Abstract: The commutation circuit of an inverter is actuated only when the main thyristors cannot be transitioned to the nonconducting state by the removal of the gate signal. If a commutation pulse is required, the commutation capacitors are charged to the appropriate voltage level immediately prior to the switching point so that they may provide a sufficient pulse of current to extinguish the current through the thyristor. The commutation capacitors are then returned to a stable voltage level until another commutation pulse is required.

Abstract: A supplemental or auxiliary portion of the commutation circuit for an inverter is actuated only during overload conditions which improve the inverter efficiency during normal load conditions. A control circuit responds to the increased time period of the commutation pulse to delay the firing of the thyristors which initiate the makeup pulse. This modified operation continues for at least three commutation cycles in order to ensure that the supplemental portion of the commutation circuit is properly initialized so that it is ready for subsequent overcurrent conditions.

Abstract: A two-stage commutation circuit for an inverter, each having a different energy storage capability, commutates the main semiconductor switches by creating a commutation pulse suitable for the different levels of load current. This improves overall inverter efficiency by decreasing the circulating commutation current at light loads due to reduced size of the commutation and makeup pulses utilized at the lower load current levels. The control circuit for the dual commutation circuits includes a voltage detector for sensing the level of the DC input voltage and current detectors for sensing the times at which the load current is essentially zero so that the transistion between the commutation circuits will occur without perturbations in the cyclic operation of the inverter.