|
|||
| Joint Astronomy Centre |
|
|
|
| Show document only |
UKIRT NEWSThe Upgrades Programme : Progress ReportTim Hawarden, Nick Rees, Antonio ChrysostomouUKIRT UpgradesObjectivesThe main aims of the UKIRT upgrades programme are essentially three-fold :
These objectives came a large step closer to reality last summer (August 1996) when the top end was replaced with a much stiffer structure with lower wind drag. It carries a fast tip-tilt secondary mirror on a precision positioning system. The Instrument Support Unit at the bottom-end was equipped with a CCD-based fast guider carried on a new, stiffer crosshead. Other elements of the programme will be realised when the control elements of the already-installed Dome Ventilation System (DVS) are implemented, which will allow us to add natural airflow through the dome to the forced ventilation already in use. Later this year insulated dome flooring will be installed and the cooling and ventilation system for the primary mirror will finally be completed, providing temperature control of the primary mirror itself.
Primary MirrorActive control of the primary mirror figure is provided by upgrades to the Primary Support System designed at RGO and ROE, installed in 1995 and brought into operation in 1996. This employs a twelve axial actuators which can apply forces of +/- 500N, with 1N resolution, to pull and push the mirror into the desired shape. It is currently operated manually, but measurements of the telescope optical performance are being accumulated to build a look-up table for automatic control.
Secondary MirrorThe precision positioning system for the secondary mirror is a hexapod, purchased by the MPIA, Heidelberg, from Physik Instrumente, Wandbronn, Germany. It provides micron-level control of X and Y translation and tilts, and of focus, tip and tilt. The 6, very stiff, legs are extended by DC motors driving precision ball screws. The secondary itself is attached to the hexapod by a fast peizo-electric tip-tilt system, also built by Physik Instrumente and provided, along with the mirror itself, by the MPIA, Heidelberg. The mirror is of 50% -lightweighted zerodur. Its deformation at maximum throw and frequency is within design specifications while most of its surface is of excellent quality, apart from a 5% -wide turned-down region at the edge. Thermal effects of its mounting pads induce some spherical and higher-order trefoil; the TDE limits the maximum Strehl ratio to about 70% at 2 mm and the thermal effects probably reduce this to ~50%. So far, however, the perfomance has been impressive, quite good enough to be going on with. Nevertheless to realise the full potential of the telescope these faults need to be corrected, while having a spare secondary will permit experiments with low-emissivity coatings which could otherwise not be considered. A replacement, to be provided by the MPIA, is therefore planned.
What are the improvements?UKIRT has experienced a dramatic improvement in pointing since the installation of the new topend, and now boasts a 1.''5 rms over the sky, while the reliability of blind tracking has also been proportionally improved and windshake markedly reduced. The new crosshead provides precision offsetting (< 0.1 arcsec) over the +/- 3.5 arcmin X-Y range of the crosshead. Image quality is markedly more stable over the sky. In addition, the tip-tilt fast guider has eliminated two more of UKIRT's long-standing problems, i.e. windshake (in all conditions) and the ``RA wobble''. One of the most impressive results has been the routine improvement of the throughput of CGS4 with the 1.''2 slit by 40%. even in indifferent conditions. The fast guider can now operate - a little more slowly , sampling at ~ 30 Hz, - down to R > 17m so these benefits are available over most of the sky (and a new, deeper guide star catalogue is urgently needed!) Most exciting of the benefits of the tip-tilt system and the optical improvements is the dramatic improvement in image quality which can be achieved when the fast guider is operated at > 30 Hz on a suitably bright guide star (R < 16.m 5, say) and the uncorrected seeing is less than ~ 0.''6. Under these conditions - supposedly median seeing at K on Mauna Kea - the FWHM of the stabilised images reduces to ~ 0.''3 or less and the peak intensity (to which the S/N is proportional) more than doubles. Unfortunately this has been a poor winter with no ``median'' seeing nights (at least when our staff members have been observing) since the tip-tilt became routinely available on the telescope in November 1996. Nevertheless, even in these conditions of poor seeing (> 1 arcsec) the tip-tilt system can return ~ 0.6-0.8 arcsec seeing. A testament to this is provided by the images of M87 which appear on the front cover of this issue. Results of the CGS4 Configuration SurveyTom KerrCGS4 Instrument ScientistFirstly, I would like to introduce myself. My name is Tom Kerr, and I am responsible for the Cooled Grating Spectrometer 4 (CGS4) at the UKIRT. Gillian Wright was previously in charge of this instrument but has now moved back to ROE and her native Scotland. I wish her luck in her new job and hope that I will be able perform the CGS4 duties half as well as she did. From now on, all of you who need information about CGS4 and its capabilities, setups, etc., should get in touch with me as their first port of call. My email address is given at the end of this article. Following a recent meeting of the UKIRT Board, it was suggested that a clear policy was needed on available configurations for CGS4. Therefore, an online survey was recently held at our world wide web site in which we asked the astronomical community which gratings they would like to use in CGS4 and why. The design of the instrument only allows two