自动化专业外文翻译--运算放大器.doc

UNIT 2 A: The Operational Amplifier One problem with electronic devices corresponding to the generalized amplifiers is that the gains, Au or A, depend upon internal properties of the two-port system (p, fl, R, Ro, etc.)? This makes design difficult since these parameters usually vary from device to device, as well as with temperature. The operational amplifier, or Op-Amp, is designed to minimize this dependence and to maximize the ease of design. An Op-Amp is an integrated circuit that has many component part such as resistors and transistors built into the device. At this point we will make no attempt to describe these inner workings. A totally general analysis of the Op-Amp is beyond the scope of some texts. We will instead study one example in detail, then present the two Op-Amp laws and show how they can be used for analysis in many practical circuit applications. These two principles allow one to design many circuits without a detailed understanding of the device physics. Hence, Op-Amps are quite useful for researchers in a variety of technical fields who need to build simple amplifiers but do not want to design at the transistor level. In the texts of electrical circuits and electronics they will also show how to build simple filter circuits using Op-Amps. The transistor amplifiers, which are the building blocks from which Op-Amp integrated circuits are constructed, will be discussed. The symbol used for an ideal Op-Amp is shown in Fig. 1-2A-1. Only three connections are shown: the positive and negative inputs, and the output. Not shown are other connections necessary to run the Op-Amp such as its attachments to power supplies and to ground potential. The latter connections are necessary to use the Op-Amp in a practical circuit but are not necessary when considering the ideal 0p-Amp applications we study in this chapter. The voltages at the two inputs and the output will be represented by the symbols U+, U-, and Uo. Each is measured with respect t ground potential. Operational amplifiers are differential devices. By this we mean that the output voltage with respect to ground is given by the expression Uo =A(U+ -U-) (1-2A-l) where A is the gain of the Op-Amp and U+ and U - the voltages at inputs. In other words, the output voltage is A times the difference in potential between the two inputs. Integrated circuit technology allows construction of many amplifier circuits on a single composite chip of semiconductor material. One key to the success of an operational amplifier is the cascading of a number of transistor amplifiers to create a very large total gain. That is, the number A in Eq. (1-2A-1) can be on the order of 100,000 or more. (For example, cascading of five transistor amplifiers, each with a gain of 10, would yield this value for A.) A second important factor is that these circuits can be built in such a way that the current flow into each of the inputs is very small. A third important design feature is that the output resistance of the operational amplifier (Ro) is very small. This in turn means that the output of the device acts like an ideal voltage source. We now can analyze the particular amplifier circuit given in Fig. 1-2A-2 using these characteristics. First, we note that the voltage at the positive input, U +, is equal to the source voltage, U + = Us. Various currents are defined in part b of the figure. Applying KVL around the outer loop in Fig. 1-2A-2b and remembering that the output voltage, Uo, is measured with respect to ground, we have -I1R1-I2R2+U0=0 (1-2A-2) Since the Op-Amp is constructed in such a way that no current flows into either the positive or negative input, I- =0. KCL at the negative input terminal then yields I1 = I2 Using Eq. (1-2A-2) and setting I1 =I2 =I, U0=(R1+R2)I (1-2A-3) We may use Ohms law to find the voltage at the negative input, U-, noting the