DocumentCode
382913
Title
Improved IDDQ testing with empirical linear prediction
Author
Bergman, David I. ; Engler, Hans
Author_Institution
Nat. Inst. of Stand. & Technol., Gaithersburg, MD, USA
fYear
2002
fDate
2002
Firstpage
954
Lastpage
963
Abstract
A new linear prediction method that improves IDDQ test effectiveness is described. The method uses statistical pre-processing of exhaustive measurements on training devices to extract principal patterns in the device IDDQ behavior and to generate a prediction model. Fitting the model to device measurements accommodates variations in the fabrication process. Comparison with the Delta IDDQ test method using the SEMATECH S-121 data shows that for nearly equal numbers of defective parts passed, the new method fails fewer defect-free parts.
Keywords
CMOS digital integrated circuits; MOSFET; VLSI; curve fitting; electric current; fault location; integrated circuit modelling; integrated circuit reliability; integrated circuit testing; statistical analysis; Delta IDDQ test method; IDDQ testing; SEMATECH S-121 data; defect-free parts; defective parts; digital CMOS integrated circuits; empirical linear prediction method; fabrication process variations; prediction model fitting; principal device IDDQ behavior patterns; statistical pre-processing; test effectiveness; training device measurements; Circuit testing; Current supplies; Data mining; Electrical fault detection; Fabrication; Integrated circuit testing; Leakage current; MOSFET circuits; Prediction methods; Predictive models;
fLanguage
English
Publisher
ieee
Conference_Titel
Test Conference, 2002. Proceedings. International
ISSN
1089-3539
Print_ISBN
0-7803-7542-4
Type
conf
DOI
10.1109/TEST.2002.1041851
Filename
1041851
Link To Document