Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.stsci.edu/ftp/documents/fos-handbook/fos-handbook.txt Дата изменения: Fri Jul 1 00:43:12 1994 Дата индексирования: Sat Dec 22 16:56:47 2007 Кодировка: Поисковые слова: guide 8.0

FAINT OBJECT SPECTROGRAPH


INSTRUMENT HANDBOOK



A.L. Kinney



Space Telescope Science Institute
3700 San Martin Drive
Baltimore, MD 21218



Version 5.0
May 1994
 Faint Object Spectrograph Instrument Handbook Version 5.0 i


Table of Contents


INTRODUCTION 1

1. INSTRUMENT CAPABILITIES 3
1.1Spectral Resolution:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::4
1.2Exposure Time Calculations :::::::::::::::::::::::::::::::::::::::::::::::::::: 4
1.3Brightness Limits:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::5
1.4Time Resolution::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::5
1.4.1ACCUM::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::6
1.4.2RAPID::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::6
1.4.3PERIOD:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::7
1.5Polarization ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::8
1.6FOS Noise and Dynamic Range:::::::::::::::::::::::::::::::::::::::::::::::::8

2. OBSERVING MODES 29
2.1Acquiring the Target ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::29
2.1.1ACQ/BINARY::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::32
2.1.2ACQ/PEAK::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::32
2.1.3INT ACQ:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::33
2.1.4ACQ: Confirmatory :::::::::::::::::::::::::::::::::::::::::::::::::::::::: 34
2.1.5ACQ/FIRMWARE::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::34
2.1.6Early Acquisition Using WFPC2::::::::::::::::::::::::::::::::::::::::::34
2.1.7Examples :::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: 35
2.1.8Acquisition Exposure Times ::::::::::::::::::::::::::::::::::::::::::::::37
2.2Taking Spectra: ACCUM and RAPID
Spectropolarimetry: STEP-PATT = POLSCAN::::::::::::::::::::::::::::::::::::::::38

3. INSTRUMENT PERFORMANCE AND CALIBRATIONS 41
3.1Wavelength Calibrations:::::::::::::::::::::::::::::::::::::::::::::::::::::::41
3.2Absolute Photometry::::::::::::::::::::::::::::::::::::::::::::::::::::::::::41
3.3Flat Fields :::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: 41
3.4Sky Lines ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: 42

4. SIMULATING FOS 43

5. REFERENCES 45

APPENDIX A. TAKING DATA WITH FOS 46

APPENDIX B. DEAD DIODE TABLES, C. Taylor 49

APPENDIX C. GRATING SCATTER, M. Rosa 54

APPENDIX D. FOS WAVELENGTH COMPARISON SPECTRA, C.D. Keyes 58

APPENDIX E. FAINT OBJECT SPECTROGRAPH INSTRUMENT
SCIENCE REPORTS 67

APPENDIX F. EXPOSURE LOGSHEETS 70

APPENDIX G. POST-COSTAR FOS INVERSE FLAT FIELDS 76

APPENDIX H. CHANGES TO THE VERSION 5.0 INSTRUMENT HANDBOOK 81
ii Faint Object Spectrograph Instrument Handbook Version 5.0


List of Figures


Figure 1.0.1: Quantum efficiency of the FOS Flight detectors:::::::::::::::::::::::::::10
Figure 1.0.2: A schematic optical diagram of the FOS::::::::::::::::::::::::::::::::::11
Figure 1.1.1: A Schematic of the FOS Apertures projected onto the sky :::::::::::::::: 12
Figure 1.1.2: FOS Line Spread Function at 2250A::::::::::::::::::::::::::::::::::::::13
Figure 1.2.1: HST + FOS + COSTAR Efficiency, E vs. :::::::::::::::::::::::::::::14
Figure 1.2.2: Light transmitted by apertures after deployment of COSTAR :::::::::::::16
Figure 1.2.3: Simulation of Detected Counts-s-1 -diode-1 for
Post-COSTAR FOS 0.900(1.0) aperture::::::::::::::::::::::::::::::::::::::::::::::::17
Figure 1.4.1: Duty Cycle versus Read-time for Period mode ::::::::::::::::::::::::::::19
Figure 1.5.1: FOS Waveplate Retardation and Polarimeter Transmission::::::::::::::::20
Figure 1.6.1: Measured count rate versus true count rate:::::::::::::::::::::::::::::::20
Figure 2.1.0: Slews Performed After FOS Target Acquisition :::::::::::::::::::::::::::30
Figure C.1: Count Rate for Model Atmosphere for a G2V Star :::::::::::::::::::::::: 56
Figure C.2: Observed Count Rate for G2V Star, with Scattered Light ::::::::::::::::::57
Figure D.1-14: FOS Wavelength Comparison Spectra :::::::::::::::::::::::::::::::::: 60
Figure G.1-9: Post-COSTAR FOS Inverse Flat Fields ::::::::::::::::::::::::::::::::::76



List of Tables


Table 1.0.1 FOS Instrument Capabilities:::::::::::::::::::::::::::::::::::::::::::::::21
Table 1.1.1 FOS Dispersers::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::22
Table 1.1.2 FOS Apertures ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: 23
Table 1.1.3 FOS Line Widths (FWHM) as a Function of Aperture Size :::::::::::::::: 24
Table 1.2.1 FOS Observed Counts Sec-1 Diode-1 (N) for Point Sources at
Wavelength (A):::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::25
Table 1.2.2 Simulated counts-sec-1 -diode-1 for unreddened objects in the
0.900(1.0) aperture at 15th magnitude in V::::::::::::::::::::::::::::::::::::::::::::26
Table 1.3.1 Brightness Limits::::::::::::::::::::::::::::::::::::::::::::::::::::::::::27
Table 2.1.1 Recommended FOS Acquisition Sequences ::::::::::::::::::::::::::::::::: 30
Table 2.1.2 Peak-Up Acquisition Based on Science Aperture::::::::::::::::::::::::::::31
Table 2.1.3 Reference for Table 2.1.2:::::::::::::::::::::::::::::::::::::::::::::::::::31
Table 2.1.4 FOS Visual Magnitude Limits with Camera Mirror :::::::::::::::::::::::: 35
Table 2.1.5 Minimum Exposure Times to be Entered in Exposure Logsheets ::::::::::::38
Table 2.1.6 FOS Exposure Times_Red Side and Blue Side:::::::::::::::::::::::::::::39
Table 4.1 Example Parameters in SYNPHOT to Reproduce a Spectrum ::::::::::::::::44
Table A.1 FOS Observing Parameters ::::::::::::::::::::::::::::::::::::::::::::::::: 47
Table B.1 FOS Dead and Noisy Channel Summary:::::::::::::::::::::::::::::::::::::50
Table B.2 FOS Dead and Noisy Channels History::::::::::::::::::::::::::::::::::::::52
Table C.1 Count Rate Ratios (Scattered+Intrinsic/Intrinsic) :::::::::::::::::::::::::::56
Table D.1 Wavelength and Indentification of FOS Comparison Lines :::::::::::::::::::58
 Faint Object Spectrograph Instrument Handbook Version 5.0 1


INTRODUCTION


The Faint Object Spectrograph and its use are described fully in the Version 1.0 FOS
Instrument Handbook (Ford 1985), and in the supplement to the Instrument Handbook
(Hartig 1989), from which much of this handbook is drawn. The detectors are described in
detail by Harms et al. (1979) and Harms (1982).

This version of the FOS Instrument Handbook is for the refurbished telescope, which is
affected by an increase in throughput, especially for the smaller apertures, a decrease in effi-
ciency due to the extra reflections of the COSTAR optics, and a change in focal length. The
improved PSF affects all exposure time calculations due to better aperture throughputs, and
increases the spectral resolution. The extra reflections of COSTAR decrease the efficiency
by 10-20%. The change in focal length affects the aperture sizes as projected on the sky.
The aperture designations that are already in use both in the Exposure Logsheets and in the
Project Data Base (PDB) have not been changed. Apertures are referred to here by their
size, followed by the designation used on the Exposure Logsheet. For example, the largest
circular aperture is referred to as the 0.900(1.0) aperture, while the largest paired aperture
is referred to as the 0.900paired (1.0-PAIR) aperture.

Section 1 presents the information that is needed for proposing to observe
with the FOS, i.e., for filling out Phase I Proposals. The overall instrument capabilities
are described and presented in Table 1.0.1. The spectral resolution is given in Section 1.1
as a function of grating and aperture. The data for calculating exposure times are listed
in Section 1.2 in three different ways. The easiest way to calculate exposure time is by
simply reading off the detected counts s-1 diode-1 for a disperser illuminated by a constant
input spectrum F = 1:0 x 10-14 erg cm-2 s-1 A-1 (Figure 1.2.3). The count rate can
then be scaled to the incident flux expected from the object of interest. The limits for the
brightest objects that can be observed with FOS are listed in Section 1.3. A discussion of
time resolution with the FOS, i.e., ACCUM, RAPID, and PERIOD modes, is given in Section 1.4.
Polarization is discussed in Section 1.5. The FOS noise and dynamic range are discussed in
Section 1.6.

Section 2 presents the information that is needed for observing with the
FOS after winning HST time, i.e., for filling out Phase II Proposals. The acquisition of
targets is described in Section 2.1. Examples of Exposure Logsheets that have been vali-
dated by the Remote Proposal Submission System (RPSS) are given for target acquisition
modes (for example, ACQ/BINARY), for the standard data taking mode (ACCUM), for the time
resolved mode (RAPID), and for spectropolarimetry (observed in ACCUM mode with the op-
tional parameter STEP-PATT = POLSCAN). The example Exposure Logsheets can be copied via
anonymous ftp from stsci.edu or 130.167.1.2 (STEIS). The Logsheets are in the subdirectory
proposer/documents/props_library, and are called fos_handbook5_example. Caveat emptor.

Section 3 describes briefly the current calibrations for wavelength, absolute pho-
tometry, and flat field calibrations of the FOS. See Chaper 16 of the HST Data Handbook
for a detailed description of FOS calibration. The HST Data Handbook is available through
the User Support Branch, and is available on-line on STEIS.

Section 4 describes how to simulate FOS spectra with the "synphot" package,
which runs in the ST Science Data Analysis System (STSDAS) under IRAF. The simulator,
developed by K. Horne, allow input of a large variety of spectra, and incorporate the current
calibration files for the FOS.

The details of data taking are given in Appendix A, along with the FOS observing pa-
rameters both in the nomenclature of Exposure Logsheets, and in the nomenclature of FOS
2 Faint Object Spectrograph Instrument Handbook Version 5.0


headers. Appendix A gives also the equations for calculating the start time of any time re-
solved exposure. Appendix B lists the dead diode tables of December 6, 1993. Appendix C,
by M. Rosa, gives a method to estimate the scattered light contribution for a number of
spectral types. Appendix D, by C.D. Keyes, supplies line lists and spectra of comparison
lamps for wavelength calibration. Appendix E is a compendium of recent FOS calibration
reports, including science verification reports. Calibration reports can be obtained by re-
questing copies from Bonnie Etkins (see below). Appendix F contains Exposure Logsheet
examples for different FOS modes.
The FOS Instrument Scientists and relevant ST ScI contacts are:


Tony Keyes, I.S. 410-338-4975keyes@stsci.edu
Anne Kinney, I.S. 410-338-4831kinney@stsci.edu
Anuradha Koratkar, I.S.410-338-4470koratkar@stsci.edu
Bonnie Etkins, Secretary410-338-4955etkins@stsci.edu
User Support Branch 410-338-4470usb@stsci.edu
Research Support Branch410-338-1082analysis@stsci.edu



The procedures for creating a Phase II proposal are being reviewed and revised as this
is written. We strongly recommend that users check the Phase II documentation carefully.
We also recommend checking on STEIS at that time for a revised version of this Instrument
Handbook.
 Faint Object Spectrograph Instrument Handbook Version 5.0 3


1. INSTRUMENT CAPABILITIES


The Faint Object Spectrograph has wavelength coverage on the blue side from 1150A to
5400A (FOS/BL), and on the red side from 1620A to 8500A (FOS/RD). There are both low
spectral resolution (= 250) and high resolution (= 1300) modes, as discussed
with examples in Section 1.1. The brightest objects observable with FOS have magnitudes
from V 6 (for a G2V star) to V 8 (for a B0V star or for an object with spectral shape
of f / -1 ; see Table 1.3.1 for brightness limits of all gratings and spectral types. For
magnitudes V 20, the target counts are approximately the same as the detector dark
counts (0.007 counts s-1 diode-1 on the blue side, and 0.01 counts s-1 diode-1 on the red
side) for a G2V star observed with the red side or for a B0V star observed with the blue
side.

These general traits of FOS blue side (FOS/BL) and red side (FOS/RD) are given in Ta-
ble 1.0.1.

The Faint Object Spectrograph has two Digicon detectors with independent optical
paths. The Digicons operate by accelerating photoelectrons emitted by the transmissive
photocathode onto a linear array of 512 diodes. The individual diodes are 0.3100wide along
the dispersion direction and 1.2100tall perpendicular to the dispersion direction. The de-
tectors span the wavelength range on the blue side from 1150A to 5400A (FOS/BL) and on
the red side from 1620A to 8500A (FOS/RD). The quantum efficiency of the two detectors is
shown in Figure 1.0.1. The optical diagram for the FOS is given in Figure 1.0.2. The FOS
entrance apertures are 3.60from the optical axis of HST.

Dispersers are available with both high spectral resolution (1 to 6A diode-1 , =
1300) and low spectral resolution (6 to 25A diode-1 , = 250). The actual spectral
resolution depends on the point spread function of HST, the dispersion of the grating, the
aperture used, and whether the target is physically extended.

The instrument has the ability to take spectra with high time resolution ( 0:03 seconds,
RAPID mode), and the ability to bin spectra in a periodic fashion (PERIOD mode). Although
FOS originally had polarimetric capabilities, the post-COSTAR polarimetry calibrations
were not exercised before the writing of this document, so the capabilities post-refurbishment
are as yet unknown. See STEIS postings for the most up to date information on the status
of polarimetry.

