more warranty claims).
I still believe that for the price, the eTower could have had a better balancer. The eTower’s in the pics above are not mine and in sure over a few months they will kinda come right. Mine own ones eventually did after an RMA and some hard cycling.
My Pylons however don’t experience the same issues. All the cells are tightly in check and never cross 3.5v. This is before stating that the Pylons have more useable capacity.
I’m not saying the eTower is a bad battery, I just think they are overpriced for what you are getting.
Agreed.
Not related, maybe in some sense… since I added 3 more cells I got this felling looking at the numbers that the BMS/BMV is not getting the coulomb count spot on. It works, but 17kWh, it “feels” like the numbers are out.
I’m timing it to run the bank to zero SOC. BMS and BMV. Weather + LS the game to watch, as it has never been below 30% SOC (BMS/BMV).
Want to see how long it takes to get to 2.9v per cel, where the inverter will switch off … then BMS/BMS can happily “count” what goes back in.
Also to see how much is really “down under”. FWIW, 2.9v is “empty”, and 3.45v is “full” - well within parameters still.
So yeah, a “good workout” now and then is an idea to contemplate. 
Yes, I agree with that 100%. Unfortunately, I think it’s the market that allows them to price them like that.
Bottom balance at 10% won’t work with a battery that relies on top balancing, if it did anything it would make it worse.
That is not bottom balancing. Bottom balancing is when you drain the batts all to the exact same level.
Versus, you start that low AND lower the charge current on a troublesome bank, the BMS has a lot more time to balance the cells all the way back to the top again, where it is then again top balanced.
Also doesn’t really work. The balancer does nothing until it actually sees a proper difference in voltage, and that only really starts to happen above 85% SOC. The best way to balance a top-balanced battery with only a passive balancer… is multiple cycles from 80% to 100%, with a low charge current.
Personally, I don’t really like how close they (Pylontech) get them to 3.6V either. The BMS asks for 53.2V on a 15-cell model (which is an extremely jumpy 3.55V cell, if all are balanced), and then a mere 0.05V later at 3.6V, all hell breaks loose as it disconnects.
Pylontech batteries do disconnect you if you exceed 3.6V on a cell. The Force-series uses a contactor, the rest uses a big FET, as far as I know anyway.
In my experience, Pylontech batteries are a lot happier below 3.5V per cell (52.5V). This trend is also used by some other 15-cell batteries, eg Exide’s offering (not really sold in SA).
O, I must then be mistaken …
Cells 16,17,18 has never been this low. All new cells.
Now to recharge the rest of the day, balancing “rustig”. Instead of dropping the charge current, I drop, increase the max the inverter can use … right now set at 1000w.
Did that for a while now … now I’m trying the above.
FWIW BMS showed 2% before I stepped in, and BMV showed 7% SOC.
AH count is wrong on the BMS.
No. Remember you turned on that controversial setting where it balances lower down too. I am talking specifically about the simple passive top-balancing BMS in most of the more affordable batteries. Between 30% and 80% SOC, they cannot really balance, as the imbalance doesn’t really reflect in voltage. Charging slowly in that space just wastes time
BMS’es tend to start balancing at 3.2v, the default setting.
Andy and me, we start balancing at 3.45v. The cells stay nicely in sync if done properly, AND they all have the same internal resistance.
Correct, bottoms balance was not the correct term. But dropping all the cell’s voltages low was the only way that I could get my 2 batteries to work with each other. With them both on 100% one would always run away to a point where the cell overvoltage was so high that the BMS would have completely disconnect the battery.
Either way, that worked for me. Keeping them to balance at 100% didn’t.
