The CCD detector Sami Dib MaxPlanckInstitute for Astronomy
The CCD detector Sami Dib, Max-Planck-Institute for Astronomy, Heidelberg Jean Surdej, Institut d’Astrophysique et de Géophysique, Liège modified by Martin Hennemann, Stefan Hippler and Jutta Stegmaier (2006) 1 Introduction 2 History of the CCD 3 How does a CCD work ? 1
1 Introduction It seems that this nearinfrared (8900 Å) picture of Uranus was the first celestial object to be photographed by a CCD in 1975 by astronomers at the JPL and University of Arizona. This image has been obtained by the 61 inch telescopes located at Santa Catalina mountains near Tucson (Arizona). The dark region in the image correspond to an absorption region with some Methane bands close to the southern 2 pole of Uranus.
2 History In 1969 Willard S. Boyle and George E. Smith, while working at Bell Laboratories, designed the first Charge Coupled Device (CCD), a working version was produced just a year later. The CCD has become the bedrock of the digital imaging revolution including digital photography and video. In January 2006 they have been honored with the Charles Stark Draper Prize 3 which is presented by the National Academy of Engineering.
3 How does a CCD work? (1) Determining the distribution of an astronomical object (star, planet, galaxy, a martian spacecraft (? )) with the help of a CCD is similar to measuring the quantity of infalling rain on a field. As soon as the rain stops, collecting buckets are displaced horizontally on conveyor belts. Then the water content of the buckets is collected in other buckets on a vertical conveyor 4 belt. The overall content is sent onto a weighting system.
3 How does a CCD work? (2) pixel (a) electrodes electrons output register (b) to output amplifier The way a CCD works is illustrated by means of a simplified CCD made out of 9 pixels, an output register and an amplifier. Each pixel is divided into 3 regions (electrodes who create a potential well). (a) For the charge collection process during an exposure the central electrode of each pixel is maintained at a higher potential (yellow) than the others (green). (b) At the end of the exposure, the electrodes’ potentials are changed and the charges 5 transferred from one electrode to the other.
3 How does a CCD work? (3) (a) (b) impurity (doping) (a) By changing the potential of the electrodes in a synchronized way, electrons are transferred from pixel to pixel. Charges on the right are guided to the output register (b) The horizontal transfer of charges is then stopped and each charge package at the output register is transferred vertically to an output amplifier and then read one by one. The cycle starts again until all the 6 charges have been read. The reading time amounts to about one minute
4 Advantages of CCDs (1) 1) Good spatial resolution 2) Very high quantum efficiency 3) Large spectral window 4) Very low noise 5) Large variations in the signal strength allowed (high dynamic range) 6) High photometric precision 7) Very good linearity 8) A reliable rigidity 7
4 Advantages of CCDs (2) Spatial Resolution Mosaic of 4 CCDs containing four times 2040 x 2048 pixels. This composite detector is about 6 cm large and contains a total of 16 millions pixels (Kitt Peak National Observatory, Arizona). 8
4 Advantages of CCDs (3) Quantum Efficiency Above you see several quantum efficiency curves of different types of CCDs as a function of the wavelength. The large domain of wavelengths for the spectral response of CCDs becomes obvious. 9
