Reports 1-1 of 1 Clear search Modify search
AOS (Beam Reducing Telescopes)
tomotada.akutsu - 21:43 Thursday 28 November 2019 (11827) Print this report
Strange coupling at TMSX IR QPDs

Kokeyama, Akutsu, (and Ushiba remotely)

Summary: we have no idea!!! crying

What we did: As reported at 11802, a strange coupling between "pitch" and" yaw", which are actually "vertical" and "horizontal" was observed. So today we came to see what was going on at the TMSX IR QPDs. During our check working, either ITMX or ETMX was (mainly ITMX, as it seemed larger signals at the QPDs) shaken at 0.1 Hz as like before. ITMX: 40cnt, ETMX: 400cnt, the difference is due to the high/low power of the coil driver.

 

  1. We partly opened the aluminum-foil shroud (wind sheild) for the IR path on the optical table to see if there were any suspicious ghost beams into the QPDs; but not found.
  2. Then, we touched aounrd the aluminum foil by expecting such touching would vary the coupling, but nothing changed.
  3. We also checked if the IR beam spot was clipped by optics but it seemed no such apparent clipping.

Additonal facts:

 

  1. There is a filter to cut the green beam from the IR beam before the QPDs; once we removed it but the situation did not change, so we got it back.
  2. On top of that, we can do centering of the IR beams on the QPDs by moving the micrometers on the QPD stages. So the QPDs seemed ok.
  3. Suspecting some electronics, we swapped the cables at the QPD driver between IR QPD2 and Gr QPD2 but nothing strange happend; the response of this swap was just as expected. So we (surely) get the cable connection back.

 

Discussion: so far, we have no idea. I (Akutsu) personally think of opening the GV beween the TMSX chamber and the ETMX chamber (EXC chamber), as the window on the GV is also clipping the light beam. Moreover, as observed at 11802, the green beams at the TMSX did not show this coupling. This was also confirmed as well today.

Hmm...

 

As we detemined to keep opening the aluminum-foil shroud, alternatively, we turned off the room light on the TMSX optical table. On the switch of the room light, we found a tape saying "Always turning on!" (see Fig), but we turned off at 18:55 after consulting with Miyoki-san. 

Images attached to this report
Comments to this report:
tomotada.akutsu - 4:05 Friday 29 November 2019 (11832) Print this report

Abstract

I calculated force-diagonalization coefficient for the observed strange coupled pitch and yaw (or, again, vertical and horizontal) signals in the two QPDs at TMSX. Here "force" means that I'm not sure the phyical interpretation yet. The data and the time used for this calculation is listed in the end of this post.

What I did

What I did is to do the following simple arithmetic. First, I download the time-series signals from every segement of QPD1 and 2. Then, compose simple pitch and yaw signals by

  • Simple_Pitch := (SEG4+SEG3) - (SEG1+SEG2)
  • Simple_Yaw := (SEG4+SEG1) - (SEG2+SEG3)

Then, compose modified pitch and yaw signals as follows

  • Mod_Pitch := Simple_Yaw + A * Simple_Pitch
  • Mod_Yaw := B * Simple_Yaw + Simple_Pitch

And then determined the A and B by looking at spectra of them to zero the corresponding peaks; this time, 0.1Hz excitation to P and Y of ITMX were used for the marker peaks, as described in 11827.

  • For QPD1: A = 0.346, B= 0.487
  • For QPD2: A= -0.916, B= -0.5535

And the actual time series and the spectra are attached; Fig 1 and 2 are for QPD1 with pitch excitation, and the spectra in Fig 2 were estimated from the data shown in Fig 1. In the acquired data, there were several times locklosses, so I selected "suitable" time window. As shown in the both figures, the 0.1 Hz fluctuation in Modified_Yaw disappears. Fig 3 and 4 are for QPD1 with yaw excitation. And the similar manner stands for the remaining Figures.

Discussion

  • When I plot the excitation signals as well, I can find some phase delay's difference for each case, and it is interesting to see.
  • Note that I did not divid the ouputs by the sum, but the important behaviors were not smeared out.
  • As shown in above, the coupling coefficient A and B have the opposite signs between QPD1 and 2.
  • Especially, QPD2's coefficient "A" is almost -1, which means Mod_Pitch2 = Simple_Yaw - Simple_Pitch = 2 (SEG1 - SEG3); those segments are aligned in diagonal... it is strange. Rotated in 45 deg?? But Mod_Yaw2 is not (SEG4 - SEG2)...Software is ok?
  • What's else???

----------------------------------

Used data

  • To find the excitation --> K1:VIS-ITMX_MN_LOCK_{P,Y}_EXCMON
  • And the conformation --> K1:VIS-ITMX_MN_DRIVEALIGN_{P,Y}_OUTMON
  • Signal channels --> K1:AOS-TMSX_IR_QPDA{1,2}_SEG{1,2,3,4}_OUT16
  • start time = Nov 28 2019 7:00:00 UTC (16:00 JST)
  • stop time = Nov 28 2019 10:00:00 UTC (19:00 JST)
  • GPStime = (Start = 1258959618.0, Stop = 1258970418.0)
  • Pitch excitation: relative GPS time (Start to be 0): 7200 to 8500 sec
  • Yaw excitation: relative GPS time (Start to be 0): 8700 to 9000 sec
Images attached to this comment
keiko.kokeyama - 9:37 Friday 29 November 2019 (11837) Print this report

Some questions:

  • Is that equivalent to rotate the four segments?
  • The last half of the figures don't show good separation between p and y. Maybe we should try a rotation matrix to apply the four segments?
  • If we are not DQing each segment, we should do so.
tomotada.akutsu - 12:33 Friday 29 November 2019 (11846) Print this report

It seems not simple rotations... Further confusion might happen if we did the same thing for exciting ETMX. Note that I only used the results for exciting ITMX.

Unfortunately yesterday the time duration for exciting ITMX yaw was shorter than that for pitch, so the resolution of the spectra are different. Anyway those arithmetics+experiment can be just done at the control room; let's see. smiley

tomotada.akutsu - 12:14 Saturday 30 November 2019 (11873) Print this report

Oh mitstake!

[Wrong]

  • Mod_Pitch := Simple_Yaw + A * Simple_Pitch
  • Mod_Yaw := B * Simple_Yaw + Simple_Pitch

[Correct]

  • Mod_Yaw := Simple_Yaw + A * Simple_Pitch
  • Mod_Pitch := B * Simple_Yaw + Simple_Pitch
Search Help
×

Warning

×