Science Goals (ref: NSF proposal) Optical Design (ref. “Optical Design”)

Design Review Spartan IR CameraE Loh, Physics-Astronomy Department, Michigan State University

East Lansing, 22 May 2001

1 Science Goals (ref: NSF proposal)2 Optical Design (ref. “Optical Design”)

–Optical alignment (ref: “Alignment” & “SOBER”)

3 System Design & Electronics (ref. “Electronics”)4 Mechanical Design (ref. “Mechanical Design”)

5 Budget & Schedule (ref. “Budget & Schedule”)

22 May 2001 DR SOAR Spartan IR Camera 2

The Team

• Jason Biel, technician– Measurements for vacuum design

– Electronics designer & technician

• Mike Davis, graduate student– Optics

• Owen Loh, Okemos High, volunteer– Finite-element analysis

– Drafting

• Tom Palazzolo, head, Phys-Ast shop– Mechanical shop, design advice, contact for mechanical designers &

job shops

• Jack Baldwin, Brooke Gregory, Ron Probst, Dan Edmunds, Phys-Ast EE, advisors

• E Loh


1. Science Goals

Tip-tilt corrected imaging in the J, H, & K bands

• To cover the wide, corrected field (5’)• To resolve FWHM of median seeing (0.15–-0.23”)• To resolve high-contrast features at the diffraction limit

(0.08” @H & 0.11” @K)


0 0.05 0.1 0.15 0.2 0.25 0.3q [arcsec]

0

5

10

15

20

ytisnetnI

[cescra

-2]

K-tt

H-tt

J-tt

V-ttK

V

Figure 2 Point-spread functions for median seeing withtip-tilt correction (solid lines) and without correction(dashed line; K and V bands only). Shown in the insert isthe tip-tilt corrected PSF with top quartile seeing.

Point-spread Function with Tip-Tilt Correction

• Point spread function is not a gaussian• Diffraction spike

Bandmedian top 25%

J 0.05 0.15H 0.12 0.30K 0.28 0.50

Strehl Ratio

0 0.1 0.2 0.30

10

20

30

40KHJ

V

Median seeing

Top quartile


Image Width

• Sub 0.5” images w/o tip-tilt• 0.15-0.23” images w tip-tilt• Telescope optics preserves images

0.5 1 1.5 2 2.5wavelength [m]

0

0.1

0.2

0.3

0.4

0.5

MH

WF

[c

es

cr

a]

Diffraction

TT/top quartileTT/median

Natural/top quartile

Natural/median

Figure 2 Image size (FWHM) for natural seeing, for tip-tiltcorrection, and for diffraction (from top to bottom). For medianseeing with tip-tilt correction, the points show the effect of telescopedegradation with the AOS specification and the Raytheon structurefunction presented at CDR.

Telescope degradation. Goodyear CDR


2. Optical Design

• Concept• Image Quality• Tolerances


Optical Concept

• Requirements– Large number of pixels [ 2 x 5’ / 0.08” = 7500 pixels ]

– Large telescope image [ 5’ x 4.2m x 16 = 100mm square]

• Rockwell 2048x2048 HgCdTe detector– 4 detectors & 7500 pixels two plate scales

• Reflective optics large telescope image• Off-axis collimator & camera mirror

– Parent design: two paraboloids

• Perfect image for 1:1 & small field

– Real design for change in plate scale

• Adjust conic constants, distances

• Field flattening lens


Design

• Four 20482 detectors• Two plate scales: 0.08 & 0.04”/pixel• 20 filters near pupil• Focal plane mask

– coronagraphy

– spectroscopy


Image Quality: Spot Diagram

• 9 Field points in a grid. Corners are corners of 4 detectors.• H band

f/11

f/21

Airy disk


Image Quality: Strehl Ratio

• 9 Field points in a grid. Corners are corners of 4 detectors.• Strehl is very high for diffraction sampled cases, f/21 in H and K

bands

Sampled for diffraction limit


Tolerances

• Error budget– Loss of Strehl of ~0.07mag

• Alignment• Manufacturing


Alignment Tolerances

Alignment Tolerances of the Optical ElementsElement

x y z x y zmm mm mm mrad mrad mrad

1 Window NA NA NA NA NA NA2 Focal surface 1.00 1.00 0.30 6.38 6.38 NA3 f/21 collimator 0.81 0.28 0.19 0.30 0.52 6.104 f/11 collimator 0.41 0.15 1.01 0.21 0.42 3.145 Fold mirror 1 NA NA 0.69 0.17 0.26 NA6 Filter NA NA NA NA NA NA7 Lyot stop NA NA NA NA NA NA8 f/21 camera mirror 0.67 0.25 0.40 0.26 0.52 4.809 f/11 camera mirror 0.27 0.17 0.49 0.17 0.23 1.86

