![Page 1: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/1.jpg)
Laboratory Emulation of Observations from Space
Roger Smith -- Caltech
Jason Rhodes, Suresh Seshadri -- JPL
Science with a Wide-field Infrared Telescopes in Space, Pasadena, Feb 13, 2012
![Page 2: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/2.jpg)
Previous culture, friendly collaboration but somewhat separate Pasadena, 2012-02-013 Laboratory Emulation
2
Requirements
Technologists study detectors
Astronomers plan missions
Publish
Measure
Discuss
Publish
Discuss
Simulate
Simplify
![Page 3: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/3.jpg)
Different paradigm needed
• Simulation is good to evaluate sensitivity of science to a given effect but is not sufficient. – At best, it is only as good as our understanding. – Simulation is often a simplified version of that.
• NASA “Technical Readiness Level” is not intrinsic to a component or technology. – It is not just environmental testing !! – Must meet requirements of the specific science case, using a
faithful model of the component or system, … including the intended operating modes, control software, timing, calibration and data processing.
• We need emulation not simulation. – Hardware not (just) software !
Pasadena, 2012-02-013 Laboratory Emulation
3
![Page 4: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/4.jpg)
Pasadena, 2012-02-013 Laboratory Emulation
4
2008-03-31
4
Typical Lab emulation
Reimage thousands of very accurate apertures onto detector using optics producing radially symmetric PSFs over 40mm x 40mm detector so that good statistics obtained over full detector.
LED
Stationary 6” integrating sphere, 2.5” exit port, 3 input ports.
Stable flux.
Mask on motorized 6 axis stage;
50nm res.
1µm min feature
Detector With rotation about optical axis Fixed
relay optics
Pupil mask,
rotatable
Analysis pipeline
Key feature: • Make “errors” (pointing, PSF, flux, etc) smaller than
required by mission so sensitivity can be verified.
![Page 5: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/5.jpg)
Why practice in the lab? …in mission definition phase.
• End to end test of real hardware is makes proposal more credible. – Numerical simulations are most useful for testing sensitivities, – Are we asking the right questions?
• Verify accuracy / precision limits.
• Develop/test mission requirements. - pointing, PSF quality/uniformity, image scale, thermal stability, etc
• Test how well calibration methods work. New ideas?
• Improve TRL: detectors, electronics, algorithms.
Pasadena, 2012-02-013 Laboratory Emulation
5
![Page 6: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/6.jpg)
Important “cultural” benefits
• Builds credibility of team as well as the technology.
• Gets scientists and technologists working together at the start.
• Focuses on the problem first and technologies second.
• Build analysis algorithms, expertise, and test facilities from start.
Pasadena, 2012-02-013 Laboratory Emulation
6
![Page 7: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/7.jpg)
Examples
To be presented:
• WFIRST: undersampled NIR survey for Weak Lensing
Other recent or planned:
• ELEKTRA: all sky NIR search for transiting planets
• FINESSE: NIR spectroscopy of transiting planets.
• ZEBRA: Zodiacal and Extra Galactic Background light.
• SPRITE: AO assisted high precision astrometry at south pole.
Pasadena, 2012-02-013 Laboratory Emulation
7
![Page 8: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/8.jpg)
SHAPE MEASUREMENT FOR WL Lab emulation of WFIRST
Pasadena, 2012-02-013 Laboratory Emulation
8
![Page 9: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/9.jpg)
Detector issues for WL
• Detector may imprint different ellipticity on galaxies and stars used as PSF calibrators.
• Detector must not prevent accurate recovery of PSF from undersampled data.
Pasadena, 2012-02-013 Laboratory Emulation
9
Hazards for CCDs ?... • CTE and its rate of change with radiation damage. Change in CTE with signal • Row-column difference in pixel size variability. • Pixel size dependence on signal? • Transient response on video. • Image persistence ? • Intrapixel response variation
Hazards for IR arrays ?... • Interpixel capacitance. • Reciprocity failure and linearity. • Image persistence when dithering. • Row-column difference in pixel size variability. • Pixel size dependence on signal. • Transient response on video.
![Page 10: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/10.jpg)
Some test objectives
• Show how well systematic ellipticity errors due to detectors can be controlled when emulate undersampling, object shape and intensity distributions.
• Same….versus image scale => mission requirements.
• Measure distribution of errors => detector requirements.
• Test effectiveness of calibrations.
• Validate analysis methods: – dither and image recombination methods. – Calibration methods.
• Test how stars and galaxies and stars are affected differently by detector behavior.
• (eventually) test correction of PSF variation (vs. rate of change)
Pasadena, 2012-02-013 Laboratory Emulation
10
![Page 11: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/11.jpg)
Key system requirements
• Exquisite image quality over full detector area (40mm) • Achromatic (400nm – 1.7µm) • Focal ratio ≥ f/8 • Low ghosting and scattered light. • Well baffled; flat illumination; flats at any time. • High resolution lithographic input masks • Precise pupil: circular or elliptical, rotatable • Low image motion (<20nm rms) • Stable lamps; wide intensity range. • Test automation • Computing hardware & analysis pipeline (144 core, 50TB)
Pasadena, 2012-02-013 Laboratory Emulation
11
![Page 12: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/12.jpg)
f/8 Offner relay
• Avoids chromatic aberrations. • To reduce already small aberrations
choose longer instrument, thus narrower angles.
