In this paper, an improved Backstepping _Port Controlled Hamiltonian (STABS_PCH) nonlinear control structure combined with a Stator Current Based Model Reference Adaptive System speed observer using neural network (VMNN_SC_MRAS) is presented. The STABS_PCH control improves and enhances the performance and robustness of SPIM drives. The combination of STABS_PCH controller with the VMNN_SC_MRAS speed estimator can compensate for the uncertainties of the machine parameters and load disturbances to improve the dynamic performance and of robustness the SPIM sensorless drive systems. The effectiveness of proposal control scheme is validated through Matlab-Simulink.