An extension of the Cartesian product operators formalism for scalar coupled I=1/2 and S=3 nuclear spins is proposed in order to predict the behavior of the function modulating the I spin echo amplitudes in spin-echo double resonance (SEDOR) experiments. With such an approach the coupling constant can be evaluated quite independently of the S=3 quadrupolar line broadening and magnetic field inhomogeneity. The method, which can be generalized to a wide class of molecules, is tested by the experimental behavior of the hydrogen SEDOR modulation by the B-10 nuclei in (Na2B12H)-B-10(11)SH.