I'm puzzled, perhaps because it happened to me for the first time in 2 years and I ave no clue what could cause it.

My power consumption was decreasing slowly over the night, so was my battery SOC. Then around 3 o'clock at night the voltage dropped to 0, as if all the power suddenly evaporated :-)

In the morning batteries started charging slowly and then after reaching like 3% around 11:00, suddenly jumped to 100%. What could cause such behaviour?

BTW, I'm using 10 x 5kWh SVolt batteries with 4 x 5 kW Deye inverters in parallel.

Probably cell imbalance / faulty cell(s) in one or more of the batteries - are you able to get detailed stats of your voltages for each battery? This data may indicate were the issue is. I am also assuming you are still under warranty? Probably a good idea to notify the installer / warranty holder of this to see if they can assist with repairs / replacements

2 hours ago, Eric007 said:

In the morning batteries started charging slowly and then after reaching like 3% around 11:00, suddenly jumped to 100%. What could cause such behaviour?

It sounds to me like BMS malfunction. BMS could have wrongly discounted Ah to reach low SOC limit and shut off the batteries. By recharging the batteries the BMS jumps to 100% SOC when the cell voltages reach bulk charge level. To find out record the batteries terminal voltage.

13 hours ago, Eric007 said:

I'm puzzled, perhaps because it happened to me for the first time in 2 years and I ave no clue what could cause it.

My power consumption was decreasing slowly over the night, so was my battery SOC. Then around 3 o'clock at night the voltage dropped to 0, as if all the power suddenly evaporated :-)

In the morning batteries started charging slowly and then after reaching like 3% around 11:00, suddenly jumped to 100%. What could cause such behaviour?

BTW, I'm using 10 x 5kWh SVolt batteries with 4 x 5 kW Deye inverters in parallel.

As the graphs from SA( power vs soc) is a good indication of Master BMS disconnect.The SOC falling off the cliff with low load on the battery bank is an indication off master bms disconnect or communication failure between battery bank and inverters. Interrogate the master bms using Pbms tools V2.5 and have a look at the bms logs for the failure period.

I'm experiencing a similar problem with a SVolt LiFePO4 battery (25.6 Volt / 2.71 kwh): the battery keeps on shutting down at random regardless of actual load or SOC. I accessed the BMS memory via the PBMS software and got a reading that's puzzling to me in that it shows instances of a "protect", but never an actual "fault":

I then checked the settings and found that the PBMS indicates a number of parameters as being "out of range" (marked by an asterix) - but when I then do a reset I get the exact same settings:

Be this as it may: from the first table I can see that the battery cells were always within the set OV/UV threshholds. Does anyone have an idea how one can solve this puzzle?

Chris

37 minutes ago, mzezman said:

Probably cell imbalance / faulty cell(s) in one or more of the batteries - are you able to get detailed stats of your voltages for each battery? This data may indicate were the issue is. I am also assuming you are still under warranty? Probably a good idea to notify the installer / warranty holder of this to see if they can assist with repairs / replacements

Thank you. It happened at night, will have to monitor the batteries and check when/if it happens again. But I would think that if a one cell fails in a single battery, the remaining batteries will still provide power? OK, even if two or more cells fail, each in different battery, the remaining batteries, with no faulty cells should still work? Or I'm missing something?

22 hours ago, TaliaB said:

.The SOC falling off the cliff with low load on the battery bank is an indication off master bms disconnect or communication failure between battery bank and inverters.

@Eric007 see above

41 minutes ago, TaliaB said:

@Eric007 see above

Yes, thank you. Sorry for dragging but which one is a "master BMS" (I have 10 batteries connected). Would it be the first one, the one directly connected to the (first) Master inverter?

If so, I can easily test it, starting from 2nd battery (so the BMS in 2nd battery will become master).

Edited October 3, 2025Oct 3 by Eric007

If the inverters are paralled they are on the same dc bus. The master battery would be the one with the communication cable running to the inverters CAN input with dip switch address binary 1. Your batteries should be daisy chained so the other 9 is the slaves. The master inverter relays all the information to the slave inverter. So you need to check all the communication wiring between the batteries and to the master inverter. Then you need to check the master battery Bms using Pbms tools and check error logs for "alarm type", "Protection type" and " fault type"

On 2025/10/02 at 4:01 PM, Chris M said:

I'm experiencing a similar problem with a SVolt LiFePO4 battery (25.6 Volt / 2.71 kwh): the battery keeps on shutting down at random regardless of actual load or SOC. I accessed the BMS memory via the PBMS software and got a reading that's puzzling to me in that it shows instances of a "protect", but never an actual "fault":

I then checked the settings and found that the PBMS indicates a number of parameters as being "out of range" (marked by an asterix) - but when I then do a reset I get the exact same settings:

Be this as it may: from the first table I can see that the battery cells were always within the set OV/UV threshholds. Does anyone have an idea how one can solve this puzzle?

Chris

A BMS protection event won't necessarily trigger a fault.

Looking at the pictures you have posted, the BMS goes into protection mode due to overcurrent events (OCP).

Based on the BMS settings the charge and discharge overcurrent protection is triggeredat 110A and then looking at the event log it seems like the load exceeded the 110A at the point the protection was triggered. The BMS will release the protection based on the OCP delay time setting.

