Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.stsci.edu/hst/stis/documents/isrs/199823.pdf Дата изменения: Tue Dec 7 22:14:48 2004 Дата индексирования: Sat Dec 22 23:34:40 2007 Кодировка:

STIS Instrument Science Report 98-23

Plate Scales, Anamorphic Magnification & Dispersion: CCD Modes
C. Bowers and S. Baum July 14, 1998

ABSTRACT This report presents values of the plate scales and anamorphic magnification in the STIS CCD spectroscopic modes. The results of the in-flight geometric distortion test are adapted as the most reliable measure of the CCD camera mode magnification, yielding an imaging mode plate scale of 0.05071 0.0007 arcseconds per pixel. Differences with ray trace results amount to an as-built increased magnification of 1.00335. The measured grating dispersions are compared with the calculated values and show very consistent agreement provided the as-built magnification past the mode select mechanism (MSM) is 1.0027 that of the design. The combination of this result and the geometric distortion plate scale imply that the as-built OTA-MSM magnification is 1.00065 that of the design. The measured dispersions at four grating settings are unusually discrepant; the calculated dispersions are probably more reliable and will be incorporated for use in the calstis pipeline. A simple model incorporating only grating anamorphic magnification gives good agreement with direct ray tracing. The values of dispersion direction plate scales, incorporating the plate scale determination from the geometric distortion test, are derived and presented using this model. An example of the use of the model to determine the plate scale for specific wavelengths is presented.

1. Background and Introduction:
Direct measurements of the plate scales ["/pixel] have not been made in the CCD spectroscopic modes. Plate scales can be determined by ray tracing or other modeling of the optical paths using the nominal design values of all components. For the greatest accuracy however, such traces or models need to incorporate the actual, as-built component parameters. During the Servicing Mission Orbital Verification (SMOV) of STIS, proposal 7131 was executed to determine the geometric distortion of the CCD imaging mode. This pro-

1


vided a very accurate plate scale determination for the camera configuration. Since all CCD imaging and spectroscopic modes follow the same path, differing essentially only in the element selected with the mode select mechanism (MSM), the camera mode results can be used to calibrate ray traces of the spectroscopic modes. Because of the numerous spectroscopic modes, settings and field positions it is not possible to ray trace all possible combinations. However the CCD modes have a relatively straightforward geometry which does allow a simple spectroscopic model of each mode to be produced. This allows the calculation of accurate plate scales for any combination of mode, setting and field position as needed to reduce a particular observation. Spectroscopic dispersions have been determined for each nominal setting of each mode from internal observations taken using the Pt-Ne/Ar line lamps (i.e. "wavecals"). Comparing these measured dispersions with predictions from ray traces or the mode models has several useful results. The basic geometry and parameters used in the models can be confirmed. The difference between the design and as-built magnification can be divided into two parts - prior to and following the MSM. Since the UV modes follow the same optical path up to the MSM, knowledge of any magnification difference up to this point can be applied to the plate scale analysis of those modes as well. Finally, comparison between expected and measured dispersion may reveal errors in the measured values.

2. The CCD imaging mode plate scales
1. From analysis of the CCD mode geometric distortion test (Malumuth et al., ISR in preparation) a plate scale for the CCD camera modes was determined of: 0.05071 (+- 0.00007) ["/pixel] valid at field center with negligible differences in the X,Y directions. This agrees with, but is more accurate than, the initial prelaunch estimate of 0.050"/pixel (Landsman, private communication). The equality of the plate scale in both directions provides a limit to any tilt error between the focal plane and CCD detector surface as well. A difference of 0.00007"/pixel is equivalent to a tilt of 0.030 degrees. 2. A ray trace of the camera modes using all design values of optics and the OTA gives a plate scale in both X and Y directions of: 0.05088"/pixel This is 1.00335x the plate scale determined by the SMOV test and differs by about 2.4 times the quoted accuracy of that determination. This result suggests that the overall magnification of the as-built OTA+instrument is 1.00335x the design. Another possibility is that some (or all) of this difference is due to an as-built CCD pixel size deviation from the nominal design value of 21.0 microns.

