Proceedings of the Mechanical Failures Prevention Technology (MFPT) Conference, 2012

Canary Approach For Monitoring BGA Interconnect Reliability Under Temperature Cycling

Preeti Chauhan, Michael Osterman and Michael Pecht

CALCE Electronic Products and Systems, University of Maryland, College Park, MD 20742, USA


Solder interconnect reliability is a major concern for electronic products since these are often the most critical elements in determining the reliability of electronics. Thermal fatigue, caused by temperature swings and mismatches between the coefficients of thermal expansion (CTEs) of the mounted semiconductor packages and the application board, is one of the leading causes of failure in solder interconnects. While simulation
models exist for predicting the thermal fatigue failure of solder interconnects, reliability prediction is challenging due to the wide variation in usage profiles. One of the approaches to address this challenge is the canary approach. The canary approach consists of creating a component or circuit incorporated into or in the vicinity of the target component, where a detectable event is driven by the same or similar mechanism
that precedes a functional failure of the target component. Thus, a canary provides monitoring and prognostics for failure of the target component.

This paper presents a physics of failure (PoF)-based canary approach for reliability prediction of ball grid arrays (BGAs) under temperature cycling. The canary is formed by a BGA resistance net consisting of outer solder interconnects and is used to predict failure in the resistance net formed by inner solder interconnects. Thus, an existing circuit in the BGA is utilized as a canary. The approach is beneficial because there is no
additional circuit needed, which provides real estate benefits along with the reliability prediction capability.

Keywords: Ball grid array (BGA); canary; diagnostics; prognostics; health management; solder interconnect reliability

Complete article is available from the publisher and to the CALCE consortium members.

[Home Page] [Articles Page]
Copyright © 2012 by CALCE and the University of Maryland, All Rights Reserved