LIGO Weekly Report

Exec. Comm. Agenda	Highlights	LSC	Administration	Hanford Observatory
Livingston Observatory	MIT	Caltech	Detector	40 Meter
TNI	LASTI	Data Analysis	LIGO II/Adv. R&D	Past Weekly Reports

1. Announcements
2. LSC Issues (Weiss)
3. Comments on Weekly Report
4. WBS 1 LIGO I Construction (Lindquist)

Special Announcements:

LSC Issues (Weiss)

Laboratory Administration (Lindquist)

The list of current actions revised to reflect the status of open actions assigned through December 12, 2002 may be found at ACTION LIST.

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. . .

WE 12/19/02	Packages	Faxes
In	20	30
Out	13	33

Advanced LIGO (Frey)

Project Web Site for posting schedule and progress related data continues to be updated with the latest and greatest.

The annual report is due to be submitted to the NSF at the end of December, covering activities from December 2001 through November 30, 2002. This Annual Report is required for the old (Construction Funds) Cooperative Agreement, and the focus is the end of Construction, Installation, and Commissioning, not Scientific Results. A first draft is nearly ready.
Meanwhile, to support the proposal early next year for a LIGO upgrade, we will be preparing an update to the Annual Report for Operations that was submitted to the NSF in September. I sent out an email on December 9 with a pointer to this report and an index of those sections that will need to be reviewed and possibly updated. So far I haven't heard much back.

No change requests are currently open. We have requested Hanford to submit a change request to cover the costs to complete the building construction effort there.

Quality/Safety (Tyler)

LIGO Hanford Observatory (LHO) and Interferometer Operations (Raab)

Mini-Run M2
This was run from 6 pm to midnight on Saturday, December 14, to provide "battle-testing" of DMT software and maybe learn something about our machines. Unfortunately H2 performed poorly, going into a day-long stretch of not locking the day of the run. Then, appoximately 10 minutes prior to the run, we had a "bootfest" to kick off the run. H1 was back on the air within an hour and ran well for the remainder of the run. Although the machine performance was a bit embarrassing, the run was a successful test of DMT software and we identified the software issues that will need resolution for E9.

H2 Commissioning
After Matt Evans' locking code software changes, we had a euphoric week or so when both interferometers locked often and well. Unfortunately, we learned that H2 still is finicky, with episodes of terrific locking and stability, interspersed with periods of no locking. We are frustrated that we have not been able to distiguish what causes us to slip from reliable to unreliable behavior. We are tracking down signals as possible between other activities to determine if they all make sense. Recently we discovered changes of gain and phase shifts in the antisymmetric port electronics that crept in as work has proceeded. However these prolems were present during both good and bad locking behavior.

Work was done on the H2 mode cleaner servo to take advantage of improved gain and bandwidth in PSL frequency stabilization. Some improvement was noted at higher frequencies in one-arm measurements, indicating better frequency noise out of the mode cleaner. It has been verified that optics on the AS port table of H2 are more acoustically sensitive than for H1. Virginio first noted this summer that the AS_Q signal was correlated with acoustic resonances measured on the H2 table optics. We need to fix this before the freeze.

H1 Commissioning
The AS_I servo was tuned up and observed to completely flatten the low-frequency fluctuations in that channel, that had previously led to overloads in the RF electronics. This and the addition of signals from multiple photodiodes helped us run with less attenuation in front of the ASPD. The higher photocurrents and improved "loopology" in the PRC and CM loops led to improvements in performance. By injecting a swept sine frequency variation onto the light, the common-mode rejection of frequency noise by the differential arm loop was determined to be > 500. This parameter, determined by the balancing of optical parameters for the two arms, is comfortably within spec. Nevertheless, we see that there is correlation between REFL_I (usually a good frequency noise indicator) and AS_Q, which we are trying to figure out. Attempts to reduce optical attenuation in the REFL path did not improve the AS_Q spectrum, but removing a periscope peak with a resonant gain stage worked. This led to a search for parasitic interferometers in various paths, which came up dry. The REFL photodiode failed later in the week and was replaced. Measurements indicate that the CM servo gain could be increased to move the UGF from 5 kHz to 10 kHz, with a reduction in noise around and below 1 kHz. Generally H1 locks well and holds lock well.

