Observations

John Baldwin, Graham Cox, Craig Mackay and the author undertook three observing runs together using L3Vision CCDs at the NOT (in the summers of 2001, 2002 and 2003). A CCD65 detector (with $624\times528$ pixels, each $20\times30$ $\mu m$) was used for the first two runs, and a CCD87 (with $512\times512$ pixels, each $16\times16$ $\mu m$) was used for the 2003 run. Both devices were front-illuminated frame-transfer CCDs. A brief log of these three observing runs is shown in Table 5.1. For each run a single lens achromat was used to convert the f/11 beam at the Cassegrain focus to a suitable image scale for the CCD detector. A diverging lens achromat was used before the focus for the first two runs, while a converging lens was used after the focus for the third run, providing a re-imaged aperture plane within the instrument. Circular aperture stops could be placed in this re-imaged aperture plane if required.

Table 5.1: Log of observing runs at the NOT involving L3Vision CCDs. For many of the individual observations only a small region of the CCD was read out.

		No. of	f/ratio	Pixel scale	No. useful
Start date	CCD	pixels	used	$mas$	nights
2001/07/05	CCD65	$624\times288$	$60$	$27\times40$	$1.5$
2002/07/25	CCD65	$624\times288$	$64$	$25\times37$	$1.0$
2003/06/27	CCD87	$512\times512$	$32$	$40\times 40$	$7.0$

For each of the runs the camera was shipped from Cambridge along with a supporting frame for fitting at the Cassegrain focus of the NOT. The optics and mechanics of the instrument were designed by John Baldwin, Craig Mackay and Donald Wilson. The frame used for the run in 2003 can be seen during assembly in Figure 5.1. The metal plate being attached on the left-hand side of the instrument is the mechanical interface to the telescope, while the camera dewar can be seen on the right-hand side of the frame.

Figure 5.1: Assembling the ``Luckycam'' instrument at the NOT in 2003. During observations, starlight enters through the central hole in the circular plate being attached to the left-hand side.

In order to measure the isoplanatic angle it was necessary to undertake simultaneous imaging observations of widely separated stars. The limited dimensions of the L3Vision CCDs did not give a sufficiently wide field of view for this, given the requirements on pixel sampling with the Lucky Exposures technique, so a special optical arrangement was designed by John Baldwin to superimpose two fields of view on the detector. This allowed stars separated by up to $30$ $as$ to be observed simultaneously. A sketch of the optical layout used for observations in July 2001 and July 2002 is shown in Figure 5.2. The light path to the detector was folded into a ``Z'' shape. The second fold mirror actually consisted of two flat mirrors butted together at a slight angle, allowing two patches of sky (typically $\sim 20$ $as$ apart) to be superimposed on the detector. In the figure, rays from two points on the sky which are superimposed on the detector are shown in red and blue. The second light path to the detector was blocked for observations of crowded fields to prevent confusion and to reduce the sky background contribution.

Figure 5.2: Schematic diagram indicating the general layout of the optics for the observing runs in July 2001 and July 2002. A diverging lens achromat was positioned before the Cassegrain image plane of the NOT increasing the f/ ratio of the beam. The light path was then folded into a ``Z'' shape. The first fold mirror was flat, while the second fold mirror consisted of two flat mirrors butted together at a slight angle so as to superimpose the images of two nearby patches of sky on the detector. In the figure, rays from two different points on the sky are indicated in red and blue. Bandpass filters were positioned just in front of the CCD. The diagram is not to scale.

For the observing run in June/July 2003, a converging lens achromat was used after the Cassegrain focus as shown in Figure 5.3, providing a re-imaged aperture plane. The light path after this lens was folded in a similar way to the observations in 2001 and 2002, and also allowed two fields on the sky to be superimposed.

Figure 5.3: Schematic diagram indicating the general layout of the optics for the observing run in June/July 2003. After passing through the Cassegrain image plane of the NOT , the light was re-focussed by a converging lens achromat. A variable aperture stop followed in the re-imaged telescope aperture plane labelled in the figure. As in Figure 5.2 the light path was then folded into a ``Z'' shape. The first fold mirror was flat, while the second fold mirror consisted of two flat mirrors butted together at a slight angle so as to superimpose the images of two nearby patches of sky on the detector. In the figure, rays from two different points on the sky are indicated in red and blue. Bandpass filters were positioned just in front of the CCD. The diagram is not to scale.

The instrument design in 2003 allowed the telescope aperture to be stopped down in the re-imaged aperture plane using a remotely controlled mechanical system. Filter changes were performed manually during the night using a rotating filter wheel. Figure 5.4 shows the fully assembled ``Luckycam'' instrument mounted at the Cassegrain focus of the NOT.

Figure 5.4: The camera and optics assembly attached to the Cassegrain focus of the NOT - Photograph by Craig Mackay

The camera used was one designed by Craig Mackay to run L3Vision detectors. The CCD was held in a liquid nitrogen dewar which was cooled to between $120$ $K$ and $140$ $K$ to minimise the dark current. An Astrocam 4100 controller was used to read out the CCD, with additional electronics providing the high voltage clock signal for the multiplication register. The CCDs were read out at frame rates between $10$ $Hz$ and $150$ $Hz$, using sub-array readout where necessary to reduce the readout time.