There is a large aperture for acquiring targets using on-board software (3:700x 3:700,
designation 4.3). A variety of science apertures are available; a large aperture for collecting
the maximum light (effectively 3:700x 1:200, designation 4.3); several circular apertures with
sizes 0.8600(1.0), 0.4300(0.5), and 0.2600(0.3); and paired square apertures with sizes 0.8600
(1.0-PAIR), 0.4300(0.5-PAIR), 0.2100(0.25-PAIR), and 0.0900(0.1-PAIR), for isolating spatially
resolved features and for measuring sky. In adition, a slit and several barred apertures are
available (see Figure 1.1.1).

The blue side sensitivity decreased at a rate of about 10% from launch until 1994.0 but
now appears to be more stable. The red side sensitivity is generally stable to within 5%, but
was observed to decrease more rapidly in cycles 1 and 2 in a highly wavelength dependent
fashion between 1800A and 2100A, affecting gratings G190H, G160L, and to a lesser degree
G270H. The flat fields for these 3 gratings have changed little since early 1992. Flat fields
will be obtained in the large 3.600x1:200aperture (4.3) for the G190H, G160L, and the G270H
gratings quarterly begining March, 1994 to continue to monitor this affect. The sensitivity
of both the blue and the red detectors is being monitored approximately every 2 months in
cycle 4.
4 Faint Object Spectrograph Instrument Handbook Version 5.0


1.1 Spectral Resolution

The spectral resolution depends on the point spread function of the telescope, the dis-
persion of the grating, the diode width, the spacecraft jitter, the aperture, and whether the
target is extended or is a point source. Table 1.1.1 lists the dispersers, their wavelengths, and
their dispersions (Kriss, Blair, & Davidsen 1991). All available FOS apertures are listed in
Table 1.1.2 with their designation as given in HST headers, their size and shape. Figure 1.1.1
shows the FOS entrance apertures overlaid upon each other, together with the diode array.
The positions of the apertures are known accurately and are highly repeatable.
The spectral resolution (FWHM) is given as a function of aperture in Table 1.1.3 in
units of diodes for a point source at 3400A and for a uniform, extended source. The FWHM
does not vary strongly as a function of wavelength, so that this FWHM, together with the
dispersion of the gratings given in Table 1.1.1, can be used to approximate the effective
spectral resolution.
o Example. Observing a point source using the red side with the G270H grating in the
3:700x 1:200aperture (4.3) gives a spectral resolution of


FWHM = 0:96 diodex 2:05A diode-1;


FWHM = 1:97A :

The same observation with the 0.2600(0.3) slit would have a spectral resolution of


FWHM = 0:92 diodex 2:05A diode-1;


FWHM = 1:89A :

Line spread functions computed from a model point spread function at 2250A through
the FOS apertures are shown in Figure 1.1.2 in units of microns, where 1 diode width = 50
microns. FOS line spread functions are available in the HST Archive.


1.2 Exposure Time Calculations

The information necessary to calculate exposure time is given here in several forms. First,
the HST + COSTAR + FOS efficiencies (Figure 1.2.1), aperture throughputs (Figure 1.2.2),
and wavelength dispersions (Table 1.1.1), are given together with a series of relations between
count rate and input spectra (Table 1.2.1). Then, count rate per diode at the wavelength
corresponding approximately to the peak sensitivity of the given grating is provided in tab-
ular form for a number of spectral types for objects with V=15 (Table 1.2.2). Finally, the
count rate per diode is shown in Figure 1.2.3 for both detectors and all gratings, assuming a
flat input spectrum (F / 0 = 1:0 x 10-14 erg cm-2 s-1 A-1 ) observed throught the 0:900
aperture (1.0).
o Example using Table 1.2.1. The count rate for a point source with flux of F =
3:5 x 10-15 erg cm-2 s-1 A-1 at 3700A using the red detector, in the 0:900aperture (1.0),
with the G400H grating, is given by equation 1 in Table 1.2.1,


N = 2:28 x 1012F ()E T :

where F = 3:5 x 10-15 erg cm-2 s-1 , = 3700A, = 3:0A (from Table 1.1.1), the
efficiency is E = 0:052 (from Figure 1.2.1), and the throughput is T = 0:95 (from Fig-
ure 1.2.2), so that
 Faint Object Spectrograph Instrument Handbook Version 5.0 5




N = 4:4 counts s-1diode-1:

The exposure time for a desired signal-to-noise ratio per resolution element is then given
by
SNR2
t = ______;
N

which for SNR = 20 (for example), gives t = 400=4:4 counts sec-1diode-1 = 91 s. For a
source with a count rate comparable to the dark count rate d, this equation becomes

!
SNR2 1 + N =d
t = ______ _________ :
N N =d


o Example using Table 1.2.2. As a comparison, count rates for objects of represen-
tative spectral type with V=15.0 are given in Table 1.2.2 at the wavelengths corresponding
to the peak response of a given grating. The example given above corresponds to an object
with power law F / -2 , V=15.0, observed with the G400H grating on the red side.
o Example using Figure 1.2.3. Alternatively, the count rate for observations in the
0:900(1.0) aperture can be read directly from Figure 1.2.3 and scaled to the appropriate flux.
For the example given above, with F = 3:5 x 10-15erg cm-2 s-1 A-1, the count rate per
diode at 3700 A is given by N = (3:5x10-15=1:0x10-14)xn counts sec-1diode-1, where
n is the count rate as given in Fig. 1.2.3. N = 0:35 x 12:0 = 4:2 counts s-1 diode-1 .
To calculate the count rate in other science apertures, the count rate must be corrected
according to the relative throughputs according to aperture, in Figure 1.2.2.
When observing in time resolved modes, the total observing time can become dominated
by the read-out time for FOS data. Section 1.4 below discusses the time to read-out the
FOS in the context of RAPID observations.


1.3 Brightness Limits

The photocathode can be damaged if illuminated by sources that are too bright. The
brightness limits of the detectors have been translated into a limit of total counts detected
in 512 diodes per 60 seconds_the overlight limit. If the overlight limit is exceeded in a 60
second interval, the FOS automatically safes_i.e., the FOS shuts its aperture door, places all
wheels at their rest position, and stops operation. The overlight protection limit is 1:2 x 108
counts per minute summed over the 512 diodes for the gratings and 3x106 counts per minute
for the mirror. The visual magnitudes for unreddened stars of representative spectral types
corresponding to this limiting count rate are given in Table 1.3.1 for all grating settings.
The restrictions on target brightness are also found in the Bright Object Constraints Table
of the Proposal Instructions (Table 5.15).


1.4 Time Resolution

The manner in which FOS data are obtained depends on which of the modes (e.g. ACCUM,
RAPID, or PERIOD) are used.
FOS data are acquired in a nested manner, with the innermost loop being livetime plus
deadtime (see Appendix A for a full description of data taking). The next loop sub-steps
the diode array along the dispersion direction (X direction), with steps one-quarter of the
diode width (12.5 micron, or 0.07600). To minimize the impact of dead diodes, this loop of
6 Faint Object Spectrograph Instrument Handbook Version 5.0


data-taking is continued by sub-stepping in steps of one-quarter of the diode width, but
starting at the adjacent diode. This over-scanning is repeated until spectra are obtained
over 5 continuous diodes, or a total of 20 sub-steps.
A typical data taking sequence would divide the exposure time into twenty equal bins,
and then perform the sequence of (livetime + deadtime), stepped four times. That sequence
would be performed 5 times, each time stepping to the next diode. As each of the 5 over-
scanned spectra are obtained, they are added to the same memory locations of the previous
spectra, so that the over-scanning does not increase the amount of data. The data taking is
then performed as (livetime + deadtime) x sub-stepping x over-scanning, or


(LT + DT) x 4 x 5:



1.4.1 ACCUM

FOS observations longer than a few minutes are automatically time resolved. Spectra
taken in a standard manner in ACCUM mode are read out at regular intervals. The red side
(FOS/RD) is read out at 2 minute intervals, while the blue side (FOS/BL) is read out at 4
minute intervals. The standard output data for ACCUM mode preserve the time resolution in
"multi-group" format. Each group of data has associated group parameters with information
that can be used to calculate the start time of the interval, plus a spectrum for each 2 minute
(for red side, 4 minute for blue side) interval of the observation. Each consecutive spectrum
(group) is made up of the sum of all previous intervals of data. The last group of the data
set contains the spectrum from the full exposure time of the observation. For details on data
formats, see Part VI of the HST Data Handbook (ed. Baum 1994).


1.4.2 RAPID

For observations needing higher time resolution, RAPID mode reads out FOS data at
a rate set by the observer with the parameter READ-TIME. The shortest READ-TIME is 0.036
seconds. RAPID data is also in group format but contains a header only at the beginning of
the data. Each group then contains group parameters with FOS related information followed
by the spectrum for one time segment. (Of particular interest among the group parameters
is FPKTTIME, which is used to derive the start time for each individual exposure, as given
in Appendix A.)
READ-TIME is equal to livetime plus deadtime plus the time to read out FOS (see Appendix
A and Welsh, Keyes, & Chance 1994),


READTIME = (LT + DT) x INTS x NXSTEPS x OVERSCAN x YSTEPS x SLICE

x NPATT + ROT :


where NXSTEPS=SUBSTEP, and is usually set to 4, OVERSCAN=COMB=MUL, and is
usually set to 5, YSTEPS=Y-SIZE, and is almost always set to 1, and where ROT refers to
the Read-Out Time. The Read-Out Time for FOS is dependent on the telemetry rate, and
on the amount of data to be read out, which is dependent on number of diodes (i.e., , the
wavelength range) being observed, as well as on the sub-stepping.

15 1024
ROT = ___x _______x NSEG(WORDS) x SUBSTEPS x YSTEPS
14 RAT E
.
 Faint Object Spectrograph Instrument Handbook Version 5.0 7


where RATE is the telemetry rate, and NSEG(WORDS) is given by


NSEG = 1 if (WORDS - 50) < 61

WORDS - 50

NSEG = 1 + 1 + INTEGER _____________ otherwise
61


where WORDS = (NCHNLS + OVERSCAN - 1) , NCHNLS is the number of diodes to
be read out (with a maximum of 512 and a minimum of 46 for an OVERSCAN of 5), and
INTEGER truncates to the next lowest integer. To achieve the fastest READ-TIMEs, the RATE
of reading data can be increased from the default telemetry rate of 32kHz to 365kHz, the
wavelength region can be decreased, and sub-stepping set to 1. The amount of data being
taken by FOS must be decreased to achieve the fastest READ-TIME because a smaller amount
of data can be read out in a faster time. The relation between number of diodes read out
and wavelength coverage can be derived from Table 1.1.1. (Table 1.1.1 is accurate to within
a few Angstroms since it is based on data that was not corrected for the geomagnetically
induced image drift, Kriss, Blair, & Davidsen 1991).

The observer should be aware of the fact that the percentage of time spent accumulating
data in RAPID can be very small depending on how the parameters are set. Figure 1.4.1,
from Welsh et al. (1994) shows the duty cycle, or ratio of time spent accumulating data
over READ-TIME as a function of READ-TIME. Given the two values of telemetry rate (32,000,
and 365,000) and the three possible values of SUBSTEP (4, 2, and 1), there are six curves
for duty cycle. The parameters should be set to maximize duty cycle, while maintaining the
resolution and wavelength coverage necessary for the scientific objectives.

o Example For an FOS RAPID observation requiring a READ-TIME of 0.2 seconds, Fig-
ure 1.4.1 shows that to achieve this short READ-TIME, the SUB-STEP must be set equal to 1,
and that the telemetry rate is automatically set to the high (365kHz) rate. This results in
W ORDS = 512 + 5 - 1, and NSEG = 1 + 1 + INT ([516 - 50]=61) = 9, leading to a ROT
given by

ROT = 15=14(1024=365000) x 9


ROT = 0:02705:


Thus the READ-TIME, made up of Read-Out Time plus (LT+DT) times the multiplicative
factors given above can be obtained by setting SUB-STEP=1, and READ-TIME=0.2.


1.4.3 PERIOD


For objects that have a well known period, FOS data can be taken in PERIOD mode in such
a way that the period is divided into BINS, where each bin has a duration of t = period=BINS.
The period of the object is specified by the parameter CYCLE-TIME. The spectrum taken during
the first segment of the period, t1, is added into the first memory location. The spectrum
taken during the second segment, t2, is added to a contiguous memory location, and so
on. The number of segments that a period can be divided into depends on the amount of
data each spectrum contains, which depends on the number of sub-steps, whether or not the
data are overscanned, and how large a wavelength region is to be read out. If the full range
of diodes are read out, and the default observing parameters are used, 5 BINS of data can be
stored. PERIOD mode data are single group, with a standard header followed by the spectra
stored sequentially, where there are BINS spectra.
8 Faint Object Spectrograph Instrument Handbook Version 5.0


The data size, which cannot exceed 12,288 pixels, is given for PERIOD by


Data size= (NCHNLS + MUL - 1) x SUBSTEP x BINS


where BINS applies to PERIOD mode only. BINS is the number of time-segments into which
the periodic data are divided. If the observer needs a larger number of BINS than 5, the
wavelength range can be decreased, or the sub-stepping can be decreased to 2 or 1. (See
Table 1.1.1 for relation between number of diodes [NCHNLS], and wavelength dispersion.)


1.5 Polarization

The deployment of COSTAR resulted in two extra reflections for light entering the FOS.
These extra reflections introduce instrumental polarization so that polarization measure-
ments have become difficult and possibly infeasible with the FOS. However, the section on
polarization is included here because the FOS team felt that the G190H and the G270H
grating may be used for polarization observations if it can be recalibrated properly. Cali-
brations in Cycle 4 will be used to quantify the polarimetric capability. Thus, proposals for
the use of FOS polarimetric capabilities can be submitted, and GO's are recommended to
survey updates to polarimetric capabilities on STEIS (see chapter 4 for logging onto STEIS).
A Wollaston prism plus rotating waveplate can be introduced into the light beam to
produce twin dispersed images of the slit with opposite senses of polarization at the detector
(Allen & Angel 1982). Although there are two waveplates available, only waveplate B is
currently recommended for use, and only in the G190H and the G270H gratings. (See Allen
& Smith 1992 for polarization calibration results.)
Although the "A" waveplate was designed to do well at Lyff 1216A, the split spectra are
not well separated by the "A" waveplate, so that the polarization at Lyff cannot be observed.
Linear polarization observations should use the "B" waveplate and gratings G130H, G190H,
and G270H.
The sensitivity of the polarizer depends upon its throughput efficiency. The detector can
observe only one of the two spectra produced by the polarizer at one time, so that another
factor of two loss in practical throughput occurs. The count rate is given by


Count rate(pol)= Count rate(FOS)x jthrx 0:5;


where jthris found in Figure 1.5.1.


