Battery Monitoring
Open Cell Voltage - DC
Voltage is function of SOC
The voltage across positive and negative terminal of battery while not discharging current to a load
The internal resistance of the battery causes the OCV to decrease when current is discharged from it.
Open Cell Voltage - AC
Voltage is function of SOC
In this model, R0 is internal resistance, R1 & C1 model the short term transient response, and R2 & C2 model the long term transient response.
Table Lookup
The table relates capacity to voltage, current, temperature, and age.
Not practical takes time to collect and test data
Does not require a cell model
Voltage Lookup
Accurate voltage measurements can be made while a cell is resting (ensure that cell has been resting for a considerable amount of time)
The relation between capacity and voltage is non-linear. A typical model is shown below
Factors that Affect Capacity Estimation
PCB component accuracy
Ideally components with low drift/offset
Trace length may affect resistance
Instrumentation accuracy
ADC resolution
Sampling rate - Is the sampling fast enough to capture waveforms and integrate
Voltage drift (RESEARCH THIS MORE)
Noise immunity
Cell model accuracy
Capacity is susceptible to degradation upon repeat cycles, how does the model accurately capture this.
Hard and time consuming to extract adequate parameters using collected data, for resistances and OCV
Temperature
Impact parameters, especially resistance, and including max capacity, OCV, capacitance
Aging
Protection
Protect against short circuits, over/under voltage, FET failure, comm failure
AFE lowers the voltage before inputting them into ADC, and can be programmed to turn off the MOSFET’s
Cell-Balancing (Example)
Each cell is controlled by FET on the voltage input pin, which is pulled to ground when a cell needs to be bealnced.
Balancing occurs by diverting some of the current to the 40 ohm resistor, which is controlled by the External FET as chargin occurs
Other Features of BMS
Communication with host using I2X, SMBus, etc
Displays with LEDs to indicate SOC or errors
Logging to record max/min measurements and data for failure analysis (similiart to blackbox
Current Integration Based Fuel-Gauging
Coulomb Counting
Involves counting the charges going into the battery by integrating the current
Requires knowledge of the initial amount of charges
Defining Q as the total capacity of the battery in Ah/mAh leaves us with the following equation sets of equations
Thus we have RM = FCC - Q
SOC is calculated as RM/FCC
FCC is updated at every full discharge
Learning when we reach 0% SOC is not ideal, since it’s too late.
Therefore certain voltage thresholds (EDV 1, EDV 2,…) can be set
Useful since these EDVs are fixed values, and can correct the SOC based on the OCV-SOC relation when coulomb counting becomes inaccurate.
Compensated EDV (CEDV)
The model used to calculate the EDV values based on current SOC, current, and temperature.
We want to predict the actual battery voltage curve
Start with OCV, then correct it using the internal resistance and current
Then correct it using the value of the low temperature, leaving us with the green curve which accurately resembles the actual battery voltage curve
CEDV Formula
Impedance Track Fuel-Gauging
Includes advantages of both voltage lookup and coulomb counting
Allows for SOC updates, where initial SOC can be measured when voltage is relaxed, in addition to self-discharge
Impedance can be measured during discharge
Total capacity can be measure without full discharge or charge