Transcript PowerPoint Presentation - Caltech Optical Observatories

NGAO Calibration/Simulation Source
T. Stalcup, M. Pollard
Requirements
• Relay NGS and LGS point sources into AO image plane
• Wavelength range
– NGS 500 nm to 2500 nm
– LGS 589 nm or 594 nm laser line
• 120 arcsecond field
• Resolved and unresolved sources at NGS and LGS 85 km and 180
km conjugates
• Flat field and spectral line sources
• 30 nm rms wavefront error
– Estimated specification. Need to produce a calibration budget
• Turbulence generator
–
–
–
–
Optional at summit, decided this shouldn’t drive design
Ground layer at cal unit secondary mirror
11.5 km conjugate at cal unit primary mirror
More details later in optical design
2
Design Goals
• Minimize motion requirements using beam splitters and multiple
source points
• Include enough focus range for NGS/LGS sources for image
sharpening and IF dichroic compensation
• Fiber-fed sources to minimize heat and access in AO cold enclosure
• Simultaneous operation of NGS/LGS sources
• Minimize surfaces
– More surfaces requires tighter figure error specification
3
Design Status
• Concentrated on difficult part – the optical design
• Choice of source sizes, brightness, and wavelengths still in progress
– Very flexible design, can put anything in input plane for calibration unit
• Flat field source still under development
• Optomechanical design in progress
4
Optical Design
• Explored a few alternatives – refractive, TMA, etc.
– Too complicated or costly due to wavelength range and wavefront
quality constraints
• Chose an Offner relay
– All spherical surfaces
– Excellent image quality
– Pupil at mirror, not free space
5
Optical Design
250 mm
Pupil at
secondary
mirror
Primary
mirror
Fold down
to AO image
plane
NGS
conjugate
Flat field
beam splitter
LGS 85 km and 180 km
conjugates
NGS/LGS
beam splitter
Flat Field and
spectral line source
6
Calibration Unit Location
7
NGS Performance
nm rms
30
15
0.5
8
NGS Wavefront Error
9
LGS Performance
nm rms
90km conjugate
nm rms
30
30
18
16
6.5
3.5
180km conjugate
10
LGS Wavefront Error
90km conjugate
180km conjugate
11
Sources
• Point sources are fibers mounted in a fixed plate
• Same pattern for NGS and LGS
• Mixture of unresolved and resolved sources
– Unresolved fiber core is 9 µm, same as current Keck system
– Resolved fiber core is 400 µm, or 0.55 arcseconds, also same as
current Keck system
• Roughly 60 fibers total
• Include 589 nm notch filter in
NGS to prevent crosstalk
Unresolved
Resolved
12
NGS Source
• Detailed analysis of fiber coupling and throughput pending…
• NGS
– Tungsten Halogen or Arc Lamp
– Some electronic intensity control possible
– If electronic dynamic range is insufficient, would need motorized filter
wheel
– Need electronically controlled shutter for dark images during calibration
– Newport Oriel lamp source
• RS232 on/off and intensity control
• Lamp elapsed hours
• Available high stability controller, uses optical feedback to stabilize output
13
LGS Source
• DPSS lasers at 589 nm are now available
–
–
–
–
–
10 mW to 2 W output
$4k to $22k
Some electronic intensity control possible
Filter wheel needed if intensity control has insufficient dynamic range
Shutter not needed, can turn on/off electronically
• HeNe at 594 nm also an option
– Lower power, 2 mW
– $2k
14
Flat Field and Spectral Line Source
• Uses pupil stop in simulator to define beam
• Based on woven fiber optic backlight
– Compact, Efficient
• Still talking to manufacturer about spectral properties
• Use broadband tungsten halogen or arc lamp for flat field
• Need to choose spectral lamps – probably hollow cathode type
15
Alternate Flat Field Source
• If woven fiber panels are not suitable
– Pattern of several fibers and diffuser
– Integrating sphere
• Davinci requested uniformity is much less strict than Contour
requirements
– 10% vs. 0.2%
Diffuse lambertian reflector
Sources
Output plane
16
Astrometric Grid
• Aluminized glass plate with micromachined 100x100 grid of 3.6 µm
holes on 360 µm centers
– 5 milliarcseconds diameter, 0.5 arcsecond spacing
– Requirement is only for 80x80. Added some for margin at edges when
shifting, but if expensive per hole could reduce.
– Design very similar to current NIRC2 plate
• Backlit with either woven fiber optic or large core fiber with projection
optics
• On in/out stage with NGS point sources
• If flat field arm extended to image plane, could place there
– Still need in/out stage
– Possible aberrations from flat field beam splitter
• Current plan includes a rotation mount for grid
– Chosen to keep aberrations from system as constant as possible, but
may not be necessary
17
Turbulence Generator
• Phase plates near mirrors, two copies of phase distribution
separated by ~2-3 mm
• Place for 11.5 km phase plate, but it would need to be very large
– Wouldn’t fit in AO enclosure with top on
– Could be used for warm checkout of tomography
Ground
Layer
Wheel
11.5 km
conjugate
phase plate
250
mm
18
Optomechanical
19
Elevation Ring Clearance
Planar design
Increased clearance design
20
Optomechanical
• All components mounted to common baseplate except AO rotator
fold
• Slide in/out of elevation bearing on rails mounted to AO bench
• Use overhead crane to lift on/off AO bench
• Not sure if commerical mounts will provide wavefront error
performance and stability required
• Mike Pollard is now past interviews, DAVINCI, etc., and will be filling
out design for PDR
21
Questions?
22