|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The LIGO Executive Committee Agenda for Monday May 6, 2002 will be:
(Meeting time: 10:30 am Pacific Time)
Open meeting 10:30 - 11:30
Special Items:
Subject: Notes from LSC Executive Meeting April 26
to: LSC executive committee
from: R. Weiss April 28, 2002
concerning: Notes from the LSC Executive Committee meeting April 26,
2002
Data analysis and "upper limit" group activities (Alan Wiseman)
-----------------------------------------------------------------
An LDAS "boot camp" will be held at the University
of
Wisconsin@Milwaukee May 13 and 14. So far about
20
people have registered for the sessions.
A new release of both LDAS and LAL will occur within the
next month.
Most of the upper limit groups will have had separate face/face meetings
by May 12. After these meetings, we will arrange for LSC wide telephone
calls at either weekly or bi-weekly intervals, the rate will be determined
at a meeting of the upper limit group chairs.
TAMA/LIGO/GEO coincidence run proposal (Al Lazzarini)
-----------------------------------------------------
The draft MOU between LIGO and TAMA had been sent to the committee
and
comments from committee members were discussed. Three issues
were brought forward. The first involved the use of GEO data in the
coincidence run. In our earlier discussions with Konda we had raised
the question of LSC use of GEO data in coincidence with TAMA and that
it
would be best for TAMA to also make an arrangment with GEO. TAMA has
contacted GEO to arrange for this. The second issue is the sub group
of
LSC and TAMA people to work together on the data. The concept of the
subgroup has been accepted but the composition of the LSC component
needs
to be addressed by the LSC. Finally, it was once again asserted that
any
publications arising from this collaboration need to satisfy the publications
policy of the LSC.
Report on the situation in Washington DC (Gary Sanders)
--------------------------------------------------------
The plan to submit the MRE proposal for advanced LIGO at the end of
2002, for review in January 2003 and a committee defense of the proposal
in early spring of 2003, was discussed at a meeting at the NSF
including Bev Berger and Tom Lucatarto. The idea that LIGO is a
continuing program of improved detectors placed in the existing
facilities has not been forgotten in Washington and we have been
encouraged to go forward. The ambiance for putting substantial
increases into the physical sciences will need
to be improved, however, and we will have to mount a campaign. A useful
direction is to give colloquia at the NSF about the state of LIGO
science, technology and the work on data analysis. Thoughtful
support for the physical sciences at a congressional level might
also be advantageous.
Al Lazzarini reported that the computing resources committee was
contemplating coordination of an ITR 2003 proposal for FTE support
at the various computing centers associated with LIGO. The most
significant need is for people in operations, maintenance and help
desk support.
Further discussion of the K. Riles shift proposal
-------------------------------------------------
Keith reports that shift support for S1 is most likely in hand but
that
more structure is needed to support S2 and, especially, S3. The Laboratory
has begun including service functions in the MOU with LSC institutions.
Keith is not optimistic and expects that the issue will remain acute.
Committee chairs reports
------------------------
David Shoemaker reported that the planned frequency of the vertical
bounce
mode in the advanced LIGO suspensions has been increased from 10 to
12 Hz.
Bruce Allen suggested that we look at the changes being proposed for
the role
of ASIS at the next LSC Executive committee meeting.
Next LSC Executive Committee meeting
-------------------------------------
Friday May 31, 2002 at 11AM EDT
LIGO Operations--Administration
There was no site teleconference held on Thursday, May 2, 2002. Instead Project Controls met to discuss the proper assignment of procurements to Equipment or Allocable Supplies, and property accounting (see below.)
The list of current actions revised to reflect
the status of open actions assigned through April 18, 2001 may be found
at
ACTION
LIST.
From: Ed Chargois <chargois_e@ligo.caltech.edu>
>From: Linda Turner - turner@ligo.caltech.edu>
Web pages for the DCC give simple how-to's for document numbering, easy access to the latest on-line documents, and search capabilities for the DCC database. Take a look. . .
ACCOMPLISHMENTS
ACTIVITY
| 05/02/02 | Packages | Faxes |
| In | 48 | 33 |
| Out | 12 | 37 |
From: Esther Cunningham <esther@ligo.caltech.edu>
Press here for ACCOUNTS PAYABLE HISTORY DATA .
From: "Brambila, Ruth" <Ruth.Brambila@caltech.edu>
From: irena@ligo.caltech.edu (Irena Petrac)
CONSTRUCTION:
SUPPORT (Baldon, Torres, Lloyd, Tischler)
Progress Period from 04.19 to 04.25
Progress Period from 04.26 to 05.02
Accomplishments:
A final draft of the quarterly progress report for the end of February incorporating all suggestions has been distributed for comment. It's almost time for the next one!
Thursday, May 2, 2002, selected Project Controls personnel met to discuss the proper assignment of expenditure type to LIGO procurements. This has been complicated by the fact that we have numerous accounts open from various funding sources, and also because many of the old legacy construction accounts, while having "Fabrication" in the description, are not truly Fabrication accounts. We also switch over this year to the $5000 threshold for equipment. We must be concerned about how to assure proper assignment in the future as well as how we should treat costs since October 1, 2001 that have been assigned to the wrong type. I propose to broach the issue during the Executive Committee meeting on Monday. We should also dust off the old plan to meet with the folks preparing purchase requisitions to make sure that the appropriate account and expenditure type are used.
The following additional change requests have been submitted:
| CR-010012
Revision B |
WBS 1.4.4.1 | Closeout Construction Budgets for Initial Computer Equipment Complement at the Sites | P. Lindquist |
| CR-020006 | WBS 1.3 (OPS) | Adjust FY 2002 budget for Technical and Engineering Support Staff | D. Coyne |
We were unable to complete the discussions concerning CR-020006 during the Executive Committee meeting April 29. Discussion will resume Monday, May 6, 2002.