2


3. Grating dispersion results:
Pre-launch grating dispersions at the measured central wavelength of each setting in all CCD spectroscopic modes were determined by Don Lindler and have been confirmed subsequently on orbit (Lanning & Hulbert ISR in preparation). The grating groove densities are known accurately as well and the angular dispersion can be calculated. The measured linear dispersion for the CCD modes [е/pixel] depends on (a) the ray separation between the detector and the point on the camera mirror (K3) at which rays from the grating strike, (b) the CCD pixel size (c) any off-nominal tilt of the detector. Figure 1 shows a comparison between the ray traced dispersions at field center for the primary settings of mode G430M and the measured dispersions based on slit spectroscopy with one of the on-board arc lamps. The solid diamonds show the ratio of measured to ray traced linear dispersion using the nominal optical design values. The open diamonds show the result of dividing the traced spectral plate scale (е/pixel) by 1.0026; this scale factor produces a single parameter least squares fit between the results of the ray trace and the measured dispersions. The linear dispersion of each CCD spectroscopic mode at each nominal setting has been computed using the dispersion relation derived from the grating equation and the grating parameters. Figure 2 shows the overall comparison of the calculated and measured dispersions of all CCD spectroscopic modes. The distance between the camera mirror and the detector center was adjusted to produce the best fit between the calculations and the measurements, producing a value of s=629.53 mm. The best fit value of this single parameter will include any non-nominal camera mirror to detector separation as well as differences of CCD pixel dimensions, small detector tilts, CCD window index or thickness variations, etc. A few discrepant points were not included in this fit; they lie at the ends of the wavelength range where few Pt lamp lines are available to provide the measured dispersion solution. In these cases, the calculated values are probably more accurate than those determined from the spectra and will be adopted in the calstis pipeline. With this selection of s, all calculated dispersions at the central nominal central wavelengths agree with the measurements to within about 0.1%. The internal consistenc between the calcuy lated and measured dispersions for all the different modes shows that the basic grating geometry is accurately known. Figure 1 also shows the comparison between these calculated dispersions in mode G430M with s=629.53 mm and the ray traced dispersion rescaled by 1.0026. The results using either method are essentially identical. From these two results, the imaging mode plate scale comparison and the dispersion comparisons, we conclude: (1) the overall as-built magnification of the OTA+STIS CCD path is 1.00335x the design, and (2) the magnification from the STIS MSM to detector in the CCD path is 1.0027x the design. We note that any out-of-nominal pixel size or detector tilts are included in both of these results. Combining these we conclude that the OTA+STIS magnification up to the MSM is 1.00335/1.0027 = 1.00065 compared to the

3


ray trace (nominal design). This value represents the combined magnification difference of the OTA+STIS corrector + STIS collimator and can be applied as a correction to the UV modes in similar plate scale analysis of those modes; it's very small value, however verifies that up to the MSM, there is essentially no magnification difference between the asbuilt instrument and the current optical ray trace.

4. Plate Scale Determination
With this correction to overall magnification of the CCD modes, we can determine accurate values of the plate scales. In the cross dispersion direction of each spectroscopic mode, the spatial plate scale should match the value obtained in the camera mode determination, namely, Plate Scale, Cross Dispersion Direction, All CCD Modes: 0.05071 (+- 0.00007) ["/pixel] There should be a slight variation of cross dispersion plate scale along the length of the slit as shown by the camera mode geometric distorition test. From the results of that test, we derive an expression for the cross dispersion plate scale along the slit length of d / dY = 0.05071 ­ 1.59122 x 10
­8

