AS2/AS3/BHD BS/LO3/LO4 moved to BHD cleanroom

Attachment 1 shows the labeled optics on the ITMY table (from Koji). I recorded the initial e-bubble reading on the ITMY table: (x,y) = (0.03; 0.17). See Attachment 5 for orientation of the e-bubble (from chamber door: x-axis left to right; y-axis front to back).

I first cleared the extra steering mirrors (marked SM in Attachment 1). The e-bubble reading after clearing steering mirrors was: (x,y) = (0.03; 0.18). Note that fluctuations of 0.01 are common for the e-bubbles, so this is a negligible change. The steering mirrors were placed on the XEND flow bench [Attachment 2].

Next I removed the BHD optics, in the order of (LO4, LO3, BHD BS, AS3, AS2). This was arbitrary but made it easy to work front-to-back and remember reverse numeric order. These were placed on the BHD cleanroom table behind the OMC platform [Attachments 3, 4]. See attachments for labeled optics. 

The final e-bubble reading was: (x,y) = (0.00; 0.20). Similar (small) change in x and y directions. Final state of ITMY table in Attachment 5.

Continued on the BHD Optics Prep:

The following mounts were prepared (they are not 100% tightened yet)
- OMC Refl Mirror Mounts: OMCiRj (i=1,2, j=1,2,3)
- HWP (fixed) -> a thread adapter is missing
- TFPs
- HWP Actuator

We'll extract LO3/LO4/AS2/AS3/BHDBS from the ITMY chamber.

Fixed HWP mount had very thin room at the center. I wasn't sure if I could mount the 1/2" HWP in it. Fortunately the HWP was quite thin (~1mm) and it just barely fit. (Attachments 2/3)

The fixed HWP mount had no 8-32 to 1/4-20 thread adapter.
I'm asking the C&B of the adapters. (Attachment 4)

- AS2 2" post is being C&Bed
- LO4 2.5" post is being C&Bed
- BHDBS spacer ring is coming back from C&B soon
- BHDBS 2.5" post is being C&Bed
- Faraday Rotater Base was ordered. Waiting for the delivery.
 

[Paco, Vittoria, Murtaza, Radhika]

LO/POP alignment

We opened the ITMX and BS chambers for this work. First, both ITMs were misaligned to isolate the forward-propagating beam into the IFO. We saw that the transmitted beam through PR2 (LO) was clipping on the small POP mirror in front of LO1 [Attachment 1]. We decided to move that mirror out of the way and focus on aligning LO; then we would proceed to align POP.

Once the small mirror was moved, the LO beam looked fairly centered on LO1. At LO2, the beam was very misaligned in yaw, missing the optic [faint beam spot in Attachment 2]. I steered LO1 until the beamspot at LO2 was about centered [Attachment 3]. At this point, the beam was reflected off LO2 towards the ITMY table. 

I went back to check that the LO beam was not clipping anywhere on the ITMX table on its way to LO2 [Attachment 4].

Next I removed the spacer that was raising the height of the small mirror [Attachment 5]. ITMY was realigned so that we would recover the backwards-propagating reflection from the IFO. I placed the small mirror back to reflect the backwards-propagating beam transmitted through PR2 (POP) [far left beam in Attachment 6]. I verified that POP was hitting the next mirror and directed towards the in-air ITMX table [Attachment 7]. We closed up all chambers.

Next steps

Fine alignment of LO beam downstream of LO2 and and of POP out of the chamber will be done after PR2 is replaced.

Lab temperature trend for 20 and 400 days:

- The lab temp significantly raised by 4 degC at Vertex and 6 degC at Xend. Yend saw no change.
  We have never seen such temp rise in the past 400days.

- The PSL temp went up at some point in the summer but we should check if this is real or any artifact.

Horizontal beam centering along XARM

I tried to improve beam centering along XARM from the control room. First I realigned IMC until counts were recovered on MC TRANS QPD. The final IMC alignment state resulted in 960 cts transmission [Attachment 1].

Next I aligned the arms until transmission in both arms reached ~0.08 [Attachment 2]. In that state I recorded ETM/ITM beam spot positions below. Note that although the beam was miscentered by over 5mm on ETMX (horizontal), it was under 1mm on ITMX. Thus it seemed like a pointing issue. I then tried to lock MICH to record the beam spot on BS, but it never locked stably and the measurements were bogus.

Optic    LSCDoF     freq.(Hz)    ampl. (counts)    gpstime    Opt. gain (counts/nm)    Opt. gain_std    v (mm)    v_std      h (mm)    h_std

ETMY    YARM    211.11    500    1385168096    112.24    4.68           -1.25   0.47  -1.48   0.41
ITMY    YARM    211.11    500    1385168173    100.55    3.18            3.39   0.74   1.71   0.40
ETMX    XARM    211.11    500    1385168993    70.39     4.36           -2.92   0.60   8.39   1.56
ITMX    XARM    211.11    500    1385169082    79.36     3.17            2.90   0.60  -1.64   0.85

I proceeded to try to walk the beam a bit along XARM by moving TT2 in yaw and compensating with PR2. I moved both by ~40 steps and XARM transmission improved to just under 0.1 [Attachment 3].  However, the beam spot measurements looked pretty much the same; mode-matching into XARM was improved while pointing stayed constant. Next I locked PRY to measure the final beam spots on PRM/PR2/PR3.

Optic    LSCDoF     freq.(Hz)    ampl. (counts)    gpstime    Opt. gain (counts/nm)    Opt. gain_std    v (mm)    v_std      h (mm)      h_std

ETMY    YARM    211.11    500    1385234260    133.31    6.03            2.28   0.75   -0.74   0.43
ITMY    YARM    211.11    500    1385234343    121.44    3.14            3.91   0.52   -0.82   0.43
ETMX    XARM    211.11    500    1385234433    88.98     3.75           -0.76   0.41    6.16   1.18
ITMX    XARM    211.11    500    1385234546    103.35    4.50            3.60   0.67   -0.90   0.23
PRM     PRY    211.11    1000    1385236336    5.57      0.73           -8.13   16.67  -2.75   2.38
PR2     PRY    211.11    1000    1385236409    8.65      1.88            2.17   0.54   -7.71   3.39
PR3     PRY    211.11    1000    1385236483    6.39      1.74            4.83   1.50    9.17   6.25

With such high transmission in the arm cavities, alignment looked good and we decided it was OK proceed with aligning the LO/POP beams.

