Focal Plane Detectors for the Dark Energy Survey

Focal Plane Detectors for the Dark Energy Survey Juan Estrada (Fermilab) and Ricardo Schmidt (CTIO) for the DES Collaboration Abstract The Science - Dark Energy Using New Probes The scientific requirements for the focal plane array for the. Dark Energy Survey (DES) are reviewed, and the status and plans for CCD testing andcharacterization for this project are presented. Image credit: Roger Smith/NOAO/AURA/NSF q Introduction to DES Science goal: A study of the dark energy using four independent and complementary techniques: Two linked, multiband optical surveys n 5000 deg 2 g, r, i and z n Repeated observations of 40 deg 2 for SN studies (10% of the survey time). q Instrument and schedule n New 3 deg 2 camera on the Blanco 4 m on Cerro Tololo n Construction: 2004 -2009 n Survey Operations: 30% of telescope time over 5 years Blanco 4 m on Cerro Tololo DES focal plane: q The angular power spectra is used as a standard ruler q 300 million galaxies q. Can be broken up into bins of photometric redshift q. Peak and baryon oscillation features provide ruler Galaxy Cluster counting q 20, 000 clusters to z=1 with M > 2 x 1014 M q Cluster angular power spectrum q Dark Energy using the gold standard probe, Type 1 a Supernovae distances q 2000 supernovae q 40 sq-degrees q. Revisit at 3 night intervals q. Photo-z for all host galaxies q. Spectroscopic-z for ~1/4 of all host galaxies. Weak lensing q 300 million galaxies q Photo-z accuracy of δz < 0. 1 to z = 1 q 10 -20 galaxies/sqarcminute DECam ~50 cm diameter • 62 CCDs (2 k x 4 k) • 500 Mpix • 3 square degrees 5 m m 3556 mm Scroll Shutter 157 Camera Filters DECam, as designed for Dark Energy Survey. The complete instrument will be built at Fermi National Accelerator Laboratory (http: //www. fnal. gov) by a collaboration of 13 universities and laboratories. CCD A key piece of this project are the CCD developed by LBNL. DECan simulated image • QE> 50% at 1000 nm • 250 microns thick • readout 250 kpix/sec • 2 RO channels/device • readout time ~17 sec See talks by LBNL group in this conference. r CCD testing and characterization g In order to ensure that the CCDs for installation in the DECam focal plane achieve the requirements. Every device will be tested and characterized before installation. Readout Electronics For testing and characterization of devices that will be installed in the telescope we will use Monsoon based DAQ system. (see presentation by CTIO group). We expect to have a working Monsoon based testing station by fall 2005. This is the readout electronics that we have selected for DES operation. This exercise will allow us to understand how to achieve our science requirements (noise + readout speed) with a Monsoon based system. Monsoon crate and power supplies under commissioning at Fermilab. A Monsoon 12 channels acquisition board is being developed for this project. As an alternative readout system and as beginning of the testing effort, we have been using the electronics from The Amateur Sky Survey (TASS http: //www. tasssurvey. org/) , that has been adapted at Fermilab for operation of LBNL CCDs. We thank Tom Droege for generously lending us this system. Full scale test vessel A system test will be performed at Fermilab with a full scale test vessel. This will be useful for understanding the mechanical design of DECam, and will allow for the operation of several CCDs (the initial plan is for a focal plane of 6 CCDs) simultaneously. We anticipate the need of testing approximately 200 devices, and rate of 1 per day during production. We estimate 7 hours per CCD, and about 1 Gbyte of data per device. i z Effect of higher efficiency in the z band The CCDs that we are planning to use in DES (produced by LBNL) have significant higher QE in the red part of the spectrum compared with typical thinned CCDs. This means we can loof further in the sky, as objects move further from us (higher redshift), their spectrum moves towards the red. Transfer efficiency for DES filters The increase in QE means the flux required for the detection gets lower by the square root of the ratio of the efficiencies (QE 1/QE 2)1/2. The power of the higher efficiency in the red, can be shown using Supernovae Ia. The plot shows the number of photons reaching the detector in the z band, and a function of the redshift of the SN Ia. The increased efficiency of the LBNL CCDs compared to a typical thinned CCD can increase the detection limit by z=0. 25. Discovery Thr thinned CCD. Discovery Thr. w/LNBL CCD vacuum LNe 5 liters Prototype CCD testing station Dewar view, with front cover removed. Motorized filter wheel (6 positions) Fe 55 source that can be hidden using this knob. Ne out Liquid Ne input Intregrating sphere 6’ Calibrated photodiode illumination CCD dewar fused silica window detachable front for easy access to CCD The existing prototype testing station has been used to readout LBNL CCDs (1 k x 0. 5 k). It is equipped with a filter wheel and broadband illumination source that allows for measurements at different wavelength. 174 173. 8 temperature stability 173. 6 Temperature (K) CCD • Active area of 170 sq-arcminutes • 15 micron pixels = 0. 272’’ • Fringe elimination CCD testing and characterization Schematic of the plane of De. Cam is shown here compared with an image of the moon taken with the existing MOSAIC 8 CCD camera at Blanco (CTIO) Fe 55 Focal Plane Lenses CCD plate 173. 4 173. 2 173 172. 8 172. 6 172. 4 cold finger 172. 2 The illumination of the CCD is done using a halogen lamp (max power 100 watt). The source has better than 0. 5% stability, 200 minutes after turn on. 172 470. 00 475. 00 480. 00 485. 00 490. 00 Time (minutes) less than 0. 1 K fluctuations achieved in the CCD plate, using a PID loop with peak heating power of 1 watt.