STAR NEWSLETTER NUMBER 47

1. From the Spokesman:

Reported by John Harris

This contribution to the STAR newsletter will be somewhat longer than usual. There are presently many exciting and important things that are going on in STAR, as evidenced at the collaboration meeting last month, many more that are planned and many others that need to go on. I will address various aspects of some of these below.

By far the most exciting news this month (and in the recent past) for STAR has been completion of the TPC assembly and the start of testing. The TPC mechanical assembly is now complete with all readout sectors installed. The insulator gas, TPC gas and laser systems are also complete and ready for operation. Two sectors are fully-equipped with front-end electronics for the testing and reconstruction of cosmic ray tracks.

Last week, the STAR TPC high voltage system for the drift field was tested for the first time. After tests with helium and argon for the insulating gas, the high voltage was raised to 37 kilo-volts using nitrogen as the insulating gas. This is ~25 percent above the anticipated operating voltage for argon-methane (P-10). At these voltages the currents monitored in the resistor chains were minimal (nano-amps).

It should be pointed out that the goal was to operate the TPC at voltages up to 55kV with nitrogen insulator gas. This would insure safe operation at 85 kV with C2F6 insulator gas for possible future use with helium-ethane as a chamber gas. The tests were stopped at a voltage of 37 kV, due to the observation of corona-like emission in the outer field cage. The cause of this corona will be investigated after some initial cosmic ray testing of the TPC using P-10 as the chamber gas. The cause of the emission is unknown, but is not critical to the operation with P-10.

Within a few days of the high voltage test, the TPC was filled with P-10 gas and the drift gas flow and purity tested. The levels of oxygen and water contamination were near the target values of 20 parts per million (ppm) for oxygen and 40 ppm for water, which depend upon the flow and purifier operation. Then everything was ready and the cosmic-ray tests began, using 8 slats of the future central trigger barrel for triggering on cosmic rays. In addition to the LBNL contingent on STAR, a number of other STAR collaborators are on-hand. Cosmic-ray tracks are being reconstructed in order to determine in detail the response and performance of the TPC. A very significant step has been taken with the successful operation of the TPC and a number of others will be necessary to fully understand the performance and operation of the TPC. This work will continue at LBNL until the TPC is "packed" for shipping to BNL in early November.

I would like to congratulate Howard Wieman, Russ Wells and the TPC crew for their TPC success. I would also like to congratulate Fred Bieser and Spencer Klein along with the TPC front-end electronics group for successful operation of the electronics. I thank the STAR group at LBNL for successfully leading the TPC effort and testing (hardware and software), and congratulate them for the positive results so far.

On another topic, as a result of the review of the RHIC Computing Facility (RCF) on July 29 - 31, we in STAR management have called together a task force to address the computing issues relevant to STAR. This information will then be provided to the RCF management to inform them of STAR's computing needs, in particular those STAR requirements which drive the design of the RCF. Peter Jacobs has agreed to chair the task force. The charge to the task force is 1) to develop the offline computing requirements (including CPU, data storage on shelf tapes, robotic tape storage and disc, network bandwidth between storage and local and remote processors, etc.) for the analysis and interpretation of STAR data for year 1 and beyond and 2) to develop a reasonable scenario for event reconstruction, data analysis (including DST and micro-DST stages) and simulations needed to correct and interpret the data and for comparisons of data with appropriate physics models. The task force is asked specifically to indicate analysis topics or specialized analysis approaches that drive particular computing requirements, so that the STAR Collaboration can consider their impact on offline computing requirements and resources. The task force, which is almost complete, will consist of STAR physicists with recent experience in the various physics analyses that STAR is pursuing. It will provide a draft document to the spokesman by October 1 and a final report by November 1.

