Victor Paulraj, Eric Paulraj
(2016)
Hybrid Diagnosis Model To Determine Fault Isolation For Scan Chain Failure Analysis On 22nm Fabrication Process.
Masters thesis, Universiti Sains Malaysia.
Abstract
With the rapid growth of Very Large Scale Integration (VLSI) in complex designs, there is high demand for Design for Testability (DFT). Vast study has proven that Scan based testing is achieving good test coverage with lower cost and smaller die area and is widely used in the industry. Scan chain fault diagnosis plays an important role as with the implementation of Scan based testing, it is reported that 10%-30% of defects in a Scan based design occurs within the Scan chain itself. Currently, there are three main types of stand-alone diagnosis models available, which are: software-based diagnosis, tester-based diagnosis and hardware-based diagnosis, where each has its disadvantages and limitations. In this project, the author proposed a hybrid Scan chain failure analysis technique that uses the proposed software-based diagnosis to obtain a list of possible failing suspect Scan cells, followed by the proposed tester-based diagnosis to further isolate the fault to a single failing device suspect. This proposed hybrid diagnosis algorithm ensures that Scan chain faults such as stuck-at and transition faults can be root-caused with lesser time and low complexity for both solid and marginal failures. Four case studies were successfully carried out to evaluate the proposed hybrid diagnosis algorithm on a 22nm fabrication process technology Device under Test (DUT) System-on-Chip (SOC) product, where the fault isolation was able to isolate a single failing device suspect for all four case studies, indicating a 100% fault isolation success rate.
Actions (login required)
 |
View Item |
![[img]](http://eprints.usm.my/style/images/fileicons/application_pdf.png)
