NOTE: If an LHO ilog entry appears run-together or corrupted, try switching (pub) in the web address to (oldilog). There are now two ilogs, one for before and the other for after a failed upgrade around January 08. I think I changed most links to the appropriate ilog, though. For the old ilog, YOU WILL HAVE TO USE READER/READONLY as your name and password instead of your ilog name and password.
CONTENTS
* Summary of transients produced by the environment during iLIGO and how they were mitigated.
* Lessons from iLIGO: prepared for the Thirty-Meter Telescope
* PEM tutorial (sensor locations, calibrations etc.)
* Coupling studies (PEM injections)
* Environmental influences studied during iLIGO
o Acoustic
o Seismic (including upconversion)
o RF
o Magnetic
o 60 Hz
o Dust
o Wind
o Other
* Environmental influences studied during eLIGO
o
Magnetic
o World Wide Electromagnetic Environment (including audio frequency)
o
Glitch
finding and mitigation
o Other
o AOS
o CDS
o PSL
o PEM
o ISI
o Other
* Guide to Environmental Disturbance talks
ENVIRONMENTAL TRANSIENTS DURING iLIGO
GENERAL
In general it is important to consider environmental transients in addition to average environmental spectra. Transients, especially seismic and magnetic, can be an order of magnitude above average levels, and can occur so often that they can be more important to the performance of the interferometer than the average environmental spectra. We also learned that transients at each station can be very different. For example, transients at EY at LHO set the expected level of Newtonian noise nearly 10 times higher than any other station would.
* Seismic percentiles for all LHO stations 06/17/20 06, LLO stations here.
* Magnetic transients at LHO 01/14/2006, LLO 10/13/2007
MAGNETIC
A) Coupling of magnetic fields to test-mass magnets (details)
* Sudden current changes in the power grid from shorts, breaker trips and capacitor insertions. Mitigation: none, just power company records, magnetic field monitoring.
* Sudden local current changes from motor and heater switching. Mitigation: shut off motors and pulsed heating, building temperature control with chilled water (not extended to LLO).
* X-arm crab-killing transient at LLO – source unknown
B) Coupling of magnetic fields to electronics (details)
* Continuous and transient (only preliminary results for transients) fields in electronics racks couple via electronics. Future mitigation: test fields and coupling for new electronics, physically separate high-current sources and digital sources from analog coupling sites.
SEISMIC
A) < 10 Hz (details) (scattering details)
* Seismic signals from microseism to wind to trucks caused in-band glitches when the actuation coil magnetic fields changed rapidly. Mitigation: replace test mass magnets
* Large relative motions between scattered light reflectors and arm cavities produced fringe wrapping noise. There were also mysterious multi-reflection paths. Future mitigation: baffles
* Servo saturations (for all seismic bands).
B) 10 to 25 Hz (details) (scattering details)
* HVAC and motor seismic signals caused glitches and continuous noise. The upconversion mechanism appears to differ from the mechanism at lower frequencies and still needs solving. Mitigation: lower turbine flows, shut down motors, but mitigation of upconversion mechanism yet to be determined.
* LN2 dewar stick-slip events produced glitches. Mitigation: insulation of dewars greatly reduced rate.
C) 25+ Hz (details) (scattering details)
*Direct coupling of local seismic signals, predominantly to the dark port, but also to the PSL and optical levers. Mitigation: float dark port tables, control site activities.
*Direct scattering noise. Mitigation: careful beam dumping, baffles.
ACOUSTIC
(details) (scattering details)
* Beam clipping, backscattering, doppler shifting, index of refraction modulation all caused coupling of acoustic signals. Mitigation: reduction of coupling (eg. larger optics, telescope damping), isolation of coupling site (e.g. acoustic enclosures), reduction of source signals (e.g. moving electronics racks).
* Microphonics. Mitigation: test electronics.
RF
(details)
* Automatic link establishment transients close to (and may have) shown up in DARM. Future mitigation: better RF shielding. However, coupling is much lower at auxiliary modulation frequencies, so DC readout may mitigate.
* Some suggestion that IFO generated RF transients may have coupled to RF electronics. Future mitigation: shielding.
DUST
(details)
* Produced glitches only at the dark port and only where the beam diameter was small. Mitigated with two purge systems and minimization of small-beam diameter regions.
List of seismic/acoustic lessons learned in iLIGO - prepared for the Thirty-Meter Telescope
1. The spring isolation of our 28 Hz HVAC turbines works as intended, so we dont have problems with vibrations at this frequency, but, worse, we had broad band noise that was several times the ambient ground motion amplitude over a band of 10-100 Hz possibly produced by turbulence from the 50 MPH wind coming out of the turbines. We have turned down the turbine flows to reduce this broad band seismic noise and installed local Mitsubishi Mr Slim type systems where we needed extra cooling.
