Reliability Testing and Statistics

Estimated reading time: 2 minutes

What is reliability? Reliability is the probability of a product performing its intended function over its specified period of usage, and under specified operating conditions, in a manner that meets or exceeds customer expectations.

The catalyst for the emergence of reliability analysis goes back to 1906, when Lee de Forest invented the triode vacuum tube. The vacuum tube initiated the electronics revolution, enabling a series of applications such as the radio, television, radar, etc. The vacuum tube is recognized by many as the technology that allowed the Allies to win the so-called “Wizard War” of World War II. Ironically the vacuum tube was the main cause of equipment failure and tube replacements were needed five times more often than all other equipment.

In 1937, Waloddi Weibull invented the Weibull distribution, and in 1951 he delivered his hallmark American paper on this subject claiming that his distribution applied to a wide range of problems. He showed several examples ranging from fiber strength of cotton to the fatigue life of steel. In 1952 the Advisory Group on Reliability of Electronic Equipment (AGREE) was jointly established by the Department of Defense (DoD) and the American Electronics Industry. In 1956 the Radio Corporation of America (RCA), a major manufacturer of vacuum tubes, released a significant report, “Reliability Stress Analysis for Electronic Equipment” (TR-1100), which presented a number of models for estimating failure rates. Then, in 1959, the “RADC Reliability Notebook” came into existence, and in 1961, a military reliability prediction handbook format known as MIL-HDBK-217 was published. Numerous other reliability documents have been published ever since.

Historically, reliability engineering of electronics has been dominated by the belief that the life or percentage of complex hardware failures that occur over time can be estimated, predicted, or modeled. Ironically, there is little, if any, empirical field data from the vast majority of verified failures that shows any correlation with calculated predictions of failure rates.

Modern reliability tests try to simulate assembly and field use conditions. These methods include solder shock, air-to-air thermal cycling, liquid-to-liquid thermal cycling, interconnect stress test (IST), and highly accelerated thermal stress (HATS, an air-to-air method). Reliability test conditions can and do vary widely by OEMs. We see resistive heating pre-conditioning cycles at 260°C for six times, followed by resistive heating thermal cycling at 190°C, current induced thermal cycle of -23°C to +220°C, and air-to-air cycles of -60°C to +160°C. Why such harsh testing conditions? What are we trying to prove? And can it be proved? At this rate, testing conditions are destined to become even more severe with the acceptance criteria shrinking to very tight tolerances. Often in reliability testing, the threshold value for determining failure is chosen subjectively.

Read The Full Article Here

Editor's Note: This article originally appeared in the March 2015 issue of The PCB Magazine