Optical Layout

Click here for a picture of the optical layout

ITL/STA CCD Operating Parameters

Blue Channel Gratings

Grating (grooves/mm)	Blaze (order; wavelength)	Resolution with 1" slit (Å)	Coverage (Å)	Blaze Angle (deg)	Dispersion (Å/pix)
300	1st/4800 Å	6.2	4825	4.3	1.96
500	1st/5410 Å	3.6	3200	8.05	1.19
800	1st/4050 Å	2.2	2000	9.67	0.75
600	2nd/4800 Å	1.4	1340	17.4	0.50
	1st/9630 Å	2.8	2680	17.4	1.00
832	2nd/3900 Å	1.0	960	19.7	0.36
	1st/7790 Å	1.9	1930	19.7	0.72
1200	1st/4830 Å	1.4	1340	17.5	0.50
Echellette	12th/4480 Å	0.4	360	42.0	0.19

Notes:

Spectral coverage is quoted for full 2688 columns. Images and throughput degrade somewhat in first and last 200 pixels.
Echellette is used in orders 6-17.

Theoretical Grating Blaze Functions

Click here for a Postscript plot of the blaze.

Cross-Dispersed Operation

The Blue Channel Echellette grating is a 6"x10", custom-ruled, 240 g/mm grating which can be used in orders 7-17 simultaneously, giving coverage from 3100-8150 Å with a 10 arcsecond long slit. Cross dispersion is provided by a large quartz prism which is inserted into the beam between the grating and the camera.

In a perfect spectrograph, the spectral resolution of this configuration is about 10,000 with a 1 arcsec wide slit (i.e. 30 km/sec). In practice, the focus degrades at the ends of the orders. During initial tests, we picked a focus value that gave a decent compromise focus over much of the frame. The resulting resolution varied from about 30 to 60 km/sec.

Sample Images

Several images which summarize the salient operating characteristics of this configuration are available as gifs (i.e. images that can be displayed by your browser). Click below to see them:

Sample Data

Some sample data is available in FITS format for interested prospective users. Five files can be retrieved by anonymous ftp:

cuhenear -- a wavelength-calibrated"one-dimensional" (actually 3072x11) spectrum of the comparison lamp. Originally stored in IRAF 'echelle' format.
HZ44 -- a 3072x11 wavelength-calibrated spectrum of the 11 usable orders of the spectrum of the standard star HZ 44. Spectrum is in photons/sec/A. Originally stored in IRAF 'echelle' format.
M67_0847p1202 -- an 1800 s exposure of a giant in M67 taken through clouds and in dire seeing. Spectrum is in photons/pixel and is wavelength calibrated. Originally stored in IRAF 'echelle' format. Since this star is so red, only orders 7-16 were extracted. Wavelength scale is approximate.
M67_0847p1202 -- an 1800 s exposure of another giant in M67. Originally stored in IRAF 'echelle' format.

To retrieve the files, ftp to sculptor.as.arizona.edu; login as anonymous, giving your email address as the password; cd pub/bc_ech_tests; binary; get the files. For easiest access to the data, after transfering the files use IRAF to read them into `echelle' format using rfits, then splot them. The different orders are in different image lines in the case of the one-dimensional data.

Throughput

Throughput measurements are not available for the cross-dispersed configuration since the initial engineering run was partly-to-mostly cloudy but a few impressions were garnered. On a photons/Å/second basis, this configuration is clearly not competitive with the other gratings -- we estimate that it may be a factor of 2 or more slower when compared to the other gratings at their blazed wavelengths. Some of this is attributable to losses from the prism; the rest (and probably the lion's share) is due to the grating. However, some of this factor is compensated for by the fact that one is atop the blaze in each of the echellette's orders as opposed to the normal gratings where the blaze can often compromise performance.

Our first impression of this grating is that it could be very useful for projects requiring moderately-high resolution over a wide wavelength range. The data are relatively easy to reduce (up to the point where adjacent orders are combined) using the `echelle' package in IRAF.

More observations are scheduled for April 1995, so we will know more about the system's throughput at that time.

Scheduling Constraints

Configuration of the spectrograph for cross-dispersed work requires that the spectrograph be taken off the telescope. In practice, this means that the spectrograph is scheduled for cross-dispersed `runs.' It is possible to leave the prism in and use other gratings, but there is significant throughput loss and the spectra are quite curved. It is also possible to mount the Red Channel on the Blue Channel and use it for backup programs in case of inclement conditions.

Prospective users of the Echellette are urged to contact Mark Wagner (520-626-3006).

