Shocking, I know. While exploring data from last night (see LHO aLOG 11026), I saw what looked to be artificially clean straight lines through the performance of the ISIs. In order to confirm, I switched the windowing function in DTT from the default Hann(ing) to "BMH" (which I think is the Blackman-Harris Window), and indeed the reported ASD at those frequencies where the spectra looked artificial changed. Sadly, with this particularly huge dynamic range of a signal, the spectral leakage artifacts hide the real signal right at the frequencies where tilt bothers us (~0.5 to 10 [Hz]). I also tried the Flat Top window, and that seems to be the best for X DOF which has high dynamic range, but in RY, which doesn't, it "over" estimates the ASD at 0.05 [Hz]. Unclear whether this is a good compromise given that depending on the hour, we're interested in one band or the other affecting the performance. So, I guess I'll have to use asd2.m found here because "[when not computationally limited,] the spectral [leakage] of a single, long-time window is much smaller than the spectral [leakage] of many short, overlapping windows (like you get with pwelch)". Which really means we should just get the live noise budget up and running. #yakshaving
Aidan, Dave H, Thomas More cabling and plumbing. The leaky laser from ALOG 11019 was replaced with a new one and a matching RF driver. We're in a position to be able to turn on the laser mid-morning tomorrow.
All view ports on the spool between BSC 6 & 10 were removed and replaced with blanks and torqued. The dome went on BSC 10 and 90% of the bolts were inserted with 4 of them snugged with the remainder to be tightend tomorrow morning.
The new ITMx monolithic welding has been successfully completed and was left suspended overnight last night. The majority of the modal measurements have also been completed today, with the exception of the violin modes, which are proving to be trickier to measure than in previous suspensions. We will continue with these measurements tomorrow with some different equipment. Stay tuned for subsequent posts with IAS and modal measurement numbers.
Modal frequencies of the test mass (ITMX) with the PUM nominally locked are: ITMX value resolution Pendulum 0.649 0.002 Pitch 1.1153 0.002 Yaw 1.0928 0.002 Bounce 6.74 0.016 Roll 9.195 0.016 Violin modes fibre serial number BR 499.9 0.06 S1400155 BL 501.2 0.06 S1400156 FR 500.8 0.06 S1400138 FL 502.2 0.06 S1400153 [just as a note to self as much as anything - remember to use the Uniform window on the spectrum analyser for the violin modes when using the chirped sweep output]
Forgot to post this last week. Final alignment numbers from the ITMx welding (all directions/rotations reported from the view of the IAS equipment, i.e. looking at the HR face of the ITMx):
08:57 Aidan and Dave H. working on TCSX table enclosure 09:14 Jeff and Andres working on SR3 09:34 Corey to end X TMS lab to bag and inventory parts 09:53 Jeff done working on SR3, Andres finishing up 10:18 Betsy going to work in LVEA test stand 10:20 Jim done taking measurements on end Y ISI, to end Y to work on cable routing 10:24 Jax transitioning end Y to laser safe 10:29 Hugh to HAM4 to install vertical L4Cs 10:30 Kyle to crane RGA over YBM (WP 4529) 10:41 end Y is laser safe 11:07 Kyle done craning 11:59 Andres, Arnaud starting transfer function measurement on SR3 12:00 Guardian code is running on ITMY ISI 12:36 Betsy and Margo leaving to go into BSC 10 12:47 Karen going to end Y to clean in VEA 12:59 Ski to end Y (not in VEA) 13:04 Tour in CR 13:45 Ski back from end Y 13:49 Dave H., Aidan and Thomas starting work again on TCSX 15:05 Karen back from end Y 15:25 Dick and Jax going to end Y to install electronics, too much other work going on, left electronics there 15:46 Apollo installing dome at end Y High dust count alarms at end Y most of the afternoon, parties involved informed
I have started an rsync of the /opt filesystem on h1boot, to the new backup boot server starting at 16:14PDT. The filesystem on h1boot may be a little slow while the transfer runs, on a timescale probably measured in hours as it is about 300GB of data.
Mitchell with Apollo help until lunch, The newly assembled optical table was moved and placed on stage 0. After lunch I shopped for the parts at the interferometer emporium, moved the Stage 1 base plate onto the granite table and attached the plate flipping gear.
Laser is on Output power is 28.1 W (should be 30 W) Watchdog is active No warnings in PSL SYSSTAT.adl other than "VB program online" PMC Locked for 4 days 3 hours (should be days/weeks) Reflected power is 12.7% of transmitted power (should be 10% or less) FSS Reference cavity is not locked Trans PD threshold is .4 V (should be at least .9 V) ISS Diffracted power is around 7.8% (should be around 10%) Last saturation event was 11 minutes ago (should be days/weeks)
J. Kissel While using the channels that project the ISI ST2 GS13s in the Cartesian basis to the Test Mass QUAD suspension point in the Euler basis for the X-arm QUADs, i.e. H1:SUS-?TMX_M0_ISIWIT_?_DQ, I noticed that the projection matrices that create these channels were a copy-and-paste from the Y-arm, and therefore wrong by 90 [deg]. I've corrected the matrices at around 2014-03-27 21:55 UTC, saved, and committed a new safe.snap such that they stick. Details for the record: ----------------------- These matrices are calculated using the script /ligo/svncommon/SusSVN/sus/trunk/Common/MatlabTools/make_ISI2SUS_projections.m which takes in the locations of each SUS from the parameter file /ligo/svncommon/SusSVN/sus/trunk/Common/MatlabTools/SEI2SUScoordinates.m based on the systems SolidWorks drawings for that suspension. They are pre-calculated and stored in the mat file /opt/rtcds/userapps/release/isc/common/projections/ISI2SUS_projection_file.mat which I loaded into a matlab and ran the function /ligo/svncommon/SusSVN/sus/trunk/Common/MatlabTools/fill_matrix_values.m to automatically fill in the epics values. The theory behind these projections can be found in T1100617 #yakshaving
I checked and confirmed that the TMSX CART2EUL matrix is also wrong, but I didn't fix it because the parameter file doesn't have the coordinates for the TMS pulled from the systems drawings yet, and I'm already distracted enough. We'll get to it later.