2. Shallow pipes etc. going under driveways/parking areas can produce bumps in the asphalt. We had one that produced a bump in the asphalt that caused large vibrations whenever it was driven over. We reburied it and repaved, but never completely got rid of the bump in the road.
4. Fake cattle guards are better than real ones.
5. Attenuation with distance for the Rayleigh surface waves that cause the large amplitude motions due to local surface sources goes roughly as:
Afar = Anear sqrt(Rnear/Rfar) e^((pi*f/Q*v)(Rfar - Rnear)) (f = frequency, v = velocity, A = amplitude, R = distance)
For our sandy desert site, propagation velocities are about 500 m/s at 5 Hz, 150 m/s at 60 Hz. We measured the Q as about 50 at 5 Hz. Typical values of Q in these bands go from 20 to 200. Since you only get improvement as the square root of distance, until the exponential term kicks in, it is useful to know the propagation velocities and Qs. For our Hanford case, the 1/e distance was about 50m at 60 Hz and 150m at 20 Hz.
6. In general it helps to use higher frequency rotating equipment (2 pole instead of 4 pole motors) to take advantage of the fact that the ground typically attenuates better at higher frequencies.
7. If possible, resonances should be away from 58-60, 28-30, 17-20, and 14-15 Hz to minimize problems with 2, 4, 6 and the occasional 8 pole motors (we had a squirrel cage hood fan running just below 15 Hz).
We have had the most problems with compressors in HVAC systems which typically run from just above 18 Hz down to 17Hz when they are loaded. These frequency sweeps cross some of our high Q optic resonances.
8. Distant buildings can be a problem if they are not beyond the exponential seismic cutoff. The HVAC chiller in a building 300 m from sensitive areas produced a moving 18 Hzish peak in ground motion that was nearly 10 times background at our site - until we put it on springs.
9. Seismic isolation of noisy sources is usually a lot easier than acoustic isolation: seismic sources just need to be put on springs or elastic pads. Acoustic mitigation is a lot harder - it is best to keep potentially noisy sources in a different room then the sensitive equipment.
10. Electronics racks can easily be put on rubber feet that are quite effective at isolating the usual >50 Hz sources in the racks. Acoustic isolation of the racks is a different story.
11. Watch for underground motors (e.g. sump pumps, well pumps). They may couple well to the ground.
12. Damping of pump pulses can be partially short-circuited by pipes that go into the ground before they reach the damper. The pulses can couple well to the ground.
13. Seismically isolated HVAC equipment was not always installed properly. The springs on hanging components were often short-circuited. The connections to equipment on springs (the connecting ducts, pipes, electrical conduits) were sometimes hard rather than soft.
14. Gravel is much worse than old worn roads, which are much worse than smoothly paved roads. We are working with the Washington Department of Transportation and requesting the smoothest possible paving when they repave near our site.
15. Trucks on a road 2 km from our Hanford site produce seismic noise in a 3-10 Hz band that reaches a factor of 30 times our normal background (our truck-free background is at a moderate level of a couple of nanometers per sqrt(Hz) in this band).
TUTORIAL
SciMon Camp web page (here) contains:
* Sensor locations
* Senor calibration factors
* Simplified coupling factors (could this DARM event have been produced by this sound?)
* Scale maps of buildings
COUPLING STUDIES (PEM injections)
iLIGO S5
S5 coupling summary Talk March 08
LHO
* Nov-Dec 2005 01/08/2006
* May 2006 05/14/2006
* April 2007 (acoustic) 04/21/2007
* Oct 2007 10/24/2007
LLO
* Dec 2005 12/10/2005
* Aug 2006 08/22/2006
* Oct 2007 10/13/2007
eLIGO S6
Early S6 coupling summary Talk September 2010
LHO
* July-Aug 2009 09/02/2009
* June 2010 (with Anamaria E.) 06/18/2010
LLO
* July 2009 08/09/2009
* June 2010 (with Anamaria E.) 06/26/2010
ENVIRONMENTAL INFLUENCES STUDIED DURING iLIGO
ACOUSTIC
S5 coupling summary Talk March 08
For pre-mitigation state, see references under mitigation
Mitigation/vetoes
* Acoustic mitigation 1st progress report April 03
* How it couples at tables 04/19/2003
* Acoustic mitigation report Aug. 03
* Acoustic mitigation progress - down by 100, Nov. 03
* Acoustic mitigation report - down by 1000, Feb. 04
* Down by 10,000 at LHO after REFL enclosures and table floating 04/21/2007
Post-mitigation problems
* Acoustic coupling to electronics in LLO LVEA electronics bay 10/13/2007
SEISMIC
S5 coupling summary Talk March 08
0-10Hz - upconversion
* Trucks produce upconversion 12/05/2003.
* Inspiral range drop with steps in the FOM seismic BLRMS trend 09/10/2004 , due to upconversion 09/10/2004.