10 Fold mirror 2 NA NA 0.16 0.31 0.47 NA11 Lens 0.48 0.32 0.33 2.27 11.34 NA12 Detector plane NA NA 0.03 0.64 0.64 NA

Positional Tolerance Angular Tolerance

1mil over 6in

6mil


Manufacturing Tolerances• Focal lengths are absorbed in

focus

• SORL can manufacture conic constants

• Surface irregularity– Peak-to-valley is /16 to /4.

= 633nm


Alignment with SOBER• Align at room temperature with

point source, SOBER, & CCD

• SOBER

– f/16 beam

– Move SOBER & shift stop to mimic pupil at 10m

– z stage mimics curved focal surface of telescope

– Tolerances 1mm & 1º

– Image in IR? TBD

Soar Beam Simulator

LED & pinhole

Lenses

R- stage

ISB surface

Sliding stopz stage


Alignment Indicator

• Intensity of 9 field points indicates error

Y-decenter of collimator 0.34mm X-tilt of fold #1 of 0.2mrad

- 1 - 0.5 0 0.5 1

- 4

- 2

0

2

4

DygrenEcnE@%D

80. 80. 80.

70.7 81.5 70.7

73.5 81.6 73.5

- 1 - 0.5 0 0.5 1

- 4

- 2

0

2

4

DygrenEcnE@%D

80. 80. 80.

70.7 81.5 70.7

73.5 81.6 73.5


Test of Alignment

-20 -10 0 10 20 30

-20

-15

-10

-5

0

5

10

E-EH

dengilaL@%D 1 2 3

4 5 6

7 8 9x-position of collimator; wrong

y-tilt of lens; right

Defect: I7<I9

Defect: I5<I8

x-tilt of fold #1


3. System Design & Electronics• System• Electronics• Software• Motors• Vacuum


System Design

PC

Stages

Detector

Data ArchiveObserver Telescope Control

Pressure Sensor

Umbilical

Camera Controller

Motor Controller

Legend

Camera Controller

Camera Controller

Camera Controller

Detector

Detector

Detector

DeviceNet

NI 6533

RS232

Fiber optic

Ethernet

RS232

RS232

On camera

In vacuum

In control rack

Elsewhere

Camera Controller

Custom

Commercial


Umbilical Card• Provenance

– CCD system

DRV11 interfaceNI 6533 interface

Laptop-type power supply

Master clock

Test pod

FIFO

NI 6533

Camera card

Serializerdeserializer

Fiber-optic tranceiver

Logic Analyzer

Existing CCD Softwareon Alpha

In FPGA

One of 4 channels shown

For debugging


Camera Card

• Provenance: CCD camera• 4 analog channels for 4

quadrants

Laptop-typepower supply

Phase-lockedloop

Test pod

Umbilical card

Serializerdeserializer In FPGA

Detector

Fiber-optic tranceiver

Amplifier &16-bit ADC(2 12-bit ADC)

Buffer

Fixed voltages(digital pot)

Timer & clock generator

TemperatureDiodes

Logic AnalyzerInstruction


Umbilical Card

• 3U 100160 mm• Tested w/ CCD software

160mm

NI 6533Fiber opticto 4 detectors

FPGA

7V in

To logic analyzer

To existing computer


Camera Card• 3U 100160 mm

• Low crosstalk– 5-mil between signal & ground layers

• Delivery expected in 2 weeks

FPGA

Signal chains

Fiber optic

7V in

Flex cables to detector Neck betweenanalog & digital circuits

• 2.5s/pixel• 4 channels• Power: 1.4W


Noise

• Detector noise is about 10e–; noise on amplifier glow is 5e–.• Electronics noise is 6e–.• Coupling from a saturated channel is about 2e–.• Coupling from clocks on cable is large.

– Sampling signal must wait 100ns after clock transition.

Source Noise e- SegregatedDetector ~10Glow w 2.5s read 5Electronics 5.7

Opamp 4.5FB resistor 3.5Offset reference 0.6Bias gate 0.06

CouplingSaturated signal on cable 0.2Clock on cable 180 yesADC (difficult to estimate) 2 ? yesSaturated signal on card 1.3 ?


