CHAPTER 1: INTRODUCTION 1.

1 Introduction:A Programmable ogic Controller, P C, or Programmable Controller is a digital computer used for automation of industrial processes, such as control of machinery on factory assembly lines. Unlike general-purpose computers, the PLC is designed for multiple inputs and output arrangements, extended temperature ranges, immunity to electrical noise, and resistance to vibration and impact. Programs to control machine operation are typically stored in batterybacked or non-volatile memory. A PLC is an example of a real time system since output results must be produced in response to input conditions ithin a bounded time, other ise unintended operation ill result. PLC and Programmable Logic Controller are registered trademarks of the Allen-!radley Company. "CA#A is idely used in industry for "upervisory Control and #ata Ac$uisition of industrial processes, "CA#A systems are no also penetrating the experimental physics laboratories for the controls of ancillary systems such as cooling, ventilation, po er distribution, etc. %ore recently they ere also applied for the controls of smaller si&e particle detectors such as the L' moon detector and the (A)* experiment, to name +ust t o examples at C,-(. "CA#A systems have made substantial progress over the recent years in terms of functionality, scalability, performance and openness such that they are an alternative to in house development even for very demanding and complex control systems as those of physics experiments.

1.! "eature# o$ P C#:-

"ig 1.1 "ingle Allen-!radley PLC /ith each module having sixteen 0points0 of either input or output, this PLC has the ability to monitor and control do&ens of devices. 1it into a control cabinet, a PLC takes up little room, especially considering the e$uivalent space that ould be needed by electromechanical relays to perform the same functions. 2he main difference from other computers is that PLC are armored for severe condition 3dust, moisture, heat, cold, etc4 and has the facility for extensive input5output 36574 arrangements. 2hese connect the PLC to sensors and actuators. PLCs read limit s itches, analog process variables 3such as temperature and pressure4, and the positions of complex positioning systems. "ome even use machine vision. 7n the actuator side, PLCs operate electric motors, pneumatic or hydraulic cylinders, magnetic relays or solenoids, or analog outputs. 2he input5output arrangements may be built into a simple PLC, or the PLC may have external 657 modules attached to a computer net ork that plugs into the PLC. %any hich appeared similar to electrical schematic diagrams. 2he electricians ere $uite able to trace out circuit problems ith schematic diagrams using ladder logic. 2his of the earliest PLCs expressed all decision making logic in simple ladder logic program notation as chosen to reduce training demands for the existing technicians. 7ther early PLCs used a form of instruction list programming, based on a stack-based logic solver. 2he functionality of the PLC has evolved over the years to include se$uential relay control, motion control, process control, distributed control systems and net orking. 2he data handling, 8

storage, processing po er and communication capabilities of some modern PLCs are approximately e$uivalent to desktop computers

"ig 1.! !lock diagram of PLCs

'

1.% &iring in a P C: 'eneration o$ (ignal In)ut

6nside the PLC housing, connected bet een each input terminal and the Common terminal, is an opto-isolator device 3Light-,mitting #iode4 that provides an electrically isolated 0high0 logic signal to the computer9s circuitry 3a photo-transistor interprets the L,#9s light4 hen there is .8: ;AC po er applied bet een the respective input terminal and the Common terminal. An indicating L,# on the front panel of the PLC gives visual indication of an 0energi&ed0 input.

"ig 1.% #iagram "ho ing ,nergi&ed input terminal <.

'eneration o$ Out)ut (ignal

7utput signals are generated by the PLC9s computer circuitry activating a s itching device 3transistor, 2-6AC, or even an electromechanical relay4, connecting the 0"ource0 terminal to any of the 0=-0 labeled output terminals. 2he 0"ource0 terminal, correspondingly, is usually connected to the L. side of the .8: ;AC po er source. As ith each input, an indicating L,# on the front panel of the PLC gives visual indication of an 0energi&ed0 output. 6n this ay, the PLC is able to interface ith real- orld devices such as s itches and solenoids. 2he actual logic of the control system is established inside the PLC by means of a computer program. 2his program dictates hich output gets energi&ed under hich input conditions. Although the program itself appears to be a ladder logic diagram, ith s itch and relay symbols, there are no actual s itch contacts or relay coils operating inside the PLC to create the logical relationships bet een input and output. 2hese are imaginary contacts and coils, if you ill. 2he program is entered and vie ed via a personal computer connected to the PLC9s programming port.

"ig 1.* #iagram "ho ing ,nergi&ed 7utput =.

>

1.* P C com)ared control #+#tem# ,it- ot-er:PLCs are ell-adapted to a certain range of automation tasks. 2hese are typically industrial processes in manufacturing here the cost of developing and maintaining the automation system is high relative to the total cost of the automation, and here changes to the system ould be expected during its operational life. PLCs contain input and output devices compatible ith industrial pilot devices and controls? little electrical design is re$uired, and the design problem centers on expressing the desired se$uence of operations in ladder logic 3or function chart4 notation. PLC applications are typically highly customi&ed systems so the cost of a packaged PLC is lo compared to the cost of a specific custom-built controller design. 1or high volume or very simple fixed automation tasks, different techni$ues are used. A microcontroller-based design ould be appropriate here hundreds or thousands of units ill be produced and so the development cost 3design of po er supplies and input5output hard are4 can be spread over many sales, and here the end-user ould not need to alter the control. Automotive applications are an example? millions of units are built each year, and very fe endusers alter the programming of these controllers. @o ever, some specialty vehicles such as transit busses economically use PLCs instead of custom-designed controls, because the volumes are lo and the development cost ould be uneconomic. PLCs may include logic for single-variable feedback analog control loop, a 0proportional, integral, derivative0 or 0P6# controller.0 A P6# loop could be used to control the temperature of a manufacturing process, for example. @istorically PLCs ere usually configured ith only a fe analog control loops? here processes re$uired hundreds or thousands of loops, a distributed control system 3#C"4 ould instead be used. @o ever, as PLCs have become more po erful, the boundary bet een #C" and PLC applications has become less clear-cut. 1.. Digital and Analog #ignal#:#igital or discrete signals behave as binary s itches, yielding simply an 7n or 7ff signal 3. or :, 2rue or 1alse, respectively4. Pushbuttons, limit s itches, and photoelectric sensors are examples of devices providing a discrete signal. #iscrete signals are sent using either voltage or current, here a specific range is designated as 7n and another as 7ff. 1or example, a PLC might use 8) ; #C 657, ith values above 88 ; #C representing 7n, values belo 8; #C representing 7n, and intermediate values undefined. 6nitially, PLCs had only discrete 657. Analog signals are like volume controls, ith a range of values bet een &ero and full-scale. 2hese are typically interpreted as integer values 3counts4 by the PLC, ith various ranges of accuracy depending on the device and the number of bits available to store the data. As PLCs typically use .A-bit signed binary processors, the integer values are limited bet een -'8,BA* and C'8,BAB. Pressure, temperature, flo , and eight are often represented by analog signals. Analog signals can use voltage or current ith a magnitude proportional to the value of the process signal. 1or example, an analog )-8: mA or : - .: ; input ould be converted into an integer A

value of : D '8BAB.Current inputs are less sensitive to electrical noise 3i.e. from electric motor starts4 than voltage inputs. E/am)le

elders or

As an example, say the facility needs to store ater in a tank. 2he ater is dra n from the tank by another system, as needed, and our example system must manage the ater level in the tank. Using only digital signals, the PLC has t o digital inputs from float s itches 3tank empty and tank full4. 2he PLC uses a digital output to open and close the inlet valve into the tank. 6f both float s itches are off 3do n4 or only the 9tank empty9 s itch is on, the PLC ill open the valve to let more ater in. 7nce the 9tank full9 s itch is on, the PLC ill automatically shut the inlet to stop the ater from overflo ing. 6f only the 9tank full9 s itch is on, something is rong because once the ater reaches a float s itch, the s itch ill stay on because it is floating, thus, hen both float s itches are on, the tank is full. 2 o float s itches are used to prevent a 9flutter9 3a ripple or a ave4 condition here any ater usage activates the pump for a very short time and then deactivates for a short time, and so on, causing the system to ear out faster. An analog system might use a load cell 3scale4 that eighs the tank, and an ad+ustable 3throttling4 valve. 2he PLC could use a P6# feedback loop to control the valve opening. 2he load cell is connected to an analog input and the valve is connected to an analog output. 2his system fills the tank faster hen there is less ater in the tank. 6f the ater level drops rapidly, the valve can be opened ide. 6f ater is only dripping out of the tank, the valve ad+usts to slo ly drip ater back into the tank. A real system might combine both approaches, using float s itches and simple valves to prevent spills, and a rate sensor and rate valve to optimi&e refill rates. !ackup and maintenance methods can make a real system very complicated.

CHAPTER %: PRO'RA00IN'

,arly PLCs, up to the mid-.E*:s, ere programmed using proprietary programming panels or special-purpose programming terminals, hich often had dedicated function keys representing the various logical elements of PLC programs. Programs ere stored on cassette tape cartridges. 1acilities for printing and documentation ere very minimal due to lack of memory capacity. %ore recently, PLC programs are typically ritten in a special application on a personal computer, then do nloaded by a direct-connection cable or over a net ork to the PLC. 2he very oldest PLCs used non-volatile magnetic core memory but no the program is stored in the PLC either in battery-backed-up -A% or some other non-volatile flash memory. B

,arly PLCs ere designed to be used by electricians ho ould learn PLC programming on the +ob. 2hese PLCs ere programmed in 0ladder logic0, hich strongly resembles a schematic diagram of relay logic. %odern PLCs can be programmed in a variety of ays, from ladder logic to more traditional programming languages such as !A"6C and C. Another method is "tate Logic, a ;ery @igh Level Programming Language designed to program PLCs based on "tate 2ransition #iagrams.

%.1 adder logic:Ladder logic is a method of dra ing electrical logic schematics. 6t is no a graphical language very popular for programming Programmable Logic Controllers 3PLCs4. 6t as originally invented to describe logic made from relays. 2he name is based on the observation that programs in this language resemble ladders, ith t o vertical 0rails0 and a series of hori&ontal 0rungs0 bet een them. A program in ladder logic, also called a ladder diagram, is similar to a schematic for a set of relay circuits. An argument that aided the initial adoption of ladder logic as that a ide variety of engineers and technicians ould be able to understand and use it ithout much additional training, because of the resemblance to familiar hard are systems. 32his argument has become less relevant given that most ladder logic programmers have a soft are background in more conventional programming languages, and in practice implementations of ladder logic have characteristics F such as se$uential execution and support for control flo features F that make the analogy to hard are some hat imprecise.4 Ladder logic is idely used to program PLCs, here se$uential control of a process or manufacturing operation is re$uired. Ladder logic is useful for simple but critical control systems, or for re orking old hard ired relay circuits. As programmable logic controllers became more sophisticated it has also been used in very complex automation systems. Ladder logic can be thought of as a rule-based language, rather than a procedural language. A 0rung0 in the ladder represents a rule. /hen implemented ith relays and other electromechanical devices, the various rules 0execute0 simultaneously and immediately. /hen implemented in a programmable logic controller, the rules are typically executed se$uentially by soft are, in a loop. !y executing the loop fast enough, typically many times per second, the effect of simultaneous and immediate execution is obtained. 6n this ay it is similar to other rulebased languages, like spreadsheets or "GL. @o ever, proper use of programmable controllers re$uires understanding the limitations of the execution order of rungs. Ladder logic is idely used to program PLCs, here se$uential control of a process or manufacturing operation is re$uired. Ladder logic is useful for simple but critical control systems or for re orking old hard ired relay circuits. As programmable logic controllers became more sophisticated it has also been used in very complex automation systems. 7ften the ladder logic program is used in con+unction ith an @%6 program operating on a computer ork station. *

%.! E/am)le o$ a #im)le ladder logic )rogram :2he language itself can be seen as a set of connections bet een logical checkers 3relay contacts4 and actuators 3coils4. 6f a path can be traced bet een the left side of the rung and the output, through asserted 3true or 0closed04 contacts, the rung is true and the output coil storage bit is asserted 3.4 or true. 6f no path can be traced, then the output is false 3:4 and the 0coil0 by analogy to electromechanical relays is considered 0de-energi&ed0. 2he analogy bet een logical propositions and relay contact status is due to Claude "hannon. Ladder logic has 0contacts0 that 0make0 or 0break0 0circuits0 to control 0coils.0 ,ach coil or contact corresponds to the status of a single bit in the programmable controller9s memory. Unlike electromechanical relays, a ladder program can refer any number of times to the status of a single bit, e$uivalent to a relay ith an indefinitely large number of contacts. "o-called 0contacts0 may refer to inputs to the programmable controller from physical devices such as pushbuttons and limit s itches, or may represent the status of internal storage bits hich may be generated else here in the program. ,ach rung of ladder language typically has one coil at the far right. "ome manufacturers may allo more than one output coil on a rung. --1 --1
)-)--

a regular coil, true hen its rung is true a regular coil, true hen its rung is true

