Feedback Noise Propagation in Multisampled DC–DC Power Electronic Converters

Abstract

This article analyzes the propagation of feedback noise in multisampled dc–dc power converters. The analytical calculations for noise attenuation, strictly valid for linear time-invariant systems, are found to offer good predictions for power converters, after suppressing the decimation effects of the digital pulsewidth modulator (DPWM). For control systems that employ a proportional-integral controller, without any digital low-pass filters, the nonlinearity caused by the DPWM decimation is found to saturate the noise attenuation properties, as the multisampling factor <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$N$</tex-math></inline-formula> is increased. Strong noise attenuation is enabled using antialiasing digital filters, with a small impact on the dynamic response. For control systems that employ a proportional-integral-derivative controller, increase of the multisampling factor, without any digital filters, causes the amplification of the noise power. Therefore, a greater attenuation of high-frequency components is required to provide a significant noise reduction in the control bandwidth. Even with these antialiasing digital filters, the dynamic response of multisampled control systems is improved compared to double-update. This is proven by the analytical and experimental comparison of various multisampled control strategies in terms of dynamic response and noise attenuation capabilities. The experimental results, from a buck-type converter, match well with simulations and analytical calculations.