自动化专业可参考的外文文献.doc

1外文原文 A: Fundamentals of Singlechip Microcomputer The single-chip microcomputer is the culmination of both the development of the digital computer and the integrated circuit arguably the tow most significant inventions of the 20th century 1。 These tow types of architecture are found in single-chip microcomputer。 Some employ the split program/data memory of the Harvard architecture， shown in Fig.35A-1， others follow the philosophy， widely adapted for general-purpose computers and microprocessors, of making no logical distinction between program and data memory as in the Princeton architecture, shown in Fig.3-5A2。 In general terms a singlechip microcomputer is characterized by the incorporation of all the units of a computer into a single device， as shown in Fig35A3。ProgrammemoryInput&Outputunit CPUDatamemory Fig。3-5A-1 A Harvard typeInput&Outputunit CPUmemory Fig.3-5A2. A conventional Princeton computerTimer/Counter SystemclockExternalTimingcomponentsSerial I/O ROM Reset Prarallel I/O RAM Interrupts CPU Power Fig35A3。 Principal features of a microcomputer Read only memory (ROM）。ROM is usually for the permanent， nonvolatile storage of an applications program .Many microcomputers and microcontrollers are intended for highvolume applications and hence the economical manufacture of the devices requires that the contents of the program memory be committed permanently during the manufacture of chips . Clearly， this implies a rigorous approach to ROM code development since changes cannot be made after manufacture .This development process may involve emulation using a sophisticated development system with a hardware emulation capability as well as the use of powerful software tools。 Some manufacturers provide additional ROM options by including in their range devices with （or intended for use with） user programmable memory。 The simplest of these is usually device which can operate in a microprocessor mode by using some of the input/output lines as an address and data bus for accessing external memory. This type of device can behave functionally as the single chip microcomputer from which it is derived albeit with restricted I/O and a modified external circuit. The use of these ROMless devices is common even in production circuits where the volume does not justify the development costs of custom onchip ROM2；there can still be a significant saving in I/O and other chips compared to a conventional microprocessor based circuit。 More exact replacement for ROM devices can be obtained in the form of variants with piggy-back EPROM（Erasable programmable ROM )sockets or devices with EPROM instead of ROM 。These devices are naturally more expensive than equivalent ROM device, but do provide complete circuit equivalents. EPROM based devices are also extremely attractive for lowvolume applications where they provide the advantages of a single-chip device， in terms of onchip I/O, etc. ,with the convenience of flexible user programmability。Random access memory （RAM）.RAM is for the storage of working variables and data used during program execution. The size of this memory varies with device type but it has the same characteristic width (4，8,16 bits etc.） as the processor ,Special function registers, such as stack pointer or timer register are often logically incorporated into the RAM area. It is also common in Harard type microcomputers to treat the RAM area as a collection of register; it is unnecessary to make distinction between RAM and processor register as is done in the case of a microprocessor system since RAM and registers are not usually physically separated in a microcomputer .Central processing unit (CPU）。The CPU is much like that of any microprocessor。 Many applications of microcomputers and microcontrollers involve the handling of binary-coded decimal （BCD） data （for numerical displays, for example） ,hence it is common to find that the CPU is well adapted to handling this type of data 。It is also common to find good facilities for testing， setting and resetting individual bits of memory or I/O since many controller applications involve the turning on and off of single output lines or the reading the single line。 These lines are readily interfaced to twostate devices such as switches, thermostats, solid-state relays, valves， motor， etc.Parallel input/output。 Parallel input and output schemes vary somewhat in different microcomputer； in most a mechanism is provided to at least allow some flexibility of choosing which pins are outputs and which are inputs。 This may apply to all or some of the ports. Some I/O lines are suitable for direct interfacing to， for example, fluorescent displays， or can provide sufficient current to make interfacing other components straightforward。 Some devices allow an I/O port to be configured as a system bus to allow offchip memory and I/O expansion。 