Statistical Performance Validation of an Embedded PSoC-Based Ball Bearing Fault Detection System

Bearings are critical components in rotating machinery, and their failure can lead to catastrophic system breakdowns and costly downtime. Early detection of localized defects
such as those in the inner race, outer race, and rolling elements are therefore indispensable for predictive maintenance. This paper presents the design and statistical validation of a novel, embedded workbench for ball bearing fault diagnosis. The system employs an impulse
excitation technique, where a solenoid induces a controlled vibrational trigger, and an
accelerometer captures the response. The core implementation leverages a Programmable
System on Chip (PSoC) for embedded data acquisition and a LabVIEW-based Virtual
Instrumentation workbench for signal processing and analysis. The fault diagnosis is
performed by examining the computed power spectrum of the vibration signals. Distinct fault
signatures are identified: new bearings exhibit sharp amplitude peaks, outer race defects show
a distinct peak rise, inner race defects generate sideband peaks, and ball defects result in a
fully distorted spectrum. To rigorously validate the reliability of this approach, a statistical
analysis was conducted on four new bearing models (6203-N, 6201-N, 6300-N, 6000-N). The
results demonstrate with 95% confidence that the developed smart workbench produces
consistent and statistically significant outcomes. The system's performance confirms its
accuracy and repeatability, establishing it as a promising and validated tool for the static
testing and fault diagnosis of rolling element bearings.