Danilo Domenici On behalf of the KLOE-2 IT subgroup Vienna Conference on Instrumentation 2013

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 2  Drift Chamber (He/iC 4 H 10 light gas mixture)  EM Calorimeter (Pb/SciFi, excellent time resolution)  0.52 T Magnetic Field (Superconductive coil)  Dafne  factory (e + e - at 1020 MeV and L int =10fb -1 /y)

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 3 LET QCAL InnerTracker 3 new detectors will be inserted before summer:  Photon taggers (HET/LET for γγ interactions)  Low angle Calorimeters (CCAL/QCAL to improve solid angle coverage)  Inner Tracker (to improve vertex resulution) resolution on K S  ππ vertex from c  S to c  S /3

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 4  4 layers at 13/15.5/18/20.5 cm from IP and 700 mm active length   r φ  250 µm and  z  400 µm in 0.42 T magnetic field  XV strips-pads readout (20 o ÷30 o stereo angle)  2% X 0 total radiation length in the active region 3 mm 2 mm Cathode GEM 1 GEM 2 GEM 3 Anode Read-out Cylindrical Triple GEM FEE boards

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF mm 300 ÷ 430 mm  First batch ever produced with a single-mask etching technique developed by CERN-TE-MPE-EM and RD51 to produce large area foils  The top side of the active area is divided in 40 sectors  The HV connections are grouped in 4 tails that are directly connectorized  Each sector can be set to down voltage, ground or float by an external jumper pinholes gas holes

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 6 few sectors with current > 1 V GEM is tested in a N 2 flushed plexiglass box to reduce RH below 10% Each Sector must draw a current < 600V Discharge rate is measured over a period of ~1h Discharge V

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 7 Epoxy glue is distributed by hand on a 2 mm wide line 3 GEM foils are spliced together with a 3 mm overlap and closed in a vacuum bag (0.9 bar) Alignement pinholes Vacuum holes Alignement pinholes Vacuum holes

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 8 Vacuum bag is closed GEM is protected with a Mylar sheet and wrapped on the cylindrical mold Transpirant tissue is placed around to distribute vacuum Final cylindrical GEM with internal and external rings

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 9 Inner layer is glues on the mold Nomex honeycomb is glued on the cathode foil Cathode is rolled on the mold and glued in a vacuum bag Final electrode

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 10 X strip V strip X pitch 650µm  X res 190µm B=0.5T) V pitch 650µm  Y res 350µm Readout plane is realized at CERN TE-MPE-EM as a kapton/copper multilayer flexible circuit. It provides 2-dimensional readout with XV strips on the same plane  X are realized as longitudinal strips  V are realized by connection of pad through conductive holes and a common backplane  Pitch is 650 µm for both

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 11 is rolled over the mold …and closed in a vacuum bag 1k strips 1M pads Readout circuit obtained by splicing 3 foils …glued…

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 12 first 90µm CF skin5mm HC second 90µm CF skin curing 24h in autoclave final readout electrode  Then the circuit is shielded with with a very ligth Carbon fiber composite structure realized by an external company (RiBa Composites, Faenza, IT)  The shield is composed by a sandwich of two 90 µm thick carbon foils prepreg with epoxy spaced by a 5 mm thick Nomex honeycomb

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 13  To avoid possible relaxation of the gaps due e.g. to thermal expansion of the foils, we fix a spacing grid on the GEMs (only for Layers 3 and 4)  It is realized by assembling 8 rings and 12 rods of 300 µm thick PEEK PEEK grid assembled grid fixed on the GEM

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 14 The GEM is placed on the Machine with its mold Everything is aligned with an axial precision of ≈0.1mm/1.5m The Readout is moved down around the GEM Anode GEM A Vertical Insertion Machine is used to assemble the 5 electrodes of a Cylindrical-GEM

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 15 Internal GEM surface with the anular FR4 flange Detector is sealed by an epoxy flow Insertion machine is rotated to seal both sides Final C-GEM detector

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 16  The 3 innermost Layers have been completed and tested with β source and cosmic rays  The Layer4 has been closed last week  At the end of February the 4 layers will be inserted one into another and mounted on the Dafne beam pipe Layer 1 Layer 2 Layer 3 Layer 4 FEE boards with Gastone chip

