I'm helping someone who observed a EZ2209 make a loud pop and emit flame while powered on by examining their EZ2209 to try and determine a root cause of the failure.
These images show the state of the driver after the heatsink was removed and initial cleaning:
Notably, the VS tracks have been largely vaporized with damage particularly on the pin 28 side of the board (towards the square cutout side). I further investigated C3 as a potential cause of the issue, as the positive side of the capacitor and its trace appears to have been completely destroyed with the opposing ground plane also blown off. The following image shows the state of C3 while in-situ:
Based on the apparent damage to C3 (missing a corner of the capacitor body besides the contact), I desoldered C3 and flipped it over to examine the corner that appeared to have been destroyed:
Considerable destruction of the capacitor body is observed. Note that the ground-side electrode separated from the MLCC while it was removed. In particular, the corner of the capacitor close to the camera exhibits considerable carbonization and is no longer geometrically intact. This supports a hypothesis that the failure was caused by a short to ground from VS through C3 caused by stress cracking originating at a corner of the VS side of the capacitor shearing the electrodes. Under this hypothesis, the damage to the capacitor (and the rest of the PCB) would be caused by the resulting large fault current as the motor capacitor discharging through the shorted electrodes in the capacitor, explaining the asymmetric damage to the capacitor centered near one of the contacts.
I suspect that this failure was caused by either mechanical force applied during assembly or through flexing of the PCB at some point during assembly, shipping, or installation. Based on this, redesign of the heatsink to redistribute mechanical loads away from the PCB, use of a flexible termination MLCC, or moving the MLCC away from potential load paths may address the issue.
I'm helping someone who observed a EZ2209 make a loud pop and emit flame while powered on by examining their EZ2209 to try and determine a root cause of the failure.
These images show the state of the driver after the heatsink was removed and initial cleaning:
Notably, the VS tracks have been largely vaporized with damage particularly on the pin 28 side of the board (towards the square cutout side). I further investigated C3 as a potential cause of the issue, as the positive side of the capacitor and its trace appears to have been completely destroyed with the opposing ground plane also blown off. The following image shows the state of C3 while in-situ:
Based on the apparent damage to C3 (missing a corner of the capacitor body besides the contact), I desoldered C3 and flipped it over to examine the corner that appeared to have been destroyed:
Considerable destruction of the capacitor body is observed. Note that the ground-side electrode separated from the MLCC while it was removed. In particular, the corner of the capacitor close to the camera exhibits considerable carbonization and is no longer geometrically intact. This supports a hypothesis that the failure was caused by a short to ground from VS through C3 caused by stress cracking originating at a corner of the VS side of the capacitor shearing the electrodes. Under this hypothesis, the damage to the capacitor (and the rest of the PCB) would be caused by the resulting large fault current as the motor capacitor discharging through the shorted electrodes in the capacitor, explaining the asymmetric damage to the capacitor centered near one of the contacts.
I suspect that this failure was caused by either mechanical force applied during assembly or through flexing of the PCB at some point during assembly, shipping, or installation. Based on this, redesign of the heatsink to redistribute mechanical loads away from the PCB, use of a flexible termination MLCC, or moving the MLCC away from potential load paths may address the issue.