|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The LIGO Executive Committee Agenda for Monday June 25, 2001 will be:
(Meeting time: 10:30 am Pacific Time)
Open meeting 10:30 - 11:30
Special Items:
no report
WBS 1.2 LIGO Operations--Administration
There was a teleconference on Thursday, June 21. The following were among the issues discussed:
From: Ed Chargois <chargois_e@ligo.caltech.edu>
>From: Cleveland Mak <mak_c@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
| Packages | Faxes | |
| In | 33 | 41 |
| Out | 10 | 40 |
Press here to access the DOCUMENT CONTROL CENTER WEB PAGE.
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)
SUPPORT (Wood)
Progress Period from 06.15 to 06.21
Accomplishments:
WBS 1.4.1.2 Project Controls (LIGO Construction)
The end-of-May Quarterly Progress Report is due. I have asked for contributions by Friday, June 22, 2001. Again, please keep it brief and report only significant accomplishments (not in progress stuff).
The following Change Request has
been submitted:
| CR-010006 | WBS 1.1.4 (OPS) | Roof Shelter for Concrete Pad Adjacent to Erosion Control Pond (Livingston) | M. Coles |
Press for the latest Contingency Needs Projection.
From: Kris Duncan <kris@ligo.caltech.edu>
From: Ed Jasnow <jasnow@ligo.caltech.edu>
No report this week.
General Items:
--------------
(F. Raab)
Seismic Systems
---------------------
H.Radkins, G.Moreno, M.Guenther, F.Raab
Coarse Actuation System
The portable coarse actuation system has been exercised and revealed
a
couple important things: The grounding of the rack is very important
to
get quiet operation of the system and we noted an intermittant
overtemperature on a few of the W (vertical) motor step drivers.
These
faults trip the driver and stop the motion. In general the system
and
components functions well.
Eucon Mining Studies
--------------------
Raab attended a meeting with the Port of Benton presenting results
of the LIGO
investigation. (No effect above ambient noise was seen at LHO TRI-NET,
consistent
with attenuation estimates obtained from field measurements by Radkins
and Rohay.)
The City of Richland has also been notified of the result (L010100-00-W)
in
conjunction with the permitting process for the mine. We have asked
the City and
the Port to include provisions in the permit for future certification
by LIGO that
any changes in equipment or in location of equipment at the mine do
not change
vibration levels coupled into the ground.
ISC
-------------------
H.Radkins, G.Moreno
All the optical levers in the corner station are now online and their
beams
and viewports are fully enclosed.
ISC/IO Table work
-----------------
C. Gray
The laying out of all the optical components (which we have) is completed; this table is ready to be positioned adjacent to HAM1.
The ISC/IO tables for the 4k do not have spacers for their respective table legs (needed because the tables are a few inches low ); have contacted Ken Mason about getting the ball rolling on that.
Optic Q measurements
------------------------------
(M.Barton, C.Gray, M.Lubinski)
Mark Barton has developed code and procedure to determine the "swinging"
mode Q's for our optics. This week was devoted to getting familiar
to this code and procedure to obtain Q-values for our optics (LOS
& SOS; 2k & 4k). Thus far, we have been able to look at some
of the 4k SUS's.
Air handlers: All air handlers are now runnning with replaced bearings. Two of the units have a second motor which is down and will be replaced. Estimated delivery is 4-6 weeks. In the meantime, everything is operational. We still need to investigate the operation of the 4-20 mA controllers which actuate the variable pitch on the fan blades. At present, these are all locked out and the fans are operating with the blade pitch fixed mechanically so that we can operate normally, but we need to look at this next.
Detector: ITMx and ITMy have been aligned. The three mode cleaner optics and the BS are in the vacuum bake oven at present. The Y end station is vented. We found fine dust in this chamber as well. ETMy was inspected to see why it exhibited a low Q in the "bullet" mode. There was nothing obvious in the way that it was suspended. Several people, including Rai Weiss, looked at it and the consensus is that is appears to be OK. The OSEMs will be replaced and the survey equipment set up for re-alignment.
GC: Installing the tape backup server software on its permanent
home in the rack. I am satisfied with the testing that I performed with
the Solstice backup suite and I will begin using it this weekend.
Installed a 16 port KVM which is working great in the communications room.
LDAS: I let the burn-in test script run over the weekend and
from what I have looked at so far, I cannot see any obvious problems. I
was able to successfully start the script from home Thursday night on all
16 nodes and Beowulf (Shannon Roddy).