Press for the latest Contingency Needs Projection.
From: Cindy Akutagawa <cindy@ligo.caltech.edu>
Summary of Commissioning Activities at LIGO Hanford Observatory (compiled
by F. Raab)
-------------------------------------------------------------------------------------
(see elog for details - if it's not in the elog, it didn't happen...)
There has been no progress on locking either the 4k or 2k interferometers, but testing has revealed some clues to what problems we may have. Occasionally we were distracted by 4k input optics - particularly MMT2 - suddenly drifting off into odd alignments over a time scale of 30 minutes and relatively frequent crashes of MY ASC and FE. It is believed that a future software fix will help avoid the crashes. So far no clue on how to exorcise the evil input optics. The possible spurious IFO edge in the 4K IFO was debunked and is now believed to have been a 30-Hz elliptic filter feature in the MC electronics, made visible by a change to damping of the mode-cleaner mirrors. A major realignment was made on the 2k IFO, resulting in improved beam centering in the recycling cavity, but no improvement in locking.
4K IFO Investigations
---------------------
The first major intervention to the 4k IFO to improve locking was a
modification to the coil drivers for 4k ETMs. This provided a 3X increase
in range at high frequencies (>30Hz), while preserving overall transfer
function. This was found not to help with locking, so we began to focus
elsewhere. Bill Kells studied how saturation occurs in the 4k ETM drivers.
He found saturation typically occurs due to large signals near 340 Hz (violin
mode) but that all coils appeared to saturate uniformly. This does not
explain reports that a yawing motion is often seen as the interferometer
attempts to lock.
A second line of investigation focused on why we should be driving hard on the ETMs. Calibrated single-arm measurements were begun by Rick on the 2K and 4k systems to allow a comparison of the relative amounts of frequency noise on light delivered by the input optics. It was found that the frequency noise was significantly worse on the 4k. This is especially problematical because the 4k IFO is far less able to handle frequency noise by driving on the ETMs. Daniel estimated the residual frequency noise for either IFO in full lock and without the CM servo, and found that there was margin for 2k but that residual frequency noise would be above the acceptance band for the 4k. This alone should be sufficient to make locking unlikely. The dominant contributions were the bounce resonances between 10-20 Hz and line harmonics at 180 and 540 Hz. The line harmonics and other noise above 100 Hz were greatly reduced by modifications to cabling and filtering, leaving the bounce mode as the largest contribution to the total frequency noise. We began working on notching out these resonances to keep them from contributing to frequency noise.
Rana is also checking noise in the 1-arm locks.
Measurements of 4k BSC stack motion in-vacuo were continued in an attempt to understand the largest components of longitudinal motion of the optical tables. This used OSEM sum signals, with a calibration that corrects for the effect of pendulum motion to give optical table motion. Nothing highly anomalous was seen here. The stacks show the 5 resonances expected from the HYTEC modeling and measurement work. The lowest frequency resonances agree with expected amplitudes. Work continues to compare the higher frequency resonances with expected values.
Optical lever noise was evaluated on 4k ITMy by Peter Fritschel, using Doug Cook's optical lever calibrations. It appears to be in agreement with observed fluctuations of arm cavity power. Improvements are expected when Richard, Josh and Rana complete modifications to the optlev servoes that should get this source of fluctuation down to ~1%.
2K IFO Investigations
---------------------
Analog notched white noise injection test of 2k electronics was begun
by Josh and Richard. This follows the scheme of an investigation done earlier
by Peter and Nergis, but is a more front-to-end test. The analog notched
noise was injected into the AS_I whitening board and measured at the coil
output DAQ monitors. All notches were seen, but with varying depths. Richard
and Josh are now investigating these variations in ongoing work.
Locking on the 2K has been short and typically "glitchy", with the single exception of a single locked section in CM on Sunday evening that lasted over an hour. (There is no known reason why this was different.) There has long been anecdotal evidence that glitching in control signals, particularly MICH_CTRL, preceeded loss of lock in the 2K. (E7 data was also polluted by glitches evident in the control signals, which eventually degraded and caused many lock losses.) Bill Kells started an investigation that was followed up by Stan, Daniel and Paul Schwinberg, looking at large glitches that failed to cause loss of lock. They found that these large glitches correlate with NSPOB going negative! (This is anomalous since NSPOB should be giving a positive-definite indication of sideband power in the recycling cavity.) It was also found that NSPOB fluctuations correlate with frequency noise as measured by MC_F. This will be the subject of further investigation.
Dave Ottaway checked the RFAM on light from the 2K input optics and found it to be near shot noise at frequencies above 250 Hz. At lower frequencies the measurement was hampered by electronics noise. In other measurements, it was also noticed that the MC_F channel shows much larger 1-Hz noise on MC_F compared to data from December. A check of earlier, higher-resolution data shows 30-40X increased 1 Hz motion since a November reference spectrum. Bill Butler is resetting damping gains on the mode cleaner to see if this improves the 1-Hz situation. We will be continuing to compare currently measured levels of frequency noise to past measurements looking for other differences.
One expectation is the 2k locks should hold better when the common-mode servo is operating since this relieves the large drive requirements at the ETMs. However, we have found lately that we are losing lock in most attempts to engage the common mode servo as we make adjustments to zero the mode cleaner offset. This had not been true earlier and we do not know what has prompted the change. In an effort to relieve the symptom until we can figure out how to cure the disease, Mike Landry is making changes in the mode cleaner autolocking script to zero these offests prior to locking the full interferometer.
