The device is being used in convertor producing +5.3V from a +12V input. The device appears to fail upon a reduction in power demand with pin 2 failing S/C to pin 1 (GND), this failure is very consistent and repeatable. I do not believe that board layout or feedback resistor values are the cause of this failure.
After looking at the waveform on the SW pin 2 (the switched side of the inductor) it appears that the low side transistor is switched off whist there is still energy present in the inductor, this conclusion was drawn from the observation of the waveform suddenly going negative by almost 1.5V.
As an experiment I connected a Schottky diode (1A rated) from the SW pin 2 to 0V (cathode to switch side), the sudden dip in the waveform is now clamped to around -200mV. With this additional diode no further failures have occurred even after rigorous testing on several units.
The design is as per the datasheet reference design but with the feedback resistors scaled for a +5.3V output. Any thoughts on this are welcome.
