A major obstacle in real-time simulation (RTS) of power electronic converters is the inaccurate sampling of pulse width modulation (PWM) signals, which leads to undesired subharmonic oscillations. Conventional solutions rely on oversampling, requiring additional time-step computations and therefore hardware capable of resolving them within very short time intervals, such as field-programmable gate arrays (FPGAs). This work introduces a method that eliminates aliasing by separating and independently solving the average and switching dynamics. The resulting model was tested on a Buck converter, a basic topology commonly examined in studies of subharmonic distortion. The proposed RTS is computationally lightweight and provides good performance even at relatively slow execution frequencies, enabling implementation on a low-cost microcontroller. The method was validated in three stages. First, it was benchmarked against a real Buck converter and a RTS model without aliasing mitigation. Although nearly four times slower than the baseline, the proposed scheme achieved an order-of-magnitude improvement in accuracy by effectively mitigating subharmonic oscillations. Second, it was compared with a state-of-the-art oversampling technique using a Simscape simulation reference. The proposed method produced comparable error while the oversampling approach required 300 times faster execution. Finally, both approaches were deployed on an FPGA, experimentally confirming the simulation results.