1.6 FOS Noise and Dynamic Range

The minimum detectable source levels are set by instrumental background, while the
maximum accurately measurable source levels are determined by the response times of the
FOS electronics.
When the FOS is operating outside of the South Atlantic Anomaly, the average dark
count rate is roughly 0.01 counts s-1 diode-1 for the red detector and 0.007 counts s-1
diode-1 for the blue detector (Rosenblatt et al. 1992). However, Rosenblatt et al. note
that the background count rate varies with geomagnetic latitude so that higher rates are
observed at higher latitudes. Furthermore, there is some evidence that the above dark rates
systematically underestimate the actual dark counts by 30%.
The detected counts s-1 diode-1 plots given in Figure 1.2.3 for an input spectrum with
constant flux of F = 1 x 10-14 erg cm-2 s-1 A-1 can be compared with the observed dark
count rate to determine the limiting magnitude for the FOS.
 Faint Object Spectrograph Instrument Handbook Version 5.0 9


o Example. For an object observed at 2600A with the red side G270H grating in the
0.900(1.0) aperture, an incident flux of F = 4:7 x 10-17 erg cm-2 s-1 A-1 would produce
a count rate comparable to the red side dark rate. For an object observed at 2600A with
the blue side G270H grating in the 0.900(1.0) aperture, an incident flux of F = 4:7 x 10-17
erg cm-2 s-1 A-1 would produce a count rate comparable to the blue side dark rate.
In the other extreme, for incident count rates higher than approximately 100,000 counts
s-1 diode-1 , the observed output count rate does not have an accurate relation with the true
input count rate. Figure 1.6.1 shows a determination of the relation between true count rate
and observed count rate, as measured by Lindler & Bohlin (1986, measured for high count
rates for the red side only). For observed count rates above 50,000 counts s-1 diode-1 , the
correction exceeds a factor of 2 and the accuracy decreases drastically. By the time a true
count rate of 200,000 counts s-1 diode-1 is reached, the error in the correction to the true
rate is of order 50%. A correction is applied in the pipeline processing to account for this
detector non-linearity at high count rates.
10 Faint Object Spectrograph Instrument Handbook Version 5.0





Figure 1.0.1: Quantum efficiency of the FOS Flight detectors.
 Faint Object Spectrograph Instrument Handbook Version 5.0 11





Figure 1.0.2: A schematic optical diagram of the FOS.
12 Faint Object Spectrograph Instrument Handbook Version 5.0





Figure 1.1.1: A Schematic of the FOS Apertures projected onto the sky. The upper panel
(a) shows the array of 0:3000x 1:2100diodes projected across the center of the 3:6600x 3:7100
target acquisition aperture. The target acquisition aperture and the single circular apertures
position to a common center. The pairs of square apertures position to common centers with
respect to the target acquisition aperture as shown in the figure. Either the upper aperture
(the "A" aperture, which is furthest from the HST optical axis) or the lower aperture (the
"B" aperture, which is closest to the HST optical axis) in a pair can be selected by an
appropriate y-deflection in the Digicon detectors. The lower panel (b) shows three more slits
that position to the center of the target acquisition aperture. The bottom of the figure (c)
shows the orientation of the direction perpendicular to the dispersion (shown as a dashed
line) relative to the HST V2, V3 axes. The FOS x-axis is parallel to the diode array and
positive to the left; the y-axis is perpendicular to the diode array and positive toward the
upper aperture. The angle between the FOS/BLUE and the FOS/RED slit orientation is
73.6 degrees.
 Faint Object Spectrograph Instrument Handbook Version 5.0 13





Figure 1.1.2: The solid curves are spectral line spread functions for various FOS apertures.
Ordinate shows relative intensity in the dispersion direction. Units are microns, where one
diode (a nominal spectral resolution element) is 50 microns.
14 Faint Object Spectrograph Instrument Handbook Version 5.0





Figure 1.2.1: HST + FOS + COSTAR Efficiency, E vs. .
 Faint Object Spectrograph Instrument Handbook Version 5.0 15





Figure 1.2.1 (cont.): HST + FOS + COSTAR Efficiency, E vs. .
16 Faint Object Spectrograph Instrument Handbook Version 5.0





Figure 1.2.2: Fraction of light transmitted by the apertures after deployment of COSTAR
for a perfectly centered point source.
 Faint Object Spectrograph Instrument Handbook Version 5.0 17





Figure 1.2.3: Detected counts-s-1 -diode-1 for the post-COSTAR FOS 0:900(1.0) aperture.
Input spectrum is F = 1 x 10-14 erg-cm-2 -s-1 -A-1 (F / -2 ; V = 13:9).
18 Faint Object Spectrograph Instrument Handbook Version 5.0





Figure 1.2.3 (cont.): Detected counts-s-1 -diode-1 for the post-COSTAR FOS 0:900(1.0)
aperture. Input spectrum is F = 1 x 10-14 erg-cm-2 -s-1 -A-1 (F / -2 ; V = 13:9).
 Faint Object Spectrograph Instrument Handbook Version 5.0 19





Figure 1.4.1: Percentage of time spent accumulating data in RAPID mode as a function of
READ-TIME, telemetry rate, where the high telemetry rate (365kHz) is marked by a solid line
and the low rate (32kHz) is marked by the dashed line, and SUBSTEP, where SUBSTEP=1
is marked by squares, SUBSTEP=2 is marked by diamonds, and SUBSTEP=4 is marked
by filled dots.
20 Faint Object Spectrograph Instrument Handbook Version 5.0





Figure 1.5.1: FOS waveplate retardation (left) and polarimeter transmission (Allen & Angel,
1982).



Figure 1.6.1: Measured count rate versus true count rate (Lindler & Bohlin 1986). The lower
curve is a plot of the upper curve expanded by 10 in the x-direction.
 Faint Object Spectrograph Instrument Handbook Version 5.0 21


Table 1.0.1

FOS INSTRUMENT CAPABILITIES



__________________________________________________________________________

Wavelength coverage1 FOS/BL: 1150A to 5400A in several grating settings.
FOS/RD: 1620A to 8500A in several grating settings.
Spectral resolution High: = 1300:
Low: = 250:
Time resolution t 0:036 seconds.
Acquisition aperture 3:700x 3:700(4.3).
Science apertures2 Largest: 3:700x 1:200(4.3).
Smallest: 0.0900square paired (0.1-PAIR).
Brightest stars observableV3 8 for B0V, V 6 for G2V.
Dark count rate FOS/BL: 0.007 counts s-1 diode-1 .
FOS/RD: 0.01 counts s-1 diode-1 .
Example exposure times4 F1300= 2:5 x 10-13, SNR=20/(1.0A), t=180s.
0.900aperture F2800= 1:3 x 10-13, SNR=20/(2.0A), t=5.8s (FOS/BL).
F2800= 1:3 x 10-13, SNR=20/(2.0A), t=4.0s (FOS/RD).
__________________________________________________________________________


1 See Table 1.1.1 for grating dispersions and wavelength coverage.
2 See Table 1.1.2 for available apertures.
3 See Table 1.3.1 for brightest objects observable, which are strongly dependent on spectral

type and grating.
4 See Section 1.2 for exposure time calculations, and Table 1.2.1 for count rates for objects

with a variety of spectral types. The example given here is for 3C273.
22 Faint Object Spectrograph Instrument Handbook Version 5.0


insert Table 1.1.1
 Faint Object Spectrograph Instrument Handbook Version 5.0 23


insert Table 1.1.2
24 Faint Object Spectrograph Instrument Handbook Version 5.0


insert table 1.1.3
 Faint Object Spectrograph Instrument Handbook Version 5.0 25


insert table 1.2.1
26 Faint Object Spectrograph Instrument Handbook Version 5.0


insert table 1.2.2
 Faint Object Spectrograph Instrument Handbook Version 5.0 27


Table 1.3.1
Brightness Limits1
__________________________________________________________________________________

Red Side Brightness Limits
Spectral _____________________________________________________________________
Type B - V G190H G270H G400H G570H G780H G160L G650L PRISM MIRROR
_________________________________________________________________________________
07V -0:32 9.3 9.8 9.0 7.9 5.7 10.3 8.2 10.5 14.8
B0V -0:30 9.1 9.5 9.0 7.9 5.6 10.1 8.1 10.3 14.6
B1.5V -0:25 8.8 9.4 8.8 7.9 5.7 10.0 8.1 10.2 14.5
B3V -0:20 8.1 8.7 8.6 7.9 5.7 9.5 8.1 9.8 14.0
B6V -0:15 7.9 8.6 8.4 7.8 5.7 9.3 8.0 9.6 13.8
B8V -0:11 7.1 7.9 8.4 7.8 5.7 9.0 8.0 9.3 13.5
A1V +0:01 5.9 7.0 8.0 7.8 5.8 8.6 7.9 8.9 13.1
A2V +0:05 5.7 6.8 8.0 7.8 5.8 8.5 7.9 8.8 13.0
A6V +0:17 5.1 6.5 7.9 7.8 5.8 8.4 7.8 8.7 12.9
A7V +0:20 5.0 6.4 7.8 7.8 5.9 8.3 7.8 8.7 12.8
A9V +0:28 4.4 6.2 7.7 7.8 6.0 8.2 7.7 8.6 12.7
F0V +0:30 4.2 6.1 7.7 7.7 6.0 8.2 7.7 8.5 12.7
F5V +0:44 3.6 6.1 7.5 7.7 6.1 8.1 7.6 8.5 12.6
F7V +0:48 2.9 5.6 7.4 7.7 6.1 8.0 7.6 8.3 12.5
F8V +0:52 2.7 5.4 7.3 7.7 6.1 8.0 7.6 8.3 12.5
G2V +0:63 2.1 5.3 7.3 7.7 6.2 7.9 7.6 8.3 12.4
G6V +0:70 _ 5.2 7.2 7.7 6.2 7.9 7.5 8.2 12.4
K0V +0:81 _ 4.3 7.0 7.7 6.2 7.8 7.5 8.1 12.3
K0III +1:00 _ 3.4 6.6 7.6 6.3 7.7 7.5 8.0 12.2
K5V +1:15 _ 3.5 6.3 7.6 6.4 7.6 7.4 7.9 12.1
K4III +1:39 _ 2.1 6.0 7.5 6.5 7.5 7.3 7.8 12.0
M2I +1:71 _ _ 5.5 7.4 6.5 7.3 7.2 7.6 11.8
ff2 = 1 6.9 7.9 8.0 7.8 6.4 8.8 7.8 9.1 13.3
ff2 = 2 5.8 7.1 7.6 7.7 6.7 8.4 7.7 8.8 12.9
ff2 = -2 -0:46 10.2 10.3 9.2 8.0 5.8 10.9 8.2 10.8 15.4
T = 50,000O 9.6 9.9 9.0 7.9 5.7 10.4 8.2 10.5 14.9
_________________________________________________________________________________
1The FOS can be damaged if illuminated by sources that are too bright. If illuminated by

targets brighter than the V magnitude limits given here, the instrument will go into safe mode,
shutting its aperture door and stopping operations. Table 1.3.1 is for objects observed in the 3.600
(4.3) aperture.
2Where F / -ff.
28 Faint Object Spectrograph Instrument Handbook Version 5.0


______________________________Table_1.3.1._Continued.____________________________
_________________________________________________________________________________
Blue Side Brightness Limits
Spectral _____________________________________________________________________
Type G130H G190H G270H G400H G570H G160L G650L PRISM MIRROR
_________________________________________________________________________________
07V 7.2 8.5 9.3 8.4 5.0 9.8 6.5 10.0 14.3
B0V 7.0 8.3 9.1 8.3 5.0 9.7 6.4 9.8 14.2
B1.5V 6.6 8.0 9.0 8.2 4.9 9.5 6.3 9.7 14.0
B3V 5.8 7.3 8.3 7.9 4.9 8.9 6.3 9.2 13.4
B6V 5.4 7.1 8.2 7.7 4.9 8.7 6.1 9.0 13.2
B8V 4.5 6.2 7.5 7.6 4.9 8.2 6.2 8.6 12.7
A1V 2.3 5.1 6.5 7.2 4.8 7.5 6.0 8.0 12.0
A2V _ 4.8 6.4 7.2 4.8 7.4 6.0 7.9 11.9
A6V _ 4.2 6.1 7.1 4.7 7.2 5.8 7.8 11.7
A7V _ 4.2 6.1 7.0 4.7 7.2 5.8 7.7 11.7
A9V _ 3.5 5.9 6.9 4.7 7.1 5.7 7.6 11.6
F0V _ 3.4 5.8 6.8 4.6 7.0 5.6 7.5 11.5
F5V _ 2.9 5.7 6.7 4.6 6.9 5.5 7.4 11.4
F7V _ _ 5.3 6.6 4.5 6.7 5.4 7.2 11.2
F8V _ _ 5.1 6.5 4.5 6.6 5.3 7.1 11.1
G2V _ _ 5.0 6.5 4.5 6.5 5.3 7.1 11.0
G6V _ _ 4.9 6.3 4.4 6.4 5.2 7.0 10.9
K0V _ _ 4.1 6.1 4.4 6.1 5.1 6.7 10.6
K0III _ _ 3.1 5.7 4.3 5.7 4.8 6.3 10.2
K5V _ _ 3.2 5.3 4.2 5.4 4.6 6.0 9.9
K4III _ _ _ 4.9 4.0 5.0 4.3 5.6 9.5
M2I _ _ _ 4.4 3.8 4.6 4.0 5.2 9.1
ff1 = 1 4.2 6.1 7.5 7.3 4.6 8.0 5.7 8.4 12.5
ff1 = 2 2.8 5.0 6.7 6.9 4.5 7.4 5.4 7.9 11.9
ff1 = -2 8.6 9.4 9.8 8.6 5.0 10.6 6.6 10.4 15.1
T = 50,000O 7.5 8.8 9.4 8.4 5.0 10.0 6.4 10.0 14.5
_________________________________________________________________________________
1Where F / -ff.
 Faint Object Spectrograph Instrument Handbook Version 5.0 29


2. OBSERVING MODES


The procedures for creating a Phase II proposal are being reviewed and revised as this
handbook is written. We strongly recommend that users check the Phase II documentation
carefully, and that users check STEIS for updates and revisions to the Handbook.