(Y ­ 512) ­ 4.78704 x 10

­10

(Y ­ 512)2 [" /pixel]

where Y is the pixel position along the slit. At the ends of the slit (Y = 1,1024 pixels) the plate scale is about 0.05058 "/pixel, about 0.26% different than at the slit center. The plate scale in the dispersion direction differs from this value primarily due to (1) anamorphic magnification in each mode due to the diffraction gratings, (2) differences in the point where the principal ray at each wavelength hits the camera mirror and the detector for each different wavelength and field position. Anamorphic magnification of diffraction gratings is due to the typical inequality between the incident and exit directions of rays of a particular wavelength. If the grating incident angle (G) and exit angle (G) in the dispersion plane are known, the anamorphic magnification for a particular wavelength can be expressed as the differential quantity Grating anamorphic magnification: (1) M
ANA

= d G / dG = cos(G)/cos (G)

For all STIS grating modes, G >G so, for example, a feature 1 arcsecond in angular extent on the sky is compressed at the detector. The plate scale [in arcseconds per pixel] in the dispersion direction will thus increase in all spectroscopic modes compared to the camera mode. Accounting only for anamorphic magnification, the conversion between imaging mode plate scale and the plate scale in the dispersion direction is given by Plate Scale, Dispersion Direction, All CCD Modes:

4


(2)

0.05071/MANA ["/pixel]

The values of the dispersion plane grating incident and exit angles have been determined for each setting allowing the anamorphic magnification and plate scale to be determined in this manner. The necessary parameters of each CCD mode grating are listed in Table 1. The grating incident angles were determined from the modified grating equation (3) m/(cos()) = 2cos()sin(G-) 2 = the dispersion angle = 20.678° for all CCD modes = grating groove spacing = central wavelength at each setting = field angle (at the grating); =0 for nominal field center m = grating order (=1 for all CCD gratings) The values of the grating dispersion plane exit angle, , may then be obtained from the normal grating equation (4) sin() + sin() = m/(cos())

The results of these calculations are presented in Tables 2-4 for all CCD modes, giving the values of calculated anamorphic magnification, MANA at field center for each nominal grating setting from equation (1) and the dispersion direction plate scales using equation (2). Each table also gives the anamorphic magnification and plate scales at the ends of the wavelength range for each setting. The shortest wavelength in a setting is indicated by L, the longest by H. The values of dispersion, plotted in Figure 2 were also calculated using these grating parameters for each setting, and showed very good consistency. The results of these calculations may be summarized as follows. Each of the low resolution modes has an anamorphic magnification of less than about 2% and consequent similar difference of plate scale in the dispersion direction compared to the imaging case. The medium resolution modes have higher grating incident angles and so higher anamorphic magnification and more substantial differences in plate scale with respect to the imaging modes. The variation in plate scale in all medium resolution modes is about 712% for the run of grating settings compared to the imaging value. Within a particular setting the variation is typically only a few tenths percent over the wavelength range. The variations of plate scale with field angle in the dispersion direction result in negligible differences from the values quoted in Tables 2-4, even at the edges of the field. As viewed from the grating, the field angle f is slightly less than 1° at the edge of the CCD field. For a particular wavelength, this effect enters as a slight change in exit angle, , through the grating equation (4) compared to field center ( = 0). Exit angle variation from

