This page will list known problems with the instrument and (hopefully)
ways to fix/recover from these problems. Some potential problems may also
be listed in the main OSIRIS
Known OSIRIS Problems (or "Features")
PLEASE REPORT ANY PROBLEMS YOU NOTICE TO:
Darren DePoy (firstname.lastname@example.org)
Bob Blum (email@example.com)
Vignetting at the corners of the array (November 1998)
OSIRIS was not originally designed to illuminate a full 1024x1024 array
(with 18.5 micron pixels) and thus this vignetting is unavoidable without
a costly modification of the optical system. This is not planned for the
forseeable future, unless you care to write us a big check.
Shading at the array quadrant boundaries (March 1999)
The HAWAII HgCdTe array in OSIRIS exhibits a time variable "shading," or
ramping of higher than average bias level at the boundaries of the upper
and lower quadrants. This feature generally subtracts cleanly to a high
degree between frames (e.g. between succesive object and sky images), but
may show residual counts sometimes (particularly between the first image
and subsequent images in a series). The unilluminated portion of the OSIRIS
images can be used to make a row-by-row correction for this. Ask the support
astronomer for details.
20 July 2000. The new array installed in
December 1999 does not exhibit the unstable shading of the previous array.
Therefore, we recommand that no line-by line bias subtraction be done.
Bias discontinuity across quadrant boundary (April 1999)
The HAWAII HgCdTe array in OSIRIS exhibits a bias discontinuity across
the quadrant boundaries (primarily across the boundary between the upper
and lower quadrants) which appears to be flux (ADU/second) dependent. The
source of the problem is not known. The effect results in a discontinuity
in the count level which does not subtract out between similar images taken
with and without illumination (e.g dome lights on - dome lights off) when
the count rate on the array is above a threshold value of approximately
1000 ADU/second. The flux should be kept below about 1000 ADU/second in
order to insure properly calibrated data. This is particularly important
for flat fields. Check with the support astronomer for details/status.
20 July 2000. A new array was installed
in OSIRIS in 1999 December. The bias discontinuity has been greatly diminished.
Fringing in spectroscopic modes (May 1999)
As with the original NICMOS III detector in OSIRIS, the new HAWAII array
exhibits fringing in the spectroscopic modes. Spectra should be taken at
multiple (7 to 10) positions along the slit to average out the effect of
Slit position changes on array (15 January 2001)
The apparrent y pixel position of the slit appears to change on the
array when the telescope is moved. It is not clear if this is just a worse
than normal flexure or if the acquistion mirror is not repeating its position
well (something may be loose). Tests with the telescope in one position
and repeated flips of the acquisition mirror indicate it reproduces its
position to +/- 0.1 pixel. We will derive a new flexure map on the next
engineering run. Until this problem is resolved, observers should pay more
attention to the slit position (we recommend checking it on each new object).
The full range of movement reported by observers is about 4 pixels.
Home Page ]
Updated: 1999 February 16 [firstname.lastname@example.org]