We are using the LM5060 with an functional input range of 12-56V. We probably don't need overvoltage protection, but since the LM5060 supports it we included it, set to nominal 62V (So it won't trip when running at VIN=56V).
In testing this circuit we ran VIN up to this trip threshold (always close to 62V), and the LM5060 fails EVERY TIME. I have repeated this test many times. If I raise the OVP threshold to effectively disable OVP I can go up to VIN = 65V (max supported) with no issues. If I lower the OVP to 48V it works properly. 62V is close to the max 65V of this part, but it appears as though this should still work.
This is with no load. I have scope captures of VIN when the trip occurs and the inductive spike is only ~150mV. We do have a schottky (SMAJ58CA).
Replacing the LM5060 always fixes the board - other components seem fine. When the IC fails pin 2 melts off. I worried that this was the sensitive pin, so added a 100-ohm series to make an RC with its 0.1uF cap - still failed, no effect (Actually pretty spectacular fireball from melting that 100-ohm resistor when the IC failed!). I have played with different input bulk electrolytics, also still failed, no effect. It seems fine when the OVP is lower, so it definitely has something to do with having the OVP trip at a high VIN voltage.
Any thoughts or suggestions?
I will try to add a snip of our sch.