4 Advantages of CCDs (4) Spectral Range FI: front illuminated BN: back illuminated, no coating DD: deep depletion 10
4 Advantages of CCDs (5) Linearity and Dynamic Range ynamic range = ratio between brightest and faintest detectable signa CCDs are extremely linear detectors, i. e. , the received signal increases linearly with the exposure time (see figure on the left). Therefore CCDs enable the simultaneous detection of both very faint and very bright objects. In contrast photographic plates have a very limited linear regime: there is a minimum exposure time for an image of an object to form. Further on during the exposure, the image gets saturated quickly (S-shape gamma curve). The dynamic range of CCDs is about 11 100 times larger compared to
4 Advantages of CCDs (6) Flat field technique (a) (b) (c) 12
5 Observations with a CCD (1) 4 exposures of the galaxy M 100 with exposure times of 1, 100 and 1000 seconds (obtained with a 11 inch Celestron telescope). 13
5 Observations with a CCD (2) 5. 1 Subtraction of the bias Processed image Raw image. . . 14
5 Observations with a CCD (3) 5. 2 The darks (1) Sn(t) = Rn 0 2(T - T 0) / T t (5. 2. 1) 15
5 Observations with a CCD (4) 5. 2 The darks (2) ST = n · S and NT 2 = (n · N 2) (5. 2. 2) ST / NT = (S / N) n (5. 2. 3) S = Sa - ST and N = (Na 2 + NT 2) (5. 2. 4) S / N = (Sa - ST) / (Na 2 + NT 2) (5. 2. 5) 16
5 Observations with a CCD (5) 5. 3 The flat field technique (1) S = S o / Sf (5. 3. 1) (S/N) = 1 / [(No/So)2 + (Nf/Sf)2] (5. 3. 2) 17
5 Observations with a CCD (6) 5. 3 The flat field technique (2) Raw image (left) from which we subtract the Bias image (middle) and the dark image (right). We then divide the obtained result by the flat 18
5 Observations with a CCD (7) 5. 4 Cosmic rays The impact of many cosmic rays are visible on this dark image 19
5 Observations with a CCD (8) 5. 5 Improving the S/N ratio of astronomical observations N = N 12 + N 22 + N 32 +. . . (6. 5. 1) S = S o + Sn + S c (6. 5. 2) N 2 = No 2 + Nn 2 + y 2 + Nc 2 (6. 5. 3) S/N = (So + Sn + Sc) / No 2 + Nn 2 + y 2 + Nc 2 (6. 5. 4) 20
5 Observations with a CCD (9) 5. 5 Improving the S/N ratio of astronomical observations S/N = (So + Sn + Sc) / So + Sn + Sc + y 2 (5. 5. 5) S/N = Co / 1 + Cc / Co + DC y 2 / Co (5. 5. 6) S/N = Co (5. 5. 7) 21
5 Observations with a CCD (10) 5. 5 Improving the S/N ratio of astronomical observations S 1 = Si = n · Si, N 1 = ( Si) = (n · Si), S 1/N 1 = (n · S i) (5. 5. 8) S 2 = n · Si, N 2 = S 2, S 2/N 2 = (n · Si) (5. 5. 9) 22
5 Observations with a CCD (11) 5. 5 Improving the S/N ratio of astronomical observations S 1 = Si = n · Si, N 1 = ( (Si + y 2)) (n · y 2) S 1/N 1 = (n · Si)( Si /y) (5. 5. 10) S 2 = n · Si, N 2 = S 2, S 2/N 2 = (n · Si) (5. 5. 11) S 1/N 1 = S 2/N 2 ( Si / y) S 2/N 2 (5. 5. 12) 23
5 Observations with a CCD (12) 5. 6 Determination of the gain and the read out noise of a CCD g nmax / 216 (5. 6. 1) N 2 = S o + S n + S c + y 2 (5. 6. 2) N 2 ADU = SADU / g + BDL 2 (5. 6. 3) 24
5 Observations with a CCD (13) 5. 6 Determination of the gain (and read out noise) of a CCD with the photon-transfer method Linear slope = CCD gain in units of e/ADU 25
5 Observations with a CCD (14) 5. 6 Determination of the gain and read out noise of a CCD (f 1 / f 2) / f 1/f 2 2 = 1 / ( f 1/f 1)2 + ( f 2/f 2)2 1 / 2( f/f)2 (5. 6. 8) f 2 = (f 2 / 2) ( f 1/f 2)2 (5. 6. 9) 26
5 Observations with a CCD (15) CCD image of Arp 188 and the Tadpole's Tidal Tail taken with Hubble’s ACS camera. 27
- Slides: 27