The OMC was locked with Moku Pro.

Attachment 1: Electrical setup. The RF part of the REFL PD signal was fed into Moku pro, while the DC part was monitored on a scope.
Attachment 2: Servo setup. The modulation amplitude is 100mV.
Attachment 3: Image rejection LPF setup
Attachment 4: Laser PZT servo during lock acquisition
Attachment 5: Laser PZT servo for stational operation
Attachment 6: Laser Temp servo setting
Attachment 7: CCD Images during lock. The REFL is still limited by the mode mismatching component.
Attachments 8/9: The REFL locked / unlocked = 340mV/5.4V = 0.06 --> Mode Matching 94%
 

- Setup the fiber protection and stray beam shields
- Replaced brand new wedged Y1-45P mirrors used in the OMC input path with stock Y1-45S mirrors. The wedged Y1-45Ps are going to be used for the optics on the BHD platform.
  Note: 45S mirrors are the same as 45Ps, but just their coating precision was not enough for 45P. So most of the cases 45S is sufficient for 45P purpose.
- All the remaining mirrors / CCDs / a lense / an RF PD (PDA-10) were setup and aligned.

- Found a DS345 is broken. The sinusoidal output does not swing and only has weird constant offsets dependent on the setting ==> need fix.

- An old analog function generator was brought for laser PZT scan test.
- Fast PZT cavity scan test: fine alignment & prelim visibility: REFL PD DS unlocked 5.50V vs TEM00 230mV -> estimated mode matching 95~96%
- When the OMC was detached from the kinematic mount and returned to the position: Relection at TEM00 230mV -> 1V. This corresponds to ~15% degradation of the alignment.
- OMC retainer screws (white PTFE screws) to hold the OMC vertically deforms the breadboard and cause the misalignment. Use something compliant or just release the screws.

Vertex is extremely hot. I'm reluctant to go into the HEPA with clean garbings....

It seems that it started about a week ago. So this is not related to the holiday.
(Although I don't know where the sensors are)

• PDR1V rotation stage was found
• 2" pedestal post x Qty 1 / top and bottom vent holes machined -> to be C&Bed
• 2.5" pedestal post x Qty 2 / top and bottom vent holes machined -> to be C&Bed

[Yuta, Radhika]

PRY locked and PRM/PR2/PR3 beam spot positions measured

We aligned the arm cavities and locked at low power, then measured beam spot positions on ITMs and ETMs to verify nominal centering.

Moving onto MICH locking, we noticed what could be AS beam clipping in pitch. We struggled to lock to dark fringe and measure a reasonable beam spot position, so we decided to move onto PRY since we already had a BS spot measurement.

We restored PRM alignment and brought it back to rougly center of the camera. PRCL started to flash and we aligned PRM to until REFLDC flashed down from 1.5 to 0.3. Next ITMX was misaligned by 1500 cts so that it wouldn't contribute to AS fringing. REFLDC flashed down to 1.3 cts at this point, and we were able to lock PRY [Attachment 1] with REFL55 demod phase: 166.02; PRCL gain: -0.4.

Beam spot positions for PRM/PR2/PR3 are below and plotted in Attachments 2, 3. Attachments 4,5 are reposts of the beam spots for ITMs and ETMs. Overall, vertical mis-centering was generally reduced across optics (except for a slight overcorrection in PRM and worsening in the XARM). Horizontal mis-centering was generally made worse in most optics, while reduced in PR3 (magnitude). We plan to open the chambers to verify by eye that the beamspots are where we expect, and we may have to make some adjustments according to the measurements. It could be that the references used at PRM/PR2 were not actually reliable, e.g. oplev beam height. 

[Yuta, Vittoria, Radhika]

PRMI flashing recovered; haven't yet been able to lock PRY

This morning the recovered REFL beam was steered onto REFL55 to maximize the REFLDC signal to 1.4 (consistent with around a factor of 10 laser power reduction).

Since PRMs alignment was changed a lot to recover REFL beam, we noticed that light was no longer hitting the PRM oplev QPD. was no longer detecting any light. We moved onto recovering PRMI flashing:

- Radhika and Yuta moved PRM around until they saw flashing at the antisymmetric port

- REFL DC read 1.4 counts, and then PRM was aligned to drop it to 0.3 counts

- While in this state of good alignment, Radhika and Vittoria went inside and centered the REFL beam on the camera

- We checked that the REFL beam was centered on the REFL RF PDs

- Then we moved on to OPLEV centering for PRM

- Then we identified the PRM QPD and couldn't find the beam

- We traced back the beam and found that it was hitting the bottom, plastic part of an optic

- We adjusted one of the steering mirrors of the input OPLEV beam so that it hit the mirror it was initially clipping on

- Then we moved around two steering mirrors to center the OPLEV.

I went through the inventory list and the parts in hand to figure out what was still missing

The list was made on E2200464, and the PDF snapshot was added to this entry.

Resolved

• 2" pedestal post with a bottom vent hole (QTY 1+spare 1) 1 in hand / to be drilled / to be C&Bed
• 2.5" pedestal post with a bottom vent hole (QTY2+spare 1) 2 in hand / to be drilled / to be C&Bed
• Thorlabs NB1-K14 mirror (QTY1 + spares 3?) 2 in hand
• Need to find 5 Y1-1025-45P collected various 45P mirrors. In principle we have all.
• Faraday Rotator Riser D2200378 (Already ordered by Don) To be delivered / To be C&Bed
• D2200122 ring spacers (Coming in soon from Caltech machining shop)
  -> They are ready for C&B on Nov 16 (JC)
• Polaris-K1-2AH on the flow bench in the C&B lab. 