5


field center to edge are typically < 0.005°. Since the plate scale depends on through cos(), such a small change in exit angle produces a negligible change in plate scale. To determine the accuracy of the models and the magnitude of any other geometrical factors, these results may be compared to the results of ray tracing. Table 5 shows such a comparison for mode G430M. The ray traced results have been rescaled by 1.0027 to produce correction for the as-built parameters. The agreement between the ray trace and model results is <0.6% in all cases; one micron differences in the ray trace results, give 0.24% errors typically. The conclusion is that the magnitude of any other known geometric factors produces differences with the simple model of no greater than 0.6%. Finally, these models can be used to predict the plate scales for specific lines. As an example, we determine the dispersion direction plate scales in mode G430M at 5007е cen=4961 and G750M at 6563е for cen=6748. These values would be needed to compare spatial extent of features in [OIII] and H for example. The procedure is as follows: 5007 (a) Determine the grating exit angle for this wavelength. From Table 1, the grating entrance angle for G430M, cen=6768 is 27.952°, groove density is 0.83333 microns. Use the normal grating equation, Equation 4, to determine the exit angle to be 7.592°. (b) Determine the anamorphic magnification from Equation 2 to be 0.8912. (c) The plate scale is then 0.05071/0.8912 = 0.05690 arcseconds/pixel. 6563 (a) Determine the grating exit angle for this wavelength. From Table 1, the grating entrance angle for G750M, cen=6768 is 22.250°, groove density is 1.6667 microns. Use the normal grating equation, Equation 4, to determine the exit angle to be 0.8666°. (b) Determine the anamorphic magnification from Equation 2 to be 0.9256. (c) The plate scale is then 0.05071/0.9256 = 0.05478 /pixel These plate scales differ by about 4%. For an extended object like the ~2" wide inner rings of SN1987a, the 5007 е image would be 35.1 pixels wide while the 6563 е image would be 36.5 pixels wide. The plate scale in the cross dispersion direction is just the nominal value of 0.05071"/pixel. A 2" object would cover 39.4 pixels in the cross dispersion direction.

6


5. Slitless Modes:
All of the calculations thus far have implicitly been assumed to operate in a slit mode. That is the input angle to the gratings, , has a fixed value for any observation. This value is that subtended by the slit center and the particular grating normal. The list of incident angles in Table 1 are just such angles at slit center. As an object becomes more extended in the dispersion direction, this incident angle begins to change from nominal. The anamorphic magnification, depending on the cosine of incident and exit angles, changes slowly for all modes since all nominal angles are small. At the maximum, open slit configuration, can be about 1є greater than the nominal values shown in Table 1. At the high wavelength settings, with incident angles of about 28 є the anamorphic magnification and plate scale could change by about 1% from the values presented.

7


Table 1. Grating Model Parameters

G430L

G430M

G750L

G750M

G230LB

G230MB

m= [µm] = 2=

1 8.33333 20.678

1 0.83333 20.678

1 14.80166 20.678

1 1.66667 20.678

1 4.16667 20.678

1 0.45455 20.678

G750L Internal Setting P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 cen 4300

G430M cen 3165 3423 3680 3936 4194 4451 4706 4961 5216 5471

G750L cen 7751 8975

G750M cen 5734 6252 6768 7283 7795 8311 8825 9336 9851

G23LB cen 2375

G230MB cen 1713 1854 1995 2135 2276 2416 2557 2697 2836 2976 3115

G
11.844

G
21.466 22.387 23.307 24.230 25.162 26.092 26.998 27.952 28.893 29.828

G
11.867 12.105

G
20.412 21.329 22.250 23.165 24.085 25.026 25.947 26.877 27.813

G
11.999

G
21.380 22.305 23.233 24.155 25.080 26.017 26.954 27.893 28.830 29.772 30.681

I1 I2 I3 I4 I5 L1

3305 3843 4781 5093

21.967 23.897 27.293 28.448

6094 6581 8561 9286 9806 10363

21.056 21.917 25.469 26.789 27.737 28.761

2794

28.547

Notes to Table 1:

cen

is the officially used wavelength use to designate the tilt of the

grating at this setting (see the STIS Instrument Handbook, Chapter 13). Internal Setting gives the internally used (for commanding purposes) position name designated for that positioning of the grating, and is relevant for internal STScI use only.