Detector Card• Card butts on 2 sides• Connects to camera card with 5 flex cables, which are

thermal resistors.• 3 layers with 5-mil G10.

Electrically isolated strapsto nitrogen dewar

Flex cables

Detector

ZIF socket


Software

• Functions [copied from Optical Imager]– Control detector

– Scripting

– Communicate with motors

– Communicate with telescope control system

– Communicate with user

• ArcView– Used for all SOAR instruments

– CTIO will debug ArcView with Optical Imager, the commissioning instrument

• LabView, “visual programming”– Independent of hardware obsolescence is obsolete

– Self documenting

– Easy to do. ArcView costs < 1 man-year


Software Tasks

• Design– Use commercial parts with LabView drivers

• Modify ArcView– Computer send commands and receives data from camera

controller through NI 6533 card.

• Replace Leach controller & driver with NI 6533 card.

• Our card has a 4k sample FIFO– 0.6ms margin for 4 detectors reading simultaneously

– Write software for summing pictures

– Change software for formatting picture

– Change motor controls

– Add temperature & vacuum sensing


Motion

• Phytron stages DT-90 & MT-85– Vacuum compatible

– Stepper motor

– Indexing switch

– Limit switches

– Open loop; controller stores position

• Controller– RS232 to computer

– LabView

– Heat

• Shutoff power? Cooler?


Vacuum Measurement

• Granville-Phillips ion gauge– Computer readout via DeviceNet

– LabView

– 12W; need to shut off


4 Mechanical Design• Cryogenic optical box

– A-frame attachment to vacuum enclosure

– Analysis of flexure

• Vacuum enclosure– Analysis of stress

– Transfer of forces from A-frame to instrument mounting box (ISB)

• Mechanisms using warm stages– Layout

– Proof of concept

• Flexure

• Heat load

• Operating temperature of stage & optics


Cryogenic Optical Box

• Symmetric box having two plates equidistant from optics.– Gravity vector is in plane.

– Optics supported by both plates.

• Torque perpendicular to plates

• Box is attached near focal surface of telescope– Rotation of optical box

causes no boresight error.


A-frame Attachment• Connect cold optical box to warm vacuum enclosure

• Complies with shrinkage of optical box– Web weak in z

• Hold box w/o sag– Web strong in x & y

• Heat load is 0.7 W for 4 A-frames.

G10 ring

G10 web

Al leg

Section removedfor clarity

Bolt tooptical box

Bolt to warmvacuum enclosure

Weakfor thermal compliance

Strongest; max sag:14 or 0.04”

Safety stop


Rotation of Optical Box• Gravity parallel to mounting plate.

(Causes boresight error)

• First approximation– Optical box rotates 40rad as a

unit– Sag is 14 at telescope focus.

• More precisely– Error is greater for gravity

perpendicular to mounting plate.– Rotation within box is 2.3rad

peak-to-peak– Boresight shifts 0.007”.

34–46rad0–155rad


Vacuum Enclosure• Aluminum plate, mostly 1/2”• Max stress is here

– Max is tensile strength / 2.2.– Code for pressure vessels is 3.5.– Is this OK?


Transfer of Forces to Bolts on ISB• Does the vacuum enclosure transfer forces between the A-frames and the

bolts on the instrument mounting box (ISB) without sag? Yes. Sag is 2.

Bolts toA-frames

Bolts to ISB

Sides of vacuum enclosure

Optical box


Mechanisms

• Two filter wheels– Loose tolerances

• Focal-plane mask– 300 along optic axis, 18 in transverse direction

• Collimator insertion– Tilt 5rad (1”) as instrument turns for boresight with tip-tilt sensor

• Camera mirror insertion– Tilt 5rad as instrument turns

• Rotate lens-detector by 112.7±0.6mrad– Tilt 0.2mrad (30 over 150mm)

• Move lens-detector assembly for focusing

Difficult


Layout of Mask & Filter Wheel• Load is balanced Easy to meet tolerances.• Phytron DT-90 rotational stages

– Integrated stepper motor, indexing switch, limit switch– Spring constant 2rad/(N-m). Wobble is ±15rad (Clarification needed.)

100 200 300 400

-200

-100

100

200

DT - 90

100 200 300 400

-200

-100

100

200

DT - 90

Vacuum enclosure

Optical box

Rotation stage

Mask wheel Filter wheel


Layout of Mirror Insertion

• Mirrors must be balanced to meet 5rad tolerance.

