• State TI Thinks Resolved
  • Locked Locked
  • Replies 12 replies
  • Answers 6 answers
  • Subscribers 63 subscribers
  • Views 407 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.

BQ34Z100-G1: Unsuccessful learning cycle. Update status never changes to 0x05

Abel Mendes
Prodigy 10 points
Part Number: BQ34Z100-G1

Hi everyone,

I’ve been trying to perform the learning cycle over a LFP battery pack, but although everything seems to be according the procedure described in SLUA903, the update status never changes to 0x05.

The battery is a LFP 48V-32A, and I’m using a scale factor of 5. After submitting the data log file to GPC tool, I got the ChemID 400 to my battery (see the attached report).

Following some suggestions found in the techincal forum, I tried many parameter adjustments over several learning cycles, but never got to the point of changing the update status to 0x05. It seems to me that all the conditions are satisfied, but this status keeps 0x04 forever.

What am I doing wrong? Could you please check the attached log files to help me overcoming this issue?

 

Kind regards,

Abel

LFP_48V32Ah-report.zip

LFP_48V32A_learningCycle.zip

  • 0
    TI__Genius 12215 points

    Hi, Abel

        For the LiFe battery, to update the Qmax, 48hours is needed at least before it happens, after that, plus additional 5hours, if Qmax update has not happened, then some hidden parameters may need to change, please send your log file and srec file for parameter modification

  • 0 in reply to Steven Yao
    Prodigy 10 points

    Hi Steve,

    Thanks for the answer. Below you can find the steps I'm going through during the learning process. Could you please describe in which steps shall I consider the times that you just mentioned?


    Step 1: IT enable + reset (Update Status changes to 0x04)

    Step 2: Discharge to empty

    Step 3: Relax state. Wait until [VOK] and [RDIS] bits in the IT status() register get clear

    Step 4: Charge battery to full

    Step 5: Wait until FC bit is set, and turn-off charger after ~60 secs

    Step 6: Relax state. Wait until [VOK] and [RDIS] bits in the IT status() register get clear, and Update Status changes to 0x05
    - Never passed step 6, because Update Status never changed to 0x05...

    Step 7: Discharge to empty

    Step 8: Relax state: Wait until Update Status changes to 0x06

    Thanks,

    Abel

  • 0 in reply to Abel Mendes
    TI__Genius 12215 points

    Abel

        As I mentioned above, for LiFE battery, you have to wait at least 48hour after FC set for Qmax update to happen, after 48hours, the gauge would start to check the voltage ramp slope is gentle enough to update DOD and Qmax

  • 0 in reply to Steven Yao
    Prodigy 10 points

    Hi again,

    I waited for about 60 hours, but unfortunately Qmax did not update at all.

    I will acquire the logs for the complete learning process, but, before I take all that long, could you please have a quick look on the srec file (config.zip)?  

    3276.config.zip

    Thanks

  • 0 in reply to Abel Mendes
    Prodigy 10 points

    Hi Steven,

    I have repeated the learning process, and again waited for more than 48h, but same result... update status never changed to 0x05.

    I've been dealing with this issue since long, so, please help me getting succeed with the characterisation of the battery. Find the complete log files below.

    Looking forward to hearing from you.

    Regards,

    Abel

    log_11_Nov.zip

  • 0 in reply to Abel Mendes
    TI__Genius 12215 points

    Hi, Abel

        Sorry for the delay on reply, I am on my time back this week.

        I found in the latest log file, the DODPassedQ suddenly changed to 0 at row#10655, this looks weird, as same thing is not seen in previous log file, this could prevent Qmax update to happen, I do not see any other dubious registers change at the same moment except the voltage suddenly raised up a little bit, I wonder if there would something exterior happened to cause the voltage raised up. Such things should be prevented during learning

        I also changed FC set and other 2 flash register setting which would make the qmax update happen easier. FC set to -1 is to have this bit synched to Valid charge detection so that you can disconnect charger once FC bit is set

    https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/196/0100_5F00_0_5F00_16_2D00_bq34z100G1_5F00_CDB_5F00_2_5F00_QME_5F00_5.srec 

  • 0 in reply to Steven Yao
    Prodigy 10 points

    Hi Steven,

    I got the same result with the file you changed. Below you can find the logs of the resulting learning cycle.

    log_17_Nov.zip

    Regards,

    Abel

  • 0 in reply to Abel Mendes
    TI__Genius 12215 points

    Abel

        I noticed that your flash update OK voltage in dataflash is set to 4000mV, this should not be the default value, please set this low enough so that the flash update during learning can happen in the first place.

        At present, the V at Charge Term is not updated at Valid charge termination(still default 4.2v), this needs to happen so that I can check other following events would happen as expected

        The default 2.8v should be OK for most applications, For LiFe battery with 16s, this can be set to 2v

  • 0 in reply to Steven Yao
    Prodigy 10 points

    I'll start re-running with those parameters next Thursday.

    In the meanwhile, I don't know if it could be the origin o the issue.... but, I just realized that parameter "Qmax Cell 0" is not initiated properly. For a 32Ah battery, and a scaling factor of 5, should this value be initialized with 6400 mAh?

  • 0 in reply to Abel Mendes
    TI__Genius 12215 points

    Abel

        From your logged file, I see that the customer should already be scaled by the factor of 5 when calibration, as the DOD passedQ is approximate to 6x00mAh already. I think setting this to 6400mAh in initial should be approprate

  • 0 in reply to Steven Yao
    Prodigy 10 points

    Hi Steven

    Here is the new learning cycle trial. I was not able to see any difference...

    log_24_nov.zip

  • 0 in reply to Abel Mendes
    TI__Genius 12215 points

    Hi, Abel

        The VOK already cleared during the relax stage after charge, the reason for Qmax being not updated is that theQmaxDOD0 being too low for Qmax Update to happen, the QmaxDOD0 - newly calculated DOD0 > 0.9*16384 should be assured, in your log file, the QmaxDOD0 is 14712, new DOD0 is 0, abs(QmaxDOD0 - newly calculated DOD0) < 0.9*16384, to make sure the requested condition can be met, you need to discharge the battery deeper before it is terminated, for a LiFe Battery, the 3.06 terminate voltage is way too high, please set it lower so that the condition mentioned above can be satisfied. A 2.5v to 2.8v terminate voltage should be common for many LiFe chemistries, please check with your cell vendor to identify the appropriate terminate voltage for discharge