In the following sections, brief descriptions and updates on progress from the subsystem managers (see Figure 7) are presented.
The integrated superconducting magnet system (ISMS) consists of the superconducting toroidal magnet which has been contracted to BWXT, the internal collimation and shielding modules which were designed and fabricated by UIUC, and a number of related projects that either interface directly with the magnet (vacuum windows, support carriage, external shielding), or else are required for performance verification and for final acceptance testing at UIUC.
The contract for the construction of the G0 superconducting toroid was awarded to BWXT on 8 May 1997. To date, the eight coils of the magnet have been assembled and essentially all of the cold-mass hardware has been fabricated. Work required for the acceptance testing of the coils is in progress. Following coil acceptance, it is expected that the assembly of the cold-mass will proceed through Q2, 1999. Fabrication of the magnet vacuum-vessel has moved onto the critical path, and there are material-selection issues that have yet to be resolved. If these issues can be finalized in a timely way, the complete system assembly and factory testing will proceed through Q3, 1999 and installation at UIUC will occur at the end of Q4, 1999. Final acceptance testing and various system trials will proceed through the year 2000. Installation at Jlab is not scheduled until Q3, 2001.
The eight Collimator Modules nest between the coils of the toroidal magnet and define the acceptance of the spectrometer as well as provide line-of-sight shielding between the target and the (external) detector array. Each module weighs about 4 tons, and is fabricated from aluminum and a special alloy of Pb. The engineering design and prototyping tests were completed in 1997, and the procurement of major components by competitive bidding began in January of 1998. The assembly of the first module at UIUC was completed on 29 June 1998, and BWXT was notified of its availability. Because BWXT was not prepared to assemble the cold-mass at that time, it was possible to reduce the over-all cost and risk of shipping by sending the modules in two groups. Four modules were completed by the end of August 1998; they were shipped to BWXT on 30 September, arriving 2 October. The second set of four were completed on 23 October, and shipped 4 November, arriving 6 November 1998. All eight collimator modules arrived without incident, and the work was completed one month ahead of schedule.
The major structural components of the Vessel support carriage were procured in June 1997. Fabrication of the minor components in the NPL machine shop will be completed in early 1999. Actual assembly of the carriage is a low priority, and will be managed by UIUC technicians in the first half of 1999. It is likely that some modification of the carriage design will be required by ongoing changes in the hall-C installation plan at Jlab.
The engineering design of the particle exit-windows is in its initial stages. The present concept is that a titanium window with a thickness in the range 0.016" to 0.020" will be employed. Issues of material properties, safety implications, fabrication techniques, and material availability are under investigation. It is anticipated that the design, prototyping and testing, and procurement will proceed through 1999. The actual fabricated windows will not be required until the spectrometer is installed at Jlab in 2001.
The final engineering design of the external Pb shielding-collar will proceed once the design of the downstream cryostat-head, to which it attaches, is finalized. The fabrication techniques will be those employed in the manufacture of the Pb components of the collimator modules. The actual collar assembly will not be required until the spectrometer is installed at Jlab in 2001.
The Optical Verification procedure will measure the locations of the coils and collimator modules as assembled in the cryostat vessel, at BWXT. The measurement will be made at 80K (LN2) at which temperature most of the ultimate thermal contraction will have taken place. The purpose of the test is to identify any necessary modifications to the assembly before the magnet leaves the factory. The Optical Verification will employ video photogrammetry, and requires the fabrication of eight window-boxes, each with three viewing ports, which will be installed over the cryostat exit-window apertures. The engineering design of the window-boxes has been completed, and procurement will begin in January 1999. Prototyping of the viewing port design is in progress. Each port consists of a central optically-flat camera-window isolated within a concentric Lucite flash-window. The viewing ports will be manufactured at UIUC/NPL in early 1999. It is expected that the ports and window boxes will be assembled and vacuum tested at UIUC in Q2, 1999.
The Magnetic Verification test is critical for establishing, under full operating conditions, that the toroid meets specified quantitative structural tolerances which reflect directly on its performance as a spectrometer. It is an integral part of the magnet acceptance procedure at UIUC. The instrumentation for measuring zero-crossings in components of the magnetic field external to the cryostat vessel will be fabricated by the TRIUMF group. Conceptual designs have been investigated, and detailed engineering design work and fabrication will proceed when the issue of vessel material has been resolved. It is anticipated that operational magnetic-verification instrumentation will be needed for final acceptance testing in Q1, 2000.