* Attempts to reproduce and reduce anthropogenic seismic upconversion 09/18/2004.
* Upconversion from wind storm due to RF saturation 04/16/2005.
* Seismic upconversion from stack motion but not DARM_EXC injections. Large amplitude motion produces backscattering-like shelf 06/11/2005, numbers work out pretty well 06/23/2005.
* Cable brushing on the optic support structures possible source of upconversion noise 07/09/2005
* Upconversion reduced 09/18/2005
* Upconversion similar at all out-stations 01/08/2006
* Amplitude of upconversion as a function of low-f rms at one frequency 01/08/2006
* Backscattering dominates at large seismic amplitudes, but not at normal levels. Upconversion is associated with large coil drive not mechanical motion 02/19/2006
* upconversion seems associated with force on optic, not velocity or displacement 06/17/2006
* different upconversion mechanism above 10 Hz 07/23/2006
* upconversion comes in bursts 12/16/2006
* upconversion bursts come at maximum change in magnetic fields 01/19/2007
* anomalous upconversion at EY for part of S5 10/24/2007
* for more recent experiments and Barkhausen details see upcon Wiki
Mitigation/vetoes
* Equinox event unlikely to be upconversion glitch according to Masahiros flags and examination of coil signal 09/28/2008
10-25Hz - upconversion
HVAC
* HVAC limits inspiral range via seismic noise 02/19/2006
* Not reproduced by servo injections – different upconversion mechanism then below 10 Hz. 07/23/2006
* Masahiros fit of upconversion noise based on seismic noise also suggests different mechanism above 10 Hz. Ito
* Scattering (see scattering) form the test mass tables may be upconversion mechanism 07/06/2008
Mitigation/vetoes
* Flow rates reduced 02/19/2006
* Seismic level as a function of HVAC flow 03/15/2006
* Flow rates sneak up and again limit inspiral range 02/11/2007
* Reducing the seismic signal from water and air turbulence in the HVAC system 08/31/2008
Dewar glitches
* Multiple lock losses come from dewar seismic events 10/12/2002
* Summary of S4 dewar glitches at LHO and LLO 08/16/2005
Mitigation/vetoes
* Weld did not stop glitches – stick/slip may be internal 05/17/2003
* Insulation of LVEA +X dewar to reduce glitches 08/21/2004
* 200 S4 glitches from dewar +X of MX: it should be insulated 09/18/2005
* Insulation working, vibrational frequency of glitch depends on LN2 level 03/07
Drifting compressors crossing optic modes
* New office building (LSB) chiller causes 10-100 Hz H2 glitches 02/18/2005
* Staging building HVAC causes 30 minute glitches 04/21/2007
Mitigation/vetoes
* LSB chiller placed on springs, resonant gain in DARM
* Staging building HVAC shut off. More mitigation needed.
Various sources
* Broad band noise from rung up 12 Hz bounce mode 11/03/2003, reproduced with mechanical shaking at 12 Hz 01/17/2004.
Seismic 25+ Hz – direct coupling
High-f seismic from local anthropogenic activities
* Coincident H1 & H2 events from fork lift activity during S4 science run 10/16/2005
* See also scattering section
Mitigation/vetoes
* Servoed pneumatic isolation of 2k dark port
o Velocity down by 10 08/19/2005
o Reduction of air pressure fluctuations 09/11/2005
o Evaluation 09/18/2005
o Instructions for "sailing" ISCT10 10/16/2005
o Fork lift signal way down after isolation 10/16/2005
* Servoed pneumatic isolation of 4k dark port
o Factor of 10 motion reduction 11/12/2006
SCATTERING
(See also seismic and acoustic sections; this section contains studies that focus on scattering)
Upconversion from fringe wrapping
* Large amplitude motion produces backscattering-like shelf 06/11/2005, numbers work out pretty well 06/23/2005. Ground shaking to produce 10 times the normal motion of the optic brings scattering shelf cutoff up to 100 Hz, so the cutoff should have been about 10 Hz with normal motion (not seen because of low sensitivity at low f).
