CCD Lab Report

CCD Camera

9/30/15

Joshua Kern

Theory

Charge-coupled devices (CCD’s) are the primary devices astronomers use to image the sky. CCD’s have arrays of pixels where each pixel can be thought of as a potential well. When photons impinge upon the surface of the pixel electrons are released that, when biased by an applied voltage, create a measurable current. However, imperfections in the individual pixels cause errors in the measured count of electrons produced by the incident light.

These errors in measurement have multiple sources. Even if an image is taken with zero exposure time, the CCD will read a small signal known as the bias level. Also, the ‘hum’ of the electronics is responsible for creating a signal that is known as the dark current. The dark current accumulates with time at a constant rate and can be accounted for by taking images with a closed shutter. The third source for signal error is from pixel to pixel sensitivity. In other words, some pixels accumulate counts easier than others or a spec of dust might block light from entering certain pixels. This can be accounted for by taking images of uniformly lit screens which are known as flat-field images. Finally, stray light from other objects in the night sky can contribute to the signal as well. All of these sources of additional signal must be accounted for and subtracted off your data in order to leave only the data from your object of interest.

Experiment

In this experiment we characterized the bias level, dark current, and pixel sensitivity of the ST-I model CCD camera. We took 10 images with an exposure time of zero seconds to determine the bias level, 10 images with an exposure time of 0.001 seconds with the shutter closed to determine the dark current, and 10 flat-field images with the same exposure time the same as our dark current images to determine the pixel sensitivity. We also took images with the shutter closed at exposure times of 1, 2, 5, 10, and 20 seconds in order to determine the rate at which the counts increase due to the dark current.

Results & Discussion

After obtaining the images, we used READFITS in IDL to reduce the data. By taking the MEAN of each image, we obtained an average bias level and dark current of 1021.4427 and 1066.6979, respectively. The table below gives all the average values for each bias, dark, and flat image which we used to obtain the previously stated values. The flat-field images were uniform with variations in counts increasing to 50,000 towards the top-left portion of the image, and decreasing to 40,000 in the bottom-right portion.

Image	Bias	Dark	Flat
1	998.4084	1096.4714	38974.117
2	1015.5076	1099.4374	48256.285
3	1035.1007	1101.3552	47070.859
4	1019.3893	1086.2279	49321.992
5	1021.3006	1094.8556	47982.125
6	996.5331	1055.5481	47343.168
7	1027.8583	1039.1448	47978.512
8	1031.2988	1029.1586	47562.883
9	1033.0098	1030.3505	47111.668
10	1035.9933	1034.4302	46365.031

To determine the rate at which the dark current changes with time, we took images with the shutter closed with increasing exposure times in order to plot the average counts in each image versus the time of exposure. The data we obtained are plotted in the graph below.

As you can see, the data that was taken does not show strict linearity however a linear trend is seen. After taking the line of best fit through the data, we obtain a rate of change of dark current equal to 0.70 ± 0.005 counts per second. The values we obtained for the bias level, dark current, and the rate of change of dark current as well as the characterization of the flat field images allow us to better reduce data obtained with the ST-I camera in the future.