Work is proceeding on bringing the NPL Helium Liquefier into full operation. It is expected that an extended test run of the system will take place in Q1, 1999. Design and fabrication of cryogenic transfer lines is dependent on decisions that will be made about G0 cryolines for Jlab Hall-C and the degree of overlap between the two that can be exploited to reduce overall costs.
The G0 experiment will use a 20 cm long liquid hydrogen target which is axially symmetric about the beam direction. The construction of the target is the responsibility of groups at Caltech and the University of Maryland.
Significant progress has been made in the last year
on the design and fabrication of
the cryogenic target loop. Because relatively low energy protons
( 
MeV) exiting the target at 
are detected in the forward angle measurement, the amount of material
in the path of outgoing particles must be minimized.
Several target cells have been fabricated with wall and endcap thickness
of 7 +/-1 mil. The cells did not fail in hydrostatic burst tests
up to pressures of 140 psi, although some deformation of the endcap
occurred between 100 and 140 psi. These tests verify that our cells,
which will be tested to 100 psi before being incorporated into the cryotarget,
are rated over a factor of two above the maximum pressure in the target
during a catastrophic vacuum failure of the surrounding vacuum vessel while
the target is full of liquid hydrogen. The flow manifold, hydrogen cell and
helium cell have been constructed at Caltech. We have received the target
heat exchanger, which was built by a commercial firm.
We are now in the process of finalizing the design of the liquid
hydrogen pump, the one major component of the cryoloop left to be
built.
Other parts of the target group's responsibilities include the the gas handling system, which is designed to automatically relieve pressure in the system in the event of target boiling; the monitoring and controls hardware and software to read and record temperatures and pressures, set and record coolant flow to the heat exchanger, monitor the performance of the liquid hydrogen pump, and control the heaters to maintain constant heat load on the cryoloop; the interlock system, which is a hard-wired system designed to prevent hazardous situations from arising; and the transverse motion mechanism to align the target in the beamline. Varying degrees of progress have been made on these parts of the target system. The target gas handling system has been completely designed and all the components specified. All hardware for the monitor/control system has been specified, a number of the modules have been purchased and tested, and substantial progress has been made in writing and testing a PC-based LabView program to control devices and read, record, and display the target data. The service module, which is the vacuum vessel that houses the target support and gas and electrical feedlines and interfaces the beamline to the magnet, is not currently the responsibility of the target group. As no group has assumed responsibility for this device since it was dropped from lab infrastructure in Feb. 1998, the Caltech group has recently begun to work on a design, with input from JLab and UMd. It is likely that Caltech and/or UMd will assume responsibility for the service module should the design be approved and the money to construct the vessel in the DOE Equipment budget be sufficient to cover its cost. The design of the target alignment mechanism is also in a preliminary stage, as it must necessarily proceed in parallel with the design of the service module.
A list of milestones for the target were identified as part of the G0 Management Plan. These are the completion of fabrication of the cryogenic target loop, the completion of target tests at UIUC, and the delivery of the target system to JLab.
The G0 target underwent a Preliminary Design Review of the
target cell, cryoloop, gas handling system, interlocks, and
monitor/controls
system at JLab on Dec. 7-8, 1998. These parts of the
target system were reviewed by a panel of technical
experts from the laboratory and John Mark, an expert on
cryogenic targets and safety who worked for years at SLAC.
Based upon valuable technical discussions with the panel members,
the target group is making several changes to improve
target performance and reliability. The outcome
of the review was that all technical designs presented were approved.
Although the service module was not within the scope of the review, they
understandably expressed concern about its design, asking that the
issue of target alignment be solved in detail and that
the service module design proceed directly. They also specified
that the G0 target be kept a ``Class 0'' installation as classified
by the Hazard Gas Standards for the lab, which requires that the
total inventory of hydrogen in the hall be 
g,
and that the target group and the G0 Project
Manager identify sources for all items required to install and operate
the G0 target in Hall C.
