Co-convenors L. Moroni (Milan), J. Russ (CMU/Fermilab)
1) Establish the feasibility of producing pixel detectors of suitable size, e.g., 30-50 micron x 100-200 micron rectangular pads, in time to begin system studies in 1999-2000.
2) If feasibility is established, then initiate the formation of a collaboration to start pixel development specifically for C0 work on a timescale permitted by funding, personnel availability, and facility availability.
3) To identify interface problems between the tracking system hardware and the needs of the trigger group, to ensure that the tracking system can operate with an online vertex trigger.
C0 Workshop Precision Tracking Group
Thursday, December 5, 9:00-12:00
Agenda
9:00 Introduction - L. Moroni (Milan) Co-convenor
9:05 Perspectives on Silicon Vertex Detector Research at Fermilab - G. Spalding (FNAL)
9:30 Operating experience with the Delphi Pixel Detector - F. Ledroit (CPPM)
9:50 Review of C0 Vertex Tracking Requirements - J. Butler (FNAL)
10:15 Pixels at C0: analog vs digital in speed, power, and radiation hardness - S. Shapiro
10:40 Rad hard detectors: silicon and diamond - M. Zoeller (OSU)
11:05 Pixels for CMS: application to C0 - G. Grim (UCDavis)
11:30 Detector requirements for C0 momentum determination - W. Selove (Penn)
1) Existing pixel development efforts are based on other drivingforces, chiefly LHC. This has three consequences for C0:
a) There is a substantial amount of funded activity for pixel development on the world scene to which C0 efforts might be joined. We heard from CMS, Atlas (Marseilles group), and the Shapiro efforts. Other active groups include Parker and Atlas(LBL).
b) The design goals of the extant efforts are not matched to the needs of C0 in rate (25 ns bunch crossing for LHC; 132 ns for BTEV); pixel size; or time scale.
c) Data flow architectures in the LHC groups are not matched to the needs of online vertex reconstruction.
2) The prospects for having a pixel precision tracking detector for C0 appear to be good, provided that Fermilab organizes a coherent pixel collaboration to focus on producing optimized devices suited to the needs of C0 and then BTEV. The groups now engaged in pixel development are interested in participating at some level. The major issues for the C0 collaboration are the following:
a) First, the formation of a collaboration between Fermilab and the other institutes to produce devices for C0. This has the standard problems of resource allocation, coordination, and facilities for production and testing devices.
b) Agreement about a central research location for pixel development, e.g., at Fermilab. The scope of the project requires many of the technical attributes of the lab for mounting, cooling, and the like. There is good local expertise from Upgrade work. Fermilab has useful facilities already on site. Futhermore, the close coupling between pixel detector data flow architecture and trigger design requires tight coordination within the larger C0 collaboration. Experience with other projects of similar scale shows that this works best with management centered at a laboratory, even though it must have major participation by universities and other institutes.
c) Development of a C0 Experimental collaboration to set the specifications for the entire vertex tracker, trigger, and DAQ system. Because these parts are intimately related in the C0 project, one cannot move beyond the initial proof of principle demonstration of pixel devices without a clear list of requirements for the system design. This requires the approval of an experimental setup for C0 and a budget.
3) The time scale to launch various steps in the pixel program is NOW.
a) The various pixel groups can profit from a pre-LHC involvement like the C0 effort to confirm radiation- hardness, test large-scale systems, and gain experience with assembly and operation. The C0 timescale is well-matched to this goal, if we begin an integrated effort now.
b) Now is the time to begin a development effort for C0. The Delphi million-pixel system - the largest now in operation - is <1% of a BTEV system. The Delphi system required 3 years of effort to move from prototype to installation. There are enough new problems for the C0 design that a 3-4 year time scale even for a few planes with trigger capability will be very tight. Although large-scale funding is not available until 1999, initial efforts to coordinate extant efforts and refocus some work toward C0 requirements have to begin immediately if the project is going to succeed.
4) Continued development of detectors should be a piece of this program.
a) Diamond detectors have made progress far more rapidly than initial silicon device efforts a decade ago. For radiation and cooling reasons, diamond detectors are very attractive for BTEV. Initial C0 detector architecture should not wait for diamond, but this growth path should be encouraged by having C0 participation in diamond R&D projects.
b) Radiation hardness of pixel detectors is of great interest for BTEV, as well as TEV33 and LHC. The C0 activities offer a unique development path to ensure that the technologies are understood and that HEP activities have access to the necessary processing methods.
1) Pixel detectors appropriate for vertex trigger capability appear to be feasible, based on a decade of pixel activities at other labs.
2) Assembling a pixel detector collaboration for C0 is feasible. There are interested groups. It requires the lab to define a focus for the effort by instituting a pixel research effort here. It requires the C0 collaboration to inspire the effort by working with the lab to develop an approved research program and funding schedule.
3) Improved devices, such as diamond, can be integrated into the development path as they are available. However, the large issues of system integration, mounting and alignment, and DAQ and trigger strategies are experiment-driven, not specific to any pixel technology.
1) Fermilab should initiate a Pixel Detector research group to work with a C0 Pixel Collaboration, to be formed as soon as possible.
2) A C0 Collaboration should be formed, to present to the laboratory a specific experimental proposal about which the system design requirements for a precision vertex tracker and trigger can be built. Various models for this have been discussed at this workshop.
3) Active work to produce a suitable basis chip for C0 pixel development, e.g., a 1 cm x 1 cm chip having rectangular pixels with analog readout, should be initiated during 1997. This will serve to crystallize the collaboration and give an initial focus to the effort. Actual specifications for this first-round device can be developed using one of the trigger models now available.
Updated 1/31/97 - M. Smith