Burnin specs

[MoEng]

Hi everybody
We've got an electric circuit which is specified to be able to operate in the temperature range -40 to +85 Celsius. My questions are:
-at what temperature (s) should we test the circuit for burnin / artifical aging reasons? Is that +85 Celsius or even above? Or is it necessary to change the temperature over the whole range to have a better chance that some parts will fail?
-what about the duration of the test, how does this depends of the temperature? Are there specs avaible to choose from, or some well known values?

Thank you for your help

 

[IRstuff]

It depends on the conditions that are specified for the projected life. If your life requirement assumes 100% operation at 85C, then acceleration would need to be at a higher temperature.

Unfortunately, you're limited to a best-case max of 125C and possibly less than that, depending on the parts being used.

The Arrhenius rate equation is usually used to determine the amount of acceleration, but it requires determination of the rate constant, which itself requires a bunch of thermal life tests to figure out.

The caveat is that the acceleration temperature should not aggravate any mechanisms that would not ordinarily appear in the operating life.

You may wish to consult the various MIL-STDs that cover topics such as MIL-STD-883

TTFN