2.1 Acquiring the Target

The HST pointing is accurate and reliable. The most common source of error in target
acquisition is incorrect user-supplied coordinates. To demonstrate the accuracy of the HST
pointing achieved after guide star acquisition, Figure 2.1.0 shows the slews performed to
center the target in the science aperture after FOS target acquisition, for observations taken
after early 1991. The position V2 = 0:0 and V3 = 0:0 in Figure 2.1.0 corresponds to perfect
initial pointing. Figure 2.1.0 shows that, using positions derived from GASP, about 70% of
the blind pointings with correct coordinates fall within 100of the aperture center. However,
an onboard target acquisition is still necessary with the FOS to center the target in the
science aperture.
Interactive acquisition mode (set with the SPECIAL REQUIREMENT INT ACQ) and three on-
board acquisition modes (ACQ/BINARY, ACQ/PEAK, and ACQ/FIRMWARE) are described below. Dur-
ing an onboard acquisition, the FOS performs the acquisition, calculates the small offset
required to center the target in a science aperture, and makes the offset. In contrast, during
an interactive acquisition there must be a real time contact with HST, and the observer must
be present at the ST ScI to interpret the image. Because of the probability of confusion when
looking at an FOS white light picture, we believe that in nearly all cases a WFPC2 assisted
target acquisition will be a better scientific choice than an interactive FOS acquisition (INT
ACQ). However, ACQ does also provide an important means of verifying, after the fact, where
the FOS aperture was positioned on the target during a science exposure, for both WFPC2
early acquisition and for onboard acquisitions of targets in complex fields.
The FOS acquisition aperture is 3.700x 3:700square (4.3). In order to have a 95% chance
of placing a star in this aperture, the star must have an RMS positional error with respect
to the guide stars of less than 1.000.
Additional acquisitions are not necessary when switching from the red side to the blue
side for the 0.900(1.0) or larger apertures, since the aperture positions are known accurately.
Once a target has been acquired into a large science aperture and observed with one detector,
a slew can be performed to place the target directly into the large aperture for the other
detector. Such "side-switch" slews would not be accurate enough to place objects in the
0:200x 1:200slit (0.25X2.0) or the 0:300aperture (0.3), however. In these cases an additional
ACQ/PEAK is required as summarized in Tables 2.1.1, 2.1.3, and 2.1.3 (see also section 2.1.3
and examples in Appendix F).
30 Faint Object Spectrograph Instrument Handbook Version 5.0





Figure 2.1.0: Slews performed after FOS target acquisition (after guide star acquisition)
to center the target in the science aperture. The average red side offset, based on 128
acquisitions, is: V2 = 0:0800 0:0500, V3 = -0:0500 0:0600. The average blue side offset,
based on 148 acquisitions, is: V2 = 0:1500 0:0500, V3 = -0:0600 0:0500. Seventy percent of
the pointings after guide star acquisitions are within 100of the target.


Table 2.1.1
Recommended FOS Acquisition Sequences for
Acquisitions starting with ACQ/BINARY


Science First Second Dimension Acquisition
Aperture Acq Acq X x Y Aperture
______________________________________________

4.3 ACQ/BINARY -
1.0 ACQ/BINARY -
0.5 ACQ/BINARYACQ/PEAK 3X3 0.5
0.3 ACQ/BINARYACQ/PEAK 4X4 0.3
SLIT ACQ/BINARYACQ/PEAK 9X1 SLIT
______________________________________________
 Faint Object Spectrograph Instrument Handbook Version 5.0 31


Table 2.1.2
Peak-Up Acquisition Based on Science Aperture
(for objects that can only be acquired with peak-up)



Aperture Number of First Second Third Fourth Throughput1
to be Used Stages Stage Stage Stage
____________________________________________________________________

4.3 2 A B 100%
4.3 3 A B C 100%
1.0 3 A B C 97%
0.5 3 A B C 95%
0.3 4 A B C D OR E 94%
0.1 4 A B C E 43%
SLIT2 4 A B C F 93%
____________________________________________________________________


1 Ratio of throughput given the centering error associated with the acquisition, over through-

put with perfect centering. 2 SLIT pointing uncertainty is larger in the direction prependic-
ular to dispersion than parallel to dispersion for the acquisition sequence given. Note that
all FOS calibration acquisitions use 4-stage peak-ups A,B,C, and D; therefore, if precision
flat fields are required, such a 4-stage peak-up should be used regardless of science aperture.


Table 2.1.3
Reference for Table 2.1.2



Type Aperture Search-Search-Scan- Scan- Critical?CenteringOverhead
Size-x Size-y Step-xStep-y Error Time
______________________________________________________________________

A 4.3 1 3 _ 1.204 N 0.600 6.6min
B 1.0 6 2 0.602 0.602 N 0.4300 12.9min
C 0.3 5 5 0.172 0.172 N 0.1200 22min
D 0.3 5 5 0.05 0.05 Y 0.0400 22min
E 0.1-PAIR-A 5 5 0.05 0.05 Y 0.0400 22min
F SLIT 7 1 0.057 _ Y 0.0400 9.4min
______________________________________________________________________


Side-switching will be allowed ONLY for those objects where the total time (both sides
combined) is less than 6 orbits. ST ScI reserves the right to change the order of the sides
(and gratings) to schedule the observation most efficiently.

Three rules apply to any side-switching specification:

1. Specify the target acquisition (TA) exposures on the Exposure Logsheet for one side,
while in a comment specifying the parameters such as exposure time, FAINT, BRIGHT, and
spectral element for the other side of FOS. Such a specification will allow easy change
of order of the detectors. If the proposer feels the TA must be performed with a specific
detector, this must be stated in the General Form, question 5.

2. The special requirement GROUP NOGAP should be used in the Exposure Logsheet to link all
exposures of the target.
32 Faint Object Spectrograph Instrument Handbook Version 5.0


3. If the proposer has a scientific need to obtain the observations in a specific grating order,
the special requirement SEQ NO GAP should be used.

See the Exposure Logsheet lines 3.0 through 4.3 (Appendix F) for an example of a
side-switching specification.


2.1.1 ACQ/BINARY


ACQ/BINARY is the method of choice for targets with well known energy distributions, but
should not be used for variable sources, sources of unknown color, or sources extended by
much more than 1 diode, or 0.300. The method has a restricted dynamic range of brightness.
Specifically, target brightness uncertainty should be less than 0.5 magnitudes for the use of
ACQ/BINARY. Objects of poorly known color should be acquired with ACQ/PEAK.

During an ACQ/BINARY, the camera mirror reimages the FOS focal plane onto the Digicon.
Acquisition of the target is performed not by moving the telescope, but by deflecting the
image of the target acquisition aperture on the photocathode until the target has been placed
on the Y edge of the diode array. ACQ/BINARY finds first the number of stars in the 3.700x
3.700acquisition aperture (designation 4.3) by integrating at three different positions in the
Y-direction. The program locates the target in one of the three strips, measures its count
rate, and locates the target in the X direction. The algorithm then positions the target
on a Y-edge of the diode array by deflecting the image across the diode array through a
geometrically decreasing sequence of Y-deflections until the observed count rate from the
star is half that when the object is positioned fully on the diode array. ACQ/BINARY is the
preferred acquisition mode for point sources.

Although ACQ/BINARY is designed to obtain the Nth brightest star in a crowded field
by setting the optional parameter NTHSTAR, acquisitions in crowded fields have not been
attempted.

There should be about 300 counts in the peak pixel for each Y-step that is on-target
for Binary Search to succeed. If the number of counts in the peak is significantly larger
than 300, the tolerances for whenpthe_target is on the edge of the diode array become very
small since they are based on N statistics. Typical centering error after Binary Search is
<~0.1500. If the Binary Search algorithm fails to converge on a position with half the counts

of the original target, the telescope slews to the last position of binary search, i.e., to the X
position of the target and the last Y-deflection.

A target must lie within the range of counts specified by the Optional Parameters BRIGHT
and FAINT. We recommend that BRIGHT and FAINT be set to allow for targets 10 times brighter
and 5 times fainter than expected. Since the maximum number of Y-steps in Binary Search
is 11, the default values for the parameters are BRIGHT = 300 x 11 x 10 = 33,000 and FAINT
= 300 x 11/5 = 660.

An Example of an ACQ/BINARY on an offset star followed by an FOS observation of the
target star is given on lines 1 and 2 of the sample Logsheets in Appendix F.


2.1.2 ACQ/PEAK


During ACQ/PEAK the telescope slews and integrates at a series of positions on the sky
with a science aperture in place. At the end of the slew sequence the telescope is returned
to the position with the most counts; no positional interpolation is performed. In the case
of an ACQ/PEAK into a barred aperture, or when using the Optional Parameter TYPE=DOWN, the
telescope is returned to the position with the fewest counts. ACQ/PEAK is a relatively inefficient
 Faint Object Spectrograph Instrument Handbook Version 5.0 33


procedure because a minimum of ~ 42 seconds per dwell is required for the telescope to
perform the required small angle maneuvers. Tables 2.1.2 and 2.1.3 list the recommended
combinations of peak-ups for acquisition of targets according the the size of the science
aperture, along with the errors in position, and the throughput errors associated with those
positional errors. Table 2.1.3 lists the overhead times involved in each stage of an ACQ/PEAK.
o Example A peak-up into the 0.2600(0.3) aperture would require a 1X3 peak-up into the
3.700x 3.700(4.3), followed by a 6X2 peak-up into the 0.8600(1.0) aperture, followed by a
5X5 peak-up into the 0.2600(0.3) aperture. The overhead time required for this three stage
peak-up is 41.50 minutes.
This mode is used for objects too bright to acquire with the camera mirror in place, for
objects too variable to acquire with ACQ/BINARY, for centering targets in the smallest apertures,
and for positioning bright point sources on the bars of the occulting apertures in order to
observe any surrounding nebulosity. For bright object acquisitions, the science grating is put
in place before the acquisition. Examples of ACQ/PEAK are given on lines 10.3 through 13.3
of the sample Exposure Logsheets in Appendix F. Tables 2.1.1, 2.1,2, and 2.1.3 summarize
recommended ACQ/PEAK sequences.
To acquire objects into the smallest FOS apertures (0.2600(0.3), 0.200(0.25-PAIR), 0.0900
(0.1-PAIR), and 0.200X 1.700slit (0.25 X 2.0)), first use a normal ACQ/BINARY acquisition, fol-
lowed by a "critical" ACQ/PEAK into the science aperture (see Tables 2.1.1, 2.1.2, and 2.1.3).
(For objects too bright to observe with the camera mirror in place, use instead a series of
non-critical peak-ups as shown in Table 2.1.2, followed by a "critical" peak-up into the small
science aperture.) The "critical" ACQ/PEAK must have a high number of counts to place the
target in the center of these smallest apertures ( 10000) and spacing between dwells of
order D/5, where D is the diameter of the Peak Up aperture. See Table 2.1.6 below for ex-
posure times. The non-critical ACQ/PEAK requires shorter exposure time and spacing between
dwells of order D/2. An example is given in Logsheet lines 3 through 4.1 in Appendix F.
Count rates must not exceed the safety limits for the mirror or the grating selected (see
Table 1.3.1 and Table 2.1.4).
An N by M pattern with steps of size X.X, Y.Y can be specified by setting SEARCH-
SIZE-X=N, SEARCH-SIZE-Y=M, and SCAN-STEP-X=X.X, SCAN-STEP-Y=Y.Y. Examples are given in the
Exposure Logsheets lines 10.3 through 13.3 in Appendix F.


2.1.3 INT ACQ

The mode ACQ, when used with the SPECIAL REQUIREMENT ``INT ACQ FOR", maps the acqui-
sition aperture and sends the image to the ground in real time. The apparent elongation
of stars in the y-direction caused by the shape of the diodes (0.2600x 1.2100) is removed on
the ground by multiplying the picture by an appropriate matrix. After the picture has been
restored, the astronomer measures the position of the target on the image. The small offset
required to move the target to the center of one of the science apertures is calculated and
uplinked to the telescope; after the slew is performed the science observations begin.
A modified form of interactive acquisition, the dispersed-light interactive acquisition
utilizing IMAGE mode, may be employed for acquisition of sources in which spectral features
of known wavelength are prominent. This method has proven quite useful for planetary
satellite acquisitions. Spacecraft overheads for this procedure are no different than the
overheads for conventional INT/ACQ.
34 Faint Object Spectrograph Instrument Handbook Version 5.0


2.1.4 ACQ: Confirmatory


ACQ can also be used after another type of acquisition to provide a picture which shows
where HST is pointed in FOS detector coordinates. The Exposure Logsheets provide an
example (lines 5-8) of an ACQ/BINARY of an offset star followed by an offset onto the nucleus
of M81. In this example, after the science observation is made, a (white light) picture of the
aperture is taken by using ACQ to verify the aperture position.


2.1.5 ACQ/FIRMWARE


ACQ/FIRMWARE is an engineering mode that maps the camera-mirror image of the aperture
in X and Y with small, selectable Y increments. The FOS microprocessor filters the aperture
map and then finds the Y-positions of the peaks by fitting triangles through the data.
Firmware is less efficient than Binary Search, and fails if more than one object is found
within the range of counts set by the observer (BRIGHT and FAINT). This mode is not generally
recommended.


