Technical Aspects of HyperSuprime

1
Technical Aspects of HyperSuprime
HyperSuprime gt1G pixels camera

• Yutaka Komiyama
• (NAOJ)

Suprime-Cam 80M pixels camera
2
How wide is the FoV of HyperSuprime?

• 2 deg? 1.5 deg? Not clearly determined so far
• Many key-points remains unclear how much weight
  can be loaded on the prime focus ? Can we grind
  some part of telescope structure ? Etc.
• Key Mitsubishi Elec. Co.
• Unfortunate Kick-off meeting was postponed since
  Mitsubishi persons had to concentrated on the
  recovery work from the earthquake occurred in
  Hawaii
• 1.5 deg FoV looks more probable
• Telescope modification is easier and minimal
• Scientific reason (FoV 10x FoV of Suprime-Cam)
• 2 deg FoV will be possible if following points
  are clarified
• Heavier instrument can be loaded on the prime
  focus
• Mounting procedure of prime focus unit to the
  telescope is established
• We therefore pursue two possibilities,
• baseline 1.5 deg, option 2 deg

3
HyperSuprime Design Concept

• FOV 2 options being considered
• 2.0 deg f 3.14 deg2 (0.5 deg f)
• 1.5 deg f 1.77 deg2
• Image Quality D80 lt 0".3 (0".22)
• Optimized Wavelength in red
• 600 nm lt l lt 1100 nm (l gt 400nm)
• FDCCD High QE in red wavelength
• Allow field curvature
• Allow some vignetting on the edge

(Suprime-Cam)
4
Comparison with Other Project
HyperSuprime Survey Speed 0.5 x LSST (2 deg
ver.) But we already have a telescope
5
HyperSuprime Technical Challenges

• Optics (Prime Focus Corrector)
• Optical design must be fit in the prime focus
• Production of large lens
• Mechanics (Prime Focus Structure)
• Lightening while keeping stiffness
• More accurate motion than current prime focus
  unit is required
• CCDs
• 200 CCDs are necessary
• Stable supply and lowering production cost are
  the key

6
1.5 deg version

• Telescope modification must be minimal
• Use current spiders, inner top-ring and top-unit
  exchange system with small modification.

New Prime Focus Unit
Inner Top-Ring
Spider
8m
7
1.5 deg corrector

• Size-constraints
• lenses must be lt f900mm
• some of central lenses must be lt f625mm
• Latest Optical Design
• Design by Dr. Nariai
• scale-down version of 2 deg version optics
• 6 lens corrector
• Silica (5) BK7 (1)
• primary lens diameter 900mm
• 3 aspheric surfaces
• Focal plane diameter 510 mm
• Plate scale 92.1 um/arcsec
• 1 pixel (15 um) 0.163
• Weight of glass 350 kg
• Adding ADC function is being
• considered

8
optical performance

• The diameter that encircles
• 80 energy inside D80
• (80 Encircled Energy Diameter)
• 0.34 _at_ 450 nm
• 0.26 _at_ 600 nm
• 0.27 _at_ 900 nm
• 0.26 _at_ 1100 nm
• The case for Suprime-Cam
• 0.30 _at_ 400 nm
• 0.18 _at_ 546 nm
• 0.31 _at_ 1000 nm
• (actual measurement)
• Seeing limited imaging guaranteed
• Vignetting
• Gradually degraded
• 80 at 0.75 degree edge

9
1.5 deg corrector alternative design

• Design by P. Gilligham_at_AAO
• Independent from Nariai design
• 6 lens corrector
• Silica (4), S-BSL7YS-NSL36 (2)
• primary lens diameter 900mm
• 2 aspherical surfaces
• Pair of cemented doublets
• functions as ADC
• Focal plane diameter 525 mm
• Plate scale 96 um/arcsec
• 1 pixel (15 um) 0.156
• Weight of glass 440 kg

10
1.5 deg corrector (alt) performance

• D80
• 0.45 _at_ g band
• 0.25 _at_ r band
• 0.24 _at_ i band
• 0.29 _at_ z band
• ADC function
• Vignetting
• Gradually degraded
• 80 at 0.75 degree edge
• Availability of special glass for ADC (S-NSL36)
  is the key
• Ohara Co. do not have large S-NSL36 blank
• Further improvement for the optical design would
  be necessary

