Abstract: A device detects an emission source of a vibration signal against a background of noise, where the vibration signal is produced by a mechanical system and has one or more frequency components. A sensor senses the vibration signal and produces a sensor signal corresponding to the vibration signal. The sensor has a mechanical frequency selection structure for mechanically resonating at desired frequencies and for mechanically attenuating a frequency component of the vibration signal having a frequency outside the desired frequencies. The mechanical frequency selection structure thus produces a mechanically-tuned sensor signal. Electrical frequency selection filters electrically attenuate a frequency component of the mechanically-tuned sensor signal having a frequency outside the desired frequencies. The electrical frequency selection filters thus produce an electrically-tuned sensor signal.

Abstract: A signal processing device for processing a machine vibration signal includes a peak value detector for determining the peak vibration amplitude values during predetermined sample time periods. In one embodiment, the vibration signal is digitally processed prior to receipt by a digital peak value detector. In another embodiment, analog processing and peak value detection of the vibration signal is employed. The peak values are then made available for further processing. In many situations, the peak values will be associated with mechanical faults which will be detected at periodic rates corresponding to the mechanical component defect frequencies without additional signal enhancement. Further enhancement can be acquired by synchronous averaging the peak values with the speed of a rotating element.