1. Fred Hartjes 1 RD51 workshop Freiburg, May 25, 2010 Gossip/GridPix testbeam experiment in SPS area T4-134 August 12 – 23, 2010 Maarten van Dijk, Martin Fransen, Harry van der Graaf, Fred Hartjes, Wilco Koppert, Sjoerd Nauta

2. Fred Hartjes 2 RD51 workshop Freiburg, May 25, 2010 256 pixels = 14 mm 1 mm Aim of the beam test uCharacterisation of Gossip detectors lGossip detectors are intended for tracking in the hottest parts (1 GHz/cm 2 ) at the sLHC lTimePix chip (55 x 55 µm) lGas multiplication InGrid lEach pixel sensitive for single e - lReduced drift gap (~1 mm) uParameters to be measured lPosition resolution lAngular dependence lTrack detection efficiency uPrimary gas to be used: DME/CO2 50/50 lVery low diffusion lGood primary ionisation Reviewed by Atlas Upgrade Steering group Very positively commented

3. Fred Hartjes 3 RD51 workshop Freiburg, May 25, 2010 Properties of CO 2 /DME 50/50 u“Cool gas” lV d ≈ 10 µm/ns @ 2 kV/cm lVery low diffusion: ~ 70 µm/√cm lRef: nAr/isobutane 80/20: ~ 250 µm/√cm uWe finally intend ~6 kV/cm lV d ≈ 50 µm/ns lDiffusion ≈ 100 µm/√cm uHigh grid voltage needed to get sufficient gas gain l400 V => 550 V lDischarges at edges (bond wires) lNew detectors passivated by Gloptop Drift field at test beam Gossip working point at sLHC

4. Fred Hartjes 4 RD51 workshop Freiburg, May 25, 2010 Illustration of DME/CO 2 properties uSlanting tracks in same detector 80 pixels (4.4 mm) Ar/iC 4 H 10 80/20 (June 2009 testbeam) CO 2 /DME 50/50 Testbeam September 2009 Limited (38%) single e- eff.

5. Fred Hartjes 5 RD51 workshop Freiburg, May 25, 2010 Fundamental properties of ionization and charge transport uPrimary ionisation uDrift velocity vs field uDiffusion uCluster density for MIPs uPossible electron attachment in DME/CO 2 uMethod lMeasuring tracks parallel to pixel chip surface lEach pixel will have good efficiency (~ 90%) for single electrons lApparatus: Gossip detector with large drift gap (19.3 mm) 4 mm 2.8 mm DICE

6. Fred Hartjes 6 RD51 workshop Freiburg, May 25, 2010 Expected performance of 1-mm Gossip uPosition resolution: 15 -20 µm expected from simulation (angular dependent) l=> RD51 telescope cannot be used lFor the time being a telescope from other Gossips will be used as a reference uPile-up lOne pixel is hit by more than one electron lStrongly angular dependent lMore prominent for DME/CO 2 nLow diffusion u=> we have to do a lot of measurements at many different angles Angle of incidence 0° Measured under poor conditions

7. Fred Hartjes 7 RD51 workshop Freiburg, May 25, 2010 Detector set-up uThree 1.0-mm Gossips + one 19.3-mm Gossip (DICE) u2 miniature scintillators for triggering u1 optical bench 1 m long l4 translation stages l4 accurate rotary stages uTo be fixed as a whole onto a CERN movable table

8. Fred Hartjes 8 RD51 workshop Freiburg, May 25, 2010 How much material do we put in the testbeam? uGossip test lDetectors: 4 Gossips, TimePix based, on PCB lMaterial budget nglass fibre epoxy with copper layers, 4 x 1.6 mm nsilicon chips 4 x 0.7 mm nTotal 4 detectors 10.1% X/X 0 l2 trigger counters 5 mm thick n=> 2.0% X/X 0 u=> total material budget 12.1% X/X 0