1
11
Corrector Key Issues Availability of Bulk Glass

• Ohara
• Meeting next week
• Shinetsu Quartz Co.
• Worlds highest-quality quartz (fused silica)
  glass
• 250kg bulk available
• (two bulk can be jointed to one, e.g.,
  200kg200kg-gt400kg)
• Homogeneity 5ppm (200kg)
• Press-mold shaping up to f1300
• They have f1.3m furnace
• They can make 1.2m glass blank
• Production 3 month
• Meeting tomorrow

12
Make it possible with Canon

• Canon have interest on this project and start the
  investigation on optics
• Canon manufactured current prime focus corrector,
  and they have excellent experience
• They have fine polishing machine (CSSP) which
  handle up to f1.5m and can polish any aspheric
  surfaces
• Manufacturability of large corrector is being
  investigated seriously
• Surface shape measurement is most difficult and
  need development
• Production 2 years after every technical risks
  are cleared
• Key company for the success of HyperSuprime

13

• Canon
• like
• Subaru !

Paper Craft Subaru dome on Canons webpage
14
Coating

• MgF2 coating method being investigated
• 2 layer coating MgF2 MgF2(solgel)
• In collaboration with Osaka Institute of
  Technology
• Canon has technique of multi-layer coating
• Large chamber available

15
1.5 deg version
New Prime Focus Unit
Inner Top-Ring
Spider
8m
16
1.5 deg version
17
Prime Focus Structure Key Components

• CFRP structure
• Lightening while keeping stiffness is the key
  point
• Nikkiso Co. (CFRP vendor) has investigated and we
  received a report
• Positioning Actuator
• High driving force and precision control are
  required
• Harmonic Drive Systems Co. is considering the
  possibility
• Main Rotator
• Lightening high-precision movement are required
• Lightened cross roller bearing (by THK Co.) is
  available
• Linear motor drive (by Yasukawa Elec. Co.) is
  promising for rotational motion

18
Positioning Actuator/Rotator

• Actuator
• Slender rod
• Proto-types is designed in detail by Harmonic
  Drive Systems Co.
• Rotator
• Linear motor developed by Yasukawa Elec. Co.

19
Structural Analysis

• FEM analysis for key components are carried out
• Feed back to the mechanical/optical design

Gravitational sag of corrector
Stress analysis of Positioning actuator
20
2 deg versionPrime Focus Structure

• Current design
• Need to replace entire top-ring structure and
  spiders
• Strong hesitation to this plan
• May not be realistic

New Prime Focus Unit
New Inner Top-Ring (double ring structure)
8m
21
2 deg versionPrime Focus Structure

• 2 deg FoV will be possible if we clear following
  points
• Heavier instrument can be loaded on the prime
  focus
• Mounting procedure of prime focus unit to
• the telescope is established
• Then, current inner top-ring
• and spiders can be used

current inner top-ring
22
2 deg versionPrime Focus Corrector

• Latest Design
• 6 lens corrector
• Silica (5) BK7 (1)
• primary lens diameter 1.2 m
• 3 aspherical surfaces
• Focal plane diameter 672 mm
• Plate scale 90.6 um/arcsec
• 1 pixel 15 um corresponds to 0.166
• Weight of glass 680 kg
• No ADC
• (can be added)

Inner top-ring
23
2 deg versionOptical Performance
0 deg 0.25 deg 0.5 deg

• Spot Diagram
• The diameter that encircles
• 80 energy inside D80
• (80 Encircled Energy Diameter)
• 0.26 _at_ 600 nm
• 0.26 _at_ 900 nm
• 0.28 _at_ 1100 nm
• 0.36 _at_ 400 nm
• c.f. The case for Suprime-Cam
• 0.30 _at_ 400 nm
• 0.18 _at_ 546 nm
• 0.31 _at_ 1000 nm
• (actual measurement)
• Vignetting
• gradually degraded
• 80 at 1 degree edge

