RickS, Darkhan

We adjusted 35.9 Hz TST (L3 stage only) line drive level from 0.08 ct to 0.11 ct.

Now the amplitude of the TST calibration line in DARM_ERR readout is close to ampltitudes of 36.7 Hz Pcal and 37.3 Hz x_ctrl calibration lines. The target SNR for these lines and for 331.9 Hz Pcal line is 100 SNR with 10 s FFTs (see T1500377).

Evan's script to automatically take frequency and intensity transfer functions is now running. 

As a reminder, the summing note B path was repurposed for this run. The interferometer won't relock unless we reconnect them. 

At 4:20 UTC we started changing the TCS X CO2 power from 0.23W to 0.4W. The rotation stage took us on some full circles, but by 04:23 we reached 0.4W. 

At 4:45 I increased the frequency noise drive by a factor of 5 to gain back coherence.

At 6:04 I decreased the power to 0.35W - trying to find the minimal frequency noise point. (The coupling sign had changed.)

At 6:34 I reduced it to 0.3W.

Shivaraj, Darkhan

Summary

We replaced whitening filters on Delta L_{res} output channel with 2 zeros and 2 poles zpk, and Delta L_{ctrl} with 3 zeros and 3 poles zpk.

Details

Madeline uses FIR filters to dewhiten Delta L_{res} and Delta L_{ctrl} in GDS scripts. To make it easier to generate shorter dewhitening FIR filters for these channels she requested to replace the existing 5 zeros and 5 poles zpk filters in both of these channels with simpler, less poles and zeros zpk's.

Shivaraj designed new FIR filters for both of these channels, and we implemented them today around 1pm.

Power spectrum of H1:CAL-CS_DARM_DELTAL_CTRL_WHITEN_OUT before applying new whitening filter plotted in attached "H1-CAL-CS_DARM_DELTAL_CTRL_WHITEN_OUT_z5x1_p5x100_MAG.pdf". After applying new whitening filter, zpk([1; 1; 1], [500; 500; 500], 1), unfortunately the interferometer wasn't stable to take a spectrum measurement after I changed the filter, so I couldn't evaluate "whiteness" of the output in this channel.

Power spectrum of H1:CAL-CS_DARM_RESIDUAL_WHITEN_OUT before applying new whitening filter plotted in "H1-CAL-CS_DARM_RESIDUAL_WHITEN_OUT_z5x1_p5x100_MAG.pdf". After applying zpk([1; 1], [500; 500], 1), the spectrum of the same channel is given in "H1-CAL-CS_DARM_RESIDUAL_WHITEN_OUT_z2x1_p2x500_MAG.pdf".

We also plan to implement similar whitening filers on new (not yet existing) channels for Delta L_{TST} and Delta L_{PUM/UIM}.

Back to locking robustly, all the way to Nominal Low Noise.  Hooray!

It was a bit of a trying day, locking-wise.  Jim was a champ, trying to get it back up after the many, many locklosses throughout the day.  We have made several changes to the guardian scripts that seem to be helping immensely.  Here are the changes that were made:

• Slower CARM offset reduction in Switch to QPDs.
  • Toward the end of the day, we were losing the lock (6+ times in a row!) during the CARM offset steps.  Sheila and I did the Switch to QPDs step by hand, in several smaller steps, which was immediately successful. 
  • We doubled the ramp time for the final CARM offset reduction, and have now successfully gone through the sequence several times.
• Lower DHARD Pitch gain.
  • Several locklosses early in the day were due to oscillations in the DHARD pitch loop, which is modified at the beginning of Noise Tunings.
  • We measured the DHARD loop, and found that the gain was a factor of 2 too high.  The ELP33 cutoff filter eats enough phase that, in combination with the too-high gain, we were unstable.
  • The DHARD gain increase from 10 to 20 in the Lownoise ETMY ESD state was commented out.  The overall gain of DHARD Pitch is now correct, with the UGF appropriately around 3 Hz. 
  • The ELP33 in FM10 turn-on has been removed from the Noise Tunings state.  Now that the gain has been measured and corrected, we can probably put it back, but I'm leaving it out for now, since I can't stay to test it. 
• ITM oplevs left on.
  • It was decided over the weekend that the ITM oplevs need to be left on, for stability, until we can finish tuning up the ASC loops.  However, I found that they were still being turned off in the Noise Tunings state.  This may not have been causing locklosses, but until we can verify this, we're commenting this out.  So, now the oplevs for the ITMs are actually left on.
• OMC DCPD whitening back in guardian.
  • Due to the giganic violin modes over the last week, the turn-on of the OMC DCPD had been commented out of the guardian.  Since we have successfully locked with it turned on by hand, I've added it back to the guardian.  Our range (according to SensMon) is back above 60 Mpc.  Nice.

