In Fig. 1 of Coelho & Rowles [J. Appl. Cryst. (2017). 50, 1331–1340] the source is shown on the goniometer circle (annotated "diffractometer radius") for the convergent beam configuration and at the focal radius for the divergent beam configuration, where the goniometer radius and the focal radius aren't necessarily equal.
For instruments using a convergent beam mirror, what is the sensible choice of primary radius (Rp)? The radius of the tube aperture or the radius of the mirror optic?