gratings to be installed at any given time. For instance, at the time of writing, the low resolution 75 l/mm and high resolution echelle gratings are installed in CGS4. The medium resolution 150 l/mm grating is currently unavailable. At the beginning of the next semester, we plan to install the long focal length (300 mm) camera. Because of the gain in resolution due to the long camera, and because the 75 l/mm grating is the most heavily used and requested grating, we will also offer a new low resolution grating of approximately 40 l/mm which will offer the same resolution and coverage as the current short camera and 75 l/mm grating. For information on resolutions and wavelength coverages, please see /UKIRT/applying/instr97b.html and the CGS4 instrument WWW pages. Two questions were put to the community: 1) which configuration of gratings would you prefer, and 2) what is the reason for you choice. The choice of configurations was as follows (with gratings in order of preference): A: 1. LW 2. MN B: 1. LW 2. HE C: 1. MN 2. LW D: 1. MN 2. HE E: 1. HE 2. LW F: 1. HE 2. MN (where LW = low resolution, wide coverage, i.e., 40 l/mm; MN = medium resolution, narrow coverage, i.e., 150 l/mm; HE = high resolution, echelle grating)
The survey finished on 7th March 1997 and we received 54 valid responses. The most popular choice was configuration A, i.e., the 40 l/mm grating as first choice with the 150 l/mm grating as choice of the second grating to be installed in CGS4. However, it was interesting to note the various responses depending on geographical location and extra-galactic or galactic research interests. I have included some charts, shown in the figure above, which demonstrate the results more clearly than any text could. The charts show:
Item (ii) differs slightly from the others in that this is the choice of single grating rather than configuration. It is easier to see from this how popular individual gratings are. The main results were fairly conclusive however. There is an approximately equal demand for the 150 l/mm and echelle gratings, and an overall, but not overwhelming, majority in favour of having the 40 l/mm grating available. In view of this, it has been decided that the gratings should each be available about 2/3rds of the time. The 40 l/mm grating will probably become the `workhorse' grating, although the actual amount of time this is in CGS4 is yet to be decided. If you have any questions about the survey, or the decisions that have been made, or have yet to be made, please get in touch with me. You can contact me at tkerr@jach.hawaii.edu. Instrument NewsCGS4 Slit NewsAs many CGS4 observers are no doubt aware, CGS4 has suffered slit problems throughout the last few semesters. Installed in CGS4 are five slits contained in a wheel which is rotated to allow the observer to change slits easily and quickly throughout the night. Currently, 1, 2 and 4-pixel wide slits are available for low resolution work (75 l/mm grating) and two others, 1 and 2-pixel slits, are available for the echelle. Unfortunately, the slit rotation mechanism persistently failed. The user was able to rotate the whole wheel to obtain the required slit. However, the motor that rotates the wheel is also used to rotate the individual slits (by means of a brake and connecting rods) so that spectral lines appear vertical on the array. It was this part of the mechanism that failed. Attempts to fix this problem met with little success. One of the great difficulties in working with cryogenic infrared instruments is the amount of down time required to investigate and repair faults. Because of this period (a minimum of 3 weeks for CGS4), it was decided in the past to `fix' the slits in place with software (this fix forced the slits to be held at their extreme point of movement in their holders) rather than cancel CGS4 observing runs for unscheduled engineering. Consequently, many recent CGS4 runs have had tilted spectral lines on the array. In most cases this did not degrade the data, but did made subsequent reduction more complex (as many readers will know from first hand experience). Attempts were made to cure the problem when CGS4 was taken off the telescope for grating changes. However, since the problem was not fully understood, these attempts were not completely successful, resulting in the slits breaking again shortly after CGS4 was back on the telescope. The good news is that we now believe the problem has been fixed. CGS4 had another scheduled engineering period in March/April 1997 specifically to address this problem. The slit mechanism was removed and dismantled. The outer bearing housing was replaced by a new housing, constructed at ROE from a material with the same thermal coefficient as the rest of the mechanism, as the cause of the problem was suspected to be that many of the bearings were crushed when CGS4 was cooled. The original bearing housing was made from different material. At the same time, new 4-point contact bearings were introduced to reduce flexure. Tests in liquid nitrogen baths at the JACH indicated that the mechanism was indeed working as predicted, and the slit wheel was placed back into CGS4. In the 1 1/2 months since the repair, the mechanism has behaved flawlessly, and has outlasted the previous periods of good behaviour. Spectral lines are vertical on the array again, and the slit wheel continues to behave rather better than its specifications. We hope this will continue into the future, especially as the slit wheel design has implications for the mechanism in the forthcoming mid-IR imager/spectrometer, MICHELLE. Thanks should go to both engineering staff at the ROE and the JAC for their tireless efforts that resulted in the repair of the slit wheel, especially Fraser Morrison (ROE) and Ian Pain (JAC).
