CMA(p,q), a blind equalization constant modulus algorithm, is widely used in digital communication. Its parameters p and q were generally selected as positive integers 1 and 2. In order to adapt to a more general environment, it is necessary to extend the values of parameters p and q from integers to ordinary positive real numbers. Different from the traditional performance analysis method which was to recursively derive the covariance matrix of the weight error vector, based on the energy conservation relationship, the error function was directly expanded by Taylor series, and the principle of separation hypothesis was used to derive the general steady-state closed analytical expression of CMA(p,q), and an ideal mathematical model for p,q values in the selectable range was established. The simulation results show that the simulation values of complex or real data match well with the theoretical values.