* Large motion from malfunctioning fine actuator causes upconversion 03/11/2007, 03/15/2007, 03/16/2007
* Large motion from malfunctioning dark port table produces backscattering glitches. 04/21/2007
Mitigation/vetoes
* S5 data quality flags for coincident H1 H2 glitches from fringe wrapping. 03/03/2008
Upconversion without fringe wrapping
* This mechanism may account for HVAC effect and other upconversion from 10-20 Hz band. 07/06/2008
* Bottom plate of optic support structures and HAM 7,8 beam dumps are probably worst surfaces for this kind of scattering upconversion 08/03/2008
* Upconversion seen in H1, thought to be scattering - model reproduces roughly, H1 peaks by shaking mid-station BSC 10/26/2008
* Worst site for upconversion is ETMX followed by ETMY, followed by ITMs 11/19/2008
* Upconversion same in DC lock 12/15/2008
Direct scattering
* Backscattering from unused ISC table responsible for large peaks 01/10/2003
* Iris and other experiments favor clipping hypothesis over parasitic interferometry. 04/19/2003
* “Black hole” beam dump responsible for 10x increase in acoustic coupling 07/10/2004
* “Black hole” as bad as white paper, black glass and razor are good. 07/17/2004
* “80 Hz bump” from backscattering from POX path. 07/09/2005
* High acoustic coupling at end stations likely to be backscattering from transmitted port. 01/08/2006
* Beam-tube bang not seen in DARM. 06/07/2006
* Motion of the vacuum enclosure for future LIGOs. 11/17/2006
* Coincident H1 H2 glitches from backscattering? 03/15/2007
* Backscattering from the vacuum envelope study I. 05/20/2007
* Backscattering from the vacuum envelope study II (including HAM5 flange). 06/20/2007
* HAM5 flange not worst acoustic coupling site. 07/18/2007
* Backscattering from the vacuum envelope study III. 08/24/2007
* Backscattering from the vacuum envelope study IV. 09/20/2007
* Photographs from inside the vacuum enclosure of potential backscattering sites ITMs. 12/23/2007
* Photographs from inside the vacuum enclosure of potential backscattering sites ETMX. 04/04/2008
* Photographs from inside the vacuum enclosure of potential backscattering sites ETMY. 04/11/2008
Mitigation/vetoes
* Original baffles removed 12/23/2007, 04/04/2008, 04/11/2008
* New baffles installed: X-manifold 07/29/2008, HAM5 08/01/2008, Y-manifold 08/02/2008
* Reducing the seismic signal from water and air turbulence in the HVAC system 08/31/2008
* New LVEA manifold and HAM5 baffles worked, scattering noise reduced 10/26/2008
* EY endcap baffle installed 12/16/2008
* EX endcap baffle installed 12/18/2008
RADIO FREQUENCY FIELDS
S5 coupling summary Talk March 08
* H2 coherence with unknown radio source. 03/15/2007
* Problems with broad band (CS) radio channels during S5 05/10/2006, 03/12/2007, 03/15/2007
* Automatic Link Establishment near LLO dangerously close to showing up in DARM 12/10/2005
* Injections of 7 modulation and beat frequencies: DARM coupling seen only for main modulation frequency. 08/05
MAGNETIC
S5 coupling summary Talk March 08
1Hz sidebands of 60 Hz from duct heaters
* Traced to HVAC 01/17/2004
* Source is duct heaters 02/13/2004
* SCRs and pulsed heating 03/07/2004
Mitigation/vetoes
* Pulsed heating turned off (building control by staged heating) 03/09/2004 without increasing temperature fluctations 04/11/2004
* Reducing loop areas of heating elements and wiring reduces magnetic field by 3. 06/12/2004
* Controlling building temperature with chilled water, not heaters 12/03/2004
3Hz sidebands from PSL chiller pulsed heating
* Source discovered at LHO 01/28/2007
Mitigation/vetoes
* New chiller type doesn’t use pulsed heating 12/05/2007
H1-H2 coincidences from power grid events
* H1-H2 events correlate with Bonneville Power Administration records, coupling is via magnetic fields 09/18/2006
Mitigation/vetoes
* DQ flags from analysis of BPA records for S5 year one 02/10/2007
* DQ flags for year two 01/25/2008
Crab killing X-arm magnetic transients at LLO
* Coherence, pedestal, failed search for source 10/13/2007
Coupling of magnetic fields to electronics (compounded by large magnetic fields in electronics racks)
* 180 Hz fields inside racks significant contribution to DARM 12/10/2004
* Coupling functions to test mass magnets and ISC electronics compared 10/24/2007
* Comparison of coupling at demod boards and optics controllers 10/24/2007
* Ambient fields inside electronics racks close to limiting sensitivity, peaks showing in DARM 10/24/2007
* Magnetic glitches inside racks show up in DARM
Mitigation/vetoes (in the future)
* We need rack magnetometers for vetoes
* Try not to crowd high current equipment together with ISC electronics
* Reduce coupling to boards
* Test new electronics for coupling and field emission
60 Hz
* Roughly ½ of 60 Hz peak in H1 gravitational wave channel can be accounted for by magnetic fields produced by currents in transmission lines 2km away. H2 60 Hz peak is twice as large and so is dominated by local sources 12/15/2006.
* Transformers near test masses can contribute to 60 Hz peak 01/18/2005
* Up to ½ of 60 Hz fields remain during power outage – confirms power transmission line hypothesis 09/28/2008
Mitigation/vetoes
* Moved transformers 15ft away from test masses (this hasn’t been done for annulus ion pump controllers).