--1 \ )-- a 0not0 coil, false hen its rung is true --2
]--

A regular contact, true hen its coil is true 3normally false4

--2 \ ]-- A 0not0 contact, false hen its coil is true 3normally true4 2he 0coil0 3output of a rung4 may represent a physical output hich operates some device connected to the programmable controller, or may represent an internal storage bit for use else here in the program. %.% 'enerall+ U#ed In#truction# 3 #+mbol $or P C Programming: In)ut In#truction

--H I-- 2his 6nstruction is Called IXC or Examine If Closed E

ie? 6f a (7 s itch is actuated then only this instruction ill be true. 6f a (C s itch is actuated then this instruction ill not be true and hence output ill not be generated. --H \ ]-- This Instruction is Called IXO or Examine If Open ie? 6f a (C s itch is actuated then only this instruction ill be true. 6f a (C s itch is actuated then this instruction ill not be true and hence output ill not be generated.

Out)ut In#truction

--3 4-- 2his 6nstruction "ho s the "tates of 7utput. ie? 6f any instruction either <67 or <6C is true then output output a 8) volt signal is generated from PLC processor. Rung ill be high. #ue to high

-ung is a simple line on hich instruction are placed and logics are created ,.g? @ere is an example of hat one rung in a ladder logic program might look like. 6n real life, there may be hundreds or thousands of rungs.

1. -----[ ]---------|--[ ]--|------( )----X | Y | S | | |--[ ]--| Z

2he above reali&es the functionJ " K < A(# 3= 7- L4 2ypically, complex ladder logic is 9read9 left to right and top to bottom. As each of the lines 3or rungs4 are evaluated the output coil of a rung may feed into the next stage of the ladder as an input. 6n a complex system there ill be many 0rungs0 on a ladder, hich are numbered in order of evaluation.

.:

1.

-----[ ]----------|---[ ]---|----( )-------X | Y | S | | |---[ ]---| Z 2. -----[ ]----[ ] -------------------( )------S X T
8. 2 K " A(# < here " is e$uivalent to M..above 2his represents a slightly more complex system for rung 8. After the first line has been evaluated, the output coil 3"4 is fed into rung 8, hich is then evaluated and the output coil 2 could be fed into an output device 3bu&&er, light etc..4 or into rung ' on the ladder. 3(ote that the contact < on the 8nd rung serves no useful purpose, as < is already a 9A(#9 function of " from the .st rung.4 2his system allo s very complex logic designs to be broken do n and evaluated. 0ore E/am)le

E/am)le-1

------[ ]--------------[ ]----------------O--Key Switch 1 Key Switch 2 Door Motor

2his circuit sho s t o key s itches that security guards might use to activate an electric motor on a bank vault door. /hen the normally open contacts of both s itches close, electricity is able to flo to the motor hich opens the door. 2his is a logical A(#.

E/am)le-! 7ften e have a little green 0start0 button to turn on a motor, and big red 0"top0 button. e ant to turn it off ith a

--+----[ ]--+----[\]----( )--| st rt | sto! r"# | | +----[ ]--+ r"#

..

-------[ ]--------------( )--r"# $otor

E/am)le &it- P C

Consider the follo ing circuit and PLC programJ

-------[ ]--------------( )--r"# $otor

.8

/hen the push button s itch is unpressed, no po er is sent to the <. input of the PLC. 1ollo ing the program, hich sho s a normally-open <. contact in series ith a =. coil, no 0po er0 ill be sent to the =. coil. 2hus, the PLC9s =. output remains de-energi&ed, and the indicator lamp connected to it remains dark.

6f the push button s itch is pressed, ho ever, po er ill be sent to the PLC9s <. input. Any and all <. contacts appearing in the program ill assume the actuated 3non-normal4 state, as though they ere relay contacts actuated by the energi&ing of a relay coil named 0<.0. 6n this case, energi&ing the <. input ill cause the normally-open <. contact ill 0close,0 sending 0po er0 to the =. coil. /hen the =.coil of the program 0energi&es,0 the real =. output ill become energi&ed, lighting up the lamp connected to it.

"ig %.1 Lamp Nlo s hen at 6nput " itch is Actuated

.'

6t must be understood that the <. contact, =. coil, connecting ires, and 0po er0 appearing in the personal computer9s display are all virtual. 2hey do not exist as real electrical components. 2hey exist as commands in a computer program -- a piece of soft are only -- that +ust happens to resemble a real relay schematic diagram. ,$ually important to understand is that the personal computer used to display and edit the PLC9s program is not necessary for the PLC9s continued operation. 7nce a program has been loaded to the PLC from the personal computer, the personal computer may be unplugged from the PLC, and the PLC ill continue to follo the programmed commands. 6 include the personal computer display in these illustrations for your sake only, in aiding to understand the relationship bet een real-life conditions 3s itch closure and lamp status4 and the program9s status 30po er0 through virtual contacts and virtual coils4. 2he true po er and versatility of a PLC is revealed hen e ant to alter the behavior of a control system. "ince the PLC is a programmable device, e can alter its behavior by changing the commands e give it, ithout having to reconfigure the electrical components connected to it. 1or example, suppose e anted to make this s itch-and-lamp circuit function in an inverted fashionJ push the button to make the lamp turn off, and release it to make it turn on. 2he 0hard are0 solution ould re$uire that a normally-closed pushbutton s itch be substituted for the normally-open s itch currently in place. 2he 0soft are0 solution is much easierJ +ust alter the program so that contact <. is normally-closed rather than normally-open. %.* Programming "or (tart4(to) o$ 0otor b+ P C:7ften e have a little green 0start0 button to turn on a motor, and big red 0"top0 button. e ant to turn it off ith a

--+----[ ]--+----[\]----( )--| st rt | sto! r"# | | +----[ ]--+ r"#

.)

"ig %.% #iagram of "tart5"top of motor 2he pushbutton s itch connected to input <. serves as the 0"tart0 s itch, hile the s itch connected to input <8 serves as the 0"top.0 Another contact in the program, named =., uses the output coil status as a seal-in contact, directly, so that the motor contactor ill continue to be energi&ed after the 0"tart0 pushbutton s itch is released. =ou can see the normally-closed contact <8 appear in a colored block, sho ing that it is in a closed 30electrically conducting04 state. (tarting o$ 0otor

6f e ere to press the 0"tart0 button, input <. ould energi&e, thus 0closing0 the <. contact in the program, sending 0po er0 to the =. 0coil,0 energi&ing the =. output and applying .8: volt AC po er to the real motor contactor coil. 2he parallel =. contact ill also 0close,0 thus latching the 0circuit0 in an energi&ed state.

.>

"ig %.* #iagram of "tarting of motor

ogic $or Continou# Running o$ motor &-en (tart 5utton i# Relea#ed

(o , if e release the 0"tart0 pushbutton, the normally-open <. 0contact0 ill return to its 0open0 state, but the motor ill continue to run because the =. seal-in 0contact0 continues to provide 0continuity0 to 0po er0 coil =., thus keeping the =. output energi&ed.

.A

"ig %.. Logic for continuous running of motor

To (to) t-e 0otor

2o stop the motor, e must momentarily press the 0"top0 pushbutton, hich ill energi&e the <8 input and 0open0 the normally-closed 0contact,0 breaking continuity to the =. 0coil0.

.B

"ig %.6 #iagram of stopping of motor /hen the 0"top0 pushbutton is released, input <8 ill de-energi&e, returning 0contact0 <8 to its normal, 0closed0 state. 2he motor, ho ever, ill not start again until the 0"tart0 pushbutton is actuated, because the 0seal-in0 of =. has been lost.

CHAPTER *: 0EANIN' O" (CADA "CA#A stands for "upervisory Control and #ata Ac$uisition. As the name indicates, it is not a full control system, but rather focuses on the supervisory level. As such, it is a purely soft are package that is positioned on top of hard are to hich it is interfaced, in general via Programmable Logic Controllers 3PLCs4, or other commercial hard are modules. "CA#A systems are used not only in industrial processesJ e.g. steel making, po er generation 3conventional and nuclear4 and distribution, chemistry, but also in some experimental facilities such as nuclear fusion. 2he si&e of such plants range from a fe .::: to several .: thousands input5output 36574 channels. @o ever, "CA#A systems evolve rapidly and are no penetrating the market of plants ith a number of 657 channels of several .:: OJ e kno of t o cases of near to . % 657 channels currently under development. .*

"CA#A systems used to run on #7", ;%" and U(6<? in recent years all "CA#A vendors have moved to (2 and some also to Linux.

*.1 Arc-itecture:2his section describes the common features of the "CA#A products that have been evaluated at C,-( in vie of their possible application to the control systems of the L@C detectors H.I, H8I.

"ig *.1 Architecture of "CA#A

Hard,are Arc-itecture

7ne distinguishes t o basic layers in a "CA#A systemJ the 0client layer0 hich caters for the man machine interaction and the 0data server layer0 hich handles most of the process data control activities. 2he data servers communicate ith devices in the field through process controllers. Process controllers, e.g. PLCs, are connected to the data servers either directly or via net orks or field buses that are proprietary 3e.g. "iemens @.4, or non-proprietary 3e.g. Profibus4. #ata servers are connected to each other and to client stations via an ,thernet LA(. 2he data servers and client stations are (2 platforms but for many products the client stations may also be /E> machines.

.E

Communication#

"erver-client and server-server communication is in general on a publish-subscribe and eventdriven basis and uses a 2CP56P protocol, i.e., a client application subscribes to a parameter hich is o ned by a particular server application and only changes to that parameter are then communicated to the client application. 2he data servers poll the controllers at a user defined polling rate. 2he polling rate may be different for different parameters. 2he controllers pass the re$uested parameters to the data servers. 2ime stamping of the process parameters is typically performed in the controllers and this time-stamp is taken over by the data server. 6f the controller and communication protocol used support unsolicited data transfer then the products ill support this too. 2he products provide communication drivers for most of the common PLCs and idely used field-buses, e.g., %odbus. 7f the three fieldbuses that are recommended at C,-(, both Profibus and /orld flip are supported but CA(bus often not H'I. "ome of the drivers are based on third party products 3e.g., Applicom cards4 and therefore have additional cost associated ith them. ;%, on the other hand is generally not supported. A single data server can support multiple communications protocolsJ it can generally support as many such protocols as it has slots for interface cards. 2he effort re$uired to develop ne drivers is typically in the range of 8-A eeks depending on the complexity and similarity ith existing drivers, and a driver development toolkit is provided for this. Inter$acing

2he provision of 7PC client functionality for "CA#A to access devices in an open and standard manner is developing. 2here still seems to be a lack of devices5controllers, hich provide 7PC server soft are, but this improves rapidly as most of the producers of controllers are actively involved in the development of this standard. 7PC has been evaluated by the C,-(-62-C7 group H)I. 2he products also provide An 7pen #ata !ase Connectivity 37#!C4 interface to the data in the archive5logs, but not to the configuration database, An A"C66 import5export facility for configuration data, A library of AP6s supporting C, CCC, and ;isual !asic 3;!4 to access data in the -2#!, logs and archive. 2he AP6 often does not provide access to the product9s internal features such as alarm handling, reporting, trending, etc.

2he PC products provide support for the %icrosoft standards such as #ynamic #ata ,xchange 3##,4 hich allo s e.g. to visuali&e data dynamically in an ,<C,L spreadsheet, #ynamic Link Library 3#LL4 and 7b+ect Linking and ,mbedding 37L,4. 8:

2he configuration data are stored in a database that is logically centrali&ed but physically distributed and that is generally of a proprietary format. 1or performance reasons, the -2#! resides in the memory of the servers and is also of proprietary format. 2he archive and logging format is usually also proprietary for performance reasons, but some products do support logging to a -elational #ata !ase %anagement "ystem 3-#!%"4 at a slo er rate either directly or via an 7#!C interface. (calabilit+

"calability is understood as the possibility to extend the "CA#A based control system by adding more process variables, more speciali&ed servers 3e.g. for alarm handling4 or more clients. 2he products achieve scalability by having multiple data servers connected to multiple controllers. ,ach data server has its o n configuration database and -2#! and is responsible for the handling of a sub-set of the process variables 3ac$uisition, alarm handling, archiving4. Redundanc+

2he products often have built in soft are redundancy at a server level, hich is normally transparent to the user. %any of the products also provide more complete redundancy solutions if re$uired.
*.! "unctionalit+: Acce## Control

Users are allocated to groups, hich have defined read5 rite access privileges to the process parameters in the system and often also to specific product functionality. 00I

