The photometer Instrument

1
PACS IQRthe Photometer case

• The photometer Instrument ICC groups
• Presented by L. Rodriguez

2
CEA DELIVERIES 4 SUB-SYSTEMS

• The detectorsFrom the early SPIRE development,
  the FM bolometer arraysrepresent the 3rd
  iteration. Most functional problems have
  beensolved to cope with specifications.

3
CEA DELIVERIES 4 SUB-SYSTEMS
The photometer assemblyA large part of the
qualification was done on the photometer asa
whole
4
CEA DELIVERIES 4 SUB-SYSTEMS
The cryocooler This subsystem is of peculiar
importance for Herschel
5
CEA DELIVERIES 4 SUB-SYSTEMS

• Readout Control ElectronicsBolometer readout
  channels, cryocooler temperaturecontrol
  electronics (covered by Martin Von Berg).

6
Part 1 Bolometer arrays specifications
achievements
7
DETECTORS optics geometry
8
Detectors requirements

• Wavelength coverageBlue 60-130 µm Red 130
  210 µm
• Sensitivityclose to background limits

9
Detectors requirements

• Wavelength coverageBlue 60-130 µm Red 130
  210 µm
• Sensitivityclose to background limits

• Low frequency kneerequirement0.1Hz ,
  goal0.01Hz.
• Yieldgood pixels above 95.
• Bandwidthgt5 Hz
• Power budget _at_ 300mK10 µW for 10 sub-arrays

10
CQM STATUS BOLOMETER ARRAYS

• Wavelength coverageOK-gt FM only 1 kind of
  detector manufacture the blue one !
• Yield 6 funct. out of 10 (5 blue / 1 red
  sub-arrays) 1 diff. Not OK 75 pixels -gt FM
  improve manufacture (double tests)
• NEPNot OK -gt noise 10 to 20 times too high!
• ATMEL cold readout BU/CL identified. -gtFM new
  batches.
• Low frequency kneeNot ? OK (0.1Hz), goal not
  measured (0.01Hz)
• Time constantOK -gt 8 Hz (protons hits)
• Power budget _at_ 300mKNot OK 20 µW for 10
  arrays. -gtFM Change readout mode!
• Others Electrical scheme and filtering not FM
  conformal !

11
CQM STATUS ILLUSTRATED Wavelength coverage
No modifications for FM
12
CQM STATUS ILLUSTRATED Yield
Portion of the focal plane with 3 Arrays
very noisy pixels are masked. Others were used
to OGSE qualification.
13
CQM STATUS ILLUSTRATED sensitivity
Noise Pop corn noise identified after some
monthsof investigations as contact noise at
"Pads" level. Revealed only below 4 kelvins!
14
CQM STATUS ILLUSTRATED sensitivity
Noise Pop corn noise identified after some
monthsof investigations as contact noise at
"Pads" level. Revealed only below 4 kelvins!

• Response 2-3 1010 V/W at 0-gt1 pW/pX.
• But 300 mK Power budget gt10µW!
• 2 consequences
• frame frequency lowered to 20 Hz,
• _at_20 µW dispersion on cold readouts.

NEP NOT OK in CQM
15
CQM STATUS low frequency Knee
NOT OK in CQM
The problem of noise on CQM R/O circuits prevent
from measuring low frequency knee of the
detectors themselves.
16
FROM CQM TO FM
Drawing by numbers
Contact noise
17
CURRENT FM STATUS

• We tested many doping parameters obtained
  fromtwo manufacture batches -MV1 1 2 2.5
  3, -MV2 2.5 3 4.
• We have already 5 sub arrays MV1-2.5 ready to be
  used asflight models. We start sorting MV2-2.5
  to complete.
• Today 3 readout schemes explored
• Bridge mode
• Standard NEP 1.5 10-16 BW 0-2.5 Hz

Integration mode standard NEP 1.1 10-16 BW
0-3.5 Hz.
Overbiased NEP 1.1 10-16 BW 0-5.0 Hz.
Overbiased TBD BW defined as frequency with
0.7 x initial amplitude
18
CURRENT FM STATUS NEP/Yield
NEP
Yield (512-5)/512 99
19
CURRENT FM STATUS Bandwidth
Relative responsivity NEP vs Frequency
Background 2pW
20
CURRENT FM STATUS Low frequency drifts
Variations of offsets overnight
12 mV 0.2 pW
Correlation with temperature
Pixel (5,5)
21
Part 2 Cryocooler Qualification achievements
22
QUALIFICATION PROGRAM
SPIRE 290µm/500 µm strings PACS all strings
500 µm
Delivery
23
OUTCOME OF QUALIFICATION PROGRAM
Proof pressure test . ? Thermal
tests - Ultimate temperature .. ? -
Cooling power curve ... ? - Autonomy
... ? problems identified strap
design (ok after modification) cooler
undercharged - 3He charge . to be
corrected on FMs 80C vacuum bake out
...... ? Vibration tests (room and low T)
. ?
At delivery both coolers were operating
according to predictions (taking into account
the underfilling aspect)
24
FM STATUS
From CQMs output
- New thermal straps (as in CQM "version 2") -
New filling procedure established

• Four units manufactured (2 FM and 2 FS)
• FM1 SPIRE and FM2 PACS went thru acceptance
  program
• FM1 SPIRE delivered to RAL Nov. 2004
• FM2 PACS available for delivery
• 2 FS partly assembled awaiting flight
  thermometers/heaters

