BQ25720: Input Ceramic Capacitor Shorting After Field Use

Part Number: BQ25720

Tool/software:

Hi TI Team,
I’m seeing intermittent failures of input ceramic capacitors in my BQ25720RSNR charger design after extended field operation. Units pass all initial tests and run reliably during validation, but after some weeks/months in the field, one or more of the six input 10µF, 50V X7R ceramic capacitors (C19–C24) will short.
Input: 24V DC
Output: Up to 8.4V, max 4A
Input capacitors: 6 × 10µF, 50V X7R MLCCs
Additional: 1Ω damping resistor, 0.47µF cap, TVS diode on input
No visible damage: No cracks, joints are good, caps >0.2" from PCB edges
Failures show up randomly after prolonged use; not all caps always fail, sometimes just some of them. Soldering/process is sound and there’s no visible physical/mechanical defect.
Could you suggest likely root causes or improvements, especially regarding input cap selection for this IC and application? Any reliability guidance for high DC bias/hot-plug/VAP scenarios would be appreciated.

Schematic snippet attached.

  • Hi, Amit,

    I guess you did not install C17 from the note. If you hotplug without a damping capacitor C17, the input capacitor will form a resonance with the cable inductance and generates a voltage spike on the input caps. It is highly recommended to have adequate input capacitances ( a mix of ceramic caps and tantalum caps), for high input voltage applications. Avoid hotplugging unless a damping capacitor is installed. 

    Regards,

    Tiger