• Size limited by mirror costs. ($70K) • λ/100 surface quality to preserve
strehl ratio >99.6% !!!
Pasadena, 2012-02-013 Laboratory Emulation
12
Input mask
M2 RoC=0.75m, Φ=95mm
RoC=1.5m Φ=460mm
150mm fold (upward)
Down looking dewar, on precision turntable bearing.
Detector
![Page 13: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/13.jpg)
13
Baffle 1 Baffle 2 Baffle 3
Two input sources, beam combiner not visible
Integrating spheres on slides to allow access to kinemtically mounted masks
Motorized 6 axis flexure stage (1µm precision over 4mm)
Primary mirror in sling mount
Secondary mirror mount NIR dewar on precision turntable with detents, above beam fold with active steering
Thermally controlled bench
Not shown: Vibration isolation, Insulating covers
![Page 14: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/14.jpg)
How ?
• Masks contain a range of known shapes and rotations. – Do we get back what we put in?
• Rotate and translate input mask. – Do different pixels give same results? – Partition ellipticity between mask and detector+optics.
• Rotate detector on pedestal. – Partition ellipticity between detector and optics+mask
• Rotate elliptical pupil. – Is Δellipticity what we expect? – Does Δellipticity depend on angle?
Pasadena, 2012-02-013 Laboratory Emulation
14
![Page 15: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/15.jpg)
4K wide CCD (fully depleted LBNL p-ch)
15
http://www.astro.caltech.edu/~goodsall/spots/presentation/ghosts2.jpg
Quad spot feature in image is a viewing resolution artifact,
not present in magnified image
S. Seshadri, JPL
3µm pinholes, f/11 at 880nm
10µm spots are undersampled
![Page 16: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/16.jpg)
Ellipticity vectors, after random dither and IMCOM
• Ellipticity ~1% across large portions of the FOV, but not everywhere • Features at the corners and in a single central column are NOT due to
reconstruction – Highly elliptical edge effects are present even in the undersampled data – IMCOM is detecting detector defect causing long single trails in central column feature
16
S. Seshadri, JPL
![Page 17: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/17.jpg)
Ellipticity map vs. pedestal rotation (CMOS imager)
Key features
• Ellipticity is changing at ~twice the pedestal angle.
• Mean is not constant.
• Nearly uniform across image – low distortion across the field
• Noise is low
17
Ellipticity map (length = magnitude)
(tilt = angle)
Histogram (magnitude)
0o
Pede
stal
an
gle
45o
90o
180o
http://www.astro.caltech.edu/~goodsall/spots/2011-09-23/rotationcomparison.png
![Page 18: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/18.jpg)
Applications of Precision Projector
• Shape measurement – NIR and optical. – Wide achromatic FoV can be used to test effects of CCD CTE.
• Flat fielding precision
• Photometric stability – pointing jitter simulation – weak signal against strong background (use beam combiner)
• Precision astrometry. – motion of individual stars vs ensemble.
Pasadena, 2012-02-013 Laboratory Emulation
18
![Page 19: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/19.jpg)
Abstract
A substantial motivation for conducting wide-field surveys from space is the desire for increasing measurement precision, and calibration accuracy. In many cases the limits of performance of detectors, optics, calibration techniques and even image processing algorithms are being pushed, while budgetary constraints drive mission designs to coarser sampling of the PSF requiring recombination of dithered images to recover the optical resolution. We describe current and (possible) future examples of laboratory experiments conducted by Caltech/JPL, which are designed to build confidence that the planned detector technology is capable of delivering the necessary control of systematic errors. We note how these experiments bring about a paradigm shift by promoting close collaboration between scientists and engineers to address real world issues, during the concept development phases. This leads to stronger proposals based on experimental confirmation that calibration methods and algorithms will (still) work, and that specifications for space mission parameters such as PSF sampling, pointing jitter, observing cadence, calibration plans, and data processing needs are neither over or under-specified. A number of these issues can have profound effects on the technical or fiscal viability of a mission.
Pasadena, 2012-02-013 Laboratory Emulation
19
![Page 20: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/20.jpg)
PLANETARY TRANSITS Lab emulation of ELEKTRA
Pasadena, 2012-02-013 Laboratory Emulation
20
![Page 21: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/21.jpg)
Setup
<Labelled picture of hardware>
<Explain how image motion is controlled>
Pasadena, 2012-02-013 Laboratory Emulation
21
![Page 22: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/22.jpg)
Allan Deviation curves
• Briefly explain analysis and what these curves are.
• Show results for range of intensities.
• Point to hint of floor.
Pasadena, 2012-02-013 Laboratory Emulation
22
![Page 23: Laboratory Emulation - IPACLaboratory Emulation Pasadena, 2012-02-013 4 2008-03-31 Typical Lab emulation Reimage thousands of very accurate apertures onto detector using optics producing](https://reader034.vdocument.in/reader034/viewer/2022042808/5f834584e55dce791d3d0e25/html5/thumbnails/23.jpg)
Test for pointing sensitivity
• Show centroids vs time
• Show allan deviation curves with and without PCA applied.
Pasadena, 2012-02-013 Laboratory Emulation
23