PEM: The BSC interface plates on BSC 1 have been installed so that the
accelerometers can be mounted when the spool piece is replaced.
The spare seismometer has been moved from the Y-end station to the
LVEA. It is temporarily using the LVEA voltage monitor channels to record
data. The spare seismometer is under HAM 1 and the LVEA seismometer has
been moved to between HAM 2 and HAM 3. The two seismometers will be used
by students in the next few weeks to compare the ground motion in the LVEA
and the instability in the mode cleaner. (Doug Lormand)
Operations: Set-up supplies and equipment to perform a particle shedding test on all the different type clean room garments we use. Assisted Joe Hanson with pick-up and delivery of air handler motors. I am attempting to streamline shipping and receiving procedures and practices. (Rich Riesen)
|
|
|
|
Installation&
Commissioning:
|
Other
Science/EngineeringActivities:
|
See also the Installation web page
This week parallel efforts went on to get both the
2k and 4k mode cleaners working and characterized.
2k mode cleaner (Fritschel, Mavalvala)
(i) High power operation: To operate at high power,
the EO shutter was commissioned on IOT7 and is set for safe operation at
1 W. The mode cleaner was found to remain locked stably as we increased
the power from 0.25 W to about 7 W . We did, however, find that the MC
did not acquire lock very easily at powers above about 5 W. This is still
being investigated. As we increased the power we noticed that the MC length
control signal was steadily growing. Working under the premise that the
MC mirrors were being pushed out by radiation pressure in the cavity, we
measured the change in control signal per increase in light power. This
turned out to be much larger (factor of ~6) than one would calculate for
radiation pressure, which is currently a mystery.
(ii) 1.06 um light coupling to new OSEMs. We measured
this coupling for the MC mirrors with the MC locked with 1W at the input
and found, in the worst cases, about 15 nm/W of input power, which is about
a factor of 300 better than similar measurements on the previous OSEMs.
(iii) Angular fluctuations at MC output: We set up
a quadrant photodetector at the output of the MC and measured the beam
pointing fluctuations. They were ~0.5 urad rms after tuning the damping
gains for the MC mirrors.
(iv) VCO phase noise measurement: this was done by
beating the VCO with another oscillator and found frequency noise on the
light due to VCO phase noise to be about 15 mHz/rtHz at 1 kHz, which is
quite consistent with the typical 2 mHz/rtHz noise floor we have been measuring
at the MC control signal.
Other 2k MC clean up activities included tuning of
POS output matrices for the MC mirrors.
4k mode cleaner (Matone, Kells, Mavalvala)
(i) The input polarization was set
(ii) The MC cavity pole was measured to be 3.3 kHz
(iii) MC resonant reflectivity was measured to be
~10%
Suspension mode Q's for the MC mirrors are being measured
(Barton, Gray, Lubinski)
Digital suspension filter measurement (Mittelman,
Fritschel)
Modeling and measurements were done to find the best
filters for filtering both sensor noise and DAC noise in the digital suspension.
New filters have been loaded for one of the 4k MC mirrors, results are
pending.
Core optics alignment (Kells, Whitcomb)
The 2k power recycled Michelson was aligned and locked.
The 4k core optics alignment was checked by positioning
the input beam on the gate valves in the mid stations and retroreflecting
the beams from the ITMs onto the reflection port. It was found, however,
that it was not possible to align the recycling mirror with the ITMs. Presently
the plan is to increase the bias range for the ITMs and RM electronically.
This week I've been at Hanford working on the suspensions.
I ended up
working mostly on the Q factors of the optics. I adapted
the input matrix
tuning software to do ringdown measurements and showed
Corey Gray and
Mark Lubinski how to take data and process it. Typical
small optics Qs
are around 7000 for position and 3000 for pitch, yaw
and side. However
the pitch, yaw and side Qs are somewhat amplitude-dependent
- more lossy
at high amplitudes. See elog entry (user: reader, password: readonly) for details.
The 4k Mode Cleaner cavity pole was measured (the
end of the past week) by
observing the intensity decay time on the transmitted
beam. During this
measurement, the cavity reflectivity was about 15%
on the average (with
minima of about 10%).
In order to observe this cavity decay, we abruptly
disabled the servo
controller. The time evolution of the transmitted
power was then fitted
(in this case by eye) to an exponential. The measurement
yielded 3.3 kHz
+- 10%.
