Pneumatic solenoid valves are widely used in nuclear power plants and their related equipment, and their failures can lead to economic losses and safety risks. Obtaining pneumatic solenoid valve condition monitoring data in real nuclear power plant scenarios requires significant manpower and resources, and it is difficult to acquire sufficient fault characteristic data for pneumatic solenoid valves. To address these issues, this study established a pneumatic solenoid valve fault acquisition platform, utilizing actual solenoid valves to simulate common solenoid valve faults in order to obtain sufficient fault characteristic data. Firstly, faulty solenoid valves were created. In this study, faulty solenoid valve prototypes were made based on classic internal leakage, external leakage, and blockage faults of pneumatic solenoid valves. Secondly, fault data was collected to construct a fault dataset. The faulty solenoid valve prototypes were operated on the pneumatic solenoid valve fault acquisition platform, and fault characteristic data such as voltage and current during the operation of the faulty solenoid valves were collected to build the fault dataset. This dataset includes four categories of data: normal, internal leakage, external leakage, and blockage, with each category containing four groups of data. Each group collected nine pieces of data using different prototypes, totaling 144 pieces of characteristic data. Subsequently, the fault data was preprocessed using a time series segmentation method, and the fault dataset was divided into a training set and a test set. The preprocessed training set was used to train an improved CNN model embedded with time-frequency transformation. Finally, the test set was used for validation. The experimental results showed that the diagnostic accuracy of the validation set was 97.7333%, indicating the feasibility of the method. The pneumatic solenoid valve fault dataset constructed in this study has reference value for fault diagnosis research on pneumatic solenoid valves in nuclear power plants.