8


Table 2. G430L and G430M Plate Scales and Anamorphic Magnification Mode G430L - Plate Scales & Anamorphic Magnification
Internal Setting P1 cen (desig) 4300 C (msd) 4307.54 MANA 0.99046 Sdisp ["/pix] 0.05120 L (msd) 2903.10 MANA 0.99324 Sdisp ["/pix] 0.05106 H (msd) 5712.00 MANA 0.98798 Sdisp ["/pix] 0.05133

Mode G430M - Plate Scales & Anamorphic Magnification
Internal Setting P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 cen (desig) 3165 3423 3680 3936 4194 4451 4706 4961 5216 5471 C (msd) 3164.12 3422.27 3679.37 3936.36 4194.57 4451.35 4700.35 4961.28 5217.04 5470.07 MANA 0.93072 0.92505 0.91937 0.91366 0.90789 0.90211 0.89647 0.89051 0.88460 0.87870 Sdisp ["/pix] 0.05448 0.05482 0.05516 0.05550 0.05585 0.05621 0.05657 0.05695 0.05733 0.05771 L (msd) 3021.90 3280.00 3537.20 3794.30 4052.20 4309.40 4558.70 4819.80 5075.80 5329.30 MANA 0.93064 0.92471 0.91878 0.91282 0.90681 0.90079 0.89491 0.88870 0.88255 0.87642 Sdisp ["/pix] 0.05449 0.05484 0.05519 0.05555 0.05592 0.05630 0.05667 0.05706 0.05746 0.05786 H (msd) 3306.10 3564.20 3821.10 4078.20 4335.90 4592.60 4841.30 5102.00 5357.4 5610.10 MANA 0.93108 0.92565 0.92022 0.91476 0.90924 0.90370 0.89829 0.89258 0.88691 0.88125 Sdisp ["/pix] 0.05446 0.05478 0.05511 0.05544 0.05577 0.05611 0.05645 0.05681 0.05718 0.05754

I1 I2 I3 I4

3305 3843 4781 5093

3304.72 3843.80 4781.20 5096.27

0.92763 0.91572 0.89463 0.88740

0.05467 0.05538 0.05668 0.05714

3162.40 3701.60 4639.70 4954.80

0.92741 0.91497 0.89299 0.88546

0.05468 0.05542 0.05679 0.05727

3446.60 3985.50 4922.10 5236.8

0.92812 0.91673 0.89653 0.88960

0.05464 0.05532 0.05656 0.05700

Notes to Table 2: This table lists the plate scale in the dispersion direction and anamorphic magnification of each nominal setting for modes G430L and G430M based on results using the geometric model of this mode. cen is the officially used wavelength use to designate the tilt of the grating (see the STIS Instrument Handbook, Chapter 13).. The columns are: (1) Internal Setting, which gives the internally used (for commanding purposes) position name designated for that setting of the grating tilt, and is relevant for internal STScI use only, (2) the officially designated central wavelength (cen - desig) used to specify the setting (see Chapter 13 of the STIS Instrument Handbook), (3) the actual wavelength at the central pixel (C-msd) measured as determined from the on-board Pt lamp spectrum, (4) the anamorphic magnification at the central wavelength, (5) the plate

9


scale in the dispersion direction at the central wavelength, (6) the determined lowest wavelength (L) at this setting (see Chapter 13 of the STIS Instrument Handbook), (7) the anamorphic magnification at the lowest wavelength, (8) the plate scale in the dispersion direction at the lowest wavelength, (9) the determined highest wavelength (H) at the setting, (10) the anamorphic magnification at the highest wavelength, (L) the plate scale in the dispersion direction at the highest wavelength. All wavelengths are in Angstroms, all plate scales are in arcseconds per pixel. Table 3. G230LB G230MB Plate Scales and Anamorphic Magnification Mode G230LB - Plate Scales & Anamorphic Magnification
Internal Setting P1 cen (desig.) 2375 C (msd) 2374.83 MANA 0.98948 Sdisp ["/pix] 0.05125 L (msd) 1672.10 Sdisp ["/pix] 0.05111 H (msd) 3077.2 MANA 0.98706 Sdisp ["/pix] 0.05137

