Repetitive perfect tracking control (RPTC) is one of repetitive control techniques to achieve high precision positioning. In this paper, RPTC with n-times learning filter is proposed. n-time learning filter has a sharp roll-off property than conventional learning filter. By using n-times learning filter, proposed RPTC can converge tracking error n-times faster than RPTC with conventional learning filter. Simulations show the fast convergence of proposed RPTC. Finally, experiments in ball screw driven stage also show the effectiveness of proposed system.