On-Line Data-Acquisition Systems in Nuclear Physics, 1969 — A Reader’s Guide
Edition facts
The report opens by tracing the evolution of nuclear data collection from 1930s binary scalers through vacuum-tube analyzers to transistorized systems, culminating in the 1960s adoption of on-line computers. The panel, appointed in March 1968, aims to provide guiding principles for laboratory directors, reviewers, and funding agencies facing rapid technological change. The document is structured around system design considerations, hardware and software trade-offs, and operational lessons from existing facilities.
Historical Trajectory and Technological Context
The preface establishes a clear chronological framework: the 1930s saw the first digital nuclear data device (the binary scaler); the 1940s introduced single and multichannel pulse-height analyzers using vacuum tubes; the 1950s brought improved analog-to-digital converters and computer-type memories (acoustic delay lines, then ferrite cores); and the transistor revolution of the late 1950s accelerated development. The 1960s are characterized as “the decade of the computer,” with on-line systems entering nuclear laboratories around 1962. This historical lens frames the report’s urgency—the field’s rapid evolution creates difficulty for decision-makers, as computer costs often represent a sizable fraction of a new laboratory’s total budget.
System Architecture and Hardware Choices
The report details specific hardware configurations, such as the Brookhaven National Laboratory multiple-spectrometer control system (MSCS). This system uses an SDS 920 computer with a 16k, 24-bit memory, a 32,010-word magnetic drum, two tape units, and nine local control stations (SCS) at spectrometers. Each SCS includes stepping motors for computer-controlled angular rotations of crystals and counters, shaft rotation encoders, manual controls, electronic counters, displays, and decoding sections. The network incorporates a Varian 620i computer and allows operations from 12 remote stations, though program loading still requires the high-speed paper-tape reader at the computer. Such concrete descriptions illustrate the state of the art in 1969.
Operational Lessons and Reliability Data
A section titled “Lessons from Operating Experience” offers candid observations from the MSCS. The system is reported to do “all things imagined to be necessary.” The computer achieved about 40,000 hours of use without a breakdown, with preventive maintenance performed mostly during the one week per month the reactor is shut down. One person serves as operator and programmer for simple jobs, also transporting magnetic tapes for off-line processing. The average user does not need to program. Machine-language programming is not considered burdensome because the system is fixed-hardware. The only noted problem is occasional program loss due to lack of hardware memory protection, estimated to cost at most a few percent of running time.
Cost Breakdown and Resource Allocation
The report provides detailed cost tables for the MSCS. Engineering design over 12 months required 64 man-months and $119,000, including $56,000 for system coordination, development, coding, and $28,000 for parameter generation and data analysis. Construction over 16 months cost $38,000 (33 man-months). Major components totaled $309,000, with the SDS 920 computer alone at $147,000. Replacement is noted as impossible because components are no longer in production. Annual operating costs are $42,700, including $11,000 for a computer operator/programmer, $12,000 for maintenance, and $4,000 for materials. Manpower requirements include one operator/programmer, one systems programmer as needed, and two part-time maintenance personnel.
Readers should approach this report as a primary source document reflecting the state of on-line data acquisition in nuclear physics at the end of the 1960s. The panel’s focus on concrete system descriptions, reliability statistics, and cost data makes it valuable for understanding the engineering and economic constraints of early computer-based laboratory automation. The excerpts do not reveal the full report’s conclusions or recommendations, but the detailed case studies provide a grounded perspective on the challenges and solutions of the era.