With the aim of controlling the motion of robots without involving complex design procedures or extensive online computations, this paper presents a strategy consisting of a linear feedback control and a self-tuning control. The main advantage of the strategy is its simplicity in design, program code and real-time implementation. The only information required in setting up the strategy is the order and the measurable states of the system, while detailed description of the model is not needed. Furthermore, uncertain effects such as static and dynamic joint frictions and external disturbances can easily be handled. In addition, all the internal signals are uniformly bounded and the control torque is smooth everywhere. These features are also verified via simulation on the GE-P50 robot.