Previous Article
Next Article
Table of Contents

Commissioning of the Loral 1200 x 800 CCD on the... (1Mar96)
1.5-m Cass Spectrograph
(from CTIO, NOAO Newsletter No. 45, March 1996)

Final commissioning of the Loral 1200 x 800 CCD (hereinafter Loral 1K
CCD) plus 1.5-m spectrograph combination was carried out during two
engineering nights on 15-16 November 1995. A previous engineering run
in May (reported in NOAO Newsletter No. 43, p.25) had allowed an
initial evaluation of the image quality, throughput, fringing
characteristics and flexure of the system at the telescope. The
encouraging results of that engineering night were amply confirmed on
the two commissioning nights in November--the performance of the Loral
1K CCD is superior in all aspects to that of the front-illuminated GEC
CCD used for so many years with the 1.5-m spectrograph, offering 40%
greater wavelength coverage (at the same resolution) and 2-3 times the
quantum efficiency.

Within the next few months we will attempt to have a new user manual
for the 1.5-m spectrograph + Loral 1K CCD available on the CTIO WWW
pages. Twelve gratings are available, which give the following
wavelength resolution and coverage with the Loral 1K:

				FWHM		   Wavelength
Gtg     l/mm    blaze(a) 	Resolution         Coverage
				(Angstrom)	   (Angstrom)

13       150     5000   	17.2            	6900
11(b)    158     8000 		16.4			6550
09       300     4000   	 8.6             	3450
32       300     6750   	 8.6             	3450
22(b)    300    10000		 8.6			3450
58       400     8000		 6.5			2590
16       527     5500		 4.8			1965
26       600     4000		 4.3			1725
35       600     6750		 4.3			1725
56       600    11000		 4.3			1725
47       831     8000		 3.1			1245
36(c)   1200     7500		 2.2			 860

(a)  Blaze is first order Littrow blaze. Effective blaze wavelength
when used in the 1.5-m spectrograph is 0.89 of the Littrow value.

(b)  Silver coated does not reflect light below ~4000 Angstroms.

(c)  Cannot be tilted far enough to be used in II order.

The Loral CCD has 15 um pixels and a slit width of 143 um (2.6")
projects to 2 pixels. In practice, however, the focus of arc lines is
observed to range between 2-3 pixels (FWHM) due to the camera optics
and charge diffusion within the CCD (for more details, see the article
in NOAO Newsletter No. 43). The wavelength resolution figures in the
above table are calculated for a FWHM of 3 pixels.

Efficiency measurements of the total telescope/spectrograph/CCD
combination were obtained with gratings 09 and 32 during the November
engineering nights. The results are given in the following table in
terms of the percentage of photons striking the telescope primary
mirror which are eventually detected by the CCD. Note that these
numbers are still not definitive due to uncertainties in the precise
value of the CCD gain on the engineering night; however they should be
accurate in a relative sense. We hope to be able to provide final
numbers within the next six months.

Grating 09      		Grating 32
Lambda QE (%)        		Lambda (QE) %

3500     7.5     		6500    14.8
4000    13.5    		7000    13.6
4500	 6.5     		7500    11.8
5000    17.0    		8000     8.5
5500    16.0			8500     6.5
6000    13.5    		9000     5.0
6500    10.0    		9500     2.9

The upper right amplifier does not perform satisfactorily, therefore
the CDD is read in single-channel mode (through the lower left
amplifier). As part of the commissioning process, a reduction of 10 ms
per pixel was achieved. The resulting gains, read noises, and readout
times are:

1/Gain  	Read Noise      Read Time
(e-/ADU)        (e-)            (seconds)

4.11            7.71            15.6
2.87		7.11            19.7
2.05            6.50            25.8
1.42            6.13            34.0
0.96		5.88            46.3

Full well of the Loral 1K CCD is 118,000 e-. Over this range, the CCD
delivers excellent linearity (gain variation = 0.26% peak-to-peak).

As reported in NOAO Newsletter No. 43, the Loral 1K CCD fringes with
substantial amplitude at wavelengths redward of 7500 Angstroms.
However, thanks to the lack of significant flexure in the 1.5-m
spectrograph and camera, it is possible to remove nearly all of
the fringing using normal dome flats.

The accompanying figure shows spectra of two type II supernovae
obtained with the 1.5-m + Loral 1K combination. These spectra are the
sum of separate blue and red observations obtained at ~8.5 resolution
with gratings 09 and 32. Total integration time for SN 1995ad was 30
minutes in the blue and 30 minutes in the red, while SN 1995v was
observed for 90 minutes in the blue and 45 minutes in the red.
Approximate magnitudes measured from the spectra are B = 16.8 and V =
15.7 for SN 1995ad and B = 18.7 and V = 17.6 for SN 1995v. Note the
lack of obvious residual fringing at red wavelengths. In the case of SN
1995ad, a signal-to-noise value of 35:1 was obtained at a wavelength of
9250 Angstroms where the CCD fringing amplitude reaches 20%. (The
signal-to-noise in the these spectra is limited more by photon
statistics than any residual fringing.)

		        [Figure not included]

With the successful commissioning of the Loral 1K CCD which is
controlled by an Arcon, the last VEB controller in service on Tololo
has now been officially retired! Que descanse en paz.

Mark Phillips, Steve Heathcote, Roger Smith

Previous Article
Next Article
Table of Contents