Premium
PastA—the characterization of the inherent fluctuations in the fabrication process for circuit simulation
Author(s) -
Müllerl Guntram E.
Publication year - 1995
Publication title -
international journal of circuit theory and applications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.364
H-Index - 52
eISSN - 1097-007X
pISSN - 0098-9886
DOI - 10.1002/cta.4490230413
Subject(s) - monte carlo method , sensitivity (control systems) , circuit design , electronic engineering , bicmos , process (computing) , equivalent circuit , computer science , fabrication , electronic circuit simulation , electronic circuit , engineering , electrical engineering , voltage , mathematics , transistor , medicine , statistics , alternative medicine , pathology , operating system
For worst‐case analysis, Monte Carlo analysis, yield optimization and design centring, the variations and correlations of the device model parameters in the electrical circuit simulation are of fundamental significance. This paper describes a method for the complete characterization of the inherent fluctuations in the fabrication process for the simulation in IC design. the emphasis is placed on analogue simulation owing to the complex connections between circuit design and technology and the multitude of devices with correlated parameters; nevertheless, statistical requirements for digital simulation can be fully covered. This characterization works with all technologies whether CMOS, BiCMOS or bipolar. the prerequisites are accurate nominal device model parameters for the circuit simulation and information about fabrication statistics, e.g. process control monitor measurements. A new approach with connection coefficients for the calculation of the correlation coefficients is presented. Dependent parameters can be easily defined through a connection hierarchy. With the presented approach the variations and correlations of model parameters for simulation can be generated from estimates or measurements of the fabrication process. Based on sensitivity analysis, the variations and correlations of design objectives can be estimated. This enables the designer to do circuit analysis and optimization (limit parameters, worst‐case distances) as well as Monte Carlo analysis.