AbstractThis paper presents a case study that was initaated by excessive monitor failures occurring during a simulated early‐life test. Statistical analysis of the failure data suggested that there should be no longterm resistor failure problem. Large increases in resistance of some metal film resistors led to many of the monitor failures. Corrosion of the resistive film by residual chlorine for a particular resistor vendor's cleaning process was responsible. The results led to process changes to reduce the contamination and the reliability testing of the ‘new’ product revealed that the process changes were successful. A statistical designed experiment indicated that the ‘old’ resistors were not all degraded in the same manner. As a result of this experiment, the failures were attributed to poor process
展开▼