Reports until 21:56, Friday 06 February 2026
H1 IOO (IOO)
masayuki.nakano@LIGO.ORG - posted 21:56, Friday 06 February 2026 - last comment - 15:55, Saturday 07 February 2026(89073)
Vertical splited beam investigation

Summary of investigation into the vertically split beam from the EOM

 

Comments related to this report
keita.kawabe@LIGO.ORG - 15:25, Saturday 07 February 2026 (89075)ISC, SYS

Horizontal beam position offset on the EOM input and output aperture on the side plates.

We realized that the nominal beam position on the EOM input and output aperture is NOT centered on the crystal cross section projected onto the side plate face, the beam is horizontally offset in +Y direction.

Look at the first cartoon (cartoon.jpg) and references therein. The beam spot offsets are 0.91mm on the input side plate and 0.54mm on the output side plate, respectively, assuming that the beam deflection angle per surface of EOM is 2.35 degrees as implied in D2500130.

0.91mm is not a small offset, it's almost 1/4 of the crystal thickness (it's 4x4x40mm).

This means that the beam should be (see nominal_sideplate.png, note that the drawing scale of the input aperture in this is twice that of the output aperture):

~3.9mm from the left (+Y) edge of the visual alignment aid notch on the input side plate,

~3.2mm from the right (again +Y) edge of the aperture hole on the output side plate.

Measurements, adjustments and measurements made the beam closer to the nominal location.

Based on the above knowledge, we took pictures of the beam position on the input/output aperture, paying attention to the errors that could arise from the parallax (which is unavoidable), i.e. the sensor card should be as close to the face of the side plate as possible and the beam spot on the sensor card should be as close to the sentor of the camera sensor as possible. This was a tougher job than you think.

Anyway, in the first round of measurements, we convinced ourselves that the beam was:

off in -Y direction by 0.7mm relative to the nominal beam position on the input plate of the EOM,

off in +Y direction by 0.5mm on the output,

give or take 0.2mm or so (the error is based on two pictures for the input beam position with random variation in parallax coming from camera position and the distance between the side plate surface and the viewer card).

We rotated the entire EOM base by using two dog clamps against the EOM base and inserting appropriate shims (EOM_rotation.png). We didn't use the YAW adjustment feature for the EOM pivot plate because there's no way to rotate it in a  controlled manner.

After the first adjustment we thought that the beam coming out of the EOM looked better (which might have been false). On the second adjustment the beam looked the same or slightly worse (which might have been false) and we reverted back to the same position as the first adjustment.

According to the pictures (input_measured.jpg and output_measured.jpg), the beam position is 
0.2mm off in -Y direction on the input,
0.6mm off in -Y direction on the output,
again give or take 0.2mm or so. Apparently the input got much better, we overshot for the output but it's not worse.
 
This is the last measurement done based on the pictures. After this, we reset the dog clamps without shims and used them as the YAW reference.
Images attached to this comment
keita.kawabe@LIGO.ORG - 15:55, Saturday 07 February 2026 (89076)INS, ISC, SYS

Multiple beams mostly in PIT coming out of EOM (pictures and history)

1W into HAM1, otherwise it's hard to photograph these clearly.

The first picture is right after the YAW adjustment was made but before adjusting PIT. The card is held just ABOVE the main beam, you can see four blobs that look like some kind of ghost beams. (If you try to picture the main beam, it's so bright these ghosts become hard to capture.)

The second picture is after the first PIT adjustment. You can only see maybe two blobs, but later we found that the rest went below the main beam (sorry no "below" picture).

So, to recap the history of the beam quality,

  1. Feb/03-04 (alog 89041).
    • The beam had horisontal line that was clearly visible even at 100mW input (horizontalstreak.jpg).
    • JAC beam was found to be higher than 1mm relative to the EOM, the latter was raised, the horizontal line was gone but I noted "but there was still a vertical streak that was hard to photograph".
    • Then the EOM was adjusted in YAW, I still noted "not sure if it's great though, it's hard to photograph but something faint might be coming out of the EOM". But note that the power was 100mW.
  2. Feb/05 (alog 89059)
    • With 1W input, Jason and Masayuki clearly saw multiple ghost beams in PIT above the main beam.
  3. Feb/06 (this alog)
    • EOM was adjusted in YAW, seemingly no change i.e.there were still multiple ghost beams in PIT.
    • Adjusting the EOM PIT angle made ghost beams go up and down relative to the main beam, but they didn't seem to go away.

Other things.

Just to make sure, we turned down the 9MHz and 45MHz RF power to 3dBm and disconnected the 118MHz and 24MHz cables and nothing changed.

We know that the crystal wedge is supposed to be horizontal and we know that the wedge orientation is correct. When we first installed the EOM in chamber, the EOM transmission was deflected horizontally in +Y direction.

Images attached to this comment