Date of Award

3-10-2010

Document Type

Thesis

Degree Name

Master of Science in Computer Engineering

Department

Department of Electrical and Computer Engineering

First Advisor

Yong C. Kim, PhD

Abstract

Errors within circuits caused by radiation continue to be an important concern to developers. A new methodology of real time fault detection and diagnostics utilizing FPGA based architectures while under radiation were investigated in this research. The contributions of this research are focused on three areas; a full test platform to evaluate a circuit while under irradiation, an algorithm to detect and diagnose fault locations within a circuit, and finally to characterize Triple Design Triple Modular Redundancy (TDTMR), a new form of TMR. Five different test setups, injected fault test, gamma radiation test, thermal radiation test, optical laser test, and optical flash test, were used to assess the effectiveness of these three research goals. The testing platform was constructed with two FPGA boards, the Device Under Test (DUT) and the controller board, to generate and evaluate specific vector sets sent to the DUT. The testing platform combines a myriad of testing and measuring equipment and work hours onto one small reprogrammable and reusable FPGA. This device was able to be used in multiple test setups. The controlling logic can be interchanged to test multiple circuit designs under various forms of radiation. The detection and diagnostic algorithm was designed to determine fault locations in real time. The algorithm used for diagnosing the fault location uses inverse deductive elimination. By using test generation tools, fault lists were developed. The fault lists were used to narrow \ the possible fault locations within the circuit. The algorithm is able to detect single stuck at faults based on these lists. The algorithm can also detect multiple output errors but not able to diagnose multiple stuck at faults in real time.

AFIT Designator

AFIT-GCE-ENG-10-04

DTIC Accession Number

ADA517271

Share

COinS