100 200 300 400

-200

-100

100

200

Mirrorout

CWout

CWin

DT - 90

Mirrorin 100 200 300 400

-200

-100

100

200

Mirrorout

CWout CW

inDT- 90

Mirrorin

Vacuum enclosure

Optical box

Rotation stageMirror

Counterweight

Background mirror

f/21 collimator f/11 camera


Proof of Concept: Insert f/21 Mirror

• Requirements. Cold mirror — warm stage — cold optical box– Support with tilt < 5rad

– Keep mirror cold

– Keep stage warm

– Minimize heat load

– Comply with thermal expansion

• Precepts– Balance load

– Use G10 A-frames to control conduction & comply with thermal expansion

– Shield stage from cold to control radiation

– Allow stage to absorb radiation from warm vacuum enclosure


Mirror Insertion

f/21 collimator

Counterweight

DT90 rotational stage

Bracket attachesto optical box

4 A-frames

4 A-frames betweenstage & bracket (hidden)

Center mass


Results for f/21 Insertion

• A-frames have 1x1x5mm legs.• Balance within 1mm.• Wrap stage in 10 layers of aluminized mylar.• Results

– Conduction is 170mW

– Tilt is 2rad; tolerance for boresight alignment is 5rad.

– Sag with mirror vertical is 8; tolerance for internal alignment is 0.8mm.

– Sag with mirror horizontal is 4; tolerance for focus is 15.

– Temperature of mirror is 88K.

– Temperature of stage is 2K below ambient. (Area of radiator is 10% that of the stage.)


5 Budget & Schedule• Budget• Contingency• Descope• Risk to budget• Schedule


Budget• Not allocated or charged: Majority of electrical engineer, mechanical engineer,

project management, drafting (done so far), and finite-element analysis.

SUMMARY BY WBSWBS ITEM M&S CONTINGENCY TOTAL

%1 MECHANICAL 218577 27 58818 2773962 OPTICS 188683 34 63729 2524123 DETECTOR 317263 5 16045 3333094 SOFTWARE & COMPUTERS 49456 59 29021 784785 INSTALLATION AND COMISSIONING49653 23 11643 612966 SUPPORT EQUIPMENT & SUPPLIES38822 22 8397 472207 MANAGEMENT, REPORTING, & DOCUMENTING50610 13 6777 57387

TOTAL 913065 21 194431 1107496

Major ItemsWBS Item Total1.3 Optical Bench, dewar, enclosure 122,5181.4 Mechanisms 147,9082.5 Collimator mirror 63,7462.6 Camera mirror 75,0353.1.1 Detector 250,0003.2 Electronics 52,8944.1 Software 48,657


Contingency vs Remaining Tasks• Tracking of tasks since budget of Aug 2000

– Electronics design is 17% over budget. ($5k of $29k)– Design of telescope simulator is 65% over budget. ($5k of $8k)– Optics design is 31% under budget. ($6k of $19k)– Overall budget dropped $100k because mechanical design firmed up, optics shortened, and mirror quotes dropped.

• Contingency is 36% of remaining tasks.

CONTINGENCY AS FRACTION OF REMAINDERWBS ITEM

% % AMOUNT1 MECHANICAL 27 59322 292 37 591102 OPTICS 10 18257 6753 41 704813 DETECTOR 89 282777 (5129) 32 109164 SOFTWARE & COMPUTERS 2 835 360 60 293815 INSTALLATION AND COMISSIONING0 0 0 23 116436 SUPPORT EQUIPMENT & SUPPLIES49 19028 (3552) 24 48457 MANAGEMENT, REPORTING, & DOCUMENTING0 0 0 13 6777

TOTAL 42 380219 (1276) 36 193155

CONTINGENCYENCUMBERED


Descope

• Descope 2nd plate scale, J, H, K, Ks filters only, spectroscopy & coronagraphy.

• Descope will be treated as contingency.– Descoped items will be added as contingency allows.

– Possible formula: spend if

Budgeted Contingency > 1.5 Actual Contingency

WBS M&S Contingency TotalTotal 128439 46958 175397Second f/ratio 83239 35658

2.5 Collimator mirror 21249 106242.6 Camera mirror 25012 12506

Mechanisms 36979 12528Spectroscopy & Coronagraphy 13200 3300

2.13 Grisms 13200 3300Spectral filters 32000 8000

2.9 Defer 8 of 12 32000 8000


Risk to Project• Number of risks covered

– A big item is $100k.