assumed current direction and the fact that ground potential is zero volts: (U-0)/ R1=ISo, U-=IR1and from Eq. (1-2A-3), U- =R1/(R1+R2) U0Since we now have expressions for U+ and U-, Eq. (1-2A-l) may be used to calculate the output voltage,U0 = A（U+-U-）=AUS-R1U0/(R1+R2)Gathering terms, U0 =1+AR1/(R1+R2)= AUS (1-2A-4)and finally, AU = U0/US= A(R1+R2)/( R1+R2+AR1) (1-2A-5a)This is the gain factor for the circuit. If A is a very large number, large enough that AR (R1+R2),the denominator of this fraction is dominated by the AR term. The factor A, which is in both the numerator and denominator, then cancels out and the gain is given by the expression AU =(R1+R2)/ R1 (1-2A-5b)This shows that if A is very large, then the gain of the circuit is independent of the exact value of A and can be controlled by the choice of R1and R2. This is one of the key features of Op-Amp design the action of the circuit on signals depends only upon the external elements which can beeasily varied by the designer and which do not depend upon the detailed character of the Op-Amp itself. Note that if A=100 000 and (R1 +R2)/R1=10, the price we have paid for this advantage is that we have used a device with a voltage gain of 100 000 to produce an amplifier with a gain of 10. In some sense, by using an Op-Amp we trade off power for control. A similar mathematical analysis can be made on any Op-Amp circuit, but this is cumbersome and there are some very useful shortcuts that involve application of the two laws of Op-Amps which we now present. 1) The first law states that in normal Op-Amp circuits we may assume that the voltage difference between the input terminals is zero, that is, U+ =U- 2) The second law states that in normal Op-Amp circuits both of the input currents may be assumed to be zero: I+ =I- =0 The first law is due to the large value of the intrinsic gain A. For example, if the output of an Op- Amp is IV and A= 100 000, then ( U+ - U- )= 10-SV. This is such a small number that it can often be ignored, and we set U+ = U-. The second law comes from the construction of the circuitry inside the Op-Amp which is such that almost no current flows into either of the two inputs. B: Transistors Put very simply a semiconductor material is one which can be doped to produce a predominance of electrons or mobile negative charges (N-type); or holes or positive charges (P- type). A single crystal of germanium or silicon treated with both N-type dope and P-type dope forms a semiconductor diode, with the working characteristics described. Transistors are formed in a similar way but like two diodes back-to-back with a common middle layer doped in the opposite way to the two end layers, thus the middle layer is much thinner than the two end layers or zones. Two configurations are obviously possible, PNP or NPN (Fig. 1-2B-l). These descriptions are used to describe the two basic types of transistors. Because a transistor contains elements with two different polarities (i.e., P and N zones), it is referred to as a bipolar device, or bipolar transistor. A transistor thus has three elements with three leads connecting to these elements. To operate in a working circuit it is connected with two external voltage or polarities. One external voltage is working effectively as a diode. A transistor will, in fact, work as a diode by using just this connection and forgetting about the top half. An example is the substitution of a transistor for a diode as the detector in a simple radio. It will work just as well as a diode as it is working as a diode in this case. The diode circuit can be given forward or reverse bias. Connected with forward bias, as in Fig.l-2B-2, drawn for a PNP transistor, current will flow from P to the bottom N. If a second voltage is applied to the top and bottom sections of the transistor, with the same polarity applied to the bottom, the electrons