ammosov

Small pad RPCs as detector for high granularity digital hadron calorimetry: 

Small pad RPCs as detector for high granularity digital hadron calorimetry Vladimir Ammosov Institute for High Energy Physics Protvino Moscow region, Russia

Outline: 

Outline Motivation FLC detector requirements Tests of RPCs - in avalanche mode - in streamer mode 4. Comparison of operation modes 5. Conclusion

Slide3: 

Pjets=Σ(Pch.part. + Pγ + Pneutr.hadr.) fraction 65% 26% 9% Precise particle reconstruction in jets is needed Physics requirements of future e+e- linear colliders (FLC) with E=500-1000 GeV for jet energy reconstruction and separation using energy flow algorithms CDF2, ZEUS, ALEPH, ATLAS To diminish the σconfusion a high 3D granularity of ECAL and HCAL is needed Motivation Tracker ECAL HCAL FLC

FLC requirements for HCAL: 

FLC requirements for HCAL Detector for sampling HCAL - insensitive to ~ 4T mag field - rather thin, ~ 6 mm - rather slow expected barrel rate <10-4 Hz/cm2 forward rate <10 Hz/cm2 (cosmics ~ 10-2 Hz/cm2) - ~50 M channels ( 38 layers, 2 cm SS sampling, 6.5 mm thick active layer, 1x1 cm2 pads) Analog HCAL is very expensive. Digital HCAL could be considered. RPC as detector is proper choice. Tesla as example

FLC requirements for DHCAL: 

FLC requirements for DHCAL RPC pad size GEANT 3.2 simulation of DHCAL response for pions 38 layers – 2cm SS absorber, 6.5 mm RPC RPC – 1.2 mm gas monogap, TFE gas, glass as resistive plates MIP/2 cut nonlinearity is taken into account 1x1 cm2 pads case is compatible when all hits are counted Preliminary RM ~ 1.65 cm

FLC requirements for DHCAL: 

FLC requirements for DHCAL RPC pad size In energy range of 1-20 Gev resolution for AHCAL and DHCAL is compatible. Further nonlinearity of DHCAL becomes important. Preliminary

RPC tests: 

RPC tests Set-up at IHEP 18T channel 5 GeV/c h+ beam RPC samples - 0.8, 1.2, 1.6, 2.0 monogaps - 1013 cm window glass - 16 pads of 1x1 cm2 - in tight box Trigger S1S2S3S4 for 2x2 cm2 area Di - preamp+disc

RPC tests: 

RPC tests Gas mixtures RPCs were tested in saturated avalanche and streamer modes For both modes TetraFluoroEthane (TFE) based mixtures were used TFE = freon 134A = C2H2F4 ~ 8 ionizations/mm Saturated avalanche mixtures = TFE/IB/SF6 IB = Iso-C4H10 as quencher, IB fraction = 5% SF6 as streamer suppresor, SF6 fractions = (2-5)% Streamer mixtures = TFE/IB/Ar or N2 IB = Iso-C4H10 as quencher, IB fraction= (5-20)% Ar/N2 as streamer developer, fractions = (2-20)%

RPC in avalanche mode: 

RPC in avalanche mode Typical Q and m distributions 1.2 mm, 2% SF6, 8.4 kV - working point, 2.2 mV thr Q ~ 107 e 2 adj pads Mean 2.8 pC RMS 1.6 pC Mean 1.47 RMS 0.58

RPC in avalanche mode: 

RPC in avalanche mode 1.2 mm gap RPC eff, <m> vs HV - 2% and 5% of SF6 For 2.2 mV Knee 8.2 kV 8.6 kV V 0.6 kV 0.6 kV Thresholds  - 0.6 mV  - 2.2 mV  - 5.0 mV 2.2 mV is best threshold eff >99% low <m> ~ 1.4

RPC in avalanche mode: 

RPC in avalanche mode 1.6 mm gap RPC eff, <m> vs HV - 2% and 5% of SF6 For 2.2 mV Knee 8.8 kV 9.8 kV V 0.8 kV 0.8 kV Thresholds  - 0.6 mV  - 2.2 mV  - 5.0 mV 2.2 mV is best threshold eff >99% low <m> ~ 1.4

RPC in avalanche mode: 

RPC in avalanche mode 2.0 mm gap RPC eff, <m> vs HV 2% and 5% of SF6 For 2.2 mV Knee 10.0 kV 11.4 kV V 0.8 kV 0.6 kV Thresholds  - 0.6 mV  - 2.2 mV  - 5.0 mV 2.2 mV is best threshold eff >99% low <m> ~ 1.4

RPC in avalanche mode: 

RPC in avalanche mode <Q> and Q behavior, 2% SF6 1.2 mm 1.6 mm 2.0 mm For all gaps Q/<Q> ~ 1    knee

