Abstract: The insulated gate bipolar transistor (IGBT) has negative Miller capacitance during switching transients. It has conventionally been attributed to the voltage dependency of the Miller capacitance. However this explanation has physical ambiguity, yet, it lacks a discussion of the conditions for the occurrence of negative Miller capacitance as well. We argue that it is the current dependence to the Miller capacitance that results in the negative case. In this paper, we provide a modification to the theoretical analysis of this phenomenon. The occurrence condition for it and the device parameters about it are discussed. It is discovered that the negative Miller capacitance must occur during the turn-off process for any IGBT, while it is relatively difficult during the turn-on process. At the device design level, the current gain of the PNP transistor in the IGBT is an important factor for the negative Miller capacitance.

Key words: IGBT, negative Miller capacitance, theoretical analysis