1. Edimburg 22-23 June 2006 Fast detectors 1 Fast detectors (for E-ELT AO) Philippe Feautrier INSU/CNRS-LAOG

2. Edimburg 22-23 June 2006Fast detectors2 Highlights of the FP6 JRA2: Fast detectors for AO SPHERE XAO VLT Controller (Marseille) Camera head (Grenoble) Peltier cooled package e2v 240x240 L3CCD First light: 2011 November 2006 September 2006 2007 November 2006 JRA1 FP6 Will be used on every 2nd generation AO instruments in Europe !

3. Edimburg 22-23 June 2006 Fast detectors 3 What are now the needs for ELT WFS detectors ?

4. Edimburg 22-23 June 2006Fast detectors4 6” ” 2.5”.75” LGS based AO challenges on ELT Spot Elongation aperture 12m away elongation of 3” LGS image Courtesy KECK Laser Launch Site Probably need ~ 20x20 pixels for each spot to properly sample the elongation. Rayleigh Backscatter Worse with Telescope Size  Large amount of pixels for spot elongation issue  With pulsed laser: electronic shutter (sodium spot only)

5. Edimburg 22-23 June 2006Fast detectors5 Exoplanets detection on E-ELT, requirements:   Higher contrast (up to C=10 9 ) (SPHERE: 10 7 )   Smaller angular distance (as close as  =30 mas) (SPHERE: 100 mas)   Limitations imposed by Temporal Delay: Halo  at small  (temp. part   5/3 ) Halo temp. part  (1 / Frame rate) 2 From 1 KHz to 3.2 KHz  Planet detected 10x faster! (when temporal errors dominate) Contribution to halo due to temporal delay Integration time needed to detect Planet  Star Halo Intensity   Limitations imposed by Atm. Chromatic effects Halo  at small  (chrom. part   5/3,  11/3 ) When  = ( obs - wfs )  then Halo intensity  WFS from R to I band  Planet detected 7x faster! (when chromatic errors dominate) Contribution to halo due to chromatic effects obs =1.6  m   Limitations imposed by actuator density: Need High SR (>90% at 1.6  m)  15-20 cm sub-aperture size  200 2 - 256 2 actuators (  42/0.2) for a 42-m telescope WFS baseline solution: Pyramid with 4 detectors Pyramid requires less pixels than SH: 4 pix. / sub-ap. but split on 4 detectors, i.e. 1 pix/ sub-ap = 256 2 pixels for each detector XAO CCDs requirements:   256 2 pixels   3 KHz (goal 4 KHz) frame rate   High QE for = 800 to 1000 nm   Low read noise: 1-2 electrons/frame E-ELT XAO WFS CCDs Requirements

6. Edimburg 22-23 June 2006Fast detectors6 Preparation activities in the frame of FP6  Funded by ESO (FP6 ELT design study): –2 feasibility studies for LGS AO detector dedicated to ELT (6 months, 50 k€ each) and one prototype (~1 year: 2007, 200 k€) –XAO best effort prototype based on PN Sensor technology: 256 2 pixels, one amplifier per column, 200 k€, testing included  At the end, 2 types of AO detectors for ELTs are required: LGS and XAO

7. Edimburg 22-23 June 2006Fast detectors7 A possible solution for LGS WFS detector ? Store Image Serial Register (courtesy Roger Smith, Caltech) Analog Front End Programmable Centroiding Logic Array Simple Sequencer Clocks & Biases MasterclockPowersupply Centroid X,Y positions, intensity x100 x100 Programminginterface 20 pixels  100x100 rectangular array of small CCDs.  Each CCD: 20x20 pixels  700 Hz frame rate, ~ 1 e noise  Electronic shutter (for spot elongation)

8. Edimburg 22-23 June 2006Fast detectors8 PN-CCD: a possible solution for XAO WFS on ELT ?  Interesting to evaluate now this technology for XAO  Propose a 200 k€ best effort contract with MPE/PN-sensors manpower  256x256 split frame transfer PN-CCD device.  512 outputs, uses four ROICs for goal frame rate of 3000 frames/sec at 1.5 e RON I I I I M M M M 128 256 Outputs 256 Outputs 128:2-3 Mux. 2-3 Outputs 9 Mp/s ROICROIC 128:2-3 Mux. 2-3 Outputs 9 Mp/s ROICROIC 128:2-3 Mux. 2-3 Outputs 9 Mp/s ROICROIC 2-3 Outputs 9 Mp/s ROICROIC 128:2-3 Mux.

9. Edimburg 22-23 June 2006Fast detectors9 Fast PN-CCD detector fabricated by MPE Main features: detector size = 27×13.5 mm 2 51 μm pixel size 528×264 pixels in total, 264×264 in image area split frame-transfer concept image transfer time ≈ 30 μs readout noise vs. frame rate: 1.8 e - @ 10.. 400 fps 2.5 e - @ 400.. 1200 fps Next step:  256x256 pixels  3 kHz Frame rate  1-2 e Noise  Good red response

10. Edimburg 22-23 June 2006Fast detectors10 Preliminary management structure

11. Edimburg 22-23 June 2006Fast detectors11 Preliminary deliverables and schedule 200820092010201120122013 Detector definition and CFT Detector LGS dev Detector LGS dev Detector XAO dev Detector XAO dev LGS test camera XAO test camera LGS detector testing XAO detector testing

12. Edimburg 22-23 June 2006Fast detectors12 Eligible costs  XAO detector cost: 1 M€  LGS detector cost: 2 M€  Controller and camera head XAO: 200 k€  Controller and camera head LGS: 200 k€  Detector testing activity: 200 k€  Total: 3.6 M€  Requested to EC ?

13. Edimburg 22-23 June 2006Fast detectors13 Conclusion  JRA2 FP6: was dedicated to 2 nd generation of AO instrument for 8-10 m class telescopes  FP7: two needs for E-ELT AO WFS detectors identified: –LGS WFS detector: very large pixel format to sample spot elongation, high frame rate (700 Hz), low noise (~1 e).TBC. –XAO WFS detector: for a pyramid WFS, 256 2 pixels are sufficient, but extremely high frame rate (3 to 4 kHz) and very low noise (~1 e) are required. TBC.  ~ 5.5 years development (2008-2013)  3.6 M€ eligible cost  For the proposed activity: at least ~ 3 European countries (Germany, Spain, France), 2 European institutions (ESO, IAC), ~ 3 laboratories (INSU/CNRS), 2 industrial partners.  AO applications are strategic for E-ELT but activity open to wider applications in the field of fast detectors.

14. Edimburg 22-23 June 2006Fast detectors14 LGS detector: could be Hybrid or Monolith  Monolith - cheaper but detector silicon determined by CMOS process; ok for LGS at 589nm.  Hybrid – expensive but detector can be optimized for good red response by use of high resistivity silicon. Package Pin Package CMOS Sequencer, AFE AND Centroid Logic Array Back Filled Epoxy Indium Bumps Array of small CCDs Bond Wire