25
SELECTED EXPERIMENTAL STATUS
FM2 PACS
NET HEAT LIFT
No difference in performance spotted before and
after the 80C bake-out and vibration tests
26
PACS FLIGHT UNIT AUTONOMY TESTS
Example 10 µW applied load - Level 0 _at_ 1.7 K
and Level 1 _at_ 4 K
Experimental 58 h 48 mn Predicted 58 h 22 mn
?
27
PACS FLIGHT UNIT TEMPERATURE STABILITY
28
Part 3 Photometer Qualification results
29
Qualification tests BFP PhFPU

• Thermal vacBake out thermal Cycling
• Vibrations6 configurations
• Thermal insulationRibbon cables kevlar wires
  
• Irradiationsgamma rays / protons alphas (HI)
• ESDin preparation with Q2
• EMCin preparation with Q2

30
CQM STATUS THERMAL VACUUM
Thermal Cycling - 26 cycles done between 300
K and 7K
Bake Out - 1 bake out _at_ 80C during 80h.
Electrical Tests - TAP test was done to
validate the integrity of the detectors between
each thermal sequence. - Optical
characterisation was done before cryo-vibration
_at_ CSL - Optical characterisation will be done
afterall the test sequence. TAP (Automatic
Pads Test equipment) Continuity insulation
between outputs.
31
CQM STATUS VIBRATIONS
32
VIBRATIONS PhFPU STM 2.2 Analysis
B BFP Q2 FM representative - No mechanical
deviation. - Electrical test to be done (feb.)
R BFP Structural Model Acc - 1 wire broken
between 2 pulleys. - Inertia masses not
representative of FM. - Slight deviation of the
focal plane position, acceptable for the optic
the 300mK ribbon cables - Survive to the z RHL
with 1 wire broken. - Thermal insulation no
affected No impact on the mission hold time.
Thermal Switch FM representative -Preliminary
tests shows that the switch is not functional
after cryo-vibrations. - Additional
investigations started in Grenoble.
33
VIBRATIONS Comparison with models
Finite Elements Model - A complete PhFPU FEM
is available results are in good accordance with
experimental data. - Simulations done with real
input levels on the R BFP structure show that
the Kevlar breaking was, at least, highly
probable.
OUR PERCEPTION OF THE TEST FACILITY - A
recurrent problem detailed analysis shows that
input levels still exceed specifications by a
factor 35! - Large injection in transverse
directions. - A Test Facility mode ?232Hz is
close to the BFP suspension one. Excitation seems
non linear ? difficult to predict during low
levels runs. For future tests, CEA requires to
decrease either the vibration hold time in all
axis, or the number of axes TBV.
34
Qualification tests Irradiations
Total dose and gamma rays measurements (0.3
K) June 2003 21 kRads in 12 hours gt rapid
drop of responsivity! annealing with one 300 K
thermal cycle. April 2004 10 kRads in 15
days gt no changes in responsivity. noise
evolution cannot be measured at needed
level. Protons and alpha particles measurements
at Orsay Tandem. July 2004 12 hours protons
6 hours alphas. protons detectors recover
baseline, after impulse when a pixel is
hit,within 120 ms, while MOS readouts recover
within minutes (fewer). Responsivity is not
affected by particle interactions. alphas no
direct hits seen probably misalignment of
cryostat (TBR).
35
Protons hits
Head on impacts Side impacts No beam
Cross talk information
36
High flux situation
Fewer impacts on CMOS
37
Gain variation
38
Conclusions (i)
Photometer System
Detectors achievements On the tested FM arrays,
we met most objectives - sensitivity, -
bandwidth, - power dissipation _at_ 300 mK
(autonomy). We still have to work on - readout
modes. - drifts (low frequency noise).
39
Conclusions (ii)
Photometer System
System concerns We have three concerns to be
solved in the next six months before delvery -
Kevlar wire breakage at BFP level (minor) -
thermal switches (major) - Cryogenic vibration
facility.
40
FINISH!
41
(No Transcript)
42
ANNEXES
43
CQM STATUS RBFP Simulation Analysis
Kevlar wire load limit 90N.
Considering the input levels and the bad inertia
configuration of the STM RBFP FEM, simulations
demonstrate that there was no security
margin. Cumulative Stress of the different axes
causes certainly the Kevlar wire breaking.
44
CQM STATUS TEST FACILITY TRAINING
Input levels higher than Specifications for some
frequencies
Random High Level X Run
45
CQM STATUS TEST FACILITY TRAINING
Problem of Rejection in Tranverse Mode Y input
? Detected level over the Spec. in X factor 35!
Random High Level Y Run
46
CQM STATUS TEST FACILITY TRAINING
Problem of Rejection in Tranverse Mode
Cumulative Stress Z input ? High level detected
in Y direction _at_ BFP level PSD 8.8 grms
Random High Level Z Run
47
CQM STATUS TEST FACILITY TRAINING
Test Facility inner mode ?232 Hz X input ? High
level detected in Y direction _at_ BFP level PSD
6.9 grms
Random High Level X Run
48
CQM STATUS RBFP KEVLAR BREAKING
Sinus High Level X Run
49
SWITCHED PEL MODE
Buffer Unit and Circuit de Lecture MOS very low
intrinsicnoise density 0,5 µV/vHz _at_ 1
Hz. "1/f" noise low, when used after 1h
stabilization time.
One Channel
50
Net advantages of Switched PEL mode
Larger current injected in readouts better
electrical bandwidth better MOS homogeneity
(far from threshold) better noise density.
Only read bolometer PELs are biased lower
power injected at 300 mK stage lower bolometer
temperature Response? -commutation noise added
?- NO!
51
PROTONS ALPHAS IRRADIATIONS
Performed at Orsay Tandem
Protons 20MeV alphas 30 MeV
52
Irradiation cryostat
53
(No Transcript)