Upload
jason
View
53
Download
4
Tags:
Embed Size (px)
DESCRIPTION
Status report on ECAL. TIM , Houston 8-12 Jan 2007 Marco Incagli - INFN Pisa for the ECAL group. IHEP Beijing, China (Group Leader : Hesheng Chen). INFN Pisa, Italy (Group Leader : F Cervelli). LAPP Annecy , France France (Group Leader: JP Vialle). International ECAL collaboration. - PowerPoint PPT Presentation
Citation preview
Status report on ECALStatus report on ECAL
TIM , Houston 8-12 Jan 2007
Marco Incagli - INFN Pisa
for the ECAL group
Marco Incagli - INFN Pisa 2
International ECAL collaboration
IHEP Beijing, China (Group Leader : Hesheng Chen)
LAPP Annecy , FranceFrance (Group Leader: JP Vialle)
INFN Pisa, Italy (Group Leader : F Cervelli)
Project responsible : F. Cervelli
Marco Incagli - INFN Pisa 3
Where is ECAL
Last detector crossed by incoming (downgoing) particlesThick detector in which particles interact in a destructive way
Minimum amount of material (X0):
Upper TOF
Tracker Entry
Tracker Exit
ECAL Exit
0.2
0
0.2
8
0.3
2
0.4
0
16
.5Tracker Total
0.0
4
ECAL Entry
ECAL Total
16
.1
Marco Incagli - INFN Pisa 4
ECAL = Electromagnet CALorimeter
A CALorimeter is a thick chunk of material in which particles deposit (part of) their energy (CÁLOR=heatenergy)
Most ground based experiments have large calorimetric systems
CMS
Marco Incagli - INFN Pisa 5
Some advantages of calorimeters
Resolution improves with the particle energy, differently from trackers in which resolution momentum
Energy loss due to finite dimensions scales with the logarythm of the particle energy (therefore very slowly)
fit
)%..(E
)%..(1062
50610
E
En
ergy
res
olu
tion
(%
)
50 GeV e-
En
ergy
dep
osit
ed (
GeV
/lay
er)
Longitudinal layer number
Marco Incagli - INFN Pisa 6
Disadvantages of calorimeters
They are heavy!Very different interaction cross sections for “light particles” (photons, electrons, positrons) wrt “heavy particles” (protons, deuterons, nuclei)Often, in ground experiments, 2 calorimetric sections are foreseen
ECAL : electromagnetic section
HCAL : hadronic section
CDF @ Fermilab
Marco Incagli - INFN Pisa 7
Calorimetry in space
due to limitations in weight, space experiments have just an ECAL section, normally with limited thicknessStandard measurement for “thickness” is the radiation length (X0) which is related to the development of the energy deposition by a particle a detector with high X0 has a good energy and angular
resolution and it is capable of measuring particles in the energy range 10GeV-1TeV with good accuracy (<5%)
e.g. : AGILE : 1.5X0 GLAST 10X0 AMS-02 : 16.1 X0
Marco Incagli - INFN Pisa 8
AMS ECAL
Thanks to the characteristics of the ISS (and of the Space Shuttle) AMS02 can take advantage of a calorimeter of unprecedented thickness and granularity, for space standards
1. Very good positron/proton discrimination
2. Good energy reolution for positrons and photons up to ~1TeV
3. Detection of photons non-interacting in material above (~75% of the total)
4. Discrimination of nuclei Z by energy deposition
5. Trigger information on photons and on e±
e+, p ?
