CE-marking specifies maximum levels of radiation, both on the cable and radiated to free air, to minimize risk that a device interferes with other devices. And when HF radiation leaks out to a cable the cable effectively becomes an antenna. So, in short, these ugly knots are from the beginning expensive and ugly quick-fixes to poorly designed PCB's, to avoid having to re-do the PCB when you are in a hurry to start selling products and quickly need that CE-marking done. So why are those on Excito products then, you would ask? That does not make sense?
Now after many years of the industry struggling with HF emissions (it gets worse as clock frequencies on boards go up), these beads have become more or less a standard in the industry. It's unusual when you buy standard PSU's (as we do) to get them without beads. (When I look around my desk, it's (as expected) only Apple who cares much enough to make sure they don't need beads, none of my apple products have them, but everything else). The B3 actually passed the emission tests with really good results so we wouldn't have needed them at all, thanks to the B3 all-aluminum casing.
The fact that some of our (standard) PSU's have beads near the PSU and other near the device is just a coincidence, the PSU with the bead near the PSU probably needed it to pass its CE test.
(And no, shielding does not always help for this, if HF leaks on the shield (which it sometimes does) the shield becomes just as good as an antenna. My Microsoft mouse has a shielded cable AND a bead.)
Sorry for the long reply, you found my sweet-spot.
