• State
  • Locked Locked
  • Replies 12 replies
  • Subscribers 64 subscribers
  • Views 442 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

BQ4050: SUV Permanent Fail Triggered Unexpectedly

TI_Lover
Prodigy 160 points
Part Number: BQ4050

Tool/software:

Hi TI team,

We use the BQ4050 in a production product.

We have found several units failing during outgoing quality control. The product is unresponsive, and we have found that the batteries are bricked due to SUV permanent fail condition.

The product passes all production tests, including a cycle of charge to full, discharge to empty, then charge to 30% for shipping. After a final system-level function check, the BQ4050 is sent to shutdown mode to prevent discharging during transit. To conduct the outgoing quality check, we wake them up by applying a charge voltage.

On the failing units, we used the BQ tool and see no obvious issues: the cell voltages are fine. The lifetime min/max voltages are in the normal range. If we clear the PF, the battery can continue to be used normally.

The CUV threshold is 2100mV with 2s delay. The SUV threshold is 1500mV with 5s delay. However, the lifetimes section reports 0 CUV events. (I guess the PF could be blocking any writes to the data flash?).

I have attached the .gg file. Can you review to see if there are any abnormalities I missed? 

I can't think of any explanation, but it seems to always happen when (attempting) to wake up from shutdown.

Thank you very much~

3527.gg for ti.csv

  • 0
    TI__Mastermind 37761 points

    Hello,

    Can you send me the srec file from the gauge. Also, can tell me how many battery cells in series you are using in your application.

    Regards,

    Adrian

  • 0 in reply to Adrian
    Prodigy 160 points

    Hi Adrian,

    Thanks for the reply - let me ask the team to download this file and I'll share it here shortly.

  • 0
    Prodigy 10 points

    Hi Adrian,

    (Posting on behalf of as we are colleagues)

    Attached is the .srec file from that particular gauge. PS: I was unable to attach the .srec file as-is so I have zipped it.

    Also, to answer your question, we have 4 series cells in our application (we use a 4S3P configuration).

    Please let us know if you have any difficulties opening the file and if you spot anything notable.

    Thanks once again.

    243204103.zip

  • 0 in reply to AN_UG
    TI__Mastermind 37761 points

    Hello,

    Thank you for sending the files over. Please allow me some time to look over them.

    Regards,

    Adrian

  • 0 in reply to Adrian
    TI__Mastermind 37761 points

    Hello,

    Can you explain more about the process of how quality control is tested. Like what exactly happens during quality control? I want to see a detailed process to make sure there is nothing in the process that will be causing this issue.

    Regards,

    Adrian

  • 0 in reply to Adrian
    Prodigy 160 points

    Hi Adrian,

    Of course. After all of the board-level and system-level test during manufacturing have been passed, we send the BQ4050 into shutdown using the command described in the TRM 5.4.2. This is our shipment mode.

    To do the quality check, the system is woken up from shipment mode by connecting a charger to our product. The charger will first power just the main PCBA and the MCU will check the status of the battery (voltage, SOC). If the battery should be charged, only then is a charging voltage applied to the pack.

    If the MCU sees the battery voltage is 0V, it assumes the battery is deeply discharged and a pre-charge routine starts. A starting voltage of 6V is applied to the pack and it's slowly ramped up until we detect a charging current. This low charge current is maintained until the battery reaches 12V. (Note: pack is 4 series LFP cells).

    However, in the case of wake-up from shipment mode, this process happens very quickly. When ramping up from 6V, once a sufficient voltage is applied to the pack pin, the pack wakes up and the voltage is immediately greater than 12V, so the pre-charge routine will terminate.

    During quality control, these failing units never exit the pre-charge routine because the gauge is stuck in PF. Therefore nothing actually happens during quality control beside attaching the charger, at which point the operator notices the system is bricked.

    I should note that we have produced >30k of this product, and have seen ~10 of this PF issue. So the issue is quite rare, but also not a one-time thing.

    Thank you very much for your time

  • 0 in reply to TI_Lover
    TI__Mastermind 37761 points

    Hello TI_Lover,

    Thank you for providing this information. This is a very strange issue and I will need to discuss with my team further regarding this.

    Regards,

    Adrian

  • 0 in reply to Adrian
    TI__Mastermind 37761 points

    Hello TI_Lover,

    I want to confirm that for the devices that are experiencing this issue, are the individual cell voltages of each cell measured to be above the SUV threshold?

    Regards,

    Adrian

  • 0 in reply to Adrian
    Prodigy 160 points

    Hi Adrian,

    On all the batteries we checked, every cell was 3.2V - 3.3V.

    The lifetime min/max cell voltage are also in the expected range.

    One interesting clue is the data in the PF Status > Device Voltage Data > Cell Voltage registers. In the attached .gg file, you can see the cell voltage are listed as 0.53V, 0.73V, 2.14V, and 6.09V.

    Thank you,

     2nd gg for ti.csvsrec file 6553 .zip

  • 0 in reply to TI_Lover
    TI__Mastermind 37761 points

    Hello TI_Lover,

    I also want to confirm that has there been any instance in production where the PF condition was met and the battery was able to exit pre-charge routine. Essentially, has any pack in production tripped the PF and was able to recover.

    Regards,

    Adrian

  • 0 in reply to Adrian
    Prodigy 160 points

    Hi Adrian,

    If the BQ4050 enters permanent fail, it will never recover unless we disassemble the product and connect the TI BMS reader tool. Our FW has no feature to recover a battery that has entered permanent fail.

    It's unclear at the moment if the BQ4050 is already in PF before we start the quality test and connect the charger, or if the PF is occurring in the exact moment we are attempting to wake up the pack. My guess is the latter, since we are able to successfully send the gauge to shutdown mode and the gauge doesn't measure voltage during shutdown.

  • 0 in reply to TI_Lover
    TI__Mastermind 37761 points

    Hello TI_Lover,

    Yes, the gauge will not trip the PF until it is in normal mode and not in shutdown mode. Thus the cell voltage must be below the PF threshold when the battery was in shutdown mode. Therefore there must be something that is drawing current that is depleting the battery. The supply current of the gauge in shutdown mode is very small so unless the battery is in shutdown mode for years I highly doubt just the supply current from the gauge is causing this. Leakage current from other devices in the application first come to mind. Possibly PCB was manufactured poorly and traces were not placed good and/or shorts in the traces that could draw larger than normal current. These are a few possibilities I am thinking of that can cause this PF to happen.

    Regards,

    Adrian