Share Email Print

Proceedings Paper

Microcontroller signal density stress prediction framework
Author(s): Sheng-Jen Hsieh; Sung-Ling Huang; Shao-Chin Chang
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Studies of electronic component stress have typically focused on temperature, humidity, and voltage stress. There has been relatively little emphasis on clock signal frequency stress. This study investigated effects of clock frequency stress on a high performance microcontroller used to control the on-off frequency of LEDs on a printed circuit board. Based on the design specification, several frequency levels between 0 and the terminal clock frequency were selected. Thermal profile samples for each stress level were collected using an infrared camera. The data were then divided into two groups for model development and evaluation. Artificial neural network and statistical regression approaches were used to model thermal profiles for each stress level. Objectives were to (1) explore impact of clock frequency stress on IC functionality, (2) observe heating rate differences under clock frequency stress over time; and (3) predict stress levels using the two approaches. Results indicate that the average prediction error is about 7.9% for the neural network approach and about 23.8% for the statistical regression approach. Future directions include thermal profile modeling using Finite Element Analysis (FEA) and development of robust hybrid analytic and experimental models for microcontroller lifetime prediction.

Paper Details

Date Published: 12 April 2004
PDF: 11 pages
Proc. SPIE 5405, Thermosense XXVI, (12 April 2004); doi: 10.1117/12.542439
Show Author Affiliations
Sheng-Jen Hsieh, Texas A&M Univ. (United States)
Sung-Ling Huang, Texas A&M Univ. (United States)
Shao-Chin Chang, Texas A&M Univ. (United States)

Published in SPIE Proceedings Vol. 5405:
Thermosense XXVI
Douglas D. Burleigh; K. Elliott Cramer; G. Raymond Peacock, Editor(s)

© SPIE. Terms of Use
Back to Top