SOK SK48V100 not discharging below 13%

[Atomykatt]

Hi,

I have a couple of SOK rack mounted batteries and used the params described here to configure them. I am able to connect to each battery individually using a serial cable connected to the master battery ot to both at the same time (using FF in the drop-down).

My setup:

- 3 * 3.3kWp strings connected to a RS 450/200 MPPT controller
- The MPPT is connected to a lynx Distributor which also connects to the 3 * Multiplus II 48/5000 inverters ruining three phase - 400V
- the batteries are connected via a lynx shunt to the distributor
- both batteries are connected with short 70sqmm cables, the positive of one battery and the negative of the other goes to the shunt.
- everything is controlled via CerboGX VE.Can, VE.Bus and BMS are connected to this.
- ESS is enabled on the system, and set to Optimized (without battery life) and minimum SoC is set to 0%.
- All cables are 70sqmm except the ones connecting the inverters which are 50sqmm

The issue I have is that the BMS cuts off the discharge at 13%. I've checked and double checked everything against the params provided by Dexter and the discharge does not go under 13%. The shunt gives a reading of 180Ah consumed and 7% SoC, but I have no idea if that is setup properly for these batteries.

The remaining Ah on both BMS'es is around 13Ah, so in sync with what I have displayed on the Cerbo. Also the cells are at around 3.2v with a 4-8mV delta between min and max cell voltages. This seems to be in tune with the 13% reading on the BMS.

Any ideas on what else can be done to get the last 13Ah to discharge?

Regards,
A.

 

[HalfBaked]

bump

 

[#40Fan]

When the BMS disconnects, are you able to check each cell voltage value immediately or is there time between disconnect and checking?

The settings in the screen shot you linked to, are all of your boxes checked in the top left of each individual section?

Any alarms in either BMS?

 

[Atomykatt]

Unfortunately I am unable to check it at the time of the disconnect as it happens somewhere between 02:00 AM and 04:00 AM. What I did however, is let it charge to 15%, shut down the MPPT, and with an ~1.5kW load try and drain it. I've attached the results, and also the parameter settings.

 

[Will Prowse]

@HighTechLab

 

[Atomykatt]

The batteries just finished charging and (as always) the BMS is returning High Voltage alerts



I was hoping these would disappear over time, once everything settles, but having configured the new params in late March, I now think they never will.

 

[HighTechLab]

I replied to your email yesterday, check spam folder - we've been having some trouble with gmail messages getting caught in spam this week.

Lot going on here and will be quickest and most organized for us to handle over email.

 

[Atomykatt]

I think you are mistaking me for someone else, as I haven't sent you any emails.

 

[HighTechLab]

I think you are mistaking me for someone else, as I haven't sent you any emails.

Ok strange! Sorry for the confusion. Feel free to drop me a line, dexter@currentconnected.com and we can get to the bottom of this.

 

[JohnL]

Would be curious to hear any updates after it's worked out. Thanks!

 

[chadwick101]

Dexter,

I think that was me you replied to. Thanks for the reply. I did find it in my spam folder after finding this thread. I'll pass what I learn on here as we go.

Thanks,

Chad

 

[Atomykatt]

Hey Dexter & co.

Thanks for your implication so far. Unfortunately progress has been slow on my end as I've not been in the country one week and retuned ill. I'll surely be posting the conclusion once it's reached.

Cheers,
A

 

[brilthor]

I'll surely be posting the conclusion once it's reached.

How's the progress going?

 

[wheeto]

Oh, I've just seen this thread after posting a question of my own. Did anyone reach a conclusion? I have debugged the CAN protocol on mine and found that the BMS is demanding charge whilst showing 13% SoC remaining. @HighTechLab @Will Prowse I see you've both commented here but if I can help by providing any of the CAN dumps I've made, please let me know! It looks a lot to me like a bug in the BMS.

 

[HighTechLab]

Oh, I've just seen this thread after posting a question of my own. Did anyone reach a conclusion? I have debugged the CAN protocol on mine and found that the BMS is demanding charge whilst showing 13% SoC remaining. @HighTechLab @Will Prowse I see you've both commented here but if I can help by providing any of the CAN dumps I've made, please let me know! It looks a lot to me like a bug in the BMS.

My team is working on this internally. I am waiting for further progress before posting anything.

 

[wheeto]

In the absence of any reply from SOK I've solved the problem for myself by intercepting CAN message 0x35C from the battery and forwarding it to the inverter system after doing a bitwise AND (&) of 0xCF on the first byte. I hope this saves someone a little time debugging until there's a proper fix.