Continue to Site

# In simulink Battery Block Voltage

#### iamkim

##### Newbie
hi i made a buck-boost method active cell balancing. And it did balancing well. but i think it has a problem. My model show me like this i settled Cell1 3.7 norminal Voltage and 3.4 Rated capacity
and 50% SoC , Cell2 3.7 norminal Voltage and 3.4 Rated capacity and 55% SoC but the battery Voltage graph shows Cell1's voltage larger than Cell2's voltage. So I wanted to write a balancing code using the volt difference, but it didn't work.
how can i fix it??? help me.... Why does the graph show this phenomenon?

To attempt a lossless active BMS, you need a ratiometric error voltage for 1/ N cells to enable small avg pulse current corrections based on the string current. If a cell corrects more than the string avg current there is a unbalanced cell failure or a design failure and must shutoff. This must also be bidirectional for chg/discharge. You may need to sequence these corrections to avoid crosstalk. Each Cell needs to be modeled for C and ESR which varies with SoC and life cycles and mismatched Ah. A good BMS prevents the exponential weak link degradation to extend battery life but should not make it worse.

State all your assumptions for each variable, new to EOL vs SoC in terms of ESR & C. Although there are multiple RC constants as you know the memory effect of a pulse charge restores near previous V after a longer time constant. Keep in mind EOL total lifespan Ah transfers can be multiplied by 10x from 200 cycles to 2000 cycles for example by using Battery University data and conservative usage by reducing Vcc max to 4.1 or 4.0V and 50% SOC min before swapping out packs.

Last edited:

To attempt a lossless active BMS, you need a ratiometric error voltage for 1/ N cells to enable small avg pulse current corrections based on the string current. If a cell corrects more than the string avg current there is a unbalanced cell failure or a design failure and must shutoff. This must also be bidirectional for chg/discharge. You may need to sequence these corrections to avoid crosstalk. Each Cell needs to be modeled for C and ESR which varies with SoC and life cycles and mismatched Ah. A good BMS prevents the exponential weak link degradation to extend battery life but should not make it worse.

State all your assumptions for each variable, new to EOL vs SoC in terms of ESR & C. Although there are multiple RC constants as you know the memory effect of a pulse charge restores near previous V after a longer time constant. Keep in mind EOL total lifespan Ah transfers can be multiplied by 10x from 200 cycles to 2000 cycles for example by using Battery University data and conservative usage by reducing Vcc max to 4.1 or 4.0V and 50% SOC min before swapping out packs.
i setting my model like this capacitance is 100e-6F inductance is 100e-6H and i post my battery model parameter script. and i give PWM in this model 350khz. so i don't understand your mean just little what information i give you? i also post my code