scope:

Life tests are run for various purposes. Tests run to detect the level of infant mortality involve short time duration; unless the percentage of devices having infant mortality is extremely high, the sample size specified in this document is not nearly sufficient. Tests to determine device lifetime for a specific application may be conducted by or for a customer; here the stress and test conditions may be specific to the application. Other life tests are run by a manufacturer and are concerned with determining the lifetime of devices under typical or extreme operation.

It is assumed that a given MMIC product is designated either as a power, general purpose, or low-noise (small signal) device, based on the intended use of the product. For low-noise devices and passive components, the RF voltage levels are small enough that it may be possible to simulate the stress accurately by imposing only DC bias.

This document applies both to packaged and unpackaged devices. Where the devices are bare die or in packages not suitable for stressing at high temperature, additional failures may occur due to the packaging, which are not considered part of the product being tested. If this occurs, such failures should be excluded from the population when calculating predicted failure rate.

Since it is unlikely that every structure and failure mechanism in a MMIC is known when these guidelines are written, some judgment must be used in applying the principles of this document to the specific tests. Where an unusual circumstance forces exceptions to these guidelines, the exception shall be stated in the life test report.