Resolution of an averaging paradox in the analysis of switched-mode DC-DC converters

The validity of an intuitive averaging method for estimating the low-frequency signals in switch-mode systems is rigorously examined It is shown that the approach may lead to erroneous results because it violates the Nyquist sampling theorem and that the resulting aliasing effects may interfere with the estimate of the low-frequency component of the signal. A procedure is proposed to estimate the expected error. >