– Labor for optical box, mechanisms, enclosure is $70k with $30k contingency

• Drafting: 3 mo.

• Internal shop: 7 mo.

• External shop: 1 mo.

• Technician: 6 mo.

– Contingency, $193k, covers 2 big risks

– Descope, $175k, covers 2 big risks.

• Descope & contingency remaining tasks of descoped instrument


Schedule Overview


Detector

• Multiplexer & engineering-grade device delivered.• Long slack time before science-grade detector is needed.

Task NameDetector

Rockwell multiplexer

Rockwell engineering device

Rockwell science-grade device

or

xer 1/19

Rockwell engineering device 11/

Rockwell science-grade device

Qtr 4 Qtr 1 Qtr 2 Qtr 3 Qtr 4 Qtr 1 Qtr 2 Qtr 3 Qtr 41999 2000


Electronics• 7 mo. slack

Task NameElectronics

Design & fabricate electronics

Computer board

Design

Fabrication

Debug

UmbA (Old cam & DRV)

UmbB (New cam & DRV)

UmbC (Old cam & 6553)

UmbD (New cam & 6553)

Camera board

Design

Fabrication

Debug

CamA (Emulate old cam)

CamB (One quad)

CamC (4 quads)

Dewar cable

Design

Fabrication

Debug

Modify test dewar

Test engineering detector

Fix problems

s

s 72%

d 85%

gn 2/7

Fabrication 3/27

Debug 23%

UmbA (Old cam & DRV) 4/6

UmbB (New cam & DRV) 1/31

UmbC (Old cam & 6553) 1/31

UmbD (New cam & 6553) 2/14

d 45%

gn 4/27

Fabrication 1/3

Debug 0%

CamA (Emulate old cam) 1/17

CamB (One quad) 1/31

CamC (4 quads) 2/14

le 58%

gn 6/7

Fabrication 2/7

Debug 2/14

Modify test dewar 2/14

Test engineering detector 5/9

Fix problems



Optical Box, Enclosure, & Mechanisms• Optical box & enclosure will soon be a critical task.• Plans for mechanisms have changed.

– Swales Aerospace’s estimate is 3 times higher than that of 1998.– New plan is to purchase high quality, warm stages & design non-precision parts.– Short slack.

Task NameDesign thermal concept

Design mechanical concept

Optical bench & enclosure

Drafting

Fabrication

Testing

Mechanisms

Write specifications

Choose vendor

Detailed design

Fabrication & Testing

pt 5/29

ept 6/7

Optical bench & enclosure 0%

Drafting 7/5

Fabrication 11/8

Testing 4/25

Mechanisms 4%

te specifications 12/15

Choose vendor 7/26

Detailed design 10/25

Fabrication & Testing 4/25



Optics

• Optics & filters are behind schedule.– Estimated time is 2–3 times longer than vendors’ quotes of 26

weeks, because of word-of-mouth tales.

– Schedule could be made up with immediate requisitions and on-time deliveries

Task NameDesign optical details

Choose vendors

Fabricate optics

Fabricate filters

Write RFQ for telescope simulator

Fabricate telescope simulator

etails 2/15

Choose vendors 4/12

Fabricate optics 4/25

Fabricate filters

Write RFQ for telescope simulator 10/25

Fabricate telescope simulator 4/25



Software

• ArcView will be fully tested by CTIO with the Optical Imager.• Scope of software task is uncertain.

– No experience with LabView.

– Need to see ArcView.

• If task is beyond students’ capability, we will seek vendor such as Imaginetics.

Task NameSoftware

Write operating manual

Design software

Write software

Test software

are 0%

nual 6/21

oftware 8/16

rite software 2/28

Test software 5/23



Integration & Installation

• There is a 16 week period for fixing problems.• Delivery is scheduled for 3/28/03.

Task NameIntegration

Integrate electronics & software

Install optics

System integration

Fix problems

Install science-grade detector

Instrument finished

Install on telescope

0%

are 8/15

s 8/15

m integration 10/10

Fix problems 1/30

Install science-grade detector 2/27

Instrument finished 2/28

Install on telescope 5/8



Risk to Schedule

• Tasks on the critical path– Optics are delayed.

– Optical box & enclosure have little slack.

– Mechanisms have a short slack.

• Delay of funding is the greatest risk.– Without starting on the critical tasks, we cannot test our estimates.

We cannot set accurate bounds on the task.

Documents

Science Goals (ref: NSF proposal) Optical Design (ref. “Optical Design”)