The Schwarzschild–Couder Telescope prototype (pSCT) INFN activity
The Cherenkov Telescope Array (CTA) will be the next generation of ground-based observatory of very high energy gamma ray sources.
The Italian Institute of Nuclear Physics (INFN) is involved in the R&D effort for the development of a possible solution for one of the Cherenkov photon camera designs, working on replacing the Hamamatsu MPPC S12642-0404PA-50 with SiPM from Fondazione Bruno Kessler (FBK), High Density Near UV SiPMs (NUV-HD) with higher detection efficiency for UV photons.
To test the feasibility and the performance of SiPM cameras, a focal plane camera prototype module, upgraded with High Density NUV – SiPMs, produced by FBK, is being assembled to be mounted on pSCT at the VERITAS site by the end of this year.
INFN is currently developing the preamplifiers and the carrier boards for the SiPM
chips that interface with the mechanics of the camera for the pSCT Medium Size
Telescope (BA, PG, NA, PI).
2016 Activity
• 2016 activity was devoted to:
– SiPM characterization
– develop a proper design of the PCB
– assembly, packaging and handling of the PCB’s
– Qualification tests procedures of the PCBs
SiPM effective area:
36.34 mm
2(taking into account bonding pads dead regions) SiPM active area:
27.64 mm
2(taking into account 76%
microcell geom. fill factor)
NUVHD 6x6
SiPM p
+-n of 6x6 mm
2with 30x30 μm
2microcells, sensitive to near UV light and breakdown voltages of
~28V, called High Density NUV (NUV-HD), produced by the Fondazione Bruno Kessler (FBK), have been selected to equip the camera focal plane for a Schwarzschild – Couder Telescope prototype update.
The sensors have been characterized both with pulsed sources and in dark conditions.
FBK provided ~2400 working SiPMs with a dedicated run within the MEMS3 agreement between
INFN and FBK, which will be used to build the pSCT modules + ~300 of a previous run
Peak 11.5 22.5 33.5 44.5 5
Normalization
0 500 1000 1500 2000 2500
Phot o-Ele ctron s 11.5 22.5 33.5 44.5 5
DLED Amplitude (mV)
0.1 0.2 0.3 0.4 0.5 0.6
P.E. PeakLinear Fit/ndf 10/3 2c 1.57e-04±Offset 5.45e-03 1.00e-04±Slope 1.15e-01
Phot o-Ele ctron s 11.5 22.5 33.5 44.5 5
0 20 40 60 80
CT 1.95e-03± 3.40e-01p 5.71e-05± 6.17e-03m Area 21574.9±Norm 23276.9100
/ndf 256/2c 2 Compound Poisson P.E. / peakP.E. 11.5 22.5 33.5 44.5 5
s Residual /
5- 4- 3- 2- 1- 0 1 2 3 4 5 Entries
58116 / ndf
2c 666.2 / 197 Prob 0 N[0] 14.2± 2328 [0] m 0.000 1 ± 0.120 8 [0] s 0.000 06 ± 0.016 54 N[1] 7.1± 638.1 [1] m 0.000 2 ± 0.235 5 [1] s 0.000 12 ± 0.018 41 N[2] 4.2± 213.9 [2] m 0.000 3 ± 0.351 6 [2] s 0.000 26 ± 0.020 17 N[3] 2.6± 85.9 [3] m 0.000 5 ± 0.466 3 [3] s 0.000 40 ± 0.020 75 N[4] 1.75± 38.04 [4] m 0.000 9 ± 0.582 8 [4] s 0.000 74 ± 0.020 67 DLED Amplit ude (m V) 0.1 0.2 0.3 0.4 0.5 0.6
Entries
1 10 2 10 3 10 Entries 58116 / ndf
2c 666.2 / 197 Prob 0 N[0] 14.2± 2328 [0] m 0.000 1 ± 0.120 8 [0] s 0.000 06 ± 0.016 54 N[1] 7.1± 638.1 [1] m 0.000 2 ± 0.235 5 [1] s 0.000 12 ± 0.018 41 N[2] 4.2± 213.9 [2] m 0.000 3 ± 0.351 6 [2] s 0.000 26 ± 0.020 17 N[3] 2.6± 85.9 [3] m 0.000 5 ± 0.466 3 [3] s 0.000 40 ± 0.020 75 N[4] 1.75± 38.04 [4] m 0.000 9 ± 0.582 8 [4] s 0.000 74 ± 0.020 67
35.0 V (+7.0 OV)
--- DLED peaks over thr.