As an aside, I happened to notice some transients in the ETMX L3 ESD AMON channels pass by on Dataviewer, and looked into them a bit.  These transients happened while we were turning off the ETMX ESD after transitioning to the ETMY.  Thankfully, we didn't lose lock, but I post the time series of the transients anyway.  You'll notice in the attached plot that we basically see no effect in DARM, although I should put these channels through the new lockloss tool from Hang, et al. to see if there are any high-frequency things in DARM that are being swamped by the low-f stuff. 

Matt E., Sheila D., Jamie R., Terra H., Hang Y.

We updated our lockloss tool. A mistake of using the filter was fixed, and all filtering now is done by using secend order sections, which are more stable than using transfer functions ('BA') directly. The algorithm of handling invalid channel names when fetching data from the nds server (which was the time limiting part of our previous version) was also updated, such that now the lock loss analysis on all valid DQ channels can be finished in a few minutes. Besides the previous code finding all sudden changes before lock loss, Terra also finished a script that could pick out saturated channels. The codes are available at

/opt/rtcds/userapps/trunk/isc/common/scripts/lockloss

with both a python version and a matlab version.

==========================================================================================================================

We applied the code to a lock loss happened at GPS time 1122945715.0, or UTC time 08/07/2015 01:21:38. The two plots shows the first 32 channels that glitched before this lock loss. In the plots, blue lines are raw time-ordered data streams of that channel, and red lines are absolute values of data filtered to a specific frequency band. Thin black lines are pop data, and thick vertical black lines indicate the starting point of that channel going wrong. In the two txt files, the one starts with lockloss shows all DQ channels that glitched for this lock loss, ordered by the time prior lock loss, and the one start with 'satLockloss' gives all channels that saturated.

Kyle, Gerardo 

Today we continued looking for the leak(s) on the Y-mid turbo, and found it/one -> there is a 2 x 10-5 torr*L/sec (for helium) leak at the factory O-ring seal for the motor wiring feed through -> This is thought to be a very thin cross-sectioned O-ring between the square-flanged feed through and the turbo motor body -> This value of leak (if same for air) would account for the observed pressure rise for time periods this pump has remained idle but, not knowing the actual dimensions of the O-ring could also be accounted for by permeation etc...  -> to test, we moved the LD and helium to the X-mid turbo and observed no indication of a leak or permeation for the same application of helium -> it is not clear how serviceable this O-ring is -> it may be that it would have to be stretched over the square flange in lieu of desoldering motor wiring -ouch!  

ALL TIMES IN UTC

15:00 IFO not locked. Appears we had a couple of Earthquakes in the Tonga and Ecuador area (~5mag).

15:10 Guardian trying to re-lock. Touched ETMY to correct alignment. Locking sequence commencing.

15:20 IFO broke lock at "Noise_Tunings"

15:35 ESD X railed

15:39 Jim Batch to restart broadcast system

15:48 ISC_LOCK set to down (in Jim W's abscence) due to ESD at EX stillbeing railed.

16:00 Ellie into the optics lab

16:00 ESD EX no longer railed

16:04 talked to Jim and Fil at EX to confirm reset of ESD driver.

16:05 ISC_LOCK re-initiated

16:35 Fill into CER to install Cosmic Ray chassis

17:05 Fil out of CER.

17:05 Robert into LVEA to do HF acoustic injections

17:10 6 Kyle and Gerardo to Y-Mid to get/leave equipment

17:20 Robert and Katie out. Lockloss

17:23 Keita and Jenne to LVEA to investigate/turm off ISS noise injection that may have ben left on.

17:30 Keita and Jenne back

17:40 After the most recent lockloss, it seems that EX ESD driver is having trouble turning back on....AND it's railed

17:58 Found a bug in the code. Was able to reset remotely!

18:02 Begin locking sequence

18:02 Nutsinee popping into LVEA to take pictures at TCS X table.

18:08 Nutsinee out

18:31 Richard out on the roof

19:03 Richard back from the roof

19:07 Cheryl going to have a quick look at the exterior of IOT2R cabling.