2.1.6 Early Acquisition Using WFPC2


We recommend using WFPC2 assisted target acquisition when there will be more than
two stars in the 3.700(4.3) acquisition aperture or when there will be intensity variations
across the acquisition aperture which are larger than a few percent of the mean background
intensity. A WFPC2 image of the field is taken several months in advance of the science
observation. The positions of the target and an offset star are measured in the image and
then (at least 2 months later) the positions are updated on the Exposure Logsheet, and the
offset star is acquired with ACQ/BINARY and finally the FOS aperture is offset onto the target.
There is only about a 30% chance that the same guide stars will be in the Fine Guidance
Sensors (FGS) when the subsequent FOS observations are made. With new guide stars, the
1oe uncertainty in any position is 0.300. Uncertainty in position of the telescope when slewing
by 10 due to the spacecraft roll is of order 0.0500. The Wide Field Camera II is made up of
three chips of size 1.250 on a side and a fourth chip 0:60 on a side. Offsets larger than 3000
should be discussed with the User Support Branch.

The first step in a WFPC2 assisted target acquisition is to use a SPECIAL REQUIREMENT on
the Exposure Logsheets to specify the exposure as an EARLY ACQ which must be taken at least
two months before the FOS observations (see lines 5 through 8 on the exposure logsheet in
Appendix F). The camera, exposure time, filter, and centering of the target in the image
should be chosen such that the picture will show both the target and an isolated (no other
star within 500) offset star which is brighter than mV = 20 and more than 1 magnitude
brighter than the background (magnitudes per square arcsecond). In order to insure that
an appropriate offset star will be in the WFPC2 image, the centering of the target in the
WFPC2 field should be chosen by measuring a plate or CCD image. The Target List for the
FOS exposures should provide the offset star with nominal coordinates and with position
given as TBD-EARLY. (See example lines 4 and 5 on the Target List in Appendix F.) The
Target List also should list the position of the offset star as RA-OFF, DEC-OFF, and FROM the
target. Alternatively, the offsets can be given as XI-OFF and ETA-OFF, or R, PA, see the Phase
II Proposal Instructions Section 5.1.4.3 on Positional Offsets.

After the WFPC2 exposure has been taken and the data have been received, the next
step is to get the picture onto an image display so you can i), choose an offset star, ii) measure
its right ascension and declination, and iii) measure the right ascension and declination of the
 Faint Object Spectrograph Instrument Handbook Version 5.0 35


target relative to the offset star. An STSDAS task (stsdas.wfpc.metric) is available currently
to extract pointing and roll angle information from the WFPC header and to convert WFPC
pixels to right ascension and declination. Upon calibration, the task "metric" will also be
available for WFPC2. If this program is not available, you will need to patch your WFPC2
image into the Guide Star Catalog reference frame. Based on your choice of an offset star,
the ST ScI will choose a pair of guide stars for the FOS observations which will stay in the
"pickles" during the move from the offset star to the target. The probability that a suitable
pair of guide stars can be found increases as the separation of the offset star and the target
decreases. So, choose the offset star as close as possible to the target (but not so close as
to violate the background rule in the preceding paragraph). The final step is to send the
position of the offset star and the positional offsets to the ST ScI to update the proposal
information for your succeeding FOS observations.


2.1.7 Examples


The following section gives examples for acquiring different types of astronomical objects
based on the strengths and weaknesses of the various target acquisition methods.



o Example: Single Stars


Stars with visual magnitudes brighter than about 12thare too bright for FOS acquisitions
with the camera mirror, and observations of objects that bright will safe the instrument.
The exact limit depends on the spectral type of the star and on the detector as shown in
Table 2.1.4 below. For a more complete list see Table 1.3.1.


Table 2.1.4
FOS Visual Magnitude Limits with Camera Mirror


O7V B0V B3V A1V A6V G2V K0III-1 -2
_________________________________________________________

Red Side Limit 14.8 14.614.0 13.112.9 12.4 12.2 13.3 12.9
Blue Side Limit14.3 14.213.4 12.011.7 11.0 10.2 12.5 11.9
_________________________________________________________



Stars that are too bright for ACQ/BINARY can be acquired by using ACQ/PEAK with one of
the high dispersion gratings instead of the camera mirror (see lines 10.3 through 10.6 on the
Exposure Logsheets in Appendix F). If the visual magnitude of a single star or point source
is fainter than limits given in Table 2.1.4 above, if the star does not vary by more than 0.5
magnitudes, and if the colors are known, use ACQ/BINARY for the acquisition.



o Example: Stars Projected on Bright Backgrounds


ACQ/BINARY can find successfully a star projected on a uniform background provided the
target acquisition integration time is long enough to give ~ 300 peak counts from the star
and the star is at least a magnitude brighter than the background surface brightness in mag-
nitudes per square arcsecond. If star magnitude and the background magnitude differ by less
than 1 magnitude, the star can still be acquired with ACQ/BINARY by increasing the integration
36 Faint Object Spectrograph Instrument Handbook Version 5.0


time. Alternatively, the acquisition can be accomplished by using an early acquisition with
WFPC2, followed two months later by an FOS acquisition and blind offset.
A different problem arises when the background varies across the acquisition aperture.
Because the logic in the ACQ/BINARY program drives the star to the edge of the diode array
by finding the position which gives half the maximum number of counts, any change in the
background in the Y-direction will bias the derived Y-position of the star. Simulations of
acquisitions of stars projected onto bright galaxies such as NGC 3379 show that the shot
noise in the star will determine the accuracy (rather than the spatially-variable background),
provided the star is at least 1500from the center of the galaxy.



o Example: Diffuse Sources and Complex Fields


The FOS onboard acquisition methods were designed to acquire point sources. Conse-
quently, diffuse sources and complex fields must be observed by first acquiring a star and
then offsetting to the desired position in the source. The most accurate positioning of the
FOS aperture on the source will be accomplished by using an early WFPC2 assisted target
acquisition. In many programs, the interesting positions in the source will be chosen on the
basis of WFPC2 images. If the imaging program is planned as described in the section on
WFPC2 assisted TAs, the science images can be used for the acquisition.



o Example: Nebulosity Around Bright Point Sources


The optimal FOS aperture position for a bright point source surrounded by nebulosity
will depend on the distribution and brightness of the nebulosity relative to the point source.
If high spatial resolution images show that the nebulosity has a scale length of a few tenths of
an arcsecond and is relatively symmetrical around the source, then the signal-to-noise ratio
may be maximized by placing the stellar source on the occulting bar of one of the occulting
apertures and observing simultaneously the nebulosity on both sides of the occulting bar.
When using this approach, you should first use Binary Search to position the source near the
center of the occulting aperture. The second step is to use a Peak Down in the Y-direction
to position the stellar source on the occulting bar. An example is given in lines 11, 12, and
13 of the Exposure Logsheet in Appendix F.
If high resolution images show that the nebulosity is rather asymmetrical, the best
approach may be to observe the nebulosity with one of the small circular apertures. In that
case the bright stellar source should be acquired with ACQ/BINARY, followed by an ACQ/PEAK,
followed by an offset onto the nebulosity.



o Example: ACQ/PEAK


ACQ/PEAK is now the method of choice for targets with variability of order 0.5 magnitudes
or greater. This method utilizes a spatial scan series of exposures to locate the target. The
position with the maximum signal is chosen; no positional interpolation is performed. This
method must be used also for targets brighter than about V=13 (see Table 1.3.1).
For a planetary object, an area of sky larger than the TA aperture (3.700x3:700, 4.3) may
have to be searched, plus the object may be too bright to acquire with ACQ/BINARY. By using
the target acquisition aperture, an effective aperture of size 3.700x 1:200(designation 4.3) is
 Faint Object Spectrograph Instrument Handbook Version 5.0 37


available and an area of 7.400x 7:400can be searched. The first two steps of the ACQ/PEAK are
to perform a 2 x 6 dwell pattern with the (effective) 3.700x 1:200aperture (4.3). Then the
ACQ/PEAK sequence outlined in Tables 2.1.1, 2.1.2, and 2.1.3 for the appropriate aperture can
be used.
The most time efficient way to acquire a bright target with the FOS is to use the 3:700x
1:200(4.3) aperture in a 1 x 3 dwell pattern, followed by a 6 x 2 dwell pattern into a 0.900
(1.0) aperture. The third step depends on the science to be done. As with the example given
above, for an object to be centered into the 0:900(1.0) aperture, a non-critical ACQ/PEAK can
be performed into the 0:400(0.5) aperture. These types of acquisitions are shown on lines
10.3 through 13.3 on the Exposure Logsheet in Appendix F and summarized in Tables 2.1.1,
2.1.2, and 2.1.3.


2.1.8 Acquisition Exposure Times

There should be about 300 counts in the peak of the Y-step that is centered on the star
in an ACQ/BINARY exposure. The maximum number of Y-steps which can be taken during
ACQ/BINARY is 11. Table 2.1.6 summarizes the total exposure time for an ACQ/BINARY, i.e.,
the time per Y-step multiplied by 11, for various types of stars. The exposure times in Ta-
ble 2.1.6, scaled to the magnitude of the target, are the times that should be entered in the
Exposure Logsheets. There is a minimum integration time that can be entered on the Expo-
sure Logsheet. The minimum is constrained by the FOS livetime limit given in Table 2.1.5.
If the exposure time must be larger than that calculated from Table 2.1.6 to accommodate
the minimum time, the values for the optional parameters BRIGHT and FAINT must be set to
reflect the total number of counts expected.


TIMETable 2:1:5
BRIGHT = 33; 000 x ______________
TIMETable 2:1:6

TIMETable 2:1:5
FAINT = 660 x ______________
TIMETable 2:1:6


For example, for a red side ACQ/BINARY of a 12th magnitude offset K0III star, 0.29s is the
exposure time derived from Table 2.1.6, but the minimum exposure time is 0.66s. The
default values of BRIGHT and FAINT must then be multiplied by the factor 0.66/0.29 = 2.28,
so that BRIGHT = 75,100 and FAINT = 1500.
The peak-up exposure times in Table 2.1.6 are calculated to produce 1000 counts in
the peak of the target image, which is the number of counts recommended for the non-
critical ACQ/PEAK described above. The ACQ/PEAK sensitivity has considerable wavelength and
aperture size dependence. A critical ACQ/PEAK into small apertures requires 10,000 counts
total to achieve a centering error that corresponds to a signal loss of less than about 2% for
the apertures smaller than 0.300. For a critical peak up, the values in Table 2.1.6 relating to
peak up must be multiplied by a factor of 10.
The times in Table 2.1.6 do not include the overhead involved in the initial setup of
parameters or the analysis time, since that overhead should not be included on the Expo-
sure Logsheet specifications. The overhead times for the lengthy ACQ/PEAK mode is given in
Table 2.1.3.
Extrapolations of acquisition exposure times for sources fainter than V=19.5 should not
be extrapolated from Table 2.1.6 because of the background noise.
38 Faint Object Spectrograph Instrument Handbook Version 5.0


2.2 Taking Spectra: ACCUM and RAPID
Spectropolarimetry: STEP-PATT = POLSCAN

Examples of exposure logsheets are included for ACCUM mode (see lines 3.0 through 4.3 in
Appendix F) and RAPID mode (see lines 11.0 through 13.0).
In RAPID mode, when a wavelength range is specified, that range will be used whether or
not there is room in memory for a larger region. Therefore, specifying a wavelength range is
not a good idea unless absolutely necessary, because it restricts the wavelength region that
is read out. The full wavelength region is often useful. For example, the background can
be determined directly from the diode array for gratings G130H, G160L, G190H, G650L,
G780H, and PRISM. The diodes below the lowest wavelength, given in Table 1.1.1, can be
used to average the actual background rate. The zero order can be monitored for G160L if
all diodes are read out. If the observer needs only a specific wavelength range to be read out,
then that range should be specified in with the keyword WAVELENGTH (column 8) of the Phase
II exposure logsheet. Otherwise, the largest possible wavelength range will be automatically
observed that is compatible with the READ-TIME requested.
The use of STEP-PATT = POLSCAN is demonstrated in the exposure logsheet lines 14.0
through 19.0. As mentioned in Section 1.5, only the G270H grating may be available for
polarimetric observations post-COSTAR; even with the G270H, full utility has not yet been
demonstrated post-COSTAR.



Table 2.1.5


Minimum Exposure Times to be Entered in Exposure Logsheets


____________________

ACQ/BINARY 0.66 sec
ACQ/FIRMWARE0.96 sec
ACQ/PEAK 0.003 sec
ACQ 3.84 sec
____________________
 Faint Object Spectrograph Instrument Handbook Version 5.0 39


Table 2.1.6
_____________________FOS_Exposure_Times_(V_=_15_unreddened)_Red_Side_____________
_________________________________________________________________________________
Spectral Peak/upPeak/upPeak/upPeak/upPeak/upPeak/upPeak/upPeak/up
_Type_____B_-_V_G190H__G270H__G400H__G570H__G780H__G650L__PRISM_MIRROR__ACQ/BIN__

07V -0:32 0.2 0.1 0.2 0.5 3.3 0.3 0.1 0.1 0.39
B0V -0:30 0.2 0.1 0.2 0.5 3.6 0.4 0.1 0.1 0.46
B1.5V -0:25 0.3 0.1 0.3 0.5 3.5 0.4 0.1 0.1 0.53
B3V -0:20 0.5 0.3 0.3 0.5 3.4 0.4 0.1 0.1 0.8
B6V -0:15 0.6 0.3 0.3 0.5 3.6 0.5 0.1 0.1 1.0
B8V -0:11 1.2 0.6 0.4 0.5 3.3 0.5 0.1 0.2 1.3
A1V +0:01 3.5 1.2 0.5 0.5 3.2 0.5 0.2 0.2 2.0
A2V +0:05 4.2 1.4 0.5 0.5 3.2 0.5 0.2 0.2 2.1
A6V +0:17 7.3 2.0 0.5 0.5 3.0 0.5 0.2 0.2 2.4
A7V +0:20 7.6 2.1 0.5 0.6 2.8 0.5 0.2 0.2 2.4
A9V +0:28 12 2.3 0.6 0.6 2.8 0.6 0.2 0.3 2.7
F0V +0:30 14 2.5 0.6 0.6 2.8 0.6 0.2 0.2 2.8
F5V +0:44 32 2.7 0.6 0.6 2.4 0.6 0.3 0.2 3.0
F7V +0:48 45 3.8 0.7 0.6 2.3 0.6 0.3 0.2 3.3
F8V +0:52 50 4.6 0.7 0.6 2.3 0.6 0.3 0.3 3.4
G2V +0:63 95 5.8 0.8 0.6 2.2 0.6 0.3 0.3 3.5
G6V +0:70 _ 5.9 0.9 0.6 2.2 0.7 0.3 0.3 3.7
K0V +0:81 _ 13 1.0 0.6 2.0 0.7 0.3 0.4 4.0
K0III +1:00 _ 30 1.5 0.6 1.9 0.7 0.4 0.4 4.5
K5V +1:15 _ 28 2.0 0.6 1.8 0.8 0.4 0.4 4.9
K4III +1:39 _ _ 2.8 0.7 1.7 0.8 0.5 0.5 5.4
M2I +1:71 _ _ 4.0 0.7 1.7 0.9 0.6 0.5 6.2
ff = 1 1.5 0.5 0.5 0.5 1.7 0.5 0.2 0.2 1.5
ff = 2 4.0 1.1 0.6 0.6 1.4 0.6 0.3 0.2 2.2
ff = -2 -0:46 0.2 0.1 0.2 0.4 3.2 0.3 0.1 0.1 0.23
_t_=_50,000O______0.2____0.2____0.3___0.5____3.0____0.4____0.1_____0.1_____0.35__

