BQ40Z50-R1: Cell Balancing not working as expected

Hi Ken,

Is it possible to share the .gg file of the current gauge settings and values so we can take a look?

Ken Cowan said:

One thing I did notice is that the Balancing Configuration register is set to 0x05 so that CBR is set, but bit 1 is a RSVD_ONE and we are writing it to 0.  Could that be messing things up?

Also, RSVD_ONE is 1 by default. Please change this back to 1.

Regards,

Anthony

Hi Anthony, thank you for the response.  I've attached the gg file.  

1) Can you confirm whether or not the behavior that I'm seeing (the cell with the highest QMax is being set to 0 Balance Time) is expected? 

2) If this is not expected, could this incorrect behavior be caused by writing a 0 to bit 1 in the Balancing Configuration register?

Ken3322882-005 Production.gg.csv

Hi Ken,

After discussing with our team, this is the expected behavior if the Qmax is much greater for cell 2 since the device will use the difference in capacity for the calculation.

We can look into the other values within the .gg file at this time.

Regards,

Anthony

Okay, I'm starting to understand this a little better but need more confirmation.  For the 8 packs I'm evaluating, I checked the Cell x DOD0 and in every case the cell with the highest DOD0 is the one set to 0 seconds of balance time so it does seem to be working correctly.  What wasn't intuitive to me was that in 5 packs, the cell with the highest DOD0 also had the highest QMax.  But on 3 packs it was the other way around.  I guess there is not a direct correlation between QMax and DOD.  My original thinking was that if you have 3 cells and Cell 1 has a Qmax of 2 Ah, cell 2 is 1.5 Ah and cell 3 is 1.2 Ah and starting from full charge if you pull 1Ah out of each cell then the one with the lowest QMax will definitely have the greatest depth of discharge.  But then considering charging, it seems that since the charge cycle will end based on the cell with the lowest QMax so the one with highest QMax will never get fully charged and start out at already the highest DOD.   So there seems to be some combination of these things happening.  My final questions are:

a) Does my explanation of how QMax and DOD are related and don't necessarily correlate make sense or am I missing something important?

b) Will programming bit 1 of the balancing configuration register to 0 cause the system to misbehave?  We will try to get this fixed as soon as possible but would like to know the risk to our system with the current settings.

Thanks,

Hello Ken,

We have received your response and working on follow up message.

Thank you,
Alan

Hi Ken,

Having the highest DOD does not directly reflect having the highest Qmax, its is based on the amount of charge seen between two DOD measurments which will be taken, with one taken in the relax before charge/discharge and one taken after. The Qmax will essentially be (Passed Charge/(DOD1-DOD0)), so if one of the cells had a smaller differential in DOD measurements, the Qmax will be different.

Programming bit one of this register to 0 will disable cell balancing.

Regards,

Anthony

Anthony, thanks for your response, the DoD/QMax makes sense now.  With regard to bit 1 of the configuration register, my reference manual for bq40Z50-R1 is different than what you posted.  In mine, bit 1 is RSVD_ONE and it says to program it to 1 and not use it.  Through an error made years ago, this bit was actually programmed to 0.  I need to understand what the consequences of this are.  I can't simply program it to 1 as there are many packs in the field programmed this way already.  Will this cause any malfunction of the cell balancing operation?

Thanks,

Ken

...

Anthony, thanks for your response, the DoD/QMax makes sense now.  With regard to bit 1 of the configuration register, my reference manual for bq40Z50-R1 is different than what you posted.  In mine, bit 1 is RSVD_ONE and it says to program it to 1 and not use it.  Through an error made years ago, this bit was actually programmed to 0.  I need to understand what the consequences of this are.  I can't simply program it to 1 as there are many packs in the field programmed this way already.  Will this cause any malfunction of the cell balancing operation?

Thanks,

Ken

Hello,

We have received your update and will follow back with a message within the week.

Thank you,
Alan