Herschel SPIE 6265-09 Photodetector Array Camera ... · PACS Instrument Overview 11. SPIE 6265-09 24 May 2006. Detector Performance. 0 0.2 0.4 0.6 0.8 1 1.2 0 50 100 150 200 250 300

PACS Instrument Overview 1

SPIE 6265-09 24 May 2006Herschel Herschel PhotodetectorPhotodetector Array Camera & SpectrometerArray Camera & SpectrometerAlbrecht Poglitsch, MPE Albrecht Poglitsch, MPE GarchingGarching

Christoffel Waelkens, Katholieke Universiteit Leuven

Otto H. Bauer, Max-Planck-Institut für extraterrestrische Physik

Jordi Cepa, Instituto de Astrofísica de Canarias

Helmut Feuchtgruber, Max-Planck-Institut für extraterrestrische Physik

Thomas Henning, Max-Planck-Institut für Astronomie

Chris van Hoof, Interuniversity Microelectronics Center Leuven

Franz Kerschbaum, Institut für Astronomie der Universität Wien

Dietrich Lemke, Max-Planck-Institut für Astronomie

Etienne Renotte, Centre Spatial de Liège

Louis Rodriguez, Commissariat a l'Energie Atomique

Paolo Saraceno, Istituto di Fisica dello Spazio Interplanetario

Bart Vandenbussche, Katholieke Universiteit Leuven


SPIE 6265-09 24 May 2006

Instrument Concept• Imaging photometry

– two bands simultaneously (60-85 or 85-130 µm and 130-210 µm) with dichroic beam splitter

– two filled bolometer arrays (32x16 and 64x32 pixels, full beam sampling)

– point source detection limit~4 mJy (5σ, 1h)

• Integral field line spectroscopy– range 57 - 210 µm with 5x5

pixels, image slicer, and long-slit grating spectrograph (R ~ 1500)

– two 16x25 Ge:Ga photoconductor arrays (stressed/unstressed)

– point source detection limit 3…20 x10-18 W/m2 (5σ, 1h)

Focal Plane Footprint

32 x 16 pixels6.4” x 6.4”

64 x 32 pixels3.2” x 3.2”


SPIE 6265-09 24 May 2006

Instrument Concept• Imaging photometry

– two bands simultaneously (60-85 or 85-130 µm and 130-210 µm) with dichroic beam splitter

– two filled bolometer arrays (32x16 and 64x32 pixels, full beam sampling)

– point source detection limit~4 mJy (5σ, 1h)

• Integral field line spectroscopy– range 57 - 210 µm with 5x5

pixels, image slicer, and long-slit grating spectrograph (R ~ 1500)

– two 16x25 Ge:Ga photoconductor arrays (stressed/unstressed)

– point source detection limit 3…20 x10-18 W/m2 (5σ, 1h)

Focal Plane Footprint

32 x 16 pixels6.4” x 6.4”

64 x 32 pixels3.2” x 3.2”


SPIE 6265-09 24 May 2006

Observing Modes• Combinations of instrument modes and satellite pointing

modes• Instrument modes:

– photometry (dual-band) – line spectroscopy

• observation of individual lines

– range spectroscopy• observation of extended wavelength ranges

• Pointing modes:– stare/raster/line scan– with/without

nodding/off-position

• Internal chopper– background subtraction– calibration

4


SPIE 6265-09 24 May 2006

FPU/Optics

sGeGaDetectorRed Spectrometer

Blue Bolometer

Red Bolometer

Calibrator I and II

0.3 K Cooler

Filter Wheel I

Filter Wheel II

Grating

GeGa DetectorBlue Spectrometer

Encoder

Grating Drive

Entrance Optics

PhotometerOptics

Calibrator Optics

SlicerOptics

SpectrometerOptics

Chopper


SPIE 6265-09 24 May 2006

FPU-Optics


SPIE 6265-09 24 May 2006

FPU-Optics

FPU


SPIE 6265-09 24 May 2006

FPU Subunits Picture Gallery


SPIE 6265-09 24 May 2006

CRE

• Two 25x16 pixel filled arrays• Extrinsic photoconductors (Ge:Ga, stressed/unstressed)• Integrated cryogenic

readout electronics(CRE)

• Near-background-noise limited performanceexpected

Photoconductor Arrays (Spectrometer)

CRE details: Merken et al. [6275-43]


