We hypothesize that one role of sensorimotor feedback for rhythmic movements in biological organisms is to synchronize the frequency of movements to the mechanical resonance of the body. Our hypothesis is based on recent studies that have shown the advantage of moving at mechanical resonance and how such synchronization may be possible in biology. We test our hypothesis by developing a physical system that consists of a silicon-neuron central pattern generator (CPG), which controls the motion of a beam, and position sensors that provide feedback information to the CPG. The silicon neurons that we use are integrated circuits that generate neural signals based on the Hodgkin-Huxley dynamics. We use this physical system to develop a model of the interaction between the sensory feedback and the complex dynamics of the neurons to create the closed-loop system behavior. This model is then used to describe the conditions under which our hypothesis is valid and the general effects of sensorimotor feedback on the rhythmic movements of this system.