• State TI Thinks Resolved
  • Locked Locked
  • Replies 18 replies
  • Answers 1 answer
  • Subscribers 65 subscribers
  • Views 3257 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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: SoH calculation requirements

IKER JOARISTI URTUSAGASTI
Prodigy 130 points
Part Number: BQ34Z100-G1
Other Parts Discussed in Thread: BQ34Z100,

In response to the already locked issue (https://e2e.ti.com/support/power_management/battery_management/f/180/t/349402?tisearch=e2e-sitesearch&keymatch=SoH%20calculation), we would appreciate to receive some feedback from your part (we tried what TEC recommended but did not get the right answer from the bq34z100). As far as we know, we fulfill all the requirements for entering RELAX mode but cannot see the OCVtaken flag update, nor the Qmax update.

  • 0
    TI__Guru* 76230 points

    Hi Iker,

    Can you summarize what the issue given that this is a new post? Are you having Learning cycle issues or wrong SOH calculation issues. Do you have log files and gg files depicting the issue?

    thanks
    Onyx

  • 0 in reply to Onyx Ahiakwo
    Prodigy 130 points

    Hi Onyx,

    We are having trouble with the Soh calculation, once the golden image has been loaded to the bq34z100 and our system is running (in fact it has been running for a year or so). The fact is that after cycling the batteries under normal operation conditions, we still cannot see the SoH updates (every single battery is showing SoH=100%). We run  the image optimization cycles for low temperature operation a few a months ago and it worked perfectly. The issue now is the SoH update, as we think we fulfill the requirements for Relax mode enetering and further flag activating but still do not see results.

    Please find attached the configuration files for the golden image and generated files.

    Regards

    3AhBattrevG.rar

  • 0 in reply to IKER JOARISTI URTUSAGASTI
    TI__Expert 6640 points
    Hi Onyx,

    could you please take care of it and advice the customer?

    As you can see, the customer is locked with this issue.
    If you need other info, please refer to me and I will make sure that you receive it.

    Thank you very much in advance.

    Best regards,
    Massimiliano De Cristofaro
    Texas Instruments FAE
  • 0 in reply to Massimiliano De Cristofaro
    TI__Guru* 76230 points
    hi Iker,
    I suggest you run a reconditioning cycle so the gauge learns qmax and updates the resistance table.
    -You can either dicharge to 50%, rest 5 hours charge to full and rest 2 hours to get qmax update and then discharge to get Ra updates. Repeat this process 3 times since you have the qmax and Ra filter values enabled.
    -or you can change update status to 00 and Raflags to ff55 and ffff and then run the optimization learning cycle. this will disable the filters and allow learning to occur as though you are starting afresh from the begining. Make sure to sent IT enable command if you set update status to 00.

    make sure to log your data and extract a gg file before and after the test

    thanks
    Onyx
  • 0 in reply to Onyx Ahiakwo
    Prodigy 130 points

    Hi Onyx,

    Thanks for the fast reply.

    1) For the first solution, do you mean that we need to perform once again the learning cycle or is it just an operational condition we need to meet to obtain the Qmax and Ra table update?

    2) For the second proposal, should we proceed with a real "optimization cycle", as explained on document SLUUBD0, uploading the 2 temperature logs + configuration file + *gg.csv file; or would it be enough with a normal learning cycle( the one that lets you reach the "LearnedStatus=0x06", as explained in the bq34z100-g1 datasheet?)

    Regards

  • 0 in reply to IKER JOARISTI URTUSAGASTI
    TI__Guru* 76230 points
    Hi Iker,

    What the learning cycle does is just update qmax and resistance. Once the golden file is created after the very first learning cycle during production is done, you just need to have the gauge learn qmax and ra in the field.
    The condition for qmax update in the field is having at least 37% of design capacity leave or enter the battery. This charge or discharge should be inbetween two relaxation periods.

    If you haven't had qmax update all this while, just one qmax update will not be sufficient because there are qmax filters in place that limits how much qmax can change during just one cycle. So that is why in point 1 i had said run several cycles to get qmax to converge to the actual value.

    point 2 involves disabling the filters, which is basically the same thing as running learning cycle from the start.

    thanks
    Onyx
  • 0 in reply to Onyx Ahiakwo
    Prodigy 130 points

    Hi Onyx,

    "The condition for qmax update in the field is having at least 37% of design capacity leave or enter the battery. This charge or discharge should be inbetween two relaxation periods."

    We usually work between 20 and 60% of SoC for batterypack lifetime prolongation but we monthly charge the batterypack to 100% SoC, as we thought it was the stimulus for entering Relax mode.