Unresolved

• PDR1V rotation stage -> in search

I took over the IFO after Radhika left. She found the REFL beam earlier and the PRM alignment offset for the refl beam on the refl camera was (P,Y)=(469,2558).
When the REFL11I responce is maximized to be -6 (presumably because of the RF residual AM), (P,Y) was (379,2548).
When the PRMI fringe like feature shows up in the AS camera, (P,Y)=(429,2028).
I left the PRM at (469,2558), so that the REFL spot is obvious on the REFL camera.

The PSL shutter was closed at the end of the work.

Today Yuta and I opened the BS chamber and found that the REFL beam was misaligned in pitch and yaw [Attachment 1]. We steered PRM until REFL and input beams overlapped [Attachment 2]. From there, we moved PRM around until we recovered the REFL beam on the camera. This means bright REFL (prompt reflection from PRM) is aligned onto the AP table. Now some steering mirror alignment on the AP table should recover REFLDC/REFL55 signal response.

Task Completed Today :

• D2200122 Spacer Rings have been machined and ready for C&B
• The North siderail of the cleanroom has been raised.
• Parts SM05RR, LMR1V, D2300352-01, -02, -03, -04 have been C&Bed.
• A New Cart has been Cleaned and added for working purposes to the cleanroom

D2200122 Spacer Rings have been machined and ready for C&B

I walked over to the GALCIT Shop today after receiving an email notifying me that the Spacer Rings are ready. We received 5 rings instead of 4 because the machinist made a mistake on one and was a couple thousanths of an in thicker. This ring is marked red on the photo in attachment #1. I have given these rings to Maty and she will take care of the C&B process since I will be out tomorrow. She said she expect them to be done sometime early next week.

The North siderail of the cleanroom has been raised.

I raised the North Siderail of the cleanroom to give us some more freedom moving the steel cart around in the cleanroom. This also allowed me to rotate the cart and add a bit more space inside. While doing this I had the particle counter running and 0.5 Micron count < 50 and the 0.5 Micron count < 10.

Parts SM05RR, LMR1V, D2300352-01, -02, -03, -04, and D2300208, D2300209, D2300210 have been C&Bed.

SM05RR, LMR1V, D2300352-01, -02, -03, -04, D2300208, D2300209, D2300210 are ready as of today. These have been placed in a plastic box on the middle space of the newly added steel cart. The Plastic boxes have the DCC/Part number on top so you will know what parts are inside.

A New Cart has been Cleaned and added for working purposes to the cleanroom

I brought the steel cart over from the shed area. I thoroughly wiped it down with IPA after cleaning with Windex and a rag. I also used the air gun to blow of any heavy dust from the wheels. I placed foil over each level of the cart and wiped down again with IPA. After, I made some AntiStat curtain around the cart to keep out anything heavy. Here are the plastic boxes containing parts SM05RR, LMR1V, D2300352-01, -02, -03, -04, D2300208, D2300209, D2300210.

[Paco, Vittoria, Radhika]

Yesterday and today I noticed that REFLDC/REFL RF PDs sense verry little light, under 0.02 counts [Attachment 1]. We opened the table with REFL PDs and observed the beam, trying to steer it onto REFL55. No increase in counts was observed. The optical response of REFL55 was verified by shining a flashlight on it (reached 2 counts). We tried touching the RF PD box and wires to ground them in case there was some weird electrical response. The counts did not increase on the PD. 

We then borrowed the power meter from the PSL table to measure the REFL beam power in watts. Oddly, measuring the beam as upstream as possible (before all the beamsplitters) only recorded 5 mW of light. This was bizarre, since by eye the beam is quite bright on the beam card. Such low power would explain the PDs not picking up the light; Koji theorized that it might be a ghost beam and the actual REFL path might be blocked inside the vacuum volume. On the other hand, PRM alignment shouldn't have changed that much to result in a misdirected REFL beam. We will investigate tomorrow.

Radhika reported that the PRM UL OSEM PD is not responding. This PD has been identified to have a shorting problem, but the short existed only at the bias pin of the PD. We disconnected the bias voltages and the PD was working with no (=0V) bias.

It seems that it lost the signal about 8 days ago and the signal intermittently appeared and disappeared.

I suggested to Radhika to remove the Al foil suspecting the other pin of the PD is not shorting.

[Koji, Paco, JC]

The door of the Input Chamber Has Been Removed



We removed the chamber door off of the input chamber. Koji and Paco worked on removed the bolts while I was preparing the crane for use. When removing bolts from the chamber doors, we were careful not to remove them all at once, Koji decided to leave 4 loose bolts remaining until the crane has a tight lift on the door. 

As for the crane, to remove this door, we needed to extend the swing arm out. The crane itself JUST BARELY makes it over to where the hook can hover nicely over the door. Next, I lifted the door until the sling was nice and tight. Then Koji process to remove the 4 bolts holding the door in place. We lifted the door all the way up and slowly swung the arm to bring the door by ITMX Chamber. Here, we lowered the door and placed it flat onto a Piano Dolley. The door is now wrapped in AntiStat and laying flat onon a piano Dolley by the ITMX Chamber. 

[Murtaza, Paco, Radhika]

After the input beam path alignment was complete (arm flashing recovered), we attempted to run Yuta's beam centering measurement script (scripts/ASC/measureALLBeamSpotPosition.py). I was able to align and lock the arms at low power following the procedure here. The resulting ITM/ETM beam mis-centering measurements can be seen below.

Next I struggled to lock MICH to measure mis-centering on the BS using Yuta's instructions. We noticed the AS beam was clipping horizontally. Paco and I aligned the AS beam using SR2/AS1 and the beam now appears unclipped. We found the demod angle that minimized AS55_I to be 86.56 deg, not far off from Yuta's value. We mananged to lock MICH with: 1. ASDC trigger matrix element: -30; trigger threshold lowered to -0.5 (enable). We then obtained the BS mis-centering measurement below. Note that the uncertainties here are on the order of the mis-centering value.

