BQ25883: Failure in IC after high current battery charging

Part Number: BQ25883

Tool/software:

Hello,

I’m working with a custom BMS board based on the BQ25883. I ran into a failure scenario that I’d like to understand better.

After disabling the I²C watchdog timer and charging the battery at around 1.5 A for a few seconds, a short circuit developed inside the IC between VBUS and VBAT/VSYS. At that point, VBUS measured 8.4 V, which was the same as the battery voltage—almost as if the high-side switching FET (QHS, Q2) had shorted.

I inspected the external MOSFETs between VBUS, PMID and SW, SYS, but none appeared to be damaged. 

After replacing the IC and limiting the charging current to 600 mA, the failure hasn’t occurred again.

My question is: Why would this failure happen in the first place, even though the charging current was still below the IC’s rated limit. Given that the inductor and cap values are of the same values recommended in the IC datasheet and EVB schematic

  • Hi,

    Did you take any waveforms during the event?

    Thanks,

    Ning

  • I didn’t capture any waveforms at the moment of failure, but the damaged chip is still available if I need to take further measurements. At the time, I was only monitoring battery voltage and current using a shunt resistor and a DMM.

    When I sent the “increase battery voltage” command, the battery voltage continued to rise as expected and charging proceeded normally—until the voltage suddenly dropped to zero. At that point, I measured the bus voltage and found it sitting at 8.4 V, the same as the battery voltage, which clearly indicated that an internal short had developed inside the IC.

  • Another observation: I took the failed IC and assembled it onto a BQ25883EVM. When I connected VBUS, the PG and Status LEDs started blinking together (No load). And it stops blinking on connecting load. 

  • Hi,

    Could you please help to measure each pin to GND pin resistance & pin to pin resistance and see if any short/low resistance on the unit alone (not soldered on board)?

    Thanks,

    Ning.

  • Hi,

    It has been quite for a while. The thread is considered solved and closed.

    Thanks,

    Ning