BQ40Z50-DEVICE-FW: R4: Voltage-Based Balancing when Relaxed not working

Mark Egloff

We are trying out the Voltgae-Based Balancing for our 4S1P LFP battery, since the normal balancing algorithm was unbalancing our cells, rather than balacing.

The voltage-based balancing during charging seems to work, but the balacing during rest doesn't seem to be triggered. [BCR = 1] is enabled and we also disabled [LFP_Relax = 0] in order not to have to wait for 2 days. Also, BCR was working when we had the old balancing algorithm in R2 active:

Could it be that BCR doesn't work at all with Voltage-Based balancing because the increased thresholds (caused by the window logic) would need to be reset constantly, but in the TRM it only says:

"The cell balance threshold is reset to the value in Voltage Cell Balance Threshold at the start of every charge and reset cycle." (I don't really understand, what "reset cycle" means though).

Thanks in advance for your help,

Mark

over 2 years ago

0 Wyatt Keller over 2 years ago

TI__Guru** 105590 points

Hello Mark,

Do you have a .gg file for your configuration and a log showing the behavior?

Many of the same conditions need to be true for voltage based balancing as dQ based balancing:

BAL_EN = 1

AND (CHARGE mode

OR ( True RSOC > Start Rsoc for Bal in Sleep)

OR (balancing active

AND True RSOC >= End Rsoc for Bal in Sleep

OR True RSOC > Min RSOC for Balancing))

Then check for cell balancing start:

BAL_EN = 1

AND GAUGE_EN =1

AND ((UPDATE STATUS = 0x06 OR FC set)

OR UPDATE STATUS = 0x0E)

Then the specific voltage based balancing checks are done.

Sincerely,

Wyatt Keller

0 Mark Egloff over 2 years ago in reply to Wyatt Keller

Intellectual 380 points

Hi Wyatt

Thanks for the quick response!

Attached is the gg and log file with a full-charge and a relaxation period over the weekend:

/cfs-file/__key/communityserver-discussions-components-files/196/R4_5F00_settings_5F00_20221114.gg-m.csv

/cfs-file/__key/communityserver-discussions-components-files/196/88735_2D00_37116567_5F00_R4_5F00_780A_5F00_Charging-m.log

I saw that during the relax-time, the GaugeStat is constantly switching between 0x4A and 0x5A, which means BAL_EN is being set and reset frequently, which is maybe the cause for the missing relax balancing. But what could be the cause for this frequent setting and resetting of BAL_EN? (we used the same Relax-Balancing settings in R2 with the normal balancing algorithm and there relaxed balancing was working properly):

Also, in your reply, you mention "Start/End Rsoc for Bal in Sleep", but I the balance settings for sleep are different from the relax balancing settings, no?

Thanks for your help!

Best,

Mark

0 Wyatt Keller over 2 years ago in reply to Mark Egloff

TI__Guru** 105590 points

Hello Mark,

I will need to look into this for more details, I will respond tomorrow with updates.

Sincerely,

Wyatt Keller

0 Wyatt Keller over 2 years ago in reply to Wyatt Keller

TI__Guru** 105590 points

Hello Mark,

Because normal impedance track operations require a relaxed voltage, we will not allow cell balancing until the dV/dt condition is met, and since this is LFP it may take a long time for this condition to be met. By the time the condition is met the voltages are probably within the voltage based cell balancing limits so it will not perform cell balancing.

Sincerely,

Wyatt Keller

Power management

Power management forum

BQ40Z50-DEVICE-FW: R4: Voltage-Based Balancing when Relaxed not working