• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 5 answers
  • Subscribers 64 subscribers
  • Views 707 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.

BQ4050: Issue of BQ4050RSMR SOC Jump

Brian Li
Genius 13850 points
Part Number: BQ4050
Other Parts Discussed in Thread: GPCCEDV, BQSTUDIO

Hi all,

This is Brian, I am writing to check with you about the SOC jump for BQ4050. Here are the details:

  • My customer uses BQ4050 as the solar monitor, which charges during the day and discharges during the night.
  • The cell they used has the function of heating mechanism, which begins when the core temperature is less than -3 °C and stop heating at 9 °C. Up to 6~7 heating movements may occur during the day.
  • The SOC jump always occurs when it is sunrise to start charging, SOC can jump from a lower charge (3% to 80%) to 100%. And at that time the voltage is generally higher, about 8.1V ~ 8.4V.
  • And also when the sun goes down, SOC can also jumps down occasionally, the jump value is usually the charge value of the sunrise charge time.
  • The jump always related to the temperature. It mostly appears when the temperature is below zero or it changes from above zero to below zero .

The attachment is partially abnormal data.

And here are two questions:

  • Whether or not the CEDV PROFILE1 has a strong constraint on RSOC? Why there’s nothing happens when a jump occurs?

 

Thanks &Best Regards,

Brian Li

电量跳变异常现象数据230209.rar

  • 0
    TI__Guru** 105590 points

    Hello Brian,

    With the type of application they have it will be very difficult to use CEDV gauging algorithm due to the temperature swings and low taper currents with solar charging.

    The log file shared does not contain any of the jumps so it's hard to tell what is going on exactly. We would need to see the full log for more detailed analysis. It is most likely due to the FCC like Dominik mentioned. CEDV learns the FCC during a qualified discharge, if the temperature at the first FCC learning is high, say 30degC, then you discharge at 0degC or lower, they will most likely be some jumps near EDV due to the voltage drop and increased resistance of the cell. Also if they are not able to get regular FCC updates during their application the error will increase as the battery ages.

    The GPCCEDV tool should always be used with CEDV gauges to get the coefficients for the application needs.

    Sincerely,

    Wyatt Keller

  • 0 in reply to Wyatt Keller
    TI__Genius 13850 points

    Hi Wyatt,

    Thank you for your reply.

    1.  I checked with my customer again and here is a clear register.

    It shows the registers before the jump, the Max Error is 26% and the Relative State of Charge is 34%.

    But after the jump, the Max Error will jump to 100% and the Relative State of Charge will jump to 99%.

    Other registers have no changes observed.

    And also we can see the cycle count is only 77, which means it is a new battery.

    (The jump also occurs on many batteries whose cycle count are only 12, 20)

    So there's nothing with battery ages. 

    Could you please help me to check the root cause of the jump?

    2.  My customer  also want to know if the CEDV PROFILE1 has a strong constraint on RSOC?

    Why there's nothing happens  when the jump of SOC occurs?

    And under what situation can it works?

    3. My customer wants to use software to reset the device to eliminate the error when the jump of SOC occurs.

    Will this reset have any impact on the subsequent normal operation?

    Thanks &Best Regards,

    Brian

  • 0 in reply to Brian Li
    TI__Guru** 105590 points

    Hello Brian,

    The Max Error should really only jump to 100% if there is a device reset, do they perform any resets in their host code now? When a reset is performed on the BQ4050 it will use the voltage to correlate to a SOC value and MaxError will go to 100% automatically.

    Cycle Count would indicate SOC jump after aging, but not jumps of Max Error to 100%.

    You could issue another reset command after the voltages are relaxed, but this will still cause max error to go to 100% until FCC is updated.

    Yes the CEDV profile that they uploaded from the GPCCEDV tool will impact how the RSOC changes near EDV. But I don't believe it is related to this case since RSOC is going up, not down.

    Sincerely,

    Wyatt Keller

  • 0 in reply to Wyatt Keller
    TI__Genius 13850 points

    Hello Wyatt,

    1.  I recommended them to check if there is a device reset, the answer is no. There's no resets in their host code.

    2.  To avoid the jump, they have tried to reset the device when their upper computer detects the wrong SOC.

          And they found there would not be another jump like that for a period of time.

          So could you please help me to check if the jump related to accumulative error?

    Thanks,

    Brian

  • 0 in reply to Brian Li
    TI__Guru** 105590 points

    Hello Brian,

    I don't believe the error is coming from accumulated error. If they are jumping to 100% then it is probably something to do with their VCT (Valid Charge Termination) settings (the CSYNC bit) which will force RSOC to 100% at VCT. It could also come from a device POR if the gauge temporarily loses power somehow it will force MaxError to 100% and recalculate SOC based on voltage.

    From the information given it seems to me to be one of these options. If they are able to capture a full BQStudio log of the issue then we could confirm root cause.

    Sincerely,

    Wyatt Keller

  • 0 in reply to Wyatt Keller
    TI__Genius 13850 points

    Hello Wyatt,

    1.  Here is their configuration settings of BQ4050.

         Could you please kindly check if there's anything wrong & how to change the settings to avoid the SOC jump?  

         520.gg.csv

    2.  Does BQ4050 has a minimum current limitation?

         (For example if the current below this limitation, the device will reset or shutdown by itself?) 

    Thanks,

    Brian

  • 0 in reply to Brian Li
    TI__Guru** 105590 points

    Hello Brian,

    From what I can tell the SOC drops is due to some type of reset, either POR or software.

    The .gg file was shared previously, it is not the same as the BQStudio log file. The BQStudio log is a register readout over time while the test is being performed. It is on the main registers page in BQStudio, not the data memory page.

    Minimum current through the sense resistor? There is no minimum current that must be flowing through the battery. If the supply voltage drops below the operating conditions stated in the datasheet then it could brown out and POR the gauge. This could also be cause by other pins being shorted causing the voltage supply to the gauge to drop, like shorting CHG/DSG pins or the PBI pin.

    Sincerely,

    Wyatt Keller

  • 0 in reply to Wyatt Keller
    TI__Genius 13850 points

    Hello Wyatt,

    They also notice another kind of SOC jump.

    When the sun sets and the temperature goes down, sometimes the SOC with jump from high to low.

    And they find the SOC value after jump is always equal to the SOC value at that day's sunrise charging time.

    Could you please help explain this?

  • 0 in reply to Brian Li
    TI__Guru** 105590 points

    Hello Brian,

    I can't provide more information without the data logs that show what the gauge is doing during the tests.

    It is not possible to determine root cause without this information, I can only give possible suggestions.

    Most common SOC jump causes for CEDV:

    1. Voltage hits one of the EDV points - forces SOC lower
    2. Charging VCT occurs and CSYNC is set, forces RemCap=FCC
    3. Gauge reset occurs (either by firmware or POR) causes gauge to recalculate based on voltage

    Max error is only an estimation and is slowly incremented during cycles.

    Sincerely,

    Wyatt Keller