What causes transistors to fail?

[pucho812]

Ok this is unusual. The issue resolved. But I have a clue as to why. Changes in thermal temperature would cause the discrete opamp to work until boom failure with dc at measured output. This is after running it doing a burn in test. So from a cold start it would be ok, then after a while boom I see and measure dc. So, while not my usual protocol I replaced all the old transistors. Luckily the parts total is like 2 dollars in total and the bc456 and 556’s are still available in thru-hole. Since then I ran the unit without issue. I can only assume that years of being on 24/7 took their toll.
While I don’t have any other explanation or details, a broken clock is even correct twice a day.

 

[clintrubber]

I can only assume that years of being on 24/7 took their toll.

As dale116dot7 also mentioned already, all electronic parts have a certain lifespan. Some live longer while charged/on (say e-caps), others the opposite.

Semiconductor-devices are designed for a certain lifespan, as in being checked for a certain max shift of parameters, given a certain mission profile (some table with various operating time & temperature combinations).
So that 24/7 usage may have seen the finish flag.


Another thing that came to mind (open door, but just in case): socket/pins of the DOA making proper contact at all times?