Time domain characterization of a system based on the magnitude of its frequency response

In EMC tests, systems are usually described by the magnitude of their frequency responce. Because of the lack of phase information, it is difficult to estimate the time domain characteristics of a system from this kind of data. Aimed to solve this problem, a minimal phase model is proposed to approximate to the real system. Based on the assumption of minimal phase system, the magnitude of frequency response can be related to phase by Hilbert transform. Thus the phase can be reciveved from the magnitude of the frequency response and the impulse response can be estimated. A discrete transfer function model is further used to establish a parametric model for the system, which can describe the system characteristics both in time domain and frequency domain. Examples are given to verify the validity of this method in estimating the phase information for the transfer function of a magnetic filed sensor. Applications of this method in system modeling, signal reconstruction and time domain response predication are also given.