19:11 Gerardo and Kyle back from the Mid station

20:52 Kyle and Gerardo to MY then MX then Y2-8.

22:55 Kyle and Gerardo back

Chris S. Joe D.80%

8/4/15 - 8/7/15
The crew installed metal strips on 120 meters of tube enclosure. That was the last of the strips that were cut to length. We had one more roll which was cut into 10' lengths today.

Scott L. Ed P. Rodney H.

8/5/15
79.6 meters of tube cleaned today ending at HSW-2-026.

8/6/15
71.3 meters of tube cleaned ending 10.7 meters east of HSW-2-023.

8/7/15
82 meters of tube cleaned ending at HSW-2-018.

As per Jim, this will be addressed on Tuesday Maintenance Day.

Let's add some statistical information on the loud glitches we've been investigating lately (see 20176, 20276, 20304).

All those glitches have a very distinctive shape: a sharp peak rising in about 1ms and decaying in about 1ms, followed by what we believe is the DARM loop response: a second slower peak of opposite sign.

I looked into all the glitches I previously identified, and run a MATLAB script to fit them with the above shape. What I'm fitting is a 100 Hz high passed DARM_IN1_DQ signal, with additional notches at 60, 502, 992, 1000, 1462 and 2450 Hz. The parameters are the rising and decay times of the first peak, its amplitude, the rising and decay times of the second peak and its amplitude. The waveform is the sum of two triangles. See for example the first two plots for a large and a small glitch.

After fitting all the 151 glitches, I actually convinced myself that 13 of them were probably of another kind, since the shape was clearly different. In two occasions (GPS 1117616943 and 1117631610) there were double glitches (see 3rd and 4th plots).

Armed with the results of the fit, I can now look at the distribution of the amplitudes with sign and of the duration. The 5th plot shows an histogram of the amplitudes. Result: there are both positive and negative amplitude glitches, but large amplitude glitches are only negative (in DARM_IN1_DQ).

The 6th plot shows the distribution of the rising and decay times of the first peak. The distribution is quite well peaked around 1 ms for the rising time and 1.5 ms for the decay time. For the second peak, rising times are peaked around 2.3 ms and decay times around 5 ms.

The last attached plot shows a distribution of the absolute value of the glitch amplitudes, in log scale. The low amplitude cut-off is likely due to my inability to detect the smallest glitches (and the threshold of 5 Mpc drop in the range). The amplitudes span a couple of orders of magnitude, with a quite smooth distribution. I see no reason why we shouldn't believe that there are many more low amplitude glitches with the same shape (and maybe origin) that we simply can't detect in this way.

The attached text file contains all the results of the analysis:

Column 1: GPS time
Column 2: amplitude of the first peak
Column 3: rising time of the first peak [ms]
Column 4: decay time of the second peak [ms]
Column 5: amplitude of the second peak
Column 6: rising time of the second peak [ms]
Column 7: decay time of the second peak [ms]

Elli, Nutsinee

We currently don't have the return SLED beam to the ITMX HWS. Not sure since when this is happened but looks like the beam splitter is likely to blame.

Sudarshan, Evan, Stefan, Shivaraj

We completed  a set of Pcal sweep measurement against the DARM for both endstation and the measurement files and the templates are stored at the following svn location.

/ligo/svncommon/CalSVN/aligocalibration/trunk/Runs/PreER8/H1/Measurements/PCAL/

Analysis will follow.

This morning after a lock-loss, the ETMX ESD railed, so Fil and I drove to the end station to reset it. To do this, Fil first powered off the chassis by the chamber (3rd chassis from the bottom on the right rack, power is red button on the left), then unplugged the right most cable on the chassis. He then powered the chassis back on and plugged the ESD cable back in. This seemed to fix the ESD, until it tripped just a little while ago.

Reinstalled PCAL AA Chassis at EX. Unit had three channels that needed repair (S1400574).

This morning, after the IFO was locked, DARM was super noisy in kHz region. ISS error point was also super noisy and the coherence between the two was big.

Turns out that the ISS got noisy at the tail end of the lock stretch from last night, at around 2015-08-07 12:00:00 UTC. That's 5AM in the morning.

We went to the floor and sure enough a function generator was connected to ISS injection point via SR560. We switched them off, disconnected the cable from the ISS front panel, but left the equipment on the floor so the injection can be restarted later.