• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 3 answers
  • Subscribers 62 subscribers
  • Views 690 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.

Original question:

BQ76940EVM: bq78350 SOC drop at some point

BQ76940EVM: bq78350 SOC estimation is wrong

user5769503
Prodigy 180 points
Part Number: BQ76940EVM
Other Parts Discussed in Thread: BQ78350-R1, BQ76940

Hello,

I'm currently working with the BQ76940EVM, EV2300 and the latest BqStudio version (1.3.86.3). I already used the GPC packer to get better SoC values. The BQ76940EVM worked quite good a while ago but somehow it turned bad without making any changes. I programmed the default firmware again and adjusted the settings again, but the SoC estimation is still wrong. It happens that the value of the SoC jumps to 0% or 100 % without making sense in the middle of a test. I have no idea why or what I can change. I attach a log file including some manual resets and test interruptions to get the correct SoC again and the actual gg file. I also attach the results of the GPC packer. I'm happy about any held.

regards Anna

report.rar

4034.TI.rar

  • 0
    TI__Guru 54715 points
    Hi User,
    I expect it will be next week before I can look at log files.
    Check that shunts are in place on the EVM. Check that wires are tight in the terminal blocks. Check for contamination or unexpected conductors on or under the board.
    Re-loading default firmware will restore all settings. Re-calibration will be needed after firmware load.
    Carefully check each parameter you load.
    If you are using average cell gauging look at the value in the log to compare to the pending EDV.
    If you are using minimum cell gauging look at the cell voltages for a low or erratic voltage and compare to the pending EDV. An erratic reading may indicate a loose wire or damaged board.
  • 0 in reply to WM5295
    Prodigy 180 points

    Hi,

    thanks that you will have a look at my log files. The shunts are in place, the wires are tight and and I couldnt find any contaminations or unexpected conductors on or under the board. After re-loading the default firmware, I re-calibrated and I checked each paramter.

    regards

    Anna

  • 0 in reply to user5769503
    TI__Guru 54715 points
    Hi Anna,
    On reset the capacity will be estimated from the voltage.
    At 2:00:24 PM there is a state of charge jump 66 to 56% with discontinuity in data.
    At 2:36:58 PM the log shows voltage about 3.6 and goes to 100% RSOC. This looks like you may have loaded a LiFePO4 chemistry file, but voltage is only slightly above the voltage above.
    You will probably need to think through each event looking at the reported status.
  • 0 in reply to WM5295
    Prodigy 180 points

    Hi WM5295,

    On reset the capacity will be estimated from the voltage: so I think the OCV curve is correct then after a reset the SoC is fine. But after a reset, when the chip uses the CEDV the SoC turns wrong. I used the chemID 1737, which should be fine, it already worked before with this chemID. I dont understand what you mean with "but voltage is only slightly above"? I already tried to think through every value but I'm nto able to find the reason.... Could you have a look at my gg file? Maybe you can find a mistake in there. I can't use the chip with this behaviour.

    regards

    Anna

  • 0 in reply to user5769503
    TI__Guru 54715 points
    Hi Anna,
    I don't recognize a configuration problem, limits are set wide but that is common for initial evaluation. It is an early version of the firmware which is not recommended. The bq78350-R1 is recommended, it can be downloaded from the product page www.ti.com/.../toolssoftware . There is also an R2 firmware available, it has some improvements/options.
    The voltage comment was that the estimations were very different with little difference in measured voltage. There were voltage transients apparent.
    On reset the part will use the measured voltage for estimation. If reset with current flow the cell voltage will be recovering from the current and will not represent an OCV at rest voltage curve. So capacity estimation will be wrong.
    Gauging should be good if the pack is calibrated and it is allowed to learn.
  • 0 in reply to WM5295
    Prodigy 180 points

    Hi WM5295,

    I had calibration problems with the bq78350-R1 firmware but I will try it again, so far I don't see any other possibilities to make the BQ76940 work. Thanks, now I understand what you mean with voltage difference, a reset during charge makes no sense. The pack is calibrated and you recommend to set the SC Bit always to 1 in the CEDV gauging config?

    Could you tell me how the RSOC is caculated? Is it RSOC = RC / DC (or FCC) * 100?

    regards

    Anna

  • 0 in reply to user5769503
    TI__Guru 54715 points
    Hi Anna,
    SC is typically left as default, 0. SC =1 will reduce the learned capacity some, it assumes the basic charger will not be as effective as the smart charger. See the last paragraph in section 9.1.3 of the technical reference manual.
    RSOC is basically RC/FCC x 100.