Status Report - 2 January 2000
During the week of 15 November 1999 a new pointing encoder system was installed on the 0.9 meter reflector. The design accuracy of the system (gearing plus encoder) is 0.225 arcseconds. Mechanical installation was accomplished without significant problems. The telescope control system was also updated at that time in order to use the new hardware.
In system testing a significant problem was detected: after moving the telescope and bringing it to a halt the encoders register motion when the telescope is stationary. When tracking the readout RA of the telescope also drifts substantially (though the telescope itself is not drifting). In mid-December the problem was traced to differential compression in the belts (polyurethane coated stainless steel) which couple the axes to the encoders. Note that the encoders can detect a motion of 1.5*10-3 mm on the gear wheels. The linkage currently uses an asymmetric idler arm to maintain tension - it is hoped that a symmetric idler arm system will correct the problem. If not it is possible to directly connect an encoder to the RA axis, but the DEC axis is more problematic. The telescope is usable with the new encoders but one must rezero the pointing system more often than is feasible for robotic operation. Operations continue as weather permits (it is winter - our poorest season) in a manual mode. Expected time to fix these problems: 2-3 months minimum.
In late November a temperature dependent failure mode in the dome slit was encountered. Basically when it is below freezing the dome will not open. The problem is a differential expansion in the members which causes the slit halves to become canted with respect to each other when opening (or trying to open). Refurbishment of the entire dome slit mechanism is scheduled to commence as soon as possible. This will be the first major work on the dome since its construction in 1957.
Our computers rolled over the Y2K without incident. Over the next few weeks we will be replacing the three main control computers.
The spectrograph design is complete and the principal optical components are on order. We should finalize detector dewar specification within two weeks. Anticipated first light for the spectrograph is about 1 April 2000.