LIGO Livingston Observatory (LLO) and Interferometer Operations (Coles)

PSL: A new, optically contacted Pre-Mode Cleaner was installed in the PSL. It has a 1.28 MHz and a visibility of 88 percent. The pre-mode cleaner servo was checked to see if a higher DC gain was possible. The installed servo board had a DC gain of about 40 dB and a unity gain frequency of about 190 Hz. The transfer function seemed to flatten at about 1 kHz instead of rolling off as 1/f. A spare version of the servo board was installed and it has the proper behaviour. It is now being tested to see if a higher DC gain and unity gain frequency is possible.

HEPI / LASTI: Winding up with the last of the machined parts for the actuators, and trying to get the delivery of the bellows shields expedited. Compiled the shop drawings with redlines as noted by the machinists. Corrections will be made to reflect actuators as-built. Actuator #2, one assembled with HYSPAN bellows is back in Louisiana and an attempt is being made to seal the leaks with low temperature solder.. (Hammond, Kern)

ETF / Advanced LIGO Pod Bases: Received the GS-13 housings back from the machine shop, with the epoxy paint machined off. Delivered them to the anodizer in Ponchatoula this morning for a complete strip to bare aluminum. These will be reassembled for installation into the ETF. (Traylor, Overmeyer, Hammond, Kern)

Optics / Laser: The ophthalmologist we've been using for eye examinations is retiring at the end of the year, so I've been contacting others and trying to locate one with the necessary equipment to perform the battery of tests the project requires. I've identified one in the same town, and have sent the our requirements for review. (Kern)

CDS: I am developing a "System Monitoring Tool" in consulation with Russ, Rana, Chethan and all others who have show an interest in having a single source of information for the vital signs of the system. This is a database centric, graphics based utility, providing information on operation of the system as well as inventory, documentation and a guide to operators when a specific action is being performed, such as rebooting a target CPU etc. The goal is to provide a point and click access to a relational database stored information including: - Tracking of the Target CPUs with the heartbeat monitoring - Tracking of the critical counts such as: total testpoints, Disk storage, Network bandwidth etc. - Provide GUI based relational information on: Epics records, Traget locations, Racks, Crates, IOCards, Cables, Inputs, Outputs, Schematics, Board Revisions etc. - Provide system installation inventory of board counts, track repairs and modifications. - Access to Documentation with search capability I'd like for this to be an open forum and take suggestion to improve upon it. (Ash Khan)

Detector/Technical Support (Coyne)

DETECTOR SCIENCE & ENGINEERING

Seismic Upgrade Project

HEPI (hydraulic external pre-isolation) installation in LASTI has been completed and the system is running. Hope to get transfer functions by the end of the week. See also the LASTI report.

(Marcel Hammond, Jonathan Kern)

Winding up with the last of the machined parts for the actuators, and trying to get the delivery of the bellows shields expedited. Compiled the shop drawings with redlines as noted by the machinists. Corrections will be made to reflect actuators as-built. Actuator #2, one assembled with HYSPAN bellows is back in Louisiana and an attempt is being made to seal the leaks with low temperature solder.

CDS Software

(Rolf Bork)

Lori has finished modifying the ASC code per Siggs request. We tried the new code at Hanford yesterday, but Pentium CPU had problems with save/restore requests. Went back to old system for now. We will test this again when we are at the sites week of Jan. 6. We will also add ASC control outputs to MMT1, which is the latest request.
New 2Gbit/sec networking equipment for Hanford DAQ should be here today. Alex will work on the software drivers over the next couple of weeks and then it will be installed at LHO week of Jan 6.
Submitted my timing test doc to DCC (T020196-00). In light of some timing problems seen between the present Variable Delay Timing module and the Penteks, I revisited using Pentek interrupts to drive the LSC software. With the latest model Pentiums, and some other software enhancements made over the past year, it seems the interrupt latency is down to ~4usec (in earlier tries this has been as much as 10-15usec). In initial timing tests here, the LSC code completes all of its ADC input, digital signal processing, and control signal outputs in a nominal 40usec. DAQ and EPICS outputs and other process housekeeping require an additional 15usec (we are up to 14 permanent 16KHz DAQ channels in the LSC). Not accounted for are GDS test points, which add up to 1usec for each signal selected. So, it's close whether this will work at the sites in the 61usec timing window. I plan to try some additional optimization to see if I can save a few more usec out of it.
Sent out a draft of the Digital Suspension Control watchdog timer requirements. Comments have been received and will be incorporated.

