In circuit simulations, calculating the turn-off switching loss in a power diode without modelling the reverse recovery current which causes the loss helps maintain good simulation speed. Ideal switches can be used and losses can be calculated or inferred from a lookup table after the switching transition. Difficulties arise with these approaches when circuit non-idealities cause the diode to commutate at zero current. The forward current before the initiation of turn-off must be known in order to determine the switching losses. This paper presents an analytical method for approximating the turn-off loss for an ideal diode model. Based on a regression to estimate the reverse recovery charge at the operating point, this method works for both ideal and non-ideal switching transitions. A low-pass filter is used to provide a finite value for the diode forward current in the event of a zero-current transition and the reverse recovery charge regression is modified to compensate for the dynamics of the low-pass filter. The accuracy of turn-off losses calculated using this analytical approach is verified by comparing the results against experimental measurements.
展开▼