While the design of the hardware and software for the target
monitor/control system was found to be sound, the Preliminary
Design Review committee raised
an issue concerning the use of LabView software instead of EPICS, which is
currently used for the Hall C and A cryotarget monitor and control
systems. This issue will have to be resolved before the monitor/control
system can be completed or the experiment scheduled.
The G0 spectrometer is composed or eight octants of detectors which will count
recoil protons from small angle e-p scattering (initial orientation of the
spectrometer) or scattered electrons from large angle e-p scattering
(180° rotation of the spectrometer). Four of the octants will be built
by a North American collaboration, and the other four by a French collaboration
(ISN-Grenoble and IPN-Orsay). Each octant will consist of 16 pairs of
plastic scintillation counters defining a specific
range of Q2. The first 15 detector pairs measure Q2 values of the
recoil proton ranging from 0.12 to 0.55 (GeV/c)2. The upper detectors serve
double duty in that they are also used to detect recoil protons with Q2
values from 0.55 to 1.0 (GeV/c)2. These protons from smaller recoil polar
angles strike detectors 12-15, and can be separated via their time-of-flight.
The 16 
detector serves as a ``guard ring'', and can also be used to
check the proper setting of the field for the spectrometer.
The azimuthal acceptance of the detectors for each octant is limited by the
upstream collimators to lie between 
10°.
The shapes of the detectors are circular and are based on an array of proton rays traced using the TOSCA program. Rays were generated for 16 values of Q2 corresponding to the detector boundaries, three target positions representing the front, center, and rear of the target, and three azimuthal angles of 0° (center of the octant), 5° , and 10.5° (one-half degree beyond the limiting aperture of perfectly aligned collimators). Those rays were used to determine the plane of each counter by their intersection with the focal surface (as defined in the G0 TDR and R. Laszewski's notes). The circular shape of each detector, both upper and lower boundary, was then found from the intersection of the rays from the center of the target and azimuthal angles of 0° , 5° , 10.5° with the detector plane. The lengths (azimuthal extent) of the counters were determined with rays from the rear of the target at 10.5° , rays which have to maximum extent in y (including misalignment of the upstream collimators). Having determined this length, the detector was extended to account for multiple scattering from the azimuthal collimators. The design and definition of the focal plane detectors has largely been the work of K. McFarlane, previously a faculty member at Norfolk State and now at Hampton University.
A concern related to the length of the detectors, was the possibility that multiple scattering of polarized protons (from spin transfer) from the collimators might create a false asymmetry. This effect was numerically simulated (Cowley, UMd, JLab, and Stellenbosch U) using phenomenological optical model potentials for lead, and although the concern is justified, an extension of the length of the detectors is of no consequence.
The above is the general description for the definition of the most of the detector shapes. However, exceptions exist. For example, the upper boundaries of detectors 14, 15 and 16 have been modified for acceptance purposes. In addition an interference problem with the lead shielding, providing line of sight shielding from the target, required that the first three detectors be moved back to somewhat larger z. By doing so, we maintain a minimum of 10 cm of lead shielding (16 cm in most cases) with minimal impact on the Q2 resolution.
The final shapes are now defined and DXF files created to cut the scintillator material in both the U.S. and France. Based on the energy loss of the protons, we have chosen to use 5 mm thick scintillator for the lowest Q2 points (first 4 detectors) and 1 cm scintillator for the remaining detectors. The scintillator material, Bicron BC-408, for the four North American octants has been ordered and delivered to Jefferson Lab. We expect to have the scintillator shapes (slightly oversized) cut by water jet in January, with final machining to follow as soon as possible thereafter. The polishing of the scintillator material is expected to begin in the spring and be finished by the end of the summer of 1999. Upon completion the scintillator will be wrapped in aluminized mylar ``tape'', fabricated by a device designed and constructed by the U. Maryland group.
In the interest of redundancy the North American collaboration has decided to view each end of each scintillator with a photomultiplier tube (four PMTs per scintillator pair). This decision has led to a high density of light guide material in the region on each side of the FPDs. As a result a great deal of design time has been invested in removing interferences between the various components, and it has been necessary to use thinner than optimum light guide material for the intermediate Q2 detectors (0.375" thick for 1 cm scintillators). The limited space has also necessitated the use of rather sharp 90° bends in the light guides (radii of 3-5 cm). Finally the large magnetic fringe fields have necessitated the use of rather long light guides to transport the light from the scintillators to the photomultiplier tubes (approximately 1.8 m for the lowest Q2 scintillators).