The characterization of the LSC controllers began
this week. The objective
is to record
of each electronic controller, including the Photodiodes,
the Mixers,
the Whitening Boards and so on. The information gathered
will then be
also implemented into an e2e model. This characterization
began with the
2k IFO at Hanford, focusing on one path only, in this
case, the reflected
port. By blocking the beam to the Photodiode on
reflection and reading
the I-phase out of the demodulator board, a noise
level of 7nV/rHz was
measured (flat at least from ~10 Hz to 6.4 kHz).
The Whitening board was then addressed in the same
way. Its input
was terminated with a 50 ohms terminator and its output
level was recorded
as a function of gain (filters bypassed). Preliminary
results give
a flat noise floor (at least up to 6.4kHz) of
This will have to be repeated with the filtering enabled.
It has been requested to add excitation/test points
to the system. As only one
of the five DSC CPUs per IFO have access to the DAQS
network, this will require
some code changes in the LSC and ASC, as well as the
DSC. Cost of adding DAQS
net interfaces to all CPUs would be prohibitive (~$200k
for all 3 IFO). The main
DSC controller in the LVEA has access to the DAQS
net, so it will handle all
TP/EX for the LVEA small optics and 2KHz signals to/from
the LVEA large optic
controllers. The ASC front end will relay 2KHz TP/EX
signals to the mid/end
stations via the ISC network. The LSC front end will
relay 16KHz TP/EX to/from
the mid/end stations, as well as to/from the LVEA
large optics. The ASC CPU has
plenty of extra time to relay 2KHz signals.
We may be cutting it close with the
LSC having to relay the 16KHz signals. We'll start
on this during the week of
July 2 (I am at LLO next week to update the vacuum
controls, therefore the
delay).
This week we are in Hanford doing Intensity Stabilization
Development. Our main objectives:
1. Frequency response measurements on both LHO lasers
to establish consistency with the other Lasers which have been tested
2. Install the ISS prototype electronics and close
the PMC-lite loop
3. evaluate performance and compare them to the Lauritsen
results
4. Make response tests on the Mode Cleaner
5. Close the loop around the Mode Cleaner
We have made frequency gain and frequency response
measurements on both 2K and 4K lasers. We have found that magnitude
and phase responses are predictably similar to each other and consistent
with Lauritsen and Livingston data. However we discovered a 10X difference
in gain between the 4K and 2K units -- the 4K being way out of wack. The
inconsistency was tracked down and resolution is in process. In the meantime
a spare current shunt module was installed on the 4K and verified.
We moved our installation from the 4K to the 2K and were successful at closing the PMC-lite loop. There is <1 mw light available to us at the PMC -- significantly less than the 6-10 we have been working with at CalTech. From the servo's perspective, gain re-ranging-and-re-distribution is necessary, but the closed loop response is reasonably close to the Lauritsen results. We will look more closely today to the noise suppression performance and move on to the Mode Cleaner.
Guido Mueller, Stacy Wise, and Michael Hannawald (postdoc
candidate)
visited LLO the past week. They worked with Betsy Weaver, Doug
Cook (both visiting from LHO) and the LLO staff on
the
input optic reinstallation. This included balancing
the small optics,
replacing OSEMs, and marking the new MMT2 position
(changing it by 2.2 cm).
This shift is the calulated amount needed to improve
the mode matching to
the interferometer. All three mode cleaner mirrors
were reprocessed , rehung and are
currently in bake.
The beam splitter has been reprocessed (new magnets, standoffs) and is currently in
The vacuum bake oven. We expect to re-install the
beam splitter next Tues.
The pendulum Q of the suspended
mode cleaner mirrors will be measured in a
bell jar vacuum system that
was qualified by searching for contamination
on witness pieces. No contamination
was found by either water break test
or by the residue that would
be seen during CO2 cleaning. The vacuum
system rests on a granite table
so that the recoil loss should allow
measurement of pendulum Q of
10^5, more than sufficient for
the mode
cleaner mirrors.
We decided to leave ITMY, which
has the low Q in the butterfly mode, in
the system and only replaced
the OSEM. The compromise to ultimate
performance with good beam centering
is not significant and, furthermore,
it is possible by moving the
beam to measure the noise contribution. ITMY
has excess glue on the cylindrical
stress concentrators and may still have
a low pendulum Q. We will be
able to measure this from the Q of the violin
modes once we have adequate interferometer sensitivity. The OSEMs were
replaced on the ITMy and it
was re-aligned.