CDS Hardware

(Rich Abbott)

1. Mohana is looking at the photon recoil calibration system with an eye to getting the prototype delivered to the site that is well documented and properly constructed.

2. Mohana is also working on a DCN change that Jay suggested associated with the variable delay.

2. Flavio is working on the ISS. He has run into difficulties with working before the PMC cavity, and also with the old version of the DCPD. He is working on solutions to these problems.

3. Finished the report from the RFI remediation review and promulgated it.

4. Re-vamped my website to make it more readable and useable.
5. Writing manual for the Pump Servo

Pentek Noise and Glitch Measurements
(Jay Heefner)

We have continued to measure the input and output referred noise of the Penteks. In addition we are investigating the glitches that have been seen on channels 5 and 6 of the DACs. The test results will be summarized in a note that should be complete in the next week. The following are some key points and observations:

The input referred noise of the pentek is increased when the module is polled. There is a broad bump that appears in the spectrum. The position of the bump depends on when the CPU begins the polling cycle.
The input referred noise of the ADCs when the timing module is polled is cleaner then pentek polling, but there are still some spikes that appear around 6KHz. The broadband input referred noise of the ADC is approximately 14uV/rtHz.

The output referred noise of the DACs is approximately 5uV/rtHz and the spectrum is clean when either the divided down 16KHz clock or the variable delay timing module is used to clock the output. The output noise increases substantially when the output is clocked at 1MHz.
The glitches that have been previously observed appear on channels 4, 5 and 6. They can be 4 discrete voltage levels and appear when the output transitions from a small negative voltage to a small posistive voltage, i.e. zero crossing. They look like DAC glitch energy, but the high level remains for the entire DAC clock cycle. The spacing is pseudorandom and is in multiples of 6msec, and 7.5msec, but the glitches do not occur on every negative to positive zero crossing.
Further work will be done this week.

Advanced LIGO DAC

(Jay Heefner)
Frequency Devices is planning to complete their tests and ship the prototype module to us at the end of this week.

(Sander Liu)
Completed preparing the external pre-isolation (EPI) electronic design requirements document draft.Will send it out for peer review before submitting it to DCC.

PSL

(PeterKing)

The optically contacted pre-modecleaner was installed on the PSL table. The beam was adjusted into the pre-modecleaner and a cavity visibility of 88 percent was achieved. No adjustments were made to the mode-matching lenses. Since the pick-off to the reference cavity is after the pre-modecleaner, the alignment to it was tweaked up as well. So far so good. The new pre-modecleaner appears to be behaving.

Sideband Camera

(Go K. Goda, Dave Ottaway, Blair Connelly)

Dave found a minor mistake in the matlab code that plots the phase and amplitude of a weak sideband. It turns out that to plot the right images, we need to make the scan area a little larger. Therefore, by use of another pair of galvos and a data aquisition board, we are retaking data for publication this week.

[#MZ this test will also pre-qualify our second set of optical phase scanner equipment, to be used to outfit LHO as LLO is now equipped.]

Optics Analysis

(Bill Kells)

I have been doing some "fall out" studies, back from LHO, on the symptoms of SB performance and how they might be interpreted for new RMs. I will talk about this some on Monday at the commissioning telecon. I have a few different perspectives than what Peter, eg, is doing: direct measurements of the SB recycling gain (which I know from much simulation is a pretty good indicator of how bad the detection efficiency (equivalent to bad CR-SB mode overlap) is likely to be. So this will be a consistency check with whatever Peter arrives at.

Optical Contamination Cavities

(Lee Cardenas, Liyuan Zhang)

OTF Lab at Lauritsen ROOM 38

Cavity #3 Reference Cavity.ready! Taking measurements every day.Ringdown duration has increased.

Cavity #2 Test cavity. The chamber is pumping with new cleaned mirrors ~ 70ppm each. We are taking RGA measurements. Optical train set-up is next.

New chamber modification drawings are done ( solder a new base and make base insulator). Chamber will be taken to the machine shop for modification.

Looking into a good glass blower shop to make a few contamination cavities. So I can have cavities ready baked for replacement. Looking into suspending the cavity with springs, the same configuration as the PSL reference cavity. Also, replace the base support.

