Hi,

If a diode fails due to over-voltage or over-current or irregular temperature effect whatever it may be, but what is the state of the diode after fails. Either it is in open or short?

 

I've blown diodes up with forward over current and inductive reverse voltage. They always failed open.

 

Although it could be either way. You should test diodes out-of-circuit or you may get a false reading. Your DMM injects a low voltage into the diode to test the forward and reverse biasing. If the diode is still in the circuit, the DMM will measure the shortest path to ground, not necessarily through the diode.

 

Yes, either kind of failure is possible. Which way is more probable depends a bit on the circumstances, but you can't be too sure. Short-circuit failure, or in less severe cases excessive reverse leakage current, is common enough with moderate overloads or overheating, or some types of deterioration of the diode itself.

Open-circuit is more likely with massive overloads or extremely high temperatures (in an extreme case, the device could break apart / melt / vaporize!) or other causes of mechanical failure. Obviously the diode goes open if it is no longer in one piece.

 

Ah yes, I have used the phrase, "the familiar crackle of expoding thermoplastic".

Thinking WAY back, I think I remember a power supply with a shorted diode. I could tell by the powerful HUMMM for about 1.2 seconds before the breaker popped.

Bad diodes aren't dependable about which way they fail. Maybe that's why they are called, "bad".