As announced at the STAR Collaboration Meeting last month, there will be a STAR review of Software and Computing. Tim Hallman has agreed to chair the committee, whose charge is a comprehensive review of the software and computing project and plans in STAR. The charge includes an evaluation of "the requirements, the design and the ongoing (hardware and software) implementation plan for the STAR Software and Computing Project." This review is expected to take place in early October at BNL.

Bill Llope has asked to step down from his responsibility as the STAR simulations project leader. I would like to thank Bill for his work in this role. We have asked Peter Jacobs from LBNL to lead the simulations effort and he has agreed. We look forward to Peter's leadership in the simulations effort of STAR.

As an editorial statement, I would like to point out that with only two years before we start to take data in STAR it is rapidly becoming critical that we understand in detail how to take and analyze data in STAR from the hardware, software and physics perspectives. This requires additional manpower, commitment and attention to STAR within the collaboration. On the hardware side, the construction plan and effort for the baseline detector project have been established (long ago) and the detector components are in the final stages of construction. Next will be to complete the assembly and then to understand the performance. Already at the stage of testing and understanding the TPC performance, it is important to have the track reconstruction software ready. From the TPC detector tests will (should) come not only an understanding of the detector, but optimization of the reconstruction software. The simulations and analysis software projects now need real concerted efforts to get us to the first day of data-taking and to extract the physics, respectively. It is time for all members of STAR to get on board and to start to focus on the many important details that are necessary for successful data acquisition and physics in STAR. I urge STAR members to contact the appropriate project leaders in STAR and your institutional representative on the STAR Council to discuss how you will contribute to the effort over the next two years. There is a particular need now for overall STAR software development work (contact Torre Wenaus at BNL, STAR Software and Computing Leader), simulations software (contact Peter Jacobs at LBNL), analysis software (contact Lanny Ray at Texas), and physics simulations (contact me to be directed to the appropriate physics working group convenor). If you have any questions or comments on this, I welcome your communicating with me about it.

Excerpted from the STAR Monthly Report for July 1997.

Project Management Summary and Highlights

At the same time, preparations are being made with both the Air Force and the air-port manager at Francis Gabrisky Field (West Hampton airport on Long Island) to fly in the TPC via Air Force C-5 Transport. TPC Engineering Management will pay a visit to Gabrisky in early August. The transport is to occur on or around November 1st.

FEE hit a snag in July. The SCA chips were temporarily "lost" when a strike occurred at the packaging factory in Bangkok, Thailand. The chips were recovered unharmed and are at another uneffected factory; they should be completed at month end. This minor delay (but major scare) will not impact chip testing nor board loading in any significant way.

Several critical manpower problems are beginning to be solved. Torre Wenaus came on board July 7th as Computing Subsystem Manager. Also, a temporary solution has been found for DAQ's engineering needs and a long term solution is in the works (an offer went out).

A Collaboration meeting took place at BNL in mid-July. A significant fraction of this meeting was taken up with discussions of software progress, plans, and issues.

TPC Summary and Highlights

Magnet Summary and Highlights

Electronics Summary and Highlights

Systems Test

Following the move, data continued to be taken at a variety of dip angles, and an improved clock and DAQ busy board have been built which includes some error handling for when readout boards/DAQ lose synchronization.

Analysis of previous month's data continued. Fits to tracks with a large crossing angle exhibit an asymmetry in their residuals, with alternate pad rows showing alternatively positive and negative average residuals. These residuals switch sign depending on the sign of the crossing angle. This problem is not yet understood, but it may be related to the crosstalk mystery described last month.

Front End Electronics (FEE)

The first SAS chips are on their way back from a different packaging house. As soon as we establish that the chips are wire-bonded correctly, mass production will begin.

FEE Board production is underway at MPI Munich. Likewise, power supply chassis production continues at LBNL. The raw power cables have arrived at LBNL.

The FEE final design review, on July 2nd, went very well. No real problems were found, and no action items were created. The most prominent comment was to put hard work into making the readout board G-link transmitter run at 60 Mhz. There are ideas on how to improve this speed, but 60 Mhz may not quite be reached.