2he products support multiple screens, hich can contain combinations of synoptic diagrams and text. 2hey also support the concept of a 0generic0 graphical ob+ect ith links to process variables. 2hese ob+ects can be 0dragged and dropped0 from a library and included into a synoptic diagram. %ost of the "CA#A products that ere evaluated decompose the process in 0atomic0 parameters 3e.g. a po er supply current, its maximum value, its on5off status, etc.4 to hich a 2ag-name is associated. 2he 2ag-names used to link graphical ob+ects to devices can be edited as re$uired. 2he products include a library of standard graphical symbols, many of hich ould ho ever not be applicable to the type of applications encountered in the experimental physics community. "tandard indo s editing facilities are providedJ &ooming, re-si&ing, scrolling... 7n-line

8.

configuration and customi&ation of the %%6 is possible for users ith the appropriate privileges. Links can be created bet een display pages to navigate from one vie to another. Trending

2he products all provide trending facilities and one can summari&e the common capabilities as follo sJ 2he parameters to be trended in a specific chart can be predefined or defined on-line A chart may contain more than * trended parameters or pens and an unlimited number of charts can be displayed 3restricted only by the readability4 -eal-time and historical trending are possible, although generally not in the same chart

Alarm Handling

Alarm handling is based on limit and status checking and performed in the data servers. %ore complicated expressions 3using arithmetic or logical expressions4 can be developed by creating derived parameters on hich status or limit checking is then performed. 2he alarms are logically handled centrally, i.e., the information only exists in one place and all users see the same status 3e.g., the ackno ledgement4, and multiple alarm priority levels 3in general many more than ' such levels4 are supported. 6t is generally possible to group alarms and to handle these as an entity 3typically filtering on group or ackno ledgement of all alarms in a group4. 1urthermore, it is possible to suppress alarms either individually or as a complete group. 2he filtering of alarms seen on the alarm page or hen vie ing the alarm log is also possible at least on priority, time and group. @o ever, relationships bet een alarms cannot generally be defined in a straightfor ard manner. ,-mails can be generated or predefined actions automatically executed in response to alarm conditions. ogging4Arc-i7ing

2he terms logging and archiving are often used to describe the same facility. @o ever, logging can be thought of as medium-term storage of data on disk, hereas archiving is long-term storage of data either on disk or on another permanent storage medium. Logging is typically performed on a cyclic basis, i.e., once a certain file si&e, time period or number of points is reached the data is over ritten. Logging of data can be performed at a set fre$uency, or only initiated if the value changes or hen a specific predefined event occurs. Logged data can be transferred to an archive once the log is full. 2he logged data is time-stamped and can be filtered hen vie ed by a user. 2he logging of user actions is in general performed together ith either a user 6# or station 6#. 2here is often also a ;C- facility to play back archived data. Re)ort 'eneration

7ne can produce reports using "GL type $ueries to the archive, -2#! or logs. Although it is sometimes possible to embed ,<C,L charts in the report, a 0cut and paste0 capability is in 88

general not provided. 1acilities exist to be able to automatically generate, print and archive reports. Automation

2he ma+ority of the products allo actions to be automatically triggered by events. A scripting language provided by the "CA#A products allo s these actions to be defined. 6n general, one can load a particular display, send an ,mail, run a user defined application or script and rite to the -2#!. 2he concept of recipes is supported, hereby a particular system configuration can be saved to a file and then re-loaded at a later date.

CHAPTER .: APP ICATION DE8E OP0ENT

"ig ..1 7il P Nas "CA#A installations 8'

..1 Con$iguration:2he development of the applications is typically done in t o stages. 1irst the process parameters and associated information 3e.g. relating to alarm conditions4 are defined through some sort of parameter definition template and then the graphics, including trending and alarm displays are developed, and linked here appropriate to the process parameters. 2he products also provide an A"C66 ,xport56mport facility for the configuration data 3parameter definitions4, hich enables large numbers of parameters to be configured in a more efficient manner using an external editor such as ,xcel and then importing the data into the configuration database. @o ever, many of the PC tools no have a /indo s ,xplorer type development studio. 2he developer then orks ith a number of folders, hich each contains a different aspect of the configuration, including the graphics. 2he facilities provided by the products for configuring very large numbers of parameters are not very strong. @o ever, this has not really been an issue so far for most of the products to-date, as large applications are typically about >:O 657 points and database population from ithin an A"C66 editor such as ,xcel is still a orkable option. 7n-line modifications to the configuration database and the graphics are generally possible the appropriate level of privileges. 2he follo ing development tools are provided as standardJ A graphics editor, ith standard dra ing facilities including freehand, lines, s$uares circles, etc. 6t is possible to import pictures in many formats as ell as using predefined symbols including e.g. trending charts, etc. A library of generic symbols is provided that can be linked dynamically to variables and animated as they change. 6t is also possible to create links bet een vie s so as to ease navigation at run-time. A data base configuration tool 3usually through parameter templates4. 6t is in general possible to export data in A"C66 files so as to be edited through an A"C66 editor or ,xcel. A scripting language An Application Program 6nterface 3AP64 supporting C, CCC, ;!