Attachments 1 and 2 compare the vertical and horizontal beam mis-centering values before and after the in-chamber alignment. In summary, beam spots on ETMs and ITMs have more or less converged towards center vertically (BS is the exception). However, the beam spots seem to have diverged horizontally from center. This means our pitch corrections were in the right direction, but we'll need to take measurements on PRM/PR2/PR3 to confirm this. It seems like trying to mitigate horizontal clipping at PR3 has propagated mis-centering downstream.

Measuring the beam spot on PRM/PR2/PR3 requires PRY locking, but it seems that little to no light is hitting REFLDC or any of the RF REFL PDs. This will be debugged next.

I checked how FB1 is and it has no alerting red LED anymore! Great

But I still can hear beeping sound. It can be from the JETSTOR intelligent storage connected to megatron, or megatron itself which has a yellow alering LED on 🤮

Megatron is Sun Microsystems Sun Fire X4600. The function of the yellow LED is "Service Action Required LED". But the manual says: "Slow Blinking: An event that requires a service action has been detected", while the LED is solid yellow in our case.
https://docs.oracle.com/cd/E19121-01/sf.x4600/819-4342-18/html/z40007ed1015419.html

[KA: Copied from JC's entry on Mon Nov 13 14:15:40 2023]

I replaced the middle of the 3 fans.

I received the fans today and went ahead to replace the fan. Koji mentioned that the middle of the 3 fans was dead. in elog 17962.

For some odd reason, when I took off the top to frame all 3 fans were spinning normally. Maybe it was a one time failure for the fan ? Well, either way, I changed the fan so that we could check the fan replacement off the list if this issue happens again.

When replacing the fan, I reused the same connector from the previous one. I just cut the cables and soldered the cables together to the new fan. I attached the new fan to the plastic mount and connected it to the system. Please note that the other 2 fan which are remaining should pick up speed when any of the 3 fans are removed. Once I connected the fan and all 3 began running, the speed drops back down. After placing the fan, I put a label of today's date.

To Do:

• Set up the OMC refl optical path and the refl PD
• Set up a trans monitor PD at CM1
• Set up a platform position marker to ensure the reproducibility
• Moku setup:
  • Temperature sweep
  • Laser PZT sweep
  • Moku RF modulation demodulation setup for PDH locking
     

What we need more to make the work better/easier in/around the HEPA table:

• Proper fiber protection
• Proper stray beam blocks (anodized Al plates)
• Electronics rack or shelves at the north side of the booth
• Parts table at the south side of the booth
• Rotate the table 180 deg and put a storage shelf beneath the table (Wire rack?)

EP30-2 Kit + Bonding Kit (= Two bottles of silica powder) were returned to Modal Lab @ Downs.

The polarization and alignment of the fiber for the OMC setup were adjusted. The polarization ratio before the PBS was 50:1. Then, the P-pol was sent to the OMC via two steering mirrors.
As a result of the beam alignment, the OMC cavity is flashing with a good amount of occasional TEM00.

Fiber alignment and polarization refinement

- NPRO power ADJ was "-29". The initial fiber output was 6mW.
- The input fiber collimator and an input steering mirror were adjusted to maximize the fiber output. The output power increased to 12mW.

- Adjusted the output fiber mount to minimize the PBS reflection.
- The initial ratio of the P/S was checked. 3mW was reflected by a PBS out of 12W. (i.e. 3:1).
- Went to the PSL table and repeated 1) rotate the fiber coupler 2) maximize the input beam/coupler alignment.
- Again, adjusted the output fiber mount to minimize the PBS reflection.
- Went back to the PSL to repeat the input side adjustment.
- Determined that I could not do it better.

- Increased the NPRO power ADJ to -20, just to have more power.
  Polarization ratio was ~50:1 (Trans 43mW, Refl 0.82mW).

The beam alignment into the OMC
- Set up the steering mirrors.
- Align the input beam so that I can see the input spot on the center of the second curved mirror (CM2).
- If the alignment is perfect, the input beam should hit the center of the first cavity mirror (FM1)
- The deviation can be adjusted by the mirror/spot position on the last steering mirror.
- So: Adjust CM2 spot by the last steering mirror, Adjust FM1 spot by the penultimate steering mirror.
- This made the cavity flashing. After a bit of alignment, the TEM00 mode was visible.
- A CCD was set at the transmission of CM2

Nice recovery of the lock!

Do you think the spot distance at PR2 makes sense if we consider the dispersion?

https://wiki-40m.ligo.caltech.edu/Optical_Layout#A.2BJbo-Effect_of_the_wedge

[Murtaza, Paco, Shoki, Radhika]

IR arm cavity flashing recovered after input optic beam centering - arms locked

Today we picked up from yesterday's work aligning input optics. First we confirmed that the beam spots were still centered vertically after yesterday's progress. Today's goal was to recover IR horizontal alignment on PR3 and BS, and further downstream. Since the PR3 suspension was translated towards the edge of the table (GTRY now unclipped), we now had to consider the risk of the PR2-PR3 path clipping on the bellows of the vacuum tube. 

1. Visually observed IR beam hitting edge of PR3

2. Adjusted PR2 yaw to shift IR beamspot on PR3 towards green spot reference.

3. Tested limits of PR2 yaw adjustment - scanned until beam was clipped by bellows (video of this in google photos album).

4. Once the IR and green spots were aligned on PR3, we tried to steer PR3 to hit BS. We maxed out the range of the CDS offsets, so physically rotating PR3 proved necessary.

5. Paco rotated PR3 and recovered its alignment. Once the IR beam was hitting the BS, fine tuning was performed to ensure the GTRY reference spot and IR coincided at each optic [Attachment 1 at BS; Attachment 2 at PR3; Attachment 3 at PR2 - note distance between green and IR].

Arm cavity flashing was recovered and the arms were locked at low power [Attachment 4].

[Murtaza, Tomo, Paco, Radhika]

LInk to google photos album for today.

VERTICAL BEAM CENTERING

We first started with vertical beam spot centering. To recap, the IR beamspot started off being high on TT2/PRM and PR2 (see Attachment 1 for starting spot position at TT2). We wanted to vertically center each beamspot while monitoring the spot position at the BS (using the YAUX green spot as a reference.

