Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stsci.edu/ftp/instrument_news/FOS/postscript/koratkar2.ps Äàòà èçìåíåíèÿ: Thu Feb 29 06:34:50 1996 Äàòà èíäåêñèðîâàíèÿ: Sun Dec 23 06:34:51 2007 Êîäèðîâêà:

FOS FILTER­GRATING WHEEL REPEATABILITY
MEASUREMENTS
ANURADHA KORATKAR and STEPHAN MARTIN
Space Telescope Science Institute, 3700 San Martin Drive, Baltimore MD 21218
Abstract. The post­COSTAR FOS Filter­Grating Wheel (FGW) repeatability was tested for both
the red and blue detectors. The test consisted of obtaining repeated FOS images in an uninterrupted
sequence through 3 different gratings on each detector. The observations consisted of exposures using
G190H, G400H and G570H with FOS/RD, and G130H, G190H and G270H with FOS/BL. All other
observing parameters were unchanged during the test.
The standard deviations of the positions of spectra with respect to each other are found to be
independent of the chosen grating. The rms deviations of the position of the spectra in the FOS
X and Y directions are of the order of 0.1 diode and 9 ybases, respectively. Deviations as large as
0.36 diodes and 18 ybases are also seen. These values are similar to those obtained from pre­launch
calibration activities. The non­repeatability of the FGW could be a major source of instrumental
inaccuracy in the wavelength calibration of spectra obtained without contemporaneous calibration
lamp spectra. It should be noted that the absolute wavelength calibration for a typical FOS spectrum
is not only affected by internal instrumental effects, such as, FGW position and aperture location,
but is also affected by the target acquisition accuracy.
Key words: FOS -- Wavelength Calibration
1. Introduction
Pre­flight measurements of the FOS filter­grating wheel (FGW) repeatability showed
that the FGW repeatability was of the order of 10 microns (12.8 ybase units in Y
and 0.2 diodes in X) in both the X and Y directions of the FOS blue detector, and of
the order of 14 microns (17.9 ybase units in Y and 0.3 diodes in X) in both the X and
Y directions of the FOS red detector (CAL/FOS­12, CAL/FOS­17, CAL/FOS­49).
This non­repeatability of the FGW leads to wavelength calibration errors.
Uninterrupted repeated observations with different gratings were obtained in the
first part of the calibration program 6146 to check the FOS FGW repeatability. The
red side detector measures used the G190H, G400H and G570H gratings while the blue
side detector measures used the G130H, G190H and G270H gratings. Observations
for each grating are listed in Tables II­VII. Each observation was in the FOS IMAGE
mode, and spectra of the internal Pt­Cr­Ne wavelength calibration lamp using the
0.3 00 aperture were obtained in a continuous random sequence of approximately 10
exposures in each grating. The FGW was the only instrument component that was
moved during the entire observing run for each detector which lasted 57 minutes for
the red side detector, and 77 minutes for the blue side detector. The exposure times
for each single exposure are given in Table I. Other exposure details such as number
of diodes used for each grating, and the wavelength range covered are also given in
Table I.

A. KORATKAR AND S. MARTIN
TABLE I
Exposure Characteristics for each Grating
Config. Grating Exposure Diode Wavelength Number of Emission
Time (s) Range Range ( š A) Lines Used
FOS/RD G190H 320 196 -- 313 2032 -- 1863 24
FOS/RD G400H 24 196 -- 313 4197 -- 3845 20
FOS/RD G570H 24 196 -- 313 5965 -- 5455 10
FOS/BL G130H 90 196 -- 313 1281 -- 1403 15
FOS/BL G190H 36 196 -- 313 1858 -- 2034 27
FOS/BL G270H 6 196 -- 313 2629 -- 2879 29
2. Data Analysis
Since each IMAGE mode observation used 118 diodes, with NXSTEPS=4, OVER­
SCAN=5, YSTEPS=24, and YSPACE=24 there were 24 spectra with 488 pixels each,
stored as 24 groups in the original data file. The 24 spectra were then converted into
two­dimensional images with 488\Theta24 pixels each using the STSDAS task rapidlook.
For each grating the first image was chosen as the reference image. The 488\Theta24
pixel images were then cross­correlated two­dimensionally with the reference frame for
each grating. The STSDAS task crosscor is used to determine the cross­correlation
(see Bracewell 1986 for details). Since the centering option is used in the crosscor
task, the maximum power is expected at the (245, 13) pixel position in the cross­
correlated image for this analysis. The shift of an image with respect to the reference
image is thus the position of maximumpower in the cross­correlated image. The cross­
correlated images are of high signal­to­noise, since many (see Table I) emission lines
with significance levels ? 3oe above the background are used in the cross­correlation
analysis.
The position of maximum power in the cross­correlated image is given in Tables
II­VII (columns 2 and 3). The offsets with respect to the reference central pixel
(columns 4 and 5) are given also in the tables. The measuring uncertainty of the
cross correlation peak is ! 0.02 pixels in X (where 4 pixels = 1 diode) and ¸0.05
pixels in Y (where 24 ybases = 1 pixel). The standard deviations (in units of pixels)
for each grating are given in the tables.
3. Results
Only the deviations from the reference pixel are important for determining the FGW
repeatability, because the absolute value of the reference pixel is arbitrary. These
deviations are of the same order for all gratings.
The 1oe rms deviations of the position of the spectra are 0.47 pixels in the FOS
Calibrating HST: Post Servicing Mission

