Temperature is widely viewed as a major influence on failures of electronic equipment. Failure Prediction Methodology (FPM), such as MIL-HDBK-217 is an often quoted reference describing the temperature influence, often depicted as being exponential in effect, affecting wide classes of electronic component parts. The FPM concept of a constant failure rate that is accelerated by various environmental influences is widely applied beyond its validity. Misapplications of the reliability models in current use may cause failure avoidance efforts such as temperature reduction and parts quality selections not to yield anticipated overall results. The cost and complexity effects can be significant: temperature reduction, for example, can result in expensive system implementations in some cases whose costs and complexities may exceed the anticipated benefits in reliability. Due to industry wide continuing improvements in component quality over the years, the continuing quest for electronics reliability should change emphasis from attention to electronic parts to activities that address assembly and processes. This paper discusses the ways the temperature ingredient of reliability and similar concepts may be currently applied, with examples to illustrate disparities between anticipations and realizations. Alternate approaches are offered and their possible implementations are discussed.

This content is only available via PDF.
You do not currently have access to this content.