---------------
Facilities
---------------
The siding for the new building is on site and the contractor installed approximately 10%. The roofing is all done. The contractor is working on the interior walls and electrical rough in. The sprinkler system is complete and tested. A potential problem came to light today, there is insufficient room in front of several electrical panels on the air handlers. This is against code and it must be fixed.
COMMISSIONING IN LIVINGSTON
Noise:
> >After last week's improvements a new noise budget was established
for the
> >interferometer. The noise in bands:
> >40 to 80 Hz is currently due to the
> >optical lever damping servos and can be reduced by further low pass
> >digital and an analog filtering.
> >80 to 300 Hz is currently strongly correlated with light frequency
> >fluctuations and should be reduced by further improvements in the
> >common mode servo. New analog filters may be required in the mode
> >cleaner to accomplish the reduction.
> >Above 300Hz is due to photodetector preamplifier and analog to
> >digital conversion noise. The noise can be reduced by increasing
the
> >light power incident on the photodetectors. This cannot be done
> >casually since the quadruture signals at the antisymmetric port
> >can be large and saturate the RF amplifiers.
> >We have found that the quadrature signals at the antisymmetric port
> >are, at least in part, due to misalignment. They are reduced
> >by incorporating the wavefront sensor 1 alignment servo. This new
> >information has given added impetus to completing the installation
> >of all the wavefront sensors.
> >
> >Operational improvements:
> >The wavefront sensor 1 alignment servo was again operated succesfully
and
> >with the tidal servo we obtained fully recycled locks of several
hours
> >duration before needing to reacquire. When well aligned, the acquisition
time
> >at night can be as brief as 1/2 minute.
> >The time domain event monitor software is running and is being used
to
> >search for the non-Gaussian breakpoints in the amplitude distribution
> >of the interferometer and the PEM signals.
> >
> >A new pre-mode cleaner with optically contacted components is ready.
It
> >has higher reflectivity coatings and should increase the light throughput
> >by about 30%. WFS2 on ISCT1 is being checked using the power recycled
> >Michelson.
> >
> >PSL glitch hunting continues. Glitches in the PSL intensity are
still
> >being investigated. It is not possible to see any glitches in the
direct
> >PSL output, before any resonant optical cavities. It was noticed
that the
> >glitches became more apparent when the intensity stabilization was
> >activated - meaning perhaps that the glitches are always present,
just
> >more obvious when the intensity is stabilized. No obvious signs
of
> >glitching was present on the master oscillator pump diode current
signal
> >when monitored on a oscilloscope.
> >
LDAS:
* Installed SSH upgrade on all Solaris machines.
* Configured metaservers to backup new L?O_BETA databases
* Playing with dataPipeline and slope DSO
Laser Safety: Last Friday we held an all-hands meeting
to explain the full system operation and answer people's
questions. The presentation is LIGO-G020221-A-L. The
Laser Safety Interlock is 'officially' on-line. We're
still discovering bugs in the logical processes as the
system was received from the vendor, but we are one by
one correcting them and they are appearing less
frequently. Because of this we have yet to enable the
electro-mechanical safety shutters but this is the last
step and we will take it as soon as we are confident
that on false alarm will close a shutter and needlessly
disturb commissioning. We are planning to install
workstation software on the Operator's PC and allow each
Operator a discrete log-on name and password so that
they may attend to the administrative processes required
to clear an event should one occur on their shift.
Jonathan is traveling this week and Harry Overmeyer and
Gary Traylor volunteered to work with the installer
while he is away in order to begin understanding the
system. (Stapfer/Kern)
Laser and Optics: Ordered and set up several pieces of
new table hardware (calcite polarizers) for Gabi and
Mike Z. Rai's Scatterometers are completed and working,
but I still have to make and install apodizing stops on
the optical systems. I've got two parallel systems
jammed into the viewport camera cans, but the 6"
viewports are just a little bit smaller than the
combined aperture and clip the entrance pupil ever so
slightly. This introduces diffraction which can be seen
on the control room monitors. Proper entrance stops will
eliminate this. I've also designed a fixture that we've
been needing for a long time; that is one which permits
someone to temporarily install a camera or PD rigidly
and safely on a spare viewport. Ordered the necessary
parts and will make them as time permits. Some time ago
we decided that if and when we ever vent we should take
that opportunity to install additional viewports. Allen
volunteered to help me with this, and working with
Varian has figured out a way to use the many 10" spare
blank-offs we removed during installation. These will be
reworked as viewports saving a few bucks and expediting
delivery. These will need to be vacuum prepped so
they'll be ready to go when called for. (Kern)
External Pre-Isolation: I was at CIT the first of this
week to meet with Dennis, Ken Mailand and Larry Jones.
My task is assume responsibility for the hydraulic
actuator for the LIGO Lab so we met and I'm learning the
details of what has been done to date, and what yet
needs to be done. I am at Stanford for the remainder of
this week to receive the data package from Corwin
Hardham and Brian Lantz and get my hands dirty with the
prototype. (Kern)
·The parts for the weldment are being machined. However, the weldment is on hold until the coarse actuation approach is finalized
·Held a meeting on the coarse actuation approach. Ken Mason presented a design with pneumatically lifted, ball transfer feet which are then driven by the z & y stepper motor and linear stages from the initial LIGO SEI system. Vertical adjustment is via the V-spring position adjustment. All felt that this approach was reasonable for horizontal adjustment, but that vertical adjustment another approach should be considered. A follow up meeting is planned.
·FEA on the machined offload spring design is in progress.
·Assembly of the spring tester at LLO is in progress.
·The procurement for 3 Streckhausen(Sp?) seismometers intended for LLO was not placed. Order is now going through.