OTF Lab. (W. Bridge)

Contamination Cavity # 1. Cavity with test sample aquamill (hydraulic fluid for the external pre-isolator) still in progress and taking measurements every day. After 2 weeks of testing it appears that the absorption change is <=about 1 ppm/year. The total loss (including scatter) is decreasing substantially perhaps as a thick layer of contaminant is pumped off of the mirrors. Tetsing will continue for another few weeks at least.

Scatterometer

Scatterometer recovering is on hold.

Suspension Stops

(JaneenRomie)

Doug Cook’s visit last week to support earthquake stop testing was successful. Mark Barton is still testing the fused silica tipped stops.

40 Meter Interferometer (Weinstein)

Kirk has headed back down under and will finish his summary report of the substantial FSS work he did once he arrives home in Australia. Jamie continues work on intensity characterization.

We have been working on system characterization issues involving the actuator. There seems to be some unexplained phase loss in the system, which makes control more difficult than expected; aspects of this behavior look unphysical, so we'e suspicious of more than just the hardware.

HEPI design, installation, testing (Abbott, Mason, Rankin, MacInnis, Coyne, Mailand, Kern, Hammond, Lantz, Hardham, Zucker)

The mechanical and electrical assembly of the HEPI pre-isolators is complete. The pump is now running fluid through all 8 actuators. During the installation one of the Parker control valves was found to have an internal leak. We had to replace it with a valve which is marginal in performance. Upon release of the hard stops the stack drifted slightly. Today we will be placing the stack back to its nominal position then resetting the actuators so that they are recentered. We recieved a lot of help this week from visitors Marcel Hammond (LLO), Corwin Hardham (Stanford), and Brian Lantz (Stanford).

[#MZ: Getting the system and actuators "wet" and operating involved a monumental push over the last week. The team really worked hard, and continues to as I write. WELL DONE!]

Data Analysis and Computing (Lazzarini)

Discussed the WFS issues and some details of the implemented version of it in SimLIGO.

We completed checking all signs and reflection matrices for various rotations of mirrors and sorted out THE correct and consistent way of doing these calculations. Made necessary changes in code and validated by a set of runs and wfs output of SimLIGO. Then cvs committed.

Continued Working on the formulation and implementation of an accurate summation cavity for scalar field for the implementation of the common mode servo.

Compiled and lightly tested field components, operators and elements (Interface and Mirror) for the new E2E field model.

(Announcement from Hiro:)
Alfi5 will be distributed as a separate package from now on. Alfi5 is updated regularly and each time a new stable version is available, it will be made available from the e2e download page, together with an announcement to users. The e2e tarbal will have the new alfi5 included, but it is less often updated. An installation instruction is included in the alfi5 tar file. The document is available separately from the e2e home page.

(Bruce)
The E2E_PATH can now be set dynamically within Alfi. Currently attempting to catch problems when loading boxes and primitives and inform user in a useful manner how to deal with these problems.

The monitor user account has now been removed from the controlMonitorAPI. All users of the client must have a valid up to date individual LDAS username and password to monitor a running LDAS system. The control user account will be removed by early next week. Then all clients of controlMonitorAPI will need to have applied and recieved an LDAS username and password. In addition, much of the controlMonitorAPI has now been converted to use the muvh faster BLT graphics package.

A rough comparision of the performance of the metaDataAPI to raw DB2 access to the table data suggests that the metaDataAPI may be only 50 percent as efficient at inserting data into tables as the raw database calls. This is preliminary and there are many questions about whether the two tests are being done in a coherent fashion and if the results reported are correct. If this is true, we should be able to improve the performance of database interactions within LDAS. Also, the interfaces were added to the C++ layer of the metaDataAPI to allow "rollbacks" of transactions which are incomplete or partial at the time of an error, thereby allowing better management of the database and preventing corrupt data from being inserted into the database.

The new SSH distributed with Solaris 9 continues to cause problems. We have made significant progress in improving the wrappers around SSH calls in the TCL layer of LDAS to handle the differences in this new SSH distribution but still have a ways to go before calling it a complete success.

The frameAPI continued to suffer from a high rate of coredumps up until Wednesday of this week when a double delete of a object bug was found in the C++ code. This seems to have fixed the core dump problems, at least for the 12 hours of testing that have gone on since the fix was introduced. There does appear to be a significant memory leak in the frameAPI which is now being investigated.

The lightWeightAPI failure mode that was present last week has been significantly improved. However, there is still a problem with a static structure in the string class that is causing problems at a rate of about once a night. This is being investigated. It turns out that this failure mode only occurs on Solaris and not on Linux. Not sure how important that observation is at this point.

