We've found that there were actually two ghost beams seemingly from the HAM1-HAM2 septum viewport for the main laser beam.
After installing nozzle baffle at for the PSL-HAM1 viewport, we've found that the ghost beams were reflected by the baffle and going down (and were easily visible). One was hitting the septum plate and the other was hitting the HAM1 table surface.
We installed two V-shaped beam dumps to catch these on HAM1. Betsy has a bunch of pictures.
Pic of the 2 septum viewport beams coming back to the new nozzle baffle in the PSL-IN viewport below.
PSL beam location is now a bit high and -X in the baffle aperture but not clipping by at least a beam diameter as seen on the card, pics attached below. This location is with the baffle pushed up and left as much as possible.
New septum-to-baffle-to-table reflection beam dump location pics below. 2 new beam dumps are centerline in middle of the table.
Pics of the beams before adding the 2 beam dumps to catch them, below. Interesting that beams likely used to come back from the septum and maybe into the PSL-IN viewport and then back onto the table somewhere. The nozzle baffle seems to just redirect them now to our new beam dumps.