A DIGITAL TECHNIQUE FOR DIAGNOSING INTERCONNECT DEGRADATION BY USING DIGITAL SIGNAL CHARACTERISTICS

Abstract

Interconnects are responsible for mechanical and electrical connection of electronic components, and they are essential in the operation of electronic components. Stress-induced substances can degrade interconnect properties and functions because they exert stress conditions such as chemical stress to interconnects. Chemical stress from moisture or contaminants causes corrosion to interconnects, and as a result interconnect failures occur in which the interconnects cannot conduct the intended functions. Since interconnect failure disturbs the connection between electronic components, it can cause ensuing failures such as electronic failures. Many approaches have been developed to detect interconnect degradation and prevent ensuing failures. In industrial fields, approaches to detect interconnect failure based on DC resistance have been widely used since DC resistance can capture electrical discontinuity. However, approaches based on DC resistance have a problem because they usually require additional sensing devices or circuitries. This study introduces a new approach to interconnect failure detection using digital signals. The proposed method using digital signals detects interconnect failures without additional sensing devices. Interconnect failure detection can be conducted by monitoring and analyzing the signal characteristics of the transmitted signal because digital signal is continuously generated and transmitted in electronics to control the electronic components and communicate between the components, and also because digital signal is in high-speed, the characteristics of the transmitted digital signal is deteriorated by physical damages on transmitted circuitries such as corrosion on interconnects. We designed accelerated life tests (ALT) of interconnects under chemical stress in order to demonstrate failure detection capability. While solder joints were exposed to chemical stress and corroded gradually, the digital signal characteristics were monitored with DC resistance simultaneously and analyzed by Sequential Probability Ratio Test (SPRT) to determine the times-to-failure of the solder joints. The test results demonstrated that the proposed approach based on digital signal can detect interconnect failures earlier than the DC resistance, which means that the proposed approach can enable electronic components to detect interconnect failures by themselves.