The most efficient way of building up energy in an oscillating system is to do it in phase with a resonant frequency of the system. If the resonant frequency changes for any reason this is difficult to obtain with a static frequency generator. It is of course possible to sweep over a frequency interval to find the maximum repeatedly, but a frequency sweep takes time and it is performed at discrete frequencies. A better approach is to set the device under test in feedback resonance. This will guarantee that the frequency is always a peak in the spectrum and the adaptation to change is immediate and continuous in every aspect. A continuous observation of feedback frequency can conceivably serve as an SHM indicator. Experiments with geophones as actuators are performed. A 9-day test revealed smooth frequency variations in the 236.901 to 237.353 Hz interval. These smooth variations are believed to be caused by thermal and humidity changes in the laboratory. An audible tension release in the test device during this period was clearly indicated by a momentary step in the resonant frequency. A second experiment revealed that as the feedback gain was adjusted in steps, the feedback frequency followed suit. This establishes indirectly the relation between deformation and resonant frequency. It is thus possible to determine non-linearity with the controlled feedback resonance method.