·Pump Station (Ken Mailand)
CONSTRUCTION
Basic station
layout is assembled , installation of the sensors, support brackets,and
test load is in progress.
Reservoir
is being assembled finished 5-7.
Final cleaning
and tightening of the screw fittings the week 5-6
DESIGN
The layout
is complete for the mounting brackets for accumulators, filters, and plumbing
supports using 'unistrut'
parts need
to be cut to size, and installed.
LAYOUT DRAWING
I have finished
a scale plan view and partial elevation [acad 2D] layout drawing, of station
component assembly.
HARDWARE ASSEMBLY
The three
resistors need final assembly 5-6, ref. we will use two, one at 40 and
one at 70 psi drop.
[One 40psi
backup for second RC assembly]
The accumulators
arrived Friday 4-19 are waiting on a seal fitting to arrive, ref one sent
to
Brian Lantz
to test performance and to compare with others previously tested.
LONG TUBE
RUN
Support frame
and coil assembly is finished.
ref. Test
taps at the 60 foot mark [shortest run] and 180 foot mark [Longest]
TEST LOAD
Parts are
prepared for assembly 5-2.
PRESSURELEAK
TEST
With the tank
valves closed 180 psi air will be slowly entered into the system
Tank leak
test @ ~15 psi or alt. / internal plugs for 180psi to check all plumbing
up to
tank side
ports.
David suggests using the PZT shaker on the BSC crossbeam to isolate the vertical bounce mode.
- Working
toward installing new Digital Suspension Control (DSC) software on
Hanford 4k
next week. Primary changes are:
-
Incorporation of the standard filter modules, including filters in MC LSC
path and optical
levers.
-
ETM servos all run at 16KHz (presently some at 2kHz).
-
LVEA large optic servos all run at 16KHz w/exception of sensor inputs
(performed
by SOS controller due to wiring). Optical lever inputs will move from
SOS controller
to LOS controllers.
-
Inclusion of new Pentek timing modules.
Status of
the upgrade:
-
Alex has Small Optics controller running; in process of modifying code
for use at
40m lab. If complete by Friday, may have Ben and Jay install SOS
system in
40m lab next week.Alex and I first
need to upgrade the 40m DAQ
system to
accept the DSC (hoping for Friday). If we can't complete Hanford code
and 40m code
by Friday, will install at 40m lab week after next (May 13).
-
I completed the EPICS code for the LVEA LOS. Khan is finishing the
Pentium code.
-
I am about 90% complete with the end station EPICS/Pentium software.
Alex, Khan
and I will be at Hanford next week to install and test the software.
We hope to
stage the installation, running the IFO after each step to verify
code doesn't
cause problems which prevent locking or introduce extra noise. Plan
is to start
with ETM controls, then LVEA.
I would also
like to install the new Epics Data Collection Unit (EDCU) for
testing next
week.This system is an upgrade to
allow up to 800 EPICS channels
to be acquired
at 16Hz and 10000+ at 1Hz.
New ASC code
for the LHO2k (same upgrade as running at LLO) and new LSC code
(same as running
on LHO2k) are almost ready to go (a few EPICS changes). If time
permits, may
load these as well next week.
- Hongyu is
working on dataviewer changes as required to read the new DMT trend
files.This will also go in at LHO next week.
Differential
Driver/Receivers: All 4 boards are at the fab house and should be back
early next week. They
will be stuffed
and tested prior to shipment to the sites.
Optical Lever
PD amp: The optical lever PD amp board has been modified to accept QD100
or QD50
photodiodes.
Boards should be back by next week.
Rich Abbott
1.Intensity
servo board is being stuffed today 2 May, the board will be tested next
week in the 40 meter lab.Site installation
is to begin at LLO during the week of the 20th of May.The
week after we will install at LHO.
2.The variable timing board has been declared a success during tests here at Caltech.The build is almost complete in the shop.
Mike Zucker
LSC photodetectors:
Repaired/restored
5 dead/failing
photodetector
modules by salvaging older-vintage photodiodes
from other
equipment.Heroic measures
to resurrect
some built with recent (fall '01) diodes just
certified
that those from that batch which haven't yet failed
spontaneously,
will nonetheless fail soon under thermal
cycling. As
a result no more of these questionable diodes will
be installed,
period. Renewed efforts to engage manufacturers
in solving
the problem; emphasis is to determine whether new
diodes procured subsequently will suffer from this defect.
The experimental
VCO circuit I have been working on still needs the
traces for
the power to be routed, after having painted myself in a corner
in order to
get to some components.Hopefully
this will be finished later
this week.
The
optical contact for the first optically contacted pre-modecleaner
was broken
and subsequently re-established.In
the past week, three
prisms and
bodies were contacted together.All
the contacts so far appear
to be good.Good
enough to survive trying to pull them apart by hand at
any rate.One
pre-modecleaner is in the final stages of assembly - waiting
for the Vac-Seal to dry on the PZT - before some further testing.
Physics Letters
A approved the paper
'Considerations
on Parametric Instability in Fabry-Perot Interferometer'
and I received
communication of the decision of the Editor (no comments).
For the E2E
collegues I checked how large a perturbation can be,
to be treated
properly by TEMmn modes with m+n<=6 and I got the value 500m. in
the change
of the radius of curvature.
I made more
several tests on the Sidebands Imbalance using the FFT-code
and I got
results far different from the ones obtained by Melody when
the recycling
cavity "only" is considered.
OTF Lab.
Contamination
Cavity # 2
We continue
taking data daily with a mineral oil sample in it.The
ring down is the same 30us. (33us before sample in) the beat frequency
is 74.42 MHz.We found a small discrepancy
on the ring down, As we re-align the cavity, the ring down went up to be
31us. (different location of the beam spot on the mirror).We
have measured the absorption loss to be ~ 1.56ppm (before 1.3ppm Jordan's
paper).
