Automating source identification of electromagnetic interference in Printed Circuit Boards

Complex, size-constrained devices require multi-layer PCB designs that are characterized by a high density of component arrangement, where cross-component electromagnetic interferences need to be taken into account.

The aim of this thesis project was to develop a system for automated source identification of near-field electromagnetic interference to complement Antmicro’s existing hardware design process. The resulting solution consists of a CNC machine, a spectrum analyser and a set of near-field probes, orchestrated with Python scripts.

The output of the system is a set of images showing relative distribution of the emissions' spectrum across the board, with places of strong radiation highlighted. Additionally, the electromagnetic emission heatmaps can be projected onto a Blender model of the PCB to create a 3D view that shows the radiated field distribution in spatial relation to the board.