Maybe there is some misunderstanding, I do not rotate my setup at all.
Just put diffraction grating with proper orientation
and take it away to measure light beam dislocation for both cases.
If incoming photon would have unusual orientation of its wave front
it would change direction of order 0 a little.