The decisions above led to concerns about the amount of light reaching the photomultiplier tubes, and a number of studies of the light transmission properties of light guides (both experimental and theoretical) have been carried out over the past year. These are discussed below. However, based on these studies we believe that we have a viable design, and ordered the UVT light guide material (Bicron BC-800). All material is now at Jefferson Lab.
The necessary photomultiplier tubes for the four octants being constructed the North American collaboration have been delivered. The tubes are 12-stage Phillips XP-2262B. All 300 tubes have been tested for linearity and their gains measured. A selection of the tubes were tested in more detail; for example, for count rate capability (in conjunction with the chosen base design). All tubes have a bar code identification.
The tube bases have been designed and built by TRIUMF/University of Manitoba and are completed. The base design was iterated several times in the collaboration, because of the dynamic range required by the experiment (minimum ionizing electrons to 60 MeV protons), and the rather high gain of the specific tubes chosen. The bases are passive, being comprised of resistors and zener diodes.
The plastic housings holding the tubes (a spring loaded design) have also been designed and are being fabricated at TRIUMF. Although the tubes sit in a low magnetic field environment, studies of the tube-base combination showed rather strong sensitivity to magnetic field. Therefore special care has been taken to ensure that the magnetic shielding fully covers the phototube.
A variety of prototypes have been fabricated and tested in preparation for the actual construction. These prototype studies have been extremely beneficial, particularly in terms of the ``do nots'' associated with the construction. Scintillators corresponding approximately to the No. 8 and No. 15 scintillators have been constructed. In doing so the methods for water jet cutting, machining, and polishing were developed. Several studies of the light output (absolute) versus position using cosmic rays and a beta source have been carried out for a variety of surface treatments and light guide shapes. These continue and include studies of the time structure of the pulses. Various components of this work have been been carried out by the Norfolk State, William and Mary, and U. Maryland groups. A series of similar important tests are now proceeding in France.
As noted above we had concerns about the amount of light reaching the photomultiplier tubes. Therefore prototype light guides were fabricated at Carnegie-Mellon (C-M is responsible for producing the full complement of light guides) and extensive transmission studies of these guides carried out by C-M and by the U. Maryland group. These studies have been compared to numerical simulations (GUIDEIT), and the agreement is quite good, lending confidence to the numerical simulations. Based on these studies of the prototypes we believe that the methods of construction are adequate to produce the required light.
A prototype support structure for one pair of scintillators (No. 8) has also been constructed, and is being assembled and tested at William and Mary. >From this we have learned valuable lessons on specification of all components, the necessity of some adjustment features, and procedures for glueing the scintillators to the light guides. We will also be able to test methods for light sealing the box. Currently we are using the prototype for additional studies of scintillator-light guide combination.
The stability of the gain and timing of the phototubes, as well as the transmission characteristics of the scintillator-light guide combination, will be monitored by flashing UV laser light directly into the scintillator. This will be done with a nitrogen laser, optical fibers, and a rotating mask allowing the illumination of a subset of fibers at a specified rate. A single fiber mounted at each end of the detector support structure will illuminate a pair of scintillators associated with a Q2 range. The ends will be alternately illuminated. This system has been designed and the components selected by the NMSU group. Construction of the system awaits funding to purchase the components. Construction will include the necessary fibers to pulse the detectors all eight octants, including the four octants being constructed in France.
The FPD support structure for each octant is comprised of a ``light tight box'' using a tedlar-mylar sandwich as the covering. A preliminary design exists and work is now underway to finalize the design with the finalized detector/light guide design. For the North American octants this support structure will also serve as the ``alignment fixture'' for glueing the scintillators to the light guides and so attention to the details of adjustment will be required. We expect the design of this to take several more months, and should be completed in March.
The French collaboration will build half of the FPD detectors (4 octants), including scintillators, light guides, phototubes (PMT) and read-out electronics. The CED will be built by the NA for the 8 octants.
The shapes based on TOSCA4, as given by the NA, have been checked
independently with 2 different ray-tracing programs by E. Rollinde/P. Vernin and
F. Merchez. They are considered as secure and will be taken as a basis for the
final design. One remaining question is the overall length of the detectors
which is only defined to 
2cm at this time.
