BQ40Z80: Cell Balancing at Rest

Hi Bhavil,

Would it be possible to share the .gg file of the gauge settings so we can look into them?

I believe monitoring the [REST] bit will allow you to check if the dV/dt condition is being met, since the gauge will not enter rest without that condition being cleared.

Regards,

Anthony Baldino

6s1p.gg.csv

Hi Anthony, please find the gg file attached. What I have observed is that the REST bit is set, the cell balancing only starts when I connect the battery with the gauge but after some time, the BAL_EN bit clears and the balancing stops. Please see the picture attached which shows the maximum delta being around 30mV and the gauge being at REST. Despite these conditions, the BAL_EN bit is clear.

Hi Bhavil,

Would it be possible to receive a log file of this situation above? I believe that one of the checks below could potentially be stopping the cell balancing operation:

Regards,

Anthony Baldino

Hi Anthony,

The gauge was last charged/discharged a couple of days ago. I will take a log of the gauge for the whole day and attach the log here.

Hi Anthony,


Please find the log files below. I had to disconnect the communication board for some time so the logs are divided.

RELAX mode log.logRELAX mode log 2.log

Hi Bhavil,

After looking into the first RELAX mode log, it seems like the cell does not enter relax since it is moving between 24802mV and 24798mV at a decently quick rate, where the REST flag does not become set. It seems like this behavior is also the same in the second relax mode log, where the voltage is still moving too much for the gauge to enter relax.

Regards,

Anthony Baldino

Thanks for the feedback Anthony,

I'll check if its the problem from the battery or the gauge.

Anthony said:

where the REST flag does not become set

I checked the log and the REST flag was set throughout the log and I did notice that cell balancing did start for a couple of times in the middle but then it stopped after shortly after, which means the gauge did meet the dV/dt condition for a brief period. How does the gauge check the dV/dt reading and how does it affect the cell balancing time calculation?

Does it check for dV/dt and takes action depending upon its value for every loop for does the gauge need to satisfy the dV/dt condition for a certain amount of time before it calculates the cell balancing time?

Hi Bhavil,

My apologies, it seems like there was an error with my calculation. You are correct, both REST and BAL_EN are set throughout the log. 

The dV/dt condition is measured over a 1 second interval to ensure that the cell is truly at rest. I believe below are the two conditions where the dV/dt condition can affect the process seen below:

For the OCV measurement to be taken, the dV/dt condition will need to be passed, which is why that point could be causing an issue as well.

If possible, can you please send me the .gg file of the gauge settings so I can look into the other conditions and see where it can be tripped up?

Regards,

Anthony Baldino

Hi Anthony,

Please find the GG file here
8623.6s1p.gg.csv

Hi Bhavil,

Thank you for sending, we will look through this.

While looking through the Relax mode log 2, the BAL_EN follows this pattern, where each spike is exactly 300s long, matching the Relax Balance interval which is the timer each check is made. If possible, can extending this interval be tested to see if the results improve? It seems like in some situations the conditions are being met in others it is not.

Regards,

Anthony Baldino

Thanks Anthony,

I'll have a look into this.

I set the timer to 300s instead of the default value of 18000s (6 hours) as I was unclear what happens in a situation when the cells are still unbalanced, but the individual cell balancing timers reach 0, before the 6 hours are over? In that case, will the gauge wait till the 6 hours are complete before recalculating the individual cell balancing timers.

Obviously, this case will arise if the Cell balancing timer calculations are not done properly, but I just wanted to clear the doubt that I had.

Hi Bhavil,

Bhavil Nagar said:

I set the timer to 300s instead of the default value of 18000s (6 hours) as I was unclear what happens in a situation when the cells are still unbalanced, but the individual cell balancing timers reach 0, before the 6 hours are over? In that case, will the gauge wait till the 6 hours are complete before recalculating the individual cell balancing timers.

I believe that increasing this value could potentially assist the issue being seen. You are correct, if the value is set to 6 hours, then it will not check to recalculate the cell balance timers until then even if the cells are still unbalanced since this is the first condition to recalculate:

If the worry is that the cell balancing will not be recalculated enough with 6 hours, then I believe 1 hour should be tested first for this value and we can go from there.

Regards,

Anthony Baldino

Hi, Anthony,

I increased the Timer to 3600 seconds, but the the pattern followed is still the same, but the duration increased to an hour. Please find the log and the GG file attached.6s1p_LS14_TI.gg.csv

6S_B3.log

Hi Anthony,

The cells are balancing much better now with a higher Cell Balancing Interval. I will try to increase the balance interval to the default value of 18000s and keep a watch on the behavior and keep you updated. I am also marking my issue resolved.

Thank you for your time and help
Cheers!