CFHT TECHNICAL ACTIVITIES
November 1988 through April 1999


The Technical Staff :

Nik Meheula resigned his position as Summit Operations Manager in order to pursue a career as a partner in an Industrial automation contractor. In Nik's one year of service he was very successful in coordinating the efforts of the summit crew to accomplish notable improvements in Preventative Maintenance and Safety compliance. (RA)
Larry Millard completed a 2 month term employment to help establish our safety plan. Larry's work helped establish the CFHT Safety manual and conduct staff training. (RA)

1. BUILDING AND DOME


1.1 Weather tower upgrades

The integrated temperature/relative humidity sensor was replaced due to a corrosion-related failure. (RC)

1.2 Windscreen


1.3 Underground diesel storage tank

In September we learned indirectly of EPA mandated upgrades to our summit underground fuel-oil storage tank. The new regulation requires the addition of spill, overfill and corrosion protection upgrades. Installation of the spill and overfill protection devices were completed. An evaluation of the tank and soil in order to gain acceptance for adding a corrosion protection system with no further re-work to the tank was conducted by an independent EPA approved contractor. Due to the age of the tank, the results analysis indicated that further evaluation of the condition of the tank, or re-lining will be necessarry in addition to corrosion protection. The Tank was temporarily taken out-of-service in order to comply with the December 22, 1998 EPA deadline. The Observatory standby generator has a 75 gallon storage tank inside the building which allows for running the generator for 5 - 9 hours of continuous operation. We have explored a number of possible actions to bring the tank to compliance and are in the process of reviewing the quotations. The deadline for tank re-work is December 1999. (RA)


1.4 Dome exterior paint


1.5 Generator and automatic power switch

The standby generator was serviced and operationally tested. A utilitypower failure simulation was conducted to test the automatic switch whichbrings the standby generator on-line and then re-transfers power back to
utility power once it is re-established. A problem was discovered as there-transfer of power resulted in the building circuit breakers tripping.A potential solution has been identified, and time during the upcoming July 1999 shutdown testing is scheduled. In the interim following a power outage, the motor-generator has to be manually switchedback on-line. (RA)

1.6 Observatory clean-up


1.7 Summit aluminizing lab

1.8 Dome shutter

The Heidenhain indicator gauges were configured to measure the run out of the dome shutter. A remote sampling circuit was built to automatically sample the indicators which were interfaced to a PC to record the data. (GB)

2. TELESCOPE

2.1 Right ascension drive gearbox failure

Assisted in initial troubleshooting which led to discovery of the failed bearing. Worked with daycrew and gearbox engineer to provide historical data on HA drive torques over several years. Made measurements of drive torque after the gearbox was reinstalled, and helped to confirm proper operation. (WC)

2.2 Declination jumps and associated sharp noises

Helped with initial problem identification and first attempts to locate its source. (WC)

2.3 Wide-Field Corrector (WFC)

The wide-field corrector was placed in service again after being disassembled to coat the middle element. A reduction in scattered light was noticed by the observers. (GB)

2.4 Cassegrain Environment

2.5 Primary Mirror

The primary mirror pneumatic support system failed and required removal of the primary mirror to repair. This opportunity was used to wash the primary mirror while it was out of the mirror cell. The mirror coating is holding up well and the washing returned the mirror to a quality nearly the same as when originally coated. (GB)
An air bladder seal within the one of the 24 primary mirror axial supports failed and was noted during observing on Friday 26 February 1999. We had faced a rash of problems in 1995 - 96 with bladder failures due to age and ozone exposure deterioration of the bladders which had been in storage for many years. A new mold was made and new bladders were manufactured which we installed in April 1996. These bladders had recently been replaced, during last July 1998 shutdown, with a freshly manufactured set as a preventative measure. (RA)
Following Telescope disassembly, Primary Mirror removal and disassembly of an axial support puck, one bladder seal was confirmed to have a leak. It was discovered that one of the bladders had ruptured. Examination under magnification revealed that the 0.5mm thin membrane had areas of uneven dispersion of the elastomer components. This had resulted in a tear. Careful examination of other seals from the batch also revealed areas of uneven elastomer dispersion. The seals were sent to the manufacturer who confirmed the problem to be non-homogenous dispersion and blamed the fact that this newest batch of bladders was made in a small batch that was more likely to result in this type of problem. They remade a complete set of bladders with a slightly revised elastomer formulation and they appear very uniform under magnified examination. A sample has been sent to an independent testing laboratory for determination if a not-destructive test can be used to further assess the integrity. Contacting manufacturers of specialty elastomers to investigate if alternative materials or processes for improvement is in progress. (RA)

