@burny was kam denn eigentlich bei dir raus bzw wie ging es weiter?

@tribesman I’m using a physical trick here to protect the cells, even without an active balancer:
The flat LiFePO4 voltage curve: A LiFePO4 cell is already about 95% to 97% full at a voltage of 3.43 volts.
The safety margin at 110 volts: If I set the lab power supply exactly to 110.0 volts, these 110 V are distributed on average to 3.43 volts per cell (110 V / 32 cells). As I understand it, this would provide self-protection against overcharging. Even if the cells are extremely unevenly charged (“unbalanced”) and three or four cells are charging faster than the rest, the voltage range with a total cut-off of 110 V is so low that no single cell can ever reach the dangerous threshold of 3.65 V. The power supply stops the flow of energy well before the critical zone. Or am I missing something here?

You are assuming that all cells inside one pack are in "good balance"... but knowing Deye balancer, I fear they are not.

There could be more than 5% SOC difference inside the same pack.

Hallo zusammen,

ich benötige eine Beurteilung von Leuten die sich gut mit HV-Batterien (DEYE GB-L) auskennen. Nachdem ich mein angeblich "defektes" bzw. stark "debalanciertes" Modul entfernt habe und eine Zwangsladung durchgeführt habe, bekomme ich folgendes Bild.  Balanciert der Turm oder wie kann ich das interpretieren?

BMS Software-Version: HVCUE3225919C1T

hab ich bei mir noch nie so gesehen.

aus welcher Anwendung ist diese Anzeige?

@benni82 Habe ich mir in HA selbst gebaut, damit ich nicht immer in die DEYE Cloud muss.

I doubt it is balancing... Deye does not do typical "top balance".
They are "trying to be smarter", and are trying to do the balancing through whole voltage range of the cell.

Don't ask me why... it is stupid, but I fear they have no idea how the balancing on LFP should be done.

If you would see the balancing going on, then on the long run, the voltages between max and min cell voltage would be coming together.

But if you look at your graph, it is quite clear the "dV trend" is basically always the same, cell voltages are not drifting together as they should be.

The only thing that is going on is the battery enables charging when max cell voltage is below 3.4V, and disables it when max cell voltage is above 3.65V. And it just goes on and on.

To achieve some kind of balancing first you have to have the balancing firmware installed on BMS, that is either version 4005 of 2AB6. And for the BMU version you should have 1317 or 1303. You can check the versions in the Deye Cloud.

Other version from my experience do not do any usable balancing.

Then you need to cycle the battery... if you just leave it sitting at 100%... this will not balance it.

You should let the battery cycle between the SOC that lowest cell voltage goes to 3V and highest cell voltage goes to 3.65V regularly.

@tribesman Deye's response: "HVCUE3223712CH1.bin, is used to recharge individual battery packs, meaning it is a special firmware for charging a single battery pack, not for balancing. Was the file you mentioned, HVCUE3226302N01, installed for you by my colleague? I have now checked, and the battery BMS version is already 2AB6. This is the firmware used for balancing the cells. When you have time, please use this firmware to perform a full charge-discharge cycle, that is, charge to 100% and discharge to 2%, to check the actual capacity of the battery. If there is still a discrepancy, I can help you update to the latest balancing firmware, 2006.

This software is designed to balance the voltage differences within the battery cells as well as between batteries. Please click on the 'Clear Relay Sticking' option on the PDU screen and then try the charge-discharge cycle again, as the SOH is already at 98%."