--- Multigaussian fit
Fit to DLED/P.E.
Multigaussian fit to DLED peak amplitude
NUVHD 6x6mm
2- NO LASER – Room Temperature
SiPM Characterization: Noise analysis – Dark conditions
S/N = μ
(1 P.E)/ σ
(0 P.E.)They show a noise rate less than
100 kHz/mm
2, an excellent signal-
to-noise ratio for single photons
and a Photon Detection Efficiency
greater than 40% in UV region.
NUVHD 6x6mm
2SiPM Characterization: V breakdown
SiPM Characterization: Light pulse analysis
Wav elen ght (n m) 250 300 350 400 450 500 550 600 650 700
PDE (%)
0 10 20 30 40 50 60 70
Photo Detection Efficiency misurata per NUV-HD SiPM paragonata a spettri di
segnale e fondo attesi
[6]Spettro di risposta di un NUV-HD SiPM a
impulsi laser a 407 nm
Fill Factors Microcell size Module Total PDE @ max
HAMAMATSU
S12642-1616PA-50 62% 79% 49% ~35%
@450 nm
FBK CTA-NUVHD 76% 80% 61% ~50%
@400 nm
FBK CTA-NUVHD HAMAMATSU 1616PA-50 S12642-
Fill factor comparison
(with Hamamatsu MPPC S12642-1616PA-50 solution)
• The pSCT camera will be equipped with 25 sensor modules, forming a “backplane”
– each module consists of 64 SiPM sensors segmented in 4 independent SiPM carriers (“PCBs”),
connected to a dedicated TARGET7-based module, that communicates with a high-speed backplane.
• The INFN effort, agreed with the SCT community, is to provide at least 25 sensor modules, i.e.
100 PCBs, each housing 16 SiPMs, for a total of 1600 SiPMs.
– Actually only 16 out of 25 modules will be mounted, leaving room for 9 additional modules to be mounted in the next year
• The module and mechanical design and the front-end electronics were adapted to equip the focal plane of the Prototype Schwarzschild-Couder Telescope (pSCT), which will be operated at the VERITAS site in Arizona at the end of 2016.
SiPM = 6,24 x 6,24
Dimensioni Modulo = 53,8 x 53,8 Spazio x Bonding= 0,51 Passo di ripetizione = 6,75 Distanza SiPM-bordo = 0,16
A4
V.Postolac he
R1
CTA MODULE
PESO: $PRPSHEET:{Peso}
$PRPSHEET:{Fine}
FR4 MODULE 6 x 6 x SiPM Separati
FOGLIO 1 DI 1 SCALA:2:1
N. DISEGNO TITOLO:
REVISIONE NON SCALARE IL DISEGNO
MATERIALE:
DATA FIRMA
INFN PERUGIA
DISEGNATO
SBAVATURA E BORDI NETTI INTERRUZIONE DEI
NO ME
VERIFIC ATO APPROVATO FATTO
FINITURA:
ANGOLARE:
QUALITA' SE NO N SPECIFICA TO:
QUOTE IN MILLIMETRI FINITURA SUPERFICIE:
TOLLERANZE:
LINEARE:
15/02/2016
6,24
0,16 6,24
0,51
6,75
53,80
0,51
53,80
0,16
SiPM 6x6 Ste p Re pet 6 ,75
TITOLO:07/03/2016R1 I.N.F .N. P ERU GIA
DISEGNATOVERIFICATOAPPROVATO INTERRUZIONE DEI
LINEARE:
PESO: $PRPSHEET:{Peso}
A3 CTA MO DULE
FATTOQUALITA' $PRPSHEET:{Fine}
ANGOLARE: FINITURA:
TOLLERANZE: BORDI NETTI
NOMEFIRMADATA
MATERIALE: NON SCALARE IL DISEGNOREVISIONE
FR4 N. DISEGNO
SCALA:2:1FOGLIO 1 DI 1 SE NON SPECIFICATO:QUOTE IN MILLIMETRIFINITURA SUPERFICIE: SBAVATURA E
PCB _Mo d_6x 6_Sim etric o
V.PostolacheA
6,75 Step Repet
6,75 Ste p Re pet 0,10 0,30
53,80 0,16
53,8 0
0,51
0,16 6,24 0,11 0,51
53,8 0
1,51
B BON DIN G P AD SCA LA 1 0 : 1 DETT AG LIO B BON DIN G P AD
SiPM 6x6