PeopleThe last year has been a time of flux for UKIRT personnel. The following is a list of people who have either recently joined UKIRT, or left for pastures new. Arrivals :
Sandy Leggett joined in August as Instrument Support Scientist for IRCAM3. She came to us from the IRTF. Tom Kerr joined in October as Instrument Support Scientist for CGS4. Tom came to UKIRT via the University of Nottingham. Antonio Chrysostomou joined in December, on a leave of absence from the University of Hertfordshire, as a Support Scientist, in particular for polarimetry. Alvin Balius joined in December from Hughes Missiles Systems, Tuscon, Arizona, as a Software Controls Engineer, working on the UKIRT Telescope Controls Systems. Maren Hauschildt-Purves joined in December as a member of the UKIRT Software Group. Maren was previously employed by the CSO. Stuart Ryder joined in May as a UKIRT Research/Operations Support Astronomer. Stuart was previously employed at the University of New South Wales, Sydney, Australia. Departures :Dolores Walther left UKIRT, on the 1st of April, to work as a System Associate for the Gemini 8m Telescopes Project, although she will continue to help out at UKIRT through the summer. She has worked at UKIRT as a telescope operator for the last 14 years. Gillian Wright left the island in January after 10 years of distinguished service. She returns to the ROE where she will continue to work for UKIRT. Kent Tsutsui, an electronics engineer at UKIRT for over fifteen years, resigned in January to take a position with the Gemini Project in Tucson. He will move back to Hilo next year. Chas Cavedoni has also moved his talents to the Gemini Project, after working for the UKIRT Upgrades Project. UKIRT and JCMT Announcement ListFrossie EconomouUKIRT Software GroupThe UKIRT and JCMT have set up a mailing list for astronomers wishing to be notified of instrument availability, newsletters, requests for feedback from our user community, and possibly other important information. This is a moderated, announcements-only, very low volume (~1 message/month) service. If you are a UKIRT and or/JCMT user, or have an interest in these telescopes, please subscribe by sending a mail message to majordomo@jach.hawaii.edu with the text subscribe news you@your.email.address in the body of the message. Contact frossie@jach.hawaii.edu for more information. Tales from the Fault System...Nick ReesUKIRT Software GroupIt all started sometime during the day of 17th April Hawaiian standard time. I was blissfully sleeping at a friend's house near Duxford, just outside of Cambridge. In Hawaii, everything was being prepared for a normal nights observing, unaware of the chain of circumstances that was about to unfold. Suddenly, Murphy, with help from his brother Sod, sprang into action and caused the following to happen:
Not surprisingly, when Dolores arrived at the summit she was ill prepared for the woes that had happened. Like the true professional she is she immediately made it her business to disturb a large number of peoples' dinners. Unfortunately it took a while to spot the silently glowing halt light, with initial work concentrating on the fast guide computer, which has a habit of spewing out nasty, but not fatal, warning messages. However, David Fuselier and Henry Stilmack managed to help isolate the problem to the mount computer and spent some time making sure there was no real reason for it to be unhappy. However, no matter what they did, whenever the mount computer was switched on it immediately lit its halt light and sulked. Nothing else - no happy signs on the console, no self tests, not even any smoke, nothing. Unfortunately there was no spare mount computer on site - the spare was in use in Hilo for development and testing of new Upgrades software. In the best of all possible worlds we would have three of everything (one in use, one hot spare, and one for Hilo development) but realistically we can only justify two. Coupled to this, the person who knows most about the system was now eating his breakfast at a friend's house near Duxford, just outside of Cambridge.... So, when I finally arrived at the Institute of Astronomy in Cambridge I found Sophie, the receptionist, in a somewhat agitated state since there had been a number of long distance calls for me. This presumably confused her somewhat since I had last worked there in 1990, and I hadn't visited for three years. Anyway, with message in hand I finally rang Dolores and put my planned meetings on hold. After reviewing the situation it seemed the only way out was to jury rig another computer to do the work. We couldn't use the fast guide computer since it also controls the crosshead and focus drives, and both of these are a bit difficult to live without. After a process of elimination I decided the primary mirror computer had to get the chop. This created a bit of a problem. The primary mirror computer is about 15-20 times slower than the mount computer. The mount computer runs about 10% loaded, so running the mount on it would load it to 150-200%. Hmmm. No problem, slow the mount servo down from 100 Hz to 40 Hz - I have always run it with quite a large margin so it probably will work. Just a few software changes - and we have to recompile the software anyway since it is a different computer architecture. It was now that the Internet problems became apparent to me. Logging into Hawaii from the UK normally takes a certain amount of patience, but this time the link was physically broken. Hmmm. No