Black-Box Behavioral DC-DC Converter IC Emission Model

Electrified, automated, connected, and shared mobility trends have led to stringent EMC requirements for in-vehicle equipment. In addition to noise countermeasure technology, it is becoming important to predict EMC risk from the early design stage. Simulation is an effective method to analyze and monitor EMI/EMC performance such that possible upstream problems can be addressed cost-effectively. A primary challenge towards such a simulation methodology is the non-availability of models for Integrated Circuits (ICs) which are the sources of noise. In this work, an operating condition-dependent behavioral model for ICs is developed using measured data on a specially designed test-PCB followed by training a machine-learning network. This is combined with an electromagnetic simulation framework to generate system-level EMC results. The proposed model-based simulation methodology is validated with measurements for a DC-DC converter system.