Introduction (Bob Goodrich)
The LRIS Reliability Plan (Paola Amico)>
Keck I Shutter Drive Upgrade (Dennis McBride)
Electron Beam Coating Chamber (Dennis McBride)
Mainland Observing
(Greg Wirth)
KOA (Keck Observatory Archive) (Hien Tran)
NIRSPEC Reliability Project (Grant Hill)
New LWS Filters (Randy Campbell)
HIRES Detector Upgrade Delay (Grant Hill &
Sean Adkins)
Welcome to the second Keck Observing Support Newsletter. The Observing Support Group consists of the Support Astronomers, Instrument Specialists, and Observing Assistants, who provide the first line of support while you are observing at Keck. The goal of this letter is to share with you significant current or upcoming changes in the operations environment.
Since the last newsletter we have been hearing a lot of good reports about the new pre-/post-observing spaces, both in the VSQ and in the Kohala building (where the Remote Ops rooms are). We have also started seeing success from the LRIS servicing mission that Paola Amico has been leading, and have started a program to reduce time lost on NIRSPEC nights. Grant Hill briefly reports on this NIRSPEC program. Dennis McBride, Keck’s Facilities Manager, provides some insight into the K1 dome shutter repair work that has been ongoing and will continue for more than a year. This is work not normally on the observer’s radar screen, and we are glad that Dennis can share with us. He also passes along some information about our newly commissioned electron beam coating chamber.
There is also an article on mainland observing from Greg Wirth, and one on the NASA archive project (KOA) that Hien Tran is leading. Randy Campbell reports the addition of two long wavelength filters for LWS.
The HIRES detector upgrade has been delayed as explained briefly in the last news item. Release of this newsletter was delayed to try and provide more up to date information, but the schedule for commissioning of the new CCD mosaic and dewar is still uncertain. We are looking forward to providing this excellent upgrade to users in 2004A, and will work with Lick to overcome these latest obstacles.
Following on the heels of the HIRES upgrade will be the new near-IR integral field unit spectrograph, OSIRIS. We are planning to start commissioning in late 2004A, and continue it during early 2004B.
We still have not been able to set up an unsubscribe/subscribe mechanism, but we have adjusted our e-mail list according to your requests. Again, feel free to send me e-mail if you want to subscribe or unsubscribe, and I will adjust the e-mail list accordingly. We hope you find the following information useful!
The LRIS reliability plan (also known as the “LRIS Fix”) was proposed in January 2003 and approved for FY04. The plan’s main goal is to minimize both on-sky losses and inefficiencies (measured with respect to available shutter time). The need for this project arises from the following facts:
The average time loss on sky in 2002 was 4.1%. The current 2003 average (up to November) is 3.0%. Our goal is to consistently drop the average below 2%, with the ultimate goal to stabilize it at ~1%.
The plan consists of two phases: the Operations phase aims at restoring the instrument to its nominal performance by servicing and optimizing the operation of various elements. During the course of servicing we will fix known problems. The Development phase aims at improving the instrument performance and its main goals are that of significantly increasing the on-sky efficiency and addressing problems that are classified as noncritical for night operations. These problems have an impact on on-sky efficiency, instrument reliability, and technician friendliness and include things like selecting and installing color correction filters to allow better internal flat fielding on the blue side, improving the MIRA script, etc. The current development plan includes a number of tasks selected and prioritized from a longer list of LRIS desiderata. Priorities were assigned on the basis of the severity of the problems and their impact on science operations and overall instrument reliability. We hope that the successful completion of the project will free up resources to work on the remaining list of desired improvements.
The completion of the LRIS Fix will also result in additional by-products: training for Instrument Technicians (IT) and backup personnel, improved, complete and up-to-date documentation, re-organization of the spares database, and creation of a preventive maintenance program.