Scintillators tested included ZA236 and BC408. The former is faster but has a larger attenuation than the latter, so BC408 will be adopted for ``long'' scintillators 5-16 but ZA236 remains an option for 5mm thick ``short'' ones 1-4.
The following BC408 prototypes have been ordered to CERN and delivered:
# 16 (length=1400 mm, thickness=10 mm)
# 9 (length=740 mm, thickness=10 mm)
# 1 (length=400 mm, thickness=5 and 10 mm).
They are all equipped with 12 mm thick light guides with fish-tail ending.
Moreover, 2 small scintillators (10 mm thick, 20x20 mm2 ) have been built and tested by the IPN-Orsay group in a 800 MeV electron beam at MAMI-Mainz (Dec. 5-6, 1998) in order to measure the number of photo-electrons (results are being analyzed). During the same run, the Disc/Mean-Timer of the IPN-Orsay electronics scheme has also been tested satisfactorily.
New dedicated labs have been installed at ISN-Grenoble and IPN-Orsay for G0 prototype testing using radioactive sources and cosmic rays. Dedicated data acquisition systems have been set-up in both labs. Measurement of prototype # 9 will start at IPN-Orsay Dec.14, 1998.
For mass production of both detectors and light-guides the French collaboration is still looking at alternative solutions: CERN, Bicron (through its French subsidiary Eurysis) or smaller local firms. We are awaiting price quotes.
The French part of the collaboration has decided, like the North American part, on a reading with one PMT at each end of a scintillator plane (4 PMTs for 2 consecutive scintillators in coincidence). The preselected PMTs are Philips XP2282 (8 dynodes) or EMI (10 dynodes) against 12 dynodes for the Philips XP2262 chosen by the NA group. It will be equipped with a X10 transistorized amplifier housed in the base mounting. The rationale is that there is a large dynamic range between the proton detection which lose up to 10 MeV for the forward angles measurement and the minimum ionizing particles which lose only 2 MeV in a 10 mm thick scintillator for the backward electron measurements. By working always at the optimal PMT current, we therefore hope to increase the PMT lifetime. Tenders have been sent and 280 PMTs will be ordered before the end of 1998.
The transistorized base has been developed at ISN-Grenoble. Design has been checked and prototypes are being currently tested for gain stability with a laser. Mass production will follow the final choice of PMTs.
There are still a number of issues that are considered important and which have not yet received our full attention. Here is a non-exhaustive list:
One important issue is the lack of progress of the mechanics which has been hindered by the uncertainties on the shape (until recently) and length (still pending) of scintillators.
We hope that a realistic schedule for the French detectors can be given at the time of the PAC.
The cryostat exit detectors necessary for the measurement of large angle electron scattering (requiring 180° rotation of the spectrometer) for all eight octants will be designed and constructed by the North American component of the collaboration. The design and construction of the CEDs is primarily the responsibility of the Louisiana Tech group. (They are currently having discussions with the TRIUMF Scintillator shop concerning the construction of some components.) The detector system will utilize 12 of the 16 Focal Plane detectors (the first four FPDs will not be used) in coincidence with a set of nine new detectors mounted just beyond the cryostat exit windows. These detectors will be viewed by photomultiplier tubes removed from the second scintillator of each FPD pair. As of this writing, the shapes of the CED scintillators are defined and work is in progress designing the light guides and simulating the expected light output. It is expected that the CED design will be completed by the end of January.
The electronics for G0 can be broadly grouped into three classes: monitoring, forward-angle, and back-angle. Of these, the monitoring electronics is the most conventional, consisting of commercial FASTBUS ADCs and TDCs which will collect information from all PMTs in the detector for a small fraction of the beam bursts. The principle function of the forward-angle electronics is to accumulate time-of-flight spectra for each of the focal-plane detectors, allowing elastically scattered protons to be separated from inelastic background. The principle function of the back angle electronics is to count the coincidences between individual focal plane detectors and associated cryostat exit detector elements, permitting separate accumulation of electrons from elastic and inelastic kinematic regions.