Contamination
cavity#1
We have tried
every way to get transmission power in this cavity and the most we got
was 13mw (power in ~ 110mw)The visibility was no more than 15%.We
made sure that we had a right mode match every time (we checked it with
beam scan to verify beam waist wanted) a very good alignment, we have changed
several times different lens and still the same result.The
ring down continue to be as before 48us.The
transmission power and the dip the same.After
all these efforts, We decided to replace for the other cavity and after
setting all up, we locked the cavity and we have an output power of 55mw
with an input power of 110mw.the
visibility is ~ 70% and the cavity is locked as right now.The
ring down is 23.32us.
We will optimize
this cavity for transmission power and visibility.
Last week
I worked with Dave Barker and Alex Ivanov to install and test a
new network
data server that has been modified to disseminate DMT trend
data> This
enables users in the control room to view DMT-generated trends
(e.g. band
limited RMS trends, etc) with the standard data viewer utility.
After modifying
the DMT trend generation class to interface to Alex's new
Network data
server and installing the new Frame builder on fb3,
everything
worked almost immediately.
A few points
are left for further work. Some mechanism must be implemented
to allow the
users to find the DMT trends more easily. At present the
trend channels
are mixed indiscriminately with the raw data channels. A
naming scheme
for the trend channels could be sufficient to resolve this
difficulty.
There is also an error somewhere in handling trends that have
zeroes in
the first few time intervals.
Daniel and
I then installed Solaris 8 on sand and stone and reconfigured
the network
attachments. This brought the two machines into a well defined
software state
necessary to assure compatibility between the Solaris DMT
nodes and
resolved some potential security issues.
Since returning
to Caltech I have been documenting operational and
diagnostic
procedure for the DMT systems and working on Time-Frequency
plots of AS_Q noise when a glitch is seen on some other channel.
This week we built, tested, and installed a second set of notch filters
and locked both arms simultaneously. We can now achieve full lock during
the day, even when the MINOS crew is using the crane, with acquisition
times of a few minutes. However, holding lock in a noisy environment
is
another matter! Hold times appear to be alignment dependent,
and we are
currently tweaking the alignment of all three cavities to maximize
our
duty cycle.
The North Cavity, with its new photodetector and notch filters, works
wonderfully. It acquires quickly and holds lock for an hour or more,
and
its noise level is about a factor of 3 higher than our best curve in
the
South Cavity.
The South Cavity also acquires quickly, but tends to fall out of lock
after
about five minutes. Moreover, its noise floor has come up by a factor
of
3 and is now comparable to the North Cavity's. We are now looking at
various noise sources to try and improve the sensitivity in both cavities.
LASTI (Abbott, MacInnis, Mageswaran, Mason, Mittleman, Ottaway,
Rollins, Shoemaker, Zucker)
=======================================================================
VACUUM ENVELOPE:
Myron closed up and pumped down after the recent round of stack
impedance measurements wound up last week.
A soft LN2 dewar caused a seal gas shortage early this week,
triggering a pump shutdown. Safety systems performed nominally
(i.e., the gate valve shut!), but it got us thinking that this is no
way to live. Myron has spec'd out a fixed air filter/dryer
system to provide a more reliable source of seal gas for the main
forepumps. We also decided to add another relay contact to the
interlock system so shutdowns like this will be less stressful
on our turbo.
EXTERNAL PRE-ISOLATORS:
Rich continues to analyze the volumes of data taken in the stack
characterization test. We've discovered at least two of the
21 sensor channels were not behaving correctly,
but it isn't yet clear whether there's sufficient redundancy to ignore
them or if we will have to go back inside and repeat some measurements.
Ken designed a coarse positioning addition to the EPI support structure,
plus
a kinematic adjustment system for centering and thrustline alignment
of the
MEPI actuator. These were reviewed and approved for detailing and fabrication.
Fabrication on the main weldments is proceeding on schedule, despite
buyout of
one of the main contractors (by a losing bidder, as it happens!)
Myron is working on a HAM installation fixture concept employing
precision
screwjacks and electronic dial indicators for position repeatability.
On the EPI electronics front, Rich Abbott reports:
1. Mohana has finished the schematic of the rack mount portion
of the
seismic retrofit electronics. PCB layout will begin shortly.
2. The field interface board for the L4C and Inductive position
sensor
is
complete and being sent to the board house. Chassis is on order
for the
field unit.
3. Pump servo conceptual design is well underway. Will finish
the
conceptual design 3 May 2002
PSL:
Jamie is working on intensity noise characterization of the PSL.
Also
spec'd out a custom EO modulator from New Focus
with angle cut ends and AR V-coatings to improve optical isolation.
Simulation and Modeling (Bhawal)
New release schedule
----------------------
(Hiro)
Next major release of the e2e package is scheduled next week.
The simulation engine of this package is greatly revamped, with new
and improved
features as well as bug fixes. This is necessary to run the next major
release of
LIGO simulation.
Another major change is that a JAVA-based alfi5 is the default GUI front-end,
instead of alfi4 written in C++. alfi5 is used internally, and is quite
stable.
This will come with several revised documents, and the installation
process
will be simplified.
LIGO I simulation system
--------------------------
(Matt) Locked SimLIGO and produced preliminary noise curve (includes
seismic noise, shot noise, and many technical noises).
Next step is to add the common mode servo and frequency noise.
Validation of Full IFO run with mode mismatches
------------------------------------------------
(Biplab, Matt, Hiro)
The Han2k model is modified to have an IFO of length 139 meter. Our
aim is to
reduce the build-up time so that a model that uses only primitive modules
(and so runs slowly but gives accurate results) can be run for shorter
duration
to get locked state results and can be compared with model that uses
summation cavity.