Vertical alignment steps (all centering below is implied vertical):

1. Opened BS chamber

2. Adjusted TT1 to lower spot position on TT2 (lowered every downstream beamspot)

3. Adjusted PR2 to correct for TT1 motion, keeping beam spot at BS fixed (relative to green reference)

4. Iteratively performed steps 2 and 3; realized our step sizes were too small and that we should work sequentially downstream

5. Opened the ITMX chamber and saw spot position was centered on PR2

5. Moved TT1 and TT2 in differential mode until beam spot was centered on TT2; hoping to keep PR2 centering [Attachment 3]

6. Needed to compensate by moving TT2 to center on beam on PR2 [Attachment 4]

7. Adjusted PR2 to center beam on PR3/BS (relative to green reference) [Attachment 5]

HORIZONTAL BEAM CENTERING

To correct the clipping of GTRY on PR3, we needed to shift PR3 along its axis. 

1. There was a counterweight on the table blocking PR3 from being shifted longitudinally (towards table edge). Paco moved a cable clamp to make room to slide the counterweight and shift PR3.

NOTE: THE COUNTER BALANCE NOW SITS ON THE EDGE OF THE TABLE WITHOUT A CLAMP. BE CAREFUL WHEN THE BS CHAMBER IS OPENED.

3. Shifting PR3 affected GTRY reaching the camera. Steering mirrors were adjusted to realign the green beam with the camera. One of the GTRY steering mirrors was shifted to prevent it from blocking the IR beam between TT1-TT2.

4. We lost the beamspot on BS after moving PR3. We'll start with moving PR2-PR3 in YAW until the IR is pointed correctly on the BS tomorrow.

CLOSEOUT

1. To wrap up for the day we realigned the arms using the AUX lasers as a reference. With this configuration, the oplevs on the test masses were centered to use as a reference point [Attachment 2].

[Murtaza, Tomo, Radhika]

In preparation for alignment of vertex optics, went through the exercise of acquiring low-power IR lock of the arm cavities. Arm transmission ndscope in Attachment 1 (max transmission 0.1).

LSC settings and oplev positions in Attachment 2.

Tomorrow morning we plan to complete the alignment work.

I borrowed a small accelerometer from 40m and walked it to the cryo lab. It is pictured in attachment 2 in its box.

There are two accelerometers I can see on the table in the cryo lab. They are both pictured in attachment 1

Here are some beam centering strategies...

Beam height around PRC:
 - From our in-vac inspection today, we noticed that the beam is high on TT2 and PRM, and was slightly clipped (or maybe not) by the earthquake stop of TT2.
 - According to the beam spot position measurements (Attachment #1; includes measurements taken today), beam spot on PRM and PR2 are vertically mis-centered by +1 cm, but vertically pretty centered on PR3, BS and ITMs.
 - These imply we need to play with TT1 and PR2 to level the beam inside the PRC.
 - This is consistent with the height issue we found during the last vent on POP_SM4 (1 inch mirror in front of LO1). We had to place a spacer to make POP_SM4 high to extract the POP beam coming from PR3 to PR2 (40m/16832). So we should be able to remove the spacer after the fix.
 - This fix should be done iteratively without losing flashing in the arms, sacrificing BHD alignment.

Horizontal beam mis-centring on PR3:
 - According to the beam spot position measurements, beam is vertically centered, but horizontally mis-centered by ~1.5 cm at PR3.
 - This is consistent with beam clipping we have in Y green transmission in YAW (Y green is clipped in YAW at PR3).
 - Since the beam is horizontally centered on PRM and PR2, we probably need to physically move PR3 to center the beam and to un-clip Y green transmission.

Beam mis-centering around LO and AS:
 - After these fixes, we probably need to re-align AS path and LO path, as the beams are not centered on LO1, LO2, AS1, AS4.
 - Hopefully these will improve LO-AS mode matching...
 - These probably will not require physical moving of the suspensions.
 - AS RF path after AS2 (90:10 splitter) is also clipped in YAW. This issue also needs to be addressed when re-aligning AS beam.
 - It is probably better to install BHD platform first, before doing LO and AS beam alignment, as BHD platform changes the balancing of ITMY stack, which has AS1 and AS4 suspensions.

I made a script to untrip software watchdogs.

/opt/rtcds/caltech/c1/Git/40m/scripts/SUS/untripSoftWatchdog.py

It is available from the blue "Damp" button in the watchdog screen attached.

Happy in-vac work!

[Yuta, Vittoria, Radhika]

All photo documentation of the open chambers can be found on the 40m google photos account (foteee@gmail.com). Here are the albums for the ITMY, BS, ITMX tables.

We opened each chamber and captured photos of each suspension below. The photos on the google drive have the name of the suspension in the description.

The digital tilt-meters (henceforth called e-bubbles) were assigned a label for each table by Yuta. We zeroed their level readings on the PSL table, assuming it is the most level surface around. We placed the e-bubbles as centered as possible on each table keeping distance from the suspensions. *Note: we didn't think to align the e-bubble with the tapped holes of the breadboard, so next time we'll make sure to do this.* These readings can be seen in Attachments 1-3 (ITMY, BS, ITMX).

Each chamber was closed and nothing was moved from any of the tables.

Thanks to Murtaza, now we have two emergency switches for the aux laser. One is on the PSL table, and the other is in the staging HEPA booth.

The fiber was put in a plastic spiral tube for protection and delivered to the staging HEPA booth. I confirmed that the light was coming out from the fiber but haven't checked the output power.

[Koji, paco]

First, ssh to fb1 and running sudo dmidecode -t1 we get

# dmidecode 3.2
Getting SMBIOS data from sysfs.
SMBIOS 2.8 present.

Handle 0x0001, DMI type 1, 27 bytes
System Information
    Manufacturer: Supermicro
    Product Name: Super Server
    Version: 0123456789
    Serial Number: 0123456789
    UUID: 00000000-0000-0000-0000-0cc47a699734
    Wake-up Type: Power Switch
    SKU Number: To be filled by O.E.M.
    Family: To be filled by O.E.M.

