An image sensor module failed in the field and was returned showing functional issues and a supply-to-ground short. After the hard lens mounted over the imaging chip was removed, the short disappeared along with the functional issues. This paper explains how the authors were able to restore the failure mode and discover the underlying defect, via backside focused ion beam cross-sectioning, with minimal intrusion into the top-side package and silicon.

This paper explains how the authors determined the cause of a fast-to-rise failure discovered during scan chain testing of an image sensor. The failed device was mounted on a portable card that facilitates transfer between test platforms in an electro-optical probing (EOP) system. Initial fault localization was conducted through backside PEM, but the results were inconclusive. The part was then analyzed on a digital scan chain tester to check for flaws in the daisy chain of shift registers. Through broken scan chain analysis, the potential cause of the problem (a failing flip-flop) was narrowed down to a few chain links and ultimately pinpointed using EOP fault isolation techniques. The failed device was then deprocessed by parallel lapping and analyzed in a SEM, revealing a broken poly gate as the physical cause of failure.

The failure analysis community working on highly integrated mixed signal circuitry is entering an era where simultaneously System-On-Chip technologies, denser metallization schemes, on-chip dissipation techniques and intelligent packages are being introduced. These innovations bring a great deal of defect accessibility challenges to the failure analyst. To contend in this era while aiming for higher efficiency and effectiveness, the failure analysis environment must undergo a disruptive evolution. The success or failure of an analysis will be determined by the careful selection of tools, data and techniques in the applied analysis flow. A comprehensive approach is required where hardware, software, data analysis, traditional FA techniques and expertise are complementary combined [1]. This document demonstrates this through the incorporation of advanced scan diagnosis methods in the overall analysis flow for digital functionality failures and supporting the enhanced failure analysis methodology. For the testing and diagnosis of the presented cases, compact but powerful scan test FA Lab hardware with its diagnosis software was used [2]. It can therefore easily be combined with the traditional FA techniques to provide stimulus for dynamic fault localizations [3]. The system combines scan chain information, failure data and layout information into one viewing environment which provides real analysis power for the failure analyst. Comprehensive data analysis is performed to identify failing cells/nets, provide a better overview of the failure and the interactions to isolate the fault further to a smaller area, or to analyze subtle behavior patterns to find and rationalize possible faults that are otherwise not detected. Three sample cases will be discussed in this document to demonstrate specific strengths and advantages of this enhanced FA methodology.