When the converter level count gets higher, the space-vector pulse-width modulation (PWM) will suffer from the calculation burden for synthesising the voltage vectors and the redundant vector selection, which can be overcome by using the carrier-based PWM, as long as the carrier signal and modulation signal can be properly selected. This study proposes a new carrier-based PWM method to achieve some important control objectives in multilevel converters by adding the optimised zero-sequence signal to the reference voltage. The control objective as well as the control method for balancing the DC-link neutral point (NP) potential is presented. Meanwhile, the relationship between the NP potential and zero-sequence voltage is comprehensively analysed and an algorithm for injecting the optimised zero-sequence signal is derived. This study also investigates how to use the zero-sequence voltage to achieve the ‘two-phase’ mode operation for reducing the converter switching losses. A common mode (CM) voltage mitigation method is then proposed which effectively mitigates the CM voltage within 1/6 of the DC-link voltage. Furthermore, this study gives a simple method to directly map the reference voltage to the converter switches by using the concept of voltage level. The experimental results with a scaled 1 kW system validate the proposed NP potential control method and CM voltage mitigation method.