Maximum recommended cell size.

[CaptainRivet]

Originally Posted by wholybee

I think the top tier would be Winston, Calb, and Sinopoly. Eve might be the top of the cheaper cells that share the same 280Ah form factor. I would consider EVE 280Ah cells, but prefer one of the top tier. I wouldn't consider any other brand of 280Ah cells.

I agree with Jedi, 4 cells in series at the capacity you need is the preferred route. Not always possible, but that would be the ideal.

Top notch is Winston and they are no LifePo4, the bigger sinopoly cells >=400AH are made by Winston for sinopoly on lower specs then Winston cells, smaller are made by sinopoly themselves.
Calb is top Tier LifePo4 for EV, EVE is top tier for storage and low spec EVs.
For >400AH the loss if one cell goes bad is really big so above that I would just build 2 banks with 400AH with its own BMS and run them in parallel. But that my personal preference, nothing wrong with 700 or 1000AH cells.

[CaptainRivet]

Originally Posted by s/v Jedi

I still did not need to balance my 8x Winston 400Ah and I’m 2 years in.

While ai do like the Cerbo GX, I don’t like the high level of Victron integration of their batteries, BMS and Cerbo GX.

Do you regularly discharge with 0.7-0.9C…catnewbee does when his wife gets the cooking Flash he pulls 7kw in 12V from the 1000Ah cells. The 3,55V combined with the REC BMS passive balancer gave it not enough time and balance capacity too small to handle the small deviation caused on this huge cells. Charging to 3,65V and using the active balancers cured that, nothing wrong here

[s/v Jedi]

Originally Posted by CaptainRivet

Do you regularly discharge with 0.7-0.9C…catnewbee does when his wife gets the cooking Flash he pulls 7kw in 12V from the 1000Ah cells. The 3,55V combined with the REC BMS passive balancer gave it not enough time and balance capacity too small to handle the small deviation caused on this huge cells. Charging to 3,65V and using the active balancers cured that, nothing wrong here

His problem is that he is above the 0.5C load. This is why I hail 24V or even 48V systems instead of keeping 12V… it lowers the current dramatically.

I still go over 0.5C but not for long periods. We do baking etc. when solar production is providing the power. I have an active balancer at hand and never connected it yet simply because I don’t need it. With 8 cells my deviation was 8mV last time I checked. My cell monitoring isn’t precise enough to measure that so I actually take out the multimeter to check this now and then

[Loredo]

"My cell monitoring isn’t precise enough to measure that so I actually take out the multimeter to check this now and then"


Just as an FYI.

The easy and cheap solution for monitoring single cell voltage(s) is this little guy : https://www.isdt.co/bg-8s.html?lang=en
It does several (un-needed for me) things, one of them is up to 8 cell voltages down to the third decimal point. As is, it's already very accurate but it can also be calibrated. What I really like about it is that it does very little buffering in the readings, it updates very quickly the output. It takes a bit of McGuyver style to install it permanently but with the help of a 3D printer nothing outta the world.

[s/v Jedi]

Originally Posted by Loredo

"My cell monitoring isn’t precise enough to measure that so I actually take out the multimeter to check this now and then"


Just as an FYI.

The easy and cheap solution for monitoring single cell voltage(s) is this little guy : https://www.isdt.co/bg-8s.html?lang=en
It does several (un-needed for me) things, one of them is up to 8 cell voltages down to the third decimal point. As is, it's already very accurate but it can also be calibrated. What I really like about it is that it does very little buffering in the readings, it updates very quickly the output. It takes a bit of McGuyver style to install it permanently but with the help of a 3D printer nothing outta the world.

Yes, I actually have that one, as well as a bunch other ones I could find
That said, even though it shows values to the mV, it’s accuracy is only 5mV. So it’s readout looks worse than when I do the readings with my multimeter, which is accurate to something ridiculous like 0.01mV