Summary:

MC alignment as well as IMs are very good. The beam hits the center of PRM and PR2. 

With this good alignment, when we steer PRM so that the REFL goes back to HAM1 without clipping, PRM alignment offsets are now 3395urad for PIT and 50urad for YAW. The largest DAC output decreased from -126000 counts (almost railing) to about -10750  -100750 counts,  so it's somewhat better but not that much. It seems like PRM really tilted.

At this point, PRM offset should be released, and after that right PRM offset should be re-established.

What was done:

1. Both pico mirrors and fixed mirror upstream of MC1 were used to center the beam on MC2 as well as on MC1 (mostly MC2).
2. MC2 was aligned so the single-bounce beam off of MC2 comes to a correct position in HAM2.
3. Used MC3 and MC1 to refine MC alignment. Excellent 00 flashes.
4. Used IM1 and IM2 so the forward-going beam goes through Faraday input and output aperture at roughly the same positions as before. No clipping.
5. Used IM3 to bring the beam to the center of PRM, and IM4 to center the bean on PR2, and iterate.
6. Align PRM to center the REFL beam on the baffle hole for HAM1 beam path.

Observation after the above:

• PRM PIT offset was reduced but it was still large.
• We didn't check if the beam was hitting PR3 (cannot, the beam is huge and the power is small), but the reflection off of PR2 was at a reasonable location in HAM3.
• The beam is too close to the center of the baffle in front of IM4. This is not a good thing according to Cheryl, as the baffle is designed such that when PRM is misaligned in YAW to direct the REFL beam to the parking beam dump, both the forward and backward (REFL) beams are equally away from the baffle edge. It seems like the baffle itself could move toward HAM1 by 2mm or so.

What was further done:

1. Misalign PRM to direct the REFL beam to the parking beam dump. PRM offsets railed and the beam was still not centered on the baffle hole for the beam dump.

What needs to be done:

1. Relieve PRM bias.
2. Align PRM so we have a nice refl beam again.
3. Confirm that we can misalign PRM to park the beam.
4. Move the baffle in front of IM4 somewhat closer to HAM1 and confirm that there's no clipping.

What doesn't need to be done:

1. Revisit beam paths to IOT2L viewports. It's impossible for these beams to miss the viewports by these tiny changes (tiny in comparison to the viewport size).
2. Revisit POP QPD sled (PRM-PR2 line is good), MC2 trans QPD (the beam is centered on MC2 so the transmission will hit the QPD), and SM2 trans QPD (lever arm is small so this change would not make the beam completely miss the QPD). We will use picos once things are under vacuum.