FOS FILTER­GRATING WHEEL REPEATABILITY
X direction and 0.36 pixels in the FOS Y direction. These rms deviations translate
to 0.1 diode in FOS X direction and 9 ybases in the FOS Y direction. The range of
deviations is 0.06 ­ 1.44 pixels in FOS X (0.02 ­ 0.36 diodes) and 0.01 ­ 0.74 pixels in
FOS Y ( 0 ­ 18 ybases). Since the deviations in the analysis are Gaussian distributed,
we can expect 3oe deviations as large as 1.5 pixels (0.36 diodes) in FOS X and 1.08
pixels (25 ybases) in FOS Y sometimes. We see that the range of deviations for the
red detector are slightly larger than for the blue detector.
We have noted in a previous instrument report (CAL/FOS­133) that the location
of the FOS spectra are affected by the ambient magnetic field in the detector. The
duration of the entire observing run for each detector in this calibration program
is long enough that magnetic field effects could be significant. Thus the scatter in
the location of the spectra observed in the present analysis could be not only due
to the FGW but also due to changes in the magnetic field environment. We note
that the amount of deviation in the position of the spectra obtained in this test are
similar to those obtained during pre­launch calibration measurements (CAL/FOS­
12, CAL/FOS­17, CAL/FOS­49). Thus, the post­COSTAR test indicates that there
has not been dramatic wear in the FGW notches that hold the grating in a given
location. Further, since the pre­launch measurements were in a constant magnetic
field and showed similar deviations as the post­COSTAR measurements, we conclude
that the scatter in the location of the FOS spectra due to non­repeatability of the
FGW is no less than the scatter induced by the magnetic field environment.
The scatter along the FOS Y direction (perpendicular to the dispersion axis) affects
the flux calibration at a level ! 2% and only for the large apertures (4.3 00 and 1.0 00 )
since the spectrum can be misdirected onto the diode array.
The scatter in the FOS X direction (along the dispersion axis) leads to wavelength
uncertainties in the range of 0.02 ­ 0.36 diodes. For a typical FOS observation, uncer­
tainties in the absolute wavelength calibration are due to; (a) the non­repeatability
of the FGW, (b) the uncertainty in the location of the FOS apertures in the FOS X
direction which ranges from ­0.21 to 0.13 diodes (CAL/FOS 131), (c) the changes in
the ambient magnetic field of the detector, and (d) the target acquisition accuracy. Of
these four sources of uncertainty, the non­repeatability of the FGW is the dominant
internal instrumental source of inaccuracy in the wavelength calibration of spectra.
A contemporaneous calibration lamp spectrum without any change in the aperture
position can reduce the effect of FGW on wavelength calibration.
Calibrating HST: Post Servicing Mission

A. KORATKAR AND S. MARTIN
TABLE II
FOS/RD G190H Grating Observations
Root name X position Y position X Offset Y Offset
(pixel) (pixel) (pixel) (pixel)
y2nc1003t 245.00 13.00 0.00 0.00
y2nc1007t 245.87 13.18 0.87 0.18
y2nc1009t 245.75 12.89 0.75 ­0.11
y2nc100ct 245.57 12.87 0.57 ­0.13
y2nc100ft 246.14 13.08 1.14 0.08
y2nc100kt 245.94 13.06 0.94 0.06
y2nc100mt 245.69 13.46 0.69 0.46
y2nc100pt 246.44 12.39 1.44 ­0.61
y2nc100tt 245.61 13.19 0.61 0.19
standard deviation 0.40 0.29
range 0.00 -- 1.44 ­0.11 -- 0.29
TABLE III
FOS/RD G400H Grating Observations
Root name X position Y position X Offset Y Offset
(pixel) (pixel) (pixel) (pixel)
y2nc1002t 245.00 13.00 0.00 0.00
y2nc1005t 244.66 13.80 ­0.34 0.80
y2nc100bt 245.01 13.30 0.01 0.30
y2nc100et 244.88 13.72 ­0.12 0.72
y2nc100ht 245.53 13.31 0.53 0.31
y2nc100jt 245.72 13.07 0.72 0.07
y2nc100lt 245.41 13.41 0.41 0.41
y2nc100ot 245.64 13.43 0.64 0.43
y2nc100qt 245.60 12.74 0.60 ­0.26
y2nc100st 245.49 12.76 0.49 ­0.24
standard deviation 0.37 0.36
range ­0.12 -- 0.72 ­0.26 -- 0.80
Calibrating HST: Post Servicing Mission

