Localization of the Equivalent Sources on the PCB Surface by Using Ultra-Wideband Time Domain Near-Field Measurements

In this paper a 2D parametric identification algorithm of stochastic electromagnetic interference (EMI) sources based on near-field measurements is presented. The unintentional radiation of electronic devices and connecting cables can be measured by using two-point near-field planar scanning probes connected to the digital oscilloscope for ultra-wideband (UWB) time-domain data recording. The characterization of the registered stochastic time-domain field distribution was implemented by evaluating the autocorrelation and cross-correlation spectra. Due to the cyclostationary property of the stochastic information signals, the ensemble averaging procedure allows evaluating amplitudes and initial phases of clock harmonics, used for the synchronization of electronic device and information transmitting through the cables. The radiating object could be modelled as a set of equivalent electrical dipoles arranged on a grid in the object plane. To determine the equivalent dipole moments, the inverse procedure for the estimated clock harmonic was realized assuming the near-field propagation model. The final localization of the equivalent dipoles for the defined clock harmonic was implemented by the application of 2D Matrix Pencil algorithm and Minimum Least Square fitting procedure. The experimental data was processed with proposed parametric identification procedure for a spatial localization of equivalent dipoles and effective sources on the surface of the radiating structure.     «

In this paper a 2D parametric identification algorithm of stochastic electromagnetic interference (EMI) sources based on near-field measurements is presented. The unintentional radiation of electronic devices and connecting cables can be measured by using two-point near-field planar scanning probes connected to the digital oscilloscope for ultra-wideband (UWB) time-domain data recording. The characterization of the registered stochastic time-domain field distribution was implemented by evaluatin...     »