Optimization methods for equivalent source identification and electromagnetic model creation based on near-field measurements

2011 
The field emission from printed circuit boards (PCB) plays an important role in the electromagnetic compatibility of electronic systems. The established field measurement methods according to CISPR suffer from the need to use large anechoic chambers. Furthermore the measurement data cannot be used for modeling concerning the calculation of the overall fields radiated from e.g. a car. Other methods which try to identify an equivalent source distribution by near-field measurements do not require large anechoic chambers, but instead an electromagnetic inverse problem has to be solved. This often leads to an ill-posed equation system due to unavoidable errors in measurement data, long computation times, and finally inaccurate results. In this paper an approach to optimize the characterization method of printed circuit boards by near-field measurements is proposed. The radiation of a PCB is modeled with a set of elementary sources, resulting in the same field like the electronic system itself. The electromagnetic fields in a plane above the PCB are measured. Optimized time domain methods are applied in order to reduce measurement time, to receive phase information, and to correlate different measurement data sets. As the current distribution on a PCB mainly depends on the conductor paths, the distribution of characterizing equivalent sources can be chosen with respect to the trace routing. Amplitude and phase correlation of the equivalent sources along each current path are taken into account. The routing information can be extracted from mechanical CAD-data or is based on high-resolution near field scans. The approaches are integrated in the equivalent source identification process. The advantages of the new method are presented and discussed.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    8
    References
    14
    Citations
    NaN
    KQI
    []