DUST
Glitches from dust crossing beam
* First noticed 01/12/2004
* Correlation observed between dust flashes and glitches in DARM; flashes were recorded on videotape at the dark port. Correlation is best where beam on table is most narrow. 01/17/2004
* Dust at other tables does not produce DARM glitches. 04/11/2004
Mitigation/vetoes
* Purged dust cover reduces glitch rate to 0.062x10e-6 glitches/sec per particle/cubic foot. Very low double coincidence rate. No AS_AC veto. 04/17/2004
* HEPA filtered blower at dark ports reduces dust to background in only a few minutes after table entry. 12/17/2004 For picture, see here.
* Differences between the four AS photodiodes should identify nearly all dust glitches in S5 for vetoes. 04/13/2008
Cleanliness issues inside vacuum envelope
* Dust measurements in BSC8 and Y-manifold suggest purge air OK – dust on floor – vacuum every time 12/20/2007
* Dust particles on optic cage spacers 12/18/2008
WIND
* Wind produces building tilt that dominates seismic signal below 0.1 Hz. STS-2 better than guralp below microseismic peak. 04/14/2001
* Effect of wind on seismic and acoustic environment – effects us via upconversion of seismic noise. 01/11/2002
* Wind speed “histogram” for a year at LHO. 01/14/2002 better table
* Mimicking the effect of wind by position, pitch and yaw injections suggests that the main upconversion problem is RF saturation. 04/16/2005
* Comparison of vault and building seismometers at various wind speeds shows that most wind seismic comes from shaking of building. 03/24/2006
SOLAR EVENTS (RF and magnetic)
* Largest S5 solar events not seen on interferometer. 07/19/2007
SUB-MILLIHERTZ ALIGNMENT EFFECTS
Low frequency angular motion of ETM from thermal and ground tilt is less than 20 uRadians. 07/04/2008
OTHERS
* RFID system
o RFID responsible for 10 Hz peaks. 10/13/2004
Mitigation/veto
o RFID shut off during science runs
* Relative humidity for one year 07/04/2008
LINE FINDING ACTIVITIES
LIST of S5 H1 & L1 lines for astrophysical groups to veto 04/17/2010
Improved list with several more added lines and publication-ready list 01/14/2011
Earlier references need to be added
* coherence study 01/14/2006
* J0537-6910 pulsar candidate likely to be instrumental line 04/13/2006
* H1 H2 line study 07/21/2006
* Identification of major features in Stochastic group’s H1 H2 coherence spectrum 09/17/2006
* Commissioning break line search 11/17/2006 and update 12/12/2006
* Report on line finding activities from January commissioning break: power supply ripple, IRIG-B, PSL beam tubes, Neslab chillers, Penteks 01/28/2007
* Thirty lines in DARM from 3 Penteks 10/25/2007
* Post-S5 line finding report: 12 of 17 pulsar candidates identified, MC lines (clamp not working), power supplies, CPU processes 12/06/2007
* S5 Stochastic group H1-L1 lines seen in ripple monitor data 09/28/2008
ENVIRONMENTAL INFLUENCES STUDIED DURING eLIGO
Seismic upconversion
Note: also see PEM injections section here
* Early S6 PEM injections find seismic upconversion at the iLIGO level even though we replaced magnets 09/02/2009
* Reducing coil current would improve range 10/20/2009
* Upconversion noise at greatest rate of coil current change 10/21/2009
* Upconversion correlates with coil current not beam jitter 10/23/2009
* Upconversion levels from individual test masses about the same as S5 pre-magnet swap values 11/22/2009
* PAM magnet adjustment motions weakly inconsistent with PAMs being the upconversion source 12/13/2009
* Requested repeat of S5 swinging test yields same results as for S5 12/19/2009
* External magnetic injections confirm that seismic upconversion is due to Barkhausen noise, ferromagnetic candidates found inside OSEM 05/16/2010
Acoustic/seismic (other than upconversion)
Note: also see PEM injections section here
Large vibration peaks in early eLIGO DARM
* Acoustic coupling in DC lock same as S5 RF lock, except greater at resonances of some structure(s) inside HAM6 12/15/2008
* Major peaks in DC DARM likely from unsuspended steering mirror causing beam jitter (active HAM has 2 passive stages fewer than iLIGO HAM) 01/17/2009
* Lack of vertical isolation of TTs may mean that replacing steering mirror will not eliminate all peaks 02/23/2009
* Replacement of steering mirror removes large families of peaks from DARM at 450 and 850 Hz but leaves bucket peaks 03/02/2009
* Remaining bucket peaks likely increased by clipping 03/12/2009
* Thinner wires may be needed for vertical isolation March talk
* Misled by tapping tests 04/12/2009
* Bucket peaks rise and fall together 04/15/2009
* Bucket peak sidebands from various HAMs and BSCs 04/18/2009
* Nic Smith identifies bucket peaks as TT bounce modes 04/30/2009
* Variations in the outside-to-inside transfer functions explain bucket peak features 05/17/2009
