We have recieved delivery of another thinned, blue optimised CCD from
EEV. It is similar in format and characteristics to that which is
already being used on the blue arm of ISIS. The commissioning of this
chip has begun on the IDS spectrograph on the INT and we hope to be
able to offer this as a fully commisioned device as soon as possible.
Before its final acceptance it will be available to observers on a
shared risk basis, and potential users should read this information
carefully and contact Stephen Smartt sjst@ing.iac.es with any further
queries. A request to use the device should be done through the usual
channels with your Support Astronomer.
The device with the ING label of EEV12 is an EEV42 type, science
grade CCD. The silicon area is 4096 x 2048 pixels, which are each
13.5 microns square. It is a thinned chip with excellent blue sensitivity
matching anything currently available on the market. A quantum
efficiency curve is available at the
RGO CCD detector groups Web pages, (or click here for a FRAMES
version) but as this is usually
meaningless to the Observer without the response curve of the
rest of the optical system, you are referred to the
Flux standard
data which is a real measure of what you expect to see while
observing with this chip.
It has two overscan regions covering the
areas [1:4200,2049:2148] and [4097:4200,1:2148]; where these coordinates
are in the form [xstart:xend,ystart:yend].
While the chip itself can support two gain settings this feature is
not currently available on the INT DAS system. It is likely it will be
implemented.
The information on this page is geared towards the prospective Observer
using IDS with this particular device mounted. Hence
it has been driven by considerations of the whole system performance.
For a fuller, more technical and general description of the EEV12
chip (and the other devices available on La Palma) see the
ING Detector Group's home-page.
2. Vital statistics - gains, readout noise, etc
In the following tables the values for gain are given in e-/ADU,
the read-out noise in e-, and
the read-out time is in seconds for a windowed
chip which covers the full IDS slit length (columns 800-1300 approximately).
If one expects saturation in some spectral features then it is a good
idea to avoid using TURBO because of possible 'blooming' effects
around the saturated pixel. On this chip, when used on IDS
we see no evidence of pick-up noise when used in either QUICK or
TURBO readout speed.
Currently only binning in the spatial direction is possible. A first
inspection shows that there is no obvious pick-up noise when
using binning by 2x1. However when binning 4x1 some fixed pattern
noise is visible at the level of +/- 7 adu. Observers use
binning (other than 2x1)) at their own
risk until we quantify possible problems .
HIGH GAIN
Unbinned
Binned 1x2 (spectrally)
Binned 2x1 (spatially)
Binned 2x2 (both)
Speed
gain (e-/adu)
Noise (e-)
Time (s)
gain
Noise
Time
gain
Noise
Time
gain
Noise
Time
TURBO
1.50
4.3
70
1.47
4.5
50
QUICK
1.22
4.0
75
STANDARD
1.10
3.2
80
3. Spectral resolutions and wavelength coverage
The table linked here gives the dispersion provided by each grating when
mounted on the 235mm and the 500mm cameras.
Observers can calculate the dispersion provided, given that the
pixel size is 13.5 microns.
The EEV12 is mounted on both cameras with it's 4096 pixel axis along the
wavelength direction, giving maximum use of the beam width leaving the
cameras. However the camera optics severely vignette the outer regions of the
dispersed light beam such that approximately only 2300 of the CCD
pixels are clear and unvignetted. The attenuation runs rises steeply
at each end of the spectra. The unvignetted portion is roughly from
pixel number 850 to 3150 in the spectral direction.
A plot of a
flat-field exposure across the chip is shown here
to illustrate this effect.
Observers should note that the 13.5 micron pixels allow a higher
spectral (and spatial) resolution to be achieved than when using the
TEKS. However this is of course at the expense of slit width.
Since the two cameras have different magnifications, and hence
different slit-width specfications for a particaular require
resolution, Observers should think carefully about what is the
best option for their programs. For example, it is now possible to
reach a dispersion of higher than 0.8 Angstroms per pixels with the
235mm camera and this chip. The best focus we have achieved so far on
the 235mm camera is a FWHM of 2.2 pixels per arc line, with a 1 arcsec
slit-width. This is unlikely to decrease to exactly 2.0 pixels during normal
setups, due to the accuracy that one can adjust the capstans, and the
slight charge spreading visivle with these thinned chips.
Observers should contact Stephen Smartt sjst@ing.iac.es if they need any advice
regarding the most effiecient use of IDS and the EEV/TEK detector choice.
4. Spatial scale on IDS
A spatial scale of
0.4 arcsec/pixel and 0.19 arcsec/pixel is achieved on the 235mm and
the 500mm cameras respectively.
In the future it should be possible to
bin in the spatial direction if one is not concerned with high spatial
resolution observations, indeed the seeing conditions at the INT need to be
excellent to allow full advantage to be taken of using an unbinned
chip with this pixel scale. See above for binning
characteristics in the spatial direction.
The maximum unvignetted
slit-length usable with IDS is 3.3 arcmin and 1.6 arcmin for the
235/500mm respectively (corresponding to a 500
detector pixels, spanning ~[800:1300, 1:4096].
5. Fringing and Cosmetic defects
These thinned chips suffer severely from fringing in the red
part if the spectrum, which limits their usefulness in this region
despite their continued good QE down to 8000Å. Some
illustrative
flat field spectra will be linked from here soon,
but in the meantime the following numbers should serve as a
reference guide to the severity of the problem :
There are a few cosmetic defects on the surface of the chip, but
nothing particularly severe. A
flat field
image plus bias image will be visible from here shortly.
6. Linearity measurements
The following plot shows a linearity test carried out on the EEV13
chip on ...... The CCD was not binned, and ???? read-out speed.
The chip is linear to better than ?? at low light levels
(below 1000-2000 ADU),
and this is the around the noise level expected from shot and read
noises at the low light levels used i.e. it is linear to within
our measured noise limits. At high count levels it is linear to
better than 0.5%.
7. Charge spreading variations and effects on spectral
resolutions
The diffusion of charges between pixels during integrations causes a
degrading of the spatial and spectral resolution. For a long-slit
spectrograph like IDS, with the INT's mean seeing (of around 1.0-1.5")
spatial degradation not a significant worry with the pixel size
of the EEV12, but it is a consideration in the spectral direction.
For a back illuminated CCD this charge diffusion (often
referred to as the Modulation Transfer Function; MTF) becomes
progressively worse for shorter wavelength incident light. For example,
using a slit-width projecting to 2 pixels on the detector results in
a FWHM measured of 2.4 pixels (measured at ~4000Å) when the
spectrograph is at best focus. Similarly
a slit-width projecting to 4 detector pixels will produce a FWHM of
~4.4 pixels (again at ~4000Å).
This effect becomes less severe towards redder wavelengths and is
negligible at around 6000Å.
8. Flux standard data and empirical through-puts
coming soon ....!
9. Quality control history of EEV12
Coming soon .....!
This page last updated: 28 Jul 1998