_____________________FOS_Exposure_Times_(V_=_15_unreddened)_Blue_Side____________
_________________________________________________________________________________
Spectral Peak/upPeak/upPeak/up Peak/upPeak/up Peak/upPeak/up Peak/up
__Type______G130H___G190H__G270H__G400H___G570H__G650L__PRISM__MIRROR___ACQ/BIN__

07V 1.3 0.5 0.2 0.3 5.9 1.6 0.1 0.1 0.80
B0V 1.5 0.6 0.2 0.3 6.1 1.6 0.1 0.1 0.90
B1.5V 2.1 0.8 0.2 0.3 6.7 1.8 0.1 0.1 1.1
B3V 4.6 1.5 0.3 0.4 6.2 1.9 0.2 0.1 1.8
B6V 6.6 1.8 0.4 0.5 7.0 2.1 0.2 0.2 2.2
B8V 14 3.9 0.8 0.6 7.0 2.1 0.2 0.2 3.4
A1V _ 11 1.8 0.8 7.4 2.5 0.4 0.4 6.1
A2V _ 13 2.3 0.8 7.1 2.5 0.4 0.4 6.9
A6V _ 25 2.9 0.9 8.2 2.9 0.5 0.5 8.1
A7V _ 26 3.0 1.0 8.2 3.0 0.5 0.5 8.6
A9V _ 41 3.4 1.1 8.2 3.4 0.5 0.6 9.7
F0V _ 43 3.7 1.1 8.7 3.4 0.6 0.6 10.
F5V _ 68. 3.9 1.3 9.0 4.0 0.6 0.7 12.
F7V _ _ 5.7 1.4 9.2 4.3 0.8 0.8 14.
F8V _ _ 6.9 1.5 9.2 4.4 0.8 0.9 15.
G2V _ _ 7.7 1.6 9.2 4.6 0.9 1.0 16.
G6V _ _ 7.9 1.7 9.4 5.2 1.0 1.1 18.
K0V _ _ 19. 2.1 11. 5.8 1.3 1.4 22.
K0III _ _ 39 3.2 11 7.3 2.0 2.0 33
K5V _ _ 40 4.5 13 9.0 2.5 2.6 44
K4III _ _ _ 6.6 16. 12 3.5 4.0 65.
M2I _ _ _ 11. 19. 16. 5.0 5.6 92.
ff = 1 20. 4.2 0.8 0.7 8.6 3.2 0.3 0.3 4.2
ff = 2 60. 12. 1.6 1.1 9.6 4.2 0.4 0.5 7.3
ff = -2 0.4 0.2 0.1 0.2 5.6 1.4 0.1 0.1 0.44
__T_=_50,000O_1.0____0.4____0.2_____0.3____6.5_____1.6____0.1_____0.1_____0.70___
40 Faint Object Spectrograph Instrument Handbook Version 5.0


Notes to Table 2.1.6


Note: Exposure time must be multiplied by 100:4(V -15).
1 Optimal exposure times for ACQ/BINARY and ACQ/FIRMWARE are calculated to detect 300 peak

counts in the peak pixel of the target.
2 ACQ/PEAK into the 0:2600(0.3) aperture requires 10000 total counts.
3 Exposure times for ACQ/PEAK into all apertures excluding the slit are calculated to detect

1000 total counts for non-critical acquisitions. For critical centering into apertures smaller
than 0.300, multiply the exposure times by a factor of 10. Note that the exposure time
for ACQ/PEAK must be multiplied by the inverse throughput of the aperture used
(T , see Figure 1.2.2). Although the exact factor depends on the input spectrum, the
approximate multiplicative factors are 1.1 for the 0.900apertures (1.0), 1.2 for the 0.400
apertures (0.5), and 1.25 for the 0.2600aperture (0.3).
 Faint Object Spectrograph Instrument Handbook Version 5.0 41


3. INSTRUMENT PERFORMANCE AND CALIBRATIONS


For additional information on instrument performance, see Calibrating the Hubble Space
Telescope: Proceedings of a Workshop, ed. Blades & Osmer (1994) and Part VI of the HST
Data Handbook, ed. Baum (1994). Both documents are available on-line via STEIS (see
section 4 for information on accessing STEIS).


3.1 Wavelength Calibrations


All FOS wavelengths are vacuum wavelengths, both below 2000A and above.

Wavelength offsets between the internal calibration lamp and a known external point
source are based on observations of the dwarf emission line star AU Mic, that have been
corrected for geomagnetically induced image drift (Kriss, Blair, & Davidsen 1992). On the
red side, the mean offset between internal and external source is +0:176 0:105 diodes. On
the blue side, the mean offset is -0:102 0:100 diodes. These offsets are not included in
the pipeline reduction wavelength calibration. With the observed dispersion reported by
Kriss, Blair, & Davidsen, velocity measurements based on single lines in FOS spectra have a
limiting accuracy of roughly 20 km s-1 if wavelength calibrations are obtained at the same
time, with no filter-grating wheel motion (i.e., NO GAP), and if the target is well centered in
the science aperture. (See Appendix D for line lists and spectra of the comparison lamps
for each detector/disperser combination.) If simultaneous wavelength calibrations are not
obtained, the non-repeatability of order 0.3 diodes in the positioning of the filter-grating
wheel will dominate the errors in the zero point of the wavelength scale (Hartig 1989).


3.2 Absolute Photometry


The post-refurbishment absolute photometric calibrations are performed by observing
some or all of the standard stars G191B2B (WD0501+527), BD+28D4211, BD+75D325,
HZ-44, and BD+33D2642 in the large (3:700x 1:200, 4.3) aperture. As of the writing of this
handbook, there is little information on time dependence of the sensitivity.


3.3 Flat Fields


Observations of two hot spectrophotometric standard stars (G191B2B and BD+28D4211)
are used to produce spectral flat fields for all usable FOS detector/disperser combinations. A
highly precise target acquisition strategy (4-stage ACQ/PEAK with pointing uncertainty of
0.03 arcsec) is used for these observations so that filter-grating wheel repeatability (0.10 arc-
sec) is the dominant source of uncertainty in photocathode sampling. All post-refurbishment
flat fields will be derived via the so-called superflat technique (see Lindler et al.- CAL/FOS-
088; flat field article by Keyes in HST Calibration Workshop, ed. Blades &d Osmer (1994);
see also Part VI, Chapter 16 in the HST Data Handbook, ed. Baum (1994)).

Appendix G provides figures showing preliminary flat field structure for all usable FOS
detector/disperser combinations derived from SMOV epoch (March, 1994) superflat observa-
tions. These figures are provided as an approximate guide since analysis of the observations
is not complete at the time of the writing of this handbook. However, no additional strong
(greater than 5% deviation from unity) features are expected to appear.

It must be emphasized that FOS flat field corrections are intended to remove photo-
cathode granularity typically on the scale of 10 pixels or less. If high precision flat fields
are required for scientific objectives, observers should attempt to attain the same pointing
42 Faint Object Spectrograph Instrument Handbook Version 5.0


accuracy (described above) used for FOS flat field calibration observations so that the science
target illuminates the same portion of the photocathode as was sampled by the calibration
observations.
During Cycle 4 several observations will be made to attempt to quantify the change in flat
field granularity structure as a function of target mis-centering perpendicular to dispersion.
Most pixel ranges in typical flat fields display deviations of 1-2% about the mean value of
unity or about a local running mean, however, some substantial (5-50%) features do occur.
Fragmentary evidence from pre-refurbishment flats indicates that photocathode granularity
in these strong features can change by 25% on the scale of a diode height (1.2 arcsec). Should
such a feature occur in the vicinity of an important spectral line and target centering be
less accurate than that of the flat field calibration observation, then the observed flux could
be affected by a currently unknown amount. We note that there is such a feature in the
1500-1550 A. range (affects C IV resonance doublet) of the FOS/BLUE G160L spectrum.
Normally one epoch of flat field measurement is made per cycle for each detector/disperser
combination to be used, however in the pre-refurbishment era some substantial temporal vari-
ation in FOS/RED G190H, G160L, and, to a lesser extent, G270H flat fields was observed.
These changes were most profound in the first two and one-half years after launch; little
variation has been noted since November, 1992. Nonetheless, these gratings will continue to
be monitored at additional periodic intervals.
Some time dependence has been observed in the red side flat fields. Red side data taken
after January 1992 and before refurbishment can be flat fielded with the data most appro-
priate to the observation. The STSDAS task getreffile will refer the user to the appropriate
flat field. Red side data taken between October 1990 and January 1992 will be difficult to
flat field because of the lack of time-dependent flat fields available between 1990 and 1992.
The time dependence (or lack thereof) of the post-refurbishment flat fields has not been
established at the writing of this handbook.


3.4 Sky Lines

The lines of geocoronal Ly ff1216 and OI 1304 appear regularly in FOS spectra, with a
width determined by the size of the aperture (see Table 1.1.3). Occasionally, when observing
on the daylight side of the orbit, the additional sky lines of OI 1355 and of OII 2470 can
also be seen. The second order Ly ff1216 appears sometimes in the G160L grating.
 Faint Object Spectrograph Instrument Handbook Version 5.0 43


4. SIMULATING FOS


A simulator developed by K. Horne is available in the Space Telescope Science Data
Analysis System (STSDAS) in the package synphot. Details about the STSDAS synphot
package can be found in the Synphot User's Guide, by H. Bushouse, Sep. 1993, STScI, and
in Appendix D of the HST Data Handbook (1994, Baum). The synphot data, which is not
part of standard STSDAS, must be retrieved and installed to run synphot, as described in
Appendix D.

o Logging on to STEIS. To log onto STEIS, type ftp stsci.edu or ftp 130.167.1.2.
If prompted for Name, enter anonymous. Otherwise enter user anonymous. The pass-
word is your full email address. You are now in a UNIX-FTP environment. Enter get
README to transfer the instructions to your home account. Most information relevant to
data reduction is located within the /instrument_news and the cdbs directories. (For detailed
information on STEIS, see the HST Data Handbook, Part XI, Chapter B, ed. Baum 1994.)

Synphot can be used to "observe" an arbitrary input spectrum with any FOS configu-
ration to produce a predicted spectrum of detector counts s-1 diode-1 . This can be done
using one of several tasks in the synphot package, including countrate, calcspec, and plspec.
The operation of each of these tasks involves specifying the desired FOS observing mode,
the input spectrum, and the form of the output spectrum. The choice as to which task to
use depends on the desired results. For producing spectra in units of counts s-1 diode-1
the countrate task is the easiest to use and its input parameters are set up to mimic those
found on FOS exposure log sheets.

For example, to reproduce the countrate spectrum shown in Figure 1.2.3 for the FOS
blue side with grating G190H and the 1.0 aperture, the parameters for the countrate task
would be set as given in Table 4.1. Note the inclusion of the argument "costar" in the
aperture parameter, which is necessary for the task to use the aperture throughput data
that are appropriate for the PSF and plate scale provided by COSTAR. In this example
the input spectrum is specified using the unit function, which produces a spectrum that has
constant flux as a function of wavelength. The two arguments for the unit function specify
the flux level and units (in this case 1:0 x 10-14 ergs s-1 cm-2 A-1 or "flam").

The task evaluates the spectrum on a wavelength grid that is automatically selected to
match the dispersion (Angstroms diode-1 ) of the chosen observing mode. The computed
spectrum will be written to the STSDAS table "spectrum.tab", which will contain two
columns of wavelength and flux values, where the wavelengths will be in units of Angstroms
and the spectrum in units of counts diode-1 . With exptime=1, as in this example, the flux
units are then essentially counts s-1 diode-1 . The spectral data in this table can be plotted
using, for example, the STSDAS task sgraph (e.g. sgraph "spectrum.tab wavelength flux").

In addition to the unit function used in this example, synphot also has built-in black-
body and power-law functions that can be used to synthesize spectra of those forms. For
example, in the countrate task you could set "synspec = bb(8000)" and "synmag = 14.5 V"
to synthesize an 8000 K blackbody spectrum that is normalized to a V magnitude of 14.5.
You could also specify "synspec = pl(3500,2)" and "synmag = 13.9 V" to obtain a power-
law spectrum of the form F() proportional to -2 (which has constant flux in wavelength
space), normalized to a V magnitude of 13.9.

The userspec parameter can also be set to read spectral data from an existing table,
such as data you may already have for a particular object. There are several spectral atlases
available on STEIS that you can use as input to Synphot. Current holdings include a library
of HST standard star spectra, Kurucz model atmospheres, a spectrum synthesis atlas from
44 Faint Object Spectrograph Instrument Handbook Version 5.0


G. Bruzual, the Bruzual-Persson-Gunn-Stryker spectral atlas which has wavelength coverage
from the near-UV to the near-IR, and the optical stellar atlas from Jacoby, Hunter, &
Christian (1985). See Appendix B of the Synphot User's Guide for information on how to
obtain these data.


