Known MOSAIC Problems

These are problems with known causes and (probably) known fixes. The instrument has other problems that we do not know how to fix; some of these are discussed in the main MOSAIC User's Manual. In most cases, we will fix the problems noted below as soon as possible (usually the next opportunity to open the dewar).


One corner of the array in f/7 imaging mode is vignetted (March 1998)

This problems is probably caused by a slight vignetting of the beam by one of the mechanisms. It requires a full optical realignment to identify and correct the offending piece.  

Low level line-to-line pick up (most noticeable in subtracted dark frames) (September 1997)

This seems to be caused by microphonics induced by the cryo-cooler. We believe that the level of the pick-up can be reduced by stiffening the detector mount, securing various passive components on circuit boards in the dewar, and running the cryo-cooler at a slightly reduced power. We will attempt these fixes as soon as the dewar is opened.

x-pupil motions fail sporadically (October 1997)

We have determined this is not due to hardware failure, but most likely caused by a bug in the motor control software. The programmer that wrote this particular piece of software has since left the group. We will attempt to determine the cause of the problem and fix it as soon as possible.

Motion of the pupil mask in the x direction is inadequate (March 1998)

The pupil mask cannot be moved in the x direction to completely cover the telescope thermal emission. Fixing this will require a full optical realignment of the instrument. Note that a temporary fix is to shim one side of the instrument (analagous to "twist-and-shout"ing old infrared photometers); on the KPNO 2.1m, the fix is to drop the east side of the instrument about 1mm off the telescope mounting surface.

The J and Kshort filters have long wavelength leaks (March 1998)

Both the J filter and the Kshort filter have significant long wavelength leaks. Be sure to use the PK-50 blocker in the prefilter wheel when using these filters. In the future we will probably put a piece of PK-50 in the same cell as the J and Kshort filters.  

The command to move the camera turret so that the entire pupil can be viewed only works at zenith (September 1998)

Dick Joyce reports that issuing the "ie campupil" command does not work when the telescope is not pointed at the zenith. The "ie campupil" command moves the camera turret to an out-of-detent position. To do this, the motor moves the turret out of a detent and about 1/4 of the way to the next. Since the balance of the turret is not neutral, the turret may slip back into the detent when the telescope is over an hour or two. We recommend that the pupil mask be aligned at the zenith.

The communications with the TCS occasionally "lock up" (November 1999)

There have been reports that the MDM TCS link can be locked up if you attempt to command an offset while the telescope is moving, or if a conflicting TCS command is issued from the xtcs tool on the sparcstation (via the GUI front-end) while the TCS is busy executing a move. Both behaviors have been reported. If the TCS locks up, issue the command TELINIT from the Prospero keyboard. This will reset the ariel/XTCS socket connection. To verify that it has been reset correctly, type TELSTATUS which asks the TCS for the current pointing info. If it responds, check the output printed in the Prospero window with the info on the TCS PC main screen. The data should all match. Note that to be able to initiate the ariel/XTCS socket connection, both the xtcs and the TCS PC computer must be running and IDLE, that is, not currently executing any moves or other telescope operations. Please report any other lock-up behavior, along with a detailed description of what you were doing when it locked, to Rick Pogge (

+/-DISK No Longer Supported (November 1999)

Due to low-level changes in the IC software, the +DISK and -DISK commands are no longer supported in this version of Prospero. The most common use of +DISK and -DISK was to turn off data storage when taking "junk" frames to clean residual image off of the ALADDIN array. Since new workarounds are being tested, please contact Darren DePoy for the latest suggestions.