Beam tests of a Micromegas Large TPC Prototype
Beam tests of a Micromegas Large TPC Prototype D. Attié SINP, January 22 th, 2010 Nov. 11, 2009 WP meeting 94 1
Overview Introduction, technological choice for ILC-TPC • The Large TPC Prototype for ILC • Bulk Micromegas with resistive anodes • Beam test conditions, T 2 K electronics • Data analysis results: – Drift velocity – Pad response function – Resolution • Comparison between two resistive modules • Future plans: – Pixelized Micromegas module – Integrated electronics for 7 modules Conclusion David. Attie@cea. fr SINP seminar January 22 th, 2010 2
How to improve the spatial resolution? • Need for ILC: measure 200 track points with a transverse resolution ~ 100 μm - example of track separation with 1 mm x 6 mm pad size: 1, 2 × 106 channels of electronics sz=0 > 250 μm amplification avalanche over one pad • Spatial resolution σxy: - limited by the pad size (s 0 ~ width/√ 12) - charge distribution narrow (RMSavalanche ~ 15 μm) 1. Decrease the pad size: narrowed. Calculation strips, for pixels the ILC-TPC + single electron efficiency D. C. Arogancia et al. , NIMA 602 (2009) 403 – need to identify the electron clusters 55 mm 1. Pixels 2. Spread charge over several pads: resistive anode + reduce number of channels, cost and budget + protect the electronics – limit the track separation – need offline computing 2. Resistive anode David. Attie@cea. fr SINP seminar January 22 th, 2010 3
Large TPC Prototype for ILC • Built by the collaboration LC-TPC • Financed by EUDET • Located at DESY: 6 Ge. V e- beam • Sharing out : - magnet: KEK, Japan - field cage: DESY, Germany - trigger: Saclay, France - endplate: Cornell, USA - Si envelope: OAW, Austria - laser: Victoria U. , Canada - Micromegas: Saclay, France, Carleton/Montreal U. , Canada - GEM: Saga, Japan - Time. Pix pixel: F, G, NL David. Attie@cea. fr SINP seminar January 22 th, 2010 4
Large TPC Prototype for ILC David. Attie@cea. fr SINP seminar January 22 th, 2010 5
Large TPC Prototype for ILC • 60 cm long TPC • Endplate ø = 80 cm of 7 interchangeable panels of 23 cm to fit in 1 T superconducting magnet 24 rows x 72 columns <pad size> ~ 3 x 7 mm 2 80 cm David. Attie@cea. fr SINP seminar January 22 th, 2010 6
Module presently avalaible 2 Resistive Kapton ~1 MΩ/□ Resistive Kapton Standard ~4 MΩ/□ Resistive ink ~1 MΩ/□ David. Attie@cea. fr SINP seminar January 22 th, 2010 7
Beam test conditions at B=1 T • Bulk Micromegas detector: 1726 (24 x 72) pads of ~3 x 7 mm² • AFTER-based electronics (72 channels/chip): – low-noise (700 e-) pre-amplifier-shaper – 100 ns to 2 μs tunable peaking time – full wave sampling by SCA – frequency tunable from 1 to 100 MHz (most data at 25 MHz) – 12 bit ADC (rms pedestals 4 to 6 channels) • Beam data (5 Ge. V electrons) were taken at several z values by sliding the TPC in the magnet. Beam size was 4 mm rms. 6 FECs and 1 FEM in its shielding David. Attie@cea. fr SINP seminar January 22 th, 2010 8
Determination of z range • Cosmic run at 25 MHz of sampling frequency time bin = 40 ns • Drift velocity in T 2 K gas (Ar/CF 4/iso-C 4 H 10, 95/3/2) at 230 V/cm: • TPC length = 56. 7 cm agreement with survey 220 time bins David. Attie@cea. fr SINP seminar January 22 th, 2010 9
Pad signals: beam data sample • RUN 284 • B = 1 T • T 2 K gas • Peaking time: 100 ns • Frequency: 25 MHz David. Attie@cea. fr SINP seminar January 22 th, 2010 10
Two-track separation φ r z David. Attie@cea. fr SINP seminar January 22 th, 2010 11
Drift velocity measurement • Measured drift velocity (Edrift = 230 V/cm, 1002 mbar): 7. 56 ± 0. 02 cm/ s • Magboltz: 7. 548 ± 0. 003 cm/ s in Ar/CF 4/iso-C 4 H 10/H 2 O (95: 3: 2: 100 ppm) B = 0 T David. Attie@cea. fr SINP seminar January 22 th, 2010 12
