Abstract
Light-emitting diodes (LEDs) play a crucial role in various practical applications. This paper presents a simple and cost-effective approach for detecting faults in LED boards with a large number. The types of faults detected include open circuits and short circuits. The proposed solution is based on a voltage threshold and a statistical algorithm to filter out noise. The paper first outlines the design concept, followed by the development and fabrication of a fault classification module. Subsequently, a control center is established to collect signals from all fault classification modules connected to LED boards in pairs. The received signals are processed and stored on a memory card. Finally, the resulting data can be accessed either from the memory card or directly via an external computer through a USB port. The experimental results demonstrate that the proposed design and fabricated device can accurately detect and classify faults on the tested boards. The data obtained can be used for statistical analysis and root cause investigation to enhance product quality.