Data pattern based electromagnetic interference modeling for IC packaging

Electromagnetic interference (EMI) is becoming more serious when the data rate of the digital system continues increasing. However, its modeling and analysis are very challenging. In this paper, a novel electromagnetic superposition method is developed to model the IC packaging electromagnetic emission. It employs the equivalence principle to obtain the electromagnetic field response over a broad spectrum. Then it uses the linear property of the passive parasitic system to superpose the contribution of different signals on the packaging. As a result, with certain pre-calculations, it is convenient to compute the electromagnetic emission efficiently from different signals with various data pattern combinations, which benefits identifying worst case scenario. In addition, data reconstruction can be evaluated through the phase shift, which simplifies identifying the EMI of any pulse bit pattern. This work provides a basic modeling framework for comprehensive radiation studies.