The objective of the study is the synthesis of an algorithm for the automatic control of an electromechanical actuator in the rotational joint of a multi-link, robotic manipulator. Control performance is assessed in terms of ability of the controlled joint to follow a commanded joint angle, or a commanded joint angular rate. Part One of the report describes the development and analysis of a mathematical (truth-)model suitable for control algorithm synthesis. Synthesis is carried out in the continuous time domain. Three control algorithms are described: classical PID control; redesign of PID control with a model-following inner loop; and a redesign based on modern control principles, involving an observer, two feedforward loops, and one feedback loop. This Part concludes with some guidelines for parameteridentification. Part Two of the report describes some exploratory numerical simulations involving continuous time PD control and discrete time PD control. Further, extensive and systematic simulations are required to determine a definite preference for one of the three control algorithms. Simulation results involving classical control already indicate that even for fairly ideal cases there is a need for control redesign.