Determination of the Pad Response Function Pad pitch • Fraction of the row charge on a pad vs xpad – xtrack (normalized to central pad charge) Clearly shows charge spreading over 2 -3 pads (data with 500 ns shaping) xpad – xtrack (mm) • Then fit x(cluster) using this shape with a χ² fit, and fit simultaneously all lines See Madhu Dixit’s talk xpad – xtrack (mm) David. Attie@cea. fr SINP seminar January 22 th, 2010 13
Spatial resolution • Resolution at z=0: σ0 = 54. 8± 1. 6 m with 2. 7 -3. 2 mm pads (wpad/55) • Effective number of electrons: Neff = 31. 8 1. 4 consistent with expectations David. Attie@cea. fr SINP seminar January 22 th, 2010 14
Field distortion measurement using laser • Two laser devices installed on the endplate to light up photosensitive pattern on the cathode (spots and line) • Deterioration of resolution at z > 40 cm : due to low field 0. 9 to 0. 7 T in the last 20 cm (significant increase of transverse diffusion) Photoelectrons from the cathode pattern B field map of the magnet 50 cm 30 cm 5 cm Beam position David. Attie@cea. fr SINP seminar January 22 th, 2010 15
Description of the resistive anodes Resistive Kapton Resistive Ink Prepreg PCB Detector Dielectric layer Resistive Kapton Epoxy-glass C-loaded Kapton Epoxy-glass Ink (3 layers) Resistive Ink David. Attie@cea. fr 75 μm 25 μm ~50 μm SINP seminar January 22 th, 2010 Resistivity (MΩ/□) ~4 -8 ~1 -2 16
Comparison at B=1 T, z ~ 5 cm Resistive Kapton • • • David. Attie@cea. fr Resistive Ink RUN 310 Vdrift = 230 cm/ s Vmesh = 380 V Peaking time: 500 ns Frequency Sampling: 25 MHz SINP seminar January 22 th, 2010 • • • RUN 549 Vdrift = 230 cm/ s Vmesh = 360 V Peaking time: 500 ns Frequency Sampling: 25 MHz 17
Pad Response Functions, z ~ 5 cm Resistive Ink Resistive Kapton xpad – xtrack (mm) σz=5 cm = 68 µm David. Attie@cea. fr xpad – xtrack (mm) Γ² = 7 mm Γ² = 11 mm δ = 10 mm δ = 13 mm xpad – xtrack (mm) σz=5 cm = 130 µm ! SINP seminar January 22 th, 2010 xpad – xtrack (mm) 18
2. 2 Ge. V Momentum • B=1 T • Beam energy set to 2. 2 Ge. V using one module in the center • Simple Landau fit MPV = 2. 2 Ge. V David. Attie@cea. fr SINP seminar January 22 th, 2010 19
Tests with Si envelope (nov. 2009) From Vienna Si Si David. Attie@cea. fr SINP seminar January 22 th, 2010 20
Synchronized events with Si envelope Resistive Ink TLU Resistive Kapton David. Attie@cea. fr SINP seminar January 22 th, 2010 21
Description of the Time. Pix chip • Chip (CMOS ASIC) upgraded in the EUDET framework from the Medipix chip developed first for medical applications Pixel • IBM technology 0. 25 m • Noise: ~ 650 e– Cin ~ 15 f. F David. Attie@cea. fr Synchronization Logic 4 1 2 Configuration latches preamp/shaper threshold discriminator register for configuration Time. Pix synchronization logic 14 -bit counter THL disc. – – – Interface Preamp/shaper • For each pixel: 55 μ m – surface: 1. 4 x 1. 6 cm 2 – matrix of 256 x 256 – pixel size: 55 x 55 μm 2 16120 m • Characteristics: 14080 m (pixel array) 55 m Counter 3 14111 m 5 55 μ m SINP seminar January 22 th, 2010 22
Time. Pix synchronization logic control • Each pixel can be configured independently in 5 different modes Timepix Medipix Mode TOT Mode 100 MHz Internal shutter • Internal clock up to 100 MHz Shutter Mask P 1 P 0 Mode 0 0 0 Masked 0 0 1 Masked 0 1 0 Masked 0 1 1 Masked 1 0 0 Medipix 1 0 1 TOT 1 1 0 Timepix-1 hit 1 1 1 Timepix Internal clock 10 ns Digital signal Analog signal not detected Charge summed David. Attie@cea. fr SINP seminar January 22 th, 2010 23
Micro-TPC Time. Pix/Micromegas • Micro-TPC with a 6 cm height field cage Windows for X-ray sources • Size : 4 cm × 5 cm × 8 cm • Read out by MUROS or USB 1. 2 devices Cover • Two detectors are available now at Saclay Windows for β sources 6 cm Field cage Micromegas mesh Time. Pix chip David. Attie@cea. fr SINP seminar January 22 th, 2010 24
