SuperB igbite Spectrometer in Hall A
Large luminosity
Moderate acceptance
Forward angles
Reconfigurable detectors
Background: photon (250-500 MHz/cm 2 ) charged (160-200 kHz/cm 2 )
Physics:
Nucleon Form Factors
SIDIS
– TMD’s
... Nucleon structure
Front tracker reused in
BigBite
1

2

3

5

6

7

9

 Use the COMPASS approach: 3xGEM, 2D readout - one significant difference: use new single mask GEM foil (instead of double mask) – cheaper and faster production
 Modular design: chambers consists of 3 independent GEM modules (40x50 cm 2 ) with thin dead area
 Electronics around the module, direct connection; 90 degree bending between modules
 External support frame in carbon fiber (long bars) to minimize thermal deformation
40x50 cm 2 module
Front Tracker Chamber: 40x150 cm 2
10

Module production fully established in INFNCatania
Electronics preliminary QA in Genoa
Module integration and characterization in INFN-
Sanità
GEM Foils
HV curing and quality test
Permaglas Frames
Visual Inspection
Ultrasound bath cleaning
Production rate
2 module in 3 months
Stretching
Gluing
Electronics
Test
Electronics integration
Test and characterization by rad. source and cosmics
Assembling gas lines
Finalization
(solder resistor, check HV)
Put together
(align on reference pins)
Glue Curing
(>24 h)
11

Small scale final system (gas,
3 Big GEMs
DESY/EUDET
Pixel Telescope
LV, HV monitored)

Main Goals:
Characterize chambers in terms of charge sharing, efficiency and spatial resolution at different HV, gas mixture.
Reference
Small GEM
1-4 GeV
Electron
Beam
AIDA-EUDET support

Figure out the gain variation of the previous test
→ Got lot's of good data with high spatial resolution information from pixel telescope
→ Stable performance (“all” conditions carefully monitored)
12

13

Produced or
Assembled
GEM foils 51
Readout+Honeycomb 15 / 18
GEM Module 5+
Front End Electronics 250*
VME Modules
Backplane
Patch Panel
26
80
50
Tested NOT Accepted Degradation after first test
25
7
7
1
3
50
5+
12+ many
0
2
0 many
-
-
2 returned to company***
• The very first 4 GEM foils did not pass the original quality checks, 3 recent GEM foils of the same bunch did not pass the quality check
• One readout + honeycomb suffered bad gluing (probably still usable)
• 2 of the GEM modules passed preliminary tests but then degraded significiantly (large dark current due to (? gas contamination ?); still chance to fix them.
• *** soldering problem, bug fixed
14

2009 2010 2011 2012 2013 2014 2015
Preliminary design
(include MC study)
Prototyping
Procurement
Pre-Production
(1 st chamber)
Production
Beam Test Dec
DESY
Sept
Mainz
June
CERN
May
DESY
Jan
DESY
Development phases in part affected by constraints in the flow of funding
Delivery delay of the GEM foils from CERN for prototyping and pre-production
(no significant impact in SBS, we started erlier)
During pre-production we made some fine optimization:
• GEM foil, details in construction procedure;
• one drastic change in foil quality check.
Development includes: GEM chambers, readout electronics (and related software)
7 July 2014 - JLab SBS Coll. Meeting - Front Tracker GEM 15

S ij
– neuron (0 or 1) – connection between two points
Track Association by NN / Simulation
Vi j
Energy
Neuron changing rate
Kalman Filter Reconstruction: simulation
Residues
16

Continue Production
Test, Characterization and Calibration of GEM and electronics
Fix damaged modules/material, replace if no fix possible
Finalization of a rubust and efficient track reconstruction algorithm
Complete and test the complex firmware of the DAQ
Study (and solve) open issues
17

18

Electronic Status
MPD v 4.0 firmware update
Revised memory mapping
D64 read only cycle implemented: MBLT, 2eVME, 2eSST
2eSST simulated peak speed: 148, 222, 296 MB/s
Event builder implemented and simulated
TBD
Multiboard block transfer
Analyze use of fiber optic protocol
MPD v 4.0 VME interface testing
2eSST cycles tested with STRUCK SIS-3104
2eSST supported by new firmware release
Readout speed measured by software: 100 transfer 4MB each. Data integrity checked for each block.
Speed limited by SIS3104 2Gb/s fiber connection
Bus speed is measured directly on VME bus
Event Builder
Implemented multi event block structure as suggested by DAQ people.
Native data width: 24 bit, packed to 32 bit on 64 bit boundary for efficiency.
Implemented 128MB FIFO data buffer using DDR2 SDRAM.
Quite complicate machinery used to arbiter read and write to/from DDR2.
Output can be read in 32/64 bit format in any of the supported VME cycles, including 2eSST.
Performed functional simulation (FPGA + DDR2 + ADC) of quite simple events to follow all the signals.
Some effort has to be put in DAQ driver to recover packed data.