Data Acquisition (DAQ)

The SCI evaluation makes the SCI technology look very promising. Outstanding issues are delivery schedule of the PMC version of the hardware (no firm date yet) and a lack of willingness on the part of the vendor to supply technical information which we consider essential to STAR's effective use of the product.

The production quantity of the DAQ ASIC has arrived at BNL. Testing of small quantities from each of the 3 production batches is planned to supplement the testing performed at the foundry.

Trigger

An agreement has been made to complete documentation of the trigger software modules used in the simulation. All the software modules are in STAR standard STAF form at this time.

Slow Controls

Controls for various sub-systems (using EPICS version 3.12.1) are in place for the coming Cosmic Ray Test.

The VME modules for the above systems are housed in 4 separate VME crates (each with one MVME167 processor) providing good system segmentation. Two of the four crates are the 6U Wiener crates. One VME Crate remains in the "system test" room at LBNL (Bldg. 70-268) for development work.

Computing Summary and Highlights

At the collaboration meeting the working group report on the computing plan and model was presented and discussed. The assumptions underlying the report are now being revisited: we have begun reworking the resource requirements estimation and refining the data processing chain (processing stages and their output content, volume and uses; meaning and role of data mining). RCF management agreed to quickly convene an event store working group to come to a design decision ASAP, as recommended by the report.

Manpower remains critically low. BNL is trying to fill five on-line and off-line positions. Institutes have been asked to increase the software participation of their members. In particular, startup is now close enough to increase involvement of post-docs and students. A task list is being assembled (in the context of developing an updated schedule and project plan) to provide concrete suggestions and expressions of need.

Institutional representatives for computing were requested to act as contacts on issues of local hardware and software requirements, code distribution, networking and connectivity, licensing, etc.; about 20 institutes have so far named a contact.

A comprehensive review of computing, on-line and off-line, will be held around the end of September. A draft charge has been written.

Simulations

Analysis Software

The basic TPC fast simulation/reconstruction chain was made operable on the aix operating system (ibm). This chain includes packages tpg, tfs, tpt, tte and tid (TPC particle ID). Array overflow checks were included in the TPC tracking evaluation module (tte) and the new package installed on the irix and sun4os5 platforms.

Software Infrastructure

An additional set of commands (mv, cp, ln, rm) have been added to the dui package in STAF and will be part of the SL97a release. Additional functionality is being added to the hbook interface (tnt) by Dave Morrison of PHENIX as part of the common RHIC software effort.

A Tcl scripting language interface to STAF, as a replacement for KUIP, was developed and demonstrated at the collaboration meeting. A decision on when to replace KUIP with Tcl in the released version of STAF has yet to be made.

On-line software

STAR Collaboration Meeting, Week of January 5th, 1998. Location TBD. For further information, please contact John Harris at Yale.

Greetings from Long Island. August was fairly typical here on Long Island weatherwise. The daily high temperatures were in the middle 80 degree F range, with typical daily low temperatures in the 70's. Interspersed through the month we had the usual assortment of thunderstorms. As August draws to a close the vast majority of the deciduous trees still have all green leaves, but here and there one can find a few trees where the leaves are just beginning to change colors.

Out at the Assembly building work continued on the assembly of the STAR magnet. The leads for the magnet coils were clamped in place and the installation of the top fifteen backleg return bars for the magnet was started.

In early September the installation of all of the backlegs should be completed. The mounting fixtures for the EMC are scheduled to start arriving at the Assembly building in early September. The installation of these EMC supports into/onto the magnet will continue for a number of months. Also in September the personnel access labyrinth on the East side of the alcove leading from the AB to the WAH will be completed, and the construction of the labyrinth on the West side of the alcove will begin.

The duration of the appointment is two years with possibility of renewel for a third year. Applications should be sent to:

7. New STAR NOTES since the last Newsletter