We had a good look at ETMY,
which has a low Q bullet mode. The suspension,
the glass, the glueing look
flawless. We could not see how to suspend the
mass better and therefore left
it in the system. It is now clear that once
the interferometer is running
again the Q measurements need to be
repeated.
There is a good indication that the dust in the system
has come from the
purge system. A large baffle of black glass was left
on the floor of the
manifold in the y end station. The black glass is
an excellent surface on
which to look for dust by glancing angle scattering.
One end of the baffle
was a few inches from the vent port and a clear gradient
in the dust with
a maximum nearest the vent port was observed. We have
tried installing dust
filters at the apparatus end of the purge tubing in
the past days to
reduce the injected dust. The filters also reduce
the annoying
acoustic oscillations in the purge system. A possible
reason that the dust
had not been observed before is that purge gas velocities
now being used
are larger. The acoustic oscillations from the bellows
tubing had not been
excited in prior purges. However, the dust filters cut the flow rate down
too much in the corner station, so many (or most?) of these filters have
been removed in the corner station. The filters give acceptable flow in the
end station.
The heating and cooling characteristics of the reference
cavity
vacuum chamber were measured after a command to change
the reference cavity
temperature by 0.5 degrees was issued. The cooling
time was faster than
the heating time by about a factor of 3. Of
note was that the text input
field entry of +0.5 degrees resulted in a cooling
down of the chamber and
-0.5 degrees, a heating up of the chamber. So
I will have to change the
EPICS records accordingly.
The shaft encoder on the laser power supply failed,
this is about the fifth
failure on this part that I can recall. Replacement
parts have been sent
by Shannon Gomes of Lightwave Electronics.
The water chiller for the laser was checked out as
the diode temperature
readback did not indicate a temperature lower than
27 degrees could be
reached - although a month ago it was 25 degrees (the
DTs display on the
laser power supply). The measured flow rate
was a little lower than
expected, so it could be that some corrosion from
the brass water fittings
has gummed up the water hoses somewhere.
We recorded long stretches of PSL microphone data
covering times with
high and low acoustic noise situations. Keisha and
Misty extracted the
spectra from these and the E4 data. These will serve
as a reference when
we compare the effect of various geometries and materials.
The students
are close to finishing the test-setup around the optical
table at the X
endstation.
We tested our microphones at the end-stations by playing
back recorded
thunders from a CD. The acoustic isolation of the
building is very
efficient in removing the higher frequencies (as they
should). Due to
this attenuation and the high noise within the building
during the
vent, it is not feasible to use the internal microphones.
This conclusion
is supported by the recordings made during Tuesday's
thunderstorm; The
outside microphones recorded the thunders very nicely,
while the
internal microphones recorded practically nothing.
This doesn't mean
that the low frequency components of the thunder do
not penetrate the
wall and can have significant effects. We decided
to install outside
microphones for the time being.
- Sigg found a problem in the latest DAQ code at the
sites with the reflected
memory network assignments that affects the LSC system
and GDS AWG. The
appropriate header file has been fixed and we will
reload the system tomorrow
(Thursday).
- Lori completed the code to use ICS110 model A in
the DAQS for the Digital
Suspension Coil readouts and fixed the gain selection
problem in the new Pentium
ADCU code.
- We modified a Clock Driver Module to provide only
external clocks and triggers
to the ICS110B ADC modules. This worked fine
in our development system. This
board has been shipped to LHO to test in the on-line
system.
- New SunBlade 1000 computers (3) arrived. One
will be sent to each site to act
as the new FrameBuilder for connection to the LDAS
T3 systems and one will
remain here at Caltech for DAQS development.
We will run the site units in
parallel with the existing Sun450 based systems until
LDAS informs us that it is
no longer necessary to store the data on CDS local
drives.
I have been talking to Rob Patterson of REO about
getting coating
runs done for pre-modecleaner mirrors.
Lihgtwave Electronics reported that NPRO #259 - the
original NPRO from the
10-W lsaer at LLO - suffered from low output power
because the pump diode
died. Something that we reported to them a few
months ago.
Mindrum Precision have reported that they are roughly
half-way through
polishing the 1 inch diameter super-polished substrates.
Apparently
they've had problems with Corning 7980 fused silica.
I was going to design the laser driver module (for the photon
calibration laser) and came up with
the block diagram. But yesterday I found out that
Wave Length
Electronics have similar drivers. And I made the order
for it
along with the evaluation board. I am also testing
the three
more of the DC readout modules. The chassis will be
in by
this week. The two more of the MC Servo Boards will
be
available to test by end of this week.
