Previous: Schematic view of the CCD systems
Up: CCD Detectors
Previous Page: Schematic view of the CCD systems
Next Page: The CCD--IPCS Detector
The table below summarises the characteristics of the types of CCD chips in current use for all instruments at the WHT. Availability of CCDs changes frequently, and updates to this information will be published regularly in Gemini. At the time of writing the CCDs available on the two arms of ISIS are: EEV3 and EEV6 which are EEV P88300s; TEK1 which is a Tektronix 1024 square; and GEC5 which is a GEC P8603. In the very near future EEV8, an EEV P88500 will become available. The CCD in FOS-II is currently a GEC P8603.
There is a choice of two readout speeds for the CCDs, known as ``normal'' and ``quick''. It is possible to switch between these modes from the ICL interface with the commands SPEEDY <channelname> and SLOUCH <channelname>, which select quick and normal speed respectively. The advantages of quick speed are a faster readout (less overhead), and a higher dynamic range. The disadvantage is a higher readout noise (column 6 of the table above). For each CCD there is a distinct value of the ``gain''; the conversion factor between the output of the analogue to digital converter and the number of detected photoelectrons per pixel, for each readout speed. The values of ``gain'', usually expressed in e/adu, will be known to local staff.
All CCDs produce an image which is larger in both dimensions than the active area of the CCD, this area is given in column 4 above. The additional image area consists of underscan and overscan and dark reference regions. The difference between these types of reference regions, and the location of the regions for CCD types in use on the WHT, are discussed by P.R. Jorden in ING La Palma Technical Note no. 79. The bias level in each data frame, which is an electronic offset with a low frequency time dependence, is estimated most reliably from the serial underscan region, in regions of the image frame which are given in the table below.
On-chip binning and windowing are available for any CCD, and the parameters can be controlled using the ICL commands BIN and WINDOW. On-chip binning is useful not only in reducing the amount of data read out from the CCD, and thus the dead time between exposures, but also in reducing the effective readout noise in the data. Because the readout noise occurs only once for each summed pixel, the effective readout noise is reduced by the a factor of the square root of the number of pixels summed. On the other hand, because the output of the analogue-to-digital converter limits the amount of charge that can be read from the CCD in any one pixel, on-chip binning reduces the maximum dynamic range that the CCD can provide by this same factor.
The Tektronix CCD, being thinned, has a higher Quantum Efficiency (QE) than the EEV and GEC CCDs, particularly at blue wavelengths. At far red wavelengths (>8000Å) it does suffer from a degree of ``fringing'', which causes a modulation of sensitivity with amplitude <5%, which is very dependent on grating position. When using a thinned CCD at red wavelengths it is vital to obtain and use flatfields taken at precisely the same grating angle at the target observations.
[ TIFF ]