sorry – I am out of sequence with my replies – you note “A wide spectrum, even if it does not affect the total flux is not ideal as you have to sum more rows, introducing more noise from the camera and sky background.”
I think thats what I found (and optimal binning doesnt play nicely).
It looks like you dont see the progressive vertical widening as you move the mirror (so the image goes down the sensor)?
I think Andy saw this?
Looking at the geometry, the narrow spectrum appears to be when there are smaller angles involved in the projection between mirror, grating and sensor. So might you expect some spreading as you increase these angles? My spectrum when at the top of the image is quite narrow… I have not tried loosening the screws on the grating holder – it sounded scary 🙂 But maybe I should try and re-tighten to get a feel for how this works