SPIE 6265-09 24 May 2006

CRE

• Two 25x16 pixel filled arrays• Extrinsic photoconductors (Ge:Ga, stressed/unstressed)• Integrated cryogenic

readout electronics(CRE)

• Near-background-noise limited performanceexpected

Photoconductor Arrays (Spectrometer)

CRE details: Merken et al. [6275-43]


SPIE 6265-09 24 May 2006

Detector Performance

0

0.2

0.4

0.6

0.8

1

1.2

0 50 100 150 200 250 300wavelength [µm]

rela

tive

spec

tral r

espo

nsiv

ity

• Distribution of spectral response within modules

• Distribution of photometric response within modules

• Distribution of absolute photometric responsivitybetween modules

Mean Responsivity of FM HS Detector Modules for Bias=70mV

0

10

20

30

40

50

60

70

80

FM15

3

FM15

8

FM15

9

FM16

0

FM16

1

FM16

2

FM16

4

FM16

5

FM16

8

FM16

9

FM17

0

FM17

1

FM17

3

FM17

4

FM17

5

FM17

9

FM18

0

FM18

1

FM18

3

FM18

4

FM18

6

FM18

7

FM18

8

FM18

9

FM19

0

r [A

/W]

0

0.2

0.4

0.6

0.8

1

0 2 4 6 8 10 12 14 16pixel number

rela

tive

resp

onsi

vity


SPIE 6265-09 24 May 2006

Detector/CRE Performance

• Near background-noise limited performance with peak DQE=0.26 achievable for high-stress detectors with FM CREs

• Analysis of proton tests on HS module indicates no significant excess noise with optimized modulation scheme, IR curing works as well as thermal curing

Mean NEP at 1.85K, C=0.14pF, P=4.2e-15W

0

1E-17

2E-17

3E-17

4E-17

5E-17

6E-17

7E-17

8E-17

0 20 40 60 80 100

bias [mV]

W/√

Hz

FM153FM160FM164FM188FM189FM190

Mean NEP at 1.85K, C=0.14pF, P=2.4e-15W

0

1E-17

2E-17

3E-17

4E-17

5E-17

6E-17

7E-17

8E-17

0 20 40 60 80 100

bias [mV]

W/√

Hz

FM153FM160FM164FM188FM189FM190

Measurement with lab equipment (not optimized for noise)


SPIE 6265-09 24 May 2006

Background- vs. Readout Noise

20 30 40 50 60 70 800

0.5

1

1.5

2

2.5

bias [mV]

signal Circles:measured noise

Dashed line:CRE noise

Squares:CRE noise subtr.

Solid lines:combined model

Dotted lines:background noise

20 30 40 50 60 70 800

0.002

0.004

0.006

0.008

noise

bias [mV]

4.7x10-15 W2.9x10-15 W

• Observed noise explained as photon noise + CRE noise• S/N (of photon noise) requires QE=0.26• With observed noise of readout (integrated with detector), near-BLIP

for HS (red) array expected, but (most likely) not for LS (blue) array


SPIE 6265-09 24 May 2006

Operation/Performance under p+ Irradiation

NEP as a function of detector/readout setting Simulated chopped observation with one ramp/chopper plateau.For each bias value, 5 ramp lengths tested: 1s, 1/2 s, 1/4 s, 1/8s, 1/16 s.The detector was in its high responsivity plateau, ~2 hours after the last curing.

Instrument model value,based on lab measurementswithout irradiation

• With optimum bias setting (lower than in lab!) and ramp length /chopping parameters, NEP close to lab values possible in space

• Curing may be necessary only after solar flare, or once per day (self-curing under telescope IR background sufficient)

Further details: Katterloher et al. [6275-42]Stegmaier et al. [6265-89]


SPIE 6265-09 24 May 2006

Bolometer Arrays (Photometer)

• Two filled arrays: 64x32 pixels (blue) and 32x16 pixels (red)• Bolometers and multiplexing readout electronics operating at 0.3K• Detector/readout noise comparable to background-noise (FM)• Cooler hold time ~59h

Pixel

Bluefocal plane

Photometer unitwith blue + red

focal planesand 3He cooler


SPIE 6265-09 24 May 2006

FM Blue BFP Performance

• Pixel yield ~98%• NEP ~1.7 x BLIP

– But 1/f noise– Best NEP only for fast

modulation (chopping/ scanning)

Further details: Billot et al. [6265-11]Simoens et al. [6275-01]