A quick google search on supermicro superserver "red LED blink" suggests fan failure.
After pulling out fb1 we confirmed the middle of three fans is dead.
The fan is a NIDEC Ultraflo (V80E12BHA5-57), 12 VDC, 0.6 A, 80x80x5 mm (see Attachments #1/#2).

We left the FB1 pulled out, but the dead fan and the lid were left as they were to keep the current airflow condition.

• FB1 has this "i" LED blinking and continuously beeping. What's wrong? 
• Also, one of the UPS at the bottom of 1X6 has the error "F06".
• Speaking about UPSs, the control room UPS has a continuous red light and frequent beeping.

Beam spot position measurements were done for the first time in air pressure at low power.
Dither amplitudes were increased to overcome lower SNR, but still the uncertainties are large.

Results:

# Optic    LSCDoF     freq.(Hz)    ampl. (counts)    gpstime    Opt. gain (counts/nm)    Opt. gain_std    v (mm)    v_std    h (mm)    h_std
ETMY    YARM    211.11    1000    1383331863    64.55    3.78    -4.33    0.37    0.47    0.40
ITMY    YARM    211.11    1000    1383331936    65.15    2.67    -0.97    0.17    -0.20    0.11
ETMX    XARM    211.11    1000    1383331957    59.96    2.10    3.40    0.30    -1.33    0.18
ITMX    XARM    211.11    1000    1383332021    62.73    2.35    0.88    0.15    0.38    0.19
BS    MICH    211.11    3000    1383333747    0.02    0.00    3.21    3.90    -3.33    2.26
PRM    PRY    309.21    3000    1383335425    6.23    6.28    13.55    29.00    -2.05    2.82
PR2    PRY    309.21    3000    1383335499    7.57    4.51    7.70    7.09    3.67    12.17
PR3    PRY    309.21    3000    1383335562    5.88    1.17    3.60    10.33    -16.91    15.18
LO1    HPC    113.13    3000    1383336782    0.05    0.01    9.12    3.44    6.50    5.91
LO2    HPC    113.13    3000    1383336842    0.05    0.01    14.51    6.63    21.34    5.13
AS1    HPC    113.13    3000    1383336901    0.05    0.01    2.62    3.36    14.91    2.70
AS4    HPC    113.13    3000    1383336963    0.05    0.01    -5.84    3.28    14.56    3.47

Locking configurations:
 - YARM and XARM can be locked by increasing the trigger matrix C1:LSC-TRIG_MTRX elements by a factor of 10.
 - MICH can be locked by changing AS55 demodulation phase C1:LSC-AS55_PHASE_R from 2.1 deg to 92.1 deg (why?). Play with trigger matrix C1:LSC-TRIG_MTRX so that it does not trigger filters.
 - PRY can be locked by changing REFL55 demodulation phase C1:LSC-REFL55_PHASE_R from 76.02 deg to 166.02 deg (why?) and by increasing the PRCL gain by a factor of 10, C1:LSC-PRCL_GAIN from -0.04 to 0.4. REFL55 is glitchy and making the measurement uncertainties big. Play with trigger matrix C1:LSC-TRIG_MTRX so that it does not trigger filters.
 - ITMY single bounce-LO can be locked by increasing the gain by a factor of 10, C1:HPC-LO_PHASE_GAIN from 1.5 to 15.
 - These changes are summarized in the following script.

/opt/rtcds/caltech/c1/Git/40m/scripts/ASC/measureAllBeamSpotposition.py

Attachment #1 is the current PRFPMI BHD alignment, and Attachment #1 is the history of the measurements.

I retrieved the nitrogen pressure regulator from x arm vacuum cabinet and returned it to cryo. Reminder to replace the 40m regulator for the next vent.

[Paco, yuta]

We continued with the alignment recovery by looking for arm cavity flashing. To do this it was very helpful to restore the OpLev centering and use the AUX lasers -- After some further alignment on both arms, we just needed to increase the LSC gains (XARM and YARM) by a factor of ~ 1.5x and lower the TRX/TRY DC matrix elements for triggering from 1.0 to 10.0 in order to engage the IR locks (See Attachment #1).
Attachment #2 shows the screenshot of MEDM screens after both arms are aligned with low power mode.

Next:
 - Measure beam spots in low power mode
 - In-vac work to center the beams on vertex optics

[Paco, yuta]

We opened the PSL shutter and verified the power behind the shutter is 92 mW. Then, we changed the IMC autolocker parameter file to lower the transmission threshold from 5000 counts to 300 using the instructions from the wiki. We then slightly aligned MC2 in YAW and maximized the flashing in transmission and also changed the parameters for the inputGain slider by +20 dB in both unlocked and locked states so we compensate for the decrease in optical gain. The changes have been pushed here.

After these changes in the config file we ssh into optimus and restart the docker container running the autolocker script in this case by running:

sudo docker restart scripts_AL_MC_1

As we found that some of the FRGs were not well calibrated, Using the driver menu, I've recalibrated the FRGs exposed to 1atm right now.

I expected that the FRGs with the calibration drifted off shows lower or higher pressure like FRG5 which showed 1000 Torr at 1atm.
However, our main FRG1 did not show significant change from 760 Torr even when JC introduced more vent air this morning. This is annoying.

We need to check if the FRG1 is failing, or this is the expected behavior.

This morning, the goal was to open up the chamber doors. Yesterday we attempted, but we did not succeed. Here is a list of the possibilties what we were thinking at the end of yesterday:

• The annulus pressure was not yet at atmosphere and the door will remain stuck until then.
• The chamber doors are just sticking to the O-Ring and we need to knock it with a rubber mallet as said in elog 15649.
• The Main Volume Pressure is NOT at atmosphere and FRG1 AND p1a is showing incorrect pressures.

After speaking on the phone with Jordan, he recommended that we should open VV1 to room air and listen if air is entering the main volume. I did so and air began to rush in.

