1
TheBeam Conditions Monitor at ATLAS experiment Abstract The Beam Conditions Monitor at ATLAS experiment is being developed as a stand-alone device allowing to measure background events induced either on beam gas interactions or by beam accidents, for example beam hitting at the collimators upstream the spectrometer. This can be achieved by measuring signals in two stations placed symmetric around the interaction point. The 25 ns repetition of collisions poses stringent requirements on the timing resolution. The optimum separation between collision and background events is 12.5 ns implying a distance of 1.9m from the interaction point. Pulses less than 3 ns wide are required with 1 ns rise time and baseline restoration in 10 ns. Combined with the radiation field of 10 15 hadrons/cm 2 in 10 years of LHC operation only diamond detectors are considered suitable for this task. Polycrystalline CVD diamond pad detectors of 1 cm 2 and around 500 μm thickness were assembled with a two-stage RF amplifier and tested in testbeams at MGH in Boston , SPS pion beam at CERN and pion beam at KEK. To increase the S/N ratio two back-to-back diamonds were read out by a single amplifier. Limiting the bandwidth at the readout to 200 MHz provided further improvement; S/N ratio of about 8:1 could be achieved with MIP's. Inclining detectors to 45 degrees with respect to the beam further increases the S/N ratio. Amplifiers of the two stages were irradiated with protons and neutrons to 10 15 particles/cm 2 . Evaluating the irradiated electronics with silicon pad diodes, 20 % degradation in S/N ratio was observed. Ten detector modules are being assembled and tested at CERN for their final installation into the ATLAS pixel support structure in the mid of 2006. V. Cindro 1 , I. Dolenc 1 , H. Kagan 5 , G. Kramberger 1 , H. Frais-Kölbl 4 , A. Gorišek 2 , E. Griesmayer 4 , I. Mandić 1 , M.Meyer, M. Mikuž 1 , H. Pernegger 2 , W. Trischuk 3 , P. Weilhammer 2 , M. Zavrtanik 1 1 Univ. Ljubljana / Jožef Stefan Institute, Ljubljana, Slovenia, 2 CERN, Geneva, Switzerland, 3 University of Toronto, Toronto, Canada, 4 Fotec, Wiener Neustadt, Austria, 5 Ohio State University, Columbus, USA Goals Provide distinctive signature of beam anomalies in ATLAS such as Beam scraping at TAS collimators Beam gas interactions Stand alone device providing information about genuine interactions Monitoring of interactions at LHC start-up Luminosity assessment Requirements Fast signals Rise-time ~ 1 ns Width ~ 3ns Base line restoration in ~ 10 ns to prevent pile-up Single MIP sensitivity S/N ~ 8 for perpendicular MIP before irradiation Installation close to the beam pipe at large η Radiation hardness up to 50 MRad and 10 15 π/cm 2 No maintenance – robust detector Working principle Distinguish interactions from background via time of flight With two symmetric stations at ±Δz/2 Interactions: in time Background: out of time on one side by Δt = Δz/c At high luminosity expect about one hit for each bunch crossing Interactions at Δt = 0, 25, 50 … ns Optimally distinguished background at Δt = 12.5 ns ⇒ Δz = 3.8 m X 2 detector stations, symmetric in z TAS events: Interactions: t=0, 25 ns, 50ns .. Δ Sensors Polycrystalline CVD diamond sensors chosen Radiation hard – shown to work at > 10 15 particles/cm 2 Fast signals – high velocity and cut-off due to trapping Small leakage current – no cooling required Procurement in collaboration with CERN RD-42 Sensors produced and conditioned by Element Six Ltd. Metallized with proprietary radiation hard process at OSU Sensor properties Size 10 mm x10 mm, active 8 mm x 8 mm (metallization) Thickness ~500 μm Charge collection distance ~220 μm Holds ~ 2 V/ μm, operating voltage 1000 V, current ~ nA Sensor assembly – two sensors back-to-back at 45° to increase signal 183cm 38 cm ATLAS 44 m Pixel BCM BCM modules Realization Two stations with four detector modules each Mounted on pixel support structure at z = ±183.8 cm Sensor at r ~ 55 mm, about 20 mm from the beam pipe Sensor with Au test metallizatio n Sensor in module box MIP Back-to-back diamonds at 45° Front-end electronics Two stage Fotec HFK500 amplifier 1 st stage: 500 MHz Agilent MGA-62563 GaAs MMIC low noise amplifier (A = 22 dB, NF = 0.9 dB) 2 nd stage: Mini Circuits Gali-52 InGaP HBT broad-band micro-wave amplifier (A = 20 dB) Amplifiers tested for radiation hardness Up to 10 15 /cm 2 reactor neutrons and protons from CERN PS Agilent: ~ 20 % amplification loss observed with Si diodes, no noise increase Gali-52: 0.5 dB amplification loss FE amplifier OK for BCM application Amplifier schematic 1 st 2 nd stage Amplifier in module box Si diode signal Si diode noise Gali-52 amplification Non-irradiated/irradiated Agilent FE comparison BCM module tests Beam test at MGH Boston Proton beam 200 MeV and 125 MeV Signal ≥ 2.3 MIPs Single and back-to-back sensors at 0 and 45° Signal increase 0->45° by ~ √2 Signal increase in double-decker by 2, noise by 1.3 Bench tests 30 MBq 90 Sr source; ~ MIPs signal QA for BCM modules Module stability tests Noise independent of HV Good reproducibility of signals Signal stable to 4 % during 24 h, longer tests running Back-to-back Single diamond 0 ° 45° single double Fotec FE (500 MHz) 200 MHz BWL S/N ~ 7.5:1 S/N ~ 9.2:1 Typical MIP pulse NINO count-rate vs. threshold for noise / signal & noise ~30 MRad ~100 kRad ATLAS PIXEL ATLAS PP2 electronics cavern Signal processing schematics ~15 m ~100 m Signal Noise Beam test at CERN SPS SPS H8 pion beam – MIPs signal with 16 m of cable: S/N (most probable) ~ 7.5:1 Timing: difference between two stations 2.7 ns FWHM Improved to 9.2:1 by implementing 200 MHz BWL on scope NINO amplifier-discriminator tests Differential timing amplifier-discriminator (1 ns peak, 25 ns jitter) LVDS output with width proportional to time-over-threshold Radiation tolerant design by CERN-MIC Signal split 1:12 into two inputs to increase dynamic range Tests confirm suitability as BCM back-end chip Beam test at KEK Few GeV/c pion beam Signal/noise ~13 at 45 o Noise occupancy measurement