Micro-TPC Time. Pix/Micromegas: Time mode • Time. Pix chip + Si. Prot 20 μm + Micromegas • 55 Fe source • Ar/Iso (95: 5) • Time mode • z = 25 mm • Vmesh = -340 V • tshutter = 283 μs David. Attie@cea. fr SINP seminar January 22 th, 2010 25
Micro-TPC Time. Pix/Micromegas: Time mode • Time. Pix chip + Si. Prot 20 μm + Micromegas • 90 Sr source • Ar/Iso (95: 5) • Time mode • z ~ 40 mm • Vmesh = -340 V • tshutter = 180 μs David. Attie@cea. fr SINP seminar January 22 th, 2010 26
2× 4 Time. Pix/In. Grid matrix module • Mother card • Mezzanine card • Guard ring card • Heat dissipation block David. Attie@cea. fr SINP seminar January 22 th, 2010 27
Resistive technology choice Further tests for resistive bulk Micromegas In 2008/2009 with one detector module Reduce the electronics to fit to the module and power consumption In 2010/2011 with 7 detector modules David. Attie@cea. fr SINP seminar January 22 th, 2010 28
7 -modules project FEC FEM PCB detector (bottom) David. Attie@cea. fr SINP seminar January 22 th, 2010 29
7 -modules project David. Attie@cea. fr SINP seminar January 22 th, 2010 30
7 -modules project • Remove packaging and protection diodes • Use 2 × 300 pins connector • Replace resistors SMC 0603 by 0402 (1 mm × 0. 5 mm) 25 cm FEC 12, 5 cm 4, 5 cm 14 cm 2, 8 cm 3, 5 cm 0, 78 cm Chip David. Attie@cea. fr 0, 74 cm SINP seminar January 22 th, 2010 31
Conclusions • Three Micromegas technologies with resistive anode have been tested within the EUDET facility using 1 T magnet to reduce the transverse diffusion • C-loaded Katpon (4 MΩ/□) technology gives better results than resistive ink technology • Data analysis results confirm excellent resolution at small distance with the resistive C-loaded Kapton (4 MΩ/□): 55 m for 3 mm pads • Data analysis of laser tests and the second resistive C-loaded Kapton (1 MΩ/□) should be done soon. • Plans are to test several resistive layer manufacturing process and capacitance/resistivity, then go to 7 modules with integrated electronics by the end of this year David. Attie@cea. fr SINP seminar January 22 th, 2010 32
Dhanyavad David. Attie@cea. fr SINP seminar January 22 th, 2010 33
Backup slides David. Attie@cea. fr SINP seminar January 22 th, 2010 34
Time. Pix chip architecture • Clock per pixel up to 100 MHz • Characteristics: – – – analog power: 440 m. W digital power (Ref_Clk = 80 MHz): 450 m. W serial readout (@ 100 MHz): 9. 17 ms parallel readout (@ 100 MHz): 287 μs > 36 M transistors on 6 layers • Pixel modes: – – masked counting mode (Medipix, Timepix-1 h) Time-Over-Threshold “charge” info Common stop “time” info David. Attie@cea. fr SINP seminar January 22 th, 2010
Time. Pix chip schematic Previous Pixel For each pixel Ref_Clkb Clk_Read Mux 4 bits thr Adj Mask Mux Preamp Input Disc Shutter THR Ctest Testbit P 0 Polarity Timepix Synchronization Logic Shutter_in t 14 bits Shift Register Conf P 1 8 bits configuration Test Input Ovf Control Ref_Clk Clk_Read Next Pixel Analogic part David. Attie@cea. fr Digital part SINP seminar January 22 th, 2010
Readout system for Medipix/Time. Pix chip • MUROSv 2. 1: – – – Serial readout VHDCI cable of length <3 m read 8 chips in mosaic tunable clock [30 -200 MHz] ~40 fps @160 MHz http: //www. nikhef. nl/pub/experiments/medipix/muros. html • USB: – Serial readout – ~5 fps@20 MHz http: //www. utef. cvut. cz/medipix/usb. html • Mosaic achitecture: David. Attie@cea. fr SINP seminar January 22 th, 2010
Detectors using Medipix 2/Time. Pix chip Solid detector Gas detector x, y, F(x, y) Drift grid x, y, z(t), E(x, y) X-ray source Ionizing particle Gas volume + - Semiconductor sensor Flip-chip Single pixel bump bonding connections readout cell + Amplification System (MPGD) - David. Attie@cea. fr Medipix 2/Time. Pix chip Medipix 2 chip SINP seminar January 22 th, 2010 +
- Slides: 38