• Analyze a possible implementation of fiber optic data link to be connected to the SSP
• Implement Multi Board block transfer (some hints needed)
• Deep debug and test (implies rewriting of some DAQ code)
20

21

Hcal
Italian activity
7 July 2014 - JLab SBS Coll. Meeting - Front Tracker GEM 22

23

24

25

26

9 Be(e,e’K) 9 Li
C)
L (G.M. Urciuoli et al. Submitted to PHYS REV
Experimental excitation energy vs Monte
Carlo Data
(red curve) and vs Monte Carlo data with radiative
Effects “ turned off ” (blue curve)
Radiative corrected experimental excitation energy vs theoretical data
(thin green curve). Thick curve: four gaussian fits of the radiative corrected data
An elementary model for the ( e,e′K+) reaction with a different balance of spinflip and non-spin-flip amplitudes might help to resolve the disagreement with theory of the relative strenght of the peaks in the doublets

Present result for K
LL in E07-002 seems to confirm previous measurement of E99-114, at a different angle. Work in progress.
Study of systematic uncertainties is undergoing.
Optimize analysis in order to conclude that at our energy regime, pQCD predictions are excluded.
With those points included, ready for publication
28

Studies on 3He as neutron effective target in SIDIS
29

INFN Group
Bari
Catania
Genova
Rome/Sanità
Researcher
(Unit / FTE)
2 / 0.5
3* / 1.5
1** / 0.5
1 / 0.5
Total 7 / 3.0
Other support: CERN / UVa / JLab
Tech.
(Unit / FTE)
3 / 2
3 / 2
Role
Gas system and beam test
GEM module assembling, mechanics, beam test, analysis
Electronics design and test
Coordination, design, test, chamber integration, analysis, DAQ and reconstruction
6 / 4
• *) one mechanical engineer
• **) electronic engineer
• INFN/Funding:
• Prototyping and Production (2008-2014): 800 kUSD
• Production and Maintainance (2015-2017): ~50 kUSD/year
2
3

• Several small but effective impromevents during last 2 years; one of the most relevent is the compression system during glue curing
• --- From Lead to vacuum ---
Old System based on Lead Bricks
Current System based on vacuum bag
Very uniform pressure for glue degassing
Original idea from LNF- Bencivenni Group
3
3

• Almost finalized
• Use small 10x10 GEM as «reference» (in series)
• Large scintillation pads for cosmic
• Test up to 3 large GEM simultaneously
4
3

Protective
SMD resistors
«spike» signal output HV connector
HV Divider and spike monitor circuit
Reasonably compact,
Easy to change
Include output signal for «spike» detection
5
3

Cluster distributions of 9 cumulated runs
Spacer
Spacer shadow of ≈ 2 mm
→ 2.2% of total dead area of a single GEM module
(Cluster finding not optimized for spacer border)
7 July 2014 - JLab
Masked channels
Spacer
SBS Coll. Meeting - Front Tracker GEM

Zoom
Reconstructed Start Time (t0)
Error on t0
Chi2 of the fit
Use double exponential function to fit 6 samples extracted with uniform random jitter (
±
12.5 ns) and poissonian amplitude distribution + gaussian noise.
Optimal latency setting: first sample around the beginning of the signal
7
3

GEM Beam Test: Spatial Scan – Start Time
Start time increase slightly
(but within uncertienties)
Start time basically constant or sligtly decreasing (opposite to y ?)
8
3

GEM Test Beam: spatial uniformity – Hit Cahrge
9
3

• Background level of:
– 400 MHz/cm2 photons
– 200 kHz/cm2 charged
• Signal width ≈ 250 ns
→ Hits/cm2/trigger ≈ 0.1
→ Ghosts/cm2/trigger ≈ 10-20 !
Three steps track reconstruction:
• Suppress accidentals and ghost with time and charge correlation
• Associate hits with global analysis
(neural network approach)
• Precise track reconstruction on remaining candidate tracks by
Kalman filter method
0
4

Pedestal
Condition:
4 cards connected to chamber (one by flat adapter)
Strips, physics order
Cards connected to VME by 20 m long HDMI cables
RMS of Pedestal
Evident noise on first few (up to 8) channels of each card;
This noise is somehow masked in the card with adapter (this is way we never put great attention to it in the past)
Card + Flat Adapter
GEM
Baseline noise at the level of 7 ADC channel.
VME
HDMI cable
7 July 2014 - JLab SBS Coll. Meeting - Front Tracker GEM 41

Electronics Low Voltage monitor
2 power lines for 2 groups of cards
Small current drop during APV configuration; the “normal” level must restore
In case of proper configuration; otherwise the electronics does not work properly
Chamber appears inefficient → this likely explain the gain issue in 2013 DESY test
7 July 2014 - JLab SBS Coll. Meeting - Front Tracker GEM 42

BARI Gas System w/Humidity Sensor and Absorber
7
Series operation
6
5
4
3
2
1
0
0 50 100 150
Gas Flow (ccpm)
Plastic Tube
Inox Tube
200
< 100 ppM of H
2
O
~ 20 m long pipe
~ 20 m long pipe
7 July 2014 - JLab SBS Coll. Meeting - Front Tracker GEM 43

