Hierarchical fault diagnosis of analog integrated circuits

C K Ho, P R Shepherd, F Eberhardt, W Tenten

Research output: Contribution to journalArticlepeer-review

4 Citations (SciVal)

Abstract

This paper introduces a hierarchical-fault-diagnosis algorithm as an aid to testing analog and mixed signal circuits. The diagnosis approach is based on that introduced by Wey and others and makes use of the self-test algorithm, and the component-connection model. The main extension to these techniques is the use of a hierarchical approach whereby blocks of circuitry are grouped together leading to a reduction in matrix size, so making even large scale circuits diagnosable. Other improvements from this approach include a novel test-point selection procedure and the fact that hard faults can also be diagnosed, provided they lie completely within a hierarchical block. The overall algorithm is described and the results from example circuits show good functionality of the diagnosis algorithm. Fault masking and sensitivity to the simulation/measurement resolution of test point values are examined and are highlighted as future activities to further improve the approach
Original languageEnglish
Pages (from-to)921-929
Number of pages9
JournalIEEE Transactions on Circuits and Systems. Part I: Fundamental Theory and Applications
Volume48
Issue number8
DOIs
Publication statusPublished - Aug 2001

Keywords

  • analogue integrated circuits
  • simulation/measurement resolution
  • built-in self test
  • functionality
  • mixed analogue-digital integrated circuits
  • analog integrated circuits
  • self-test algorithm
  • matrix size
  • fault diagnosis
  • hierarchical fault diagnosis
  • hard faults
  • integrated circuit testing
  • mixed signal circuits
  • large scale circuits
  • diagnosis algorithm
  • test-point selection procedure
  • circuit simulation
  • component-connection model
  • test point values

Fingerprint

Dive into the research topics of 'Hierarchical fault diagnosis of analog integrated circuits'. Together they form a unique fingerprint.

Cite this