FOS FILTER­GRATING WHEEL REPEATABILITY
TABLE IV
FOS/RD G570H Grating Observations
Root name X position Y position X Offset Y Offset
(pixel) (pixel) (pixel) (pixel)
y2nc1001t 245.00 13.00 0.00 0.00
y2nc1004t 245.06 13.31 0.06 0.31
y2nc1006t 245.28 13.73 0.28 0.73
y2nc1008t 245.10 13.48 0.10 0.48
y2nc100at 245.50 13.16 0.50 0.16
y2nc100dt 245.24 13.64 0.24 0.64
y2nc100gt 246.19 12.99 1.19 ­0.01
y2nc100it 246.11 13.31 1.11 0.31
y2nc100nt 245.76 13.74 0.76 0.74
y2nc100rt 245.51 13.50 0.51 0.50
y2nc100ut 245.48 13.74 0.48 0.74
standard deviation 0.40 0.29
range 0.00 -- 1.11 ­0.01 -- 0.74
TABLE V
FOS/BL G130H Grating Observations
Root name X position Y position X Offset Y Offset
(pixel) (pixel) (pixel) (pixel)
y2nc0z03t 245.00 13.00 0.00 0.00
y2nc0z07t 244.35 13.17 ­0.65 0.17
y2nc0z09t 244.79 13.15 ­0.21 0.15
y2nc0z0ct 244.52 13.44 ­0.48 0.44
y2nc0z0ft 244.37 13.00 ­0.63 0.00
y2nc0z0kt 244.36 13.68 ­0.64 0.68
y2nc0z0mt 244.23 13.17 ­0.77 0.17
y2nc0z0pt 245.31 13.67 0.31 ­0.33
y2nc0z0tt 244.16 13.51 ­0.84 0.51
standard deviation 0.39 0.31
range ­0.84 -- 0.39 ­0.33 -- 0.68
Calibrating HST: Post Servicing Mission

A. KORATKAR AND S. MARTIN
TABLE VI
FOS/BL G190H Grating Observations
Root name X position Y position X Offset Y Offset
(pixel) (pixel) (pixel) (pixel)
y2nc0z02t 245.00 13.00 0.00 0.00
y2nc0z05t 244.86 13.00 ­0.14 0.00
y2nc0z0bt 244.92 13.20 ­0.08 0.20
y2nc0z0et 244.84 12.32 ­0.16 ­0.68
y2nc0z0ht 245.04 12.72 0.04 ­0.28
y2nc0z0jt 245.48 12.50 0.48 ­0.50
y2nc0z0lt 244.51 12.44 ­0.49 ­0.56
y2nc0z0ot 245.47 12.57 0.47 ­0.43
y2nc0z0qt 244.79 12.73 ­0.21 ­0.27
y2nc0z0st 245.04 12.54 0.04 ­0.46
standard deviation 0.30 0.28
range ­0.49 -- 0.48 ­0.68 -- 0.28
TABLE VII
FOS/BL G270H Grating Observations
Root name X position Y position X Offset Y Offset
(pixel) (pixel) (pixel) (pixel)
y2nc0z01t 245.00 13.00 0.00 0.00
y2nc0z04t 245.51 13.48 0.51 0.48
y2nc0z06t 244.56 13.13 ­0.44 0.13
y2nc0z08t 245.50 12.63 0.50 ­0.37
y2nc0z0at 245.49 12.71 0.49 ­0.29
y2nc0z0dt 245.01 13.37 0.01 0.37
y2nc0z0gt 245.13 12.84 0.13 ­0.16
y2nc0z0it 245.99 13.16 0.99 0.16
y2nc0z0nt 245.73 12.46 0.73 ­0.54
y2nc0z0rt 245.95 13.08 0.95 0.08
y2nc0z0ut 244.54 12.57 ­0.46 ­0.43
standard deviation 0.50 0.33
range ­0.46 -- 0.99 ­0.54 -- 0.37
Calibrating HST: Post Servicing Mission

FOS FILTER­GRATING WHEEL REPEATABILITY
References
Bracewell, R. N.: 1986, The Fourier Transform and Its Application, 2nd ed., New York.
Dahlem, M. & Koratkar, A.: 1994, `Post­COSTAR FOS Aperture Wheel Repeatablilty Measure­
ments', CAL/FOS 131,
Hartig, G.: 1984, `FOS Filter­Grating Wheel Repeatability', CAL/FOS 12,
Hartig, G.: 1985, `Improvements in the Filter­Grating Wheel Repeatability', CAL/FOS 17,
Hartig, G.: 1988, `FOS Filter­Grating Wheel Repeatability', CAL/FOS 49,
Koratkar, A., Keyes, C. & Holfeltz, S.: 1995, `Location of FOS Spectra: Cycle 4', CAL/FOS 133,
Calibrating HST: Post Servicing Mission