GEM Test Beam: spatial uniformity – Charge Sharing
Charge asymmetry = 2 (CxCy)/(Cx+Cy) ≈ 0.4÷0.5
HV=4100
Ar/CO2=70/30
→ x/y Charge Sharing ≈ 1.5-1.6
y
NOTE: some x/y hits can be not correcly associated or have pileup due to events with 2 or 3 particles/pulse
Charge asymmetry spread ≈ 0.3 x
7 July 2014 - JLab SBS Coll. Meeting - Front Tracker GEM

Good run
Efficiency holes in some of the runs: beam instability ?
Lost synch with trigger ?
Analysis work in progress, recently master student from
Catania joint.
Several data to be analyzed
7 July 2014 - JLab SBS Coll. Meeting - Front Tracker GEM 45

T0 (A
0
)
T0+25ns
(A
1
)
T0+50ns
(A
2
) cluster
T0+125ns x - Strip y - Strip
Readout dual layer
(x bottom and y top) strip plane
Combination of T0,
T0+25ns,
T0+50 ns
Fit of hit signal evolution

y x
(run 499) cluster charge max strip charge beam position cluster width max peak sample
# clusters x/y
Note: beam crosssection ≈ 3x2 mm 2 , approx. 30% with 2 or 3 hits in beam pulse

y x
(run 499)
Integral Start time (t0) Leading Const.
Trailing Const.
Chi 2 x/y tau0 ≈ 20 ns, tau1 ≈ 90 ns, RMS(t0) ≈ 4-5 ns

Excellent (surprising ?) start time spread (from fit) < 5 ns expected at the level of 25/sqrt(12) ≈ 4 ns (in case of negligible contribution from electron drifting in GEM) x/y start time reasonably correlated no apparent effect of MPD clock synch time spread depends on choice of APV latency (see next slide)
Test beam pulse period 160 ms, multiple of APV clock period
Time spread < 5 ns!

GEM Beam Test: Spatial Scan – Start Time
Gas flow rate: 54.3 ccm
(0.5V/h)
Peak Sample
Start Time (fit)
Look reasonably consistent
Peak Sample vs V/h:
2 V/h: 1.2 ±0.4
1 V/h: 1.2 ±0.4
0.5 V/h: 1.3 ±0.5

7 July 2014 - JLab SBS Coll. Meeting - Front Tracker GEM 51

• If no fix available: change cabling topology
• Test in high bck needed !
• To be investigated
• Implement optical fiber and SSP communication (?)
52 7 July 2014 - JLab SBS Coll. Meeting - Front Tracker GEM

VME intrerface
Event Builder
7-8 July 2014
Paolo Musico and Evaristo Cisbani
SBS Meeting @ JLAB 53

Front Tracker
Chamber
Module
GEM Foils (CERN)
Drift Foil
GEM foils
Readout
Honeycomb
Mechanics (RESARM)
Frame 0 – Cover
Frame 1
– Top
Frame 2 – Drift
Frame 3
– GEM
Frame 4 – Readout
External Frame (Plyform)
Electronics (EES)
FE
MPD
Backplane
Patch Panel (digital + analog)
Electronics Cabling (LINDY)
HDMI 3m cable
HDMI 20m cable
LV cable
HV cable
HV Divider
LV Power Supply (6Vx20 Amp)
HV Power Supply
Gas System
Controller + Flowmeters
(MKS)
Connectors (Legris)
Patch Panel Gas
General Patch Panel
Module Chamber Front Tracker In house Ordered
1
1
3
6
18 3
1
1
1
2
1
1
3
1
1
18
4
8
5
4
7
1
1
10
24
15
12
21
3
1
3
3
3
3
6
3
1
54
4
12
8
3
9
3
3
36
1
324
24
72
48
144
90
72
126
18
6
18
18
18
18
36
18
6
18
54
18
18
1
212
6
1
20
20
20
40
20
250
26
80
50
18
51
15
18
80
45
40
8
24
1
250
4
2
1*
100
**
**
Most of the material already in house or procured
Remaining parts will be procured in 2014
Assembling line operational
Testing line under finalization
*) Design completed, waiting first proto
**) 2014 funding
54

7-8 July 2014




APV Front-End cards


210 delivered (some used for DESY test)
140 in progress (SMD ok, to be bonded)
Backplanes
 74 delivered (some tested)
MPD
 28 delivered and tested: 2 with marginal problems (to be fixed)
Patch Panels


25 Analog Patch Panels delivered
35 Digital Patch Panels delivered
SBS Meeting @ JLAB 55

7-8 July 2014
CYCLE
2eSST160
2eSST267
DATA period
6 ck = 54 ns
(3s/3h)
4 ck = 36 ns
Bus Speed
142 MB/s
213 MB/s
Readout
Speed
117 MB/s
124 MB/s
56