SummaryIn order to control the speed or the position of an electrical drive, the torque of the machine has to be set accordingly. The torque is a function of the machine current which is set commonly by means of a voltage source power converter. Because of the simplification of the set-up procedure, cascaded control structures are used. For a stable operation it is important that the speed of the inner control loop is much higher than the outer ones. Hence the speed of the current control limits the dynamics of the drive control. This means that the speed and the accuracy of the current control is decisive for the overall performance of the drive control.In this paper a new control method will be presented which is capable to reach the setpoint value in the shortest possible time of only one control period (comp. fig. 7) but without the knowledge of the machine parameters. This is the fastest possible response for digital controls which can be obtained theoretically by deadbeat controllers. However, high dynamics can hardly be achieved by deadbeat controls in practice because they are very sensitive against variations of the control parameters. Therefore the machine parameters like the inductance and the resistance as well as the course of the induced voltage have to be determined by a measurement for a proper operation. Furthermore the control has to be adjusted manually by the user. This leads to an high effort for the set-up procedure. In addition these parameters may vary during operation because of warming for instance, which can lead to a maladjustment of the control.The new control will adapt onto the load automatically which leads to a significant simplification of the set-up procedure. The identification of the control path is based on the detection of the current slopes of the load current ripple, caused by the alternating of the switching states of the power converter. Hence no additional test pulse are necessary for the identification of the system behaviour. This permanent adaptation leads to an optimal dynamic behaviour of the drive system. In contrast to many model predictive control methods, the new control has a low computation effort and no extensive model of the control path is necessary.The principle function of this new control approach will be explained and verified for an armature current control of a d.c.-drive system. However the principles of this control can also be applied on three-phase application which will be demonstrated by experimental results.