Changed gain parameters appropriately to lock the model.
For a case which considers as-built differences in arm mirrors and some
amount
of thermal lensing as well, some differences between primitive and
summation runs
are found in a mirror position and in power at some points.
Code development and maintenance
---------------------------------
(Biplab) Committed several recent changes to CVS.
Updated primitive reference document.
(Ed Maros) worked on build-e2e script to simplify building of e2e package
for
end users.
Alfi
-----
(Bruce) Out sick 2 days.
- Completed reimplementation of connections in Alfi5.
- Misc. problem report bug fixes.
(Melody)
-Fixed more problem reports for alfi5.
-Updating the 'ALFI5 Tips and Shortcuts' document to include the latest
enhancements.
LIGO Data Analysis System
Software Systems (Blackburn)
The first beta release of LDAS (0.2.0) made it out of Millikan this
week. It is now running on all LIGO LDAS systems (except the LDAS-DEV
system which is already developing towards the next releas). In
addition, this release of LDAS is running on the UWM cluster and
being used by the LSC. This release was tested in conjunction with
a new release of LAL and LALwrapper. Many of the search code authors
participated in this joint endeavor and greatly improved the final
product. This is a major milestone for the LDAS software and many
thanks and congratulations are in order to allthose that helped make
it possible. There will be one more release of LDAS in preparation
for the Science Run.
As part of the new release of LDAS, all LIGO LDAS systems had their
various counters reset to one, job and log file directories cleaned
out, and a new database created.
One of the most enlightening aspects of the testing that went into
this release of LDAS was the discovery of several issues with the
structure used to store detector geometry in the frames and the
handling of this within LDAS. It was discovered that the latitude
and longitude values were reversed in the Hanford and Livingston
frames, an artifact which exists within the E7 frames. A "fix" to
this was added to LDAS so that it could correctly pass along these
values to downstream code such as the DSO search codes. It was also
discovered that all frames had the detector geometry labeled the
same and that one of the azimuth angles were incorrect. There was
not enough time to resolve these in the release. But the framebuilder
has subsequently been fixed to provide correct values for all these
attributes.
Lazzarini note: We are now working to amend the specification to correct the shortcomings. This will involve an iteration with Virgo.
A lot of effort went into improving the performance of LDAS as well
as the reliability for this release. We are optimistic that we will
see a higher percentage of jobs complete successfully than was
encountered in the E7 run. We also worked to improve the performance
of the new diskCacheAPI by two orders of magnitude between the pre-
release versions and the final release of LDAS. It has also been
reported by Greg Mendell that the Periodic search code used to
generate Short Fourier Transforms (SFTs) is running between 1.5 and
2.0 times faster than it did in the E7 run.
Another significant new feature in this release is the ability of
LDAS to generate reduced data sets in Processed Frame format. This
functionality has been demonstrated to work and the University of
Oregon team has participated in testing it. There are a couple of
issues that should be addressed in the next release of LDAS which
will make these reduced data objects more useful. However, one of
the major concerns will need a modification in the Frame Spec to
fully correct.
Hardware Systems (Anderson)
Caltech
-------
(Dan Kozak)
* Continued HPSS migration from Redwoods to 9940s (stopped 3590 work
for
the moment as per James Patton's instructions), HPSS user directory
reconfiguration (moving towards no data in user home directories)
and
account maintenance (mostly retiring accounts).
* Still working to get StorageTek to make good on replacing the Ultra 10.
* LIGO SAN testbed:
Stress testing SAM-QFS, filled disk space and ran
it out of
tapes and it failed gracefully. Successfully
built new
filesystem using existing .inodes file. Now
testing other
failure modes and disaster recovery mechanisms.
(Al Wilson)
* Compiled a test version of aide (tripwire) for Linux testing on m22.
* Set up system for new hire Eric McWhorter - Welcome Aboard!
* Troubleshooting the differences between cfdef4 and ver 5, have reached
a
dead end. additional input needed.
* Revisited the Kickstart setup with the latest version of Red Hat.
(Stuart Anderson)
* Helped manage the LDAS-0.2.0 software release.
* Attended the Teragrid meeting at Argonne.
* Finished activating zlib security patch on all LDAS Solaris servers
at Caltech.
* Brought new datacon server on-line in the LDAS-CIT system.
MIT
---
(Keith Bayer)
* Installed ssh (zlib) patch on solaris boxes.
* Ran cable from ldas lab to highbay control room.
* Acquired wall boxes to punch down cable.
* Moved lancelot DMT machine to ldas for conference.
* Loaded SRDS data onto T3s.
Livingston
----------
(Igor Yakushin)
* Installed zlib security patch on all Solaris machines.
* Configured metaservers to backup new LLO_BETA databases.
* Playing with dataPipeline and slope DSO.
Hanford
-------
(Greg Mendell)
* Applied the zlib security patch in 112611-01.zip on all Sun boxes
on
the LDAS network at LHO. (Reboot of these was done with
the new release
of LDAS.)
* Assisted with the release of the new 0.2.0 version of LDAS at LHO.
* Worked with Dave Barker to check that the frame header/detector
structures are now correct. In general, we are discussing
various
issues to prepare for E8 and S1.
Lazzarini:
Lazzarini and Anderson represented LIGO Lab. at an applications workshop
for the NSF Teragrid program. The meeting was informative in that we came
away with a much better awareness of the Teragrid plans and timelines.