I attended the GriPhyN/iVDGL Technical Planning Meeting this week. The LIGO/GriPhyN Demo that was presented at the Supercomputing 2002 Conference was discussed in several sessions and was actually demo'ed in an afternoon session on the last day. This is a demonstration that combines LDAS and GriPhyN software and hardware within the GriPhyN/iVDGL and LIGO Projects to extend the periodic search codes used within LDAS to the grid beyond LDAS, thereby bringing additional computational resources into the task of performing a comptationally demanding periodic source search. The demo also has a wonderful graphical front end showing the Milky Way Galaxy with white dots where each known pulsar exist on this two dimensional image. As the code runs, it updates the white dots to be green or red depending on a significance factor provided by Greg Mendell's search code. The demo is a wonderful illustration of our ability to combine LDAS technology with GriPhyN/iVDGL technology to extend our capabilities to perform scientific analysis.

No major data analysis work to report this week. I wrote a report with the PULG group for the periodic search section of the LIGO Annual Report, and worked on system administration issues.

LIGO II/Advanced R&D (Sanders)

Back from LHO (partial week) I have been continuing simulations, analysis of the thermal lensing problem for Advanced LIGO.
There has been a lot of other work on this now at MIT. So I am revving up old results of mine, from somewhat different perspective, for consitancy, and to get a clearer idea of what some results of theirs mean. This is related to the general loss and distortion (in interferometers) work I've been doing at LHO, etc. Its also related to the GinGin facility design which I've also been involved with this week.

Working on the magnet standoff assembly. The ITM drawings, except for the magnet standoff assy and the dummy mass, were submitted to the machine shop 12/17. CES will be making the common parts for the ITM and ETM.

Discussed layout interferences with SEI and SUS groups. The planned footprint for the ITM and FM quad SUS structures with reaction chains causes interference between the FM structure and a BSC Support Tube. Phil Willems will be evaluating whether an abbreviated reaction chain on the FM SUS system would meet the requirements.

Also, the anticipated mass levels and table positions for the ITM and FM SUS systems require substantial balance mass and expected substantial payload levels for the SEI design. We plan to set mass budgets for significant table-mounted IFO components, with margins, and track component mass development with time.

This week I worked with Doug Cook (visiting from LHO) to set up an experiment to test earthquake stop designs. We set up an SOS in a vacuum chamber in the OTF and organized feedthroughs and electronics to allow it to be servoed. We got the servo to work stably for a brief period but it's extremely fiddly, we suspect because the gains are too large (having been hardwired for MOSs at the 40m). I'm probably going to have to get Jay to install gain knobs. I also disassembled Doug's electric field meter and wired it up for remote operation (with the sensor inside the chamber and the case and electronics outside). According to the manufacturer this should work but I'm not getting sensible readings yet.

Low Frequency Seismic Isolation:
(From: Riccardo DeSalvo <desalvo@ligo.caltech.edu> )
(Stoyan and Xavier)
It is their last meeting for a while; they gave a good LIGO talk, available at http://www.ligo.caltech.edu/~desalvo/report-xav-stoy/

(Mike)
Back from Drexel for 3 weeks, will work on sapphire thermal conductivity measurements.

(Alessandro )
Sent Francesconi for driving lessons on Ultra-Sound machine. Delivery in Pisa expected this week.

(Xavier)
The presented strain energy study in its preliminary form, will complete and refine the study in the next six months in INSA before coming back for DEA. Interim results: determined max bending angle of Flex Joint, localized the energy flow, braze should not be an issue in TN, studied the need for fillets.

(Stoyan)
Will also continue analysing the system in INSA. Determined the effective bending length, it is <40 microns, studied the natural modes of wires, vertical and horizontal resonances, studied the influence of junction masses, compared mechanical performance with fused silica fibers.

(Allyson)
Trained in fitting procedures.
Found some improvement in determining Vickers indenting sizes by painting the sample with markers before indenting. Will probably use electron microscope for precision measurements.

(Charles, Stoyan)
Charles is inheriting the automatic sputtering valve system to implement.

(Charles)
Spent three days in Malibu testing sputtering on MoRuB, Vitralloy and CuBe. Sputtering Ni or Ti followed by AuSn braze evaporation.

Weekly Report for Week Ending December 19, 2002

Special Announcements: