<<

MAS vacuum attains 5 x 10~11 torr, perimental areas are being ex­ mesons and of pairs of pions. Capi­ with the whole ring undergoing a tended, in collaboration with the talizing on the copious supply of week-long bakeout at 300C. Fast German GSI Laboratory, Darm­ polarized particles, the production kicker magnets pass beams to the stadt, and with the Italian National of neutral pions is found to be main ring in less than a microse­ Institute for Nuclear (INFN). highly spin sensitive. cond at 12.9 MeV/nucleon. The Studies using the SPES IV spec­ Saturne is one of the two Labor­ two radiofrequency accelerating trometer have shown how the pro­ atories operated jointly by the Insti­ cavities operate between 160 kHz duction of the delta (excited nu- tute of Fundamental Research of and 2.48 MHz for accumulation cleon) particle changes with differ­ the French Atomic Commis­ and subsequent acceleration, while ent nuclear targets, with the delta sion and the National Institute for a special decelerating mechanism appearing to show an affinity for, and Particle Phy­ facilitates injection (at 187.5 and increased stability with, nuclear sics of the National Centre for keV/nucleon) and stacking. matter. Other have Scientific Research, the other being To reap the benefits of the in­ shown an interesting correlation the GANIL heavy ion accelerator at creased supply of particles, the ex­ between the production of eta Caen.

Control systems for experimental physics

At an international conference last mance require highly sophisticated gence at the device level. Control year at Villars-sur-Ollon (Switzer­ control systems, necessarily more architectures are mainly based on land), scientists from all over the expensive, conflicting with the cur­ two-level networks, with a back­ world looked at the problems of rent economic restrictions. Never­ bone, token ring or Ethernet, con­ controlling complex physics instal­ theless a continual effort ensures necting workstations to mostly lations, including particle accelera­ that controls, aided by advanced VME-based front-end microproces­ tors, nuclear reactors, large tele­ technology, keep pace with emerg­ sors, or concentrators, in turn driv­ scopes and high energy physics ing requirements. ing local distribution networks (e.g. detectors. Main issues and trade-offs were 1553B) to individual devices. The meeting, organized by the covered by invited papers giving In such environments software is European Physical Society's Inter- overviews of major control sys­ becoming increasingly complex and divisional Group on Experimental tems, and by contributed papers, expensive, and the meeting empha­ Physics Control Systems, EPCS, workshops, tutorials and exhibits. sized software . With brought together 180 scientists To stimulate debate on topical mat­ few well-integrated software engi­ from the world's leading exper­ ters, panel workshops (Architec­ neering tools for large projects on imental physics research laborato­ ture, Interfacing and Intelligent Pro­ the market, a number of recom­ ries, universities and industries, in­ cess Equipment, Applications Soft­ mended strategies and techniques cluding substantial delegations ware, Operational Aspects, Auto­ were presented. Structured engi­ from Czechoslovakia, Iraq, Israel, mation, Software Engineering) cov­ neering methods have proved their Japan, the People's Republic of ered a substantial part of the pro­ worth in a number of control pro­ China and Poland. The absence of gramme. jects, and, though still in the proto­ the USSR delegation was much re­ While more complex machines type stage at accelerator Laborato­ gretted. need more complex controls, rapid­ ries, object-oriented concepts As physics installations get larg­ ly advancing technology makes show some promising signs of er and more complex, manual oper­ hardware costs dive. Current con­ success. ation becomes increasingly difficult. trols tend towards fully distributed Process control systems should Safety, reliability and high perfor­ systems with more local intelli­ be reliable and flexible. Mainte-

CERN Courier, May 1988 7 Discussing systems architecture at the re­ cent experimental physics controls meeting in Villars, Switzerland, are left to right, Mi­ chael Shea (Fermilab, US), Shin-lchi Kuroka- wa (KEK, Japan), Rudolf Steiner (GSI, Darm­ stadt, Germany) and Winfried Busse (Hahn- Meitner Institute, Berlin). nance eats into resources and pre­ cious operation time, so it is es­ sential to include monitoring to trace system behaviour, or misbe­ haviour, and software should be largely data driven. On-line data, needed for real time activities, re­ lates to off-line data describing operational constraints or providing management information. The da­ tabase is thus an essential part of the control system. Good management requires solid planning with identified milestones, realistic schedules and defined bud­ gets. Gone are the days when inge­ nious engineers or could timber together some ad hoc solu­ tion in their 'spare time'. Smaller installations with fewer resources run into problems. There was a clear call from the smaller in­ stitutes for help from bigger broth­ ers, for example with subsystems available in kit form. Along these lines, an initial working group aims at standardizing the operational invest in appropriate engineering At the workshop on Automa­ control protocols of power con­ tools developed in parallel with the tion, a number of levels were pro­ verters. Seven institutes are partici­ real-time system. posed to define successive de­ pating to this first collective EPCS Efficiency, both in operation and grees of sophistication, useful endeavour. exploitation, calls for a single con­ when reconciling goals and avail­ Applications software, a major trol room with standard consoles. able resources. The lowest level is fraction of today's development Traditional consoles with alpha­ data acquisition, for monitoring work, was looked at in detail. Re­ numeric and graphic displays, operation and for developing con­ commendations included - a touch panel, tracker balls, etc, are trol algorithms. The second is more layered structure of modules each tending to be replaced by modern concerned with supervisory con­ catering for a specific type of activ­ workstations with powerful operat­ trols, and the third incorporates ity; separation of code (a library of ing systems and graphics pack­ modeling programs to test pro­ reusable routines) and data (the on­ ages. Experience with these has posed changes. The fourth level line data base); and housekeeping, shown that their extreme flexibility performs continuous controls with scheduling and similar activities in requires rules for consistent ev­ the setpoints of the control loops standard modules. This should pro­ eryday use. Questions included - available to the operator, but main­ vide basic process-independent Should alarms be included on the tained by the system to assure packages into which specific appli­ main console workstation? Can a steady operation. At the next level cations can be embedded relatively single workstation display be mod­ the system is able to change the cheaply using data tables and short ified for multiple access? Recom­ state of the process according to a special-purpose procedures. How­ mended standard features included master sequence, involving auto­ ever further development towards summary status display, chronolog­ matic start-up and shut-down with appropriate structures and standar­ ical recording of events, covering sequential control algorithms for all dization is still necessary . Future hardware faults and operator re­ process devices. The penultimate large controls projects should also quests, etc. level extends this sequential control

8 CERN Courier, May 1988 LB Landolt-Bornstein Numerical Data and Functional Relationships in Science and Technology Editors in Chief: K.-H. Hellwege • O. Madelung Comprehensive Index ~ HALL EFFECT for 6th Edition 1950-1980 and New Series 1961-1985 DIGITAL TESLAMETERS Editor: K.-H. Hellwege

By incorporating digital linearization and A general index for Landolt-Bornstein has long been a optional temperature correction these Hall effect desideratum. Orientation within an individual volume Teslameters measure magnetic fields to 0.01% is not too difficult due to the clear organization and a precision with 10 ppm/°C temperature coeffi­ detailed table of contents in each volume. The growing cient. Digital readout in Gauss or Tesla on a number of volumes, however, and their increasing spe­ 7-character LED display. Built-in peak hold, rms cialization have often made it problematical to find the ac and digital filtering functions, with RS-232-C required data. In this respect the Comprehensive Index plus fiber optic serial or IEEE-488 interface op­ tions. Analog output proportional to field with will be a help to the user of the work. frequency response from dc to 3kHz (optionally The index covers all volumes of the 6th Edition and the 9kHz). volumes of the New Series available by the end of 1985, a total of 126 volumes with more than 74.000 pages and DANFYSIK A/S - DK-4040 JYLLINGE, DENMARK 56.000 figures, published during the last 35 years: thus a TEL INTERNATIONAL +45 2 38 81 50 TELEX: 43136 - FAX: +45 21315 51 product of a whole scientific generation. In USA: GMW Associates . Tel.: (415) 368-4884 DANFYSIK During the long period of scientific progress covered by Landolt-Bornstein, the development of some fields has* led to radical shifts of emphasis and changes in the models used and the nomenclature. Also new areas of research and application have cropped up, others are no longer of major interest. Consequently some changes in the logical structure of the work were inevitable, and these are reflected in the Comprehensive Index. Since - with only rare exceptions - the 6th Edition has been published in German, the New Series in English, keywords in both languages had to be used, and the index consists of two separate parts in alphabetical order. The index has been prepared by members of the Fach- informationszentrum Energie - Physik - Mathematik (FIZ Karlsruhe) and the Landolt-Bornstein Editorial Office, Darmstadt. 1987. Approx. 300 pages. Hard cover DM 600,- ISBN 3-540-16675-0

Please order through your bookseller or write to: Springer-Verlag Berlin Heidelberg New York London Paris Tokyo 4005-Marketing Heidelberger Platz 3 • D-1000 Berlin 33 —. GRATINGS • STAIRCASES lOit) GITTERROSTE • INDUSTRIETREPPEN HYDRAULIC PUMPS — CAILLEBOTIS • ESCALIERS GAS BOOSTERS • AIR AMPLIFIERS POWER PACS • TEST STANDS

TypeS Type SF Type SPS

for numerous applications of high and very high pressures up to 14 000 bar Send for more information stating your require­ ments! AMMAIMN-TECHNIK AG CH-5742 KOLLIKEN P.O. Box 80 Tel. 64 43 49 39 Ing. Herbert Panne Telex 982177 Fax 64 43 88 43 GmbH u. Co. KG D-6349 Greifenstein 2 Manufacturers of equipement to produce, control West Germany and store high pressure oil (+ many other fluids) and gas Tel. 6478-513 Tx 484 219 Fax 6478-1205

Rugged and dependable. The HP 8562 A/B Spec­ trum Analyzer passes MIL-T-28800C require­ ments for vibration, pulse shock and transit drop. Special features include full HP-IB control, Demodula­ tion, Markers, Frequency Counter, Peak search, THE NEW PORTABLE SPECTRUM ANALYZERS FROM HP WITH UNBEATABLE PRICE/PERFORMANCE Plotter dump and Non-volatile memory for Traces and Analyzer set­ tings. A special Test and Adjustment Module provides extensive servicing capabilities for on-site maintenance. Select from: HP 8562A, HP 8562B: Synthesiced, 1 KHz - 22 GHz (A is preselected) HP 8590A: 10 KHz -1.5 GHz (1.8 GHz optional) HP 8592A: 50 KHz - 22 GHz (preselected) Exceptional low price! Special options (models 90A and 92A): EMI and CATV measurements.

For more information, or a demonstration, please contact your local HP office.

HEWLETT PACKARD to fault diagnostics and possible that industry is ahead of the phy­ design and implementation. system recovery. The highest level sics labs, both hardware- and soft­ Visits to CERN and to Ciba-Geigy of sophistication is automatic op­ ware-wise. Several tutorials ad­ illustrated differences between in­ timization through setpoint adjust­ dressed this problem: dustrial and scientific control sys­ ment and state changes, a field buses, vital for connecting intelli­ tems. The former operate in well ripe for artificial intelligence. gent physics equipment, were the defined, stable and understood Present particle accelerators are object of two lively sessions, in­ production environments, while the not highly automated, as it is felt cluding one on the rising star of latter, as exemplified by particle ac­ that measurements are not reliable VME; celerators* are less well defined, enough, there are insufficient con­ the underlying principles of expert subject to change and have to trol parameters and their settings systems, a new discipline, highly cope with process effects often are often not sufficiently reproduci­ promising for the experimental phy­ not fully understood. ble. sics field, drew a large audience, Reflecting its success, it was By special request, an additional and was followed by examples agreed that this type of conference session was improvised on VMS from industry and a demo; should find a venue once every two versus UNIX, highlighting the differ­ highly relevant for controls was ob­ years, in the intervening time com­ ences between a highly flexible ject-oriented software design; plemented by workshops and proprietary operating system and a and finally SASD (structured analy­ study groups on specific subjects. public one. sis and design), a computer as­ It is increasingly becoming clear sisted methodology for software From A. Daneels and B.Kuiper

With CERN's prototype large diameter li­ Around the Laboratories thium lens for antiproton work on its testbed at the Soviet Institute for Nuclear Physics at Novosibirsk Laboratory are (left to right) Robert Betlone, Gregory Silvestrov, Tanja Vsevolozhskaya, Peter Sievers, Juri Petrov, Victor Volokhov, Boris Bayanov and Sacha Chernyakin. The lens underwent 500,000 pulses at 800 kA.

CERN/NOVOSIBIRSK Lithium lens colla­ boration

The idea of lithium lenses to boost the collection of antiparticles was born at the Soviet Institute for Nu­ clear Physics at Novosibirsk and subsequently exploited for use with antiprotons at Fermilab and at CERN, where 2 cm-diameter lenses are employed. For the ongoing consolidation programme for CERN's antiproton supply, a larger (3.6 cm diameter) lens is foreseen and a first proto­ type has been built at CERN. To

CERN Courier, May 1988 11