Paper deals with NN compensated odd symmetric nonlinear actuators. Most actuators are nonlinear sporting odd symmetric nonlinearities such as deadzone and saturation. One of ways of dealing with such actuators is to try to compensate for nonlinearities by static neural networks. Compensated actuators can improve behavior of the system, but then arises the problem of stability analysis because compensated nonlinearity is now complex nonlinearity not described in common literature. One way of dealing with such problem is to perform stability analysis via describing function method. Paper describes the method for compensating nonlinearities, recording describing function and performing stability analysis.