1、.1英文原文Data ConvertersThe current upsurge in the use of digital processing techniques ,brought about by developments in very large scale integration, has forced the role of the interface elements between analog and digital systems to the forefront of attention. Data converters are these interface ele

2、ments. Although they were once regarded by electronics engineers as e*pensive, rather specialized pieces of equipment , they are now monplace . They are available as self-contained functional units, in both modular and low cost integrated circuit from . No one with an interest in modern electronic i

3、nstrumentation techniques can afford tube without a sound knowledge and understanding of the operation, capabilities and limitations of these devices.USEFULNESS OF CONVERTERSIn many of the real world systems studied by scientists and engineers the system parameters are continuously changing quantiti

4、es (analog variables)and when electronic measurement techniques are used data is derived in analog form as the electrical output signal of a transducer .It is perfectly possible to process ,manipulate and store analog data using a purely analog electronic system : negative feedback techniques can ma

5、ke analog systems perform very precisely. 1However , the accuracy of a purely analog system is often not usable because of the difficulty of rending , recording and interpreting analog data with high accuracy . Also , when large amounts of analog data are involved the task of analysis and storage as

6、sumes mammoth proportions .Digital electronic systems can be made to process rapidly and accurately , to manipulate and to store large amounts of data . The advent of low cost digital microprocessor systems has drastically reduced the cost of implementing digital data processing . Microprocessors ma

7、ke it possible to e*tend the use of electronic digital techniques into areas where they were formerly not considered practicable because of economic considerations . However ,m digital circuits only operate on digital data the scientist who wishes to avail himself of the power of digital techniques

8、must first transform his analog data into digital form . A system called an analog to digital converter,ADC ,is used to perform this function and a system called a digital to analog converter ,DAC ,performs the opposite type of conversion , transforming digital data into analog from . ADCs and DACs

9、provide the essential interface which is required between analog and digital systems . The ADC allows a digital system to take in information from an outside analog system-the digital system can then rapidly process and analyses this information . The DAC allows the results of such analyses to be mu

10、nicated back to the analog system perhaps to modify or control its action .ADC and DAC systems are not new but until paratively recently they were e*pensive to implement and have consequently been regarded as rather specialized pieces of equipment . The advent of low cost monolithic IC converter dev

11、ices has changed this position: they make awide range of versatile signal processing techniques economically available to the measurement scientist who is prepared to invest some of his time in familiarizing himself with the latest devices and their capabilities . The material which follows is inten

12、ded to serve as a first step in such a familiarisation e*ercise.Rather than attempt a survey of available converter devices , which , in a fastchanging area such as this ,would be doomed to obsolescence even before it was pleted , we concentrate on the principles underlying the most popular conversi

13、on techniques . Specific IC devices are mentioned , but only as a means of relating the discussion to practical circuits which you can e*perimentally evaluate for yourself-there is no substitute for hands on practical e*perience in this type of learning e*ercise . However , it should not be assumed

14、that the devices referred to are he only ones , nor indeed the best ones available (from a cost /performance standpoint). The choice of the best converter for a specific application can only be made from athorough study of the manufacturers latest product guides .ANALOG TO DIGITAL CONVERSION TECHNIQ

15、USparison Of A/D Conversion TechniquesA brief summary is now given in which the main characteristics of devices which use the various conversion techniques that have been discussed will be pared . There are normally there dominant factors which govern the choice of the A/D converter to be used in an

16、 application : speed , accuracy and cost . It is convenient to pare the different techniques in relation to these three factors .Speed limitations are inherent in the particular techniques but arrange of speeds can be e*pected for different converter devices using the same conversion technique . A s

17、peed/accuracy design promise is inherent in all the conversion techniques . Increase accuracy is obtained at the e*pense of a reduction in conversion speed . In general , cost is directly related to speed and accuracy , but the cost of a converter device(indeed of any device) , is greatly influenced

18、 by market factors which might be quit unrelated to the performance of the device . The bigger the market is , the more firms there will be attempting to get a slice of it , and petition brings down the cost .As far as inherent speed imitations are concerned the techniques discussed in this chapter

19、in order of decreasing speed (increase in conversion time )are : parallel conversion , successive appro*imation conversion , tracking conversion and integrating conversion .If sheer conversion speed is a dominating factor in an application , the designer will have to choose a converter which uses th

20、e parallel conversion technique . At the time of writing there are not many ultra-fast converters on the market and they tend to be e*pensive , particularly the higher resolution devices . Some firms which make very fast converters are TRW LSI Products , Datel-Intersil, puter Labs and Motorola . Ava

21、ilable parallel converters achieve an 8-bit conversion time of the order of 30ns . Applications in which conversion time must be very short are increasing for e*ample, radar , digital TV and fast transient event recorders . 11,12 No doubt this growing range of applications will interest more manufac

22、turers in very high speed converters and will result in a greater number of less costly devices appearing on the market.In converters which use the successive appro*imation technique conversion time is reated to the number of bits(to the resolution ) . The fastest successive appro*imation converters

