Alexa Dan Evan Lisa Sheila

Today the ETMY bounce mode was ringing up during our many lock attempts.  We rediscovered Nic's work to damp this mode using DARM feedback to the pitch drive of ETMY L2.  We have been able to damp the mode while hanging out in ALS_DIFF, but progress is slow.  We suspect that the high noise floor of DARM when locked on ALS_DIFF is limiting our ability to damp the mode.  Transitioning to RF DARM may let us damp the mode more quickly, but this carries it's own inconvenience.

Anyways the way to damp the mode is to enable ETMY L2 DAMPDARM MODE1 --> PIT, with FM1-3 turned on.  (FM3 is a 206dB gain stage, I have never seen so many dB in one place before.)  With the bounce mode very high we had to reduce the overall filter gain to -0.04 to keep the SUS L2 drive from saturating.  (Nic recommends -1, probably this is good once the mode has damped a little.)

 - Dan as Sheila

Evan, Lisa, Sheila, Dan, Alexa, Kiwamu

Summary

With DARM on ASAIR_RF45_Q and CARM on analog REFLAIR_RF9_I at zero CARM offset, we were able to transition the DRMI 3f signals from REFLAIR_B to the 1f POPAIR_B signals. 

The locking sequence is scripted in the guardian up until this point.

Details

By exciting PRM, we adjusted the demod phase of POPAIR_B_RF9 and POPAIR_B_RF45. In both cases we maximized the I signals, and found 66.6º and -30º for RF9 and RF45 respectively. Then, we excited PRM, SRM, and BS at 137 Hz, 137 Hz, and 53 Hz respectively.  The folder /ligo/home/alexan.staley/Public/Spectra_3F2POP contains the DTT files, which have the power spectra for all the POPAIR_B signals, as well as transfer functions to the respective REFLAIR_B signals. The input matrix is as follows:

When locked with the POP signals, we measured all three loops and found that they were all at the appropriate UGFs with good phase.

This morning, the IMC had a difficulty in relocking it.

It turned out that most of the IMCASC DOFs had accumulated some large offsets which had pushed the alignment so bad that it was not able to catch a 00-mode. This was due to the fact that we had turned on both the offsets (e.g. IMC-DOF_1_P_OFFSET) and the integrators, which off course simply accumulated the offset in the up front even when the IMC was unlocked. I disabled all the offsets as they are not necessary for now.

Over the last few days, we have been occasionally been locking the ISI to the position sensors below 90mHz along the beam direction when the wind is high.

The general idea is that wind causes the ground to tilt, at low frequencies this shows up as a huge translational signal in the interial sensors, so we are better off controlling the ISI with position sensors to higher frequencies when the wind is high.  This means that we give up some isolation at the microseism, so it might not be a good idea to do this when the microseism is high.  The last few days the 0.1-0.3 Hz BLRM has been hovering around 0.4 um/sec, in this situation blending high along the beam direction when the wind is steadily above 20 mph seems to help us. 

I've attached a screen shot of the configuration that I've just switched the Y end to, this only takes two button clicks, switch the Y blend from 45 mHZ to 90 mHz, and turn off the output of the sensor correction in the Y direction. 

At X end we have been doing the same blend switching along the beam direction, but here we have the BRS, which can be used to subtract the wind induced tilt from the ground seismometers so that sensor correction can give us some benefits even durring windy times.  So the X end procedure is three clicks:  raise the X blend to 90 mHz, turn off the unocrrected X sensor correction by typing zero in the ISI_CUST_CHAMBER_STS2CART screen at the X X2 position, then turn on the corrected sensor correction by typing a 1 in the X X3 position.  Before and after screen shots attached. 

7:00 Jodi to LVEA West Bay 
7:30 Mitchell joined Jodi 
8:15 Aiden to LVEA
8:45 Bubba to LVEA (checking on 3IFO containers)
8:45 Karen to End Y 
8:45 Thomas and Sudarshan to End Y (PCal)
9:15 Corey to Mid-Y
9:37 Jodi and Mitchell back
9:40 Karen left End Y
9:47 Bubba back from LVEA
9:49 Corey back from Mid-Y
9:53 Rick to LVEA (for a few minutes)
10:01 Thomas and Sudarshan back
10:10 Aiden to High bay
~12:30 DRMI locked, commissioning started
A fire truck came in and went to Mid Y at one point in the afternoon (forgot to write down the time, my bad). Nothing was on fire.
~ 4pm, another lock (Hooray!)

This morning I turned on HEPI feedforward to BSC St1 on all of the BSC's. I took measurement as I went, generally improvements between 5x and 10x, between 1 and ~10 hz. The attached plot from ITMY is one of the better ones, but typical. Green is the ground, black is before, red is after. I still have to get some time to bring down FF to get some gain matching measurements, but I can probably fit this in on some morning before the commissioners come in.

