I’ve built a buck circuit with MP9943 and right now upon testing with a electronics load I’ve found a very strange behavior that under CC mode the circuit works perfectly fine but under CW mode the circuit will when into unknown protection mode or some sort of failed to regulate by shorting itself but not fully shorted, and the switching waveform is short bursting. This will only happened when the load is connected first and power up. If I power up the circuit first then connect the load it would not have this problem. This issue also exists when an actual load is connected. And it brokes a few of the MP9943 chips. However, another MP9943 buck circuit within the same pcb board but with 5V output would not have this issue. Furthermore, some I believed broken MP9943 chips I’ve tried also experience a slightly different behavior, the switching waveform is completely messed up rather than a typical short burst. And the problem on exists when input voltage is above 24V, anything lower than 24V is seem to be fine. This is very confusing.
For all wattages, what load conditions did you end up running? I have a feeling that there may be some input current issues. Perhaps this is a symptom of some input conditions?
For all wattages, was the load connected first and then powered up? If you want to have the case in which the IC to power up and then have the load supplied would be to adjust the soft start time in order to have this adjusted. In the case this is a main load condition for your application, then this would be understood.
On the subject of input conditions as I initially mentioned, if possible, ensure the power supply is supplying a sufficient enough current or could be varying switch current between 5V and 9V at the output. Not doing so could induce spikes that can potentially damage the IC. Switch nodes here should be observed at these outputs to compare and perhaps isolate the issue. Additionally, this may correlate to the damage that you observed from what you have described.
Let me know if you have tried what I have said and check the input and switch characteristics at these voltages and methods on corrective current measurements within threshold.
I ask all of these questions because upon reviewing your schematic, I see no outstanding issues. If you happen to have a layout to show as well, which depending on how the trace widths, switch node placement, and certain component replacements, this may also have the potential to influence spikes as well.
Here are the PCB Layout Guidelines as shown in the bottom of the datasheet:
