We are experiencing what appears to be a thermal shutdown of the LM5022 device. It was designed (using WebBench) for 30W power delivery (48V from 18Vin), but if we attempt to draw >22W, then the switcher will shut down after some time period. As the load increases, the time period decreases, strongly suggesting a thermal condition. However, measuring the body temperature of the device using an electrical temp probe shows around 62C, but the data sheet indicates that shutdown happens when the die is >165C. I have verified that the UVLO pin is not coming anywhere close to the shut-down voltage threshold, and the other control pins are pulled to the inactive state. Also, it runs with the internal oscillator.
The PCB layout has good thermal pads for the coil and FET, but the pads to the GND pin, VIN pin, and FET drive pin of the LM5022 are 0.25mm wide only.
Questions: Could the internal die temperature be significantly higher than the body temperature due to poor thermal coupling?
If so, would adding thermal pads to the corner pins of the LM5022 be sufficient to alleviate this problem?
I don't have any tools for thermal simulation, so I don't want to spend resources on a new layout if it doesn't fix the problem.
Regards,
Brian J Carrigna