Presentation is loading. Please wait.

Presentation is loading. Please wait.

June 6, 2001 Paul O’Connor, BNL ATLAS Cathode Strip Chambers.

Similar presentations


Presentation on theme: "June 6, 2001 Paul O’Connor, BNL ATLAS Cathode Strip Chambers."— Presentation transcript:

1 June 6, 2001 Paul O’Connor, BNL ATLAS Cathode Strip Chambers

2 CSCs are the forward precision muon system of ATLAS 32 four-layer chambers 2.0 < |  | < 2.7 |Z| ~ 7m, 1 < r < 2 m 4 gas gaps per chamber 31,000 channels 32 four-layer chambers 2.0 < |  | < 2.7 |Z| ~ 7m, 1 < r < 2 m 4 gas gaps per chamber 31,000 channels IP

3 Principle of operation Determine muon position by interpolating the charge on 3 to 5 adjacent strips Precision (x-) strip pitch ~ 5.6 mm Measure Q1, Q2, Q3… with 150:1 SNR to get  x ~ 60  m. Second set of y-strips measure transverse coordinate to ~ 1 cm. Position accuracy unaffected by gas gain or drift time variations. Accurate intercalibration of adjacent channels essential. S = d = 2.54 mm W = 5.6 mm

4 Chamber construction detail

5 Gas Gain vs Anode-Cathode gap uniformity Gas Gain, to first order, not important for Chamber precision (relative charge measurement) Must be reasonably uniform for operational stability Assume need to keep gain to +/- 30% -  150  m gap tolerance

6 Flatness of two gaps in Module Zero Early Panel while still developing technique Min to Max variation ~6-7 % One of the later panels Min to Max variation ~4%

7 Gain measurement station

8 Corresponding Gas Gain Variation Gas Gain Variation tracks well anode- cathode gap variation Agrees well with calculations Confidence for robust operation 

9 CSC Electronics On-chamber (BNL) –Amplifer/shaper(rad-hard custom ASIC) –Analog memory( ‘’ ) –ADC(rad-hard COTS component) Off-chamber (UCI) –Generate all control signals needed on chamber –Sparsify raw chamber data, build event, interface to T/DAQ –Initialization, calibration, monitoring Interconnect and services (BNL) –~ 1000 optical links, ~ 800 Gbits/s bandwidth overall –LV distribution, 2.6 kA @ 7V at chambers (18 kW) –HV distribution –Liquid cooling for on-chamber electronics –(gas system)

10 Reliability Goal: < 0.1% data lost due to hard or soft fails Strategy: on-detector electronics boards must be as simple and robust as possible –Minimal parts list: ~ 10 component types –All intelligence to reside off-chamber (SCA Controller in ROD) –ASM boards to have no –Configuration registers –Programmable logic –FIFOs –RAM –Counters –State machines –Thorough radiation test of all components –ESD protect inputs at board and chip level –ATLAS policy on grounding and shielding –Extra attention to amplifier stability –No DCS –Minimal monitoring

11 On-chamber Electronics Chain To/from ROD Amplification, Sampling, digitization, and Multiplexing into optical fiber performed on detector. 70 ns shaping 10 bit dynamic range 3 custom, radiation-tolerant ASICs

12 Electronics Location in Faraday Cage ASM- I ASM-II

13 Readout organization – one chamber ASM Board 4-layer Chamber module y -strips x-strips bidirectional optical fiber links ASM Board ASM Board ASM Board ASM Board 864 channels per chamber Optical links: 10 downlinks, 5 uplinks per chamber

14 ASICs for CSC Electronics CHIPTYPEFOUNDRYTECHNOLOGY P/SanalogAgilentAMOS14TB SCAmixedATMELDMILL MUXdigitalTBDTBD POS. VREGpowerSTRHBip1

15 Preamp/Shaper (P/S) Technology: Agilent 0.5 um CMOS Function: 12-channel preamplifier and 7-pole shaper No. of chips required: 5,120 (full scope of 64 chambers) MPW runs: 3 Tested with full size chambers in beam Radiation testing: preliminary ionizing test to 1 Mrad, very encouraging result Milestones: –Beam test09/01 –Final DR and PRR10/01 –Production12/01

