SUSI2  CCD Information        HOME INDEX SEARCH HELP NEWS

Data Format

The two EEV chips are identified as ESO CCD #45 and ESO CCD #46. They are read into a single Multi-Extension FITS file, with chip #45 having the lower x values (i.e. it is on the left on a standard display). They can be combined into a single FITS image using the MIDAS script insusi.prg. The two chips are read out in parallel, each through a single port, at 2 x 10⁵ pixels/sec. On the RTD, the default orientation has North at the top and West (chip #45) to the left. This is illustrated in the following figure:

The space between the two chips has been "filled" with overscan columns so that the respective geometry of the two chips is approximately preserved. The format of the file is 4288 x 4096. Along the x axis there are 50 prescan, 2048 active, 46 overscan pixels for chip #45, followed by 50 prescan, 2048 active and 46 overscan for chip #46.

The scale at the Nasmyth focus of the NTT is 5.4 arcsec/mm. The measured pixel scale is 0.0805 arcsec/pixel (0.161 arc/sec in the 2 x 2 binned mode). Each chip of the 2x1 mosaic covers a field of 5.5 x 2.7 arcmin. The gap between the chips corresponds to ~ 100 real CCD pixels, or~ 8 arcsec. In the default position (rotator=0), the gap between the chips runs in the N-S direction. The optical axis (which is also the reference for pointing) is at about 60 pixels left of the center of the mosaic (i.e. in chip #45) very close to the inter-CCD gap.

CCD QE

Measurements of the QEs for the two chips in the ESO detector laboratory show that chip #45 is up to 5% more efficient than chip #46.

The table below gives the QE (averaged over both chips) at different wavelengths:

Angular Misalignment - Astrometry

The angular misalignment between the two chips is smaller than 6 arcmin. Very accurate astrometry requires an independent calibration of the two subfields. The table below lists the mean astrometric parameters (in milidegrees per pixel) over the two chips:

Other CCD basics

Binning: 
There are two supported binning options (1x1 and 2x2), though one can also use 3x3 and 4x4 binning at own risk (i.e. we cannot guarantee the stability of the gain and bias, nor the non-linearity of these modes since we do not monitor them).

Windowing: 
The chip can be windowed, and this window may be defined so that it overlaps both chips.

Read Out Times: 
The read-out times (interval from the closure of the shutter to the complete transfer of data) are:

Binning:	Read-out Time
1x1	56 seconds
2x2	16 seconds

Read Out Noise: 
The measured read-out noise is:

Chip:	Read-out Noise
#45	4.7 e-
#46	4.6 e-

For the latest measurements of the SUSI2 CCD parameters, please check the Summary of latest SUSI2 CCD tests. To follow the evolution of these parameters, check out the NTT detector page which is updated ~weekly.

Saturation and Linearity: 
Pixel saturation occurs at ~150000 e-. With the adopted gain of ~2.25 e- /ADU this approximately corresponds to the full range of the ADC converter in the unbinned mode. The CCDs have been found linear within + 0.15% over the full range 0-60000 ADUs.

Charge Transfer Efficiency:
Both serial and horizontal CTE are better than 0.999999.

CCD dark current:
<0.5 e-/pix/hour, i.e. negligible for all practical purposes.

Bias Level: 
There is a slight dependence on the mean level of the charge in each row of the CCDs, therefore, the bias level should be derived from the average of prescan and overscan regions of each.
A 0 second bias exposure averages ~300adu, but check the summary of the latest SUSI2 CCD tests for up to date values.

Cosmetics:
Chip #45 (left) has ~15 bad columns, while #46 has some blemishes that look quite bad but actually flat-field out very well. See the SUSI2 BADPIXEL MASK page for more details and the masks themselves.

Send comments to : ls-susi Last modified: Tue Nov 4 2003