Transcript A powerful tracking detector for cosmic rays: the magnetic

The PAMELA Silicon Tracker
 INTRODUCTION
 MAGNETIC SPECTROMETER
 PERMANENT MAGNET
 SILICON TRACKING SYSTEM
 (MECHANICS)
 PERFORMANCES of the tracking system
 CONCLUSIONS
Lorenzo Bonechi - PAMELA collaboration
INFN Sezione di Firenze - Dipartimento di Fisica dell’Universita’ di Firenze
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
The PAMELA experiment
MAIN TOPICS:
 Flight model
 antiproton and positron spectra
delivered
 search for light antinuclei
 Launch from
Baikonur
SECONDARY TOPICS:
Satellite-borne experiment
 Modulation of GCRs in (Kazakhstan)
the Heliosphere
Semi-polar
orbit  low ene
@
end
2005
!!!
 Solar Energetic Particles (SEP)

Earth
Magnetosphere
3-years
mission
 high
stat
PAMELA
>
3.104 antiprotons
80 MeV/c - 190
GeV/c
> 3.105 positrons 50
MeV/c RESURS
- 270 GeV/c
DK1
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
Satellite and orbit
Resurs DK1

Earth observation

350 / 610 km

Inclination = 70.4o
 Soyuz 2 launcher
Pamela operational
During launch /
orbital manoeuvres
 Baikonur Cosmodrome
 Launch date = end 2005
 3 year mission
 Housed in an atmospheric pressure vessel
 Temperature = 5oC ÷ 35oC
 All subsystems must withstand launch
vibrations!
350 - 610 km
Firenze, 06 October 2005 - RD05
 Electronics must withstand up to ~3 krad
Total mass ~ 470kg / 345W power budget
Lorenzo Bonechi
The PAMELA subdetectors
GF ~20.5 cm2sr
Magnetic
spectromete
r
 Magnetic
rigidity:
R = pc/Ze
 Charge
sign
Requirements:
MDR = 740 GV (4
mm spatial resolution)
Spillover limits:
 Antiproton up to 190
GeV
Firenze, 06 October 2005 - RD05
 Positron up to 270
Lorenzo Bonechi
The permanent magnet
 5 magnetic modules
 Permanent magnet
(Nd-Fe-B alloy)
assembled in an
aluminum mechanics
 Magnetic cavity sizes
(132 x 162) mm2 x 445
mm
 Geometric Factor:
20.5 cm2sr
MAGNETIC
 Black FIELD
IR absorbing
MEASUREMENTS
painting
 Gaussmeter
(F.W. Bell)
Magnetic shields
equipped with 3-axis
probe mounted on a
motorized positioning
device (0.1mm
precision)
 Measurement of the
three components in 67367
points 5mm apart from
each other
 Field inside the cavity
0.48 T at the center
 Average field along
Firenze, 06 October 2005 - RD05
the central axis of the
Lorenzo Bonechi
The silicon tracking system
DESCRIPTION of the SILICON SENSORS
•
•
•
•
Double Sided (x & y view)
Double Metal on the n side (No Kapton Fanout)
AC Coupled (No external chips)
Produced by Hamamatsu
Geometrical Dimensions
Thickness
Leakage Current
Decoupling Capacitance
Total Defects
p side
Implant Pitch
Readout Pitch
Biasing Resistance (FOXFET)
Interstrip Capacitance
n side
Implant Pitch
Readout Pitch
Biasing Resistance (PolySilicon)
Interstrip Capacitance
Firenze, 06 October 2005 - RD05
70.0 x 53.3 mm2
300 mm
< 3 mA
> 20 pF/cm
< 2%
25.5 mm
51 mm
> 50 MW
< 10 pF
67 mm
50 mm
> 10 MW
< 20 pF
Lorenzo Bonechi
structure
of the
tracking
6 detector planes
composed
by 3
system
ladders
ladder : - 2 microstrip
silicon sensors
- 1 hybrid
with
front-end electronics
 silicon sensors
(Hamamatsu):
 300 mm, Double Sided
- x & y view
 Double Metal - No
Kapton Fanout
 AC Coupled - No
external chips
 FE electronics: VA1
chip
 Low
noise
