[Sheila, Robert, Eric, Daniel, Karmeng]

By blocking and unblocking the beam from the OFI heading into HAM7, we see 1uW on H1:IOO-OFI_PD_A_DC_POWER, which is measuring power after a 99% splitter (B:BS1, E2000450, 99% reflective for s pol light at 45% AOI).  

Scattered light entering HAM7: average of 25.66uW and STDev 1.066uW. Single bounce from IFO, with input power of 9.2W. With 9.2W on PRM we expect 9.2W * 0.03 (PRM) *0.25 (BS) * 0.32 (SRM) = 22mW incident on the OFI, so this implies that the leakage towards HAM7 is 0.1% of the beam incident on the OFI; 29dB.  This implies that the isolation of this path from the main beam is better than what we infered assuming 1% transmission from B:BS1 in 87350, and is close to the isolation assumed in T1800447.  

We also measured the polarization of the beam by measuring the beams after they hit the polarizer on SFI2.  Consulting the diagram on page 8 of T1900649, this polarizer is set to transmit vertically polarized light and deflect horiztonal to hit the aperature which isn't labeled in the diagram.  We measured the beam between the rotator and the second polarizer (B:P1),after the aperture, with average of 14.07uW and STDev of 1.587uW.   This implies that the beam coming from the OFI is 55% vertically polarized.

Revisiting 87318 in light of these measurements.  If the polarization and splitting ratio of B:BS1 in full lock is the same as single bounce, this implies that we had 0.75mW scattered towards HAM7 from HAM5 in full lock.  Using the 4% transmission of this diode (and assuming that 30% of the AS light is carrier, which I want to double check), we need a reflectivity of 3% from some scatter source after the beamsplitter to explain the 12nW of scattered light reaching the DCPDs.  

  The thin film polarizer on the OFI should have an extinction ratio of 1:10,000 E1900285.  Also, in 88105 I moved the temperature of the OFI while we were locked and saw that we couldn't significantly decrease the scattered power that way.