You mentioned in your post that the battery shuts down irrespective of the load, the log seems to tell a different story (although I am not 100% on how to read the log, assume by example the 13570 current in the second line is 135A, not to sure if you can confirm).

I checked one of my logs (also PACE) and the log shows the current with the decimal indicator at the correct place which seems to be different to your picture)

Edited October 3, 2025Oct 3 by I84RiS

2 hours ago, I84RiS said:

I am not 100% on how to read the log, assume by example the 13570 current in the second line is 135A, not to sure if you can confirm).

All values in the log represents mv, ma and mah so the 13570ma = 13.57A

18 hours ago, TaliaB said:

All values in the log represents mv, ma and mah so the 13570ma = 13.57A

I thought that might be the case, but saw the column header was A, which is why i then checked my logs (which does show A, but with the decimal on the correct spot)

Thank you I84ris for clarifying the protection vs fault trigger. Got it!

And TaliaB is right: the values in the "Pack Current" column are all in milli amps - meaning that the load registered in the second line was indeed 13.57 Amps. Which is well below the 110 Amp limit for triggering an OC protect...

In fact, due to the small size of the inverter connected to the battery it would be pretty much impossible for me to even reach the 110 Amp: I am using a 2 kw RCT inverter which under normal conditions would draw a max of 2000 Watts / 24 Volt = 83 Amp.

Anyhow: the BMS does register an OCP. There's three possible explanations for this to happen that I can think of.

1) The BMS is simply faulty or set incorrectly.

2) Something between the battery and the inverter triggers the OCP. I do have a Victron solar charger for my second panel array which independently can feed a max of 24 Volt / 20 Amps into the battery - but it will of course only do so during daytime. The OCPs registered in lines 1-11 have all occurred at night though, so I decided to rule out the Victron as a potential cause for the OCPs which in any event would have to be an over charging event, not an over load event. But all the OCPs registered in the log have occurred under load condition, not under charge condition.

3) In theory this inverter model will tolerate a surge of up to 4000 Watt for a maximum of 5 seconds in its default setting. However, I set the relevant parameter (= overload bypass) so that the RCT will automatically switch over to utility the moment 2 kw load are exceeded (a scenario which I then tested: it works faultlessly). The transfer time of the RCT for such a switch over is given as 10-20 milliseconds - which again is well below the delay times of the battery BMS:

With options 2 and 3 being ruled out my conclusion is that in all likelihood the BMS is the culprit. But then maybe there's something wrong in my logic...

Edited October 6, 2025Oct 6 by Chris M

6 hours ago, Chris M said:

Thank you I84ris for clarifying the protection vs fault trigger. Got it!

And TaliaB is right: the values in the "Pack Current" column are all in milli amps - meaning that the load registered in the second line was indeed 13.57 Amps. Which is well below the 110 Amp limit for triggering an OC protect...

In fact, due to the small size of the inverter connected to the battery it would be pretty much impossible for me to even reach the 110 Amp: I am using a 2 kw RCT inverter which under normal conditions would draw a max of 2000 Watts / 24 Volt = 83 Amp.

Anyhow: the BMS does register an OCP. There's three possible explanations for this to happen that I can think of.

1) The BMS is simply faulty or set incorrectly.

2) Something between the battery and the inverter triggers the OCP. I do have a Victron solar charger for my second panel array which independently can feed a max of 24 Volt / 20 Amps into the battery - but it will of course only do so during daytime. The OCPs registered in lines 1-11 have all occurred at night though, so I decided to rule out the Victron as a potential cause for the OCPs which in any event would have to be an over charging event, not an over load event. But all the OCPs registered in the log have occurred under load condition, not under charge condition.

3) In theory this inverter model will tolerate a surge of up to 4000 Watt for a maximum of 5 seconds in its default setting. However, I set the relevant parameter (= overload bypass) so that the RCT will automatically switch over to utility the moment 2 kw load are exceeded (a scenario which I then tested: it works faultlessly). The transfer time of the RCT for such a switch over is given as 10-20 milliseconds - which again is well below the delay times of the battery BMS:

With options 2 and 3 being ruled out my conclusion is that in all likelihood the BMS is the culprit. But then maybe there's something wrong in my logic...

Tend to agree with your logic. Are you able to write to the BMS (change the settings) using PBMS tools.

Not all versions of the software can write to the BMS, also think the PC to BMS connections needs to be to the RS232 serial port (as opposed to the RS485 mod port)

Edit 1: a reset won't change the parameters, you need the correct version able to write to the BMS as described above

Edit 2: also interesting to see that the Full Capacity is reducing every time there is a protection event (put differently, the SOH is reducing each time an OCP event is triggered). You can reset the SOH back to 100% using the correct version of PBMS tools, it might void your warrenty if you still have one. Perhaps contact the supplier for a firmware update. It can be loaded using the same version of PBMS tools that can write to the BMS. I have this version which I can share with you if you are interested but as said it might void your warrenty.

Edited October 6, 2025Oct 6 by I84RiS

Hi,

thanks for your suggestions. My PC is connected via RS232, so a write should work - but I'll test with the SOH first.

As for a firmware update: the warranty has expired anyhow. About the version you can share with me - are you refering to the PBMS tool or to the firmware?