Characteristics of a Current-Fed Inverter with Commutation Applied Through Load Neutral Point

Abstract

A current-fed inverter which is commutated by an auxiliary commutating circuit acting through the load neutral point is investigated. The auxiliary commutation circuit consists of a single commutating capacitor and two auxiliary thyristors which are used to sequentially commutate the main thyristors. The properties of the inverter are considered for the case of an ac machine load, although the results are general and can be easily applied to cases where the inverter is interconnected with an ac power system. A ``voltage behind subtransient reactance'' model results in a good prediction of the inverter operating modes for either a synchronous or an induction machine. It is shown that if conventional simple-sequential gating is used, the inverter can only operate in the motoring region, while a commutation failure will result in the generating region. This negative feature is overcome by introducing a ``delayed gating'' scheme for the thyristor, which allows the peak commutating capacitor voltage to be actively controlled; thus, commutation for various inverter conditions is maintained. A scheme of parallel commutation circuits is introduced which allows the value of commutating capacitance to be actively selected as a function of load condition. While the main thyristors are subjected to approximately 50 percent more applied voltage compared to the more conventional auto-sequential type of inverter in present use, the thyristors may be rectifier grade since a substantial di/dt inductance is naturally provided by the motor leakage inductance: a fact which means that negligible snubber components are needed compared to the auto-sequential circuit.