. • ê . .. I OF I ORNL P 1723 No, 14 5 250 FEE IN 3.6 L25 L4 L5 MICROCOPY RESOLUTION TEST CHART NATIONAL BUREAU OF STANDARDS -1963 ORNU prin23 Conf- 651001-8 HOV 1 8 1906 1 - .. # - privately owned rights; or with the Commission, or his employment with such contractor. disseminates, or provides accomo to, any information pursuant to his employmont or contract such omployee or contractor of the Commission, or employee of such contractor prepares, ployoc or contractor of the Commission, or omployee of such contractor, to the extent that As used in the above, "person acting on baball of the Commission” Includes, any em- Uko of any information, apparatus, method, or procob8 dsclosed in this report. B. Asnumos any liabilities with respoct to the use of, or for damages resulting from the of any information, apparatus, method, or process disclosed in this report may not infringe racy, completeness, or usefulness of the information contained in this report, or that the use A. Makes Lay warranty or representation, expressed or implied, with respect to the accu- Statos, nor the Commission, por any person acting on behalf of the Commission: This report was prepared as an account of Government sponsored work, Nelther the United LEGAL NOTICE TE USE OF A DIGITAL COMPUTER IN TIME-OF-FLIGHT EXPERIMENTS* R. Gwin, I. W. Weston, and R. Ingle Oak Ridge National Laboratory Oak Ridge, Tennessee A general-purpose PDP-4C digital computer** has been incorporated as an integral part of a data acquisition and processing system ior keutron time-of-flight experiments. The system, which includes a dis- play unit, has been used in experiments to measure the neutron capture and fission cross sections of 235T and 239 Pu as a function of' neutron energy, and in "Rossi-alpha" experiments2,9 to measure the time correla- vion of neutrons emitted from an assembly of fissionable materials. The decision to use a computer rather than a wired programa analyzer . the analyzing system for these experiments was based on the flexibility a which the large core memory of a computer could be used and on the act that the computer could be programmed to provide variable channel withs. The variable channel wiatns permit the investigation of a large vime region, and thus a wide neutron energy range, with channel widths appropriate for the time intervals covered. The investigation of commercially available computers began late 1963. A program to process the data was written for each computer con- wered, and when it was established that satisfactory speed require- ments for data processing ana storage and for data display were obtainable, une PDP-4C computer was selected. . . . . . OSS i. 4 " . 2 ... *esearca sponsored by the U. S. Atomic Energy Commission under contract . with the Union Carbide Corporaticn. ***Vanufactured by the Digital Equiqment Corp., Maynard, Mass. · . . . . . RISTORASID FOR ANNOUNCEMENT I MUISTARAR SCIENCE ABSTRACTS -20 : In order to facilitate the use of the computer, fast-data-break hard- ware, which was a standard option, was purchased along with it. This equip- ment consists of three channels for inserting data directly into the core memory. Each channel consists of 18 data bits, 13 address bits, and a data- ready signal. In order to insert information into (or extract information irom) the core memory, a data-ready signal is fed to the computer. The com- puter completes its current instruction and then, in the next computer cycle, the 18 data bits on the input lines are inserted into the word of the memory addressed by the 13 address bits. Since the address for the incoming data must be specified, a special feature was added to the computer to provide an address for the incoming data. This item consists of two sets of registers (set by toggle switches), one of which specifies a beginning aääress "begin" and the other a linal address. In operation, the first data word is inserted into address "begin." The next data word is inserted into the address "begin + 2" and so on until the final address is reached. The next address is then "begin," and the cycle is repeated. The memory region between "begin" and "final" forms a buffer zone for the incoming data. The data-processing program searches the buffer zone of the computer between "begin" and "final" for data. If there is no data to process, display or computational programs are executed. One other special feature obtained with this computer enables the input om data into the computer in the same fashion as a conventional analyzer. That is, the meinory word whose address (channel number) is presented to the computer on the address lines of the fast-data-break channels is incremented oy one. This feature was not used in the cross-section