9. Fred Hartjes 9 RD51 workshop Freiburg, May 25, 2010 Services: self supporting uGrid voltage lISEG supplies (2 units of 2 CH each) lRemote control via RS232 by LabView on dedicated DCS PC uDrift field by 4 ch NIM HV unit (Wenzel) lRemotely controlled by NI DAQ unit uTriggering lDedicated very fast trigger box from scintillator coincidence lNIM crate with trigger logic uLow voltage: standard lab supplies uDAQ lby MUROS unit using Pixelman software (MediPix) lUsing dedicated DAQ PC uEnvironmental lRecording temperature and atmospheric pressure in HV log file

10. Fred Hartjes 10 RD51 workshop Freiburg, May 25, 2010 Gas: also self supporting uGas system completely in beam zone including gas bottle lUsing single premixed light weight SS bottle under low pressure n≤ 20 bar l=> 42 g H 2 eq for DME/CO 2 50/50 (10 bar) l=> 29 g H 2 eq for Ar/iC 4 H 10 mixtures (≤ 20 bar) lVery well below the 400 g limit of CERN Safety Class 1 nRisk of small local flash fire or explosion lLeak check by measuring gas flow deficit u=> many advantages lNo gas mixing system lNo long pipes l=> rapid and easy installation uRequired gas flow ruled by inflow of oxygen by diffusion l3.5 l/h required getting O 2 level < 30 ppM nRequired for 19.3-mm drift Gossip n1-mm Gossips no problem (< 500 ppM) l=> we will using this flow only during data taking to save gas To be approved by CERN FGSO

11. Fred Hartjes 11 RD51 workshop Freiburg, May 25, 2010 Diagram Nikhef gas filling system uComputer controlled (LabView) gas mixing by pressure monitoring uUsing accurate (+/-35 mbar) electronic pressure sensor uMixtures with up to 3 components 11

12. Fred Hartjes 12 RD51 workshop Freiburg, May 25, 2010 Gas from premixed Nikhef bottles uFilling system now well operational uEach bottle carries two self-sticky labels lLabel with serial# lLabel (automatically printed) with nmixture identification nFilling pressure (max 20 bar gauge) nPreparation date uFor each bottle a log file is automatically stored containing mixing data, place of use and name of user uBottles for external use will be formally certified by Nikhef director uSystem will receive CE classification

13. Fred Hartjes 13 RD51 workshop Freiburg, May 25, 2010 To be approved by CERN FGSO

14. Fred Hartjes 14 RD51 workshop Freiburg, May 25, 2010 Block diagram Gossip beam test uCompletely stand-alone system including DAQ and gas lOnly beam, 230 VAC and movable X_Y table needed lComputer control from barrack using Remote Desktop lBetween barrack and experimental area: 2 Ethernet cables + 2-3 BNC cables

15. Fred Hartjes 15 RD51 workshop Freiburg, May 25, 2010 Installation schedule for August 12 testbeam uMD period July 26 – 30 lLaying Ethernet and BNC cables between testbeam zone and barrack lMake design for a support block on the movable table for mounting the optical bench with detectors uAugust 12 (start testbeam period) lIn testbeam area nMounting optical bench with detectors onto movable table nAlign detectors to beam nInstall 2 PCs, DAQ box, trigger logic incl. NIM crate and gas system lIn barrack nInstall 2 PCs, some NIM logic

16. Fred Hartjes 16 RD51 workshop Freiburg, May 25, 2010 Measuring program in August testbeam uGossip characterisation lSingle electron efficiency vs Vgrid (plateau curve) lPosition resolution nMany different angles lTrack detection efficiency nLow drift field, high gain uDME/CO 2 parameters lUsing parallel tracks (Gossips at ~ 0 deg, DICE at 90 deg) lDrift velocity vs E field lDiffusion/sqrt(cm) vs E lElectron absorption lCluster density lIonisation density uMeasure drift time spectrum at various gains lTime slewing TimePix