This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

BQ76952: Balance question

Part Number: BQ76952

Tool/software:

Hi expert,

My customer (CosMx, Auto BMS project) found there is a passive balance issue in their project as below, please help to confirm.

They found the passive balance period sometimes would be shorter and it don't match the configuration after running passive balance several minutes. Customer configured the passive balance as odd-even balance and the interval is 2s. This issue happens occasionally a few minutes after passive balance is turned on.

Best Regards,

Ryker

  • Hello Ryker,

    Do you mind sharing the schematic and .gg file for this? When this occurs, do you know if it’s happening for specific cells, or seemingly random?

    Best Regards,
    Alexis

  • Hi Alexis,

    Please refer to below attachments of schematic and configuration. Customer and I don't know what's the .gg file, please help to check if the configuration attachment is enough to analyze. 

    Besides, it's happening for seemingly random and all BQ76952 projects. Please help to analyze ASAP.

    https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/196/4024.Schematic.7zhttps://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/196/Configuration.7z

    Best Regards,

    Ryker

  • Hello Ryker,

    Thank you for sharing these files. Please give me some time to look over them.

    Best Regards,
    Alexis

  • Hi Alexis,

    Please help to confirm by today. Because this issue influents every BQ76952 projects and it is critical. Thanks.

    Best Regards,

    Ryker

  • Hello Ryker,

    Thank you for your patience. Your schematic and configuration look fine.

    The shorter cycle is most likely due to the COV/CUV checks that occur during balancing. Every one second, cell balancing is disabled for ~20ms to all COV and CUV checks to run on all cells.

    This is mentioned in Section 2.5 Voltage Measurement Accuracy During Balancing from the Cell Balancing Application Note. This is further explained in the last few paragraphs in Section 10.1 Cell Balancing Operation of the Technical Reference Manual

    Best Regards,
    Alexis

  • Hi Alexis,

    I think the issued isn't caused by COV and CUV. Because in below balance waveform, the abnormal balance time are 0.5s and 3.5s while the normal balance time is 2s (2s is configured by software). It means sometimes the balance time is 0.5s and sometimes is 3.5s (it happens once or twice in 1 to 3 minutes). And the balance command which sent by MCU and cell's balance statue which read by MCU (0x0083 command) all are 2s when the actual balance time is 0.5s or 3.5s. Please help to confirm this issue.

    Besides, could you help to confirm if there any problem of BQ76952's configuration with above Configuration file. Thanks.

    Best Regards,

    Ryker

  • Hi Alexis,

    Please help to analyze this issue ASAP. Thanks.

    Best Regards,

    Ryker

  • Hello Ryker,

    Just to clarify, the waveform showing just 1 AFE switching between an odd/even cell for ~2s each, but is this just happening for the top AFE or bottom AFE or both? I see from the configuration file that it was mentioned that there is a ‘Low-side AFE enable balance’ and ‘High-side AFE enable balance’. Also, how many cells are being balanced at once?

    Best Regards,
    Alexis

  • Hi Alexis,

    • This issue is happening for both AFE. There are 2 AFE and each AFE monitors 10 cells. So there are 5 cells are being balanced by one AFE at once.
    • Customer also found all BQ76952 projects have this same issue even if they disable all code and only keep balancing function.
    • Customer suspect the issue is caused by Subcommands of 0x0083 CB_ACTIVE_CELLS(). According to the BQ76952 manual, when writing the balance command 0x0083, it takes about 1 second to take effect. However, in reality, this 1-second interval fluctuates and gradually decreases. When it decreases to almost 0, it resets back to 1 second. This switching process causes the balance cycle to be inaccurate, resulting in an additional 1-second delay. Please refer to below waveforms for the details. And help to confirm the reason and how to avoid this issue.

    Best Regards,

    Ryker

  • Hello Ryker,

    Thank you for clarifying.

    Customer suspect the issue is caused by Subcommands of 0x0083 CB_ACTIVE_CELLS(). According to the BQ76952 manual, when writing the balance command 0x0083, it takes about 1 second to take effect. However, in reality, this 1-second interval fluctuates and gradually decreases. When it decreases to almost 0, it resets back to 1 second.

    This is possible as the command does not always take 1 second to take effect. Unfortunately, there isn't a known way to necessarily avoid this as we cannot control the timing of the subcommand. 

    Best Regards,
    Alexis