0.7 deg 0.9 deg 1.0 deg
24
Filter Exchanger

• Filter Exchange System
• draw bridge turn-table
• Available filters per night 4 or 5
• Filters can be replaced during daytime
• without dismounting PFU from the telescope
• Filter (including holder, mech.)
• 600800mmf 40 kg / filter
• Motor is embedded in the filter holder and
• the holder moves semi-automatically
• Exchange time 5 min
• Only at rest position (zenith)
• (necessary for secure filter transfer)

Turn-table
25
Filters

• Size f600800mm
• No vender to manufacture such large filters
• Asahi Spectra Co. or Barr Associates are the
  candidates
• Asahi will develop large chamber
• Barr has a capability to manufacture f500 filter
• for Pan-Starrs
• Mosaic filter is another option
• Narrow-band filters?
• requires more homogeneous transmission
• But vignetted area becomes large
• Cost gt 100k / filter
• Suprime-Cam equips with 8 standard filters
  (B,V,Rc,Ic,g,r,i,z)
• We probably have to choose one system, i.e.
  Johnson-Cousins or SDSS
• LSST, Pan-Starrs choose SDSS system

26
Shutter

• Roll type shutter
• Compact and thin design
• Fit in limited space at Prime Focus
• Proto-type is being designed
• Motor and control OK
• The material for the roll-sheet being
  investigated
• Min. exposure time 1 sec
• Accuracy of exp. time lt0.1

27
CCDs

• Fully depleted CCD (FDCDD) Merit
• High QE at long wavelength (gt60 _at_ 1um)
• Suppress fringes
• High yield
• Thinning process for back-illuminated
• device is not necessary
• High yield ( lowering cost) is important
• for HyperSuprime.
• If 20k/CCD, 4M for 200 CCDs
• Hamamatsu Photonics Co.
• Develop FDCCDs in collaboration with NAOJ
• Fabrication of back illuminated test device was
  very successful
• High QE (70 at 1um) was achieved
• Large format (2k x 4k) device is under
  development (close to final)
• 10 CCDs will be in hand next spring and installed
  on Suprime-Cam

28
CCDs

• Quantum Efficiency

29
CCDs 2k x 4k
2048 x 4096 (15 micron) device in process!
New Suprime-Cam dewar is waiting for 10 CCDs to
deliver. It will be tested on Subaru Telescope
next summer.
30
CCD package

• 4 side buttable package
• CCD is glued to AlN ceramic plate
• Flex cable comes from back side of the package
• 3 locate pins for the alignment on
• the focal plane

Front Back
guard block
31
Focal Plane
SDSS phot. camera

• Focal Plane
• Diameter 510670mm
• 120180 CCD packages installed (11.4 Gpixels)
• 0.16 / pixel sampling (for 15um pixel)

510670mm
32
Dewar
Suprime-Cam

• Extend the Suprime-Cam dewar
• From 10 CCDs to 180 CCDs
• 180 CCDs are screwed on the AlN ceramic plate
  and cooled down to -100 ?
• Cooler high power pulse tube cooler or GM cooler
  are candidates

33
Readout Electronics

• Readout
• 1 exposure 3 Gbyte
• readout time 30 sec (2ch readout / CCD)
• Analog electronics MFront2
• Subaru standard (MFront)
• Being developed to handle FDCCDs.
• To make it light and compact is important
• ADC compact, inside dewar
• Digital Electronics Messia5
• Already in use Subaru standard
• 24 Messia5 can handle 120180 CCDs
• (3Gbyte/exposure)
• Data transfer will be a bottle neck.
• 10 Gbps Ethernet is necessary.

34
Data Flow
35
Data Flow

• Up to 1Tbyte data will be produced per night
• Semi-automatic reduction pipe-line is required
• We have started a development (NAOJ, UT, KEK
  collaboration team)

M31 covered with 180 CCDs of HyperSuprime
36
Summary

• Mechanics
• need detailed discussion with Mitsubishi Elec.
  Co. ASAP
• proto-types of key elements are being designed
  and manufactured
• Optics
• Canon start the investigation
• CCD
• first 10 2kx4k CCDs will be delivered next spring
• Electronics
• looks O.K.
• Data Pipeline
• development started