Firenze,
06 October
2005 - charge
RD05
Lorenzo Bonechi
Silicon sensors defects
Request to Hamamatsu: Defects < 2%
Defects:
Short Circuit of AC coupling (Most common, not destructive)
Short between adjacent strips
Open circuit on metal lines
# total defects
It seems to be ‘ perfect ’
12
10
8
6
4
BUT…
More
2
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
0
2
The first batch was OK (Prototype ladders were ‘perfect’, bad strip < 2%)
We started the mass production… Huge number of bad strips (>10%)!!!!!
After a big ‘fight’ we discovered in many sensors short circuits between
adjacent strips at the level of implantation (p side).
Hamamatsu replaced all the bad sensors (few months of delay)
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
Implantation procedure problems!
Transverse ‘cuts’ on the junction side
reduce the interstrip resistance
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
The mechanical assembly
Requirements:
•
•
•
•
•
1 plane made by 3 ladders
no material above/below the plane (1 plane = 0.3% X0!!!)
survive to the launch phase (7.4 grms, 50 g shocks!!!)
good alignment precision
thermal stresses (5-35 0C)
Solution: Carbon fibers stiffeners glued laterally to the sensors
• very high Young module carbon fiber (300 Gpa)
• pultrusion technology
Elastic + Rigid gluing
A very thin (2.5 mm) Mylar foil is glued on the plane to increase
the safety of the whole spectrometer during integration
and flight phases
No coating on the bonding
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
The first silicon plane
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
Mylar film protecting the plane
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
Test plane lodging on the magnet
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
The flight model of the
magnetic spectrometer
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
Detector performances (1)
Strip
noise
Y view
MIP
50
<SIG>GOOD = 9.2
X view
<SIG>GOOD = 4.4
signal
GeV/c
proton
(CERN-SPS
Firenze, 06 October 2005 - RD05
2003)
Lorenzo Bonechi
Detector performances (2)
Spatial resolution
sx = (2.77 ± 0.04) mm
ETA
4
ETA3
ETA2
sy = (13.1 ± 0.2) mm
ETA2
40-100 GeV pions
(CERN-SPS 2000)
beam-test of a
small trackingsystem
prototype
Firenze, 06 October
2005 - RD05
Simulation of silicon
detector:
best p.f.a. 
angle-dependent nonlinear
Lorenzo Bonechi
Momentum resolution
2003 
Last beam-test
of PAMELA
f
40-150 GeV/c protons
Track selection cuts:
 Nx  5
Ny  4
 Hit views 1x and 6x
 95%-efficiency cut on
c2
ultiple scattering
Nx & sx
MDR ~
Firenze, 06 October 2005 - RD05
1 TV
Lorenzo Bonechi
On-ground muon results
2005  acquisition of
atmospheric particles during
PAMELA test before delivering
 Check of spectrometer systematics
positive and negative muons
with
Very preliminary results:
- no efficiency correction
- first-order alignment
- no ETA p.f.a.
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
Conclusions
 PAMELA
apparatus integrated and
delivered to the Russian space agency

launch foreseen for the end of
2005
 Detectors tested with particle beams and
atmospheric muons during integration
phase
Spectrometer:
 sx~3 mm at 0o, sx< 4 mm
10o
 MDR up to 1 TV
up to