- The production quantities of the custom boards needed
for LLO and the LHO 2K
are arriving. The only boards not in house are the
LOS Coil Driver (due this
week) and the optical lever interface (due next week).
Three fab techs have been
hired to assist in stuffing these boards. Two will
start this week and the third
will start on July 9.
- The rack and wiring diagrams for the LHO 2K are
complete and the cross connect
lists will be sent to LHO by 6/21.
- Rus has started fabrication of the cross connect
harnesses for the LLO 4K
upgrade.
New osems for spares, 40m, TNI, test stands, etc.: Ordered misc. parts for Todd to start production.
GDS-wide software release. This required installing
new code that has been
submitted (e.g. a new version of OSC from Dave Chin),
merging two
parallel branches of the socket code (one using TCP,
the other uses
UDP) so that the can coexist, and reorganizing the
sources in a structure
agreed on by Daniel and myself. In addition I have
made a change to the
data access class to allow more rapid response to
monitor data requests.
The D970308 integrated layout drawings (six sheets)
for the LHO interferometer are being revised to reflect the current as-built
design configuration. A current layout of the PSL components was received
from Dave Ottoway and is being incorporated.
This week we have been working on our North arm cavity in an attempt
to
duplicate last week's success in locking the South cavity. (We had
planned, as a first milestone, to lock each arm separately.)
Unfortunately, the North arm cavity had become badly misaligned a few
weeks ago, and attempts to correct this alignment via the OSEM
controllers' dc offsets were not successful. We vented, opened the
chamber, and realigned the North cavity without incident. We are
currently checking the stack level and watching for drift, and we plan
to close up and pump down over the weekend.
no report
Simulation and Modeling (Bhawal)
Weekly Physics meeting
----------------------
Bill Kells discussed the FFT simulation for studying thermal lensing
effects
in LIGO I mirrors and how the radius of curvature of the recycling
mirror was
chosen to compensate the effect in the hot state. Implications of introducing
a
lower refractive index in input mirror in E2E model to have approximately
correct simulation of the hot state were discussed.
Detector characterization for LSC modeling
------------------------------------------
(Luca) gathered various data at LHO for the modeling of LSC, and is
analyzing now. The details are given in the detector section.
Mode Mismatch
-------------
(Biplab & Hiro) After validating symmetric cavity cases, more studies
were
done on asymmetric cavities in which case making theoretical predictions
is
difficult in many cases. Considered cavities with perfectly reflecting
end
mirror to study simplified situations. Some results are good but more
refinement in understanding is needed for large mismatches.
PSL studies
-----------
(Biplab) At Rick's suggestion, we introduced proper lens and distance
in front
of RFPD measuring reflected light from the reference cavity of Frequency
sensor. Ran various cases of transverse and longitudinal motion in
presence
or absence of nonuniformities in finite size detectors and estimated
noise
levels. Made a table for various Zernike polynomials representing
nonuniformities in PD responses and corresponding noise levels.
Dual recycling code
-------------------
(Malik) has delivered the formulation and matlab code of the dual recycling
summation cavity. The documentation is available in
http://www.ligo.caltech.edu/~e2e/actions/dualrec-Malik/
Validation of the formulation is going on.
Alfi
----
(Hiro) Further research was done to support the choice of JAVA for
the next
alfi development. The language is reasonably accepted by the community
based
on various observations, including the fact that gcc 3.0 now supports
JAVA
compilation. I started writing a requirements and implementation document
to
clarify the goal, and several tests will be done in the early stage
of the
development to ensure that there is no show stopper using JAVA as the
development language.
(Bruce)
- Creating class document summarizing the proposed implementation of
Alfi
in Java.
- Installing new JGo library to be used for new version of Alfi.
- Designing the interface between the JGo library and the non-graphical
elements of the new Alfi system.
(Ed Maros)
1. Installed Java SDK 1.3.1 for Solaris and Linux under /ldcg
2. Continued to work on reducing the number of refreshes done
by
Alfi.
(Melody)
- Evaluating the JGo Java library and figuring out how it would
be used
for the new implementation of Alfi
- Worked on the makefiles to build the Java files for Alfi and
generate
the Java API documentation (an example of this type
of documentation
can be found at http://ligo.caltech.edu/~maraya/api
)
- Started on the code organization for new Java code.
- Did some performance testing between the JDK 1.2.1 and 1.3.1
- Had a brief discussion with Bruce regarding the new Alfi design.
- Investigating on using JUnit as a the testing framework for
the
upcoming Java code.
