One feature includes a method for implementing a fixed point recursive filter that reduces or eliminates steady state error. The method comprises obtaining a first filter state value, processing the first filter state value to remove a scaling factor to obtain a second filter state value, ascertaining that the recursive filter has reached a steady state, determining a nonlinear drift parameter based on a difference between the first filter state value and the second filter state value multiplied by the scaling factor, and adjusting the second filter state value with the nonlinear drift parameter to reduce steady state error of the recursive filter. Ascertaining that the recursive filter has reached the steady state may include determining that a filter output value at time n is equal to a filter output value at time n−1.