Electrical Breakdown of Power Rectifiers for Electric Gun Applications

Experiments have been performed with a 100-kJ pulse-forming network (PFN) to characterize the transient behavior of semiconductor diodes serving as capacitor-protecting devices. In addition to the experiments, computer techniques are used to illustrate and predict the dynamic behavior ofPFN diodes. From analyses of the data collected, it was determined that in a PFN under specific loading conditions, the diodes are subject to elevated transient high frequency voltage waveforms. In some experiments, the magnitude of the rate of change in voltage (dV/dt) across the devices was such that catastrophic failure was observed due to this phenomenon alone. This paper focuses on the determination of boundary conditions necessary for reliable device performance and the solutions that will circumvent diode operational failures. Our solutions include semiconductor device layout, choice of diode reverse recovery time, PFN switch timing, and selection of capacitive and inductive circuit parameters, all of which are presented in detail in this study. Information describing fundamental physics of semiconductor diodes under transient conditions and their role in electric gun propulsion technology is provided first to clarify the applicability of these areas of concern to the general field of very high power electronics. INTRODUCTION Power Diode Usage in Pulsed Power Systems Diodes are successfully used as pulse power components in several facilities currently involved in electric gun research including the U.S. Army Research Laboratory (ARL)l-3. Since they shunt the capacitors during a voltage reversal, semiconductor diodes are referred to as "crowbar" diodes; i.e., they clamp the capacitor for the duration of the voltage reversal. Diodes will effectively extend the lifetime of the capacitors, which have been shown to fail under voltage reversal situations4. Diode failure occurs as a result of excessive junction temperature due to heating which leads to increased device conductivity; this results in additional current flow and further temperature increases. The conductivity is directly proportional to the concentration of mobile current carriers in the device, and is itself a function of operating temperature. The relationships between conductivity (cr) and mobile carrier concentrations for electrons (n) and holes (p) and carrier concentration and temperature are given, respectively, in Equations 1 through 3. Here q represents the charge of an electron, and J..L n and J..Lp are the electron and hole mobilities: 0' = qJ..Lnn + qJ..LpP. (1) ForT > 30 K, the semiconductor becomes intrinsic with n=p=ni and: cr = q(J..Ln + qJ..Lp)ni, (2) and ni expressed as a function of temperature is given as: ni (T) = 3.88 x to16 T3/2 exp (-7000/T) cm·3 (for silicon above 30 K). (3) lf a semiconductor is not allowed to dissipate heat quickly enough, this process will continue until the melting point of the device is reached and it is ultimately destroyed. This process is often referred to as "thermal runaway." Diode Reverse Recovery Time ('trrl The reverse recovery time ('trr) is that time required for the diode to switch from an "on" or conducting state to an "off" or high impedance state. The 'trr consists of the time associated with the minority carrier removal at the depletion edge, which is referred to as the "storage" time ('ts), and a "fall" time component ('tf) due to the junction depletion capacitance (Cj). This is expressed in Equation 4. The storage time ('ts) depends upon the effective carrier lifetime ( 'teff) which is approximately the minority carrier lifetime ( 'tp) of the semiconductor device (for long base diodes where the neutral base width, W n >the diffusion length, L), or the transit time ('tt)