Astronomy Imaging by CCD Camera

A MX5 CCD camera with a (printer) port accelerator module from Starlight Xpress Ltd. using Pix-M5 control software operated through a 233 MHz PC with 32 Mb SIMM memory is the core of the imaging equipment. For eyepiece projection, typically with 6 mm orthoscopic and 10 mm Plossl eyepieces, the camera is used in combination ith a Orion Universal Combined Camera Adapter. The adapter also accepts standard filters. Filters usually employed, separately or together, are a general sodium light pollution filter (LPF) and a True Technology infra-red (IR) filter.

1. The CCD camera is rigidly located in the draw tube, and such that the image reflects true North-South and East-West star field orientation, which makes it much easier to quickly compare the image field with the sky maps (Sky Map Pro 5 and The Sky).

2. The 10 metre cable is not allowed to drag on the ground and is attached to the camera via the telescope cradle for support by Velcro tape. The cable readily twists and is carefully rewound after use onto a large diameter spool.

3.After NCP / Polar alignment a bright star e.g. Vega is used in conjunction with a Hartmann Mask to find focus. The Hartmann mask, placed on the front of the telescope, has two diametrically opposite 25 mm diameter holes with centres 155 mm apart. A 1 second exposure allows rough focussing, by racking the draw tube in and out, which decreases to 0.1 second as focus is neared. Download time between images is 7 seconds, but time between images can be controlled for focal change purposes.

4. There is no need for time wastage in finding the focal point, time after time. The nominal draw tube focus positions for a variety of eyepiece and filter combinations are readily recalled by employing wooden or plastic markers or spacers, appropriately labelled, which are cut to the length of the draw tube displacement, from a known and marked, fixed point. Once made, the appropriate marker is selected and the draw tube quickly racked to the marker length.

1. Having chosen a target the nearest visible star is located and centred in the image field using 1.5 or 2x siderial rate motor speeds on the dual-axis motor drive control unit (MDCU) situated adjacent the computer screen. 10 metre RA and Dec. coaxial cables lead from the MDCU to the telescope.

3. The RA and Dec. data are fed into a simple DOS QBasic program (Teltime.bas), written by the author. The program then calculates the time for which the RA and Dec. drive buttons on the MDCU require to be depressed, at times of 2x or 32x siderial rate, to move from the star to the target. The key mathematical factor in the program is the calculation of the speed of closure between the star and the telescope at drive speed Mx siderial rate. The closure speed is (M-1) and (M+1) times siderial rate for a target West (lower RA) and East (higher RA) of the star, respectively. Purists, may also wish to correct the closure times by division by 1.00274, for the difference between the solar day (24 hours) and the siderial day (23h. 56m. 4.1s.).

5. Generally, drive times at 32x siderial rate do not need to exceed 120 seconds to hop to the target. This is important, since failure to locate the target first time will test your patience to the limit. Reaction time to depress and release the MDCU buttons is more than adequate to find the target centrally imaged.

1. A first image is quickly taken at 1 or 2 minutes exposure to confirm the central location of the target, and give the system time to settle. Target misses are usually due to mis-representation of the starting star in the finder scope, particularly if of small magnitude. The image is rarely saved.

3. An exposure time of 4 minutes is generally chosen if the target is deep-sky. The image is saved and the file name, e.g. "M51_4m1.raw" noted in the diary, with seeing conditions. The histogram is examined and the image is usually non-linearly stretched in PIX-M5 to determine the nearness to saturation, and noise content. This determines whether more 4 minute or longer exposures are viable, or the need is to decrease the exposure time.

4. In principle, the exposure times are as long as possible, since the seeing conditions can change dramatically in minutes. With the very low dark current of the MX5 CCD, dark field images are not always vital, and are always the last images. Flat field images, with an improvised light box, are only essential on eyepiece projected images, if all is kept clean, and invariably are made the next day.

1. NCP alignment is crucial to prevent star trails, and the Drift Method is a recognised process to correct alignment. However, blurred images can result from poor focus. The criterion is generally a shift of less than one half of a pixel diameter. The depth of focus, where it can be difficult to detect the difference from perfect focus, is calculated as the width of a CCD pixel times the focal ratio. At prime focus, this calculates as 9.8 *4 i.e. 0.04 mm movement from perfect focus point before a blur occurs. This is very small which necessitates a smooth, but rigid rack and pinion focusser. You need to make it so.

2. Nyquist theory requires that if the CCD camera is to capture the diffraction limited capability of the telescope then two pixel diameters (2d) must cover the telescope's resolution limit. The relationship, dependent upon a star's Airy disc diameter, is given by 2d = 1.02wN, where w is the light wavelength, and N the focal ratio. At 520 nanometres for green light, maximum sensitivity of the MX5 CCD, and 9.8 microns for pixel diameter, the optimum focal ratio is 37.

3. Planet imaging is therefore optimum at f/37 for good seeing, but will be less for poorer conditions. This requires eyepiece projection for the f/4 Schmidt-Newtonian, but only a 2x Barlow lens for the 75mm f/16 refractor. The Schmidt-Newtonian generally achieves f/15 and f/22 with 10 mm Plossl and 6 mm orthoscopic eyepieces, respectively.

1. Colour filter sheets, with spectral transmission curves were obtained from Philip Harris school supplies. (Colour imaging and filters are discussed by M Hart). Red, green and blue sheets were laid onto adhesive backed vinyl floor tiles containing a 200 mm diameter central hole. The filter sheets could be readily attached to the front of the telescope via the dewcap.

2. The transmission of each of the RGB filters, was integrated over the 400 to 750 nanometres spectral sensitivity of the combined IR filter and Sony ICX055AL chip using a Mathcad program. The same program was used to determine the relative response of the chip to red, green and blue wavelengths from it's spectral sensitivity curve in combination with the IR filter transmission curve. This allowed the relative sensitivities of the chip plus filters to red, green and blue wavelengths to be determined as 6.6:4.2:10.8. Thus required relative exposure lengths for Red : Green : Blue are 5:7:3 minutes. This differs from the light box calculations, and no IR filter, of 3:5:4 minutes, where there is a large infra-red pickup by the CCD.

3. The focus point is noticeably different for the three filters, but too small for markers to be used. The filters are spectrally crude but, on the whole, suffice for reasonably tolerable RGBL images, and provide a start to colour imaging.