Theory of drift step-recovery diodes

A theory of drift step-recovery diodes as current interrupters in inductive-storage generators is elaborated. The theory includes the nonlinear dependence of the base resistance and barrier capacitance on the current passing through the diodes. Simple relationships are obtained for the diode parameters (the thickness and doping level of the base, the charge of nonequilibrium holes extracted from the base for the time T-B of high-reverse-conductivity phase, and the surface area and number m of series-connected diodes) and parameters of the loop (the capacitance and inductance of the energy storage and the initial voltage U-C0 across the capacitance) that provide the generation of a voltage pulse with a desired rise time t(B) and amplitude U-m on the load. For a given diode efficiency k, the maximal values of the overvoltage factor U-m/U-C0 and pulse sharpening coefficient T-B/t(B) are shown to depend on a factor proportional to k(omega)(1-k)E-B/E-s, where omega=0.27 (for U-m/U-C0) or -0.3 (for T-B/t(B)); E-B is the breakdown field; E-s=v(s)/mu; and v(s) and mu are, respectively, the saturated drift velocity and mobility of holes in weak fields. The maximum rate of rise of voltage obtainable with a single-diode (m=1) structure equals 0.3v(s)E(B). The characteristics of the Si and 4H-SiC diodes are compared. Numerical simulation of the recovery process substantiates the theory. (C) 2004 MAIK "Nauka/Interperiodica".