This facility is potentially useful as a product range develops， since successive enhancements may become too big for onchip memory and it is undesirable not to build on the existing software base。Serial input/output .Serial communication with terminal devices is common means of providing a link using a small number of lines。 This sort of communication can also be exploited for interfacing special function chips or linking several microcomputers together 。Both the common asynchronous synchronous communication schemes require protocols that provide framing （start and stop) information 。This can be implemented as a hardware facility or U（S)ART（Universal（synchronous） asynchronous receiver/transmitter) relieving the processor (and the applications programmer） of this low-level， timeconsuming， detail. t is merely necessary to selected a baudrate and possibly other options (number of stop bits, parity， etc.） and load （or read from） the serial transmitter （or receiver) buffer。 Serialization of the data in the appropriate format is then handled by the hardware circuit。Timing/counter facilities. Many application of singlechip microcomputers require accurate evaluation of elapsed real time 。This can be determined by careful assessment of the execution time of each branch in a program but this rapidly becomes inefficient for all but simplest programs .The preferred approach is to use timer circuit that can independently count precise time increments and generate an interrupt after a preset time has elapsed .This type of timer is usually arranged to be reloadable with the required count 。The timer then decrements this value producing an interrupt or setting a flag when the counter reaches zero. Better timers then have the ability to automatically reload the initial count value。 This relieves the programmer of the responsibility of reloading the counter and assessing elapsed time before the timer restarted ,which otherwise wound be necessary if continuous precisely timed interrupts were required (as in a clock ,for example)。Sometimes associated with timer is an event counter。 With this facility there is usually a special input pin ，that can drive the counter directly. Timing components。 The clock circuitry of most microcomputers requires only simple timing components. If maximum performance is required,a crystal must be used to ensure the maximum clock frequency is approached but not exceeded. Many clock circuits also work with a resistor and capacitor as low-cost timing components or can be driven from an external source. This latter arrangement is useful is external synchronization of the microcomputer is required. WORDS AND TERMSculmination n.顶点 spilt adj。分离的volatile n. 易变的commit v。保证albeit conj。虽然custom adj。定制的variant adj.不同的piggy-back adj.背负式的socket n。 插座B:PLC1PLCs （programmable logical controller） face ever more complex challenges these days 。 Where once they quietly replaced relays and gave an occasional report to a corporate mainframe， they are now grouped into cells， given new job and new languages, and are forced to compete against a growing array of control products。 For this years annual PLC technology update ,we queried PLC makers on these topics and more 。Programming languages Higher level PLC programming languages have been around for some time ，but lately their popularity has mushrooming。 ”As Raymond Leveille， vice president & general manager, Siemens Energy Automation 。inc； Programmable controls are being used for more and more sophisticated operations， languages other than ladder logic become more practical, efficient, and powerful。 For example, its very difficult to write a trigonometric function using ladder logic 。Languages gaining acceptance include Boolean， control system flowcharting， and such function chart languages as Graphcet and its variation .And theres increasing interest in languages like C and BASIC。PLCs in process controlThus far， PLCs have not been used extensively for continuous process control .Will this continue? ”The feeling that Ive gotten, says Ken Jannotta, manger, product planning, series One and Series Six product ，at GE Fanuc North America ,is that PLCs will be used in the process industry but not necessarily for process control。Several vendors -obviously betting that the opposite will happen -have introduced PLCs optimized for process application 。Rich Ryan， manger, commercial marketing, Allenbradley Programmable Controls Div。， cites PLCss increasing use such industries as food ,chemicals ，and petroleum。 Ryan feels there are two types of applications in which theyre appropriate。 ”one,” he says, is where the size of the process control system thats being automated doesnt justify DCSdistributed control system。With the starting price tags of chose products being relatively high, a programmable controller makes sense for small, low loop count application .The second is where you have to integrate the loop closely with the sequential logical 。