New watchdog and suspension summary MEDM screens were made, and IFO_ALIGN screen was updated.
We also implemented new misalign/restore scheme using C1:SUS-$(OPTIC)_BIASPIT_OFFSET instead of using C1:SUS-$(OPTIC)_TO_COIL OFFSETs in matrix elements.

Script for generating screens with multiple suspensions:
 - A script was made to help generating MEDM screens with multiple optics. A template adl file is used to place each optic in a defined grid.

/opt/rtcds/caltech/c1/Git/40m/scripts/medm/genAllSusMon/genAllSSusMon.py

New misalin/restore scheme:
 - Instead of turning on OFFSETS in C1:SUS-$(OPTIC)_TO_COIL matrices in pitch, C1:SUS-$(OPTIC)_BIASPIT_OFFSET is now used for misaligning.
 - The optic is considered to be misaligned if C1:SUS-$(OPTIC)_PIT_OFFSET and C1:SUS-$(OPTIC)_PIT_BIAS read different value. This can happen with BIAS(PIT|YAW)_OFFSET, C1:SUS-$(OPTIC)_(PIT|YAW)_OFFSET_ON being 0, C1:SUS-$(OPTIC)_BIAS(PIT|YAW)_GAIN being 0 etc.
 - When restoring, these channels are turned on in both pitch and yaw to make sure it is restored.
 - Updated script lives in

/opt/rtcds/caltech/c1/Git/40m/scripts/ifoAlign/sosAlign.py

New MEDM screens:
 - New C1SUS_WATCHDOGS.adl (Attachment #1), IFO_ALIGN.adl (Attachment #2; does not look new) and C1SUS_SUMMARY.adl (Attachment #3) are attached.
 - In the attached watchdog screen,
   - LO1 is aligned but tripped, not ready to damp because of PD RMS values are larger than MAX_VAR. C1:SUS-LO1_LATCH_OFF=1 and C1:SUS-LO1_LOGIC=1.
   - LO2 is aligned and tripped, but ready to damp. C1:SUS-LO2_LATCH_OFF=1 (this being 1 means it is tripped and not untripped) and C1:SUS-LO2_LOGIC=0 (ready to damp).
   - AS1 is aligned and tripped, but ready to damp. C1:SUS-AS1_LATCH_OFF=0 (this being zero means someone untripped, but does not mean any other) and C1:SUS-AS1_LOGIC=0.
   - AS4 is misaligned and damped. Misaligned meaning (C1:SUS-AS4_PIT_OFFSET != C1:SUS-AS4_PIT_BIAS) || (C1:SUS-AS4_YAW_OFFSET != C1:SUS-AS4_YAW_BIAS).
   - C1:SUS-$(OPTIC)_LATCH_OFF cannot be turned in this screen to show that it does not mean much.
 - For the suspension summary screen, I intentionally kicked AS4 to make PDVARs and alignment PIT offset red. Note that screenshot was taken after the vent, and some oplev values are also red.

After swappiong out the cylinders, I went to review the pressure to check the rate of the N2, but It looks like we already hit atmospheric pressure?

I'm assuming the room air must have entered the system some how while I was swapping the cylinders. I must've have made a mistake, but I don't understand how room air could've entered while the main volume was isolated during the swap. Could somebody review the steps I took and let me know where I made a mistake? If there is no issue with the procedures, I must have made the mistake and not have noticed. There was a steep rise in the pressure after 200 Torr.

[Murtaza, JC]

We started preperation of the vent at 7:30 am.

First, to avoid room air entering the system, Murtaza and I purged the line to see if there were any leaks. It turns out that the regulator was the issue! We used soap and water when purging the line to check if we could visibly see an air bubbles, but none were apparent. Yet, there was still a steady hissing somewhere and couldn't figure out where it was coming for. It turned out to be from the control knob of the regulator. I pulled out another regulator from the Vac Cabinet on the X-End, and it worked out perfect.

At 8:30

We finished the leak hunting and moved to begin backfilling the IFO with N2. We are letting in 10 PSI of pressure right now and the Vacuum pressure is steadily increasing approximately 0.02 Torr every 10 min. We are doing great so far

We noticed that the alignment offset indicator of the suspension screen only responded to the ON/OFF switches. It was the source of confusion.
What we wanted was for the indicators to react to the switches and gains of the internal filterbanks, in addition to the external ON/OFF switches.

This was done by editing SUS_SINGLE_OVERVIEW.adl in /opt/rtcds/userapps/release/sus/c1/medm/templates/NEW_SUS_SCREENS

The indicator logic was updated like

"dynamic attribute" {
        vis="calc"
        calc="(A!=0)&&(B&4)&&(C&1024)&&(D!=0)"
        chan="$(IFO):SUS-$(OPTIC)_POS_OFFSET_ON"
        chanB="$(IFO):SUS-$(OPTIC)_BIASPOS_SW1R"
        chanC="$(IFO):SUS-$(OPTIC)_BIASPOS_SW2R"
        chanD="$(IFO):SUS-$(OPTIC)_BIASPOS_GAIN"
}

Now the indicators will become green only when the offset switch is ON, the internal input/output switches are both on, and the gain is non-zero.
Enjoy!

[Koji, Yuta]

We implemented software watchdog for BHD optics.
In order to make the logic the same for all of the optics, software watchdogs for vertex optics, ETMX and ETMY were also updated.
Now all the OSEM PD VAR are calculated from the fast channels INMON, and turns off suspension damping loops only when tripped.

Software watchdog for BHD optics:
 - Created a template db file for c1sus2 optics to do software watchdog. Now C1:SUS-<OPT>_<OSEM>PD_VAR are calculated based on RMS of C1:SUS-<OPT>_<OSEM>SEN_INMON, and turns off SUSPOS, SUSPIT, SUSYAW, OLPIT and OLYAW dampling loops (but not alignment offsets nor disable coil outputs C1:SUS-<OPT>_<OSEM>_COMM) if triggered.
 