* Targeted bucket peaks eliminated by replacing thick wire with thinner wire 05/23/2009
TCS (all with C. Vorvick)
* 147ish peak from TCS parabolic mirror 06/09/2009, peak removed from DARM by clamping and damping 06/10/2009
* 93 Hz DARM feature removed by turning TCS photodiodes 06/11/2009
* Minimizing 158 parabolic mirror peak, noise dominated by fluid flow down to 30 Hz, 70 Hz associated with cooling 07/31/2009
* Peak from laser periscope damped, suggestion for expansion bladder near table 12/18/2009
* 74 and 123 peaks from big periscope mirrors 02/11/2010
* 74 Hz peak driven by organ pipe
* New scheme for aligning TCS to test masses by minimizing vibrationally excited peaks 02/19/2010
Below 10 Hz
* Wooded Island earthquake swarm 02/24/2009
* 9 Hz ISI problem from office area HVAC fan belt 04/13/2009
* Fun with big waves and why the microseismic peak frequency is at twice the wave frequency 11/11/2009
* How to identify activity in LVEA: no human motion found during data period for GRB051103 04/17/2010
Repaving of 240
* Update on dates, planned treatment (BST), how much it would cost to improve treatment, and attempts at improving smoothing quality control 03/10/2009
* Schedule update 04/20/2009
* Large signal from earth-moving near Y-emd used to calculate ground attenuation 06/13/2009
Other
* A first look at PSL acoustic coupling. 06/08/2008
* Acoustic coupling to H1 is 10x worse than S5 08/31/2008
* PSL is dominant coupling site in RF 10/26/2008
* X-arm TCS acoustic coupling down dramatically after backscattering reduction by 180 degree polarizer offset 02/04/2009, coherence 02/09/2009
* Duct noise excites DARM peak 02/09/2009
* 147.3 Hz peak in DARM from scroll pump 06/05/2009, propagated through air and coupled through jitter 06/07/2009
* Low frequency TT2 drift contributes to DARM jitter peak non-stationarity, cable wiggling attempt to fix 01/21/2010
* Hungarian infrasound microphone 08/19/2010 10/13/2010
Magnetic
Note: also see PEM injections section here
* Fat 60 Hz sidebands likely from coupling of ambient fields to TT magnets, sidebands can be reduced by canceling 60 Hz motion on OMC QPD 07/10/2009
World-wide electromagnetic background (including audio frequency)
* Status on Sept. 16th, with links to other observatories 12/11/2010
Line finding and mitigation
At LHO
* Lines in DARM that are coherent with rack magnetometer signals 02/09/2009
* 5Hz comb in DARM eliminated by separate power to laser controller 05/17/2009
* H1-L1 coincident lines at 54.496 & 108.992 tracked to 7851 VME CPU 09/06/2009
* Broad peak at 122 Hz from ISCT4 iLIGO shutter backscatter 10/18/2009
* Occasional broad-band peak in DARM eliminated by tightening buzzing duct joint in duct above HAM6 (with John W.) 11/17/2009
* Peak from PSL periscope 12/17/2009
* 158 Hz peak from Foundry switcher 01/24/2010
* 2 Hz harmonics found in power supply ripple 01/24/2010
* 199.5 Hz family of jitter lines not vibrationally excited 01/24/2010
* We are no longer sensitive to rack door position (we were thermally sensitive in S5) 01/24/2010
* DetChar’s crab pulsar killing peak identified and eliminated (with Anamaria E.) 06/10/2010 06/11/2010
* 873 Hz peak from Sorenson power supply fan 09/12/2010
At LLO
* Family of peaks in DARM from fans on top of LVEA electronics bay racks 08/06/2009
* Largest rack fan peaks eliminated by seismically isolating fans 08/07/2009 08/09/2009
* 60 Hz sidebands are from magnetic fields around HAM6 – feed forward servo should work 08/09/2009
* Peaks at 306 and 390 are from PSL periscope 08/09/2009
* Peaks in DARM from leaving chiller closet doors open 06/26/2010
* Acoustic coupling at POX diode 06/26/2010
Glitch finding and mitigation
* Gremlin glitch preceded by 600 Hz oscillation visible only on PD1&2 (no longer seen after 600 Hz notch added to OMC length control) 06/08/2009
* Glitches associated with high values of side sensors at ETMY (John Zs ETMY side glitches) attributed to servo saturation – gain lowered 07/08/2009
* MICH and PRC gain increases eliminate Glitch Groups 30 Hz glitches 04/12/2010
* Testing of hypothesis that ISCT1 glitches are from RBS piezo mirrors 05/06/2010
* ISCT1 glitches eliminated by shutting down power supply for RBS piezos 06/14/2010
* “Grid” glitches, investigations pointed to OMC piezo and heater driver. Glitches stopped immediately when board resoldered 07/18/2010, 07/20/2010, 08/15/2010, 08/17/2010
Others
M10 Studies
* Coherences with DARM 02/06/2009
* Glitches correlate with vibration sensors 02/07/2009
* Duct noise excites DARM peak 02/09/2009
* Rack magnetometer results (coherent lines) 02/09/2009
ISCT4 maintenance
* Non-physical accelerometer signals 10/14/2009
* Gains adjusted 10/15/2009
* IST4 refloated after maintenance 04/13/2010
Channel maintenance
* Seismometer and radio 10/22/2009
Other maintenance
