Keita, Camilla. Followed on from alog 58875.
We injected only s-pol light and then checked the output through SFI2 was also s-pol. As expected: see T1900789. To do this we put a flipped a PBS tor reflect any p-pol down to the table.
With 5mW s-pol out of fiber we measured 7μW out of SFI2- much too small. We realized that our previously used SFI2 input angle of 16deg was for outgoing p-pol so for incoming s-pol should be offset 1.7deg ~ 1.8mm for 58mm distance. We already nearly clipping B:M3 so shifted B:M3 and the bema using the fiber to move the beam to have less of an incoming angle to SFI2 and to center the beam more of B:M3. We found horizontal alignment is easier with the beam above the aperture. Then drop the beam down using B:M3 once well aligned horizontally. Now had improved SFI2 output of 65μW. Power through SFI2 has loss by factor of 73.
| Location | Power |
| Out of fiber, s-pol only | 5mW |
| Before B:BS1 | 5mW |
| After B:L2 (and B:BS1) | 4.4mW |
| After SFI2 | 65μW |
This could be from clipping or misaligned apertures?
We took a beam profile measurement which showed slight divergence and a beam much wider than the apertures. Data attached and photo of set up. D4sigma beam width at SFI2 1st aperture (689mm from launcher) is ~3.3mm. Aperture width ~1-1.5mm. So some loss is expected but not a factor of 73.
Keita photographed the height of apertures for if we decide to remove them.
B-path in the week of May/10:
Problem 1a:
With both of the alignment helper apertures on, when injecting S-pol light from output side of SFI2, after tweaking the alignment to maximize the power transmission through B:P2 and B:F1, the beam is not level with the platform, it is off by ~5mrad in PIT. That's not a huge number (e.g. cos(5mrad)~13ppm) but is quite noticable (1/16" over 12").
We always ended up injecting at a correct height at B:P2 (picture 2, nominal height should be 2.5") but the transmitted beam is about 1/16" lower at the position of B:P1 (picture 1). Something like that happened when Sheila and I were checking the retro-reflection of A path through SFI1 even though the apertures weren't on, so we spent some time to see what was going on, but unfortunately haven't found anything obvious.
Aperture height is well within 1/100" from nominal 2.5" judging from the picture 3 and 4 (see D2000605 and D2000610 respectively, note that the ruler is on 3/8" thick SFI sled in picture 3 but not in picture 4, also note that the ruler is upside down in picture 4).
When we force the beam to be level, the transmission significantly drops (~1mW out of ~5mW injection) but that's still usable for setting the roll/yaw/pit angle of B:P1, so that's what we'll do in the week of May 17.
We're not tricked by wrong polarization, we're getting what we expect (S-pol going toward SFI1 when we inject S-pol).
Problem 1b:
With "optimized" alignment, we're only getting ~3.2mW out of B:F1 when 5mW S-pol is injected from B:M4.
Beam size we're using is not great but not terrible either (the beam radius is about 116um at B:P2 and ~430um at the input side aperture while the aperture radius is 0.025"~640um) so it's not like a huge amount should be clipped by the aperture.
For now we're not that worried, we'll see if, after alignment is done and the alignment apertures are off, the throughput etc. are good.
Problem 2:
When we tried to set the B:P1 angle, we noticed that the AR reflection is blocked by the vertical "hand" of the mirror holder unless the main beam spot is drastically offset to the side.
Currently, wedge holder assy is held inside a cylindrical 0.75" to 1" adapter, which is held by a right handed 1" Siskiyou. But nominally this should be 0.75" left-handed Siskiyou without 0.75" to 1" adapter (https://dcc.ligo.org/LIGO-D2000087). We're trying to locate the correct part.
In order to decrease the beam size to get less light clipping on the SFI2 aperture, we added a 257.5 ROC lens just before B:M4. We measured the beam profile with the lens in approximately the same position and expect the waist to be around the first (B:M3 side) SFI2 aperture. See beam measurement and approximate positions of apertures attached made using Nick Smith's alm. The apertures are ~1.3mm in diameter so we now do not expect clipping on input or output aperture with this temporary lens.
By injecting 5.6mW onto platform, we had 1.2mW before adding the lens and 3.2mW after SFI2 with lens.
Problem 2 fixed and wedge aligned to the horizontal (rather than maximized through SFI2) beam - see alog 58999.
