The optics (marked in pic 1) as 1 and 2 were changed. Previously we had some mirrors(perhaps 5104), and they were causing the polarization to change too much. So, they were replaced with Silver mirror PF10-03_P01-10. Then the polarizer was installed in rotation mount at place (a) shown in pic.
I have taken the Thorlabs cube polarizer from LC table. The previously used one in PCI was newport one, and it was not properly installed in the thorlabs cage rotator.
The polarizer alignment has to be tuned slightly more, but for now the polarization at the camera goes from azimuth (-90,90), and ellitpcity (0-2) degrees.
After moving the mirrors, I had to realign the beam. The alignment was tuned to obtain horizontal and vertical to be within 0.027deg and -0.016deg. See the plot 1 and 2 showing the horizontal and vertical alignment.
Actually the reason to install the metallic mirrors was to remove the polarizer to avoid possible power discrepancy at different rotation angle.
The HWP is mounted inside the motorized rotator and provides good azimuth coverage with only minimal ellipticity (~<1 deg)