The CCD detector

1
The CCD detector

• Sami Dib and Jean Surdej
• 2006 online edition and modifications by Stefan
  Hippler
• Dont forget to read the notes pages of each
  slide
• 1 Introduction
• 2 History of the CCD
• 3 How does a CCD work ?
• 4 Advantages of CCDs
• 5 The CCD as a 3 dimensional detector
• 6 Observations with a CCD

2
1 Introduction
It seems that 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). It has been made at the
8900 Å wavelength in the near Infrared. The
dark region in the image correspond to an
absorption region with some Methane bands close
to the southern pole of Uranus.

3
2 History

Determining the brilliance distribution of an
astronomical object (star, planet, galaxy, a
martian spacecraft ?) with the help of a CCD is
pretty much similar to the measurements of the
quantity of infalling rain on a farm. 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 belt. The overall
content is sent onto a weighting system.
4
3 How does a CCD work ? (1)

Output register
Pixel
(a)
(b)
To Output amplification
Electrodes
Electrons
The way a CCD works is illustrated by means of a
simplified CCD made of 9 pixels, an output
register and an amplifier. Each pixel is divided
in 3 regions (electrodes who serve to create a
potential well). (a) when an exposure is made,
the central electrode of each pixel is maintained
at a higher potential (yellow) than the others (
green) and the charges collecting process takes
place. (b) At the end of the exposure, the
electrodes potentials are changed and charges
transferred from one electrode to the other.
5
3 How does a CCD work ? (2)
(b)
(a)
Impurity (doping)

• By changing in a synchronized way the potential
  of the electrodes, 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 charge packages at the output
• register are transferred vertically, one by one,
  to an output amplifier and then read one
• by one. The cycle starts again until all the
  charges have been read (reading time of about
• 1 minute for a large CCD).

6
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 each 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

Quantum efficiency curves of different types of
CCDs as a function of the wavelength compared to
the one of other detectors. We can see on this
plot the large domain of wavelengths for the
spectral response of CCDs.
9
4 Advantages of CCDs (4)

• Spectral
• Range
• FI front
• illuminated
• BN back
• illuminated, no
• coating
• DD deep
• depletion CCD

10
4 Advantages of CCDs (5)

• Linearity and Dynamic Range

Dynamic range ratio between brightest and
faintest detectable signal
CCDs are extremely linear detectors, i.e the
received signal increase linearly with the
exposure time. The CCD thus enables the
simultaneous detection of very faint objects and
bright objects. In contrast photographic plates
have a very limited linear regime. First of
all there is a minimum exposure time below which
no image of the object forms. At some higher
degree of exposure, the image gets quickly
saturated (S-shape gamma curve). The dynamic
range of CCDs is about 100 times larger compared
to Film.
11
4 Advantages of CCDs (6)

• Flatfields

(b)
(c)
(a)
flat field technique (see text below)
12
5 The CCD as a 3 dimensional detector 6
Observations with a CCD
As can be seen from this series of 4 exposures (
figures above next page ) of 1, 10, 100 and
1000 sec, of the M100 galaxy, obtained with a 11
inches Celestron telescope, the signal to noise
ratio changes in a crucial way as a function of
the exposure time
13
6 Observations with a CCD (1)
Additionally to the improvement of the S/B ratio
as a function of the exposure time, we can also
clearly see the change in the regime of the
noise, mainly caused by the readout noise of the
CCD in the shorter exposure, and to the photons
noise in the sky for the longest exposure.
14
6 Observations with a CCD (2)

• 6.1 Subtraction of the bias

Processed image
Raw image ...
15
6 Observations with a CCD (3)

• 6.2 The darks (1)
• Sn(t) Rn0 2(T - T0) / ?T t. (6.2.1)

16
6 Observations with a CCD (4)

• 6.2 The darks (2)
• ST N S et BT2 (N B2),
  (6.2.2)
• ST / BT (S / B) ?N .
  (6.2.3)
• S Sa - ST et B ?(Ba2 BT2),
  (6.2.4)
• S / B (Sa - ST) / ?(Ba2 BT2).
  (6.2.5)

17
6 Observations with a CCD (5)

• 6.3 The flat field technique
• S So / Sf,
  (6.3.1)
• (S/B) 1 / ?(Bo/So)2 (Bf/Sf)2.
  (6.3.2)

18
6 Observations with a CCD (6)

• 6.3 The flat field technique (1)

Raw image (left) from which we substract the Bias
image (right) ... and the dark image
(below) (see also next page).
19
6 Observations with a CCD (7)

• 6.3 The flat field technique (2)

We then divide the obtained result by the flat
field image (above) and obtain The final image
(right).
20
6 Observations with a CCD (8)

• 6.4 Cosmic rays

The impact of many cosmic rays are visible on
this dark image
21
6 Observations with a CCD (9)

• 6.5 Improving the Signal to Noise (S/B) ratio
  of astronomical observations
• B ??B12 B22 B32 ...?.
  (6.5.1)
• S So Sn Sc,
  (6.5.2)
• B2 Bo2 Bn2 y2 Bc2,
  (6.5.3)
• S/B (So Sn Sc) / ?? Bo2 Bn2 y2 Bc2?.
  (6.5.4)

22
6 Observations with a CCD (10)

• 1.6.5 Improving the S/B ratio of astronomical
  observations
• S/B (So Sn Sc) / ?? So Sn Sc y2?.
  (6.5.5)
• S/B ?Co / ??1 Cc / Co n y2 / Co?. .
  (6.5.6)
• S/B ?Co.
  (6.5.7)

23
6 Observations with a CCD (11)

• 6.5 Improving the S/B ratio of astronomical
  observations
• S1 ?Si N Si, B1 ?(?Si) ?(N Si), S1/B1
  ?(N Si)
• 
  (6.5.8)
• S2 N Si, B2 ?S2, S2/B2 ?(N Si).
• 
  (6.5.9)

24
Observations with a CCD (12)

• 1.6.5 Improving the S/B ratio of astronomical
  observations
• S1 ?Si N Si, B1 ?(?(Si y2)) ? ?(N y2),
• S1/B1 ?(N Si)(?Si /y)
  (6.5.10)
• S2 N Si, B2 ?S2, S2/B2 ?(N Si).
  (6.5.11)
• S1/B1 S2/B2 (?Si / y) ?? S2/B2.
  (6.5.12)

25
6 Observations with a CCD (13)

• 6.6 Determination of the gain and the read out
  noise of a CCD
• g ? Nmax / 216.
  (6.6.1)
• B2 So Sn Sc y2,
  (6.6.2)
• B2ADU SADU / g BDL2.
  (6.6.3)

26
6 Observations with a CCD (14)

• 6.6 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
27
6 Observations with a CCD (15)

• 6.6 Determination of the gain and read out noise
  of a CCD
• ?(f1 / f2) / ?f1/f2 ?2 1 / ?(?f1/f1)2
  (?f2/f2)2? ? 1 / ? 2(?f/f)2?,
• 
  (6.6.4)
• ?f2 (f2 / 2) (?f1/f2)2.
  (6.6.5)

28
6 Observations with a CCD (16)
CCD image of Arp 188 and the Tadpole's Tidal Tail
taken with Hubbles ACS camera.