2dF and its Spectrographs

1
(No Transcript)
2
2dF and its Spectrographs
2dF spectrographs mounted on top end ring of AAT

• Designed for
• 2dFGRS
• Limited in
• Efficiency
• 4-7 peak
• Resolution
• Rlt4000
• Detector area
• 1K x 1K
• Stability
• few pixels

3
The AAOmega Project

• 2000-2005
• Refurbished 2dF top end
• 392 ? 35m MOS fibres, 2.0" diameter
• 8 guide bundles plate rotation
• New, stable, high efficiency, bench-mounted
  dual-beam, articulating spectrograph with VPH
  gratings and 2kx4k CCDs
• Major overhaul of 2dFDR
• Can also use with existing SPIRAL IFU at
  Cassegrain 11 x 22" field, 0.7" /lenslet/fibre

4
Science Requirements

• High spectral resolution need R 10,000 to
  reach velocity precision 1 km/s
• High S/N require S/N 50-100 to get best velocity
  precision
• Good Sky Subtraction sky subtraction often to 1
  or better, sometimes 0.1
• High stability lt1/20 pixel over several hours,
  for velocity precision and sky subtraction.
• Uniform PSF few for velocity precision and sky
  subtraction.

5
Design Drivers

• To use VPH gratings, with their excellent
  efficiency and flexibility.
• To accommodate as many fully resolved fibres as
  possible
• To maximise the number of spatial spectroscopic
  resolution elements
• To have excellent optical performance for
  370nm-950nm.
• To allow use at R1500-10,000 at any wavelength.
• To have spectral stability better than 1/20 pixel
  over 4 hours.
• To minimise scattered light and ghosting.
• To be highly efficient.
• To use as few and as simple optical components as
  possible.
• To have uniform and well-sampled PSF
• Within this constraint, to have cameras as fast
  as readily achieved
• To allow use with the existing SPIRAL (IFU) front
  end.
• To do all the above within the limited budget and
  with minimal risk.

6
Dual Beam Schmidt Spectrograph
7
Spectrograph layout
8
Optical Parameters

• Schmidt Collimator
• Fibre slit 145 mm 392 (MOS) or 512 (SPIRAL IFU)
  fibres
• Field lens in contact with fibres
• f/3.15 (f/3.4 2dF prime focus FRD
  non-telecentricity
• 190 mm collimated beam diameter.
• Dichroic acting at 570nm
• Dispersers
• VPH gratings used at angles 0-47
• Schmidt Cameras
• f/1.3, giving 2.4 x demagnification.
• 3.2 pixels/FWHM MOS, 2.0 pixels/FWHM IFU
• Articulation angle range 0 - 94
• Detectors
• 2 x 4096 (spatial) x 2048 (spectral) E2V, 15?m
  pixels

9
Optical performance

• Optimised for use 370-580nm (bue arm) and
  560-950nm (red arm).
• Spot size lt7.8?m rms radius for all wavelengths
  and configurations,
• gt PSF uniform to few across field

30 ?m
IFU
MOS
10
PSF, FWHM, resolution

• Aberrations reduce FWHM!
• MOS PSF sampling 3.2 pixels/FWHM
• IFU PSF sampling 2.0 pixels/FWHM
• PSF uniform to few
• MOS PSF occupies at most 7 pixels
• True resolution better than FWHM indicates

11
VPH Grating Set
Name Blaze Wavelengths ??
Disp Resolution nm nm nm
nm/pix (MOS) 580V 450 370 to 580 210
0.1 1400 385R 700 560 to 880 320 0.16 1400 160
0B 400 370 to 450 80 0.038 3500 1500V 475 425
to 600 80 0.037 4000 1000R 675 550 to
900 120 0.057 3700 3200B 400 360 to
470 30 0.014 8500 2500V 500 460 to
600 40 0.018 8500 2000R 650 570 to
750 50 0.023 8500 1700I 860 670 to
880 60 0.028 8500 1700D 860 670 to 880 30 0.024
11000

• Grating set covers 370-950nm at resolutions
  R1400, 3500, 8500.
• R11000 possible at specific wavelengths using
  Dixon gratings.
• Resolution in IFU mode 1.6 times higher
• Other gratings (1400Z, 6000B ?) can be added

12
MOS Spectral resolution
Blaze wavelength ? sin? Dispersion ?
sec? Resolution ? tan?
13
VPH gratings
Peak efficiency 80-90 FWHM ??/? 0.42 vs (2048
? 15?m)/247mm0.12 Superblaze FWHM ??/? 1
14
Throughput

• Worst losses above spectrograph
• Peak total throughput 18(blue), 22 (red)
• Total throughput 10 at 400nm and 950nm

15
Comparison with 2dF

• At least 2-3 times better than 2dF
• More in red and at high dispersion

16
Sky Subtraction methods

• Traditionally poor with fibres
• Mean-Sky Method (MSM) 20-30 fibres are dedicated
  to sky. Target 1 accuracy (vs few for 2dF and
  6dF)
• Beam-Switching/ Cross Beam-Switching telescope
  nodded between object sky on 10-20 min
  timescales good to 1. Factor 2 S/N penalty.
• NodShuffle combine beam-switching with charge
  shuffling good to 0.1 but only 200 fibres.

• Mini-shuffle charge shuffle by few pixels to
  give blended ON/OFF pairs, good to 0.3, and can
  use all fibres.

17
NS on Real Objects
Nov 2003, 9.5 hours 2dF NS data, galaxies with
r22.5m

• Sources 50 times fainter than sky
• Reduced spectra have no sky residuals, Poisson
  limited
• Will reach this depth with AAOmega in 4 hrs

18
Signal/Noise estimates

• Signal/Noise for point sources, per Å per hour,
  in Dark/Grey/Bright time
• Anticipated median seeing lt1.3 with dome aircon
• S/N ? 2 for NS

19
AAOmega and A?

• AA? 2dF 6dF Sloan OzPoz VIMOS
  W-Field
• AAT AAT VLT VLT
  Gemini
• Mirror Diam (m) 3.9 1.2
  2.5 8 8 8
• Mirror Area (m2) 11.9 1.1
  4.9 50.3 50.3 50.3
• Spectro Efficy 0.4 0.15 0.3 0.3
  0.3 0.4 0.3?
• FoV (deg) 2 6 3 0.42
  0.50 1
• FoV (deg2) 3.1 28 7.1
  0.14 0.20 0.8
• No of objects 392 400 150 640
  132 560 4000?
• No of photons 1860 700 50 940
  1990 11300 60,000
• per unit time
• Time to cover 1 2.7 1.7 1.7
  8.2 4.2 1.3

20
Conclusions

• AAOmega high throughput, spectral resolution,
  stability, good sky subtraction
• Can do projects not possible with 2dF, or other
  instruments high quality spectra of faint
  objects over a wide field
• Physics of the Local Universe