The Operations phase is 2/3 completed. Few mechanisms remain to be serviced while the documentation part has only just started. Some delays were experienced due to unexpected breakdowns (the blue shutter, dichroic carousel harmonic drive, blue camera oil leaks and red CCD electronics) and the discovery of some badly damaged blue transport mechanisms, which required preparation of an emergency recovery plan. A repair mission scheduled for May 2004 will replace the grism and dichroic mechanisms with a modified, more robust design. The Development phase has just started for all those items that can be addressed in parallel to the operations tasks. We plan to complete the project according to schedule by the end of the current fiscal year.
Even if we prefer to delay the evaluation of results until the project completes, the initial results are already encouraging, as can be seen in the figure below. This plots the metrics data over the last 15 months.
Fig. 1. LRIS time lost by month.
The yellow bars show the total number of hours lost to LRIS faults each month.
The blue line is a three-month running average.
(Editor’s note: The Keck I shutter drive was historically underdesigned. Because of this it has undergone a couple of catastrophic failures, the latest in August 1999. The Shutter Drive Upgrade project will address these deficiencies, and provide a safer, more modern shutter for Keck I.)
The Keck I Shutter Drive Upgrade Preliminary Design Review was held on September
23rd and 24th. The response from the review board and committee members was
favorable. The CARA Board has given the green light to proceed with the detailed
design. The Critical Design Review is scheduled for late February 2004. With
a positive outcome from the CDR, we will start procurement and fabrication
next Spring. The project plan calls for a telescope shutdown in
late August of 2005 for the final installation.
This project, being undertaken primarily by the Mechanical and Facilities Groups, will
replace our existing separate slew and tracking drive systems with a single
variable speed vector drive system. In addition, major components of the drive
train will be upgraded, including brakes, gearboxes, couplings and wire ropes.
The principal objectives are to improve the robustness and reliability of the
top and bottom shutters, and to give the ability to use the shutters as
windscreens. Presently, the shutters are restricted from windscreen
operation due to the low equipment safety margins in the original design and
the resultant failures that have occurred.
We are working with outside consultants Tom Hoffman of Hoffman Engineering for
mechanical support, and Steve Dangel of Dangel Robots & Controls for
drive and control system support. Both are located in Boston.
We have commissioned our new electron beam coating chamber. The first mirrors were coated on November 18. This chamber provides both filament and electron beam capabilities. It provides quick turnaround times for small mirrors (up to 20 inches in diameter), and provides capability for exotic materials and multilayer coatings. It will be used for interferometry beam train mirrors and for tests that may eventually lead to new coating techniques with our large chamber.
Although virtually all Keck observing currently takes place from the Remote
Operations rooms in Waimea, observers and others will want to operate our instrument
software from other locations. Reasons include:
To meet these requirements, people need not only the ability to run instrument
software outside of Remote Ops, but also to share the main observer’s
windows.
For all of these reasons, CARA is spearheading an effort known as the “Distributed
Observing Initiative” with the mission of enabling our instrument software
to run virtually anywhere, and for the control windows to be shared among disparate
locations. This ambitious project has several specific goals:
After considering several possible plans, our current vision is to support distributed observing by employing a software suite called Virtual Network Computing (VNC) to create a sharable set of instrument control screens. Developed as a joint project between AT&T and Cambridge University, VNC allows users on different machines to view and manipulate the same virtual desktop. Our tests so far indicate that this approach is feasible, even for sharing windows with a mainland site.
In addition to installing and testing VNC, the project also involves modifying summit computers to provide the Observing Assistants with enough screens to eavesdrop on the instrument control software, and also rewriting many of the instrument software scripts in order to make them VNC-compatible.
This work is complemented by developments at several remote operations sites on the mainland. The goal is to provide multiple facilities which are equipped with the computers and video conferencing facilities required to support mainland-based observing, including backup computer links which allow observing to continue uninterrupted in case of Internet failures. Mainland observing sites are now in operation at UCSC and UCSD, and Caltech’s facility should be online for the start of semester 2004A. Bob Kibrick (UCO/Lick) has been the driving force behind the development of the mainland observing facility at UCSC, and both Patrick Shopbell (Caltech) and Pete James (UCSD) have coordinated the assembly of comparable facilties at their respective institutions.