Will load these tomorrow & correct sdf

Laser Status: 
SysStat is good (except VB program online = known issue)
Output power is 32.4 W (should be around 30 W)
FRONTEND WATCH is Active
HPO WATCH is red

PMC:
It has been locked 20 day, 2 hr 23 minutes (should be days/weeks)
Reflected power is 2.1 Watts  and PowerSum = 25.3 Watts.
(Reflected Power should be <= 10% of PowerSum)

FSS:
It has been locked for 7 h and 17 min (should be days/weeks)
Threshold on transmitted photo-detector PD = 1.12 V (should be 0.9V)

ISS:
The diffracted power is around 5.4 % (should be 5-15%)
Last saturation event was 7 h and 19 minutes ago (should be days/weeks)

Over the last week, we've made a few tweaks to the SUS DriftMon such that it can be a quick tool to use when checking for drifting or misaligned suspensions.  I've linked a smaller, more user friendly version of the monitor to the sitemap which can be tucked in a corner if useful.  If channels are yellow or red and further investigation is needed, pop open the larger screen via the button in the lower right corner.  From the larger screen you can also update the alarm levels to capture latest-and-greatest alignment.  I have updated the monitor to a full lock time from this last weekend (2/09 14:10 UTC, 1107526216 GPS).

I have adjusted the alarm thresholds of the SUS channels such that they tell you:
YELLOW = you are close to the previously saved alignment, so likely just need some minor alignment tweaks
RED = you are too far away from the previously saved alignment.  

If values are RED, and you are starting an initial alignment, it might mean:
- The GUARDIAN for the SUS is in a misaligned state
- WFS pointing offsets or FM integrators have been left on and need resetting
- Possible temperature change in the VEA

Note, RM1,2 and OM1,2,3 thresholds are pretty wide so I wouldn't expect them to go into alarm.  These suspensions are not used to point beams into the IFO but instead direct beams onto WFS/PDs so we are more forgiving of their pointing.  The OM chain might see more commissioning so we will tweak these alarm levels as needed.

The following maintenance tasks have been approved by Daniel for tomorrow:

1. restart h1susetmx model (attempt to fix decimation problem)
2. add some slow channels to TCS corner station model (h1tcscs)*
3. update PEM DAQ channels to list approved by Peter Fritschel*
4. add missing PCAL channels*
5. add ALS models test points to DAQ*
6. patch and reboot all CDS Linux workstations
7. reboot h1guardian0 machine (patch and remove NFS client file attributes caching)

* - task requires DAQ restart

We have had several reports of user accounts whose applets on the top right menu bar of their desktops have disappeared. We don't know why they disappear, but we know how to bring them back.

Attached is an image of what the top right should look like. If you are missing this, follow this procedure

1. Position mouse in top menu bar, not on an icon.
2. Press 'ALT' keyboad key and right-hand mouse key at the same time
3. Select  'Add to Panel...'
4. Scroll down, select 'Indicator Applet Complete'
5. Press 'Add' button

If the applet icon is not in the correct position, move it by positioning your mouse on the icon and pressing 'ALT' + right-hand mouse button

1. Select 'Move'
2. left mouse on the desired location

no restarts reported. Conlog reports not available.

Items for Monday by Subsystem

SEI:  Hugh/Jim reporting:  Test Blend Filter changes, Feed Forward Filters ready to go,

SUS:  no report

CDS:  no report

3IFO:  Richard reporting:  Working on Baffle PDs in LVEA

facilities:  Tumbleweed clean up until 10am inquired about for the week (Daniel mentioned tumbleweeding didn't affect too much)

PSL/Pcal:  Rick reporting:  EY visit for Pcal, ITM camera image work (work permit being made)

TCS:  3IFO inventory work

Items for Maintenance Tuesday (tomorrow)

• Daniel inquired about any possible bootfests tomorrow in the hopes for a quick recovery & being ready for Commissioning as soon as Maintenance ends at nooon!
• HEPI Pump Control (SEI)
• Capacitor checks (SEI)
• Seismometer [STS2] on floor (Richard),
• Current fluctations on HAM4/5 system (Richard)
• Oplev enclosure for BS (Doug)
• Long term storage container work (Bubba),

Sheila, Alexa, Evan

Summary

We have transitoned CARM from digital normalized REFLAIR9I to analog REFLAIR9I, with a 4 kHz bandwidth and 50 degrees of phase. An OLTF is attached [the last data point is spurious, so ignore it]. This lock started at about 2015-02-09 12:24:00 UTC. We are leaving the IFO locked.

There is plenty of phase to push the bandwidth higher, but we have encountered large offsets induced by switching on the common-mode and summing-node boards.

We can also improve the low-frequency fluctuations of the CARM error signal by introducing an integrator somewhere; we need more dc gain.

Details

Analog REFLAIR9I is plugged into input B2 on the summing-node board (SNB), with polarity "off". The output of B goes into input #2 on the common-mode board (CMB), with positive polarity. Digital normalized REFLAIR9I is plugged into input A1 on the SNB, with polarity "on". The output of A goes to input #1 on the CMB. [See LHO#16489 for a review.]

