A prognostic method of assessing solder joint reliability based on digital signal characterization

Abstract

Electronics are subjected to thermal, mechanical or chemical stress conditions during their operations. These stress conditions accumulate damages to the electronic system, adversely affecting its components and interconnects including solder joints. Solder joint degradation can cause system malfunctioning, which eventually results in catastrophic failures. In order to prevent system failures, it is required to manage the system health continuously by monitoring damages such as solder joint degradation. Regardless of stress conditions, solder joints often start to degrade from their surface where high speed signals in an electric circuit are concentrated due to the phenomenon referred to as the skin effect. Thus, high speed signals such as digital signals can serve as a sensitive means of detecting solder joint degradation at early stages. This paper presents a diagnosing method of assessing solder joint reliability based on digital signal characterization. Accelerated life tests are conducted to generate solder joint failure. The test setup consists of a circuit board with solder joints, a digital signal transceiver, an environmental test chamber and a stress application fixture. Constant mechanical shear stress is applied to the solder joints of a circuit board at an elevated temperature. A digital signal transceiver generates high speed signals travelling through the solder joints, and continuously monitors the signal characteristics which indicate signal integrity. The test results confirm that the signal characteristics show statistically significant variations between intact and cracked state of the solder joint, and the parameter variations indicate the deterioration of the signal integrity. These results suggest that digital communication signals can be used as a non-destructive diagnostic tool of physical degradation. It can be improved as a tool which provides early warning of impending system failures without installing additional fault sensing instruments. This diagnostic approach may serve as a proactive prognostic module that allows for real-time health management of electronic products and systems.