Reducing the semiconductor losses in a matrix converter

The authors look at the implementation of a matrix converter switching strategy under microprocessor control and using IGBTs. The effects of different switching methodologies for the converter are investigated. The power losses in the switching devices of a matrix converter are analysed and quantified. The effects of different switching frequencies and approaches to PWM generation are considered. Comparisons are drawn with a conventional inverter in terms of efficiency and relative cost. The conduction losses in the matrix converter are higher than the conduction losses in a standard inverter. Methods of reducing the switching losses in a matrix converter are proposed. The implementation of these methods means that the total losses in the matrix converter can be less than those in an inverter drive at high switching frequencies. It is shown that even at lower switching frequencies the losses in the matrix converter are not significantly greater than those in an inverter drive