Development of statistical model for baseband MOSFETs

• Main Author: Lim, Guan Hui
• Other Authors: Zhou Xing
• Format: Theses and Dissertations
• Language: English
• Published: 2010
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Semiconductors
• Online Access: https://hdl.handle.net/10356/41399
• Institution: Nanyang Technological University

In this work, backward propagation of variance (BPV) methodology is used to develop a statistical MOSFET model for 0.35um technology to account for global process variations based on electrical data from measurements of MOSFET test structures in scribe line module (SLM) of production wafers. BPV, coupled with Pelgrom model for MOSFET mismatch, is also used to develop a statistical MOSFET mismatch model for 0.13um technology to account for local process variations such as random dopant fluctuation (RDF) in channel and line edgelline width roughness (LERILWR) of gate based on MOSFET test structures that were specially designed and fabricated on multi-project wafer (MPW) for this study. The statistical MOSFET models developed in this work enable the designer to perform Monte Carlo circuit simulation using circuit simulator such as HPSICE so that better and accurate assessment can be made on the potential circuit performance and functional yield of the circuit at the design stage. MOSFET mismatch test structures were also specially designed, fabricated and characterised for 0.1 8um, 0.1 3um, and 90nm technologies to study the impact of backend-of-line (BEOL) metals, multi-fingered MOSFET layout design and gate protection diode against plasma charging damage (PCD) on MOSFET mismatch characteristics.