AlUjari Observatory at the Kuwait Science Club (KSC) (Long: 48 00 06, Lat: 29 16 08) is the largest observatory in Kuwait and the gulf area. It was commissioned in 1986 and was named in the honor of Kuwaiti veteran astronomer Dr. Saleh AlUjari.
By todays standards, and in fact, even in 1986 standards, it is considered a small-to-medium size observatory, housing an equatorial fork mounted base home to two telescopes as illustrated in the image below.
The primary long focal length (6350 mm) telescope is of a Schmidt-Cassegrain design with 50cm diameter mirror. The secondary is a 50 cm wide angle telescope optimized for astrophotography with a focal length of 1700 mm. Three additional 15 cm smaller scopes serve as finder scopes and for solar photography. The total gross weight of the telescopes & mount is reported to be 9 tons.
The observatory is protected by a 7.5m Kirby dome and a 3m wide slit door. A Swiss company (Institute for Computer Assisted Research in Astronomy, Alterswil, Switzerland, or ICARA hereafter), now long defunct, developed and commissioned the observatory. Due to the lack of any insulation materials within the dome, a grave design error overlooked by the Swiss company, the dome turns into an oven in the peak summer heat despite three dedicated air conditioning units that can barely keep up with the inferno outside.
Over the years, normal wear and tear and poor maintenance of the observatory mechanical and electrical systems reduced its scientific capabilities. To make matters worse, new towns developed nearby and the effect of light pollution drastically cut the capability of the observatory. Currently, the observatory is located in AlZahra suburban town just 15 km south of Kuwait City. Twenty years ago, the whole area was a pristine desert with crisp clear skies.
All these factors added up to transform the once-vibrant observatory into a museum for visitors and students alike. Nowadays the observatory is mainly used for occasional planetary and lunar imaging.
The system was delivered with state-of-the-art computers. But you'd have to bear in mind that it was 1986 when the system was delivered and the refrigerator size computers were used for control & monitoring in the early stages of commission. However, ICRAS failed to interlink the observatory control system with the computer for a GOTO solution. Instead, astronomers had to use a heavy duty keypad to move the telescope in RA & DEC to the approximate target position, then by using the view finder, the astronomer moves it closer to the target until it enters the eye piece field of view. This process is tedious but for the few bright objects in the sky, and even for these objects, it takes a good five minutes to have it properly centered in a medium powered eye piece. High power eye pieces on the primary telescope (yielding 400x-500x) were out of the question. Furthermore, the sidereal drive was broken, rendering it useless to track objects, and thus the user has to constantly correct for the earth rotation.
When I joined the Kuwait Science Club in mid 2006 as a volunteer, I started working on developing a new control system based on KStars and INDI. I estimated the project to take 12 weeks, or 3 months but I was overly optimistic, the project consumed 7 months ending, officially in October 2007.
The main challenge, as we shall see, was the hardware part and the complete lack of any documentation on the telescope. I had to reverse engineer the whole system starting from the control keypad up to the motors, relays, and the various electromechanical auxiliary systems. Each wire, button, knob, circuit, and signal path had to be investigated and documented.
The new Ujari control board is connected to the telescope command wires at one end, and to the National Instruments (NI) terminal block on the other. The shielded NI terminal block is connected to NI 6509, a digital I/O (TTL) PCI card with 100 channels. This card is well supported under Linux and the control server runs OpenSUSE 10.1 and KDE 3.5.6.
BEI encoders are also connected to the terminal block. Starting in May 2007, I started developing an INDI driver to achieve a closed-loop control with the goal of completely automated scripted system. Employing INDI and KStars accelerated the development process further. Thanks to the absolute encoders, the telescope requires no calibration on startup, and using KStars, the user can navigate the sky map and slew the telescope to any object, either by clicking on it, or by entering the desired RA/DEC of the target. When the user completes their observation, they can simply issue a park command and the telescope slews to its slow position.
Additionally, I developed an AutoDome feature where the slit of the dome is synchronous with the azimuth bearing of the telescope within tolerance limits. Unfortunately, I wasn't able to fully control the slit itself due, again, to a design error by ICARA as the dome slit motor relies on direct connectivity to power & control lines installed at the dome base. That is, either the dome slit or the dome motor can be operated one at a time. Once the slit is open, the operator must remove cable before rotating the dome itself.