SPIE 6265-09 24 May 2006

FM Blue BFP Performance

• Pixel yield ~98%• NEP ~1.7 x BLIP

– But 1/f noise– Best NEP only for fast

modulation (chopping/ scanning)

Further details: Billot et al. [6265-11]Simoens et al. [6275-01]


SPIE 6265-09 24 May 2006

Grating Efficiency Verification0th ORDER (INCIDENT AND REFLECTED RAYS @ 19°)FTS MEASUREMENTS AND THREE CALCULATIONS USING PCGRATE

0

10

20

30

40

50

60

70

80

90

100

30 50 70 90 110 130 150 170 190 210

wavelength (mu)

Effic

ienc

y (%

) DE( 0) 200 prec UNPOL

Balance 200 prec UNPOL

DE( 0) 100 prec UNPOL

Balance 100 prec UNPOL

DE( 0) 100 prec UNPOL (detail)

Balance 100 prec UNPOL (detail)

GRATING MEASUREMENT

• Measurement done on production sample in different optical configuration than PACS; however, it gives strong indication that– Grating groove profile is correct– Numeric E&M model prediction is correct


SPIE 6265-09 24 May 2006

Filter Performance

Wavelength [ µm]

Filte

r tra

nsm

issi

on

50 60 70 80 90 100 110Wavelength [ µm]

Filte

r tra

nsm

issi

on

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Wavelength [ µm]

Filte

r tra

nsm

issi

on

100 120 140 160 180 200 220 240 260

Wavelength [ µm]

Filte

r tra

nsm

issi

on

Wavelength [ µm]

Filte

r tra

nsm

issi

on

Wavelength [ µm]

Filte

r tra

nsm

issi

on

70 80 90 100 110 120 130 140

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

80 100 120 140 160 180 200 2200

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

50 60 70 80 90 100 110 1200

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

50 55 60 65 70 75 80 85 90

Spec

trom

eter

P

hoto

met

er

• Requirement of 40% in-band transmission fulfilled or exceeded in most bands


SPIE 6265-09 24 May 2006

QM Instrument Test: Optical Performance

“+” beam “-” beam

Blue photometer, chopped/nodded

• Source: ~ diffraction PSF size hole mask in front of external blackbody, contrast ~1%of background

• PSF wider than expected (25-45%); attributed to defocus of test optics

• Lower limit to spectral resolving power from measurement of saturated water lines ok

Wavelength [µm]

BlueSpectrometer

R>1500

Water spectrum(lab air)

Blue spectrometer, partial x-y rasterof focal plane


SPIE 6265-09 24 May 2006

QM Instrument Test: Photometer Flatfield

CS1CS2 BB1

BB1 (contrast up)

BB2

BB2 (contrast up)

• Chopper scan over full FoV, including internal calibration sources• Chopper scan over full FoV, including internal calibration sources• Illumination: cryogenic BB sources at 2 temperatures• Flatfielding algorithm (through redundant observations) works• Structure/spots in optics (filters? mirrors?) revealed

– elimination by dithering/scanning with satellite


SPIE 6265-09 24 May 2006

QM ILT: Calibration Sources Performance

• Emissivity– Measured against

cryogenic OGSE blackbody

– Values are within specified range.

– But CS 1 value is likely off due to inhomogeneity

• Homogeneity– Ok for CS 2

CS 2 CS 1

Chopper deflection angle


SPIE 6265-09 24 May 2006

Predicted Instrument Performance

• Calculated optical efficiencies– telescope main beam (diffraction + WFE)– grating in respective diffraction orders– diffraction losses due to IFU (image slicer)

0

0.2

0.4

0.6

0.8

Wavelength [µm]

Tele

scop

e ef

ficie

ncy

Wavelength [µm]60 80 100 120 140 160 180 200

0

0.2

0.4

0.6

0.8

Diff

ract

ion

thro

ughp

ut

60 80 100 120 140 160 180 20060 80 100 120 140 160 180 2000

0.2

0.4

0.6

0.8

Wavelength [µm]

Gra

ting

effic

ienc

y


SPIE 6265-09 24 May 2006

Predicted Instrument Performance

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

50 100 150 200 250

PhotometerSensitivity

PhotometricBands

wavelength [µm]

filte

r tr

ansm

issi

on(10'x10'in 1 day)

on-array chopping/line scanning

off-position chopping

wavelength [µm]

[mJy

] (p

oint

sou

rce;

5σ/

1h)

