Automated Tool Condition Monitoring (TCM) often relies on additional sensors sensitive to tool wear to achieve robust machining processes. The need of additional sensors could impede the implementation of tool monitoring systems in industry due to the cost and retrofitting difficulties. This paper has investigated the use of existing position encoder signals to monitor a special face turning process with constant feed per revolution and machining speed. A signal processing method by converting encoder signals into a complex-valued form and a new vibration signature extraction method based on phase function were developed to analyze the encoder signals in the frequency domain. The cumulative spectrum indicated that the spectral energy would shift from the lower to the higher frequency band with increasing cutting load. The embedded vibration signatures extracted from the encoder signals provided additional detectability of the machining condition with distinguishable spectral modes. This paper confirms the sensitivity of the encoder signals and more signatures could be extracted for tool wear detection in the future work.