In the framework of our research on high-gain observers for nonlinear systems we have reached the final step of the development of our solution, which consists of real-time implementation in Linux. The results presented in this paper were obtained using RTAI-lab and the Scilab/Scicos CACSD environment on a series-connected DC motor. We introduce the three observers we implemented as 1) the classic high-gain Luenberger, 2) the high-gain extended Kalman filter, and 3) our newly proposed adaptive-gain extended Kalman filter (AEKF). Finally we close the control loop with a classic control scheme for DC motors in order to give a complete view of the performance.