Because the monitoring electronics is conventional, relatively little work has been invested in preparing this part of the electronics. Apart from the FASTBUS electronics, this will require only cables and splitters. In addition to monitoring PMTs for signs of gain shifts or degradation, other information is available because of the fact that this part of the electronics will capture data event-by-event. This makes it useful for monitoring many properties of the detectors such as efficiency and time resolution as a function of position within the detectors, alignment of backup detectors with focal plane detectors, coincidences between focal plane elements, etc. Additionally this will sample the time spectrum (with a relatively simple dead-time correction), allowing the front-end electronics to be cross-checked, calibrated, and accurately corrected for dead-time effects. Acquisition of the commercial electronics is expected to be spread across FY 2000 and FY 2001 apart from individual units to be used for testing and to ensure suitability for the intended purpose.
There are several parallel efforts in development of forward-angle electronics. The North American and French groups will each produce electronics to read out their half of the detectors. In addition to providing a backup plan, should either design encounter unforeseen obstacles, this dual approach will allow some cross checking to ensure that neither design is undermined by unsuspected systematic errors. The French development project is further sub-divided into two parallel, but very distinct, efforts being carried out at IPN-Orsay and ISN-Grenoble. Both of the French designs will be prototyped, but it is intended that only one will be selected for production. All three designs have the same basic function, accumulation of time spectra for individual focal plane detector meantimes, in coincidence with a hit on the corresponding backup detector. These role of these detectors is distinct from the FASTBUS TDCs in two important ways. First, the forward-angle electronics must accept almost all of the coincidences, at a mean rate of over 1 MHz and an instantaneous rates of over 10 MHz (as opposed to the FASTBUS electronics which samples only a tiny fraction of the hits). Secondly, in order to achieve this very high encoding rate, only mean-time spectra are accumulated. No information is stored concerning correlations or different detectors, nor of time-differences between the PMTs of a single detector.
The North American and ISN-Grenoble designs are both based upon the same concept. High-speed shift registers, shifted by a clock-train synched to the beam burst are used to encode the arrival time of the hit as the depth to which the data shifts before the end of the clock-train. The parallel outputs of the shift resistors are then strobed into scaler channels, incrementing the time spectra stored there, before the shift registers are reset for the next beam burst, 32 ns later. The differences lie mainly in the implementation, with the North American design being more modular and the ISN-Grenoble design more integrated. In the North American design the meantimers and scalers are in separate units connected by ribbon cables to the time-encoding board. ISN-Grenoble is developing custom ASICs, one to perform the meantiming, another to perform high-speed scaling, and possibly a third to serve the shift-register function. Their single board will serve the functions of meantiming, determining coincidences, time-encoding, scaling, and VME interface. Only constant fraction discriminators (CFDs), and splitters are required as separate modules. The North American design, on the other hand performs only the coincidence-detection and time encoding on the custom board (in addition to support services, such as dead-time monitoring) and employs separate modules for CFDs, meantimers, and VME-scalers. Assuming the ISN-Grenoble ASIC-development project is successful, the North American design may benefit from low-cost scaler and meantimer modules in place of the originally-planned commercial units.
The IPN-Orsay design differs substantially from that described above. It even more integrated than the ISN-Grenoble design, in that it includes the CFDs on the single board being developed. More significantly, the heart of the time encoding is a completely different, using flash TDCs passing data (through FIFO buffers) to a set of DSPs which assemble the spectra in memory to be read out between macropulses by an additional DSP which passes the data to VME-accessible memory. The time resolution of the flash TDCs is 250 ps, compared to the 1 ns resolution of the shift register technique.
The North American time-encoding board has been prototyped (along with the support board needed to produce clock trains) and tested. Additional features have been added to aid in monitoring deadtime, to simplify external cabling, and to detect errors caused by noise or bad cables. A first ``production'' board has recently been printed and stuffed and is now undergoing testing. Assuming only minor modifications are required, full scale production of these boards can begin in early 1999, subject to funding. Acquisition of commercial scalers and meantimers will be delayed until at least FY2000, allowing the possibility that a more economical solution can be found, using boards incorporating the ISN-Grenoble ASICs.
ISN-Grenoble is pursuing two ASIC designs for the meantimer. The first has already been prototyped and has been found to perform well when provided with suitable temperature feedback. The second design is presently being prototyped, and is expected to be more intrinsically stable. The prototype ASIC for the 32 channel 100 MHz scaler has also been delivered, and will be tested in the near future. Testing will include production of a VME board which may eventually be used as the external scaler module for the North American design. This VME board is expected to be tested and ready for use in January of 1999. The time-encoding board incorporating the meantimer and scaler ASICs and the time-encoding shift registers is expected to be ready for testing by February of 1999. This may draw upon the experience from tests of the North American board.