According to our reckoning, the shape of the digital CARM loop at the start of the transition is roughly 1/f above a few hertz. At a few hertz and below, it has a number of boosts and integrators which make it tricky to engineer a stable crossover with the analog signal.

Transition procedure is as follows, starting right after the guardian has brought the interferometer into resonance.

• With the B2 gain at 0 dB on the SNB, input B2 engaged on the SNB, and the input #2 gain at 0 dB on the CMB, we engage input #2.
• Step up B2 gain to 8 dB on the SNB.
• Turn off double integrator in digital CARM path (FM2 in REFLBIAS: it is two zeros at 2 Hz and two poles at 0.01 Hz).
• Step up B2 gain to 14 dB on the SNB.
• Remove zero-pole pair in digital CARM; it is a 1.6 Hz pole and 20 Hz zero. We remove this pair by engaging FM1 on REFLDCBIAS.
• Step up B2 gain to 17 dB on the SNB.
• Engage analog compensation filter in the CMB. It is a pole at 40 Hz and zero at 4 kHz.
• Ramp down digital CARM gain all the way to zero using the input matrix.
• Increase the analog CARM gain by 10 dB using the CMB slider. This is the point at which we saw disconcertingly large offset jumps. The total jump was 6 ct of REFLAIR_A_RF9_I_NORM. For comparison, the peak-to-peak fringing of this signal during digital CARM is 6 ct pkpk.
• We also saw a huge offset jump when turning off A1 on the SNB, which analogizes the digital CARM loop. In retrospect, this is probably because we had an integrated history on the output of this loop (since we turned the loop off at the LSC input matrix, not at the loop's output).
• To tune the error signal offset to zero, we have shifted around the common offset slider on the CMB

Also, the use of DHARD WFS (pitch and yaw) has removed the need for touching up the ETMs. However, since the AS36Q WFS feedback to the BS has not worked for the past couple of days, the BS had to be touched up by hand every so often.

Evan, Alexa, Lisa, Ryan, Sheila

Today we were able to lock about 7 more times todat, for at most a half hour.

• The gain needed for the transisition to RF CARM changed from an 8 in the input matrix to a 5.  We checked several settings, but found no explanation for this change.  The arm transmitted powers today have been 5-10% lower than on saturday morning.  
• The guardian now takes us all the way to on resonance, in about 15 minutes most of the time.  It fails often, maybe about half the time we drop the lock durring CARM offset reduction. 
• We were able to measure the ratio of the in vac refl 9I signal to the normalized (digitized) signal, using a 785 with an exictation into the CM board and test points in the summing junction (plot attached). The gain on the in vac path was 31dB, while the normalized path had a gain of 0 dB (these gains are both before the measurement point).  We made several attempts to transition to this signal, but haven't suceeded so far. 
• DRMI lock acquisition has been much faster today than in the past, probably due to the BS limiter.  
• We adjsuted the dark offsets in AS45Q
• ITMX oplev still seems to have significantly more drift than other optics, screen shot attached. If this is real motion it is too much for us to stay locked with, this is about a half a micron drift in less than 10 minutes. 
• We haven't been using the WFS to servo the BS alingment, we stopped doing this because engaging the WFS would sometimes cause mode mopping and lock losses.  Without the WFS we have slow drift tof the beamsplitter (as seen in AS 90) which might be causing some of our lock losses.  
• We drove a line in DARM at 43.1 Hz and measured the response in ASAIR 45 and saw that the signal in Q is only 5 times larger than I. Evan tried to adjust the demod phase to improve this, but the situation didn't change.
•  We added a boost to RF DARM, an integrator with a 2 Hz zero. The guardian turns this on before transitioning to CARM to REFL.  We are also no longer using FM4 (which was a low pass with extra gain around the suspension resonances, which seemed unusable on friday night.)  This needs further investigation.

We have many examples of locklosses from today.  We have added lock loss states to the guardians, so we should be able to get a list of these off the Detchar summary pages or the command line utility.  

The HWS on ITMX was running during last night's 1 hour lock. It saw a lens of ~6 micro-diopters (double-passed) form while the IFO was locked. The transient data is in the attached plot.

HWSY is not currently operational.

At the moment, I can't reliably estimate the absorption coefficient, but the absorbed power is about 1.06 mW per micro-diopter (double-passed), so roughly 6mW of power was absorbed.

If I go with 2W requested from the PSL rotation stage and do the same calculation we did at LLO:

2W (input power) * 30 (recycling gain) * 0.5 (beam splitter) * 286 (arm cavity gain) = 8.6kW stored arm power,

which yields

  ~700 parts per billion absorption in ITMX coating

I would definitely take this number with a healthy grain of salt ... I'd still like to run some diagnostics on the ITMX HWS following yesterday's code update.

(Also, the data in the attached plot was rescaled in dataviewer, by 0.00326, to what is currently displayed to account for the fact that the HWS magnification was set to 1x at the time of this measurement, rather than 17.5x. The spherical power scales as 1/M^2 = 1/17.5^2 = 0.00326).