Control over velocity of a model train may be determined based upon the speed of rotation of a control knob. A processor receives electronic pulses indicating rotation of the knob beyond a predetermined increment of angular distance. The processor calculates the amount of power ultimately conveyed to the model train based not only upon the number of pulses received, but also upon the elapsed time between these pulses. The shorter the elapsed time between pulses, the greater the change in power communicated to the train. Initially, a user can rapidly rotate the knob to attain coarse control over a wide range of velocities, and then rotate the knob more slowly to achieve fine-grained control over the coarse velocity. Utilizing the control scheme in accordance with embodiments of the present invention, in a compact and uninterrupted physical motion, a user can rapidly exercise both coarse and fine control over velocity of a model train.