And now the plots makes sense, and we can do science again. Look at the tilt (ISI RY) contribution to Longitudinal Motion at the suspension point! This is now, at no particular time. Now on to look a Sheila's data from yesterday (see LHO aLOG 11026).
Installed these this morning, leveled, torqued, wired, protected, labeled. No issues except the TCS rack 3" from Pier1. That rack will need to move.
Richard, Sheila
Richard noticed that the arm was locked pretty stably on IR when he came in this morning, and made a trend over night (first screen shot). There was a half hour lock, second screen shot.
No restarts reported.
Tonight the COMM handoff is stable even without the OpLev damping on. This is interesting because we have farily high microseism (above the 90% percentile all night) so it is possible for us to have a good night even with high microseism.
I was able to make some measurements of the COMM noise with and without the oplev dampng on, and it seems that the OpLev damping helps. I repeated the measurement twice, the second time the ground motion was larger. The RMS down to 0.02 Hz was reduced from 300Hz to 144Hz the frist time, and from 600 Hz to 235Hz the second time. In this calibration I have removed the cavity pole at 42 Hz. ( This file is saved in sheila.dwyer/ALS/COMM/COMM_Noise_March27.xml)
I also tried reducing the gain of the ALS COMM loop. Since 18dB gave us a 35 Hz ugf, I turned off the common comp and moved the gain down to -2dB, where we should ave a ugf of 350 Hz. This reduced the gain at high frequencies, but there must be an instability around 10Hz in this configuration. A screen shot is attached, and the data is saved as COM_Noise_March27_reduce_gain.xml
I have now turned off the OpLev damping so Fabrice can have some data to look at without damping in the morning. I plan to leave the guardian locking the arm on IR over night. This should be usefull data for anyone from DetChar who wants to look at the arm locked on IR, or seismic.
Valera called and asked for a measurement of the arm losses. I used REFL_AIR, meaured 0 counts when the IMC was unlocked, 1.539 when the arm was locked with ALS off resonance, and 1.471 when the arm was locked on resonance. The IR trans power was only 70% of the nominal, so the arm was not well aligned and this only gives us an upper limit on the losses, which is 159ppm, assuming the ITM transmission is 1.4%.
A more accurate measurement shoud probably be done.
Correction: This is really a lower limit, anyway not a very good one.
Spent all day trying to go back to the blue trace in the attached (which is from yesterday afternoon), but we came back to red, which is basically the same as this morning except the frequencies of three finger structure.
We re-routed the cables twice, and neither of these attempts recoverd the Q of 1.75 and 1.4Hz thing. So much for my "bundling=bad Q for translational" theory which sounded good.
Unrubbing excersize reliably eliminated the elevated noise floor 0.8Hz-6Hz that probably came from rubbing, and made the Q of some modes good, but did not do it reliably for 1.75Hz and 1.4Hz.
At this point I have to say it's unlikely that we can reliably make it like the blue trace even if we go in the chamber tomorrow, so if nothing special comes up we need to button down. I asked Arnaud to run his TF. Tomorrow Jim should run his stuff again.
As for the three finger structure (0.6-0.8Hz), we need to live with it, it's not unique to this pilot unit (see the X and Y free swing spectra, due to low resolution it looks like two fingers for Y, but the point is that there are three resonances between 0.6 and 0.8Hz for X).
BTW we eliminated another potential rubbing problem, which is the cable loop around the table cloth. We found that the clearance was less than a mm, probably something like 0.5-0.7mm. This depends on cable routing again, due to some strange way the cable goes in to the top mass, comes out, loops around the table cloth, goes in the top mass again, and finaly going to the ISI table.
Anyway, it was not touching but the head room was small, we pulled the cable on the cable clamp inside the top mass and made the clearance more like 2 or 3mm.
Have a few photos of the finicky beast (aka TMS). Photos are of cabling work (for run from Upper Mass to ISI Optics Table & for the loop around the Upper Mass), and final locations of the "fine balance masses" (which are a pair of 1/4-20 screws on the front end of TMS Optics Table).
Yesterday, we made two attempts at getting to the "blue" state from Tues:
Not sure what the deal is with my photos, but for some reason alog chooses to not let some of my photos upload to an entry. Have used original names of images, and changed names of files. Have tried from home and from work to no avail. Will keep trying. Attached is the error message I get afer I try uploading the file I select.
Entries made via MacBook Pro & with Firefox.
The issue was fixed by adjusting a parameter on the server.
Dave O, Gerado, The performance of the Output Faraday Isolator was tweaked and a transmission of 96.5 +/- 0.5% was recorded and an isolation of 9e-5 was measured. See attached report for more details.
We also used the Shack Hartmann Sensor to put and limit on the spherical power introduced by the Output Faraday Isolator (See attached document)