Table 4.1: Example parameters in Synphot countrate task to simulate an FOS observation
of a flat-spectrum source with flux=1.0 x 10-14 ergs s-1 cm-2 A-1 , as in Figure 1.2.3.
Parameter SettingDefinition
______________________________________________________________________

output spectrum.tab Output table name
instrument fos Science instrument
detector blue Detector used
spec_elem g190h Spectral elements used
aperture 1.0,costarAperture / field of view
cenwave INDEF Central wavelength (HRS only)
userspec unit(1.e-14,flam)User supplied input spectrum
synspec Synthetic spectrum
synmag Magnitude of synthetic spectrum
refwave INDEF Reference wavelength
reddening 0. Interstellar reddening in E(B-V)
exptime 1. Exposure time in seconds
verbose yes Print results to STDOUT?
count_tot INDEF Estimated total counts
count_ref INDEF Estimated counts at reference wavelength
refdata Reference data
______________________________________________________________________



Acknowledgements

I would like to thank Ralph Bohlin, Ian Evans, Ron Gilliland, Tony Keyes, Anuradha
Koratkar and Rex Saffer for their careful reading of the Handbook. I would also like to
thank Tony Keyes for updating the FOS simulator, for producing both bright limits and
target acquisition times, and for making many useful suggestions. I would like to thank
Cindy Taylor for producing most of the tables and the plots. I would like to thank Howard
Bushouse, Jen Christensen, Anne Gonnella, Keith Horne, Buell Jannuzi, Pete Reppert, Sue
Simkin, William Welsh, Rogier Windhorst, and Meg Urry for providing comments, criticisms,
and corrections.
 Faint Object Spectrograph Instrument Handbook Version 5.0 45


5. REFERENCES


Allen, R.G., & Angel, J.R.P. 1982, FOS Spectropolarimeter Performance, FOS Instrument
Handbook, Version 1, ST ScI, page C-1.

Allen, R.G., & Smith, P.S. 1992, FOS Polarimetry Calibrations, Instrument Science Report
CAL/FOS-078.

Baum, S. 1994, ed. HST Data Handbook, ST ScI.

Bazell, D. 1990, Synphot Users Guide, ST ScI.

Blades, J.C., Osmer, S.J. 1994, ed. Calibrating the Hubble Space Telescope: Proceedings of a
Workshop, ST ScI.

Burrows, C., & Hasan, H. 1991, Telescope Image Modelling User Manual, ST ScI.

Bushouse, H. 1993, Synphot User's Guide, STScI.

Caldwell, J., & Cunningham, C.C. 1992, Grating Scatter in the FOS and the GHRS, Science
Verification 1343 Interim Report.

Ford, H.C. 1985, FOS Instrument Handbook, ST ScI.

Harms, R.J. 1982, The Space Telescope Observatory, ed. D.N.B. Hall, (Special Session of
Commission 44, IAU General Assembly, Patras, Greece, August, 1982; NASA CP-2244).

Harms, R.J., Angel, R., Bartko, F., Beaver, E., Bloomquist, W., Bohlin,R., Burbidge, E.M.,
Davidsen, A.,F., Flemming, J.C., Ford, H., & Margon, B. 1979, SPIE, 183, 74.

Hartig, G.F. 1989, Faint Object Spectrograph Instrument Handbook Supplement No. 1, ST ScI.

Hartig, G.F. 1989, FOS Filter-Grating Wheel Repeatability: Dependence on Motor Selection,
Instrument Science Report CAL/FOS-060.

Jacoby, G., Hunter, D., and Christian, C., 1984, ApJS, 56, 257.

Kriss, G.A., Blair, W.P., & Davidsen, A.F. 1991, In-Flight FOS Wavelength Calibration -
Template Spectra, Instrument Science Report CAL/FOS-067.

Kriss, G.A., Blair, W.P., & Davidsen, A.F. 1992, Internal/External Offsets in the FOS
Wavelength Calibration, Instrument Science Report CAL/FOS-070.

Lindler, D., & Bohlin, R. 1986, FOS Linearity Corrections, Instrument Science Report
CAL/FOS-025.

Morris, S.L., Weymann, R.J., Savage, B.D., & Gilliland, R.L. 1991 Ap. J. (Letters), 377,
L21.

Mount, G., & Rottman, G. 1981, The Solar spectral irradiance 1200-3184A near solar
maximum: July 15, 1980, J. Geophys. Res. 86, 9193.

Neill, J.D., Bohlin, R.C., & Hartig, G. 1992, Photometric Calibration of the Faint Object
Spectrograph, Instrument Science Report CAL/FOS-077

Rosenblatt, E.I., Baity, W.A., Beaver, E.A., Cohen, R.D., Junkkarinen, V.T., Linsky, J.B.,
and Lyons, R.W. 1992, An Analysis of FOS Background Dark Noise, Instrument Science
Report CAL/FOS-071.

Wegener, R., Caldwell, J., Owne, T., Kim, S.J., Encrenaz, T., & Comber, M. 1985 The
Jovian Stratosphere in the Ultraviolet, Icarus, 63, 222.

Welsh, W.F., Chance, D., & Keyes, C.D. 1994 High Speed Spectroscopy Using the FOS in
Rapid Mode, Instrument Science Report in preparation.
46 Faint Object Spectrograph Instrument Handbook Version 5.0


APPENDIX A.


Taking Data with FOS

Two sets of nomenclature are used to describe the taking of FOS data_those used in
the exposure logsheets to command observations, and those used in the FOS data headers.
Table A.1 gives the translation between the two, together with defaults and definitions.
FOS observations are performed in a nested manner, with the innermost nest being the
livetime of the instrument plus the deadtime (LT + DT). Table A.1 lists the parameters
in the order in which FOS observations are nested. Standard spectra are taken by sub-
stepping the diode array along the dispersion in the X direction, and then by performing the
sub-stepping five times over adjacent diodes to minimize the impact of dead diodes. The
sequence is then
(LD + DT) x 4 x 5:

The minimum livetime is 0.003 seconds. The minimum livetime plus deadtime is 0.030
seconds. Using the minimum livetime results in very inefficient observations, since data are
being taken only 0:003=0:03 = 0:1 of the time.
The user has access only to those parameters that can be set in the exposure logsheet.
For example, the user cannot set the livetime, but the user can set the product of livetime
and INTS (STEP-TIME = LTx INTS). Likewise, the user cannot explicitly set the deadtime,
but in PERIOD mode, the user can set the ratio of livetime to deadtime (DATA-RATIO = LT/DT).
For the most common mode, ACCUM, an FOS integration is constructed in the order
(LT+DT), INTS, NXSTEPS, OVERSCAN, YSTEPS, NPATT, and finally NREAD. The
total elapsed time of an integration is then given by



t = (LT + DT) x INTS x NXSTEPS x OVERSCAN x YSTEPS x NPATT x NREAD :


where NXSTEPS = SUB-STEP, and YSTEPS = Y-SIZE. This equation also gives the elapsed
time for the observation, which for standard ACCUM mode is equal to


t = (LT + DT) x INTS x 4 x 5 x 1 x NPATT x NREAD :


The number of patterns, NPATT, is set after the setting of sub-step (NXSTEPS),
OVSERSCAN, and YSTEP, to achieve the exposure time requested. When NPATT has
reached the maximum that it can be set to (256), then INTS is incremented. Obviously, this
must be done in an optimal way to ensure that the efficiency (/ LT/DT) remains high. The
maximum value for INTS is also 256.
For a RAPID observation, an FOS integration is built up in a slightly different order;
(LT+DT), INTS, NXSTEP, OVERSCAN, YSTEPS, NPATT, and finally NMCLEARS. The
total elapsed time of the observation is



t = (LT + DT) x INTS x NXSTEP x OVERSCAN x YSTEPS x NPATT x NMCLEARS :


which is usually equal to


t = (LT + DT) x INTS x 4 x 5 x 1 x NPATT x NMCLEARS :


However, the sub-stepping, the overscan values, and the wavelength range can be lowered
in RAPID to accommodate shorter time between the taking of spectra.
 Faint Object Spectrograph Instrument Handbook Version 5.0 47


For a PERIOD observation, an FOS integration is built up of


t = (LT + DT) x INTS x NXSTEP x OVERSCAN x YSTEPS x SLICES x NPATT


where SLICES = BINS. As with RAPID, x step and overscan values can be lowered to result in
a greater number of SLICES (BINS).
These equations give the elapsed time of an observation and so they can be used to
calculate the actual start time of any observation, by subtracting them from the first packet
time (FPKTTIME) which is given in the group parameter at the beginning of every group
of "multi-group" data.


Start Time= FPKTTIME - t

The start time of the entire observation is also given in the data header as EXPSTART. All
times in the header, including the first packet time, and the start time, are given in units of
Modified Julian Date, which is the Julian date minus 2400000.5. The Modified Julian Date
for 1993 is given by:


MJD = 48987:0 + day of year+ fraction of day from0hUT:



Table A.1

FOS Observing Parameters
Listed in Order of Execution
_________________________________________________________________________

Exposure FOS Default Definition
Logsheet Header
_________________________________________________________________________

_ LIVETIME 0.500 sec(LT) Time FOS is integrating.
_ DEADTIME 0.010 sec(DT) Overhead time.
_ INTS _ Number of times to execute (LT+DT)
SUB-STEP NXSTEP 4 Number of steps of size diode/NXSTEP
in direction of dispersion.
COMB OVERSCAN YES Whether or not to execute x stepping to
remove the effects of dead diodes.
For COMB= YES, MUL=5.
For COMB= NO, MUL=1.
Y-SIZE YSTEPS 1 Number of steps perpendicular to dispersion.
BINS SLICE 5 For PERIOD only, equal to 1 otherwise.
Number of bins to divide one period into.
_ NPATT _ Number of times to execute the pattern so as
to achieve the exposure time.
_ NREAD _ For ACCUM only, equal to 1 otherwise.
For readouts short enough to correct for GIMP
_ NMCLEARS _ For RAPID only, equal to 1 otherwise. Number
of times to clear data so as to read new data.
_________________________________________________________________________
The FOS header value for OVERSCAN is equal to the value for MUL.
48 Faint Object Spectrograph Instrument Handbook Version 5.0


Table A.1. Continued

Additional FOS Observing Parameters


__________________________________________________________

Exp.Log. FOS Default Definition
__________________________________________________________

STEP-TIME LTxINTS 0.5 Available in RAPID and PERIOD.
DATA-RATIO LT/DT Maximum Available in PERIOD only.
__________________________________________________________
 Faint Object Spectrograph Instrument Handbook Version 5.0 49


APPENDIX B.


Dead Diode Tables



C. Taylor


Occasionally one of the 512 diodes on the red or the blue side becomes very noisy, or
ceases to collect data. Since launch, the FOS has lost 3 diodes on the blue side and 2 diodes
on the red side. In addition, several diodes on each side have become noisy and have been
disabled. When a diode goes bad in orbit, there is a delay before that diode behavior is
discovered, and another delay time before that diode is disabled so that its effect is removed
from the data. Table B.1 lists the current (as of December 6, 1993) disabled diodes. Table B.2
lists the history of the diodes that have been disabled, when they were discovered to be bad,
and when they were removed from action. The channels are numbered in this table from 0
to 511, while they are numbered in the STSDAS tasks from 1 to 512.
50 Faint Object Spectrograph Instrument Handbook Version 5.0





insert table B.1
 Faint Object Spectrograph Instrument Handbook Version 5.0 51





insert table B.1 Continued
52 Faint Object Spectrograph Instrument Handbook Version 5.0





insert table B.2
 Faint Object Spectrograph Instrument Handbook Version 5.0 53





insert table B.2 continued
54 Faint Object Spectrograph Instrument Handbook Version 5.0


APPENDIX C.


Grating Scatter



M. Rosa
Space Telescope European Coordinating Facility,
Garching bei Munchen, Germany



1. Dispersion and diffraction of light in the FOS

The FOS is a single pass spectrometer with blazed, ruled gratings. Both the blue and the
red side detectors cover wide spectral ranges. Therefore, the FOS is subject to "scattered"
light which has its origin primarily in the diffraction patterns of the gratings and the entrance
apertures, as well as the micro roughness of gratings due to their ruled surfaces. These
limitations are brought about by physical principles.
Additional scattering due to contamination of optical surfaces or unbaffled stray light
worsens the situation. However, the analysis of laboratory and in-flight FOS data shows
that the actual instrument performance is very close to the performance anticipated from
ideal optical surfaces. Therefore, the contamination of observations by scattered light can
be predicted with reasonable accuracy.
For illustration of the above arguments, let the target spectrum be the model atmosphere
appropriate for the Sun (Kurucz 1993), observed in the FOS BLUE G190H mode through
the 0.9 arcsec round aperture. The detector covers a range of 1.47 degrees of the diffracted
angle, corresponding to the wavelength range 1573 A to 2330 A . The 3 panels of Figure C.1
cover the range -10 to +23 degrees in diffracted angles (0 A to 7800 A in first order).
Figure C.1 shows, in logarithmic count rates (offset by +1 in the y direction),
o the "ideal" spectrum as observed by an unphysical instrument that relates wavelengths
one-to-one to diffracted angles;
o the "grating" spectrum as dispersed by the blazed grating; orders visible on the graph
are 0, 1 and 2;
o the "model observations", i.e., the ddispersed spectrum convolved with the additional
scattering imposed by the finite size aperture, the ruled surface of the grating and a
minute amount of dust on the optical surfaces.
The shapes of the zero order peaks in the lower panel of Figure C.1 reflect the actual line
spread function (LSF). The far wings of this LSF carry light from the peak of the original
distribution into domains where the target spectrum, filtered by the total throughput of all
optical elements and the detector efficiency, produces very few intrinsic counts. In addition
this LSF moves photons from the zero order peak into the adjacent parts of the 1st order
seen by the detector_although the zero order peak itself is correctly baffled.
In Figure C.2 are shown the actual observed count rates for the star 16 Cyg B, very
similar to the Sun, overlaid with the "ideal" and the "model" observations from Figure C.1.
For a solar-like target spectrum, the scattered light component ranges between 0.999 and
0.01 of the observed signal in the BLUE G190H mode.


