BNL December 1 2005 The STAR MRPC TOF

BNL December 1, 2005 The STAR MRPC TOF Project Geary Eppley, with Jo Schambach, for the TOF Group

Design considerations • The integration space available for TOF in STAR is that currently occupied by the CTB. • The space is about 2. 2 m from the beam line: requires precision time measurement. • The clearance between the TPC and EMC is about 10 cm. • The space has a uniform 0. 5 T magnetic field parallel to the beam. • There is limited space for “nearby” electronics outside the magnet steel.

MRPC development • • Multigap glass RPCs developed in the late 1990 s at CERN, C. Williams, et. al. STAR TOF group designed STARspecific MRPC modules that would fit existing CTB trays and tested them at CERN in 2000 & 2001.

MRPC prototypes in STAR • • • One tray of MRPC modules has been installed in STAR for Runs 3, 4, & 5. An improved tray design each year but some modules have been installed for all 3 years. Runs 3 & 4 used CAMAC electronics. Prototype HPTDC based electronics used in Run 5 NINO/HPTDC based electronics for Run 6/7.

Electronics integral to tray • • 95% of the electronics resides on the tray inside the cover 4 external electronics boxes located on the magnet steel connect the 23 k-channel system to STAR trigger and DAQ

Tray specifications • • There are 32, 6 -channel modules per tray. The active area covers about 87% of -0. 9<eta<0. 9 Occupancy in central Au. Au is ~12%. HV provides +-7 k. V. Gas mixture: 95% r 134 a, 5% isobutane. Average noise rates are ~15 Hz/channel with greater than 100% correlation. Detection efficiency is ~95%.

Time measuring • • • Absolute times are recorded and buffered by the HPTDC for all hits above threshold. A single 40 MHz oscillator provides the clock for the 2882 HPTDCs in the system. Both start and stop detectors use the HPTDC to record times eliminating systematic uncertainty from the choice of counter. Each HPTDC has a 21 bit clock counter providing a 51 us clock period. The HPTDCs are reset by a signal with a common origin at the beginning of each run giving each HPTDC a different phase that may be learned from the data. Since this phase is always the same, the phase becomes a powerful tool to monitor the integrity of the data through buffering, trigger matching, event building, transmission to DAQ, and global event building.

Total time resolution from Run 5: 200 Ge. V Cu. Cu, 105 ps 62 Ge. V Cu. Cu, 124 ps

Time resolution by channel

Timing results, Runs 3, 4, &5:

Manufacturing overview • • Module production and test, electronics production and test, and tray assembly and test each have a duration of ~2. 2 years. Module production begins ~3 months before the start of tray assembly and electronics production ~1. 5 months before the start of tray assembly. Modules are thoroughly tested in China with integral HV wire and signal cables. Electronics cards are tested in groups of 17, 1 tray, with their integral cables and sent to the tray assembly site as a set. Modules are assembled into trays, the electronics added, and tested for 2 -4 weeks as a complete detector unit. Tested trays are shipped to BNL as complete detector units ready for insertion into STAR. Trays are re-tested at BNL: HV current, noise rates, gas leak test Test beam at BNL for selected trays?

STAR TOF publications: • Physics – Open charm yields in d+Au collisions: STAR Collaboration, Phys. Rev. Lett. 94 (2005) 062301. – Cronin effect of identified particle at RHIC: STAR Collaboration, Phys. Lett. B 616 (2005) 8. • Technical – – – – B. Bonner et al. , Nucl. Instr. Meth. A 478 (2002) 176. M. Shao et al. , Nucl. Instr. Meth. A 492 (2002) 344. W. J. Llope et al. , Nucl. Instr. Meth. A 522 (2004) 252. F. Geurts et al. , Nucl. Instr. Meth. A 533 (2004) 60. J. Wu et al. , Nucl. Instr. Meth. A 538 (2005) 243. Y. Wang et al. , Nucl. Instr. Meth. A 538 (2005) 425. Y. E. Zhao et al. , Nucl. Instr. Meth. A 547 (2005) 334.