The IPN-Orsay board is being laid out, and the first prototype is expected by January of 1999. The components of this board are existing ICs, but some programming and testing will be needed. It is expected to be ready for testing in March of 1999.
Comparisons of the two French designs are expected to be carried out, using realistic scintillator signals, in March of 1999. By the end of April, a decision will be made on which of the two solutions to use to instrument the four octants being instrumented by the French. Whichever solution is selected will then go into production, and is expected to be ready for delivery by February of 2000.
Three solutions are also being pursued for the back-angle electronics. The time-frame for their development is less critical since they will not be used for early running of G0. The demands on the back-angle electronics are also less strong, since time-encoding will not be required. Production of the North American back-angle electronics is expected to begin in early FY 2000 and to be completed in November of 2000. The two French designs for the back-angle electronics are modifications of the forward-angle designs, and will be pursued in parallel. Whichever design is selected is expected to be completed along with the corresponding forward-angle electronics by February of 2000.
JLab is providing material, equipment, and services to G0 as a part of the infrastructure of Hall C. In addition, some modifications will be required to existing equipment in the hall.
The planning, material, services, and labor required to install the collaboration deliverables (SMS, target, detectors, etc.) and to establish an interface to the Hall C infrastructure is covered by this task. The physical location of the G0 experiment will be down stream of the Hall C spectrometer pivot assembly. This has the advantages of making G0 minimally invasive to the other Hall C experimental programs and allowing the simplest G0 staging arrangement. Based on preliminary radiation calculations it is also the best siting for minimizing overall beam produced backgrounds in Hall C.
Infrastructure services and equipment related to the electrical systems of G0 includes: 1) signal and control cables connecting the experiment to counting room patch panels (their fabrication and installation and mechanical support), additional patch panels for intermediate break-out; 2) AC power in Hall C for vacuum, cryogenic systems instrumentation, target instrumentation, gain monitoring system instrumentation etc.; 3) high voltage supplies installed In Hall C along with cables, cable mechanical support and high voltage patch panels (If needed), and 4) the construction of a G0 electronics and counting room with its required infrastructure including AC power, air conditioning, racks, and modular electronics crates.
To date all signal cables have been purchased and delivered at a cost significantly below the budget estimate.
Careful measurement of the parameters of the electron beam will be necessary in order to control systematic uncertainty in the G0 measurement. This will require beam monitors including: 1) RF cavity position/Intensity monitors, and other beam position monitors (stripline, harp wire scanners). Also included in this item are electronics systems associated with the electron beam such as the raster system and the helicity readout from the polarized source in the counting room. Also included in this task is the work and material needed for the modification and adaptation of the Hall C Moller polarimeter to the needs of the G0 experiment. Finally, mechanical modifications to the beam line such as well as the vacuum system for the SMS and valves to permit the isolation the SMS are included here.
An instrumentation girder is under design which will contain the rf cavities, stripline monitors and a harp wire scanner. This self contained girder package is very similar to the new girder package currently being completed for the upcoming Hall C program. The G0 girder package will be located in upstream of the G0 target system which is itself upstream from the magnet/detector assembly. This self contained instrumentation girder will be removed and stored between G0 runs. By locating G0 downstream of the pivot and employing a dedicated beam monitoring instrumentation package G0 will have the advantage of obtaining access to signals from three separate beam monitoring packages. Specifically, the one in the tunnel to Hall C, the package in front of the standard Hall C scattering chamber and the G0 package. This should result in excellent monitoring of all beam properties.
The G0 SMS, the target, and the Moller polarimeter will require cryogens from the end Station Refrigerator. To accommodate G0 the cryodistribution system must be upgraded to provide cryogens to the target and the SM. Flexible ``U-Tubes'' for delivery of cryogens to the SM and the target will be fabricated.
Since G0 is now located downstream of the standard Hall C pivot a satellite cryogenic ``well head'' will be installed near the planned G0 location. Since this satellite ``well head'' will likely be used for future Hall C experiments (beyond G0 ) the funds for this work are coming from the Hall C general operations budget and a not part of the scope of the G0 experiment. Engineering design work for the G0 specific (and funded) transfer lines is also under way.
