IRIS TMT Week

1
IRIS TMT Week

• James Larkin
• Keith Taylor, Tim Davidge, Ian McLean, Eric
  Prieto, Mike Brown, Darren Erickson, Laura
  Ferrarese, Andrea Ghez, Glen Herriot, Tommaso Treu

2
Overview of Requirements
3
IFU Design Families
Focal Plane Feed to Spectrograph Detector
Image Slicer Fiber Bundle Lenslet Array
4
Design Description

• Existing AO-fed IFUs (OSIRIS, SINFONI and soon
  NIFS) are either lenslet (OSIRIS) or slicers
  based (SINFONI NIFS).
• Both are reasonable options for IRIS and we are
  following two design paths
• Slicer Keith Taylor and Eric Prieto
• More efficient pixel usage
• Lower detector noise
• Lenslet James Larkin
• Better image quality
• Easy to expand to many thousands of field
  locations.
• We are also designing an internal imager like
  OSIRIS.
• Optical design James Larkin

5
Slicer - Summary

• All reflective design.
• Each optical section contains 60 slicer mirrors
  that are 1.4mm wide and 84 mm long.
• Four sets of slicer mirrors provide the full
  120x120 effective field with spectral bandwidth
  of 20.
• Current design uses spectral stepping to achieve
  R4000. A non-stepping version will be
  investigated next which would reduce the number
  of slices to 30 per unit, or would reduce the
  spectral bandwidth to 10.
• Supports scales from 0.04 to 0.20 per pixel.

6
IRIS SlicerPrieto (LAM) Taylor (Caltech)
7
Lenslet Design

• Packing geometry makes it difficult to allow a
  large spatial width in the dispersion direction
  for a bandwidth of 20.
• OSIRIS supports both a 20 bandwidth mode with
  16x64 lenslets and a 5 bandwidth mode with 48x64
  lenslets both with R3800.
• 5 bandwidth is easily expanded to larger
  roughly square fields.
• Science Team was asked about spectral bandwidth
  vs. field of view and universally favored field
  of view.
• Lenslet design now focusing on 5 bandwidth
  spectrograph
• Slicer now investigating 10 bandwidth mode.

8
Wavelength Range

• Goal of 0.6 to 5 microns is extremely challenging
  for both lenslets and slicers
• Changes in pupil and slit diffraction with
  wavelength
• Need for a wide range of spatial sampling
• Differences in detector performance (read noise
  and dark current increase with long wavelength
  cutoff).
• Longer wavelengths prefer pixels with larger
  pixels to minimize some of the above effects.
• For lenslet design, Ive broken the design into a
  short wavelength (1-2.4 microns) and long
  wavelength arm (2-5 microns)

9
Lenslet Basic Parameters

• Short wavelength spectrograph 1-2.4 microns
• 4096x4096 detector (mosaic of 2048x2048) with 18
  micron pixels
• Plate Scales of 0.004, 0.009, 0.020
• Lenslet pitch 300 microns
• Spatial fields of view 136x136 gt 0.54x0.54,
  1.224x1.224, 2.72x2.72
• F/15 focus, physical sizes 1.19mm, 2.68mm,
  5.94mm
• Spectrograph Optics
• Camera focal ratio F/4.42 including corners
• Collimator focal ratio F/2.41 including corners
• Output is 18,500 spectra each with 400 spectral
  channels
• Long wavelength spectrograph 2.0-5.0 microns
• 2048x2048 detector with 27 microns pixels
• Plates Scales of 0.009, 0.020, 0.050
• Lenslet pitch 600 microns
• Spatial fields of view 68x68 gt 0.61x0.61,
  1.36x1.36, 3.40x3.40
• F/15 focus, physical sizes 1.34mm, 2.97mm,
  7.46mm
• Spectrograph Optics
• Camera focal ratio F/3.18 including corners

10
Lenslet Optics
Fold Mirror Lenslet Array
Filters
AO Focus
Reimaging Cameras
Reimaging Collimators
Grating
Spectrograph Collimator Mirrors (TMA)
Detector
Fold Mirror
Spectrograph Camera Mirrors (TMA)
11
5 Bandpass OSIRIS Spectra

• White Light (3072 spectra) Arc Lines
  (3072 spectra)

12
Design - Imager

• We have designed a diffraction limited camera
  based on a possible 4096x4096 Rockwell Detector
  with 9 micron pixels. 15 arcsec field.
• All Spherical
• 4 reflections
• For 25 nm of wfe, each mirror
• can have 6 nm of rms surface
• error. This is l/20 PV surface quality.
• Small spherical mirror is at well formed pupil.
• Transmission gt80 (85 filter, 98.5 per mirror).
  Detector TBD.

0.88 m
0.22 m
Filter Pupil Fold
NFIRAOS Focus
Detector
13
Trade Studies Mounting/Rotating

• Were assigned down-looking port of NFIRAOS
• Rotation of the entire instrument
• Decreases need for back focal distance
• High throughput
• Fixed gravity vector
• Requires active alignment
• to NFIRAOS

14
Trade Study Tip/tilt sensors

• Baseline is now 3 infrared tip/tilt stars.
• Originally, we hoped to use sub-arrays of the
  imaging detector(s) as quad-cells.
• Unfortunately, it now seems essential to be able
  to patrol the entire 2 arcmin AO corrected field.
• We are just beginning to look into deployable TT
  sensors (not part of contract).
• TT Sensors will also provide alignment to NFIRAOS.

15
Potential Use of Focal Plane

• Central IFU Field up to 3.4
• Four 15 imaging fields
• Equivalent to WIRC
• Leaves 90 of field for patrol region.

IFU
TT Stars
Imaging Fields
2 Patrol Field
16
OSIRIS

• Part of our team is still in the middle of
  commissioning OSIRIS (AO-fed IFU for the Keck
  Laser AO system) which definitely takes time, but
  is very valuable experience.

17
OSIRIS
OIII (500.7nm) Steps27 km/s
1

• 4c48.48Radio Galaxy
• (z2.343, 11.1 Billion light years)

2
3
4
1.6 14 kpc
HK image Carson et al 2001
18
TitanMoon of Saturn

• 0.02 arcs/lenslet 130 km
• ly2.1 microns

0.8
Surface
Stratosphere
19
Variability at Sgr A
First Laser Assisted Spectroscopy of GC ever.
0.32 2500 A.U.
Sgr A flare
Telescope Crossing
15 minutes per frame
20
Work Plan / Schedule

• IRIS Study
• 15 September 2005 Delivery of draft IOCDD
• 31 October 2005 Delivery of draft IFPRD
• 15 November 2005 Delivery of final IOCDD
• 1 December 2005 Delivery of final IFPRD
• 15 February 2006 Final Feasibility report
• 1 March 2006 Review meeting
• 14 April 2006 Revised Final Feasibility report

21
Cost Targets (ROM cost)Example Table
22
Cost Targets (ROM Cost)
23
Summary

• Weve had a late and slightly slow start.
• Settled on basic design parameters (local minima
  in instrument parameter space)
• Beginning to flesh out real optical components
• Mechanical design will start in early November.
  Wrapping metal around the glass.
• Gaining experience with OSIRIS.
• LGS spectroscopy of faint and complex objects
• AO backgrounds/ pupils/ modes
• Improved thermal modelling for extrapolation to
  IRIS
• Estimated Final Cost 20 Million not including
  TT sensors.