TOF electronics schematic

Electronic board status Board Current Status Future Plans Timeline TPMT Used in Run 5 Rev design: add discriminators Ready for production by Q 4 FY 06. TINO Prototype designed and built. Testing with cosmics. Test in test beam. Decision milestone by Q 2 FY 06. Production by Q 4, FY 06 TDIG Used in Run 5. Rev design by Q 3 FY 06: delete comparators & 1 TDC, add L 0 -mult. Production by Q 4 FY 06 TCPU Used in Run 5 Rev design: add TCPUTHUB I/F, delete SIU, TCD I/Fs Production by Q 4 FY 06 THUB Prototype currently in preliminary design stage Prototype with 3 I/F channels, redesign after bench test Production Q 1 FY 07

Interface to L 0 trigger • • • Provides multiplicity at 9. 4 MHz with <~700 ns latency. The multiplicity range is 0 -12 for each half-tray, where a bit is added to the sum if any of the 8 TOF channels in a NINO chip is above threshold. The multiplicity sum is formed asynchronously and sent to L 0 where it is gated and readout either by the current CDB or by the DSMI directly. The point of this is that the RHIC clock is not used anywhere by the TOF electronics (other than to readout trigger commands). It is not clear that the noise rates will be low enough for the TOF system to be a useful trigger for UPC. If it does work well enough for UPC, it will still take an enormous amount of manpower to calibrate and commission such a trigger. The barrel EMC is already in L 0 and has a granularity of 300 compared to 240 for the CTB. It could serve as the complement to the ZDC for centrality triggers.

Interface to L 2 trigger • • The intent is to provide a 23 K bit map of the TOF hits to L 2 for each “L 0 accept” command At the STAR review of TOF in April 2004, the L 2 connection was to be implemented by sending this information to the TOF DAQ receiver which would pass the information to L 2 over a network connection. This plan has since been dropped by STAR DAQ. The trigger group is now implementing a custom-designed fiber connection (STP) from L 0 to L 2, and TOF could send its information to L 2 over the same type of connection. There is no budget in the TOF project to implement this connection, however, the THUB card has not yet been prototyped and it might be relatively straightforward and low-cost to add the necessary components to THUB for an STP fiber connection

L 2, continued. • • STAR Trigger could use the CERN/ALICE DDL link that will be the fiber interface between all STAR detector subsystems and DAQ in the upgraded STAR detector to receive L 2 trigger data from detector subsystems. A similar scheme will be implemented in ALICE and just requires splitting the fiber. The software to receive data using the DDL protocol is already implemented in STAR DAQ and could easily be implemented in trigger as well. The existing trigger STP fiber concentrator is already full just from L 0 so a second custom concentrator would need to be installed to accommodate TOF, and any other subsystems.

Interface to DAQ • • • The TOF system needs to be faster than the upgraded TPC so as not to introduce any additional dead time. The TOF information is only useful in a STAR event if the TPC is also readout in that event. The system will handle L 0 accept commands at >10 k. Hz. The system will handle L 2 accept commands at >2 k. Hz. A design issue with a large expense consequence is the required size of the pre-L 2 buffer. Or, how many tokens are allowed in the system for events with TPC/TOF readout? The ALICE/CERN ALTRO has buffering for 3 events. Is STAR DAQ planning extra buffer space for TPC events somewhere to allow for asynchronous L 2 trigger processing? TOF is currently thinking of allowing for a 256 -event buffer and this costs ~$10 k for memory. Is there a minimum dead time between L 0 accept commands because of TPC readout limitations?

Current project status: • • The official closeout report from the DOE review of the TOF project in August has been received and all the issues dealt with. A revised management plan has been accepted by the DOE. Cosmic ray testing of the TINO front-end board is ongoing. Automated TINO board production has not been achieved. Trying a new vendor. Upgraded 38 -channel start detector under construction. Based on a recommendation from the review, the China TOF project built a few new modules with highly resistive electrodes. Cosmic testing is underway. Layout of the first prototype THUB is underway. Waiting for the DOE go-ahead to start the project.