Physics with ECAL
Marco Incagli - INFN Pisa 10
Dark Matter searches in positrons
~30% of the universe is made of matter of unknown origin: DARK MATTER
Most theories predict a surplus in the production of positrons, due to DARK MATTER
Hints from HEAT and AMS-01
AMS-02 can measure this distribution with very small statistical error (<0.1%) … BUT …
Marco Incagli - INFN Pisa 11
…background subtraction
…BUT… what about systematic error?A particular concern is the “background subtraction”Proton flux higher by 104 ; to measure positrons at 1% need suppression factor of 106 TRD provides a factor 102:103
The ECAL standalone provides a factor 103
Including the energy-momentum match, a further factor 10 is gained
protons
positrons
Marco Incagli - INFN Pisa 12
Photons can be detected in AMS field of view by two devices
TRACKER: uses conversion in 0.25X0 TRD material (max opening angle ~40o; conversion probability ~18%)
ECAL: uses the shower shape and asks for no hits in TOF counters (opening angle ~20o ; ~78% of the photons reach ECAL)
e+ e-
ECAL structure • Sampling calorimeter with lead foils and scintillating fibers
• Basic block is superlayer: 11 lead and 10 fiber layers
• 9 superlayers with alternating x and y readout
• Total thickness is 166mm, corresponding to 16.2 X0
• Total weight 634 kg
Lead foil(1mm)
Fibers(1mm)
y x
z particle direction1.73mm
p e
FIBER
LEAD
Marco Incagli - INFN Pisa 14
Supporting structure and readout of AMS ECAL
18.5 mm
• The active part is inserted in a supporting structure which connects ECAL to USS
• Light is readout with PhotoMultiplierTubes (PMTs) connected to fibers through light guides
yx
z66 cm
Marco Incagli - INFN Pisa 15
Pictures of ECAL production
ECALECAL
Marco Incagli - INFN Pisa 16
Status of construction in short
Flight hardware: detector : ready front end electronics : ready and mounted on detector readout electronics : QM tested, FM under construction,
ready by end of April 2007 (boards + mechanics) must undergo SQ tests (vibration + thermo-vacuum)
HV power supply : ready except for heaters and thermostats
must undergo SQ tests (vibration + thermo-vacuum) LV power supply : under construction in Taiwan
Marco Incagli - INFN Pisa 17
Detector + front end electronics
Completed in September 2006
Tested in CERN North Area experimental hall Oct 15-30 with protons of 150GeV and electrons of 6-210 GeV
Currently stored in class 100K clean area
Marco Incagli - INFN Pisa 18
Readout electronics : FM
Production started november, 2006 Check of drawings : ok Check of material :
PCBs all done Electrical components all in stock except for few
connectors to be soldered on ECAL Back Plane (EBP) Mechanical parts front panels done and anodized;
to be treated with alodine for conductivity. Crate production in Pisa: Feb,1 May,1 . Surface treatment (white paint): May,1 May,15
Marco Incagli - INFN Pisa 19
3 EBP (backplane), 2FM+1FS , mounted and inspected to be tested in Pisa
All electronics ready by end of April
Marco Incagli - INFN Pisa 20
HV power supply
Flight HV ready and mounted on temporary stand
Missing parts: heaters (2) thermostats (4) reflecting tape
Offer requests sent this week to CGS 3/4 months for delivery
Marco Incagli - INFN Pisa 21
LowVoltage: ERPD crate
FM
•Electronics:
• DC/DC were tested in Taiwan by J. Marin.
• Board with failures repaired by CAEN now at CIEMAT (Madrid).
• Ready to be sent to Taiwan for final test and coating.
•Mechanics:
• Items 5,6,7 ready
• Material for the rest of items 1,2,3,4 sent to IAC.
Marco Incagli - INFN Pisa 22
ECAL test beam
ECAL FM has been mounted on a frame and equipped with FM or QM electronics; work done at LAPP, Annecy (FR) on sep-oct 2006
ECAL mounted on rotating table and carried to CERN
Marco Incagli - INFN Pisa 23
ECAL at test beam
ECAL in the test beam area covered to protect flight hardware (color chosen by Sylvie…)
Incoming beam
Marco Incagli - INFN Pisa 24
Not only ECAL test beam
Cherenkov countersTrigger countersTracker:4 ladders
ECALEcrate 0
Ecrate 1NIM crate
NIM LVDS
PCI AMSWIRE
JTCRATE
JLV
1
JIN
J
JTB
OX
Marco Incagli - INFN Pisa 25
A complex DAQ system
JLV1
ETRG
8 AMSWire cables10 m long
PCI-AMSWIRE
External Trigger
EthernetCable50 m
E0-crate E1-crate T-crate
JT-crate
JINF
ETRG
JINF
JINF
JTBX
JINJ
EPPCAN
Linux PC
PARALLEL PORT
E T SH WE IR TN CE HT
• Software for readout written by Stefano Di Falco with the help, or based on the programs, of P.Azzarello (Perugia), D.Haas (Geneve), A.Kounine (MIT), A.Lebedev (MIT)• Request from Helsinky University of Technology (Finland) to provide raw data taken at test beam in order to test data transfer with real AMS data
2 AMSWire cables
Marco Incagli - INFN Pisa 26
Trigger system : rather elaborate!
Fast trigger window: ECAL,
counters, Cherenkov
“TOF” window: used by external
scintillator counters
Level 1 window
Internal oscillator: pedestals
evaluation (asynchronous
trigger)
Marco Incagli - INFN Pisa 27
To “calibrate” ECAL means to know its response to an impinging particle of known energy
We use “Minimum Ionizing Particles” (MIPs) : hadrons, tipically protons, or muons (cosmic rays at sea level) which cross ECAL without interacting with the lead nuclei but just with the external electrons of the material
Trajectory and energy of MIPs are marginally modified by this process: a MIP proton deposits 0.125GeV crossing ECAL.
MIP
Main goal of test beam: ECAL calibration
Marco Incagli - INFN Pisa 28
Proton runs with 20k events along y and x axisEach run hits the center of a PMT columnTotal of 72 runs per each cross (12 hours of data taking)4 MIP scans:
o Nominal voltageo Nominal voltage+50Voltso Nominal voltageo Nominal voltage+80Volts
Changing voltage varies signal amplification
MIP cross scan
5 rows
4 ro
ws
y
x
Marco Incagli - INFN Pisa 29
MIP distribution
Typical distribution of energy deposited by a MIP in an ECAL cell, in ADC counts, fitted with a Landau distribution
The different peak position reflects the different voltage (=amplification)
Note that a too large amplification limits the dynamical range of the amplifier
Nominal HVPeak = 17.2 ADCcounts
Nominal HV + 50VoltsPeak = 32.7 ADCcounts
Marco Incagli - INFN Pisa 30
Mip distribution
MIP peak (ADC counts)
Problem in “zero suppression” algorythm
We know how to fix this software problem; new software tested with cosmic rays
Marco Incagli - INFN Pisa 31
Beam profile using tracker layers
4 tracker ladders, in front of ECAL, have been used to measure the beam profile and to know the particle hit position event by event
The beam profile changes with particle type (electrons vs protons), beam energy, primary beam polarity, …
Beam dimensions ~2×2cm2; each strip is ~100m
BLACK = 3 laddersRED = 4 ladders
BLACK = 3 laddersRED = 4 ladders
1 cm
1 cm
Marco Incagli - INFN Pisa 32
Tracker-ECAL correlation
0.9cm : ECAL cell size
Beam profile : xview Beam profile : yview
EC
AL
cel
ls
Marco Incagli - INFN Pisa 33
Testing the trigger
ECAL trigger ANALOG part on ECAL Intermediate Board (EIB)
DIGITAL part (=trigger logic) in ETRG board
+
-10 +
-comp
thresh
FF
D Q1RPMT dynode
EIBto ECAL crate
OutputSignal
from PMT Programmablethreshold
Marco Incagli - INFN Pisa 34
Test of analog part of trigger
Trigger efficiency as a function of deposited energy in single cell
Varying the programmable threshold, the trigger threshold moves
Deposited energy (ADCcounts)
100
80
60
40
20
0
Eff
icie
ncy
(%)
Trigger threshold
Trigger threshold
Trigger threshold
Marco Incagli - INFN Pisa 35
Electron runs
Electron can release, in a given cell, an amount of energy which is several thousands larger than that of a MIP
In order to cover the full energy range, the signal is split in 2 and 1 branch is divided by a factor ~30 before ADC
PMT cellHigh gain
Low gain
Z
Y
X
PMT (PhotoMultiplier Tube)
Marco Incagli - INFN Pisa 36
High gain vs low gain
We expect a straight line, but on some cells we see a spurius population under study
Hig
h G
ain
Low Gain
Hig
h G
ain
Low Gain
Marco Incagli - INFN Pisa 37
Linearity plot : ADC counts vs beam energy
As a consequence Anode ADC counts vs
energy shows 2 slopes, up to 30GeV and from 50GeV up . No corrections for energy leakage applied
Linearity plot with dynode (again no corrections) shows a linear behaviour
Under investigation
Marco Incagli - INFN Pisa 38
Conclusions
ECAL ready for integration
FM electronics in production; ready end of April 2007 need to do SQ tests for all crates (Ecrate+HV+LV) !
Test beam shows good results in terms of MIP (protons) analysis;
some unclear results for more energetic (tipically electrons) energy deposits under investigation
very useful to debug and calibrate flight hardware and firmware ; positive test of trigger system and of DAQ procedure
More results at next TIM
BACKUP SLIDESBACKUP SLIDES
Marco Incagli - INFN Pisa 40
Problems in zero suppression
In order to limit the bandwidth, only channels with an ADC count value above some limit (>2 in our case) are written out
This procedure is called zero suppression and it is authomatically run at the beginning of each data taking
Due to the large beam rate, some times a spurius event enters in the evaluation of the pedestal shifting its value
SMALL pedestal shiftDistribution cut at 9 ADC countsFit is still possiblePeak = 17.5 counts
LARGE pedestal shiftDistribution cut at 26(!) countsFit is not possiblePeak = ?
Marco Incagli - INFN Pisa 41
Mip distribution
We know how to fix this software problem; new software will be tested with cosmic ray runs
The effect is uncorrelated with the PMT number, so >90% of ECAL cells (1324 in total) have been equalized using the two scans at nominal voltage
MIP peak (ADC counts)
Remaining channels can be equalized with ground level cosmic rays (muons instead of protons, but same properties for equalization)