3. INSTRUMENTATION

3.1 Gumball / Cassegrain Bonnette interference problem

The new Gumball optics have been seen to interfere with the Cassegrain Bonnette central mirror drive. The solution to this problem which has been decided on to relieve this problem is to drive the central mirror such that it avoids the area of interference. This solution allows the central mirror to access all ports except for the east port in an up-looking configuration. The solution has not yet been implemented, but the interference is not encountered by moving the mirror into the position in which it is typically used.
Data has been taken using Gecko to begin to characterize the Fabry-Perot sources in Gumball. Extremely high resolution spectra of the channel specta from both the red and blue etalons has been taken and is now being reduced. The reduced data will be sent to the Observatoire de Lyon for use with the Oasis data reduction software. (GB)
Discovered this problem during TCS IV testing of cass bonnette functionality. Have proposed a solution in software where the bonnette's central mirror will be intelligently rotated so that it will not encounter the interference. This will be done when the cass bonnette software is moved to the TCS IV computers. (WC)


3.2 AO Bonnette

The repair of the multiple valid home problem on the atmospheric dispersion compensator (ADC) revealed problems with the calibration of the dispersion and the position angles being sent to the ADC. These problems were noticed by the Oasis observers from the Observatoire de Lyon. A better dispersion relation was obtained through modeling of the ADC position using the atmospheric dispersion model in Zemax. This relation has been implemented and test and seems to be working well. A correction to the position angle calibration was identified by the Observatoire de Lyon. This correction has been implemented and test data has been taken using AOB visible imaging, but it may require the greater sensitivity of Oasis to see if the calibration is completely valid yet. (GB)
Problems have been encountered with the fiber optic data link between the WFS electronics on the 5th floor and the WFS electronics on the 4th floor. The original fibers were replaced during the last shutdown with fibers routed through the Cassegrain cable wrap. The fibers worked well for several runs, but problems were encountered during setup for recent runs. The major problem is that the signal from the 5th floor electronics is marginal, so small changes to the fiber characteristics can cause large signal loss. It is planned to replace the fiber transceivers currently used with some higher signal transceivers. In the mean time, the fibers have been replaced with a new set with a hopefully longer lifetime. (GB)
A problem with low APD counts as seen by the 4th floor acquisition system was traced to the data fiber run through the cassegrain wrap. An additional cable was pulled through the wrap, solving the problem. The cause of the failure is TBD. (RC)


3.3 MOS/OSIS

The efficiency for all of the MOS/OSIS grisms has been determined using the spectrophotometer. The efficiency curves for the grisms have been published on the Web to allow access to this data to the observers. (GB)


3.4 OASIS


3.5 The Coude f/4 spectrograph - Gecko

A new shutter system has been designed and fabricated for the exposure meter and slit shutters. The new shutters should be far more reliable that the old shutters which normally required maintenance at least once during every run. The new shutters are based on a latching solenoid mechanism and so will have zero heat dissipation. The shutters have not yet been integrated into Gecko, but should be ready for the May run. (GB)
The image quality of both spectrographs were checked using Loral3 (a thick CCD was used to get rid of the line spreading which occurs with thinned CCDs). The image quality for both the red and UV spectrographs was found to be well within the specifications for the spectrographs. (GB)
The counter module for the exposure meter was relocated from the 3rd floor outer Coude area to the 4th floor observing area. (RC)


3.6 FTS

A long-standing problem with frequency stability of the HeNe laser has been resolved. The laser can unlock and mode sweep due to external thermal perturbations. Using a commercial PID temperature controller and existing heater elements, the temperature at the body of the laser is now maintained within +.2/-.1C of the set point. Initial runs show no evidence of the instability problem. (RC)


3.7 LAMA

Safety interlocks have been installed on the Lama doors which will shutdown the lasers when the doors are opened. It was possible to implement this safety feature now that the auto-focus mechanism has been installed and is working since it is no longer necessary to focus the laser by hand. (GB)