LIGO Data Analysis System
Software Systems (Blackburn)
The testing and bug fixing for the LDAS 0.0.18 release continued
this week. As of today, there are no known bugs that need to be
exterminated for this release and we will be spending the day
carrying out the suite of "canned" tests for LDAS infrastructure
and from previous MDCs (and a few others that we think of along
the way) to validate the readiness of the software for a release
to be announce to the community hopefully today or tomorrow.
The errors found in the documentation for building the LDAS
baseline development environment know as ldcg have been fixed
and the missing configuration options required for the stand-
alone wrapperAPI have been added to the documention.
The huge memory leak found in the dataConditionAPI on Monday
of this week (>5MB per job) was traced to an improper use of the
ILwd::LdataContainer::pushback function in user classes. These
were all corrected and the memory leak has been removed.
The database tables were identified during the last MDC as
lacking a needed field for capturing the ldas job-id and the
ldas system (domain) where the job was executed. This has been added
to the tables and the wrapperAPI, dataConditionAPI and the
database test scripts have been updated to use these extra
tables. All database instances, including the production databases
at LHO and LLO which contain metadata from the engineering runs,
were updated accordingly. The LDAS 'dbaccess' package was update
to fill in this information when the dataCondAPI and eventMonAPI
send data to the metadataAPI.
The problem with not being able to insert multiple rows of
data or insert data into multiple tables from the dataConditionAPI
was just today traced down and fixed. The version of LDAS being
tested today has this fix.
The normalization issues between the power spectral densities
generated in the dataConditionAPI and what was expected in the
LAL search codes run under the wrapperAPI that prevented the
final "challenge" test from taking place during the last MDC
was fixed this week. However, since then it has been discovered
that there are several inspiral template issues in the search
code which also need to be fixed before the "challenge" tests
can be performed.
Several minor changes were made to the factors appearing in the
higher order moments of the statistics class, along with a few
other minor bug fixes for array sizes in the dataConditionAPI.
LDAS staff worked very closely this week with the key figures of
the LSC community to prepare a document specifying a standard for DFTs
and spectra. This document will be presented to the larger LSC community
in the next day or two.
Hardware Systems (Anderson)
CIT by Stuart Anderson:
1) Intermittent tape failures in the Caltech Cybernetics robot have
been
reproduced after following the initial vendor instructions. Now that
the initial tests have been done, Cybernetics should accept the unit
back for repair.
2) A problem with new Solaris machines synchronizing their Unix time
over the network via NTP has been identified:
a) For Sun Blade100 the
internal Sun bug number is 4460876 and
a patch
is expected soon.
b) For Sun Blade1000 the
internal Sun bug number is 4379876 and
a new
case number for LIGO support will be opened today to
track
this problem.
3) Reorganization of the computer supply storage in room 605 Millikan.
LHO by Greg Mendell:
1) Worked with Otto on plans to enclose the ldas area in the staging
building.
2) Purchased a label maker, media converters, and cooling fans for the
Sun T3
StorEdge cabinets.
3) Setup a GC machine (lynx) with copies of LDCG, LDAS, LAL, and LALWrapper
software for visitors who wish to develop scientific search code.
4) Set up an ssh agent to run a backup script of all file systems on
the
ldas network.
5) Studied backup plans for live DB2 databases.
LLO by Shannon Roddy
A 72hr burn-in test of the new Beowulf cluster has passed.
Data Analysis Activities
[A. Vicere']
I have received requests for providing a frame producing code to the
stochastic group for the next MDC. As this is going to be relevant for
several people, I have decided to make an effort in finalizing the content
of the library, robustify it, create a documentation and provide two stand-alone
tools and one skeleton program,the latter to be tailored by the users.
This enables everybody who has access to the LIGOTOOLS to produce frame
files. The library is currently based on the VIRGO Fr library, which is
part of the LIGOtools. With Albert and Kent we decided to move to the LIGO
framecpp, while retaining the option to link against the Fr library to
be able to check in the future that LIGO and VIRGO keep aligned the formats:
this is work in progress.
[P. Shawhan]
I made channel lists for each of the datasets stored in the HPSS
archive, and also assembled an exact catalog of contiguous stretches
of valid data for each E2, E3 and E4 dataset ("full", RDS, and trends).
This information is available from a web page,
<a href="http://www.ldas-sw.ligo.caltech.edu/ligotools/dataflow/UDN_List.html">
http://www.ldas-sw.ligo.caltech.edu/ligotools/dataflow/UDN_List.html</a>.
