Predictable yield-driven circuit optimization

Abstract

The authors present a comprehensive approach to predictable yield optimization. They utilize a new physics-based statistical GaAs MESFET model which combines the advantages of the DC Khatibzadeh and Trew model and the small-signal Ladbrooke formulas. The yield of a broadband amplifier was significantly improved after optimization. Predicted yield over a range of specifications was verified by device data. The benefits of simultaneous circuit-device yield optimization assisted by yield sensitivity analysis are demonstrated.<>