DESIGN AND EVALUATION OF FILTERS FOR REDUCING OVER VOLTAGES ON ASYNCHRONOUS MACHINES FED BY PWM INVERTER

Abstract

It is not quite uncommon, in general, to use DC motors in industrial applications, but after the introduction of power electronics, AC motors grabbed much importance. Of these, induction motors are widely used in industrial and domestic applications due to the advantages like simple construction, ruggedness, flexible control, good efficiency and low cost. In order to control the stator voltage and speed, PWM inverters are used. But the usage of PWM inverters has the effects like non-uniform voltage distribution and increase in the temperature. Recent advances in power electronic switching device technology have resulted in dramatic improvements and cost reduction of pulse-width modulated AC adjustable speed drives. Concomitant with the better performance enabled by the high switching speed and increased switching frequency they have also raised several concerns related to the consequences of high speed switching. One of these concerns is the over-voltage that appears at the motor terminals due to the impedance mismatch between the power cable and the motor. In this paper a developed accurate simulation models for power cables and motors that allow a better understanding of the over-voltage problems. The models can be readily implemented using computational tools like MATLAB, thereby providing a convenient method to develop the best dv/dt filter solution for a particular drive. The power cable is modeled using several lumped-parameter segments of a lossy representation of transmission line. An algebraic analysis is developed to choose an adequate number of lumped-parameter segments. The number of required segments is a function of the pulse rise time, cable characteristic parameters and cable length. Simulation results are presented analyzing the over-voltage problem for a wide range of cable lengths for 3 hp, 15 hp and 40 hp motors. The most important filter networks are also designed and the solutions are investigated using the simulation programs.