RPC in avalanche mode: 

RPC in avalanche mode Eff and <m> vs pad spacing No any prominent dependence for 0.3 -1.0 mm spacings

RPC in avalanche mode: 

RPC in avalanche mode <m> vs anode thickness Should be as small as possible

RPC in avalanche mode: 

RPC in avalanche mode Eff and <m> vs beam incident angle No any prominent dependence for 900 - 450 angles

RPC in avalanche mode: 

RPC in avalanche mode Noise is increased as function of E 1.6 mm - ~0.2 Hz/cm2 1.2 mm - ~0.5 Hz/cm2  - 1.2 mm  - 1.6 mm  - 2.0 mm 1.6 mm knee 1.2 mm 2.0 mm Noise

RPC in streamer mode: 

RPC in streamer mode Ar/N2, IB and SF6 fractions were varied to find the mixtures with minimal <Q> and RMS/<Q> values for the 0.8, 1.2 , 1.6, 2.0 mm gaps. Features: 1. Ar and N2 are similar. Best range (0.1-0.2) 2. Best IB range (0.1-0.2) 3. On knee <Q>1.2~200pC, <Q>1.6~300pC, <Q>2.0~400pC 4. RMS/<Q> ~(0.5-0.6) 5. Efficiency ~95% for 1.2 and 1.6 mm  No 100% avalanche -streamer transition . 6. For 2.0 mm eff. ~80% due to larger Q 7. 2% SF6   <Q> by 2 times,  RMS/<Q> by 20% Ar(N2)/IB/TFE ~ 10/10/80 are the best for the streamer mode

RPC in streamer mode: 

RPC in streamer mode Typical Q and M distributions, 200 V above knee 1.2 mm gap, TFE/Ar/IB=80/10/10 FWHM=20% RMS/Q=0.6 No ways to suppress multi streamer tail

RPC in streamer mode: 

RPC in streamer mode 1.2, 1.6 mm gaps at knee with 2% of SF6 variation of Ar/N2 fractions thr = 50 mV best choice - Ar/N2=10% eff. ~95% RMS/Q ~ 0.6-0.8 Q ~ 100 pC for 1.2 mm Q ~ 150 pC for 1.6 mm

RPC in streamer mode: 

RPC in streamer mode 1.2, 1.6 mm gaps at knee Ar/N2 = 10% variation of IB fraction thr = 50 mV best choice - IB=10% eff. ~95% RMS/Q ~ 0.6 Q ~ 200 pC for 1.2 mm Q ~ 300 pC for 1.6 mm

RPC in streamer mode: 

RPC in streamer mode Eff for 0.8, 2.0 mm gaps best Ar10,IB10 mix thr = 50 mV eff. <70% for 0.8 mm due to lack of ionization eff. ~80% for 2.0 mm due to larger Q

RPC in streamer mode: 

RPC in streamer mode Eff, M and Q vs HV for 1.2 and 1.6 mm gaps Ar10 mix for different thresholds best choice - thr = 300 mV

RPC in streamer mode: 

RPC in streamer mode M and RMS/Q vs Q for 1.2 and 1.6 mm gaps Growth of both is seen with increasing of Q

RPC in streamer mode: 

RPC in streamer mode Eff ~95% and <m> ~ 1.2-1.3 for 300 mV thr 1.2 mm gap

RPC in streamer mode: 

RPC in streamer mode 1.2, 1.6, 2.0 mm gaps, thr > 50 mV Noise ~0.1 Hz/cm2 for 1.2 and 1.6 mm

Comparison of avalanche and streamer modes: 

Comparison of avalanche and streamer modes Eff vs M streamer Ar10 mix 100 mV thr 300 mV thr N210 mix 100 mV thr Avalanche - solid line There is some region for low M (1.1-1.2) with eff~95% comparing with avalanche mode

Comparison of avalanche and streamer modes: 

Comparison of avalanche and streamer modes Rate capability streamer <4-5 Hz/cm2 avalanche <300 Hz/cm2 It is hard to work in streamer mode even for usual beam conditions Streamer is suitable only for very low rates like e+e- FLC

Comparison of avalanche and streamer modes: 

Comparison of avalanche and streamer modes As example, for 1.2 mm gap

Comparison of avalanche and streamer modes: 

Comparison of avalanche and streamer modes Avalanche mode is preferable due to: 1. higher efficiency (>99%) 2. smaller charge deposition (~102) - no observed ageing effects - higher rate capability (~102)

Conclusion: 

