We propose an experimental-theoretical method to identify and compensate static errors in the phase coordinates of the elements of an optical encoder. The technique developed is applied to the compensation of systematic errors in angle measurements in the modern optical encoder.