A New Optimization Based Frequency Piecewise Small Signal Model for GaN HEMT

In this paper, a new optimization based frequency piecewise small-signal model for Gallium nitride (GaN) high electron mobility transistor (HEMT) is proposed. Different from the classic modeling technique, which use a single set of model parameters to describe the behavioral of the device across the whole broad frequency band, the new proposed model uses piecewise modeling method. Due to the inner nonlinearity of GaN HEMT at high frequency, the behavior of the device across the band is decomposed into few different pieces, based on the nonlinearity of the device. The model is extracted and optimized with commercial tool at each different piece of frequency band. The verification is carried out on an 8 × 125 μm GaN HEMT with a 0.25 μm gate feature size. The proposed model shows a big improvement in accuracy of S-parameter behavior prediction. The relative error of S
₁₁
and S
₂₂
is greatly reduced and can be controlled under 2%.