(Also reachable from the
<a href="http://www.ldas-sw.ligo.caltech.edu/ligotools">
LIGOtools home page</a> by clicking on the link called "Universal
Dataset
Names (UDNs), time ranges, channel lists, and links to servers"
The most obvious cause of lost data was downtime of the DAQ system,
which naturally caused a gap in the "full" (one-second-long) data frames,
and if the downtime was longer than a minute, there was also a gap
in
the second-trend data (which is stored in minute-long frame
files). However, a less obvious consequence is that each trend
frame
which includes a DAQ downtime contains ZEROs up to the time of the
DAQ restart. This means that each DAQ downtime, even if brief,
has
an associated loss of up to one minute of second-trend data, and up
to one hour of minute-trend data. There have been many DAQ downtimes
at
Hanford during the engineering runs.
In addition, several files of minute-trend data from LHO were found
to have been truncated in the process of moving them to the HPSS
archive, causing them to be unreadable. This resulted in a loss
of
4 hours of E2 data, 3 hours of E3 data and 7 hours of E4 data (only
for
LHO minute-trends). Stuart has figured out the cause (a "race
condition"
between writing the file and rsync'ing it to Caltech) and has now applied
a fix.
General Computing (Wallace)
MIT:
-Investigating VA Linux system imager for ldas cluster install
-Investigating DHCP / NAT routing for possible upcoming MDC at MIT
-Built Solaris 2.8 + patches on Ultra1 for CDS
-Ordered new Sun Blade 100's for gc and CDS
-Ordered new batch of IP addresses for gc
Livingston:
(Shannon)
-Installing the tape backup server software on its permanent home in
the rack.
-I am satisfied with the testing that I performed with the Solstice
backup suite
and I will begin using it this weekend.
-Installed a 16 port KVM which is working great in the communications
room.
Hanford:
(Christine)
- Spent a good deal of time trying, unsuccessfully, to fix a PC with
a
Windows protection error.
- Helped the SURF students with various problems.
- Purchased more ethernet cables, disk drive replacements, SCSI cables,
etc.
CIT:
(Lisa)
- Created some new accounts, most surf students. I created a
group and a
mailing lists "surf" to help better manage the surf situation.
- Got the cybernetics jukebox to be recognized by Solstice Backup software.
This required some original documentation from spectralogic.
That
document has
been sent to all of sysadmin.
- Installed authorization codes for Solstice Backup software.
- Setup 2 blade 100s on GC network in the 40m.
- Worked on some pc problems in the 40m. The interesting one
involved different
ways to use exceed to access epics screens on their Martian network.
- Set up a blade1000 on the CDS network in Wilson House.
- Installed solaris 8 on an ultra 60 in Wilson House.
- setup an automount to start using on the cdssol boxes for the luna
disks.
- Started looking at getting the protel license situation under control.
- Epics is installed on the mitfb ultra 60. I tarred this up
and burned it to
cd so that we can install it else where. This will work if and
only if the
installation path is /cvs/cds/epics.
- Activated a port on the 3d floor for Phil Willems.
(Mike)
-Fred Mann, his hard drive had a corrupted fat allocation table. I
ran
Norton utilities this seemed to work he is up and running.
-James Covington, I rebuilt him another computer and reloaded all software.
His old computer C20646 needed an over haul.
-Setup hardware for PAC meeting with Lisa.
-Had to flash the BIOS, C20646 and reloading OS and software.
-DCC, troubleshoot a RICHO IS420 scanner; it was not scanning a PDF
document. I had to install an OCR program then reset scanner settings.
-Biplap Bhawal, had a problem printing. I had to install print drivers,
and
setup email. This was on his PC card installed in his SUN box.
-Janeen Romie, ran out of disk space on her SUNPC card. I had to uninstall
office Y2K and reinstall this program on her extended partition.
-DCC, troubleshoot a HP ScanJet ADF this had a problem with the Auto
Document Feeder. I had to download some updated software and reset
settings
in order for this feature to work properly. To get the ADF to scan
multiple
documents in PDF format, you must run this program from within Acrobat
choose the scan feature under file menu then choose import then across
to
scan option. This seems to be working fine.
-Went to Millikan to test out VRVS in conference room. Video seems
to be
working but still must test audio. Suresh seems to say that everything
is
fine but I am working on testing out audio myself in between onsite
user
support.
