This study aims to extract the stationary features from a non-stationary data using continuous wavelet transform. For this purpose a Hall effect sensor is used to obtain current measurements during the welding operation. A welding process includes several stages referred to as the initial case, transient case and operation case. In this manner, some frequency components can be determined for these cases thereby defining behavior of a particular welding machine. Regarding to the wavelet analysis results, fundamental frequency at 50 Hz is determined to be a dominant characteristic for the considered application. Moreover, side band effects are observed around the fundamental frequency. A sinusoidal waveform of 50 Hz is localized with huge amplitude values in very short time of the time-scale plane and this duration is required during the welding operation in terms of contact between the electrode and material surface.