Diagnosis of compressor product's malfunctions based on the vibration wave pattern data of the compressor and the FFT analysis data is developed. The feature of the analysis technique is two points. One is "Evaluating the ratio of the FFT integral calculus value of low frequency area and the high frequency area of the vibration wave pattern for the feature of abnormalitys", and the other is "Separating and evaluating the wave pattern region at the time of the start and the steady part as correspondence to the phenomenon that an abnormal tendency appears in the non-steady early start region". The feature of each abnormality is extracted. A diagnosis algorithm that distinguished normality and abnormality is developed. Moreover, the cause is distinguished about abnormality. The verification examination of the scroll compressor is done in the production line. And the effectiveness of this diagnosis algorithm is validated for both the alternating-current motor type scroll compressor and the direct-current motor type scroll compressor.