LM2676-3.3 Circuit failures, the magic smoke is escaping.

[barttech]

For both my MOD5234 and MOD5213 3.3V supplies I've been using a circuit pretty similar to the one on the MOD5234 Dev Board. It has a LM2676-3.3 switching regulator, inductor, a few caps and a diode. I ran the design through National Semis web-based design program and used their values and have had pretty good luck with it, but not a perfect as I need.
Occasionally the diode, a 1N5819, rated 1A, 50V, will fail short. It sometimes happens during an accidental overload (my shaky hands have trouble probing these small chips), other times there seems to be no reason, but smoke ensues. This usually causes the voltage to go out of regulation, often high, and destroys various things, like MOD5234s.
Anyone else using this circuit who has a recommendation for a better diode?
Thanks,
Sam

[v8dave]

Hi Sam,

I use the following for the same Nat Semi device.

30WQ04FN

This is a 3.5A device and less likely to fail if you draw more than 1A which is likely with the 1N5817. My old designs used that and it was also failing due to shorts or excess current. I have had no issues since switching to the 30WQ04FN.

Cheer
Dave...

[barttech]

That is exactly what I wanted to hear! Thanks, Dave, you've saved me from rebuilding a small inventory of boards.
Sam

[v8dave]

Your only downside is that the 30WQ04FN is SMD but you should be able to fit it on the current boards and rework for a revision for later.

Dave...