4. DETECTORS

4.1 Clean room

Fibers allowing Detector Host communication were run from the Computer Room to the Clean Room. (RC)
Air conditioning was installed for the inner area of the Clean Room. (RC)


4.2 KIR

KIR was disassembled to install indium foil between the cold finger and the focal plane. (RC)
A thermal problem (focal plane temperature ~10 C higher than normal) has been diagnosed as a fault in the LN2 fill procedure. A new procedure was drafted and implemented. (RC)

4.3 UH8K


4.4 Loral 5


4.5 EEV

The contact problem with the ZIF (Zero Insertion Force) socket on the original focal plane board has been remedied with construction of a new board utilizing LIF (Low Insertion Force) pin sockets by Ron Johnson at UBC. An output amplifier earlier diagnosed as nonfunctional on the engineering grade device became operational with installation in the new LIF socket design. (RC)
The EEV engineering and science grade arrays were installed in the Luppino dewar, cooled and imaged. Preliminary measurements (unoptimized) on the science grade device yield 8e- read noise. (RC)


4.6 REDEYE

The Redeye failure was manifested by "0"'s for the image element values and loading of the bias voltages. These problems continued in Waimea with a warm dewar. The loading of the bias voltages continued until the focal plane board was removed. With no device in the focal plane board socket and the board reinstalled, the bias voltages returned to nominal values. (RC)
The NICMOS3 device used in Redeye was sent to Rockwell for examination where it demonstrated normal performance at room temperature without the bias loading seen earlier. (RC)
The device was returned to CFHT for reinstallation. We will reexamine the focal plane board and wiring harness, and replace of the focal plane board connector. (RC)

5. COMPUTERS AND SOFTWARE

5.1 Network security

We have significantly tightened external access to the CFHT network due to security concerns. All restrictions apply only to accessing CFHT from the Internet. There are no CFHT restrictions on going to any other Internet destination from a CFHT machine. We presently have two ways of logging into the CFHT network from the Internet. These two methods are via the "secure shell" software and our new SecurID Token system. The secure shell is a public domain software package which encrypts and compresses the communications between two machines. While this method has some shortcomings it vastly improves remote connection security. The SecurID and now the preferred method for logging into the CFHT network from the Internet is by using the SecurID Token system. This system, consisting of both software and hardware, provides very secure remote access (telnet and ftp). The hardware consists simply of a smaller-than-a-credit-card-sized token which is carried by a traveling CFHT staff member. The token has a associated PIN number and displays a 6 digit number which changes every 60 seconds. Basically, logging in from the Internet requires one to enter both the PIN number and the current 6 digits on the token, after logging into a normal user account. (BG)


5.2 Support for large detector arrays

Additional data storage capacity has been acquired and installed at the summit to keep up with the large amount of data produced by the CFH12k detector system. The present capacity has been increased to 100 Gbytes in a mirror configuration to increase system redundancy. (BG)


5.3 Video conferencing

Continuing with the plan to provide support for Waimea remote observing a Picturel video conferencing system has been acquired and installed at the summit and Waimea. The system allows also to connect to other places around the world using an ISDN line, This option has been currently used to conduct video conferencing with DAO to discuss the work in progress of the MegaPrime project. We hope to use it in the near future to conduct remote meetings with members of the Megacam project at CEA. (BG)


5.3 General data acquisition


6. MAJOR PROJECTS

6.1 TCS IV

An initial release version of TCS IV came on line in early February. The release was scheduled for the end of the month, but the initial experience with the system was so positive that a decision was made to continue running it for all observing. Since then we have run entirely with the new system. Of course all is not perfect with this infant system. Two month's experience has resulted in about 90% up time fro the system, with problems frequency gradually decreasing. Overall the system has required less direct support than expected (feared), and this too is decreasing. (WC)
This initial release still depends on TCS III for bonnette and dome services, as well as for balancing the telescope. A few features, such as a proper reference star catalog, are also missing. These deficiencies are being tackled as time permits, with the goal of eliminating usage of the HP 1000 set for early this summer. The project has officially been closed and the staff redistributed, but "project" work will continue until the end of the year. (WC)