This issue can be fixed by using a 2 HP 220v motor, a 12v rechargeable battery, and a DC-to-AC converter, but we unfortunately ran out of the very limited budget allocated the project. Amazingly, we were able to undertake this upgrade project using a very limited budget, which I cannot disclose publicly, but one that is probably less than 5% of what would have been quoted given the scope of services for this upgrade. Two key factors kept the cost down. First, everything was accomplished using open source technologies. Second, I wasn't paid to perform this upgrade, it was a volunteer effort.
For the coordinate transformation, I used a modified version of libnova to carry on the tasks. Except for the odd convention that 0 Azimuth is South, instead of North, libnova is a very helpful and reliable astrodynamic library. Another critical requirement was playing sounds to indicate events, like slew is complete or telescope is entering emergency park mode. I resisted temptations to include Aliens "Attention! You have T-Minus 15 minutes to evacuate this facility".
As mentioned before, the gross weight of the telescopes & base is 9 tons. Therefore, there are minimum allowable altitude safety guards in the driver, auto & emergency park in case something goes wrong, and a watchdog that shuts the telescope even if the whole operating system crashes. But would you bet your life on software alone?
There are also two limits switches for each axis that shuts the telescope down if it goes too far, and if all that fails, I installed an emergency stop button connected directly to mains. All in all, there are four safety layers, two of which reside in the software, and two in the hardware.
On October 29th, I discovered that one of the tension cables attached to one side of the declination axis was snapped. All work was halted until technicians from the production workshop in KSC replace the cable. Unfortunately, despite many efforts by Mr. Zaid AlGawaee, we failed to locate the missing cable after a couple of days. The current hypothesis regarding the function of the cable is to help in reducing possible flexure by the DEC tube since it was, presumably, attached to a small counter weight for this very purpose. The telescope continues to operate reliably.
System security is another critical part, and when the control server is online, external users can only login to the server via SSH public key authentication. With Linux, the threat of viruses and worms is also greatly reduced, so the system is more secure by default. But even the most stable systems are subject to failures, and this is where the NI watchdog comes into play. It protects the system against complete OS crash, and even in case of PCI bus failure. When the watchdog loses communication with the driver, the NI board is reset to a user-defined safe state.
Furthermore, complete authority resides with the driver, and this protects the instruments from human error and rouge clients. Before any command is performed, a series of interlocking checks are conducted to insure that the operation is sound and within the system limits.
In addition to the control system, we performed major maintenance for all the mechanical and electrical systems. In fact, anything with joints was maintained in one form or another. The slit door was particularly in bad shape and it hangs midway due to friction. The dome suffered from similar problems as heat and dust is collected in the summer months and this hindered its motion. The KSC production workshop team worked hard by removing all panels, applying grease and cleaning the chains, wheels, and springs at the dome base and at the door slit. However, due to the limited resources of KSC in terms of funds and man power, there is barely any regular preventive or predictive maintenance performed.
The critical part of maintenance involves the mirrors and lens of the primary and secondary telescope. We removed the dew shield only to be welcomed by a Schmidt lens covered with a thick layer of dust, we gasped in horror! We cleaned the mirrors and lens thoroughly during the following few days, and a new check list was made for DASS employees to add to their maintenance schedule. Due to frequent sand storms every year from April to June, dust seeps through minute holes in the dome, and once it does that, it gets everywhere. Therefore, proper preventive maintenance is necessary on frequent basis to avoid costly repairs and replacements.
The 71 arcsec RMS means that 60% of the time, the object will fall within a 71 arcsec diameter circle. Giving the age of the instruments & the many uncertainties involved, it's an excellent start especially when we take into consideration the light pollution that will render most faint objects useless. The following is the TPoint generated diagram based on the selected model. It appears that there are still sources for systematic error. It was suggested by Wallace that the large ME needs to be treated by mechanically raising the polar axis.
The control system was subjected to thorough & rigorous testing since the inspection of the project. A simulation mode was developed in the driver to permit testing of new cases as they arise before they are tested in the live system. We conducted one of the primary tests in Oct, 2007. The following video is an excerpt of the test: