I could really use some help figuring this one out, please!
I’ve done some testing and took plenty of screenshots along the way, so you can follow the entire scenario while reading (check the timeline in the screenshots for reference).
System setup:
Victron MultiPlus II 48V / 5000VA / 70A
2 × Pylontech US5000 (4.8 kWh each)
Lynx Distributor
Cerbo GX MK2 (No solar panels connected yet — charging only from the grid.)
You can see all my settings in screenshots #ESS+DVCC
Now, when the system is showing 100% SOC:
ESS: Optimized with BatteryLife
DVCC: On
BMS control: Pylontech
CCL: 0.0 A (so it’s fully charged)
Battery voltage: around 52.3 V
Everything seems fine so far
However, there’s a noticeable gap between the lowest and highest cell voltage — somewhere between 0.020 V and 0.030 V, and it just stays there. (screenshot #2)
Then, as soon as I disconnect the grid, things start changing:
For the first few minutes (SOC99%), voltage drops to around 51.7 V, but the cell-voltage gap gets smaller, down to around 0.010 V (screenshots #3).
After a few more minutes (still at SOC 99%), the voltage drops sharply to about 49.7 V, and stays there. The gap between the lowest and highest cell voltage now drops to 0.003 V. At 90% SOC it even goes as low as 0.001 V — which seems perfect. (#4)
So… why does this happen?
Why are the cells so unbalanced when fully charged but perfectly matched once the voltage drops?
Am I missing something in my setup?
Just to mention — I bought the system back in May, but only got around to installing it a few weeks ago.
It doesn’t normally sit at 100% SOC; I only charged it fully this time to show how it behaves in that state.
Any insights would be greatly appreciated!
PS: at 100%, the system keep changing its status between Idle, Charging and Discharging very often.
Because when a LiFePO4 is above 3.45V is considered 100% charged and therefore, on Pylontechs, the balance is done only when the voltage on any cell is above 3.36V and the difference between min and max cell voltages is bigger than 30mV and stops when it gets below 30mV.
It’s normal.
This is how the Pylontechs are set from factory and all your observations (1,2,3) are consistent with proper functioning of these batteries.
We are “polluted” with behavior and settings from other BMSes (used in DIY rigs) where you can set the “imbalance” as low as a few mV, which is unnecessary low. As long as the cell voltage is above 3.4V, the cell is considered 100% charged (or above 99%) and a difference of 20-30mV between cells is negligible as you are on the end very steep part of the charging curve.
The same is on one of my system that is functioning this way for more than 2 years, with charge and discharge daily and all is OK.
See below yesterday behavior…
Before posting, I actually read quite a few topics where you had contributed — I was looking for similar cases and learned a lot from your input.
Based on your name, I assumed you might also be from Romania. I was hoping to get in touch with you directly, but unfortunately, there’s no DM option on the Victron forum.
Is there any chance I could contact you somehow? I’ve been looking for an expert to discuss a few more questions with, and I’d really value your insight.
To be more precise a I have a few more questions regarding the ESS configuration needed for my application and so on.
Of course, I’m more than happy to compensate you for your time if that’s possible and needed.
Also, regarding the previous statement: “At 100%, the system keeps switching between Idle, Charging, and Discharging very frequently.”
Can you please help me with an explanation or resolution for this behaviour? Because at other states of charge, the batteries go into Idle mode normally, but when they reach 100%, the system keeps changing its status constantly.
Where do you see that switching (idle, charging, discharging)?
On VRM, inside the battery box?
If yes, that it’s programmed to show idle for current = 0, discharging when current is negative and charging when the current is positive.
And at 100%, even if the CCL = 0, there is a small, continuous, variation of the battery current, once it hits the programmed CVL (not battery CVL), because:
the system has that PowerAssist/PowerControl active all the time
the inverter is drawing some power from the batteries
there is some “fighting” between the desired voltage and the computed voltage on one of the DVCC scripts that causes the voltage to oscillate. Some time ago I’ve pointed for this small discrepancy, but…