6.2 Coude Fiber Feed

The project to provide a fiber feed from Cassegrain to coude, replacing the coude trains, is well underway at the Observatoire de Paris-Meudon (OPM). The project, which has been contracted to OPM, will provide a stand-alone fiber feed system including a thorium-argon calibration lamp and a halogen flat field lamp. CFHT is only providing the fiber routing and storage and related mechanics, a very small amount of electronic integration, and software integration. (GB)
The fiber chosen should provide as good or better throughput from 1100 nm down to about 400 nm as the current mirror trains. The throughput, unfortunately, will drop quickly from 400 nm down to 300 nm with a predicted throughput at 310 nm of about 18%. Although the UV mirror train will still be used for the near future, it will also be retired when MegaPrime is completed. (GB)
The project is still on schedule and is expected to be completed at OPM by about the beginning of June 1999 and will hopefully be installed onto the telescope before the end of July 1999. (GB)

7. WIDE FIELD IMAGING PLAN

7.1 CFH12K

CFH12K Technical activities for software and CCDs
The last issues on reliability of the SDSUII controller have been successfully addressed in October and early November. Several thousands of readout with power on and off sequences were obtained without a failure.
The 12 CCDs have been implemented in the cryostat in November at IfA and the whole system has been shipped to Waimea where optimization of the CCDs took place for 3 weeks. Apart of two defective CCD (on the far left of the mosaic: one with bad CTE and one with bad linearity), the 10 remaining CCDs have been optimized for a use in sky background noise limited conditions. Noise is about 5 electrons, linearity is better than 1% over the whole range of the ADC converter (16 bits), CTE is better than 0.99999 and the dark current is low enough to be neglected or at least handled in a easier way than the UH8K. Readout time is 58 seconds (two controllers in parallel).
The CCDs have been carefully checked in various conditions on the sky since January and appear to be behaving perfectly, confirming results obtained in the laboratory.
The CFH12K can be controlled by the observer either from a command line interface, a graphical interface or scripts.
Various observing tools have been tested: focus, offset, shift-and-add. They greatly reduce the overheads, making the CFH12K a very efficient instrument (for 10mn exposures and a sequence of 6 exposure in a shift-and-add script, the overhead is only 11%).
The multi-extension FITS (MEF) format is now available to the users. Each file is 200 Mb and contains 12 extensions (13 soon, the 13th being the binned by 8 image of the mosaic to allow very quick look without having to read the whole file).
A large amount of data has been collected during the engineering time and will allow a complete characterization of the camera and the prime focus environment: improvement of the scattered lights, astrometry, guide probe mapping, residual image, shutter ballistic, Z value for each filters, bonnette alignment, noise contributions.
CFH12K Software - Status Report
Scope of effort
Read out reliably the 12Kx8K mosaic in less than a minute. Run reliably the camera for several weeks at the summit without any failure. Provide proper observing tools to ensure that the higher quality data are acquired off the sky.
The available hardware at the time of the CFH12K project start were the SDSU GenII CCD controller (became available spring 1998), and the current UH8K acquisition system (a Sparc 20 with EDT SBUS boards and the custom IfA fiber interface). These two main components had enough throughput to ensure the system would be limited by the detectors performances.
The CCD controllers had to be configured so they allow the optimal use of the detectors (DSP code development). The data acquisition software had to be developed along the line of 8kcom (UH8K interface) and DetI to allow both low level engineering capabilities as well as high level commands for normal operations on the sky.
The data acquisition software had to be developed to allow remote troubleshooting and along the same line: allow remote operation. Other tools were required to allow efficient remote observing down from Waimea.
State of completion
CCD controllers:
The SDSU GenII controller has been carefully looked at and troubleshooted in the CFH12K configuration. All the issues on power consumption, state initialization have been addressed.
DSP code has been developed to allow optimal operations of the CFH12K MIT/LL 2K4K CCDs (readout modes)n.
Synchronization of both controllers has been successfully implemented to allow a reduction of the readout time by a factor of two.
Control of CFH12K parts has been implemented: ACE IOs (shutter, filter wheel, temperature,...).
Data acquisition host:
A new version of director has been developed to allow multiple agents running at once. This allows the CFH12K interface to control from the same level the CCD controllers (pixels acquisition) and other functions allowing proper control of the various CFH12K parts (power supplies,...).
12kcom, the agent for the data acquisition, is a rich interface to the camera based on a command line interface but it allows also commands to be sent from a Graphical Interface based on RPM or from a script. 12kcom provides all the low-level commands for engineering development as well as troubleshooting capabilities. It also proposes all the essential commands to run the camera in normal observing mode.
Observing tools:
Various tools have been developed to allow efficient operation on the sky:
focus, shift-and-add, offsets,...
Efforts needed for completion
An automatic and efficient display tool is not yet available and is terribly needed to improve observing efficiency. As a first step, a binned by 8 image can be created to allow the user a quick look at the general aspect of the image. Then a real display of the whole image (or ways to access it) must be implemented. The binned by 8 image is the higher priority since it allows very quickly to check the integrity of the data before starting a new exposure. The NOAO display seems to appear the only solution at the moment for a mosaic display but it is still under development at NOAO.
The operation of the camera is a bit limited by the low amount of disk space available on the session host. This should be upgraded to 100 Gb (currently 50 Gb).
To ensure the best scientific return out of the CFH12K data, a support to the data calibration frame should be provided by CFHT. This implies further development and tests of tools presently available (FLIPS for example). The main element remains the way the good recipes are implemented. Recipes are being now established in part of the instrument verification phase.
The instrument needs to be documented and advertised: technical documentation, user's manual, cookbook, web site, publications, public relations.
Minimal acceptable package
The binned by 8 image and the automatic display are a must. Sidik needs to spend at least 2 or 3 weeks on this. This is a scientific requirement.
The present CFH12K code (DSP code, C code, C++ code) is well commented and does not need further works.
All the work on the documentation of the instrument is crucial. I need to spend several weeks over the next 3 to 4 months to conduct this work.
Providing the calibrations, or at least a help for the calibration is essential (as advocated by the SAC). This is a large amount of work to organize and install properly the packages. The basics of the software exist but the pipeline needs to be developed and tested. This fits perfectly in the future with the service observing and block scheduling modes. This work will extend easily over the summer (jcc).