This task covers the engineering and drafting effort associated with the mechanical supports, cryogenic plumbing, shielding, and other additions and modifications to the Hall C environment that are specifically required by the G0 experiment, i.e. engineering and drafting for the other Hall C Infrastructure tasks.
This task includes all labor and material required for the fabrication, and assembly of the mechanical support of the G0 detectors. The support consists of individual CED and FPD octant support frames, associated with the North American and the French detectors, which ``plug in'' to a detector superstructure (the so-called Ferris wheel). The alignment of CED and FPD octant modules is accomplished with adjustment degrees of freedom provided by the Ferris wheel. Internal alignment of scintillator elements within octant supports is established when the octant modules are assembled. Lead and borated polyethylene shielding an well beam line material Integral to the Ferris wheel support will be fabricated as part of this task. A rail system will permit the Ferris wheel to be retracted from the SMS for, servicing. The support frame (Detector Assembly Support Frame) to hold both the North America and French detector modules is under final design. The drawing package should be ready for bidding around March 1999.
Since the original G0 concept was for the stationary placement of the experiment and the selection of the downstream placement option requires that is be easily moved out of the beamline so other experiments can run in Hall C a rail motion system is under development. This rail system is for moving the G0 magnet and DASF out of the beam line and the separation of the Magnet and DASF in order to allow the SOS to reach its full angular range. Time to complete a drawing package for the rail system is 2-3 months. Since this is a requirement for other experiments to run easily in Hall C the costs of designing, building and installing this rail system will be absorbed by the general hall operations budget and consequently is not part of G0 's scope.
The software effort associated with the G0 experiment has been divided into four parts: slow controls, data acquisition software, analysis, and simulations. Separate teams will focus on the creation of software to meet the requirements of the experiment in these areas.
The slow controls group is charged with the integration of slow controls hardware and software, provided as part of other subsystem deliverables, into a unified system. This will involve communication with stand-alone systems associated with the spectrometer magnet/cooling control system the target and the accelerator. Mechanisms for dealing with these slow controls front-end systems programmed with Lab-view and EPICS will be provided. In addition, specialized software for communication with beam diagnostic systems (such as current, beam position, and helicity monitoring) as well as electronics which requires external parameter down-loads will be written. User-friendly displays will be provided to indicate status and alarm conditions.
Leadership for the slow controls effort will come from the NMSU group. The design of the slow controls system is still at the planning stage. Completion of a design report awaits the definition of the control systems for the target and magnet. The initial version of slow control software will be provided for testing at UIUC late in 1999. Subsequent versions will then be tested and finally used during the experiment commissioning at JLab.
The creation of CODA software necessary for reading out the electronics to produce a data stream, for storing that data stream, for backing up that storage to tape and for providing hooks for on-line analysis of the data stream are all within the scope of the G0 data acquisition effort In addition an interface to the slow controls software must be maintained to allow slow controls parameters to be inserted into and stored along with the data stream.
The current effort is to produce a final definition of the hardware to be read out by the DAQ software. This is expected later this spring following the decision on the electronics for the French octants. The first version of the code will be developed to provide a test stand for detector octants as they are assembled.
Another area of computation involves extraction of physics and diagnostic results from the data stream produced by the experiment. The analysis effort is dominated by the calculation of asymmetries and the determination of systematic uncertainties. An on-line code must be prepared in time for the commissioning experiment. To achieve this goal, the software will be tested ahead of time with pseudodata produced by simulation codes. As the experiment matures, it is expected that the on-line analysis code will evolve to become the production off-line code. The bulk of the off-line analysis will be carried out by thesis students. The analysis effort will therefore involve coordination of parallel analysis tracks and documentation of the analysis for eventual publication.
The development and maintenance of standard simulation codes and their application to predict the behavior of the apparatus and to contribute to its design, is another vital computational task. There are currently six areas of study for the simulation of the G0 experiment.
The simulation task includes the maintenance of standard versions of the above codes. This will include documenting them to a level which would allow the average graduate student to use them with only minimal consulting with the authors. The task also involves running the codes to answer a list of questions/issues provided by the Experiment Coordinator.