Integrated circuit fault prognostics based on the delay characteristics of the clock network | IEEE Conference Publication | IEEE Xplore

Integrated circuit fault prognostics based on the delay characteristics of the clock network


Abstract:

This study proposes an integrated circuit fault prediction technique, which is based on delay characteristics of the clock network. A typical digital integrated single ro...Show More

Abstract:

This study proposes an integrated circuit fault prediction technique, which is based on delay characteristics of the clock network. A typical digital integrated single road was used as the example to study the responses of the clock network delay signals to the integrated circuit failures, and analyzed the feasibility and effectiveness of the method in detail. The experimental results show that the fault diagnosis rate of the method was as high as 92%; after spectrum analysis, the delay signals were converted to the amplitudes and phases, the relation between the phases and the degradation rates was a single exponential function, and the phases of the device was increased, the degradation rate was reduced, with decreasing the sweep frequency. The phases of the range of the degradation rates is from 50% to 80%, can be defined the fault diagnosis threshold or warning interval of the device. At lower frequency, the gradient of the phase was more slowly, so the error of prognostics is smaller at low frequency, the accuracy of prognostics is higher.
Date of Conference: 15-18 November 2016
Date Added to IEEE Xplore: 16 January 2017
ISBN Information:
Conference Location: Paris, France

I. Introduction

In order to ensure the reliable and safe operation of the electronic equipment in the special working environments, it is necessary to monitor working status and reliability of equipment in real time. Prognostics and health management (PHM) technique is collecting all kinds information of the system with sensors, using different algorithm to estimate conditions of the system, real-time monitoring and predicting the failure or malfunction trend, and to combine with various resources to provide maintenance support measures. Recently, PHM is receiving more and more attention. And the health statuses of key components, modules and equipment of the underlying system directly affect the performance of the whole system, therefore, diagnosing and prognosticating failures of the key modules of the underlying system, especially the large scale integrated circuit that is the core circuit function, is the crux of implement that based on the state of equipment maintenance technology.

References

References is not available for this document.