Abstract: Planetary gearboxes are widely used in large electrical equipment with low-speed and heavy load, and the fault diagnosis of planetary gearboxes is particularly important. Currently, fault detection techniques mainly relies on vibration signal; however, due to the weak fault-related impacts at low-speed operating conditions and the difficulty of separating faulty impulse, it is difficult to detect the fault. In view of this, a low-speed planetary gearbox fault diagnosis method based on the encoder signal was proposed. Firstly, the method obtains the fault information via the built-in encoder, avoiding the adverse effects caused by the lengthy vibration transmission path. After that, a sparse low-rank decomposition model was established and the fast principal component pursuit (FPCP) algorithm was introduced to solve the extracting planetary gearbox fault impulse at low speed. The experimental results of planetary gearbox indicate that the proposed method can not only obtain the fault information at the input shaft speed of 30r/min, but also effectively separate and extract the fault impulse. Therefore, it can provide an effective tool for fault diagnosis of low-speed rotating machinery.