* Got microseism feedback servo going again 11/19/2009
* HVAC shutdown does not improve range 02/18/2010
Sept. 16th investigations
* Status of PEM system 10/17/2010
* Status of world-wide EM environment on Sept. 16th 12/11/2010
TCS
* New scheme for aligning TCS to test masses by minimizing vibrationally excited peaks 02/19/2010
* Shaker alignment of TCS beam 08/19/2010
INVESTIGATIONS FOR aLIGO
Coupling to squeezed eLIGO interferometer
* Squeezer table motion 01/12/2011
* Motion of 67” table 03/25/2011
* Grouting reduced squeezer table motion by 5 07/20/2011
* The impact of squeezing on vibration coupling to H1 10/23/2011
* The peaks squeezing added to DARM were removed by damping squeezer table optics 11/14/2011
AOS/SLC
* Accelerometer measurements on ITM and a pre-prototype design that moves only 3 times as much as the floor instead of 9 times (our current ITM design). 02/29/2008
* Angular noise from >0.1 Hz pier and ground motion in ITM optical lever pitch signal. 04/13/2008
* Yaw noise for the ITM and pitch and yaw noise for the ETM optical lever. 06/08/2008
* Sub-millihertz angular noise from thermal pier distortions and ground motions is less than 20 uRadians. 07/04/2008
* Angle noise from motion of 4k BS optical lever pier 08/03/2008
* Resonance of first leaning pylon is at 35 Hz – with no payload, 38 Hz 07/22/2011
* Arm cavity baffle installation, glint search, and contamination control 01/20/2012
* Beam splitter beam spot views for investigating scattering 4/10/2013
CDS
* Large magnetic fields from switching power supplies in first aLIGO I/O chases 08/18/2011
* No microphonic electronics found in aLIGO test stand 09/19/2011
*
Fans in aLIGO I/O chassis show up on channels through power supply
ripple 3/16/2012
* More on I/O chassis fans, power supplies, and magnetic field coupling sites in CDS racks 05/18/2012
ISI
* A first look at magnetic fields from actuators and electronics. 05/11/2008
PSL
* PSL table motion for adLIGO planning 12/17/2009
* Chiller closet design gave 100x acoustic reduction at 100 Hz, may work for aLIGO PSL 07/19/2010
* First test of aLIGO PSL table legs 08/22/2010
* Update on aLIGO PSL table leg installation techniques 09/12/2010
* PSL table installation successful at LLO 12/17/2010
* Interim check of LLO acoustic enclosure 2/11/2011
* Helped install and tested H2 PSL table 4/18/11
* Status of acoustic isolation of H2 PSL enclosure 09/19/2011
* Excess vibrations on H2 PSL table from crystal cooling circuit 10/24/2011
* H1 PSL table passes test 01/18/2012
* Acceptable magnetic fields from PSL chillers 5/11/2012
*
Hose quick-connects are
noisy: PSL crystal chiller vibration investigation 11/16/2012
*PSL table vibration from chiller solved by removing quick-connects 12/16/2012
PEM
* Hungarian infrasound microphone 08/19/2010 10/13/2010
* Building slab follows ground below a few Hz 04/18/11
*Attempts to explain S5 H1-H2 coherence features 08/02/2012
*High magnetic coupling at ITMY 08/12/2012
*Investigations of excess magnetic coupling 08/27/2012
*Magnetic coupling at quad levels 1 and 2 and potential mitigation 11/08/2012
*Mitigation of ~9 Hz chilled water turbulence peak in EY seismic level 11/16/2012
*Magnetic coupling investigations: masses are magnetic, fields and
gradients around actuators 01/22/2013
* Vacuum cleaning does not increase dust counts in nearby clean rooms 1/14/11
* Particle levels in cleaned BSC8 down by about 10 from 2007 12/28/2011
* Contamination control associated with ACB installation 01/20/2012
*
Black chamber-cleaning
residue can be removed but we might want stiffer gloves instead 06/14/2012
*
Particulate From New MC
Beam Tube Segments 12/16/12
* Replacing latex with nitrile gloves would reduce the chance of transferring black chamber cleaning residue 12/16/2012
Other
* HEPI motor comparison 12/17/2008
* Calibrated HAM6 table motion 02/23/2009
* TCS table motion for adLIGO planning 04/15/2010
* Tip tilt damping should be active for adLIGO 06/18/2010
* Ferromagnetic parts in AOSEMs and BOSEMs 08/22/2010
* Decommissioning of ISCT10 floating table 11/19/2010
* Disposing of 2k ETMY test mass 11/19/2010
* Vacuum cleaning does not increase dust counts in nearby clean rooms 1/14/11
* Elimination of power grid Time Error Correction should not hurt Crab detection 08/19/2011
ENVIRONMENTAL DISTURBANCE TALKS
The LIGO E2 Investigation of Non-Stationary Noise - LSC Meeting March 14 -17 2001 Livingston, LA
Air compressors, water chillers, tank shots at YFC, human activity
Environmental Disturbances: E3 and E4 Investigations
Table of environmental frequencies, proto-PEM injections
Environmental Disturbances, E5, E6 and E7 Investigations - LSC Meeting, March 20 - 23, Livingston LA Locking during wind, close construction, anthropogenic noise from vitrification plant (first sign of upconversion).
