Compact Models for the I-V Characteristics of Double Gate and Surround Gate MOSFETs

Abstract

The models presented by Lu and Taur, [1], for lightly doped double gate and surround gate MOSFETs each require numerical solution of a transcendental equation. In this paper we present compact solutions for the equations based on the Lambert function, [2]. These solutions are shown to be accurate compared with exact numerical solutions.

An Improvement to the Numerical Robustness of the Surface Potential Approximation for Double-Gate MOSFETs

“…The approximations described in [1] and [15], as well as our proposed enhancement described in the next section, were implemented in Scilab and investigated for accuracy, numerical robustness, and computation expense in terms of CPU time. The investigation was carried out on an AMD Opteron 2.4-GHz Linux-based PC featuring 16-GB RAM.…”

“…it is the lowest bias point at which the approximations discussed in both [15] and [1] work. Table I summarizes the results (the accuracy is shown in Fig.…”

“…To implement the model in a circuit simulator, it is desirable to have closed-form expressions. In [1] and [15], a closedform approximation procedure for β 1 was presented. However, these procedures have some drawbacks, which we discuss in the next section, after which we suggest an enhancement to [1] to overcome the drawbacks.…”

Symmetric and Asymmetric Double Gate MOSFET Modeling

“…The surface potential, channel charge and current models that were obtained in [2][3][4]12] require the solution of a transcendental equation (see equation (22) in Appendix A). In [5,6], this transcendental equation is solved iteratively with very fast convergence for circuit simulation application.…”

“…Most of the published compact modeling work is focused on symmetric and undoped DG-MOSFET devices (see [2][3][4][5][6][10][11][12][13]), whereas real devices are lightly doped, about 10 15 cm -3 , and asymmetric. In the numerical simulation of [14] it is shown that even a small doping density of 10 15 cm -3 could cause a large shift in the surface potential.…”