VErtex LOcator (VELO)

Size: px

Start display at page:

Download "VErtex LOcator (VELO)"

Valerie Sharp
5 years ago
Views:

1 Commissioning the LHCb VErtex LOcator (VELO) Mark Tobin University of Liverpool On behalf of the LHCb VELO group 1

2 Overview Introduction LHCb experiment. The Vertex Locator (VELO). Description of System. Commissioning the subsystems. Commissioning the full system Status. 2

3 Motivation for LHCb LHCb designed to study: Heavy flavour physics CP-violation in b sector Constrain unitarity triangles Rare B decays Search for New Physics B meson production: bb-pairs correlated in space. 3

4 LHCb detector design 27th July,

5 LHCb detector rotated 27th July,

6 LHCb detector (Reality) Muon det Calo s RICH-2 Magnet OT+IT RICH-1 VELO Installation of all detectors is complete 27th July,

7 Vertex detector requirements Event reconstruction: Precise tracking Low mass Vertexing: Need to separate primary/secondary interactions Close to beam in extreme radiation environment. Trigger: Fast reconstruction of primary vertices. R-Φ sensor geometry. LHC beam tolerance: Detector must be 30mm from beam at injection 7

8 VErtex LOcator 2 retractable halves 5mm from beam when closed 30mm from beam during injection 21 R-ФФ modules/half Operates in secondary vacuum (<10-4 mbar) 300μm foil separates detector from beam vacuum (<10-8 mbar) 2 phase CO 2 cooling system 27th July,

9 Vacuum system VELO sensors PileUp sensors RF-foil Detector separated from beam vacuum by 300μm Aluminium i foil Maximum allowed differential pressure is 5mbar Shape allows overlapping sensors 27th July,

10 n strip sensor technology (Micron) Double metal layer for signal routing. R sensors: 45 degree quadrants Pitch=40-101μm Ф sensors: 2 regions Short/long strips Pitch=38-100μm Stereo angle. Sensor Design 10

Radiation tolerance type inverted type in nverted not

+ in n-bulk sensors Strip isolation via p + spray 300μm

3x1010 14 n eq /cm 2 /year at 8mm 5x10 12 n eq /cm 2

11 Radiation tolerance type inverted type in nverted not after 1 year of irradiation [2 fb -1 ] R [cm] U dep [V] n + in n-bulk sensors Strip isolation via p + spray 300μm thick Harsh non-uniform radiation environment x n eq /cm 2 /year at 8mm 5x10 12 n eq /cm 2 /year at 42mm Run partially depleted after 3-4 year (8fb -1 ) Reduced CCE 11

12 Module design Double sided hybrid to balance stresses due to bi-metallic effects. Thermal pyrolytic graphite core removes 24W of heat (ΔT=20C). Carbon fibre cover for rigidity. 2 phase CO 2 cooling system Sensors operate at -5C Non-planarities ~250μm Sensor-sensor < 10μm Readout with 16 Beetle chips: 2048 channels/sensor channels in VELO. 12

13 Material budget 13

14 Material budget/component 14

15 Module production (to be repeated) 6Vi Visual linspections, 6M Metrologies, 7 Electrical l Tests, 4V Vacuum Tests 15

16 VELO CERN 16

17 Installation 17

18 Installation 18

19 What the LHC sees during injection 19

20 Overview of VELO electronics All components installed. Each system commissioned stand-alone first. 20

21 ARX ARX ARX ARX Readout Board (Tell1) CC CPC TTCRx GBE Analog input from 64 links 10 bit 40MHz [A- Rx] FPGA for pre-processing processing Cross-talk, common mode suppression, clustering Zero suppression of data Data sent out by Gigabit ethernet t Max output rate is 1.1MHz Control via credit card PC 27th July,

boards [Tell1] Control boards Readout slices Timing

22 Control system and DAQ All cables and boards installed Extensive testing programme: ADC cards [A-Rx] Readout boards [Tell1] Control boards Readout slices Timing Cable checks Noise levels adjustment t of timing i to 0.65ns 22

23 Standard Easy3000 CAEN system Low voltage system 1 mainframe controller 3 48V power convertors 22 8V/9A 12 channel power supply boards. 2 modules per power supply board. 23

24 Problems with LV system Reliability of A3009 8V/9A supplies Many were sent back to CAEN Mean = 0.58 Bad connections to front of supply. Solder leakage on manufacture of connectors. 24

25 Safety problem with A3009 modules Tried many safety test. One problem found. Communication cable was pulled. Channels stayed on. Jumpers from interlock were pulled. Channels stayed on. Simultaneous loss of cooling Rapid heating of hybrid. Firmware fix from CAEN Currently not used. 25

26 High Voltage System 5 Iterations of OPC Server from ISEG An example problem: Ramp channel one Turn on channel two Channel one jumps to target voltage Remaining problem OPC server crashes ~ once per week ISEG EHQ 607n-F Company engineers extremely helpful but a long (and not yet finished) process

Commissioning the VELO (I) Single module test repeated for all modules on 1st side 5 th May onwards Warm cooling (8C) Modules operated at 25C Differential pressure between een 2 and -5mbar Operation

27 Commissioning the VELO (I) Single module test repeated for all modules on 1st side 5 th May onwards Warm cooling (8C) Modules operated at 25C Differential pressure between een 2 and -5mbar Operation of 15 modules on data taken in assembly 15 th May 2008 Single module operation under Neon atmosphere 18 th March Final cooling, vacuum, LV, HV, interlocks, DAQ and control software Looked at IV scans Data taken in various configurations Noise level compared with previous data taken in assembly R-side 27th July, 2008 not CM corrected CM corrected 27 Dif fferential pressure e

28 Commissioning the VELO (II) Single module test of 2 nd half from 2 nd June. Full half powered for first time June 10 th. Many problems found. Failure and replacement of broken read-out boards. Problems related to fabrication of Tell1 boards. Preparation for operation under vacuum. 27th July,

29 VELO under vacuum Full detector operated under vacuum after beam evacuated. Operated cooling fully loaded at -25C 10 -5C -5 Cooling of rf foil affects beam pressure. Majority of data taking during commissioning at C 14C. Minimize thermal cycling of modules. 29

30 full power 30

31 Data taking Non-zero suppressed data: events 10Hz with limited event builder. Zero suppressed data (with test pulses) 1kHz with limited event builder. 1/100 th of design rate. 31

32 Noise performance Phi sensor R sensor Raw noise Common mode subtracted noise 32

33 Noise all sensors PHI R Mean = / Mean = /

34 Noise comparison between single module power up and system power up 34

35 Status Installation Module production and assembly of detector halves March 2007 All other systems installed: Cooling Vacuum Positioning system. High voltage Low voltage Installation of detector halves October 2007 Connected the cables to the modules. Full system operation June 2008 Outstanding issues Integration into LHCb global running. 3 broken Tell1 boards to be replaced. Data taking at high rates. Software improvements: Online monitoring. Wait for first beams. 35

36 Conclusions All components fully installed and tested. Extensive commissioning i i programme over last year. First operation of whole detector under vacuum 99.2% of channels fully operational. LHCb VErtex LOcator in good shape for first beams 36

Commissioning the LHCb VErtex LOcator (VELO)

Commissioning the LHCb VErtex LOcator (VELO) University of Liverpool E-mail: Mark.Tobin@cern.ch The LHCb VErtex LOcator (VELO) is designed to reconstruct primary and secondary vertices in b-hadron decays. It is a silicon microstrip detector situated