ABSTRACT
A mathematical description of the harmonics of the carrier-based methods is difficult because the intersection moments are not linearly or equally spread over the reference cycle. Carrier-based pulse width modulation (PWM) algorithms have emerged in analog hardware support. Understanding the principles of carrier-based PWM generation with discontinuous reference functions later helped define reduced loss space vector PWM methods. Digital implementation of the uniform-sampled PWM method requires software preparation of the overmodulation operation based on calculation of the function to be added to the reference. Even if overmodulation can be carried out with one of these methods, the performance is strongly affected during overmodulation operation and it is recommended only for dynamic or transient operation. Inverter operation is unfortunately affected by minimum pulse width, deadtime selection, and maximum available voltage. Carrier-based modulation provides a linear characteristic up to the modulation index of 0.785.