and Rossi-alpha experiments because of its flexibility; however, it has been tested and äunctions properly. Tae complete time-cf-Ilight system was not purchased as a unit from one manufacturer. The scientific instruments and their interfacing with The computer were the responsibility of Oak Ridge National Laboratory. The requirements for using the last-data-break channels of the computer are simple and no difficulty with them has been encountered. There was some problem in obtaining a digital time-interval meter ("ciock") that met the scientific requirements. The "clock" used in the cross-section experiments was assembled froza mošules obtained and wired by the computer manuiacturer. It has a channel width of 0.1 usec, a range of 104.8 msec, S and a dead time between multiple events of 3.2 hsec. An additional "clock"*** provides channel widths of 0.005 usec and a dead time of 0.8 usec between events. This faster clock was not completely tested and depugged in time for the cross-section measurements, but it was used in the Rossi- alpha measurements. An existing pulse-height analyzer has also been inter- faced with the computer to proviáe the capability for multidimensional experi- ments. In all these cases the interfacing with the computer was readily accomplished, the major difficulty consisting in proper functioning of the scientific instruments themselves. A picture of the computer is shown in Fig. 2. The memory and arithmetic 3 unit occupy the equivalent of two standard electronic racks. A separate rack contains the tape reader, tape punen, oscilloscope, and some electronic equipment associated with the experiments. . *** Purchased from Eldorado Electronics, Concord, California., . . . ... . .. ...... ***** Typical results from the cross-section experiment using the variable cacnel wiath program are shown in Fig. 2, waich gives the count rate for 2051 as a function of the time of flight. The times at which the channel width changes were made are readily distinguished by the abrupt change in counts per channel. Incorporated in this processing program were various control features; for example, in the event that a channel overflowed, the address of the channel was recorded. In aādition, the data input rate was scaled, as was the processing rate, and an alarm was given if the average rates were not balanced. The word.3 used for data storage (a total of 7180) were punched on paper tape in bina::y format without addresses. The data tape was then read and checked with the contents of the computer. With the computer, all operat:.ons, including simple ones such as clear- ing the data memory region, have to be programmed. The time required for learning the programming and operation of a small computer must be weighed against the advantages of using the computer. Although the present computer was obtained primarily because oz its flaie data collection capabilities, we have spent some time in preparing more elaborate programs than was re- quired for the basic operation. Ar. example of one of these is a program to display data on an oscilloscope. (The display unit displays up to 1024 points along the x axis and along the y axis.) In this program the number of data points being displayed, the scale, and the memory overlap region are controlled by carisole switches. The location and number of points being displayed are available to other programs and thus data output or data processing can be performed on the data displayed. The use of the computer in these experiments has proved to be very satisfactory. To conclude on a practical note, the computer has been säipped by commercial carrier three times, and in each case the system was in good operating condition on arrival at its destination. - -- . r . ' * .. . * T. LT ELLE . .. . . . . . . . . . . . . . REFERENCES i. The cross-section experiments in which this system was used were sixilar to those described previously: G. de Saussure, L. W. Weston, 2. Gwin, J. E. Russell, aná P. W. Hockenbury, Nucl. Sci. Eng. 23, 45 (1965). 2. J. Orndori, Nucl. Sci. Eng. 2, 450 (1957). 3. J. T. Mihalczo, Nucl. Sci. Eng. 20, 60 (1964). LENZ9 OJOHD; ל"כלכלי .1 A • : • :כן r::::: :: נינ»: ::::: צי :•* : . ) :ז)-::: • . צ.צ • • ..:::::•••נו :::: :. : : : : :;. *: .. .:: :.. ..... ;: : ' . .- :י 1 ::::. :. י: ;{ {.... !: : .::::.. :..: . . : : ! י. ::: א " 20,000 ORNL-097G 65-4628 U235 16,000 COINCIDENCE EVENTS (FISSION) 00 counts/channel 8000 4000 . •111 30,000 . 11:17 24.000 ANTICOINCIDENCE EVENTS (CAPTURE EVENTS * 13 OF TOTAL FISSION EVENTS * BACKGROUND) counts/channel 6000 mont ; 1000 30CO 2000 CHANNEL..UWER 3600 .:?. . -'lingi 1*?( uz Yin in 23?U. 1: END Roinnin - 4 . i DATE FILMED 12/ 0 /65 1. i --- - - --- - -- - - - 1 PS