2. Predicting the contamination

UV observations of intrinsically red spectra are subject to severe contamination. A rough
idea of the contamination can be obtained from Table C.1, where the log of the count ratio
 Faint Object Spectrograph Instrument Handbook Version 5.0 55


(Scattered+Intrinsic/Intrinsic) is listed for a variety of target spectra and high dispersion
FOS modes.
As a guide to the wavelength range where scattered light will dominate over the signal
for a given target spectrum, one determines the signal count rate spectrum:


N = F E

where N is the count rate per diode, F is the incident spectrum, and E is the efficiency as
a function of wavelength for the spectral range (FOS mode) of interest, and for the adjacent
modes (wavelength ranges) towards the red. Scattered light will dominate the signal in
regions were N falls off more rapidily than the LSF. The medium range (10-500 diode)
wings of the LSF can be approximated by an inverse square function (diode-diode0)-2 , the
conversion from space into diode space is provided in Table 1.1.1 (FOS Dispersers).
In order to accurately predict the contamination by scattered light for a given grat-
ing/detector/aperture combination, an appropriate estimate of the intrinsic energy distribu-
tion of the target is required; and the properties of all optical components have to be taken
into account in detail. Software to model the resultant count rate spectra will soon be made
available in the IRAF/STSDAS FOS analysis package.


3. Advice to proposers

o UV spectra shortward of 2500 A of very red targets may be obtained almost free of
scattered red light in the low resolution modes of the GHRS.
o Use the BLUE digicon for very red targets (later than K3) for G190H observations to
reduce the amount of far red scattered photons.
o Consult table C.1 to find the wavelength range where the scattered light starts to dom-
inate. Observations of continuum sources shortward of this range are absolutely use-
less, unless the intrinsic spectra flatten off. For example, the coronal emission in ff Ori
(M0Iab) can be traced in a G130H spectrum, but a quantitative assessment is impossible.
o Contaminated data with a ratio of scattered counts over intrinsic counts of up to 5 can
likely be corrected provided the intrinsic target spectrum is known for longer wave-
lengths, and provided the exposure times are chosen such as to give enough S/N for
the weak signal in the total of signal+scattered counts. It is also advisable to obtain a
target spectrum at longer wavelengths, at least with the adjacent FOS range.


Additional References

Ayres,T.R.: 1993, "Scattered Light in the G130H and G190H Modes of the HST FOS" ,
CAL/FOS-0115, STScI

Caldwell, J. and Cunningham, C.C : 1992, Science Verification 1343 Interim Report

Kinney,A.L., Bohlin,R.C.: 1993, "Background due to scattered light", CAL/FOS-0103,
STScI

Rosa, M.R.: 1993, "Scattered light in the FOS: An Assessment using science data" ,
CAL/FOS-0114, STScI = ST-ECF Newsletter No. 20, p. 16

Stroke, G.W.: 1967, "Diffraction Gratings", in Encyclopedia of Physics - Handbuch der
Physik, S.Fl"ugge (ed.), Springer, Berlin, p.427-754
56 Faint Object Spectrograph Instrument Handbook Version 5.0


Table C.1
Logarithmic ratios of count rates (Scattered+Intrinsic)/(Intrinsic)
for unreddened stars. FOS, blue detector


___________________________________________________

A0 V G5 V K 3 III
G130H G190H G270H
___________________________________________________

log[(S+I)/I)] log[(S+I)/I)] log[(S+I)/I])
___________________________________________________

1170 0.98 1600 2.92 2250 1.49
1215 1.73 1700 1.00 2350 1.15
1250 0.18 1800 0.41 2500 0.52
1300 0.02 1900 0.19 2700 0.24
1400 0.01 2000 0.07 3000 0.02
1600 0.00 2300 0.01 3300 0.00
___________________________________________________



Figure C.1: FOS Blue G190H Count rate spectra for a G5 V model atmosphere in the
detector plane. See text. Note that the real detector only covers the wavelength range
marked by a thick bar.
 Faint Object Spectrograph Instrument Handbook Version 5.0 57





Figure C.2: FOS Blue G190H data for the G5 V star 16 Cyg B. The count rate spectrum
due to intrinsic photons and the composite of intrinsic and scattered photons are overlaid.
58 Faint Object Spectrograph Instrument Handbook Version 5.0


Faint Object Spectrograph Instrument Handbook Version 5.0 59


60 Faint Object Spectrograph Instrument Handbook Version 5.0


Faint Object Spectrograph Instrument Handbook Version 5.0 61


62 Faint Object Spectrograph Instrument Handbook Version 5.0


Faint Object Spectrograph Instrument Handbook Version 5.0 63


64 Faint Object Spectrograph Instrument Handbook Version 5.0


Faint Object Spectrograph Instrument Handbook Version 5.0 65


66 Faint Object Spectrograph Instrument Handbook Version 5.0


 Faint Object Spectrograph Instrument Handbook Version 5.0 67


APPENDIX E


Faint Object Spectrograph Instrument Science Reports

April 1, 1994


060 FOS Filter-Grating Wheel Repeatability: Dependence on Motor
Selection , G. Hartig - 5/89
067 In-Flight FOS Wavelength Calibration - Template Spectra G.A.
Kriss, W.P. Blair, and A.F. Davidsen - 2/91
068 FOS Red Detector Plate Scale and Orientation, B. Bhattacharya
and G. Hartig - 11/91
069 FOS Red Detector Flat-field and Sensitivity Degradation, G.
Hartig - 11/91
070 Internal/External Offsets in the FOS Wavelength Calibration
G.A. Kriss, W.P. Blair, and A.F. Davidsen - February 1992
071 An Analysis of FOS Background Dark Noise - E.I. Rosenblatt,
W.A. Baity, E.A. Beaver, R.D. Cohen, V.T. Junkkarinen, J.B.
Linsky, and R. Lyons - 4/92
072 Aperture Calibrations During Science Verification of the FOS
L. Dressel and R. Harms - May 1992 (reworked)
073 Scattered Light Characteristics of the HST FOS
F. Bartko, G.S. Burks, G. A. Kriss, A.F. Davidsen, R.D.
Cohen, V.T. Junkkarinen and R. Lyons - April 1992
074 On - Orbit Discriminator Settings for FOS
R.D. Cohen - February 1992
075 FOS Spectral Flat Field Calibration (Science Verification
Phase Data), S.F. Anderson - February 1992
076 Analysis of FOS On-Orbit Detector Background with Burst
Noise Rejection, E.A. Beaver and R. W. Lyons - April 1992
077 Photometric Calibration of the FOS
J. D. Neill, R. C. Bohlin, and G. Hartig - June 1992
078 FOS Polarimetry Calibration [update of CAL/FOS 055]
R.G. Allen and P.S. Smith - March 1992
079 FOS Operation in the South Atlantic Anomaly
W. A. Baity, E. A. Beaver, J.B. Linsky and R. W. Lyons -
April 1992
080 FOS On-Orbit Background Measurements
R. W. Lyons, J. B. Linsky, E.A. Beaver, W. A. Baity, and
E. I. Rosenblatt - April 1992
081 FOS Onboard Target Acquisition Tests
S. Caganoff, Z. Tsvetanov, and L. Armus - April 1992
082 Lab Test Results of the FOS Detector Performance in a
Variable External Magnetic Field
E. A. Beaver and P. Foster - June 1992
68 Faint Object Spectrograph Instrument Handbook Version 5.0


084 Photometric Calibration of the Faint Object Spectrograph and
Other HST Scientific Instruments - R.C. Bohlin and J.D. Neill
7/92
085 FOS Aperture Throughput Variations with OTA Focus - D. Lindler
and R. Bohlin, 8/92
086 Analysis of Photometric Standards following July 1992 FOS Over-
light Stafing Event, C. J. Taylor and C. D. Keyes, 12/92
087 FOS Blue Dectector Plate Scale and Orientation; A. Koratkar,
5/93
088 FOS Flats From Super Spectra; D. Lindler, R. Bohlin, G. Hartig
and C. Keyes, 3/93
089 Primary author: T. Keyes
090 FOS Flat Field Reference Files: A Quick Reference Guide to the
Appropriate File for a Particular Date and Instrumental Con-
figuration; C. Keyes and C. Taylor
091 A Rough Photometric Calibration for FOS,BLUE,G160L,ORDER0, Keith
Horne and Michael Eracleous, 8/93
092 The Post COSTAR Rotation Matrices for Calculating V2,V3 Offsets
in Mode 2 FOS Target Acquisition; A.P. Koratkar and O. Lupie


093 FOS Inverse Sensitivity Reference Files: A Quick Reference Guide
to the Appropriate File for a Particular Date and Instrumental
Configuration, Cynthia J. Taylor and Charles D. (Tony) Keyes,
6/93
094 FOS Calibration Plan for Cycle 3; Charles D. (Tony) Keyes and
Anuradha Koratkar, 6/93
095 Location of FOS Polarimetry; Anuradha Koratkar and Cynthia J.
Taylor, 6/93
096 Location of FOS Spectra: Cycle 1 and Cycle 2 Results, Anuradha
Koratkar and Cynthia J. Taylor, 8/93
097 Light Loss in FOS as a Function of Pointing Error, R. C. Bohlin,
8/93
098 Correction of the geomagnetically-induced image motion problem
on the Hubble space telescope's faint object spectograph, John
E. Fitch, Dr. George F. Hartig, Dr. Edward A. Beaver and Dr.
Richard G. Hier, 8/93
099 Serendipitous Background Monitoring of the Hubble Space Tele-
scope's Faint Object Spectograph, John E. Fitch and Glenn
Schneider, 8/93
100 Cycle1/Cycle2 Discriminator Settings, Cynthia J. Taylor and
Anne L. Kinney, 2/94
101 authors: Roberto Gilmozzi and Ellyne Kinney
102 FOS Aperture Throughput Variations due to Focus Changes, D.L.
Lindler and R.C. Bohlin, 8/93
103 Background Due to Scattered Light, A.L. Kinney and R.C. Bohlin,
9/93
 Faint Object Spectrograph Instrument Handbook Version 5.0 69


104 Pre-COSTAR FOS Point Spread Functions and Line Spread Functions
from Models, I.N. Evans, 9/93
105 Pre-COSTAR FOS Aperture Throughputs from Models, I.N. Evans,
9/93
106 Pre-COSTAR FOS Aperture Transmissions for Point Sources and
Surface Brightness of Diffuse Sources, R. C. Bohlin, 10/93
107 Pre-COSTAR FOS Aperture Throughputs for Mis-centered Targets
Derived from PSF Models, I.N. Evans, 11/93
108 FOS Calibration Plan for SMOV, A.Koratkar,C.Keyes, A.Kinney,
I.Evans and C. Taylor, 11/93
109 FOS Calibration Plan for Cycle 4, A.Koratkar, A.Kinney, C.Keyes,
I.Evans, and C.Taylor, 11/93
110 Location of FOS Spectra: Cycle 3, Anuradha Koratkar, 12/93
111 Positions Prepeatability of Spectra Obtained with the FOS, Lyons,
R.W., Beaver, E.A., Cohen, R.D., & Junkkarinene, V.T., 2/94
114 Scattered Light in the FOS: An Assessment Using Science Data,
Michael R. Rosa, 11/93
115 Scattered Light in the G130H and G190H Modes of the HST
Faint Object Spectograph, 11/93
116 SMOV Report I: Location of FOS Spectra, Anuradha Koratkar,
Cynthia Taylor, Anne Kinney and Charles (Tony) Keyes
117 SMOV Report II: FOS Coarse Alignment 4907, A.L. Kinney,
A.P. Koratkar, O. Lupie, C.J. Taylor and C.D. Keyes
118 SMOV Report III: FOS Baseline Sensitivity, Charles (Tony)
Keyes, Anne Kinney, Anuradha Koratkar, Cynthia Taylor, 1/94
119 The Faint Object Spectograph Binary Search Target Acquisition
Simulator BS4, I.N. Evans, 2/94
120 FOS Aperture Transmissions for Point Sources, R.C. Bohlin,
2/94


Draft version - not yet released.


Standard Calibration Source Instrument Science Reports



001 Updates to HST Standard Star Fluxes, R. Bohlin, & D. Lindler,
July, 1992.
002 Preliminary Comparison of the HST and White Dwarf Absolute
Flux Scales, R. Bohlin, December, 1993.
70 Faint Object Spectrograph Instrument Handbook Version 5.0


APPENDIX F


Exposure Logsheets

The RPS version of the Exposure Logsheets given in Appendix F can be copied from
anonymous ftp (stsci.edu, or 130.167.1.2). The Logsheets are in the subdirectory pro-
poser/documents/props_library. They are called fos_handbook5_example.
Faint Object Spectrograph Instrument Handbook Version 5.0 71


72 Faint Object Spectrograph Instrument Handbook Version 5.0


Faint Object Spectrograph Instrument Handbook Version 5.0 73


74 Faint Object Spectrograph Instrument Handbook Version 5.0


Faint Object Spectrograph Instrument Handbook Version 5.0 75


76 Faint Object Spectrograph Instrument Handbook Version 5.0


APPENDIX G


Title
Faint Object Spectrograph Instrument Handbook Version 5.0 77


78 Faint Object Spectrograph Instrument Handbook Version 5.0


Faint Object Spectrograph Instrument Handbook Version 5.0 79


80 Faint Object Spectrograph Instrument Handbook Version 5.0


 Faint Object Spectrograph Instrument Handbook Version 5.0 81


APPENDIX H


Changes to the Version 5.0 Instrument Handbook



o Figure 1.2.1. Updated HST + FOS + COSTAR efficiency.
o Figure 1.2.2. Updated aperture throughput.
o Figure 1.2.3. Updated observed counts per second per diode.
o Figure 1.4.1. Addition of duty cycle plot, for RAPID mode.
o Figure 2.1.0. Updated slews performed after FOS target acquisition.
o Figures 3.3.1, 3.3.2, and 3.3.3. Updated flat fields.


o Table 1.2.2.. Updated observed counts per second per diode.
o Table 2.1.1, 2.1.2, and 2.1.3. Updated FOS acquisition sequences.
o Table 2.1.6. Updated FOS exposure times.
o Table 4.1. Updated examples for simulation of FOS spectra.


o Appendix C, Scattered light, by M. Rosa.
o Appendix D, FOS Wavelength Comparison Spectra, by C.D.Keyes.
82 Faint Object Spectrograph Instrument Handbook Version 5.0