We expect the Distributed Observing Initiative to produce its first tangible results in the coming months. By the start of semester 2004A in February, we hope to start regular summit eavesdropping by the Observing Assistants on several of our most popular instruments (LRIS, HIRES, ESI, and DEIMOS), meaning that observers will run the instrument software under VNC by default. Later in the year, observers will be able to use the mainland facilities for observing runs with these same instruments.
Observing teams interested in using one of the mainland sites for either eavesdropping or full remote operation in 2004 should bear the following considerations in mind:
Other restrictions on mainland observing also apply. Please consult Keck’s full policy on mainland observing for details.
Most of you know that HIRES will get a new mosaic of three very efficient CCDs
in early 2004, but many may not be aware that coincident with this event, another
milestone will also be taking place: the inauguration of a science data archive
at Keck Observatory. Funded by NASA and developed as a joint effort between
Keck Observatory and the Infrared Processing
and Analysis Center (IPAC) of the Michelson
Science Center (MSC) in Pasadena, the Keck Observatory Archive (KOA), will
initially include data from HIRES, with the possibility of expanding to other
instruments in the future. All HIRES data obtained
by all Keck users will be archived and released after a proprietary period,
which is proposed to be 18 months.
Timed to coincide with the original commissioning date of the HIRES detector upgrade, KOA is proceeding well and on track to begin accepting HIRES data for ingestion in early February 2004. Implementation and testing of the KOA system are currently underway, following a technical status review in August and completion of the design earlier this summer. Users will have full access to publicly released data through a web user interface maintained at IPAC. By late next year the archive will also include a data reduction pipeline for HIRES, allowing it to serve both raw and calibrated data.
Starting last summer, we began working on several fronts to reduce time lost to server crashes. The initial focus of these efforts was on faster diagnosis and recovery while searching for the underlying cause. Within the observatory, the work gained project status in October, allowing significant dedication of resources and accelerated progress. Our efforts and results to date can be followed in detail via a web page dedicated to the project:
http://www2.keck.hawaii.edu/inst/nirspec/projects/ri/index.html
For those less interested in details of the project but wanting information that directly affects them please refer to the the NIRSPEC news page:
http://www2.keck.hawaii.edu/inst/nirspec/nirspec_news.html
In particular, we have made some noticable improvements in diagnosis and recovery from server crashes and observers are urged to read the relevant news page items.
In late October 2003, the LWS cryostat was removed from the Forward Cassegrain Module (FCM) in Keck I and transported to the Instrument Lab at the Keck summit facility. The dewar was opened and two new filters, centered at 24.5 µm and 23.1 µm, were installed. The LWS filter page has the manufacturer’s band-pass information. The delivery of these filters completes a consortium procurement of Q and N band filters from the University of Reading (UK) Infrared Coating Lab. The two new filters increase the wavelength parameter space of the instrument’s capabilities and will create new scientific opportunities for observers. The instrument was reinstalled in the FCM in early November 2003 and tested on a subsequent engineering night. The intended sensitivity and image quality measurements of the new filters were not obtained due to poor seeing and other problems. However, the indication is that they are of high quality and the new filters are now available to all LWS users.
To this point the HIRES upgrade has proceeded well and is close to completion. But difficulties in the home stretch appear likely to delay delivery and commissioning. The main problem is the appearance of condensation in the dewar, possibly due to some kind of contaminant. Work is underway to identify and eliminate the source of the condensation but as we “go to press,” the extent of the delay and its impact (if any) on the 2004A observing schedule is undetermined.
We commend the upgrade team on their work so far, are confident in their ability to quickly resolve these issues, and eagerly await a new detector that will keep HIRES at the forefront of high-resolution spectroscopy.