I noticed some cut-n-paste IPC errors in the h1asc model which needed resolving before we upgrade to rcg2.8.3.
1. IPC from h1asc to sus PR2,PR3 and PRM: RFM parts instead of Dolphin (PCIE). These were done today, so it was easy to hand edit H1.ipc and remove these 6 parts from the end of the file. I then edited h1asc.mdl and replaced the RFM parts with PCIE parts.
2. IPC from ASC to SUSHTTS: Dolphin (PCIE) parts used instead of Shared Memory parts (SHMEM). Slightly more difficult to fix as subsequent PCIE parts have been added. I removed the ASC parts and shuffled the later parts down in index. Then changed h1asc.mdl to replace PCIE with SHMEM.
Finally using rcg2.8.3 I performed a "make h1asc" which populated H1.ipc with the correct ASC part types. I verified the additional parts were correctly indexed.
In preparation for an RCG upgrade to version 2.8.3 tomorrow morning I am recompiling (but not installing) all models against 2.8.3. Please do no make any model changes or builds for the remainder of today.
In detail:
/opt/rtcds/rtscore/advLigoRTS-2.8.3 is a new checkout of the tags/advLigoRTS-2.8.3 released by Rolf today
/opt/rtcds/rtscore/release relinked to 2.8.3
New build area created
/opt/rtcds/lho/h1/rtbuild-2.8.3 and release pointer relinked to it
inside of rtbuild-2.8.3 the advLigoRTS-2.8.3/configure script was ran.
No issue to report other than missing zippers on the SEI Ceiling Sock.
Clean room back up with the Cartridge ready to go down at lunch. Torqued to Support Tubes and closed up by 1430pst.
Thanks to Apollo Scott, Mark & Bubba, SUS Travis and SEI Jim.
Corey helped too and took lots of photos.
Congratulations all for reaching this milestone safely. That's five H1 cartridge insertions, plus the two for H2 chambers BSC6 and BSC8. Nicely done! Remaining in-chamber installation includes:
ETMY Cartridge Install all day at EY
630 Cleaning crew working at EY
845 Filiberto installing HV supplies at EX
901 ETMY Cartridge being lifted
905 Jeff B adjusting dustmon at EY to provide better local alarms
1000-1200, 1300-1600 Jodi/Chris S working at MY
1107 Travis working on ITM at LVEA Test stand
1243-1500 Alexa working at EX
1245 Brief dust alarm in OSB optics lab....Joe Derenzis working in that area.
1315 Jax rummaging for hardware in Squeezer Bay
1427 Craig C working in H2 enclosure
After discovering that the MPC polarization controller creates a peak at 27kHz in the EX PLL noise spectrum, I decided to take some more measurements at the end station with the controller off.
PLL Servo Board
PDH Servo Board
PLL error signal measured out of PFD IMON, and PDH error signal measured out of demod IMON.
The lowest RMS comes from the PLL Boost 1: OFF, and the PDH Boost 2: ON. With the PLL Boost 1 off there is no gain peaking and the the OLTF of the PLL is stable with a UGF of 22kHz and a phase margin of 50 deg. I need to take an OLTF with the PDH Boost 2 on.
Adjusting EX PLL boost 1 ON/OFF and EX PDH boost 2 ON/OFF as above, the various Comm PLL Error signals (PFD IMON) were measured.
[Evan, Yuta, Kiwamu]
We worked on the alignment automation this morning. We closed the PR2 pitch and yaw loops manually.
In order to have the excitation only on the bottom mass of PR2 while keeping the feedback on the top mass, we installed a high pass filter at the bottom drivealign matrix and installed a notch filter at the top stage LOCK_P and LOCK_Y.
I plan to add these channels to the DAQ during maintenance tomorrow and create an alarm handler for them.
The slow feedback kept causing large values of the IMC VCO set frequency, which causes an error for the VCO. To get around this I turned up the UGF on the IMC VCO slow loop to 0.1Hz, and now the slow feed back gain (In COMM PLL screen) is set to 40000Hz/V. This seems stable for as long as the cavity stays locked, which is currently a few minutes. I also updated the screen ALS_CUST_COMM_PLL.adl because of a mistake with the reset button.
I was looking at the ALS-C_REFL_DC_BIAS output at the floor. The signal appeared on output 3 instead of 1 of the D-sub breakout panel. This was traced to a wiring error in the lsc model. Now corrected in the model, but not recompiled yet. This output has been connected to the second input of the common mode board and is operational. When the lsc model gets recompiled, the signal needs to be reconnected to output 1. The corner ALS overview screen has been updated to show the new filter modules.
The 27kHz line in the green locking was from the polarization controller, as Bram suggested. Attached are plots of the PLL and PDH error signals with the MPC powered on and off.
There is also a wandering feature in both signals that goes away when the MC is unlocked.
Again this afternoon the single shot cavity alignment was good, but the dither alignment kept arriving at an alignment where the cavity axis was different from the single shot beam.
There was an error in Sheila's matlab script in calibrating the PLL error signal. I have attached the new matlab file and plot. I have also attached the data concatenated.
Set the local audible alarm on dust monitor #1 at End-Y. The 0.3 alarm is set for 18 particles (@500 MEDM counts) and the 0.5 alarm is set for 10 particles (@300 MEDM counts). These values may need adjusting as more data points are collected.
Green team:
Red team:
Blue team (ALS WFS):
Blue team (ISCTEY, delayed):
TMS:
SEI/SUS team:
Kyle will be craning the portable RGA over the Y-arm in the vicinity of the elevated slab and test stand areas. Work should be complete by noon. See WP 4455.
Summary: Acoustic coupling at HAM6 (dark port) is likely to be a borderline noise source for aLIGO, near the expected noise floor in the several hundred Hz region, as it was at the end of eLIGO. As risk mitigation, I began investigating the resonance features that couple acoustically driven external vibrations to the table surface with transfer functions approaching unity in the several hundred Hz region. The resonance features consist of broad resonances that are likely ISI blade spring and flexure and/or table resonances, with higher-Q resonances riding on them that are likely from components: the GS13 pods, the mass pegs, the individual panels, and the optic structures on the table. The damping planned for other reasons is unlikely to significantly change the Qs of the peaks. We may want to begin developing damping schemes in case this coupling is a problem at HAM6 or elsewhere.
Vibrational coupling, driven by external acoustic noise, was a problem in eLIGO, requiring us to suspend HAM6 steering optics and install blade springs on tip tilts (reduction in 400-550 & 750-900 Hz peaks with suspension). Even with this work, DARM was still somewhat contaminated by acoustically-driven vibrational coupling at the end of S6 (here). Figure 1 shows that there was coherence between the HAM6 geophones and DARM in the 400-550 and 750-900 Hz band late in S6 after our HAM6 interventions. I had hoped that the passive damping planned for the HAMs (for other reasons) would also help reduce the higher frequency motion. However, Guillermo Valdes (UTB), took a recent look at LLO HAM3, which had the damping installed, that made me want to conduct a damping experiment at HAM4. I found little or no decrease in the Qs of the high-frequency peaks even after I more than doubled the planned tabletop damping (7 dampers instead of 3). Thus I wanted to gain a better understanding of the source of this vibrational coupling in case it is a problem in aLIGO.
The ISI transfer functions from ground motion to table motion suggest an isolation of less than 10 at certain bands in the several hundred Hz region (for example see 500 Hz here). The ISI transfer functions typically show broad horizontal resonances in the 400-550 and 750-900 Hz regions, even before the tables are populated (Figure 2). Modeling by High Precision Devices (G-0701156-00-R) suggested lowest table resonances between 300 and 400 Hz. There are resonances in this region, but the higher frequency resonances are typically more pronounced on the geophone signals. A competing, and, to me, more likely possibility, is that the broad resonances are resonances of the blade springs and flexures, possibly matching table resonances (see resonances of the BSC ISI blade springs here). Geophone spectra show these broad resonances, as well as more narrow resonances riding on top of them (Figure 3). Thus the tallest peaks, the ones that are most likely to contaminate DARM, could be reduced by damping either the broad resonances or the peaks that ride on them. As an example of how the features appeared in DARM during early eLIGO, Figure 4 shows DARM from early S6 with the broad ISI resonances and narrow peaks riding on top. While reducing the broad ISI peaks would have been the best option, reducing the narrow peaks looks like it might have reduced the maximum peak height by a factor of 2 or 3.
In an effort to identify the sources of some of the narrow peaks that ride on the broad resonances, I did a series of tap tests on HAM4. Figure 5 shows that taps on the ISI side pegs for balance masses, the centers of the “X” shaped cutouts in ISI side panels, optic supports on the tables and GS13 pods all excited individual peaks or families of peaks that rode on top of the broader resonances. Figure 6 is a photograph showing these structures.
Figure 7 shows that each GS13 pod has its own characteristic family of several high-Q peaks. The pods consist of GS13s inside vacuum enclosures. The vacuum enclosures can be heard ringing long after they are tapped. These pod peaks are evident in geophone signals even when there is no tapping. The resonance peaks for a particular pod are largest in signals for the geophone in the pod, but are also evident in signals from other geophones. Fortunately, the frequencies of the geophone pods are, typically, slightly below the broad 400-550 Hz resonance, so it is likely that only a few of the higher frequency pod peaks will be among the highest amplitude peaks that could show up in DARM. Nevertheless, we may want to consider a passive damping scheme for the geophone pods, or a scheme to move the resonances a little lower in frequency.
The individual panels that make up the body of the ISI also have resonances that may coincide in frequency with the broad 400-550 and 750-900 ISI resonances. I took two of the several panel types and suspended them by “strings”. One had resonances at about 400, 700 and 800 Hz, the other at 490 and 670 Hz. Some of the resonances may be associated with the “X” shapes left by the cutouts (see Figure 6). Fabrice and I imagined a couple of passive damping schemes for these, if needed, but access to all of them would be tough and we might just want to instead put damping that is tuned to the 400-550 and 750-900 Hz resonances onto the tabletop structures that touch the beam.
Finally we might want to be thinking about a simple scheme for passive damping of individual optic structures or the ISI mass pegs.
There may also be ways to damp the broad resonances, such as tuned damping on the blade springs, as the SEI group developed for the bucket peaks in eLIGO, or even active schemes. I would like to repeat my HAM4 work at HAM6 just before it is closed up, so that we have a record of the frequencies of the pods and optics on the surface of HAM6 in case features show up in DARM. It might also make sense to investigate whether blade spring and flexure resonances could produce such high transfer functions. In conclusion, I think we should begin to consider mitigation routes in advance in case this coupling is a problem at HAM6 or elsewhere.
I've made several changes to the TwinCAT code and medm screens for ALS friday and this morning. The most important thing is that I added the slow feedback from the COMM PLL contrl signal to the IMC VCO. At the bottom of the COMM PLL screen now there is a section for the Sow feedback, if this is running it replaces the IMC-VCO_SETFREQUENCYOFFSET with an error signal that cancels the error signal from the frequency comparator and replaces it with the PLL control signal, multiplied by a gain. The gain is currently 10000Hz/V. This also checks to see if the arm cavity is locked, if the SHG status is OK and if the beatnote stength is above some minimum, and will stop the feedback if there are error conditions.
This seems to be working well as long as the arm cavity has the slow feedback to the top mass active.
Other updates were adding an error conition to the end station laser locking pll for the noise eater oscillation, fixing the corner state machine, and adding an option to the PDH autolocker to allow it to control the second boost. medm screens updated were mostly in ALS, except for ISC_CUST_DUAL_PFD.adl
First attempt failed.
OL whitening.
See Thomas's entry. ITMX and ETMX are good now.
Demod phase and sensing nonsense.
After Thomas was done with OL, we had to measure sensing matrix again as it uses OL data. Found that demod phase changed, not reliable from measurement to measurement. In the end, I've found that:
The measurement was done by wiggling ETMX at 3.5Hz and ITMX at 1Hz, first in YAW, measuring the demod phase and the sensing matrix for segment 1 and 3, then switch to PIT for 2 and 4. Angle to length shouldn't be a problem as PDH should have a large gain there to squash the length error.
New sensing matrix.
The attached shows one snapshot of demod phase itself on the left screen and the demod phase/sensing matrix measurement on the right.
Two dtt sessions show PIT and YAW. In each dtt window, left is ITM and right is ETM, top shows how Q phase is minimized, middle shows the sensing matrix amplitude in cts/urad (WFS/oplev), bottom shows the sensing matrix phase (should be 0 or +-180) as well as relative phase between diagonal elements (should be +-180).
The sensing matrix phsae still has 10 deg-ish systematic at 1Hz (ITM) and 25 deg-ish at 3.5Hz (ETM). Maybe a decimation filter for OL DQ, maybe something else, definitely not 1:10 whitening, I don't worry about this for now.
PIT.
ITM | ETM | |
WFSA | -1922.46 | -1651.43 |
WFSB | -655.52 | -393.65 |
YAW.
ITM | ETM | |
WFSA | -3243.17 | 1042.83 |
WFSB | -2947.28 | 931.62 |
That was it for Friday. I will not work over the weekend at the site.
Corresponding input matrices:
Pitch:
Hard | Soft | |
A | -0.000568 | 0.000588 |
B | -0.00228 | 0.00776 |
Yaw:
Hard | Soft | |
A | 0.0526 | -0.0582 |
B | -0.0274 | 0.0299 |
These have been updated accordingly.
I've also attached a matlab script that will turn the sensing matrices into input matrices, just to make sure the method exists somewhere other than my notebook.
HAM2 and HAM3 as well as the HEPIs tripped at the exact same time (actuator trip for all of them) @ gps 1077061619 (15:46 PT)
It happened again this evening when Sheila restarted the Beckhoff at 1077078129. HAM2 and HAM3 ISI actuactors tripped because of a spike in the act signal, as well as the HEPI. cf attached plots
Jamie and I tried this a few more times..
Stoping and restarting the EPICs database doesn't unlock the IMC.
Stoping PLC2 does unlok the IMC (and FSS) , although in the three times I tried it it didn't trip any HAMs today. Restarting the other PLCs doesn't cause any problems, and it doesn't matter wether or not the Guardian is running.
Performed first round of builds on all models, four did not compile:
h1isiham3, h1iopseib1, h1susitmy, h1iscey
we are investigating why these are failing.
H1.ipc file did not get modified after the first build of h1asc before the full build.