The screening and detection of failures in high reliability electronics is crucial to mission success. The governing standards for microelectronic components prescribe effective screens to address known failure modes during the manufacture of the components. When failures are detected outside of the normal component screening process such as in the next higher assemblies, the analysis of such failures to root cause is even more crucial. Effective root cause analysis requires collaboration among personnel having expertise in many areas, both technical and organizational. Too often, the analysis stops at the ‘how’ a part failed, rather than the ‘why’ and true root cause determination. Failures may not only represent defects in the technical aspects of the design, materials or application of the hardware, but may also be related to human organizational factors such as the competing interests of time, quality and cost. The analysis must extend beyond the forensic data into the human organization and the interaction between people, process and hardware.
In this paper, the author illustrates an effective root cause corrective action (RCCA) analysis process through the discussion of a microelectronics packaging case study. The case involved failure of high reliability hybrid microelectronic components at the next higher assembly level. The case study provides insights into the root cause analysis process as well as awareness that component level high reliability screening is not impervious. It is critical for the next level user to have the expertise in their organization to understand both the how and the why of such failures.
The case study describes a failure related to loose metal particles inside a hermetically sealed multilayer ceramic package which had passed Particle Impact Noise Detection (PIND) screening at the supplier. Root cause analysis determined the cause for the screening escape as well as the cause of the metal particles, and effective corrective actions were implemented at both the component supplier and Honeywell. Technical causes include improper package design, weld schedule and screening methods. Organizational causes include inadequate documentation, lack of training and failure to react to prior failures. The root cause investigation team was a multi-discipline group from Honeywell, our customer, and the component supplier, with additional consulting input from other suppliers, equipment manufacturers and industry experts.
A Case Study of Fishbone Sequential Diagram Application and ADREP Taxonomy Codification in Conventional ATM Incident Investigation