Annual Report on Modeling Advanced FDSOI for IC Design

  • Authors:
    Chenming Hu (UC/Berkeley), Sayeef Salahuddin (UC/Berkeley), Yogesh Singh Chauhan (IIT Kanpur), Juan P. Duarte (UC/Berkeley), Yen-Kai Lin (UC/Berkeley), Pragya Kushwaha (UC/Berkeley), Harshit Agarwal (UC/Berkeley), Huan-Lin Chang (UC/Berkeley)
    Publication ID:
    P090784
    Publication Type:
    Deliverable Report
    Received Date:
    1-May-2017
    Last Edit Date:
    3-May-2017
    Research:
    2671.001 (University of California/Berkeley)

Abstract

This report summarizes the latest developments in the BSIM-IMG models. There are three new model features in this report. First, the improved gate current model with back gate bias dependency is proposed. The voltage drop in the front gate oxide is altered by back gate electric field which makes gate current back bias dependent (VBG). Second, the VBG-dependency of GIDL is also included in the model because the electric field in the overlap region is changed with VBG. Since back gate voltage in FDSOI MOSFETs effectively modulates the conductance of channel, its effect must be accounted for calculation of noise power spectral density (PSD). Also, in FDSOI MOSFETs front and back gate interfaces will have different vertical fields and interface qualities. So, to accurately model channel conductance and thermal noise, different effective mobilities at two interfaces should be considered. Thus, the thermal noise model is improved. All the three enhanced models are implemented into BSIM-IMG and are validated with experimental data.

4819 Emperor Blvd, Suite 300 Durham, NC 27703 Voice: (919) 941-9400 Fax: (919) 941-9450