A virtual predictive environment for monitoring reliability and maintainability of printed circuit boards

There are a number of ways to express reliability, but one commonly used method is mean time between failures (MTBF). To quantify reliability, a test was conducted based on a set of time-to-failure (TTF) sample data. This test can then be fitted to a probability cumulative function, F (t). Thus the "reliability" is expressed explicitly by: R(t)=1-F(t). In this paper, a new test methodology to estimate the total life time and reliability of the printed circuit board (PCB) is presented where new predictive sets of TTF were found to represent the predictive upcoming TTF for each PCB component. The approaches "to repairing" or "to replacing" failed components were used during this methodology simulation and the effect of such on the total life and reliability of that PCB was determined. A virtual reliability multi-meter was created in which probes can be placed on any two nodes in a PCB. Predictive cumulative reliability is calculated for all components between those two nodes and eventually appears on the virtual reliability multi-meter screen. Quantitative accelerated life tests are designed to quantify the life of the PCB under different thermal stresses and produce the data required for life time data analysis.