Effects of Random Modulation on the reduction of EMI emissions of a DC-DC Converter

Nowadays, the energy sector is giving much attention to renewable energy sources since they are clean and inexhaustible. However due to their intermittency, they require smart power electronics converters for their sustainable operations. Therefore, power electronics engineers are focusing on finding power converters that have high efficiency, less size, and low cost. The latest advancement in power semiconductors, especially wideband gap semiconductors (GaN and SiC), has dominated over a conventional Si IGBTs and MOSFETs. SiC MOSFETs are getting much application due to their higher voltage rating, higher thermal operation, and higher switching frequency operation. However, the increasing switching frequency (up to 150 kHz), and high dv/dt and di/dt nature of SiC MOSFETs cause a potential trait from the EMC point of view.

Electromagnetic Interference (EMI) filters are widely used for filtering out the EMI noise generated by power converters. However, when size is a great concern like aircraft and automobile applications, there should be another option to minimize the conducted emissions. One of the practical solutions to suppress conducted emissions is to work on the converter's modulation schemes, especially considering the availability of cyber-physical systems, which facilitating implementations and cooperation between different systems and controllers. Pulse Width Modulation (PWM) is used for most converters. In conventional PWM schemes, the harmonics power is concentrated on the deterministic or known frequencies with a significant magnitude, which leads to mechanical vibration, noise, and EMI. One solution to this problem is to use random modulation schemes. By applying randomness on the conventional PWM scheme, the harmonic power will spread out so that no harmonic of significant magnitude exists, and peak harmonics at discrete frequency are significantly reduced. In random PWM (RPWM), one of the switching parameters of the PWM signal shown in Fig.1 is varied randomly.


Figure 1. Switching signal.

where α_{k} is the pulse width, T_{k} is switching period and ε_{k} denotes the pulse delay or position.

       Existing RPWM modulation strategies are classified depending on the parameter which is made random as Random Frequency Modulation, Random Pulse Position Modulation and Random Pulse Width Modulation.

In this blog the Random pulse position is highlighted, and its implementation is presented for spreading the harmonic peaks of a DC-DC Converter.  

The basic principle of the implemented RPWM scheme

The basic principle of the implemented RPWM scheme is illustrated in Fig.2. The implemented RPWM scheme is based on randomly varying the pulse width of the switching signal. Therefore, two triangular carrier waveforms with the same switching frequency but 180-degree phase shift are used. Based on the output of a random number generator (randomly providing “0” or “1” as output), the random-carrier selector block selects one of the two carriers at the time. As a result, the carrier waveform at the selector output is a mixture of the two triangular waveforms at the input, and this waveform is compared with the reference signal in the comparator to get the required RPWM signal. This results randomness on the pulse width or the duty cycle of the switching signal.

Fig. 2.(a) Block diagram of RPWM implementation. (b) RPWM vs standard PWM: Carrier waveforms and modulation signalSteps of model-based design of generic system.

The output voltage harmonics of the DC-DC converter is shown blow in Fig.3. The comparison clearly puts in evidence that the application of Random Modulation on the DC-DC converter shows better spectral suppression compared with the conventional PWM modulation in the frequency range of interest, with consequent damping of CE peak values just over 10 dB.


Fig.3 output voltage harmonics RPWM Vs PWM

© Copyright 2020 ETOPIA

Designed with Mobirise