Michiel van Noort Nagaraju Krishnappa Joerg Bischo · 2014. 10. 7. · Nagaraju Krishnappa Joerg Bischo Observing the Sun. Observing the Sun through the Earth atmosphere. Observing

Microlens SpectrographMichiel van Noort

Nagaraju Krishnappa

Joerg Bischoff

Observing the Sun

Observing the Sun through the Earth atmosphere

Observing the Sun in detail

Evolving on a timescale of 10s.

Understanding=Spectra

Evolution

How to observe a 3D data cube with a 2D detector?!

I Slice: Use time as the 3rd dimension (scan)I Narrow band imagerI Slit spectrograph

of Resolution, Signal to Noise and Cadence

I High spatial resolution → many slit-spectra

I High spectral resolution → many ”images”

I High Signal to Noise → long exposures

I Rapidly evolving → available time is limited (1-10s).

−→ A good compromise is difficult

I By eliminating the need to scan, 1-2 orders of magnitude canbe gained

I Problem: How to detect a 3D data cube with a 2D detector?

Mapping 3D −→ 2D

Making space for the 3rd dimension

Door number 1

I Make space for spectral dimension by shrinking pixels

I Disperse at a small rotation angle to the pixel grid

I Truncate using a narrow prefilter to avoid overlap

I 3D cube recorded in a single exposure

I De-magnification factor N: N2 spectral ”pixels”

Targets

To be useful we need:

I Critical sampling in image space

I High throughput (∼50%)

I Spectral resolution ∼200000

I Spectral range ∼ 4A. (∼ 350 pixels incl. prefilter, N ≈ 18)

I At least 100x100 image elements

I High frame rate: small image elements (fast CCD)

Instrument Concept

�������

�������

������������������������������������������������������������������������������������������

Microlens ArrayReimager Spectrograph

1. re-imaging optics

2. image reformatter

3. high resolution spectrograph

The image reformatter is the key experimental part of the system,the rest is ”standard”.

Proof of concept

I Instrument uses array of “dots” instead of slit

I Dots can also be created with pinhole mask

I Test of concept with pinhole array...

Test setup:

I No re-imaging

I Pixels ”shrunk” with mask with 22x22 pinholes of 25µm

I Prefilter 4.4A FWHM @ 6302A.

I ”Ordinary” spectrograph (SST/TRIPPEL)

Pinhole array masks almost all light −→ very inefficient ( 0.25%)!

Spectrograph test: Pinholes

0 500 1000 1500 20000

500

1000

1500

2000

−20 −10 0 10 20

X [arcsec]

−20

−10

0

10

20

Y[a

rcse

c]

+

+

+0.5

1.0

0 50 100 150 200 250 300 350 400

wavelength [pixel]

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Inte

nsi

ty[H

SR

A]

Microlens Assembly

Single microlens array was tested in 2000 (Suematsu et al.):

I Array of 50x50 600x600µm microlenses

I 10A FWHM bandpass filter

I 1536 x 1024 CCD

They found:

I Too much straylight → mask needed

I ”Very hard” to align microlenses and pinholes

Project abandoned (built Hinode instead)...

Single lens solution

I Single microlens with [pinhole] mask: pixels imagedon the grating

P1F1 F2

I Image constrast is large → spectralresponse is scene dependent

I Image constrast is wavelength dependent → spectralresponse wavelength dependent

Dual lens solution

Dual microlens array design to:

I Image pupil on the grating

I provide two planes to mask straylight

P1 F2F1 P2

Microlens array 1 images ”pupil” on pupil mask P1P1 imaged on the colimator plane P2.Primary pinhole mask inserted at F2

Modeling

I Critical sampling of the image: focal ratio degradation of afactor 2

I Diffraction effects dominate the microlens assembly

I Mainstream optical design packages do not work (Fresnelcondition violated)

I Propagation calculated by numerical evaluation of

E (x , y , z) =z

iλ

∫ ∫E (x ′, y ′, 0)

1

r2e

−i2πrλ dx ′dy ′

Dual Microlens Assembly

6

0.3

F=0.3

F=6

P F

0.017

0.017

L1 L2

Properties

I Sensitivity to image contrast

Low contrast High contrast

I High transparancy (70-80%)

I Low parasitic light (∼0.03%)

0 20 40 60 80 100Grating size at L=500 [mm]

0.0

0.2

0.4

0.6

0.8

1.0

Cont

aine

d lig

ht fr

actio

n

I Low sensitivity to surface errors

Properties Cont’d

I High sensitivity to angle of incidence: Pupil motion on thegrating ∼ 100mm/deg

I Incoming beam [almost] perfectly telecentric (<0.1 deg.)!

I High sensitivity to microlens co-alignmentI Displacement amplified by Lspec/FL2 (∼5000)!

Alignment crucial −→ Monolithic design

Coupling between exit beam speed and image element size

I Critical sampling: F/=2

I Pupil “apodization”: F/=2

Total beam speed-up: F/=4Small CCD pixels −→ Additional scaling speedup

Prototype

Monolithic design:

n=1.45709

Mask

Rear viewFront view

mµR=2039 mµ

R=

87

mµ6690 mµ196

mµ

70

mµ325

mµ325

I Thick substrate (6.5mm)I Maximum feasible sag ∼ 15µm

I Quality of second lenslet array must be highI Secondary mask in spectrograph focus

I Alignment error < 1µm (array 42× 42mm → 0.01”)

Prototype manufacture

High precision −→ Fraunhofer Institute for Applied Optics Jena

I Front: Lithography + reactive ion beam etching

I Back: Reflow lenslets

I Mask: black chromium

Prototype layoutPrototype layout

Prototype testing

I Delivery in November 2014I To be tested

I Front-back ML alignmentI pupil co-alignmentI transparencyI Contamination (crosstalk + straylight)I pupil sensitivity to constrast

Lab setup:

42x42mm

500mm 310mm 310mm

25x25mm

500mm

3750mm

Spectrograph properties (TBD)

I ”Normal” spectrograph can be used

I Projected grating size ≥ 50x50mm

I Smaller FOV −→ Faster spectrograph

I F-ratio may need to be as low as 5 (possible?)I Fast beam: short spectrograph at high order

I Small FSR allows up to order 1500I Large blaze angle gratingI Increased sensitivity to angle of incidence

I Effects of the non-uniform illumination?I Closely packed multiple identical modules (compact)

I Transmission spectrograph possible?

Field splitter + re-imaging optics

I Larger FOV → multiple modules

I Multiple modules → Field splitter?

I Control of beam telecentricity?

I Large magnification → Low light levels (stray-light problems)

F

x 4

PF

x 5

F

x 4

PF

x 5

Still to come

I Lab tests (Q1 2015)

I Re-imaging optics (Q1-Q3 2015)

I Telescope test (Q3 2015)

I Field splitter (2015-?)

I Spectrograph (2015-?)