Rapid Precipitation and Mitigation of Intermittent Faults

Gary Hazard and Alpana Gangopadhyaya

Intermittent faults have no exact pattern or frequency, and can adversely affect system performance. They are the primary driver of false alarms, which constrain testability development and compromises detection reliability. Additionally, intermittent faults can create Return Materials Authorization (RMA) diagnostic uncertainty with status like No Fault Found (NFF), Cannot Duplicate (CND), and No Problem Found (NPF), etc. Many of us face the challenge of delivering products of a highly advance nature in an accelerating Time-To-Market window. Robust hardware combined with high reliability BIT diagnostics can significantly increase availability. This presentation will focus on a Highly Accelerated Life Testing (HALT) augmented BIT detection methodology that rapidly precipitates and mitigates intermittent faults. Examples of how the HALT augmented intermittent fault detection model can be used to significantly improve system reliability will be presented.