meets Lorenzo
the
Firenze, 06 The
October 2005 spectrometer
- RD05
Bonechi
-------------------------------------------------------------
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
The PAMELA experiment
MAIN TOPICS:
•
•
e / p fluxes measurement
Antiproton
flux
Positron
charge
ratio
Search for light Antinuclei
SECONDARY TOPICS:
•
•
•
•
Modulation of GCR’s in the Heliosphere
Solar Energetic Particles (SEP)
Earth Magnetosphere
…
p spectra
e+ spectra
80 MeV/c … 190 GeV/c
50 MeV/c … 270 GeV/c
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
Expected Fluxes in 3 Years
Particle
Number (3 yrs)
Energy Range
Protons
3.108
80 MeV – 700 GeV
Antiprotons
>3.104
80 MeV – 190 GeV
Electrons
6.106
50 MeV – 2 TeV
Positrons
>3.105
50 MeV – 270 GeV
He
4.107
80 MeV/n – 700 GeV/n
Be
4.104
80 MeV/n – 700 GeV/n
C
4.105
80 MeV/n – 700 GeV/n
Antihelium Limit
7.10-8
80 MeV/n – 30 GeV/n
•‘Semi-Polar’ orbit (700)  Low energy particles
•Wide energy range + 3 years mission  Reliable measurements
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
Pamela Subdetectors
TRD
Anticoincidence system
• Threshold device. Signal from
e±, no signal from p,p
• Defines acceptance for
tracker
• 9 planes of Xe/Co2 filled
straws (4mm diameter).
Interspersed with carbon
fibre radiators  crude
tracking.
• Plastic scintillator + PMT
1.2 m
• Aim: factor 20 rejection e/p
(above 1GeV/c) (2. 105 with
calorimeter)
• Timing resolution = 120ps
• Measures rigidity
Si-W Calorimeter
• 5 Nd-B-Fe magnet
segments (0.4T)
• Measures energies of e±.
DE/E = 15% / E1/2 + 5%
• 6 planes of 300mm thick Si
detectors
•+/-10 MIP dynamic range
Firenze, 06 October 2005 - RD05
• Trigger / detects albedos /
particle identification (up to 1
GeV/c) / dE/dx
• Plastic scintillator + PMT
Si Tracker + magnet
• ~3mm resolution in bending
view demonstrated, ie: MDR
= 740GV/c
Time-of-flight
• Si-X / W / Si-Y structure.
Mass ~450 kg
Acceptance ~20.5 cm2sr
• 22 Si / 21 W  16X0 / 0.9l0
• Imaging: EM - vs- hadronic
discrimination,longitudinal and
transverse shower profile
Lorenzo Bonechi
The PAMELA Magnetic Spectrometer
• Magnetic System
– It produces an intense magnetic field region where charged
particles follow curved trajectories
• Tracking System
– It allows to determine six points in the high field region to
reconstruct the particle trajectory and so its momentum and
charge sign
•
•
e+
B
Firenze, 06 October 2005 - RD05
Momentum p = m g v
Charge sign (e+/e-) (p/p)

If B uniform and
perpendicular to p,
then
p = qBr
Lorenzo Bonechi
A glossary of magnetic spectrometers
for cosmic rays studies
•
•
•
Momentum
Rigidity
Deflection
p = qBr (r=radius of curvature)
R = p/q = Br
h = 1/R = q/p
•
DR/R = Dh/h = R Dh (Dh = constant  point’s measurement error)
spatial resolution
•
Maximum Detectable Rigidity (MDR) :
 DR 