    1) What you are trying to explain is that we need to enter one Relax mode (lets say at SoC=0%), then charge to SoC=100% and wait to enter the other Relax mode; and then discharge to SoC>= 37%?

    2) What does to "enter the battery" mean?

    Regards

  • 0 in reply to IKER JOARISTI URTUSAGASTI
    TI__Guru* 76230 points
    Hi Iker
    1. Yes
    2. I meant passed charge going into the battery or out of the battery.i.e charge or discharge.
    thanks
    Onyx
  • 0 in reply to Onyx Ahiakwo
    Prodigy 130 points

    Hi Onyx,

    Thanks for the reply. One last question over your reccomendation:

    Is it necessary to reach 0% to take the OCV measurement or is it possible to cut charge and discharge processes at lets say 10%?. (This would benefit us in terms of final field application).

    We have no problem on reaching the SoC=100% for entering one of the Relax modes but as I said, is it possible to enter the other Relax mode at SoC=10%? (I cannot see it written anywhere in the datasheet)

    Regards

  • 0 in reply to IKER JOARISTI URTUSAGASTI
    TI__Guru* 76230 points
    Hi Iker,
    There has to be 90% of design capacity passed charge for the initial golden learning to occur OR 37% of design capacity for field learning to occur.

    so if your battery is 1000mA, you have to have 900mA accumulated charge either in the charge or discharge direction. It has to be either charge or discharge and not both
    thanks
    Onyx
  • 0 in reply to Onyx Ahiakwo
    Prodigy 130 points

    Hi Onyx,

    I am still not able to figure out a clear indication on how to preoceed.

    A few posts ago you said that the Qmax update had to occur between two relaxation periods, to allow the algorithm to converge.

    Those two (or more) Relaxation periods, when do they need to occur? Both need to occur in either the discharge or charge process?

    i.e. during discharge process from full charge:

    "Relax 1" at SoC=100%

    "Relax 2" at SoC=60% (as it is higher than SoC=37%)

    Would that scenario work properly for field SoH updates?

    Regards

  • 0 in reply to IKER JOARISTI URTUSAGASTI
    TI__Guru* 76230 points
    Hi Iker

    Relax 2 should be at 50% since your battery may have aged and 60% may not be actual 60% since your qmax has never updated.

    Also check to make sure you don't relax at the flat zone of your chem id.
    Onyx
  • 0 in reply to Onyx Ahiakwo
    Prodigy 130 points

    Hi Onyx,

    We have already made the recommended test. We started the test at SoC=100% and waited for flag OCVtaken to change state from '1' to '0'. Then discharged the pack at 0.4C discharge rate to SoC=50% and waited for OCVtaken flag to update. But the flag did not update and we could not see any SoH update.

    Our battery-pack chemistry is LiFePO4 with and optimized cycle already run (if this could help you to with the issue)

    Any guess on what the error could have been?

    Regards

  • 0 in reply to IKER JOARISTI URTUSAGASTI
    Expert 2475 points

    One thing to keep in mind about LiFePO4 cells is that they have a very flat discharge profile with a wide disqualified voltage range. Qmax cannot update if the cells rest inside the disqualified region and this will probably require a 100% discharge to update Qmax on your cells. You can read Qmax to see whether it is updating. SOH should update, if Qmax updates.

  • 0 in reply to TEC
    Prodigy 130 points

    Thanki you Tec,

    We sort of guessed that could happen due to flat regions. We will try lowering the SoC down to low voltage regions to check if it works.

    Regards

  • 0 in reply to IKER JOARISTI URTUSAGASTI
    Prodigy 130 points

    Hi all,

    We did the above described process with sloped regions (0-10% and100%) of the chemistry curve and still see no response.  We have monitored variables related to qmax updating sequence such as OCVtaken flag, FIRSTdod, LearnedStatus, QEN, VOK but these do not change values as expected.

    Regards

  • 0 in reply to IKER JOARISTI URTUSAGASTI
    TI__Mastermind 25955 points
    Hi Iker,

    Thank you for performing the above described process. Please send a datasheet of the cell, full log from your test and your senc file for further analysis.

    Sincerely,
    Bryan Kahler
  • 0 in reply to Bryan Kahler
    TI__Mastermind 25955 points
    Hi Iker,

    I'm closing this thread due to the time elapsed since the last response. If you get a chance to take those logs, please start a new thread, reference this thread, and attach the datasheet of the cell, configuration details, full log from the test, ChemID used (if any), as well as your gg.csv file.

    Thanks!

    Sincerely,
    Bryan Kahler