Just to add that the first HWP is slightly tilted so that the reflected beam (~1mm size from sensor card) is at the side of the laser aperture.

The beam sampler was actually replaced by a 2 inch polarizing beam splitter from CVI (TFP-1064-PW-2037).

The previous issue with power control was solved by tuning it closer to its Brewster angle (56deg) allowing to change the transmitted power from 16 to 370mW (maximum power at 2A).

We reduced the laser current to 2A to prevent potential damage of this 30years old laser following Akutsu-san recommendation.