16 SCA Technology: DMILL Function: analog memory Design responsibility: LARG No. chips required: 5,120 for full scope (64 chambers) This is the same chip used in LARG FEB Design was modified at our request to allow pin-selectable “MUON mode” Effectively increases the readout rate by X6 –4 channels x 3 gains => 12 channels –Duty cycle 50% => 100% per SCA –Read clock 5 MHz => 6.67 MHz Tested in MUON mode in lab, 3/01- 5/01 LARG in control of radiation qualification, procurement, and production testing Production forseen for Jan. 02

17 MUX Function: data concentration between ADC and G-link Technology: TBD No. of chips required: TBD Milestones: –Start design10/01 –MPW fab02/02 –Final DR10/02? Voltage Regulator LHC4913 positive voltage regulator being developed in framework of RD-49 by ST Microelectronics Catania Location: ASM-1 and ASM-2 Quantities required: 640 ea. 5V and 3.3V (for 32 chambers)

18 ASM 1 Prototype with P/S chip

19 Noise vs. capacitance Linearity Simulated:  Measured: o

20 Preamp/shaper 60 Co irradiation results Supply currentGain Noise Waveform

21 ASM-2 prototype with SCA chip

22 ASM-2a waveforms

23 ASM-2a Gain and Linearity

24 ASM-2a Noise Measurement IC50 preamp/shaper (unpulsed) into SCA on ASM-IIa board Channel 1 results displayed 24 SCA cells written and read out Measurement repeated for 1000 cycles. Mean and s.d. of each cell Histogram of all cells

25 CSC Off-detector electronics optional

26 ROD Functions : –Generate control signals and transmit to chambers –Receive and process raw chamber data –Assemble events and transmit to Concentrator –Calibration and monitoring –ROD serves 2 chambers (1920 channels, 12.8 Gbit/s) 9U VME plus 220 mm Transition Module, 100W Modular design using TI DSP-based daughterboard (GPU) –70 x 70 mm, 3W –> 100 Mword/s data BW –Component cost ~ $300 –12 GPUs per ROD Status : –GPU: working prototype on hand –ROD: motherboard layout started 5/1/01 Milestones : –Prelim. Design Review 05/01 –First prototype in hand 08/01 –Integration with ASM-204/02 –Final design review 10/02 –PRR 07/03

27 Low Voltage Power and Cabling Design for full scope (64 chambers) Uncertainties: –Power supply location –Cable run details –Local regulator characteristics Working assumptions: –Space for LV supplies available on UX15 floor/gallery –Use LARG-like supplies if neutron flux too high for standard DC-DC converters –25m run to patch panel on outside of endcap wheel –12 m from edge of endcap to chamber pigtail Total electronics power: 18400 W (2630A @ 7V) Cabling: 64 pairs AWG 4, 480 cm 2, 1000 kg 100 W dissipated per cable, 6400 W total for full scope Fiber optic cables –960 fibers from USA15 (120m run) –90 cm 2 cross section, 860 kg

28 Prototype Recirculating Gas System System design similar to final Will be used in test beams Important for long term aging tests foreseen using X5/GIF Design and Construction by L.Kotchenda (PNPI) and his crew at BNL

29 Support frame, y-strip readout location, Faraday cages

30 High Rate Performance Overall background rate 10 7 Hz per chamber 50% charged particles, 50% neutron and  Charged particle background is rejected by –Timing window around trigger (out of time) –Or by pattern recognition of non-projective tracks Neutrals can deposit high charges –50% of neutrals above Q FS –1% of neutrals above 6* Q FS Neutrals produce short-range electrons usually confined to 1-layer But a neutral hit anywhere in chamber induces crosstalk onto all strips by anode-cathode crosstalk:

31 Beam Test Measurments

32 Summary Chambers ready for start of production Two of 4 ASICs ready for production MUX and Voltage Regulator ASIC must be resolved as quickly as possible Location of power supplies must be resolved System test with Module 0 chamber, ASM boards this Fall System test of optical links and ROD next Spring

33 The CSC Team BNL: V. Polychronakos P. O’Connor A. Kandasamy S. Junnarkar A. Gorodeev V. Tcherniatine V. Gratchev UCI : A. Lankford D. Stoker M. Schernau S. Pier D. Hawkins N. Drego J. Dailing B. Toledano


Download ppt "June 6, 2001 Paul O’Connor, BNL ATLAS Cathode Strip Chambers."

Similar presentations


Ads by Google