The program is prototyping new 64 bit intel hardware at each of four sites
and LIGO will likely be invited to participate in initial activities to
develop working data analysis codes for teragrid. A simple, productive
approach appears to be to port the Hough transform paradigm. To this end,
in preparation for the meeting GEO principals developing the Hough code
contributed information on its implementation and how it might fit into
a Teragrid computing model. In addition, it appears that, through NCSA,
it may be possible to obtain FTE support in porting the code to the new
architecture.
Data Analysis Activities (Lazzarini)
Mendell:
1) Worked with LDAS at Caltech to assist with the debugging of the
reading of frequency series data from the frame proc structure.
This
was not entirely successful, and more work will be needed to get this
to
work before the June release.
2) Started regenerating E7 SFTs at LHO and LLO using the corrected clean
locked segments. A new bug (PR 1534 submitted by Peter Shawhan)
in
which processing of a job begins before a file is transferred from
a
remote site has caused some jobs run at LLO to fail when submitted
by a
script running at LHO. I am currently working to get around this.
3) Working on an update of the SFT specification, and preparing for
travel to Milwaukee May 13-17 to work on the knownpulsardemod DSO.
Shawhan: Preparing to travel to UWM for the inspiral upper limit meeting
Lazzarini; Ditto for UTB and stochastic background meeting
General Computing (Wallace)
MIT:
(Keith)
-Activated wall jacks in NW22 rooms
-Working on ghostscript filter converter for inkjet printer
-Received 2 SB1000's for virtual control room
-Packed and moved some equipment to NW22
-Installed 2nd hdd in sb100 for system backups
Livingston:
Nothing to report at this time.
The sysadmin will be out for a few days. He had a little incident with
a
nail-gun at home.
http://www.ligo-la.caltech.edu/~sroddy/xray1.png
Hanford:
(Christine)
- Installed fast ethernet into some of the offices.
- User support for pcard software problems.
- Continued setting up new PCs and cleaning off files from old PCs.
- Continued with inventory of software and hardware on site.
(Larry)
- Had another meeting with NoaNet concerning the WAN connection at
the Hanford
Observatory. This time Albert was able to attend. A couple of different
connection options are being investigated by NoaNet. We should have
more
information by the end of the month.
-Going over the network statistics and information that Christine generated.
The
T1 line is definitely not enough bandwidth.
CIT:
(Mike)
-Loaded two PC's for visitor's workstations that included full reloads
with
Operating System and General Computing software.
-Setup an office for a visitor workstation that needed a PC, in place
of
the Sun workstation.
-Check the following documentation on the internal web.
http://docuserv.ligo.caltech.edu/docuserv/computing/win_y2k.html
This is to insure that users who are loading their own computers will
follow these few guide lines to help make sure our Network will not
go down
due to Master Browser Elections and to keep your computer from trying
to
register your address to the DNS table. If anyone would like me to
back up
their Laptop or Computer by getting a Ghost Image, please contact the
sysadmin group and we will make the necessary arraignments to work
this
into our schedule.
-Finished up loading and creating Ghost Images of two laptops and have
them back into the loaner pool.
-Started loading another laptop but still have GC software to load.
-Showed Veronica and Erica how to used VRVS to prepare them for a LIGO
Seminar this Friday 5-02-02.
-Started loading a New PC for Wilson House User.
-Onsite User Support,
This included some hardware failures, software issues, networking problems
that were connectivity issues.
(Lisa)
- Began building an independent test string so that I can test changes
to
mail,
web servers, new versions of dns, etc., without using the production
environment. In addition to creating a good testing environment,
this
information will also provide a procedure for doing a full disaster
recovery
of
sirius, acrux and ligo.
- After some extensive testing, I made changes to the way spam &
virus email
is
blocked at the mail server. Noticed some anomalies with triadpm.com
trying to
route mail through acrux.
- The usual round of new accounts, mail aliases, etc.
(Veronica)
- LIGO website: Working on the Press Kit website. Met with Linda and
Ryan
on the status of the press kit, both the online version and the handout.
Discussed with Linda the need to implement a document management tool.
Linda offered to contact an independent assessment contractor, to
determine what software best to be used for this.
Posted various updates to the LIGO website (Fellowships, Elba Conference
webpages, MOUs, Resreach Bulletin Boards).
Worked with Mike and Larry on the VRVS use for the upcoming LIGO seminars.
- CaJAGWR: updated the seminars listing.
(Larry)
-Working a number of procurement items. Purchases for 40M, LDAS and
a few
items
for GC. Working on upgrading the video switch in the server room.
-Worked with Mike and others on setting up the CIT Control room. Testing
of
the
system will be in the Small Conference Room, that room is now reserved
for the
next three months for that purpose.
-Reviewed some of the computer audit procedures that Caltech performs.
These
will be checked out at MIT and the Observatories over the next few
months.
-Worked with Mike and others on setting up the VRVS for the upcoming
conferences.
-Reworked the computer room. Presently, there is no more room for more
computers. Also, will be purchasing another UPS to handle a couple
of the new
servers.
-Working with Thomas F. on a new calendar tool. The tool is called
Calcium and
appears to be platform independent.
Lazzarini;
Wallace and Lazzarini held a teleconference with representatives of
Washington State's NoaNet telecommunications organization. Discussion was
what upgrade paths exist for LHO to move beyond T1, their relative merits
and costs. Next step is to arrange a 3-way meeting among LIGO, NoaNet,
and PNNL/DOE. Tentative date for this meeting is 31 May.