DETT AG LIO A SCA LA 1 0 : 1
2,27 1,70
0,30 0,30
0,16
0,51 0,10 0,11
2,27
6,24
64 FBK 6x6mm2
25.80 mm
2 5 .8 0 m m
53.80 mm
5 3 .8 0 m m
2 x 25.80 mm = 51.6 mm
~ 2 x 25.8
0 mm = ~ 51.6 m m
Module unit: formed by 4 independent quadrants (“PCB”)
Quadrants (”PCB”)
Homogeneous coverage of camera area
1,50
26,80
Gluing & Bonding PAD
I.N.F.N. PERUGIA
V.Postolache 07/04/2016
VERIFICATO
A3
INTERRUZIONE DEI
PESO: $PRPSHEET:{Peso}
$PRPSHEET:{Fine}
FR4 DISEGNATO
FOGLIO 1 DI 1 SCALA:5:1
N. DISEGNO TITOLO:
REVISIONE NON SCALARE IL DISEGNO
MATERIALE:
DATA FIRMA
CTA MODULE
APPROVATO
SBAVATURA E BORDI NETTI
PCB_Mod_16_6x6_Simetrico
NOME
FATTO QUALITA' ANG OLARE:
FINITURA:
SE NON SPECIFICATO:
QUOTE IN MILLIMETRI FINITURA SUPERFICIE:
TOLLERANZE:
LINEARE:
R1
Scheda Loc X Loc Y Dimensione
A1 3,28 3,28 0,30 PASSANTE
A2 6,66 2,78 0,30 PASSANTE
A3 6,66 3,78 0,30 PASSANTE
A3
0
0 A1 A2
X Y
A
0,68
2,10
5,20
6,75 STEP x 4
5,20 1,55 13,50 STEP x 2 0,68 5,20 1,55
1
23,10
26,80
26,80
R3
23,10 1,90
DETTAGLIO A SCALA 10 : 1 R0,30
0,68
0,30
0,52
0,30 0,62 0,63
5,20
1,55
1,40
R0,30
0,16 1,40 1,80 R0,35
2,30
2,30 0,51
0,30
CTA MODULE
07/04/2016
R1
I.N.F.N. PERUGIA
DISEGNATO VERIFIC ATO APPROVATO FATTO
SOLDER MASK
INTERRUZIONE DEI
LINEARE: $PRPSHEET:{Fine}
FR4 QUALITA'
ANGOLARE:
PESO: $PRPSHEET:{Peso}
FINITURA:
TOLLERANZE:
BORDI NETTI
NOME FIRMA DATA
MATERIALE:
NON SCALARE IL DISEGNO REVISIONE
TITOLO:
N. DISEGNO
SC ALA:5:1 FOGLIO 1 DI 1
A3
SE NON SPECIFICATO:
QUOTE IN MILLIMETRI FINITURA SUPERFIC IE:
SBAVATURA E
PCB_Mod_16_6x6_Simetrico
V.Postolache
B
BONDING PAD BONDING PAD
SOLDER MASK G LUING PAD-BIAS
SC ALA 10 : 1 DETTAG LIO B SOLDER MASK
GLUING PAD-BIAS
26,80
1,50 5 A
26,80
0,16 6,24
6,75 REPET x 4 5
6,24 0,16
6,75 REPET x 4 26,80
DETTAG LIO A SC ALA 10 : 1
2,27 0,62
0,62
0,51 0,62 0,62
1,75
0,16 0,62
2,27 0,62 5
0,51
The PCB layout was designed to optimize the geometry of the SiPM pads and bonding pads (equi-spaced SiPM’s), in parallel with BA group for the design of the PCB
SiPM carrier boards, produced by ARTEL, are ready.
They are being equipped with some passive components for filtering purposes (ready for 25°
of July)
one quadrant with 16 FBK 6x6mm
2SiPM
(*) 2 x 26.80 mm = 53.6 mm
(*)
• A copper block is used to thermally and mechanically couple the PCB to the camera pods to form a module in a backplane
• It is placed on the PCB back-side with high precision in both X, Y and Z
coordinates, before the SiPM placement. This is crucial for the performance of the camera.
• More than 100 copper blocks are available for the assembly
• The blocks have been placed on the dummy PCBs at ARTEL facility with a precision of <100 micron in XY plane,
<0.1° degrees in Z coord (16 available)
• The requirements for alignment precision are of ~300 mm in XY plane and < 2° in z (vertical) axis
• Custom mechanical holders are being
produced with holes and position pins
to achieve a high accuracy for the
alignment (~10 mm) in the xy plane
and z direction (<0.1°).
