On the determination of the effective channel length of MOSFETS

N. Patel , M. Garcia , S. Titcomb , R. Anderson

The conventional method used to determine Δ L, the processing induced channel length shortening in MOSFETs at room temperature gives invalid results at 77 K. This error is found to result from the incorrect assumption that the channel resistance is a linear function of effective channel length for the values of gate overdrive normally used for the extraction of Δ L. By restricting gate voltages to the range where this assumption is shown experimentally to be valid, values of Δ L extracted at 77 K are smaller than those extracted at room temperature. This attributed to the reduced source-channel and drain-channel depletion region widths at 77 K at and above threshold.

Domaines: Articles anciens

N. Patel, M. Garcia, S. Titcomb, R. Anderson. On the determination of the effective channel length of MOSFETS. Journal de Physique IV Proceedings, 1994, 04 (C6), pp.C6-25-C6-29. ⟨10.1051/jp4:1994604⟩. ⟨jpa-00253098⟩