BMS- Which methods to calculate the battery SOC is more accurate?

Jwilmesbrady

New member
Joined
Dec 12, 2020
Messages
7
SEPLOS smart BMS gets the battery real capacity at its initial full charge of the battery through time plus charging current.
And will calibrate through voltage in the using process in case if the battery never fully charged.

And now they updated the firmware, to get the real capacity at a full discharge cycle.
And calibrate through voltage in the using process if the battery never fully discharged.
 

Attachments

  • 190725768_191023752894795_6354499434730929882_n.png
    190725768_191023752894795_6354499434730929882_n.png
    704.2 KB · Views: 19
  • BatteryMonitor interface.png
    BatteryMonitor interface.png
    726.1 KB · Views: 19
Last edited:

daromer

Moderator
Staff member
Joined
Oct 8, 2016
Messages
5,647
Basing SOC on voltage level is tricky and in many cases not even possible. First of all the battery must have been resting for atleast 5 minutes and you need to know the exact Voltage to SOC curve including aging.

The best way and to be honest the only "real" way is to calculate the energy going in/out and have a meter that also can take into account losses due to how hard you push in terms of C. (C factor as in often translated to Ampere)

Most systems dont have the callibratio based on C and only do colloumb and for most users thats fine.
Then you calculate the amount going in and out and you always know the SOC. You then reset this to either 0 or 100% based on reaching your top voltage level or bottom level and that becoms your ref point.

Hope its understandable.
 

OffGridInTheCity

Well-known member
Joined
Dec 15, 2018
Messages
1,592
Agree with @daromer. In my case - I've given up worrying about SoC to any specific degree. My focus is off-grid power production. So I worry about max charge settings (hi voltage) and Inverter off (low voltage) and overall health (balance/capacity/long-life) of the battery bank.

I do track many statistics, so I have a pretty good idea of battery bank capacity and a general idea of SoC - but accurate SoC at any given moment is fuzzy and I've found that it's not important for general operations day after day :)

To answer the question - I infer SoC by the voltage along the charge/discharge curve of the specific battery + a compensation factor for extreme amp levels (compared to routine) that significantly pull down or push up the current voltage.
 
Last edited:

Jwilmesbrady

New member
Joined
Dec 12, 2020
Messages
7
Basing SOC on voltage level is tricky and in many cases not even possible. First of all the battery must have been resting for atleast 5 minutes and you need to know the exact Voltage to SOC curve including aging.

The best way and to be honest the only "real" way is to calculate the energy going in/out and have a meter that also can take into account losses due to how hard you push in terms of C. (C factor as in often translated to Ampere)

Most systems dont have the callibratio based on C and only do colloumb and for most users thats fine.
Then you calculate the amount going in and out and you always know the SOC. You then reset this to either 0 or 100% based on reaching your top voltage level or bottom level and that becoms your ref point.

Hope its understandable.
Yes, they calculate the real capacity that going in/out from the shunt.
But the real case is, the battery may not get to 0 or 100% for months, without the calibration through voltage, the SOC would be deviation over time.
The detailed calibration as follows:
Adjust the SOC to 95%, when the overvoltage occurs, and the SOC is lower than 90%.
Adjust the SOC to 70%, when the voltage reaches 3.4v*series.
 

daromer

Moderator
Staff member
Joined
Oct 8, 2016
Messages
5,647
Adjusting to lower voltage than top voltage is quite useless on LiFePo4. 70% and 85% can be exact same voltage as example

You have to live with it drifting and or you have to charge it fully. Thats the life of using this type of system im afraid and the only way to reduce the drift is to get a high quality meeter for it
LiFePo4 should only have the voltage used above 3.5V or below 2.9V or around. Thats where you potentially know the SOC ;)
 

Jwilmesbrady

New member
Joined
Dec 12, 2020
Messages
7
Test report of SOC-VOLTAGE at different temperatures.
 

Attachments

  • soc voltage at different temperature.png
    soc voltage at different temperature.png
    186.2 KB · Views: 8
Top