ISSN 1594-1906 Padova and Asiago Observatories The Echelle E2V CCD47-10 CCD H. Navasardyan, M. D'Alessandro, E. Giro, Technical Report n. 22 September 2004 Document available at: http://www.pd.astro.it/ Vicolo dell Osservatorio, 5 35122 Padova Tel. +390498293411 Fax +390418759840
This page is left intentionally blank
1 Introduction Here are presented the physical parameters, operational characteristics and some testing results of the Asiago echelle detector E2V CCD47-10, which has been mounted on July 2004. The testing procedures are the same as they are described in technical report 21. Images for the testing procedures have been taken with the CCD mounted on Echelle spectrograph. In order to escape from non linear effects due to echelle order strips, small suitable region is chosen for the linearity analysis, gain and readout calculation. Of course it is not an ideal condition for the testing, but at least it can give some estimation in the real operational conditions at the telescope. 2 Physical parameters E2V CCD47-10 is a AIMO back illuminated CCD Imager, with 1100 x 1100 13µm pixels, which corresponds to 0.49 arcsec/pxl on the Echelle focal plane. Echelle CCD can be run at one of two readout modes : 1 or 2. Gain and readout noise are calculated in the section 4 for both readout modes. The default gain setting for Echelle CCD is readout mode 2. The fringing pattern is becoming visible at Hα and it is about 17% at 8000 Å. 3 The bias Three sequences of 10 bias images taken on 2 August 2004 have been used for procedures presented in this section. Fig. 1: A sample of a row bias 1
Each image is examined carefully by eye, looking for: excessive numbers of cosmic rays: 4-5 cosmic rays are found in 512x512 sub-region wavy patterns: No any pattern is visible. The Fig. 1 illustrates a sample of raw bias frame taken on 2 August 2004 slopes: Not found hot regions: Not found dark spots: Not found The mean and standard deviation of suitable region of each bias frame is measured. The mean is within 1 ADU of the same value for every frame in the sequence, but there is an offset (less than 4 ADU) among sequences, which is possible to subtract using overscan. The histogram of a sub-region ( see fig. 2) of raw bias shows form of a Gaussian, distribution. 35000 30000 25000 Frequency 20000 15000 10000 5000 0-31799.50-31789.50-31779.50-31769.50-31759.50-31749.50-31739.50 ADU Fig. 2. Simple histogram of a subregion of a single bias image taken on 2 August 2004 A median stack of 10 bias frames is constructed. No any significant pattern and hot pixels seen in the median bias image. Fig. 3 shows horizontal and vertical cuts of a median stack of 8 bias frames taken on 3 July. No significant pattern noise is available. 2
Fig. 3: Horizontal and vertical cuts of a median stack of 10 bias frames 4 Linearity In the analysis of linearity, a sequence of flat images for each readout mode (the integration time from 10 600 sec for readout mode 1 and 10 300 sec for readout mode 2) is used taken with 1kw lamp positioned to the screen. The used setup is : CD 300, slit 200µm, decker 3. The Fig. 4 shows the conventional linearity curves - median counts versus exposure time for two readout modes and best fitted linear curves to those points. 5 Gain and Readout noise As shows transfer curve ( see fig. 5 ) the variance is near to linear and gives the gain value 2.33 for the readout mode 1 and 1.18 for readout mode 2. 3
Fig. 4 Linearity curve for the e2v CCD47-10, August 2004 The gain and readout noise for both readout modes have been measured also with findgain task of IRAF. findgain uses Janesick's method for determining the gain and read noise of a CCD from a pair of dome flats and a pair of bias frames. Where the gain is given in electrons per ADU and the readout noise in electrons(see IRAF help pages or technical report 21 for details ). A sequence of pairs of flat images taken in the same manner as for linearity analysis, such that exposure times of both images of a pair are the same and the exposure times of the image pairs generate wide range illumination, has been taken. ( 2x40sec, 2x80sec, 2x120sec, 2x180sec,2x240, 2x500sec for the readout mode 1 and 2x20 sec, 2x40sec, 2x60sec, 2x100sec, 2x140sec, 2x220sec, 2x300sec for the readout mode2). A suitable section with the size of 150x70 pixel is selected, which is free of cosmic rays and where illumination can assume be constant. For the images taken on 25 August calculated gain and readout noise are presented in Tab. 1 and 2. Table 1. Gain and readout values for readout mode 1 exp.time Gain Readout 40sec 2.315 9.082 80sec 2.305 9.042 120sec 2.37 9.297 180sec 2.321 9.105 240sec 2.344 9.195 4
Table 2. Gain and readout values for readout mode 2 exp.time Gain Readout 20sec 1.187 8.713 40sec 1.145 8.408 60sec 1.156 8.49 100sec 1.168 8.578 140sec 1.198 8.793 220sec 1.125 8.261 300sec 1.19 8.735 Fig. 5. The transfer curve for readout mode 1 and 2, for the e2v CCD47-10, August 2004 References Ref. 1: Janesick et al. 1987, Optical Engineering, 26, 2, 156 Ref. 2: Abbott T.M.C., 1995, In situ CCD testing Ref. 3: H. Navasardyan The AFOSC Tektronix TK1024 CCD testing procedures Technical Report n o 21 5