question

there could be an efficiency setting in the BMS firmware that needs tweaking. However, SOC is an integrated value (computed over time) NOT measured, so should not be used for protection or gen start functions. These should always be based on battery cell voltages. Normally, the SOC is reset at either Low or High cell voltage, sometimes there is a convergence algorithm that steers the SOC to what it should be for the current cell voltage - with allowance for charge / discharge condition and age of cells.

If you don't get to the reset condition then without the complex convergence algorithm, the SOC will drift away from the "correct" value.

Best to look at how your gen start is configures, and switch this to a voltage measurement.

As the communication is setup between the battery and Inverter the SOC is automatically used as this is how they were designed to be used as this can adequately control the battery according to the manufacturers desired parameters. Using voltage to control a Lithium battery that has communication is just not a solution. If this is the solution then the battery is not functioning as it is meant to as all the other Lithium batteries with comms I have seen do not have any such problems.

Is there a way I can tweak the efficiency settings to make this battery behave as it was meant to?

Since the Dyness battery communicates its own SOC to the victron gx device this is really a Dyness support issue. However I see this often with other battery manufacturers as well. We make sure to use the cell or battery overall voltage to determine when the battery is empty or full, not the SOC communicated via the BMS. (Just like Chezzz said).
Have you tried letting the battery sit at 100% for an extended period of time? When we see this behavior usually its because the battery is becoming imbalanced. If the battery is fully balanced the SOC algorithm should revert to 100% and be accurate for awhile.

I am experiencing the same issue with a single Dyness BX51100 connected to a MPII, a Cerbo and an MPPT 150/60. The battery SOC is dropping with zero load (MPII off) and the charge controller is not charging the battery as it thinks the voltage set point has been reached due to the bms signal..

However I have another Dyness system working normally.

I haven't fully got to the bottom of it yet.

I would be interested to know more about the specific model of battery and the other components in your system? Anthony

I have exactly the same problem with my Powerwall connected to a Venus GX and a Multiplus. My firmware on the battery does not yet support the single cell voltages as on the new Pylontech batteries. Anyone know if it's possible to update the battery firmware or just how to get the battery back to a balanced state without having to remove it and take it in to a dealer which is 3 hours drive away from me? Sorry I know this is supposed to go on the Dyness community, but since you've got it going...

@Stingray I resolved the issue with my Dyness BX51100 battery by updating the battery FW to V1.07 which the GX reports as V136.3 incidently.

What battery FW is your GX currently reporting? Did you mean to say Powerbox and not Powerwall?

I wish I could update mine. The GX device does not detect any firmware versions, but it will probably be the factory version. How did you update your batteries? I have a feeling I would have to take the whole thing apart and do each battery separately.

The GX device is getting the parameters from the battery correctly, so I don't think it's a comms error.

Yes, Powerbox :)

Hi, This is how my battery displays on the GX. Are you sure your battery to GX comms cable is made correctly in terms of pin out assignment.

Please note Dyness have advised that the Firmware displayed by the GX (v136.3) is not reported by the battery which begs the question where is it coming from?

The image below shows the interface used. It is available from Aliexpress. You will need to source a Usb Type A to Usb Type B cable (printer cable) and an RJ45 to battery CAN cable separately. The software, firmware and parameter file came from Dyness support. They are a hit and miss in terms of contactability and support.

Check the battery voltage and current and see if on occasion the Charge Voltage Limit (CVL) is attained with the charge current dropping to almost zero. I had a similar issue, and the voltage was not rising sufficiently for the BMS to reset the SoC to 100%.