Purpose <p>Power transformers are essential components of every power system. Failures in power transformers can result in significant damage, including prolonged power outages for consumers and industries. Consequently, their condition must be continuously monitored to maintain their reliable operation. In recent years, researchers have focused on vibrational properties of power transformers as a promising online and non-invasive method for their fault diagnosis. This paper aims to provide a comprehensive review and classification of vibration-based fault detection techniques for power transformers, addressing their origin, signal propagation, measurement techniques, signal processing methods, and vibration modeling.</p> Method <p>This study summarizes recent research and developments in power transformer fault diagnosis using vibration signal processing. It reviews and classifies existing vibration-based detection techniques, emphasizing practical applications. The study also categorizes methods based on the origin and source of vibrations, the propagation of signals, and practical measurement and signal processing techniques. Furthermore, the paper examines approaches to transformer vibration modeling, providing a structured perspective on existing methods.</p> Conclusion <p>The findings of this study highlight the current state of vibration-based fault detection techniques for power transformers. The conclusion shows future research directions and emphasizes the need for novel vibration signal processing techniques to improve fault detection accuracy and reliability. These advancements could significantly enhance transformer monitoring and extend their operational lifespan.</p>