• State TI Thinks Resolved
  • Locked Locked
  • Replies 14 replies
  • Answers 6 answers
  • Subscribers 63 subscribers
  • Views 818 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.

BQ25790: System shutdown when removing battery

Hardy Xue
Expert 3640 points
Part Number: BQ25790
Other Parts Discussed in Thread: BQ25792, BQ25703A, BQ25713, BQ25710, BQ25700A, BQ25708, BQ24780S, BQ24800

Hi Team,

My customer uses AC(adapter)+DC(battery) to supply power to the system. But when they remove battery, system will shutdown.

The adapter can supply enough power to the system. How can we avoid system shutdown when removing battery by EC or hardware?

Best regards,

Hardy

  • 0
    TI__Guru**** 167196 points

    Hardy,

    What do the I2C fault and status registers report immediately at system shutdown?

    Regards,

    Jeff

  • 0 in reply to Jeff F
    TI__Expert 3640 points

    Hi Jeff,

    It is shutdown suddenly and can’t get the I2C log for this moment.

    The registers status of charger before we unplug the battery as attached.

    F7A Charger register setup for _bq25790 20201118.xlsx

    Also, we check the VBUS, VSYS and VBAT waveform as below. could you help to find any clue.

    Best regards,

    Hardy

  • 0 in reply to Hardy Xue
    TI__Guru**** 167196 points

    Hi Hardy,

    With charge enabled and the minimum recommended capacitance at SYS and BAT, when the battery is removed, the charger SYS and BAT voltage overshoots.  If you can disable charge before battery removal or add more capacitance on BAT prevents the overshoot.  In theory, the BAT pin capacitance needs to be at least CBAT=ICHRGmax*(0.0005) / (VBATREG-VBATLOWV), to prevent overshoot with charge enabled.  I have not tested on the bench but with test either later today or tomorrow and get back to you no later than Monday.

    Regards,

    Jeff

  • 0 in reply to Jeff F
    TI__Guru**** 167196 points

    Hardy,

    My equation doesn't really work.  There are too many variables, like load current on SYS and input current limit, etc. 

    One clarification I should have made previously.  The EVM artificially has the TS pin report that the battery is present and at a safe temperature for charging.  In your real application, if you remove the three terminal battery, with the third terminal connected the battery thermistor (which would connect to TS pin), the charger will report a fault and auto disable charge.  With charge disable, SYS will not collapse.

    Regards,

    Jeff

  • 0 in reply to Jeff F
    TI__Expert 3640 points

    Hi Jeff,

    Thank you for your reply.

    I let my customer to try disable EN_CHG and do the test again. But still need to wait for result.

    If it doesn't work, do you have any other suggestion for doing any test or need us to provide other information?

    Best regards,

    Hardy

  • 0 in reply to Hardy Xue
    TI__Guru**** 167196 points

    Hardy,

    Disabling charge via the register bit or the CE bit OR configuring TS pin appropriately for the application should work.

    Regards,

    Jeff

  • 0 in reply to Jeff F
    TI__Expert 3640 points

    Hi Jeff,

    When we disable charger, the system will not shutdown. The waveform is as below.

    According to previous waveform, it look like not VSYS OVP cause system shutdown. Why VSYS will not drop when we set bit of EN_CHG = 0?

    But there is no way for EC to know when the battery will be removed. So, It is impossible for EC side to disable it “before” battery be removed.

    Do you have other suggestion?

    Best regard,

    Hardy

  • 0 in reply to Hardy Xue
    TI__Guru**** 167196 points

    Hi Hardy,

    V(BAT) overshoots and V(SYS) follows it up, causing BATOVP and SYSOVP.  Both OVPs cause the converter to shut down, to prevent what it thinks is overvoltage.  If you disable charge, BATOVP doesn't occur.  If you check the  I2C fault registers, you will see both faults.

    If battery has a thermistor that is supposed to be connected to TS but is not connected to TS, a TS fault occurs and charge is auto disabled. 

    The only other way to prevent the BATOVP with charge enabled and TS pin disabled is to add additional (>470uF) capacitance on the BAT pin or use ILIM_HIZ pin resistor to limit input current, which limits output power to both the battery and the system.

    Regards,

    Jeff

  • 0 in reply to Jeff F
    TI__Expert 3640 points

    Hi Jeff,

    1. Why VBAT will overshoot when removing battery? Charging to CBAT?

    2. According customer's register setting, do we have other way to prevent the BATOVP with charge enabled by setting register?

    3. In my opinion, BQ25790 look like doesn't support without battery, right?

    Best regards,

    Hardy

  • 0 in reply to Hardy Xue
    TI__Guru**** 167196 points

    Hardy,

    The charge current loop turns on hard and fast and charge voltage loop responds too slowly with the minimum capacitance on the board.  With additional capacitance (>470uF) on BAT to absorb the overshoot, there will be no fault.  Also, 3 terminal battery with thermistor that is connected to TS pin will cause a TS fault, disabling charge, when battery is removed.  The EVM has TS artificially set to 10kohm thermistor using shunt on JP18.

    Regards,

    Jeff

  • 0 in reply to Jeff F
    TI__Expert 3640 points

    Hi Jeff,

    1. Customer can't add 470uF on their board because there is no space.

    2. According to the TS pin setting as below, I think the suggestion does not seem to be relevant.

    3. The system cannot set a too low input current limit because the customer will not let battery in discharged state when adapter is inserted. So, do we have other methods to avoid system shutdown?

    Best regards,

    Hardy

  • 0 in reply to Hardy Xue
    TI__Guru**** 167196 points

    Hi Hardy,

    There is no other way to prevent OVP when no battery when attached charge is enabled.  This week I will be evaluating new versions of the BQ25792 with some additional features.  One of those features is supposed to prevent OVP and therefore system shutdown.  I will let you know if these later versions prevent the problem.

    Regards,

    Jeff

  • 0 in reply to Jeff F
    TI__Expert 3640 points

    Hi Jeff,

    Thanks for your reply.

    Furthermore, TI have many charger solution like BQ25700A, BQ25703A, BQ25708, BQ25710, BQ257018, BQ25713, BQ24780S, BQ25790 and BQ24800. It is mentioned in the datasheet of BQ25790 that remove battery is not supported. 

    1. Which device will have same OVP situation with BQ25790?

    2. Why system with BQ25713 will not shutdown when remove battery?

    Best regards,

    Hardy

  • 0 in reply to Hardy Xue
    TI__Guru**** 167196 points

    Hi Hardy,

    The charge current startup time is faster and the OVP response timing is different on the BQ2579x compared to the BQ25713.

    Regards,

    Jeff