Minutes of telecon May 14 th 2002:
participants: morning call: Jesper Munch, Maik Frede, Martina Brendel,
Ralf
Wilhelm,Carsten Fallnich, Benno
, evening call: David Shoemaker, Shally, Ralf, Martina, Rene, Benno
1) updates
Adelaide:
- decision was made to buy new fibres
- laser head is together, resonator aligned, expect lasing experiments
to start tomorrow
- interferogram shows straight fringes (maybe small positive horizontal
lens which should
not be a problem) up to pump powers of 100W
- new jig designed for laser head assembly
- 10W system performs well, first lasing experiments yield 13W in multimode
operation
for 40W pump power
LZH
- started to include depolarization in modelling effort, first test
will be to try to describe the
saturation of a test set up for the 20W class front end (GEOtype laser
with undoped end
cap rods)
- first test with a high power rod pumped through both end faces with
125W (250W pump
power together) gave 65W output power with an M 2 between 5-7, next
step is to insert
Faraday rotator into the standing wave cavity to try to reduce depolarization
effects and
improve spatial profile and output power
- experiments to spatially homogenize the beam behind a bundle of fibres
with a quarz rod
of 5cm length show very good results, it is planned to use such a homogenizer
to pump
the high power stage
Stanford
- power supply for Lightwave 20W amplifier repaired
- single passing this amplifier with 4.5W out of the LIGO type laser
as an input gave 20W
output
- first Stanford amplifier head is aligned and Shally will start experiments
with that head
today: goal is to get about 40W for 180W pump power
- after a characterization of that beam he plans to inject it into
the second Stanford head to
get about 100W
next conference call: June 6th
From: Mark Barton <mbarton@ligo.caltech.edu>
This week I worked with Phil Willems to do some detailed validation
of
the thermal noise predictions of my Mathematica model. We found one
embarrassing error common to the quad, triple and toy models: I had
mixed
up the tags specifying the damping, so that wire bending elasticity
was
having longitudinal damping attributed to it and vice versa. Once that
was corrected, we found that for each degree of freedom in a one-mass,
four-wire toy model, the thermal noise predicted from the model agrees
very closely with a very simple analytic prediction using the fluctuation
dissipation theorem for a one-dimensional mass-spring system with obvious
substitutions for mass, elasticity and damping. We also looked at a
quad
system similar to the MIT prototype except with silica fibres at the
lowest level and found that the thermal noise was very large at low
frequencies, but asymptoted very quickly to exactly the thermal noise
that would be expected from considering the last stage alone. I've
posted
the updated v2.2 model used for these tests on my models page (<http://
www.ligo.caltech.edu/~mbarton/SUSmodels/asus4/>) along with a Mathematica
notebook with just the thermal noise results <http://
www.ligo.caltech.edu/~mbarton/SUSmodels/asus4/ThermalNoiseTest.nb>.
From: ctorrie <ctorrie@ligo.caltech.edu>
Advanced LIGO Suspensions
On Monday Janeen and I held a review of the drawings and ideas for the Mode Cleaner Prototypes for LASTi. Norna, Phil, Mark, Larry, Dennis, Ken and David Tanner and Dave Reitze from Florida contributed comments and suggestions.
For the rest of Monday and Tuesday I worked with Larry Jones discussing in more detail some of the ideas from the drawing review.
On Wednesday 1st of May I leave for Scotland and the Institute at Glasgow University. My next weekly report, for 3 weeks, will be on the 22nd of May.
The various parts I have in workshop, prototype parts for the MC suspension,
should be ready for my
return. Some of these will be shipped to me in Glasgow by Janeen.
From: Janeen Romie <romie_j@ligo.caltech.edu>
AdLIGO Suspensions
Held a drawing review for the MC suspension so far. This was in preparation
for Calum's trip to Glasgow today. David Tanner and Dave Reitze brought
up some issues that Phil W. addressed. The MC optic is 15cm dia x 7.5 cm
thick. As designed, the optical aperature is not obstructed. No comments
on Phil's MC reaction chain (or lack thereof) paper so far.
Larry W. is putting all of Calum and my Solidworks and AutoCAD files
in one location on sirius so we can access them together without confusion.
Working with Norna on the RM layout.
From: Riccardo DeSalvo <desalvo@ligo.caltech.edu>
Michael, Hareem
More trouble with cryostat’s vacuum. Waiting for replacement electronic
board for cryostat to control valves. Will test it asap.
Michael
Mapping the cryostat measurement setup, preparing for thermal
capacitance measurements.
Hareem
Old braze found polluted by oxyde, re-purified gold from braze, made
new
batch (Au71Sn19)99Ti1. Will test it next week.
Ordered some AgSnTi braze from commercial source.
Aso
Calculating folded pendulum suspension thermal noise with analytical
formula. Parameters to be introduced, then check validity.
Eventually will repeat exercise with MSE II and cross check.
Kelin
Molded several MoRuB samples into ribbons, prepared several beads for
levitation and splatting, received replacement splatter ceramic sample
supports. Ready to test splatting again.
Studying thermal capacitance theory of samples.
Akiteru
The cavity is still largely misaligned. I am implementing
another feedback using the suspension platform actuators, which I did
not do because I was not very sure of their reliability, but it was
a
stupid
attitude. Let's see how it goes... The problem is that the initial
alignment of the cavity was much more difficult than I expected. In
any
case, I hope to send you 'real' good news soon.
Ric
Pushing papers.
Collected the comments of the final draft and forwarded to NIM the
two
sensor and actuator papers.
Paper pipeline: IP measurements (Szabi), MGASF measurements
(Virginio), MGASF simulations (Virginio, Giancarlo), Horizontal
accelerometer (Alessandro), 3 m hyerarchical controls (Akiteru), AID
remedial option (Virginio, Szabi), SAS pre-isolator remedial option
(Riccardo).
Set a target/deadline of end of May for the first 4 papers. Akiteru’s
paper(s) have an automatic deadline given by the defence of his
thesis,
the other papers will come later.
For additional information about this report, contact sanders@ligo.caltech.edu