If you are submitting a proposal to use this configuration, please specify very clearly that this is the case in order to alert the schedulers of the telescope.

Aperture Plates

0.75 arcsec slit, 180 arcsec length

1.00 arcsec slit, 180 arcsec and 20 arcsec lengths

1.25 arcsec slit, 180 arcsec and 20 arcsec lengths

1.50 arcsec slit, 180 arcsec and 20 arcsec lengths

2.00 arcsec slit, 180 arcsec and 20 arcsec lengths

3.50 arcsec slit, 180 arcsec and 20 arcsec lengths

3.75 arcsec slit, 180 arcsec and 20 arcsec lengths

5.00 arcsec slit, 180 arcsec and 20 arcsec lengths

"Comb" -- line of 1 arcsec circular holes on 10 arcsec centers

3 arcsec diameter single circular hole

1.0, 1.4, and 5.0 arcsec diameter double circular holes

1.0x2.6 and 2.0x3.0 arcsec double slits

1.25x90 arcsec slit for use with cross-dispersing prism

Slit Length

The finite size of the filter wheels vignettes light near the ends of the long slits. The unvignetted slit is about 150 arcsec long, as shown in this plot of the intensity along the slit.

In this example, the CCD was binned by a factor of two in the spatial direction, giving a spatial image scale of 0.6 arcsec/pixel, twice that of the nominal unbinned scale of 0.3 arcsec/pixel.

Variation of Line Width with Slit Width

The following table gives the FWHM of a Gaussian fit to an unresolved comparison lamp emission line as a function of the width of the entrance slit. A low-resolution grating was used, so the anamorphic factor is near unity. The CCD was not binned on readout.

Click on the individual table entries to see the line profiles.

Note that the comparison lines are not well approximated by a Gaussian for wider slits so the quoted FWHM are not particularly accurate.

Slit Width (arcsec)	FWHM (pixels)
1.0	3.1
1.25	3.6
1.5	4.5
2.0	5.3
3.5	7.9
5.0	11.5

Order-Sorting Filters

Blue-Blocking Filters:

The following filters are available to block light from higher orders. Those marked with asterisks are normally mounted in the spectrograph. Click here for transmission curves and here for a Postscript plot.

UV36*	L-38*	L-42*
LP-495* (Y-50)	LP-530	R-63

Red-Blocking Filters:

The following filters are available to block light from lower orders. Those marked with asterisks are normally mounted in the spectrograph. Click here for transmission curves and here for a Postscript plot.

CuSO4*	U330
C-500

Echellette Order-Selecting Filters:

Postscript plots of the echellette order-sorting filters are available below. Note that these are not usually resident in the spectrograph, so prospective users must alert the Instrument Specialist well in advance of their run.

8th Order Filter	9th Order Filter
10th Order Filter	11th Order Filter
12th Order Filter	13th Order Filter
14th Order Filter	15th Order Filter

Hoya Filters: The following filters standard Hoya Glass 2"x2" filters that are kept on hand at the MMT. Consult a Hoya catalog for transmission curves. Note that these are not usually resident in the spectrograph, so prospective users must alert the Instrument Specialist well in advance of their run.

ND 0.3	ND 1.3	ND 2.5
ND 4.0	ND 5.0	ND 7.0
U 360	UV 28	O 54
O 56	B 390	R 60
R 68	IR 85	LB 120
LB 145	LB 165	LB 200
LA 120	RM 90	Y 48
V 10	G 533

Spectrograph Throughput

The data referred to below are spectra of standard stars observed through a 5 arcsec slit in photometric, good-seeing conditions. We do not have data for all gratings at all possible wavelength settings.

No corrections have been made to the data except to divide them by the exposure time, multiply by the CCD gain and divide by the average pixel width (in Angstroms). Therefore, the spectra show measured number of photons per second per Angstrom. No corrections for extinction, telescope reflectance, slit losses, etc. have been applied. We list the mean pixel size and the airmass and plot the AB magnitude versus wavelength in the lower panel.

In the cases of high resolution gratings, data from several grating settings are presented on a single plot.

Note: The plots below were constructed with data taken on the 4.5-m MMT. Given that the 6.5-m telescope has two fewer reflections and more than twice the light gathering power of the 4.5-m telescope, the numbers in the plot should be multiplied by a factor of 2.5 or the effective AB magnitude should be increased by one magnitude when estimating exposure times for the new telescope.

Click on the appropriate configuration for a Postscript plot.

The Blue Channel of the MMT Spectrograph

Contents