problem, whip out the trusty modem card for the PC and find an unbarred telephone line. Now, I don't know whether any of you have seen 'Craig's Lab' - the area where instrument development in done at the Institute of Astronomy. It is a typical University development lab with a single unbarred telephone line (can't trust those Ph.D. students, after all). However, those Ph.D. students, being a resourceful lot, have been undeterred by this. Over the years, this single line has been wired to just about every place in the building with a variety of rat's nest like schemes. Thus, you can just about hear the other person speak, but for digital communication it is useless. Phone jacks are typically six wire connectors, with only two being used. However, with this one you could just about get a dial tone on any of the six lines, but none was of a particularly high quality. It took a while to sort this all out. We swapped PC's, changed modem configurations, changed modem cards, etc, etc, but to no avail. Finally someone remembered that the Fax machine was also on an unbarred line. We connected into that and finally got a connection. After all this the result was a bit of an anti-climax. I recompiled the software, slowed the servo loop down and told Dolores the incantation to use to change the boot parameters on the primary mirror support computer so that it would wake up next thinking it was the new mount computer. She then swapped the two computer boards, connected up the Ethernet and console, and turned the power on. We all held our breath and bingo, the console sprang into life and the system started loading. Well, nearly. There was one error on bootup. This turned out to be due to a card which hadn't been reseated properly after some diagnosis earlier in the night. So, no problem, power down,. reseat the card and power it up again. Except that this time nothing came out on the console. Oh dear! Maybe it isn't the CPU after all, but the power supply is dodgy and it only worked the first time since the system had been switched off for a while. But the power supply status lights are fine. But it might just be a killer ripple. Oh dear again! Fortunately, after about 15 minutes of this Dolores sheepishly admitted to having used the 'Hold Screen' key to test my patience (and maybe stop the initial error messages scrolling off the screen). Hit the button again and bingo - working system! Well, nearly. The system came up fine and was using about 60% of the CPU which was about what I had estimated. However, 15 minutes later I got another call. Every time the telescope slewed it shut down again, apparently from CPU starvation. I logged in and disabled all telescope limits checking. I then told Dolores to slew. A whole load of warning messages appeared and the telescope shut down again. Ah ha! There is a reasonableness check on the encoder readings which is based on knowing the maximum amount the telescope can move between servo cycles. Since the servo cycle was now 2.5 times longer this check was no longer valid. I disabled the check, asked Dolores to slew again and finally, finally everything started working again. The time was about 1.00 am Hawaiian Standard Time. Did we lose 6.5 hours of observing or gain 5? Life would indeed have been easier if there had been a plug in spare available at the summit, but whilst commissioning risks need be taken, I think we coped pretty well with a fairly major problem. I would like to thank the observer [WHO WAS THIS?] and Dolores for being so patient and bearing with us as we diagnosed the problem.
A View from the TopThor WoldUKIRT OperatorWell, this has to have been the worst winter for the past few years up here on Mauna Kea. Myself, I cannot remember when I last spent so many days and nights, prisoner of that Hotbed of Fun called Hale Pohaku, twiddling my thumbs, reading... I do not need to tell those of you who came halfway around the world to enjoy the Resort Atmosphere of HP how wonderfully boring the place can be! You know who you are. Those of you bruised in the crush to grab a terminal at HP. Those who spent joyous hours, wandering aimlessly about the commons building... Take heart! Winter has receded and Spring has Sprung upon us. Soon, the mamani trees around HP will bedecked in golden blossoms. Crowds of baby quail and their parents will again pose traffic problems on the summit road. The pastures are greening, and Humuula Ranch will soon bring up their herds of all-black cows with all-black calves to harass us on the Saddle-HP road in the thick fog at night. (I will not take this time to make any comments about where those steaks at HP come from!). Those of you who are heading this way might--just might--luck out in the Weather Lottery. Me, I am heading for a much-needed vacation from the HP excitement of the past three months, only to return to begin training our first new PDRA/TO in May. Our second one will be arriving in July. (This will undoubtedly mean I will be needing another break later in the year). In the meantime, I will merrily continue tipping and tilting on V=16 quasars in 0.5-1'' seeing at K and having no problems. Photons are flying onto our detectors. I can only wish you the same good fortune! As they say in the Southeast Colonies; `Y'all come back now, y'hear?'. We will try our best to keep you happy. Just remember: sacrificing your Operator to the goddess Poliahu will NOT bring better weather. I guarantee this.
Previous: Research Articles | Up: UKIRT Newsletter | Next : And Finally... |