/cvs/cds/caltech/target/c1susaux2/template_soft_watchdog.db

 - Ran the following for <OPT>=AS1, AS4, LO1, LO2, SR2, PR2, PR3 to create db files for BHD optics.

python /cvs/cds/caltech/target/c1susaux2/createDbFromExcel.py <OPT>

 - Ran the following on c1susaux2

sudo systemctl restart modbusIOC.service

Software watchdog for vertex optics and ETMX:
 - Found that, for ITMX, ITMY, BS, PRM and SRM, oplev loops were not turned off even if they trip. They also disable coil outputs (ETMX as well).
 - Added the following to respective db files.

 record(calcout,"C1:SUS-<OPT>_OLP_DAMP")
{
    field(DESC,"Turn off OPLEV PIT SUS damping")
    field(SCAN,"Passive")
    field(INPA,"C1:SUS-<OPT>_OL_PIT_SW2R")
    field(CALC,"(A-1024)<0?A:(A-1024)")
    field(OUT,"C1:SUS-<OPT>_OL_PIT_SW2S PP NMS")
}

record(calcout,"C1:SUS-<OPT>_OLY_DAMP")
{
    field(DESC,"Turn off OPLEV YAW SUS damping")
    field(SCAN,"Passive")
    field(INPA,"C1:SUS-<OPT>_OL_YAW_SW2R")
    field(CALC,"(A-1024)<0?A:(A-1024)")
    field(OUT,"C1:SUS-<OPT>_OL_YAW_SW2S PP NMS")
}

 - Also commented out five lines below (ETMX as well) so that it does not turn off the coil output switch (we want them to be on so that alignment offsets are not turned off).

record(dfanout,"C1:SUS-<OPT>_PUSH_ALL")
{
    field(DESC,"Fanout for pushing all buttons")
    field(SCAN,"Passive")
    field(DISV,"0")
    field(SDIS,"C1:SUS-<OPT>_LATCH_OFF.VAL PP NMS")
    field(HOPR,"2")
    field(LOPR,"-2")
    field(OMSL,"supervisory")
#    field(OUTA,"C1:SUS-<OPT>_UL_COMM PP NMS")
#    field(OUTB,"C1:SUS-<OPT>_LL_COMM PP NMS")
#    field(OUTC,"C1:SUS-<OPT>_UR_COMM PP NMS")
#    field(OUTD,"C1:SUS-<OPT>_LR_COMM PP NMS")
#    field(OUTE,"C1:SUS-<OPT>_SD_COMM PP NMS")
    field(OUTF,"C1:SUS-<OPT>_LATCH_OFF PP NMS")
}

Software watchdog for ETMY:
 - Copied what was done for ETMX to ETMY db file. NOTE that it is /cvs/cds/caltech/target/c1auxey1/ETMYaux.db, and not c1auxey without 1 (40m/17948) !!!!!
 - Ran the following on c1auxey1

sudo systemctl restart modbusIOC.service

Next:
 - Make a new watchdog MEDM screen
 - Make a new suspension summary MEDM screen
 - Update the script to untrip watchdogs
 - Update the scprit to align and mis-align suspensions (use C1:SUS-$(OPTIC)_BIASPIT_OFFSET instead of using sneaky offset in coil output matrix?)

[Koji, Yuta]

Yuta was working on the watchdog unification for all the suspensions. He reached ETMY's one and tried to restart the ETMY's SUS slow machine.

We initially thought that the machine is c1auxey, BUT the correct name is c1auxey1 (horrible name). We lost ~30min because of this confusing name.

There was a folder named c1auxey and c1auxey in /cvs/cds/caltech/target. This was the source of confusion. I've compressed c1auxey into c1auxey.tar.gz. target_archive. We will eventually rename c1auxey1 to c1auxey in order to end this confusion.

Then we found that c1auxey1 could not come back after rebooting. This required the following actions.

... logged into c1auxey

$ df

Filesystem        1K-blocks       Used  Available Use% Mounted on
udev                1992120          0    1992120   0% /dev
tmpfs                402404       6044     396360   2% /run
/dev/sda1         113871044  108170996          0 100% /
tmpfs               2012012          0    2012012   0% /dev/shm
tmpfs                  5120          4       5116   1% /run/lock
tmpfs               2012012          0    2012012   0% /sys/fs/cgroup

modbusIOC was not loaded because the NFS filesystem /cvs/cds was missing.

$ sudo mount /cvs/cds

sudo systemctl daemon-reload
sudo systemctl enable modbusIOC.service
sudo systemctl status modbusIOC.service
sudo systemctl start modbusIOC.service

This made the Y-end epics service back in operation.

We are ready for the vent

Lasers:

• I've closed all the shutters for the four lasers.
• The PSL/End green shutters were also closed.

Vacuum:

• Closed V5 to isolate TP3
• Stopped TP1 then closed V4 (TP2 isolated)
• Stopped TP3
• Stopped TP2 manually (Note that the serial connection broken)
• Closed TP3/TP2 manual backing valves
• Stopped AUX RP for TP2/TP3
• Closed main valve V1 / manual valve to the main volume
• Closed the annuli volme by VA6
• Closed VM3 to isolate RGA

I measured 883 mW of PSL power before the PSL Shutter. I then set the waveplate AXIS 1 to 36.82 deg to make the power 93 mW.

Beam spot position measurements were done, oplevs aligned.
Attachment #1 shows IFO alignment with PRFPMI BHD aligned.
Attachment #2 shows oplev values.
Attachment #3 shows the history of beam spot position measurements since Saturday.

Updates on beam spot position measurements:
 - PR2 mass and moment of inertia were updated using the numbers from Solidworks model (40m/17937).
 - Script to measure all of them were prepared. This script still requires human to check the alignment and locking status. /opt/rtcds/caltech/c1/Git/40m/scripts/ASC/measureAllBeamSpotposition.py
 - Script to plot the history: /opt/rtcds/caltech/c1/Git/40m/scripts/ASC/plotBeamSpotPositionMeasurements.py

FROM THE POINT OF VIEW OF IFO, WE ARE READY TO VENT