• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 63 subscribers
  • Views 787 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.

BQ20z65 dataflash failure

Michael Houzer
Prodigy 100 points
Other Parts Discussed in Thread: BQ20Z65, BQEVSW

Dear Members,

I have a problem with a large number of battery packs returning from customers. The packs are based on BQ20Z65 and we are producing a huge number of packs with this chip since a  long time.

Within the last time we get a lot of packs back from our customers that report a permanent fail with DFF (dataflash fault). We are able to unlock the device and reset the fault, but when we try to rewrite the dataflash with the TI bqEasy-Software we immediately get the DFF-error again. Further investigations show that at least one page in the flash is defective and can not be written. Only solutiion is to replace the bq20z65 with a new one.
 All boards are manufactured within the same time. The error is independent of the cycle count of the pack. Age of the packs is about 6 months.

Can anyone at TI help or give some information what the cause of this error could be?


Michael Houzer

  • 0
    Prodigy 100 points
    Dear TI-Team,

    the number of battery packs returning from customers is increasing. Meanwhile more than 30 out of 700 batteries returned whith DFF-Fault. In the last years we built some thousend of batteries with BQ20Z65 and never had this fault. Can TI give some information what can cause this fault and what can damage the chip in this way?
    Regards
    Michael Houzer
  • 0 in reply to Michael Houzer
    TI__Expert 6575 points
    Hi Michael,
    The DFF gets set if the gauge notices that the values being written into the Flash are not matching the read back from the gauge.
    This can happen usually if the gauge has had too many flash writes and the Flash has worn out, or
    Some value in the Flash was changed during a Flash update from the gauge and the value read back from Flash is different from what was written.
    the third possibility is that the checksum for the firmware, was written to or somehow corrupted.
    Can you please send me senc files from the failed packs and also the golden senc so that I can compare the changes?
    This may give us some clues.
    Regards,
    Swami
  • 0 in reply to Swaminathan Ramanathan75031
    Prodigy 100 points

    Hi Swami,


    thank you for your answer.

    For me it looks like the flash is worn out. I found out that when I try to update a value in a page of the dataflash with bqEvSw the value gets not written an the DFF is asserted. If I install another DFI-File from a good pack, the page with the error changes to another page. So for one image the error could be at eg. RA-Table and with an other image the error could be at eg. calibration page.

    When I reset the DFF-error the chip is working until it writes the DFF again. So the chip self is fully functional.

    But how can the flash get worn out in about 6 months?

    Meanwhile I found out that we produced different boards with the chips comming from the same reel. But only one kind of these boards have the DFF-error. The main difference is that the boards with the DFF have a microcontroller that is connected to the SMBus. This controller is only reading the actual current value from the BQ20Z65 to control some additional peripherals. When the actual current is below a certain value, the controller is reading the value every 60 seconds. Between the read cycles the SMBus is idle (both lines low) to conserve power when the battery is not used.

    This feature is implemented in out batteries since years, but the time between the read cycles in former batteries was 20 seconds. To conserve more power the cycle time was changed to 60 seconds. Now only batteries with the longer cycle time are affected.

    Could this be a reason? Do you need more information beside the DFI?

    I will see that I can get the DFI-Images of bad chips an send it to you.

    Regards

    Michael