In this paper, an experimental validation of a low-cost prototype for fault detection and classification of a photovoltaic (PV) system is presented. The prototype can detect, classify faults and displaying the results on the website via Internet of Things (IoT). The experiments result have been carried out at MIS Laboratory of Picardie Jules Verne University, (France). The code is developed under Matlab/Simulink environment and then implemented into a dSPACE- 1104 for experimental verification. The obtained results of the developed prototype offer similar results as the one obtained using a dSPACE-1104. This confirm the accuracy of the implemented fault detection and classification algorithm (the investigated fault are: short-circuit, open-circuit, shading effect, and dust accumulations on PV modules). \textcopyright 2020, The Author(s), under exclusive license to Springer Nature Switzerland AG.