Environmental Disturbances (Including S1-Stoppers) - LSC Meeting, August 19th - 22nd, Hanford WA Wires warmed not melted, fans – electronic, hood, and nuclear plant
Progress Since March on S3 Environmental Disturbances - LSC Meeting, August 16th-19th, 2004, Hanford WA Dust mitigation, duct heater mitigation, acoustic mitigation and dewar glitch mitigation, no intersite transients seen.
S4 Environmental Disturbances - LSC Meeting, March 20-23, 2005, Livingston LA Drifting chiller, RFID 10Hz, transformers near test masses, crab protection, upconversion, why airplane signals are so bursty,
Environmental Disturbances: S4 and Pre-S5 Seismic upconversion produced on stack, peak identification, acoustic/seismic mitigation: H2 dark port floated, REFL port acoustic enclosures.
Environmental Disturbances: Early S5 and S4 - LSC Meeting, March 19-22, 2006, Hanford WA HVAC flow reduction improves range, likely from turbuence, seismic upconversion not from backscatter, not from RF sat, predicted from acceleration, reproducible with injections, H1 H2 coincidences from fork lift, table floating should help, PEM injections, rack magnetometers.
S5 Environmental Disturbances: August ‘06 - LSC Meeting, August 14-17th, 2006 @ Louisiana State University HVAC upconversion not reproducible with injections, H1-H2 events from magnetic bursts coupling to test-masses, J0537-6910 instrumental, dam sets inspiral range.
S5 Environmental Disturbances: To March 07 - LIGO/VIRGO Scientific Collaborations Meeting, Hilton Capitol Center, Baton Rouge LA, March 19-22, 2007 4k AS float, backscatter reduction, microseismic servo, dewar insulation working, extra rack cooling, chiller sidebands, HVAC update, crab protection, H1-H2 events and 60 Hz peak from power grid, dam update
S5 Environmental Disturbances: March to July '07 - LIGO/Virgo Collaborations Meeting, July 23-26, 2007, MIT Floating 4k dark port, drifting compressor and roll mode, wind farm studies, largest solar event, backscattering studies.
Environmental coupling during S5 What PEM injections showed about coupling during S5, important channels, and coupling epochs.
Mitigation of environmental coupling for eLIGO 1. Baffles to reduce vibrational coupling, 2. Reduction of seismic signal from HVAC system 3. Recent reduction of vibrational coupling at the OMC.
PEM injections for S6a and some resulting improvements (September 2009) Environmental coupling measurements and reductions in environmental coupling that resulted.
Line related commissioning and some upconversion observations (March 2010) Upconversion was not improved with magnet swap, but likely Barkhaussen.
Why LIGO's range is limited by ground motion: the likely source of upconversion (Sept 2010) Seismic upconversion reproduced with external magnetic injections that narrowed location of source to region of magnetic screws in actuator.
Environmental noise-related lessons from S6 (March 2011) Top ten list of environmental noise problems from S6 and resulting lessons for adLIGO.
Early work to reduce aLIGO environmental coupling (September 2012) Baffles, HVAC, electronics etc.
Update on PEM - related activities: March 2012 (March 2012) PEM upgrade, suggested PEM projects, PSL, CDS etc. investigations
Update on
PEM-related activities: September 2012 (Sept 2012) Magnetic coupling to quads, magnetic coupling to
electronics, other hardware activities, Maggie’s PEM web page, PEM hardware & software projects
Investigations of magnetic coupling to the quads (March 2013) Measurements of magnetic moments and relative permeabilities of quad components, and estimates of displacement noise.