MANA 0.99221

Mode G230MB - Plate Scales & Anamorphic Magnification
Internal Setting P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 cen (desig.) 1713 1854 1995 2135 2276 2416 2557 2697 2836 2976 3115 C (msd) 1712.76 1854.32 1995.72 2135.69 2275.60 2416.77 2557.33 2697.42 2836.43 2975.54 3108.89 MANA 0.93125 0.92555 0.91982 0.91413 0.90840 0.90258 0.89674 0.89087 0.88500 0.87906 0.87330 Sdisp ["/pix] 0.05445 0.05479 0.05513 0.05547 0.05582 0.05618 0.05655 0.05692 0.05730 0.05769 0.05807 L (msd) 1635.30 1776.70 1918.10 2058.20 2198.20 2339.40 2480.10 2620.30 2759.50 2898.70 3032.30 MANA 0.93119 0.92523 0.91926 0.91331 0.90734 0.90127 0.89519 0.88908 0.88297 0.87678 0.87080 Sdisp ["/pix] 0.05446 0.05481 0.05516 0.05552 0.05589 0.05626 0.05665 0.05704 0.05743 0.05784 0.05823 H (msd) 1790.20 1931.70 2073.00 2213.00 2352.80 2493.80 2634.20 2774.10 2913.0 3051.90 3185.10 MANA 0.93158 0.92613 0.92065 0.91521 0.90972 0.90415 0.89856 0.89293 0.88729 0.88159 0.87606 Sdisp ["/pix] 0.05443 0.05475 0.05508 0.05541 0.05574 0.05609 0.05644 0.05679 0.05715 0.05752 0.05788

I1

2794

2794.52

0.88678

0.05718

2717.50

0.88482

0.05731

2871.30

0.88900

0.05704

Notes to Table 3: This table lists the plate scale in the dispersion direction and anamorphic magnification of each nominal setting for modes G230LB and G230MB based on results using the geometric model of this mode. The columns are the same as for Table 2.

10


Table 4. G750LB G750M Plate Scales and Anamorphic Magnification Mode G750L- Plate Scales & Anamorphic Magnification

Internal Setting P1 P2

en (desig.) 7751 8975

C (msd) 7765.73 8977.25

MANA 0.99031 0.98881

Sdisp ["/pix] 0.05121 0.05128

L (msd) 5269.30 6479.40

MANA 0.99309 0.99151

Sdisp ["/pix] 0.05106 0.05114

H (msd) 10262.6 11475.1

MANA 0.98785 0.98642

Sdisp ["/pix] 0.05133 0.05141

Mode G750M - Plate Scales & Anamorphic Magnification
Internal Setting P1 P2 P3 P4 P5 P6 P7 P8 P9 cen (desig.) 5734 6252 6768 7283 7795 8311 8825 9336 9851 C (msd) 5735.24 6251.62 6768.24 7279.82 7792.22 8314.28 8822.85 9334.41 9846.38 MANA 0.93722 0.93157 0.92589 0.92024 0.91456 0.90873 0.90302 0.89722 0.89138 Sdisp ["/pix] 0.05411 0.05444 0.05477 0.05511 0.05545 0.05580 0.05616 0.05652 0.05689 L (msd) 5450.60 5967.00 6483.70 6995.20 7507.80 8030.00 8539.00 9050.70 9563.40 MANA 0.93743 0.93152 0.92559 0.91969 0.91376 0.90768 0.90173 0.89569 0.88961 Sdisp ["/pix] 0.05409 0.05444 0.05479 0.05514 0.05550 0.05587 0.05624 0.05662 0.05700 H (msd) 6019.20 6535.70 7052.20 7563.40 8075.70 8597.20 9105.60 9616.40 10129.4 MANA 0.93728 0.93188 0.92645 0.92106 0.91562 0.91004 0.90457 0.89902 0.89343 Sdisp ["/pix] 0.05410 0.05442 0.05474 0.05506 0.05538 0.05572 0.05606 0.05641 0.05676

