This works, more or less;
https://www.seventransistorlabs.com/Images/DistAmp2.jpgThe important part being the metal plate underneath everything, and the LISN on the left, into which the power supplies connect. Its RF ports are terminated as shown, since I was using it as a filter in this case, but yeah, that's where you'd measure things of course.
Testerizing conducted susceptibility requires a modestly powerful amplifier and a sweep generator; which, interestingly enough, is the main intended application of the above mentioned amplifier (the shiny stuff on the right side). You will need more than a few watts to reach useful output levels (3 or 10V are typical), times a fair factor to account for the attenuator required (to help isolate the amplifier from reflected power at unlucky frequencies), and you will also need a cable clamp signal injector of some sort, to deliver common mode conducted energy to cables that aren't simply wired into a LISN.
A similar setup is used for EFT testing, and surge (where applicable) is coupled in with a CDN as well.
Tim