R


= 1
R = MDR
e+
B
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
The PAMELA Magnetic Spectrometer
MAGNETIC SYSTEM
•
5 magnetic modules
•
permanent magnet assembled in
an aluminum mechanics
Permanent magnet elements
Geometry
of a magnetic
block
The “Magnetic
Tower
”
– Nd-Fe-B alloy
•
magnetic cavity sizes:
•
field inside the cavity:
•
places for detector planes and
electronics boards lodging
•
Geometric Factor: 20.5 cm2sr
•
Black IR absorbing painting (not
shown in the picture!)
– (132 x 162) mm2 x 445 mm
– 0.48 T at the center
Firenze, 06 October 2005 - RD05
Base Plate
Aluminum
prototype
frame
Lorenzo Bonechi
The PAMELA Magnetic System
Magnetic field measurement
•
Gaussmeter F.W. Bell equipped
with 3-axis probe mounted on a
motorized positioning device
(0.1mm precision)
•
Measurement of the three
components in 67367 points
5mm apart from each other
•
Average field along the central
axis of the magnetic cavity:
0.43 T
•
Good uniformity !
Firenze, 06 October 2005 - RD05
Main field component along the cavity axis
Main field component for z=0 (II)
Main field component for z=0 (I)
Lorenzo Bonechi
The PAMELA Tracking System
The “ladder”
The detector planes
The TRACKER
The silicon sensor
•
6 detector planes
•
each plane:
“ladders”
•
the “ladder”: 2 microstrip silicon
sensors + 1 hybrid circuit with
front-end electronics (VA1 chip)
•
silicon sensors: double sided;
double metalization; integrated
decoupling capacitance
•
resolutions: s x  3mm, s y  13mm
•
MDR > 740 (GV/c)
composed
Firenze, 06 October 2005 - RD05
by
3
Lorenzo Bonechi
Few words on the electronics….
Requirements:
Solutions:
•
Very small power consumption
•
(60 W all included for 36864 readout channels)
CMOS low power analog and digital
electronics
•
Very low noise
(3 mm resolution required!!!!)
•
VA1 chips: ENC = 185 e- + 7.5 e-  C(pF)
Small input Capacitance (<20pF)
Decoupling between front-end and read-out
•
Redundancy and safety
(satellite experiment)
•
Big modularity, hot/cold critical parts
•
Protection against highly ionizing cosmic rays
(Mainly Single Event Effect tests)
•
Selection of components (dedicated tests)
Limiting circuits on the power lines
Architectural `tricks’ (error correction
codes, majority logic etc.)
•
Very big data reduction
(4 GB/day of telemetry, 5 Hz trigger rate, 30
GB/day of data, >90% reduction is mandatory)
•
12 dedicated DSP (ADSP2187) with highly
efficient compression alghoritm
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
Tracker front-end: thermal test
MAGNETIC FIELD
MEASUREMENTS
 Test were done in the
following conditions:
Mechanical aluminum
frames
with
iron
replacing
the
magnet;
Different paints;
System
closed
in
a
vacuum
chamber
to
avoid air convection;
Tracker
planes
replaced
by
mechanical
planes
(mechanical silicon +
raw alumina) with the
front-end
electronics
Firenze,
06 October 2005 by
- RD05resistors;
replaced
Lorenzo Bonechi
Siliconic glue
Silicon gluing points
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
Vibrations tests in Galileo (Florence)
First resonance frequency: 340 Hz!!!!
Test plane survived to +6db spectrum (10.4 g rms)
and repeated 50 g/5 ms + 40g/1 ms shocks
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
ZOP compression algorithm
No Zero Suppression (Losses of particles in case of bad
strips or change in the pedestals!!!)
We use a reversible alghoritm (Zero Order Predictor,
ZOP)
Deventstrip = ADC eventstrip - PEDstrip - CNevent
Deventstrip is distributed around 0
First word is transmitted
Following word is transmitted if above/below n s
.
.
A word is transmitted with the corresponding address if
the preceding one was not transmitted
If a cluster is identified (Deventstrip > N s) 
+/- 2 strips are transmitted
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
Some results on the compression…
Compression Compression
time<1ms factor>96%
• Decompressed data
o Non compressed data
First Plane
Signal/Noise
• Decompressed data
o Non compressed data
Last Plane
Resolution Dx (mm)
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
2002: production of flight model detector planes
Performances obtained with cosmic rays in Firenze : s/n for MIP
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
July 2000: CERN SPS
•
FINAL LADDERS
•
FINAL ELECTRONICS
•
SMALLER MAGNETIC SYSTEM
Spatial
resolution
hDISTRIBUTION
Dp/p versus p
(July 2000 beam test
h =with
1 / R 5= ladder
q / p prototype MS)
s x = 2.77  0.04mm
s / n  50
s y = 13.1  0.2mm
s / n  20
Firenze, 06 October 2005 - RD05
Lorenzo Bonechi
July 2002: CERN SPS
During the last test (June
2002) the spectrometer flight
model has been tested to
determine the performances
Firenze, 06 October 2005 - RD05
300 GeV/c
Electron event
Signal/Noise
Signal
nons/n
bending
 26 view
bending
view
bending
view
s/n
 52
Lorenzo Bonechi