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 17 Technology0.35 CMOS - no radhard Sensitivity (pF)20 mV/fC Z IN 400 Ω (low frequency) C DET 1 – 50 pF Peaking time90 – 200 ns (1-50 pF) Noise (erms)800 e e - /pF Channels/chip64 ReadoutLVDS/Serial Power consum.≈ 0.6 mA/ch 128 channels GASTONE Board  Mixed analog-digital circuit  Low input equivalent noise, low power consumption and high integrated chip  4 blocks: 1. charge sensitive preamplifier 2. shaper 3. leading-edge discriminator 4. monostable Visit the Gastone poster by Flavio Loddo

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 18  All the detectors have been tested with a cosmic-ray test-stand equipped with an external Tracking System provided by 3 planar GEMs 10x10 cm 2  Final signal and HV cables, FEE and DAQ systems have been used in the tests  The test-stand is also equipped with a 90 Sr movable source to perform fast measurements without tracking

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 19 Z resolution r- φ resolution (bending plane) KLOE field 0.52 T  Space resolution has been measured as a function of a transverse Magnetic Field to reproduce the KLOE situation  In the bending plane the electrons are spreaded by the field with a consequent increase of space resolution  At the KLOE field of 0.52 T the resolution  r φ  250 µm is still within the experiment requirement  The field effect is not visible in the non-bending plane as expected

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 20  The capacitive coupling between the GEM3 Down and the Readout plane could result in induced currents causing high-multiplicity «splash events»  The effect is strongly suppressed by coupling the GEM3 Down to Ground through a series RC circuit with large capacitance and small resistance without BC C = 2.2 nF R = 10 Ω

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 21 The profile of the source in 6 different positions is reconstructed by triggering the DAQ with a clock signal This fast test allows to check the cabling and the uniformity of the detector

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 22 Cosmics hits are reconstructed by requiring a track in the 3 planar GEMs

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 23 Z vs X (Lego View) noise cosmic tracks

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 24

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 25

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 26  Last October in Frascati we organized a Cylindrical GEM Mini-workshop dedicated to this novel technology  Among the participants there were: Rui de Oliveira from CERN workshop where the GEM foils are produced Qun Ouyang from IHEP Beijing Lev Shekhtman from INP Novosibirsk  A very strong interest have been shown by our chinese and russian collegues to exploit the Cylindrical GEM technology on the future upgrades of the BESIII and VEPP2000 experiments

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 27  We exploited the intrinsic lightness and flexibility of a GEM detector to build a fully cylindrical Inner Tracker without frames in the active area and an X 0 ≈ 2%  After more than one year of construction time we have completed the 4 Layers that will be mounted as upgrade of the KLOE detector  The detectors have been extensively tested showing a good operational stability and the expected performance  We are collaborating with other groups that have expessed interest on Cylindrical GEMs in order to improve such technology and expand its operational field

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 28

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 29 Material Radiation Length (cm) Copper1,43 Polyimide - Kapton28,6 Carbon fiber28 Argon14000 Isobuthane17000 Epoxy - Araldite ,5 Honeycomb - Nomex1250 Fiberglass - FR416 Air30500 Aluminum8 Gold0,33 Thickness (µm) Radiation Length (%) Copper31,68E-04 Polyimide501,40E-04 Copper31,68E-04 GEM foil564,76E-04 Copper32,10E-04 Polyimide501,75E-04 Honeycomb30002,40E-04 Polyimide501,75E-04 Copper32,10E-04 Cathode foil31061,01E-03 Gold0,13,03E-05 Copper52,45E-04 Polyimide501,75E-04 Copper51,05E-04 Epoxy12,53,73E-05 Polyimide1254,37E-04 Epoxy12,53,73E-05 Polyimide501,75E-04 Copper32,10E-04 Gold0,13,03E-05 Anode Foil2631,48E-03 Carbon fiber903,21E-04 Honeycomb50002,40E-04 Carbon fiber903,21E-04 CF Shield32009,54E-04 Total 1 Layer4,93E-03 Total 4 Layers1,97E-02 The KLOE-2 requirement of X0 < 2% is fulfilled Gas (90% Ar – 10% iC4H10)90006,29E-05

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 30

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 31  Gas gain measurement performed in current mode using as normalization a refernce detector with known gain  Discharge measurement performed with a 141 Am α source  KLOE2 chosen mixture Ar/Isobuthane (90/10) is compared with the standard gas mixture Ar/Co 2 (70/30)  Our working gain will be 2 x 10 4

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 32 V Strip# Shorts Made by a 100 ps precision Time Domain Reflectometer Transmission line length and its damages evaluated by measuring the delay of the reflected signal

VCI 2013 – 12/02/2013 D. Domenici – INFN-LNF 33