23、 currently available are 8-bit devices which have a conversion time of the order of 0.8us. General purpose 8-bit devices may be e*pected to have conversion time of the order of 30us . High resolution 16-bit successive appro*imation converter modules have conversion times of the order of 400us .Succe

24、ssive appro*imation is probably the most widely used technique in the A/D converters which are currently available ; there are a large number of manufacturers making many different successive appro*imation devices . The successive appro*imation technique has more possible sources of error than the s

25、lower integrating techniques . The highest performance successive appro*imation converters are usually the discrete ponent modular type . Costs rise steeply if the application calls for state-of-the-art speed and accuracy . Successive appro*imation converters must be preceded by a sample/hold module

26、 when they are used to digitize rapidly changing analog signals(see section 5.3 ).Tracking converters , unlike successive appro*imation types , do not require an input sample/hold . Although they use simpler digital logic circuitry than successive appro*imation tyes they are subject to similar error

27、s , but do not have the speed of a successive appro*imation converter preceded by a sample/hold(see section 5.8.5). There are paratively few tracking or counting-type converters on the market pared with successive appro*imation and integrating types.Integrating converters are the slowest type but th

28、ey are capable of very high accuracy . Conversion times of currently available integrating types range from 0.3 ms for a fast 8-bit converter to 250 ms for a slow 16-bit one .The integrating technique has the advantage of averaging-out noise ponents of the analog input signal whose frequencies are h

29、igher than the inverse of the integration period . Integrating converters do not require sample/holds. If the application is one in which the analog signals are slowly varying an integrating converter will provide the greatest accuracy for the least cost . Integrating converters are available from a

30、 large number of manufacturers .SUCCESSIVE APPRO*IMATION A/D CONVERTERThe successive appro*imation A/D conversion technique provides a more rapid conversion than the other two feedback techniques . The digital logic circuitry used in this technique instead of incrementing the D/A converter output on

31、e LSB at time , performs a series of trial conversions, In the first trial the control logic applies the MSB to the D/A converter and the analog output of the D/A converter (1/2 full scale) is pared with the analog input signal by the parator . If the DACs output is less than the analog input the MS

32、B is retained ; if the DACs output is greater than the analog input the MSB is switched OFF. The control logic then goes on to apply the nest MSB which is retained or discarded depending on the result of the parison between the DACs output and analog input . The process of trying the addition of suc

33、cessively smaller bits and retaining or discarding them depending on the result of the trial , goes on until the LSB is reached ; the conversion is then plete .A simplifiedrepresentation of the timing sequence which occurs in a typical 4-bit successive appro*imation A/D conversion is shown in figure

34、 .DIGITAL TO ANALOG CONVERTER APPLICATIONSUntil quite recently there has tended to be a clear distinction between analog and digital electronic systems . Functional operations such as amplification , wave-form generation , filtering , modulation and demodulation , etc ., have traditionally been rega

35、rded as purely analog operations ; In many cases the implementation of such functions has been founded on operational amplifier based circuitry .1 Logical operations involved in process control and mathematical operations per-formed on abstract digital numbers have been the main area of application

36、for digital electronic systems . Digital systems are founded on circuitry using gates , flip flops , counters , shift registers ,etc .One of the main reasons for the divisions of electronic systems between analog and digital has been the high cost of data converters . data converters are the essenti

37、al interface ponents which allow analog and digital devices to word together . The ine*pensive monolithic data converters which are now available make it economically viable to bine analog and digital devices . Such a converters permit a precise digital control to be added to analog functional appli

38、cations ; they also allow many of the traditionally analog functional operations to be performed with circuitry based on digital devices . This chapter gives e*amples of some of the many analog/digital applications that are possible with the use of data converters .中文翻譯數(shù)據(jù)轉(zhuǎn)換器超大規(guī)模集成電路的開展，掀起了使用數(shù)字處理技術(shù)的高

39、潮，促進(jìn)人們必須重視模擬和數(shù)字系統(tǒng)間接口部件的作用。數(shù)據(jù)轉(zhuǎn)換器就是這種接口部件。他們雖曾被工程師一度認(rèn)為昂貴而專門轉(zhuǎn)化的裝置，而現(xiàn)在已經(jīng)平凡無奇。他們可作為整體的功能單元，以模塊形式或廉價的集成電路形式提供。對現(xiàn)代電子儀器技術(shù)感興趣的人，都應(yīng)該熟悉或了解這些器件的操作、能力和圍。轉(zhuǎn)換器的用途在科學(xué)家和工程師所研究的許多現(xiàn)實(shí)系統(tǒng)中，系統(tǒng)的參數(shù)是連續(xù)變化的量。在采用電子測量技術(shù)的情況下，數(shù)據(jù)是以傳感器電輸出信號那樣的模擬形式取得的。要對模擬數(shù)據(jù)進(jìn)展處理、操作和存貯，是用純模擬的電子系統(tǒng)是完全可能的：負(fù)反響技術(shù)能夠使模擬系統(tǒng)十分準(zhǔn)確的工作。然而，由于高精度的模擬數(shù)據(jù)難以讀出、記錄和解譯，純模擬系統(tǒng)的

40、精度常常不便于使用。而且當(dāng)涉及到大量模擬數(shù)據(jù)的時候，數(shù)據(jù)的分析與存貯要花費(fèi)極大的工作量。數(shù)字的電子系統(tǒng)可以做到處理快速而準(zhǔn)確，且操作和存貯的數(shù)據(jù)量可很大，廉價的數(shù)字為處理器系統(tǒng)的出現(xiàn)以急劇地降低了實(shí)施數(shù)字?jǐn)?shù)據(jù)處理的費(fèi)用。以前由于經(jīng)濟(jì)上的緣故，認(rèn)為不能推廣應(yīng)用電子數(shù)字技術(shù)的領(lǐng)域，微處理器以使之成為可能。不過，數(shù)字電路只能操作數(shù)字的數(shù)據(jù)，科學(xué)家想要利用數(shù)字技術(shù)的能力必須首先將模擬的數(shù)據(jù)轉(zhuǎn)換成為數(shù)字的形式。所謂“模數(shù)轉(zhuǎn)換器ADC的系統(tǒng)可以實(shí)現(xiàn)上述功能，而稱為“數(shù)模轉(zhuǎn)換器DAC的系統(tǒng)則實(shí)現(xiàn)相反的轉(zhuǎn)換，即將數(shù)字?jǐn)?shù)據(jù)轉(zhuǎn)換為模擬形式。ADC和DAC為模擬系統(tǒng)與數(shù)字系統(tǒng)之間提供了必要的根本接口。ADC可是數(shù)字

41、系統(tǒng)從外界的模擬系統(tǒng)取得信息，數(shù)字系統(tǒng)就能很快處理與分析這些信息。DAC可將這種分析結(jié)果回送給模擬系統(tǒng)，調(diào)整或控制其動作。ADC和DAC并非新器件，只是不久以前他們的價格頗為昂貴，一直被認(rèn)為是設(shè)備中相當(dāng)專門化的局部。廉價的單片IC集成電路轉(zhuǎn)換器件的出現(xiàn)改變了這種情況：他們可以使通用圍較廣的信號處理技術(shù)為量測科學(xué)家經(jīng)濟(jì)地所利用，只要他們愿意花費(fèi)一些時間去熟悉最新的器件及其功能。轉(zhuǎn)換器的開展是如此之快，往往在其打到完備以前就陷于舊。我們與其介紹一些常用的器件，還不如著重于一般通用的轉(zhuǎn)換技術(shù)的原理。具體的IC器件也要有所涉及，但只是便于聯(lián)系實(shí)際線路進(jìn)展討論。在這類學(xué)習(xí)的訓(xùn)練中“親手實(shí)踐是最好的方法。

42、要為特定的應(yīng)用選擇“最好的轉(zhuǎn)換器，只有全面了解制造廠的最新產(chǎn)品指南以后才能做到。模數(shù)轉(zhuǎn)換技術(shù)A/D轉(zhuǎn)換技術(shù)的比較對各有關(guān)器件的主要性能做出比較。選用A/D轉(zhuǎn)換器一般要考慮三個主要因素：速度、精度與價格,用這三個因素就便于對不同的轉(zhuǎn)換方法進(jìn)展比較。對于*一轉(zhuǎn)換方法，有著其固有的速度限制；但采用同一種轉(zhuǎn)換方法時，不同的轉(zhuǎn)換期間也有可能有不同的速度圍。各種轉(zhuǎn)換方法在設(shè)計上均存在著速度/精度的權(quán)衡。要提高精度，勢必降低轉(zhuǎn)換速度。一般來說，價格與速度和精度直接相關(guān)，但是轉(zhuǎn)換器件的價格實(shí)際是任何器件的價格，很大程度又受市場因素的影響。而市場因素也可能與器件性能無多大關(guān)系。市場越大，越有許多廠家要去瓜分，競爭就是價格下降。就固有的速度限制而論，假設(shè)以速度減慢轉(zhuǎn)換時間增加的次序?qū)Ω鞣N方法做一排列，應(yīng)為：并行轉(zhuǎn)換、逐次逼近型轉(zhuǎn)換、跟蹤型轉(zhuǎn)換和積分型轉(zhuǎn)換。假設(shè)應(yīng)用中高速轉(zhuǎn)換速度是主要因素，則設(shè)計這就只有選擇采用并行轉(zhuǎn)換方法的器件。到編寫時為止，市場上尚沒有許多超高速的器件，而且他們過于昂貴，尤其是高分辨率的器件。制造高速轉(zhuǎn)換器的一些工廠有：TRWLSIProducts, Datel-Intersil, puter Labs及Motorola。常用的并行轉(zhuǎn)換器8位轉(zhuǎn)換時間可到達(dá)30ns左右。要求