Livello SiPM
5
9
1,84
9
102
14 9 120
380
5 0,80 9 14
1,84 120
380 B
B A A
UNI EN 22768-1-f per quote senza tolleranzaUNI EN 22768-1-f
27/ 06/ 2016
01 Gluing &Bonding JIG_03 CTA_SiPM_Module
+/ - 0.01 mm. per quote dei fori
PESO : $PRPSHEET:{Peso}
{Fine}
Alluminio A2
FO G LIO 1 DI 1 SC ALA :1/ 1 N. DISEGNO TITOLO :
REVISIO NE NO N SC ALA RE IL DISEG NO
M ATERIALE:
05/ 07/ 2016 V.Postolache
FIRMA NO ME
INTERRUZIO NE DEI
DATA PROG ETTATO
Q UALITA ' FATTO
SBAVA TURA E BO RDI NETTI FINITURA:
A NG O LARE:
R1 I.N.F.N.PERUGIA
V.Postolac he DISEG NATO VERIFIC ATO A PPRO VATO SE NON SPEC IFIC ATO:
Q UOTE IN MILLIM ETRI FINITURA SUPERFIC IE:
TOLLERANZE:
LINEARE:
UNI EN 22768-1-f per quote senza tolleranza SEZIONE A-A
SC ALA 2 : 1
7
0,20
1,50
5,60 SEZIONE B-B
SC ALA 2 : 1 18,90
9
9
14
• The custom jigs to handle the PCB have been designed and are being produced at the PG mechanical workshop (ready by 25 July)
• They have been designed such that up to 27 PCB can be hosted and assembled at the same time
• The jigs will be used to manage the PCBs in all the phases of assembly:
• To place the copper block on the backside, before SiPM placement
• To place SiPM chips with the pick&place machine
• To ship them around
• To bond the SiPMs to input/output lines
• To coat the SiPm surface with UV transparent resin
• Two pairs different jigs were built since the PCBs are of two different,
symmetric types
• As mentioned, in order to speed up the SiPM placement and glueing on the PCBs, a “pick & place” machine is used to first distribute the conductive glue to the PCB top layer metal pads and with a vacuum suction cup to take the sensors out of the custom holder and place them on the PCB with high precision.
• A test to validate the accuracy of the pick & place machine has been run using some NUV-HD SiPM placed over dummy PCBs produced for testing the procedure.
• Different patterns of conductive glue dispensing were used to optimize the alignment precision during assembly (no slip)
With the jigs, to place and glue SiPm on
104 PCB will require 1 day
Best results with one drop of glue and with a layer of glue
The layer has also better planarity, therefore will be the pattern used
Three different pattern were tried: 1) one drop of glue, 2) a layer of glue, 3) four drops of glue
Alignment precision is of the order of
~30 mm and < 0.5° rotation
XY Metrology
XY Metrology
Global Alignment results
Much better than the requirements for pSCT!
Z planarity
Z planarity
• The final step of the assembly is the epoxy dispensing over the SiPM to protect bonds and surface. The epoxy used to cover the modules has to have UV transparency properties to minimize the signal light absorption and mechanical properties for efficient dispensing. Since the UV light absorption probability increases with depth, the epoxy layer has to be dispensed with the highest possible level of uniformity all over the module, avoiding border effects.
• Preliminary tests of resin deposition have been done.
• Good UV transmittance
• Spin coater ti dispense the resin
– SiPM Border effects to be taken into account to get uniform thickness
– Further tests are being performed varying both the spin coater parameter and the resin viscosity
– Difficult task Have to think about NOT using resin coating
Electrical Qualification Tests
PCB multiplexer
Test Station
…
1
16
• After the assembly, the PCBs are electrically tested
• Tests are made using Semtek multiplexer with 16 channels connected through the PCB connector
• Circuit design ready, boards
under construction (ready by
middle of august)
CTA time distribution–data processing prototype
White Rabbit facility in Roma Tor Vergata laboratories has been used to test the CTA Software Array Trigger (SWAT)
Based on CTA-ACTL documents, with the CTA-INAF group, a test-bed of the time distribution – array data processing has been defined and it is currently under implementation
D.D'Urso - INFN and ASDC 25
11/27/21
Prototype Scheme
D.D'Urso - INFN and ASDC 26
11/27/21