Conclusion 1. RPCs in avalanche mode are in favor to be used for FLC DHCAL 2. Working conditions: -gas gap 1.2 -1.6 mm - gas mixture TFE/IB/SF6 - average induced charge ~2 pC (107 e) - efficiency > 99% - pad multiplicity ~ 1.4 - rate capability ~ 100 Hz/cm2 - noise 0.2-0.5 Hz/cm2 3. RO electronics (thr>1-2 mV) is challenge ( cost should be at ~0.1 Euro level)

Conclusion: 

Conclusion Geant3 simulations It seems eff down to 80% does not hurt resolution much It seems <m> up to 1.4 does not hurt resolution much Preliminary Preliminary

Conclusion: 

Conclusion 1 m3 DHCAL prototype SS cup SS cup Cathode strip PB Anode pads PB glasses Al bar spacer g. v. 40 layers 2 cm steel absorber 960x960 mm2 RPC active area 1x1 cm2 pads 96x96 = 9216 pads/layer 368640 pads in total Construction and beam tests within 2 years

RPC in streamer mode: 

RPC in streamer mode 1.2, 1.6 mm gaps at knee variation of Ar/N2 fractions thr = 50 mV best choice - Ar/N2=10% eff. ~95% RMS/Q ~ 0.6 Q ~ 200 pC for 1.2 mm Q ~ 300 pC for 1.6 mm

RPC design for DHCAL: 

RPC design for DHCAL

RO electronics: 

RO electronics General scheme for 64 channel read out Two parts: Conditioning (analog) FPGA (digital)

RO electronics: 

RO electronics Requirements for FLC - All FEE should be on board - One channel < 1 cm2 - Anode PCB with pads should be multi layer PCB

RO electronics: 

RO electronics Steps for design and usage For 1 step - 2 approaches A) single PCB IC conditioning on pads within 1 cm2, FPGA on side of anode PCB B) two PCBs one for signal transportation from pads on side, second for all FEE

RO electronics: 

RO electronics Two versions for conditioning 1) IHEP version Preamp(IC)+comparator(IC)+gate(IC) each single channel IC 2) Minsk version special 8 channel chip included preamp, amp and comparator (very sensitive ~0.2 A)

RO electronics: 

RO electronics IHEP version, threshold > 0.5 mV Preamp, 10x Comp, 5 mV Gate, 100 ns TTL pos signal for FPGA Was tested successfully in Dec02 beam run as separate board (approach A)

RO electronics: 

RO electronics Test of IHEP version with RPC signals using  source Comparison of counting rate with old, calibrated FEE 1 mV threshold is achieved

RO electronics: 

RO electronics Test of Minsk version with RPC signals using  source Comparison of counting rate with old, calibrated FEE 0.5 mV threshold is achieved

RO electronics: 

RO electronics IHEP version, Approach A single 6 layer anode PCB for 64 channels Layer meaning 1. Anode pads 2. Shield GND 3. Signal CMOS lines 4. Power layer 5. Shield analog GND 6. Component layer FPGA on the same PCB, out of RPC ALTERA EP1K50 is used as FPGA All components (ICs) in SOT-23-5 packages

RO electronics: 

RO electronics Anode pad layer component layer IHEP version, Approach A single 6 layer anode PCB for 64 channels

RO electronics: 

RO electronics One channel 1x1 cm2 Component layer Six PCBs are ordered IHEP version, Approach A single 6 layer anode PCB for 64 channels

RO electronics: 

RO electronics Preliminary tests do not allow to reach lowest threshold - below 5 mV threshold it is channel generation due to the cross talk from the gate with TTL signal to pads. IHEP version, Approach A single 6 layer anode PCB for 64 channels

RO electronics: 

RO electronics 1. Play with IHEP version (single PCB) to find possible minimal threshold for one RPC plane (64 channels) 2. Use the two PCB approach further - to simulate chamber PCB with signal transportation to find proper impedance etc - to design FEE PCB based on Minsk version 3. For beam tests of 20 layer electromagnetic cal (Dec03) use Minsk version of two PCB approach 4. It is proposed to use for the 1 m3 prototype also two PCB approach. It is planned together with Minsk to design special 16 simplified chip based on on the current Minsk 8 channel analog chip. PLANS

R&D plans: 

R&D plans December 2003 Beam tests of the 20 layer 'electromagnetic' calorimeter with the 64 channel small RPC planes and the 2 cm steel sampling June 2004 We are ready for Production and assembly of RPC planes for the 1 m3 DHCAL prototype

R&D plans: 

R&D plans  R=4.5 cm 98% trans. cont.  Gas electromagnetic calorimeter with 20 layers sampling: 2 cm steel + 0.65 cm RPC plane sensitive area 9x9 cm2 ( 8x8 pads of 1x1 cm2, 1 mm spacing) GEANT3 simulation of transverse containment pe=1,10,40 GeV/c Trans. cont. Diff. distr.