Currently, the procedures that places things (grating facets, detectors) on the Rowland torus place do so exactly at the center. That means that most of a facet / detector etc is not on the torus. The average deviation could be reduced slightly by placing it such that the respective Rowland circle intersects the element in question twice.
Currently, the procedures that places things (grating facets, detectors) on the Rowland torus place do so exactly at the center. That means that most of a facet / detector etc is not on the torus. The average deviation could be reduced slightly by placing it such that the respective Rowland circle intersects the element in question twice.