Designing of Low-cost Acoustic Analyzer Device with Arduino for Monitoring Hydropower Generator

Abstract

Low-cost acoustic monitoring can support condition-related data collection for hydropower generators where commercial instruments may be costly or difficult to deploy. This study developed an acoustic acquisition and monitoring arrangement using a KY-038 sound sensor, a parabolic reflector, and an Arduino Mega 2560 architecture for a 15 MW Kaplan hydroturbine alternator. Approximately 100 measurements were collected in each of three defined alternator sectors during normal operation. The sensor output was converted from analogue voltage to pressure and sound level for recording and interpretation. The measurements showed a spatial increase in sound level with proximity to the rotor. Reported average levels for the non-pressed configuration were approximately 73.05 dB in the cooling region, 75.83 dB in the air-gap region, and 79.05 dB near the rotor. With the reflector pressed against the measurement location, the reported sector averages were 74.24, 76.27, and 80.69 dB, respectively. The two acquisition configurations therefore showed a similar spatial pattern, while the pressed configuration produced slightly higher measured levels. These findings indicate that consistent sensor position and contact are important for repeatable baseline acoustic mapping. The proposed low-cost arrangement provides a practical basis for repeated monitoring under comparable operating conditions. However, measurements were not obtained under confirmed fault conditions, and the parabolic-reflector mounting technique requires further improvement and validation.