7.2 The MEGAPRIME Project (previously referred to as MEGACAM)

Project Management
Manpower (this section to include time spent by all agencies involved and the manpower from CFHT)
Contracts
CEA

Observatoire de Paris, Meudon

Division Technique de INSU

CCD procurement

MegaCam

Wide-Field Corrector

Image Stabilizing Unit (ISU)

Guiding and Focusing Unit (GFU)
Worked with MegaPrime group and with William Rambold of DAO on conceptual design of the GFU and its interfaces with TCS IV. (WC)

Prime Focus Environment
A preliminary study on the heat extraction at the prime focus was completed at CFHT. The analysis studied the potential for heat extraction using a suction air handler located off the Telescope. The study concluded that the narrow ducts in-line with the spider supports resulted abnormally high air velocities that could result in vibration. An analysis of cooling the MegaPrime instrument enclosure and electronics enclosures using glycol heat exchangers is in progress. (RA)


Conclusion


7.3 CFHT-IR

After signature of the contract between CFHT and Universite de Montreal (UdeM) on October 14, 1998, the project is now progressing reasonably well.
Project Status
Optics:
The optical design has been approved by CFHT and the corresponding purchase order has been issued to Janos Technology Inc. on March 25. Delivery of the complete set of lenses with coating is expected for July.
Cryostat and cold mechanisms
The final design of the cryostat and the cold subsystems has been submitted to CFHT by UdeM end of March and is currently being reviewed. A first quotation has also been received from Infrared Laboratories, based on that final design. We expect to issue the purchase order by the beginning of May for a delivery of the cryostat by September. UdeM is also ready to start the manufacturing of the cold subsystems.
Interfaces with OSIS and the Cassegrain Bonnette
The preliminary design of the interface between CFHT-IR and OSIS has been received by CFHT at the end of March and will be reviewed in the coming weeks. UdeM is presently designing the interface with the F/8 bonnette.
Acquisition system and electronical subsystems
Due to important loads in other projects at CFHT, the integration of the acquisition system and the design of the electronical subsystems has been delayed to May. We expect a completion of this part of the project by October, in time for the integration and tests of the camera.
Detectors
We received the HAWAII engineering grade array in February and we are currently discussing with Rockwell the selection of the science grade array.
Schedule
Following the current development schedule, we now expect to have CFHT-IR ready for commissioning by February 2000.