Batch controllers are prime example ,where the sequence and maintaining the process variable are intertwined so closely that the benefits of having a programmable controller to do the sequential logical outweighs some of the disadvantages of not having a distributed control system。Bill Barkovitz， president of Triconex, predicts that all future controllers that come out in the process control system business will embrace a lot of more PLC technology and a lot more PLC functionality than they ever did before .”Communications and MAPCommunications are vital to an individual automation cell and to be automated factory as a whole。 Weve heard a lot about MAP in the last few years ，and a lot of companies have jumped on the bandwagon.2Many， however, were disappointed when a fully-defined and completed MAP specification didnt appear immediately 。Says Larry Komarek： ”Right now， MAP is still a moving target for the manufacturers, a specification that is not final .Presently, for example。 people are introducing products to meet the MAP2.1standard 。Yet2。1based products will be obsolete when the new standard for MAP3。0 is introduced.Because of this, many PLC vendors are holding off on full MAP implementations. Omron， for example, has an ongoing MAPcompatibility program；3but Frank Newburn， vice president of Omrons Industrial Division ,reports that because of the lack of a firm definition ,Omrons PLCs dont yet talk to MAP。Since its unlikely that an individual PLC would talk to broad MAP anyway, makers are concentrating on proprietary networks. According to Sal Provanzano, users fear that if they do get on board and vendors withdraw from MAP， theyll be the ones left holding a communications structure thats not supported.Universal I/OWhile there are concerns about the lack of compatible communications between PLCs from different vendors, the connection at the other end-the I/Ois even more fragmented 。With rare exceptions, I/O is still proprietary .Yet there are those who feel that I/O will eventually become more universal .GE Fanuc is hoping to do that with its Genius smart I/O line。 The independent I/O makers are pulling in the same direction. Many say that I/O is such a highvalue item that PLC makers will always want to keep it proprietary 。As Ken Jannotta， says： The I/O is going to be a disproportionate amount of the hardware sale. Certainly each PLC vendor is going to try to protect that。 ”For that reason, he says， PLC makers wont begin selling universal I/O system from other vendor。 if we start selling that kind of product， says jannotta, ”what do we manufacture？With more intelligent I/O appearing, Sal Provanzano feels this will lead to more differentiation among I/O from different makers. Where the I/O becomes extremely intelligent and becomes part of the system, he says, it really is hard to define which is the I/O and which is CPU. It really CPU, if you will， is equally integrated into the system as the I/O.Connecting PLC I/O to PCsWhile different PLCs probably will continue to use proprietary I/O， several vendors make it possible to connect5 their I/O to IBM PCcompatible equipment. Allebradeley, Could， and Cincinnati Milacron already have， and rumor has it that GE is planning something along these same lines .4Bill Ketelhut, manage of product planning at GE Fanuc North America ，sees this sort of thing as alternative to universal I/O.”I think the trend ,instead of toward universal I/O, will be multiple host interface ，” he says 。Jodie Glore ，director of marking， Square D Automation Products， Views it as another indication that PLCs are， and have been for some time, industrial computers。PLCs VS PCsIf the IBM 7552, the Action Instruments BC22，and other computers are appearing on the factory floor, wont this mean new competition for PLCs？ Rich Ryan： ”There are some control functions that are better jobs for computers。 Programmable controllers have been forced to fit into those applications。 Yet， the majority of vendors we surveyed dont like the PC invasion will pose a problem for them .Most said that PLCs and PCs are enough apart in architecture that they will usually do the control。 They dont feel that PCs will take jobs from PLCs just because PLC I/O modules can now be connected to PCs； they believe this simply means that PLCs and PCs will be able to share the same data.There are inherent architectural differences between a general purpose computer, says Rich Ryan， ”and a programmable controller 。There are hardware constructs built into almost every manufactures programmable controller today that customize the hardware to run ladder logic and to solve machine code。 ”One fundamental difference he cites is called state of the machine 。Ryan： ”When you shut the machine off, or interrupt the cycle， or you jump to another spot