50 100 150 200

5

4

3

2

1

0

60 80 100 120 140 160 1800

1000

2000

3000

4000

5000

SpectrometerResolving power

wavelength [µm]

wavelength [µm]

60 80 100 120 140 160 180 2000

5.10 18

10.10 18

15 .10 18

20.10 18

SpectrometerSensitivity [W/m2](5σ, 1 hour)

(…but cosmic rays)


SPIE 6265-09 24 May 2006

Observing Modes and AOTs

• Point source photometry– Targeted at observations of sources which are completely

isolated and point-like or smaller than one blue matrix. This AOT uses chopping and nodding, both with amplitude of 1 blue matrix, and dithering with a 1 pixel amplitude, keeping the source on the array at all times.


SPIE 6265-09 24 May 2006


• “Small source” photometry– Targeted at observations of

sources that are smaller than the array size, yet larger than a single matrix. To be orientation independent, this means sources that fit in 2’ × 2’. This AOT uses chopping and nodding, but the source cannot be kept on the array at all times.


SPIE 6265-09 24 May 2006


• Large area photometric mapping– This mode is necessary to map sources larger than the array

size, or to cover large contiguous areas of the sky (photometric surveys). There are two ways to perform this kind of observations:

• Scanning (without chopping)– Filled arrays allow arbitrary scanning orientation

• Rastering– Note: Rastering without chopping probably precluded

by 1/f noise


SPIE 6265-09 24 May 2006


• Line Spectroscopy: observation of individual line(s)– Chop/nod or wavelength switching– POINTED: single satellite pointing– POINTED WITH DITHER: small spacecraft movements

perpendicular to the chopper direction to compensate for slicereffects in case of slightly mispointed targets

– MAPPING: limited to rectangular small regions with a maximum extension of 2.8 arcmin to allow for clean chopper off-positions for each raster point; fixed large chopper throw; map parameters in spacecraft coordinates

– Wavelength switching: For one spectral line, the grating will befrequently switched between on-line and off-line. The same pattern will be repeated a few times at slightly shifted wavelength

– Up to 10 lines can be specified in one observation, but all lines have to be reachable without a filter change

– Spectral sampling >3 samples/FWHM (by small up/down scan)– Minimum execution time: 192 s


SPIE 6265-09 24 May 2006


• Range Spectroscopy: observation of extended range(s)– POINTED: single satellite pointing + chop/nod– POINTED WITH DITHER: small spacecraft movements

perpendicular to the chopper direction + chop/nod– MAPPING with chop/nod: limited to rectangular small regions with

a maximum extension of 2.8 arcmin to allow for clean chopper off-positions for each raster point; map parameters in spacecraft coordinates

– MAPPING with off-position: crowed fields and extended spectral structures; chopping between sky and internal CS; map parametersin sky coordinates.

– Up to 10 (freely defined) ranges can be specified in one observation, but all lines have to be reachable without a filterchange

– Spectral sampling: high (>3 samples/FWHM)Nyquist (optimized for speed; minimum timefor full first+second order scan: 1440 s)


SPIE 6265-09 24 May 2006


• SED Mode: full PACS wavelength range with Nyquist sampling– Full grating scan in first order, which covers also the complete

second order in Nyquist sampling. Then, a filter switch is required, followed by a full scan of the third order (+ part of first order)

– POINTED: single satellite pointing + chop/nod– POINTED WITH DITHER: small spacecraft movements

perpendicular to the chopper direction + chop/nod– MAPPING with chop/nod: limited to rectangular small regions with

a maximum extension of 2.8 arcmin to allow for clean chopper off-positions for each raster point; map parameters in spacecraft coordinates

– MAPPING with off-position: crowed fields and extended spectral structures; chopping between sky and internal CS; map parametersin sky coordinates

– Minimum execution time: 2280 s


SPIE 6265-09 24 May 2006

How to Estimate Observing Times?

• Under preparation: HSPOT (developed from Spitzer SPOT) will give accurate time budgets

• Now: A simple ‘Time Guesstimator’ is available for feasibility checks and rough estimates, follow the link ‘Time Guesstimator’ at http://pacs.ster.kuleuven.ac.be/

• Be aware of confusion limits! (0.6, 3, 11mJy)

Documents

Herschel SPIE 6265-09 Photodetector Array Camera ... · PACS Instrument Overview 11. SPIE 6265-09 24 May 2006. Detector Performance. 0 0.2 0.4 0.6 0.8 1 1.2 0 50 100 150 200 250 300