ith

..! E7olution:"CA#A vendors release one ma+or version and one to t o additional minor versions once per year. 2hese products evolve thus very rapidly so as to take advantage of ne market opportunities, to meet ne re$uirements of their customers and to take advantage of ne technologies. As as already mentioned, most of the "CA#A products that ere evaluated decompose the process in 0atomic0 parameters to hich a 2ag-name is associated. 2his is impractical in the case of very large processes hen very large sets of 2ags need to be configured. As the industrial 8)

applications are increasing in si&e, ne "CA#A versions are no being designed to handle devices and even entire systems as full entities 3classes4 that encapsulate all their specific attributes and functionality. 6n addition, they ill also support multi-team development. As far as ne technologies are concerned, the "CA#A products are no adoptingJ

/eb technology, Active<, Qava, etc. 7PC as a means for communicating internally bet een the client and server modules. 6t should thus be possible to connect 7PC compliant third party modules to that "CA#A product.

..% Engineering:/hilst one should rightly anticipate significant development and maintenance savings by adopting a "CA#A product for the implementation of a control system, it does not mean a 0no effort0 operation. 2he need for proper engineering can not be sufficiently emphasi&ed to reduce development effort and to reach a system that complies ith the re$uirements, that is economical in development and maintenance and that is reliable and robust. ,xamples of engineering activities specific to the use of a "CA#A system are the definition ofJ

A library of ob+ects 3PLC, device, subsystem4 complete ith standard ob+ect behavior 3script, se$uences, ...4, graphical interface and associated scripts for animation, 2emplates for different types of 0panels0, e.g. alarms, 6nstructions on ho to control e.g. a device ..., A mechanism to prevent conflicting controls 3if not provided ith the "CA#A4, alarm levels, behavior to be adopted in case of specific alarms.

8>

..* Potential bene$it# o$ (CADA:2he benefits one can expect from adopting a "CA#A system for the control of experimental physics facilities can be summari&ed as follo sJ A rich functionality and extensive development facilities. 2he amount of effort invested in "CA#A product amounts to >: to .:: p-yearsR 2he amount of specific development that needs to be performed by the end-user is limited, especially ith suitable engineering. -eliability and robustness. 2hese systems are used for mission critical industrial processes here reliability and performance are paramount. 6n addition, specific development is performed ithin a ell-established frame ork that enhances reliability and robustness. 2echnical support and maintenance by the vendor.

8A

CHAPTER 6 : CONC U(ION

"CA#A is used for the constructive orking not for the destructive ork using a "CA#A system for their controls ensures a common frame ork not only for the development of the specific applications but also for operating the detectors. 7perators experience the same 0look and feel0 hatever part of the experiment they control. @o ever, this aspect also depends to a significant extent on proper engineering.

8B

RE"ERENCE(.

.. A.#aneels, /."alter, 02echnology "urvey "ummary of "tudy -eport0, 62C75E*-:*-:E, C,-(, Neneva 8Ath Aug .EE*. 8. A.#aneels, /."alter, 0"election and ,valuation of Commercial "CA#A "ystems for the Controls of the C,-( L@C ,xperiments0, Proceedings of the .EEE 6nternational Conference on Accelerator and Large ,xperimental Physics Control "ystems, 2rieste, .EEE, p.'>'. '. N.!aribaud et al., 0-ecommendations for the Use of 1ieldbuses at C,-( in the L@C ,ra0, Proceedings of the .EEB 6nternational Conference on Accelerator and Large ,xperimental Physics Control "ystems, !ei+ing, .EEB, p.8*>. ). -.!arillere et al., 0-esults of the 7PC ,valuation done ithin the QC7P for the Control of the L@C ,xperiments0, Proceedings of the .EEE 6nternational Conference on Accelerator and Large ,xperimental Physics Control "ystems, 2rieste, .EEE, p.>...

8*