I1 I2 I3 I4 I5 L1

6094 6581 8561 9286 9806 10363

6097.85 6581.44 8558.81 9285.94 9805.13 10362.7

0.93325 0.92794 0.90599 0.89778 0.89185 0.88543

0.05434 0.05465 0.05597 0.05648 0.05686 0.05727

5813.00 6296.90 8274.70 9002.60 9522.10 10082.1

0.93328 0.92774 0.90482 0.89627 0.89010 0.88343

0.05434 0.05466 0.05604 0.05658 0.05697 0.05740

6381.80 6865.80 8841.60 9568.10 10088.0 10643.3

0.93349 0.92842 0.90741 0.89955 0.89388 0.88771

0.05432 0.05462 0.05588 0.05637 0.05673 0.05712

Notes to Table 4: This table lists the plate scale in the dispersion direction and anamorphic magnification of each nominal setting for modes G750L and G750M based on results using the geometric model of this mode. The columns are the same as in Table 2.

11


Table 5. Comparison Between Ray Traced and Model Results - G430M
Ray Trace Results rescale Internal Setting P1 cen. (design.) 3165 c mes 3164.12 Model Results divided by Ray Trace Results Sdisp 0.99872 1.00129 0.99890 1.00147 1.00052 1.00052 1.00577 1.00577 SX-disp 0.99921 1.00178 1.00158 1.00416 0.99921 0.99941 0.99921 0.99941

Model Results

Field ctr edge

Sdisp 0.05455 0.05441 0.05455 0.05441 0.05768 0.05768 0.05721 0.05721

SX-disp 0.05075 0.05062 0.05063 0.05050 0.05075 0.05074 0.05075 0.05074

Sdisp 0.05448 0.05448 0.05449 0.05449 0.05771 0.05771 0.05754 0.05754

SX-disp 0.05071 0.05071 0.05071 0.05071 0.05071 0.05071 0.05071 0.05071

3021.90

ctr edge

P10

5471

5470.07

ctr edge

5610.10

ctr edge

Notes to Table 5: This table compares plate scales determined by ray tracing with those obtained using the model of grating anamorphic magnification only. Values are given for two settings of mode G430M c=3165, 5471 at central and extreme wavelengths in the settings, and for field positions at the slit center and end. All differences between ray traced and modeled results are <0.6%. First three columns as are in Table 2.

Figure 1: A comparison of the ratio of the measured to ray traced dispersions of mode G430M. The dispersion at the central wavelength of each setting is shown. The ratio using the nominal optical design is shown by solid diamonds. A single scaling parameter of 1.0028 produces the results indicated by the open diamonds. The open circles show the similar ratio comparison using the grating model for this mode and the best fit distance between the camera mirror and the detector center. Figure 2: A comparison of the measured and calculated dispersions for all CCD modes at all nominal positions. The slit-detector magnification was varied to produce the best fit by altering the camera mirror (K3) focal length. The dashed lines indicate a difference in the measured/calculated ratio of 0.1%. The three longest wavelength points of mode G750M and the shortest point of G230MB appear significantly discrepant with the rest of the data and are probably in error due to the lack of Pt lines from the calibration lamp at these settings.

12


Mode G430M: Comparison of Ray Traced, Measured and Calculated Dispersions
1.010

1.005

1.000

0.995

0.990 3000

3500

4000

4500

5000

5

13


CCD Modes: Msd/Calc Dispersion Ratios
1.01

1.005

G 750M G 430M 1 G 430L G 750L G 230M B Camera fl changed from nominal 625 mm to 629.53 mm. G 230LB

0.995

0.99 1600 2600 3600 4600 5600 6600 7600 8600 9600 10600

Wavelength [A]

14