A comprehensive, multi-modal system designed for the detection and analysis of Partial Discharge (PD) in high-voltage transformers. This repository contains the complete ecosystem of the project, spanning from hardware data acquisition (STM32 & Raspberry Pi) to a PC-based analytical dashboard.
The project is divided into three main operational modules to ensure safe, isolated, and highly accurate signal acquisition and analysis.
This module utilizes a Raspberry Pi paired with a USB soundcard for acoustic/analog signal acquisition. It processes the raw partial discharge waveforms at the edge before sending them to the main dashboard.
For direct electrical measurements, a custom high-voltage isolated front-end was developed. The high-voltage coupling system utilizes a total series capacitance of 1048 pF across C1 and C2 for precise signal extraction. The isolated signal is then processed by an STM32 microcontroller.
A custom Python-based digital oscilloscope and analysis dashboard. This software retrieves the acquired data, applies signal normalization, and visualizes the partial discharge pulses in real-time, comparing 35kV oscilloscope readings with the normalized Raspberry Pi data.
