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.

BQ34Z100-G1: Learning cycle issues with LiFePO4 cells (Update Status stuck at 0x05)

Part Number: BQ34Z100-G1

Hello all,

I have issues to complete the learning cycle for a 2p LiFePo4 battery pack (A123 APR18650M1A, 1100mAh). The gauge was programmed with ChemID 404 and data flash parameters configured to match LiFePo4 voltage range.

When performing the learning cycle, update status successfully updates from 0x02 to 0x04 and 0x05. Qmax also updates as expected. FC was set after charging. But after the final discharge cycle and relaxing for more than 9 hours, update status still did not change from 0x05 to 0x06. RUP_DIS and VOK both were clear at this point, only OCVTAKEN was set.

When looking at the diagram below, I wonder if the relax times after charging / discharging even need to be increased further (currently around 8 hours) since there is still some change in cell voltage noticable.

There are some remarks I need to add. Can you please comment, wether one of these items might be the reason for the failed learning cycle and what needs to be changed?

  • I did not change the "Flash Update OK Cell Voltage", still at 2800 mV
  • Not charging with strict CC/CV profile; charging circuit goes into thermal foldback, so charging current starts at 2000 mA but is significantly reduced, as the charger chip heats up (clearly visible in diagram)
  • Although the battery is rated for cut-off and charge voltages of 2V and 3.6V, it is only discharged to 2500mV and charged to 3530mV. For the learning cycle, is it required that the full voltage range as given by the cell datasheet is crossed?

Do you have any hints how to continue, before starting the next learning cycle?

See Learning Cycle log and dataflash export files for reference below.

Thank you and best regards,

Daniel

Learning cycle log:

2019-08-16 Learning Cycle 1.zip

Data Flash Auto Export Files:

Data Flash autoexport.zip

  • Hi,

    Please allow us 2 days to analyze your logs and get back to you.

  • Hi Daniel,

    Looking at your log, when you entered chg relaxation your control status never got set to 0x0011. That means while a DODatEOC was taken a DOD0 was not taken when passed chg cleared. So, when you had Ra updates, you are not getting an update status change. Please run another dsg cycle, where you chg and stop chg at 60mA taper current, then relax until VOK clears and then start a dsg. At dsg relaxation when VOK clears again it will update to 0x06.

  • Hello Batt, thanks for analyzing my logs, really appreciate it!

    If I understood you correctly, I need to modify the setup to cut off charging at a higher current. Will do that and try again.

    In the meantime I did some more testing:

    • running a ChemID cycle: this gave me ID 482 as best match (opposed to 404 as selected by the cell vendor datasheet)
    • programmed the new ChemID 482, lowered FLASH update OK voltage to 2500 mV and restarted a learning cycle (see logs for reference below)

    Also in this case, Update Status did not change to 0x06 after DSG and relax. But when I charged the pack to full later on, at some time it changed to 0x06. I am unsure now, if this is expected behavior and the learning cycle now really is completed or if I should restart with your suggested changes.

    Thank you & best regards,

    Daniel

    Files:

    2019-08-20 Learning PackB Test1 ChemID 482.zip

    2019-08-20 Learning PackB Test1 ChemID 482 flash auto export.zip

  • Hi Daniel,

    It looks OK to me. However, just to be safe, if this is going to be your production pack you may want to reset it and run another learning cycle with the unlearned gauge.