-Mike Smith, had problems with AutoCAD 2000i & plotter. I had to
uninstall
then reinstall AutoCAD 2000i then delete plotter drivers and reinstall
plotter and reset all settings.
-Ken Mailand, I had to reinstall AutoCAD 2000i because on the 19th
in the
evening the program was not saving his updated work. Apparently he
lost
hours of work. I did a search for updated work and ran Norton to check
for
a virus but came up with nothing.
I reinstalled AutoCAD 2000i. Ken ran a test on AutoCad to see if his
updated work is being saved, which passed he is back in business.
(Barbara)
- Continued efforts to break out staff benefits, indirect costs, and
contingency on the web costbook reports. Fixed the activity detail
pages
to return to the correct calling page; this feature wasn't working
correctly. Created a style sheet for the reports and made other
housekeeping changes.
- Loaded the SUS data into the costbook.
- Web site changes for LDAS boot camp page, accounting, LSC, and meetings.
(Larry)
-Spent time dealing with procurement issues. Placing an order for a
couple more
Blade1000 units and trading in a number of old units.
-Working with Donna to get a number of PC issues resolved. Getting
data restored
from a blown disk and a new unit in for Barry.
-Resolved a couple of minor network problems. Resetting the cables
in most cases
took care of things.
-Helped a user with an application issue. The problem turned out to
be there was
not enough disk space in their home account. That has been temporarily
resolved
and that entire disk should be moved to a larger disk next week.
-Working with the Sysadmin group and Gary creating a computer incident
report
form. Barbara is going to put it on the WEB.
-Met with Albert and Richard McCarthy from Hanford to do more work
towards getting more
bandwidth to the Observatory. Also, went over getting a second T1 line
in at
Livingston with Albert, Tom and Shannon, so far all are in agreement
that it is
needed.
-Spent a little time with the DCC going over things. As a item to note
the
scanners and other equipment used by the DCC are not for general usage.
They are
to be used by the DCC group. There is a scanner and CD burner located
on the 3rd
floor of Bridge that is available for general usage.
-Going over a number of security issues with the Sysadmin group.
Silicate Bonding
Helena Armandula
Evaluating silicate solutions of different strength.
LASTI
Jay Heefner & Rolf Bork
- Jay will be travelling the MIT next week to install the DAQ and EPICS
server.
He will also complete the installation and test of the suspension rack.
- the LASTI DAQ system is being boxed up and shipped.
- The EPICS server has been preloaded with the software and screens
for the
suspensions and PSL. All have been tested here at CIT prior to shipment.
Hopefully it will be a matter of unpacking and installing the system
and
powering it up.
Lee Cardenas
Lee has packed a number of optics and optical mounts and shipped
them off to LASTI.
VME ADC Module Development
Jay Heefner
- We have received the initial responses to our draft spec for a VME
ADC module.
We will incorporate suggestions and polish up the spec and then go
through a
formal RFP for development.
From: Janeen Hazel Romie <romie_j@ligo.caltech.edu>
3. Other Engineering and Science Issues
New osems for 40m, LASTI, TNI etc.: Ordered parts for Todd to start
production.
Advanced LIGO: I'm working on action items from the SUS cost/schedule
review.
From: Riccardo DeSalvo <desalvo@ligo.caltech.edu>
Lia Stellacci
(SURF student) arrived in Pasadena will work with Fidecaro on
acoustic
emission.
Virginio
in Hongo for two weeks.
Riccardo
Moving to Virgo next week to work with superattenuator group.
The
weekly meetings will switch times to allow the different longitude
correction. Evenings for Hongo, mornings for Pasadena.
Hongo activity
One tower getting ready for commissioning, finishing vacuum cabling,
testing connections, possibly pumping down and starting diagonalization
of IP tonight. Next step will be suspension alignment and
characterisation.
Szabi
Shipped circuits of stepping motor controllers to Hongo.
Alessandro
Next week defending thesis.
Tested 30 KHz actuator at half power (150W) using Silicon Carbide grit
instead of Boron Carbide with Aluminum horn and a Allen screw head
as a
master on it. Cut out a glass Allen wrench in a few minutes.
Testing
titanium horn and good grit soon. Cutting speed about 5 mm^3/minute
despite low power, wrong grit and aluminium horn. Very adequate
for NC
milling.
Hareem
Analysing old Q-measurement data for GASFs and MGASF.
Brett
Measuring detwinning of MM springs.
Found 90% pullback after detwinning pulse (partly expected).
Starting quantitative measurements.
For additional information about this report, contact sanders@ligo.caltech.edu