already flowing through the bottom N section will promote a flow of current through the transistor bottom-to-top. By controlling the degree of doping in the different layers of the transistor during manufacture, this ability to conduct current through the second circuit through a resistor can be very marked. Effectively, when the bottom half is forward biased, the bottom section acts as a generous source of free electrons (and because it emits electrons it is called the emitter). These are collected readily by the top half, which is consequently called the collector, but the actual amount of current which flows through this particular circuit is controlled by the bias applied at the center layer, which is called the base. Effectively, therefore, there are two separate working circuits when a transistor is working with correctly connected polarities (Fig. 1-2B-3). One is the loop formed by the bias voltage supply encompassing the emitter and base. This is called the base circuit or input circuit. The second is the circuit formed by the collector voltage supply and all three elements of the transistor. This is called the collector circuit or output circuit. (Note: this description applies only when the emitter connection is common to both circuits known as common emitter configuration.) This is the most widely used way of connecting transistors, but there are, of course, two other alternative configurations - common base and common emitter. But, the same principles apply in the working of the transistor in each case. The particular advantage offered by this circuit is that a relatively small base current can control and instigate a very much larger collector current (or, more correctly, a small input power is capable of producing a much larger output power). In other words, the transistor works as an amplifier. With this mode of working the base-emitter circuit is the input side; and the emitter through base to collector circuit the output side. Although these have a common path through base and emitter, the two circuits are effectively separated by the fact that as far as polarity of the base circuit is concerned, the base and upper half of the transistor are connected as a reverse biased diode. Hence there is no current flow from the base circuit into the collector circuit. For the circuit to work, of course, polarities of both the base and collector circuits have to be correct (forward bias applied to the base circuit, and the collector supply connected so that the polarity of the common element (the emitter) is the same from both voltage sources). This also means that the polarity of the voltages must be correct for the type of transistor. In the case of a PNP transistor as described, the emitter voltage must be positive. It follows that both the base and collector are negatively connected with respect to the emitter. The symbol for a PNP transistor has an arrow on the emitter indicating the direction of current flow, always towards the base. (P for positive, with a PNP transistor).In the case of an NPN transistor, exactly the same working principles apply but the polarities of both supplies are reversed (Fig. 1-2B-4). That is to say, the emitter is always made negative relative to base and collector (N for negative in the caseof an NPN transistor). This is also inferred by the reverse direction of the arrow on the emitter in the symbol for an NPN transistor, i.e., current flow away from the base. While transistors are made in thousands of different types, the number of shapes in which they are produced is more limited and more or less standardized in a simple code - TO (Transistor Outline) followed by a number. TO1 is the original transistor shape a cylindrical can with the three leads emerging in triangular pattern from the bottom. Looking at the base, the upper lead in the triangle is the base, the one to the fight (marked by a color spot) the collector and the one to the left the emitter.2 The collector lead may also be more widely spaced from the base lead than the emitter lead. In other TO shapes the three leads may emerge in similar triangular pattern (but not necessarily with the same positions for base, collector and emitter), or in-line. Just to confuse the issue there are also sub-types of the same TO number shape with different lead designations. The TO92, for example, has three leads emerging in line parallel to a flat side on an otherwise circularcan reading 1,2,3 from top to bottom with the flat side to the right looking at the base. With TO92 sub-type a (TO92a): 1=emitter 2=collector 3=base With TO92 sub-type b (TO92b): 1=emitter 2=base 3=collector To complicate things further, some transistors may have only two emerging leads (the third being connected to the case internally); and some transistor outline shapes are found with more than three leads emerging from the base. These, in fact, are integrated circuits (ICs), packaged in the same outline shape as a transistor. More complex ICs are packaged in quite different form, e.g., flat packages. Power transistors are easily identified by shape They are metal cased with an elongated bottom with two mounting holes. There will only be two leads (the emitter and base) and these will normally be marked. The collector is connected internally to the can, and so connection to the collector is via one of the mounting bolts or bottom of the can.A运算放大器对应于像广义放大器这样的电子装置，存在的一个问题就是它们的增益AU或AI,它们取决于双端口系统(、Ri、R0等)的内部特性。器件之间参数的分散性和温度漂移给设计工作增加了难度。设计运算放大器或Op-Amp的目的就是使它尽可能的减少对其内部参数的依赖性、最大程度地简化设计工作。运算放大器是一个集成电路，在它内部有许多电阻、晶体管等元件。就此而言，我们不再描述这些元件的内部工作原理。运算放大器的全面综合分析超越了某些教科书的范围。在这里我们将详细研究一个例子，然后给出两个运算放大器定律并说明在许多实用电路中怎样使用这两个定律来进行分析。这两个定律可允许一个人在没有详细了解运算放大器物理特性的情况下设计各种电路。因此，运算放大器对于在不同技术领域中需要使用简单放大器而不是在晶体管级做设计的研究人员来说是非常有用的。在电路和电子学教科书中，也说明了如何用运算放大器建立简单的滤波电路。作为构建运算放大器集成电路的积木晶体管，将在下篇课文中进行讨论。理想运算放大器的符号如图1-2A-1所示。图中只给出三个管脚：正输入、负输入和输出。让运算放大器正常运行所必需的其它一些管脚，诸如电源管脚、接零管脚等并未画出。在实际电路中使用运算放大器时，后者是必要的，但在本文中讨论理想的运算放大器的应用时则不必考虑后者。两个输入电压和输出电压用符号U+、U-和U0 表示。每一个电压均指的是相对于接零管脚的电位。运算放大器是差分装置。差分的意思是：相对于接零管脚的输出电压可由下式表示U0=A（U+-U-） (1-2A-1)式中 A 是运算放大器的增益，U+ 和U-是输入电压。换句话说，输出电压是A乘以两输入间的电位差。 集成电路技术使得在非常小的一块半导体材料的复合 “芯片”上可以安装许多放大器电路。运算放大器成功的一个关键就是许多晶体管放大器“串联”以产生非常大的整体增益。也就是说，等式(1-2A-1)中的数A约为100,000或更多 (例如，五个晶体管放大器串联，每一个的增益为10，那么将会得到此数值的A)。 第二个重要因素是这些电路是按照流入每一个输入的电流都很小这样的原则来设计制作的。第三个重要的设计特点就是运算放大器的输出阻抗(R0)非常小。也就是说运算放大器的输出是一个理想的电压源。我们现在利用这些特性就可以分析图1-2A-2所示的特殊放大器电路了。首先，注意到在正极输入的电压U +等于电源电压，即U+ =US。各个电流定义如图1-2A-2中的b图所示。对图 1-2A-2b的外回路应用基尔霍夫定律，注意输出电压U0 指的是它与接零管脚之间的电位，我们就可得到-I1R1-I2R2+U0=0 (1-2A-2)因为运算放大器是按照没有电流流入正输入端和负输入端的原则制作的，即I- =0。那么对负输入端利用基尔霍夫定律可得I1 = I2，利用等式(1-2A-2) ，并设 I1 =I2 =I，U0 = (R1 +R2) I (1-2A-3)根据电流参考方向和接零管脚电位为零伏特的事实，利用欧姆定律，可得负极输入电压U-： (U-0)/ R1=I因此U-=IR1 ，并由式 (1-2A-3)可得： U- =R1/(R1+R2) U0 因为现在已有了U+ 和U-的表达式，所以式(1-2A-1)可用于计算输出电压，U0 = A（U+-U-）=AUS-R1U0/(R1+R2)综合上述等式，可得： U0 =1+AR1/(R1+R2)= AUS (1-2A-4) 最后可得： AU = U0/US= A(R1+R2)/( R1+R2+AR1) (1-2A-5a)这是电路的增益系数。如果A 是一个非常大的数，大到足够使AR1 (R1 +R2)，那么分式的分母主要由AR1 项决定，存在于分子和分母的系数A 就可对消，增益可用下式表示这表明， AU =(R1+R2)/ R1 (1-2A-5b)如果A 非常大，那么电路的增益与A 的精确值无关并能够通过R1和R2的选择来控制。这是运算放大器设计的重要特征之-在信号作用下，电路的动作仅取决于能够容易被设计者改变的外部元件，而不取决于运算放大器本身的细节特性。注意，如果A=100,000， 而(R1 +R2) /R1 =10，那么为此优点而付出的代价是用一个具有100,000倍电压增益的器件产生一个具有10倍增益的放大器。从某种意义上说，使用运算放大器是以“能量”为代价来换取“控制”。对各种运算放大器电路都可作类似的数学分析，但是这比较麻烦，并且存在一些非常有用的捷径，其涉及目前我们提出的运算放大器两个定律应用。1) 第一个定律指出：在一般运算放大器电路中，可以假设输入 端间的电压为零，也就是说，U+ =U-2) 第二个定律指出：在一般运算放大器电路中，两个输入电流可被假定为零： I+ =I- =0 第一个定律是因为内在增益A的值很大。例，如果运算放大器的输出是1V，并且A=100,000, 那么(U+ = U-)=10-5 V这是一个非常小、可以忽略的数，因此可设U+ = U-。第二个定律来自于运算放大器的内部电路结构，此结构使得基本上没有电流流入任何一个输入端。B晶体管 简单地说，半导体是这样一种物质，它能够通过“掺杂”来产生多余的电子，又称自由电子（N型）；或者产生“空穴”，又称正电荷（P型）。由N型掺杂和P型掺杂处理的锗或硅的单晶体可形成半导体二极管，它具有我们描述过的工作特性。晶体管以类似的方式形成，就象带有公共中间层、背靠背的两个二极管，公共中间层是以对等的方式向两个边缘层渗入而得，因此中间层比两个边缘层或边缘区要薄的多。PNP 或 NPN (图 1-2B-1)这两种结构显然是可能的。PNP或NPN被用于描述晶体管的两个基本类型。因为晶体管包含两个不同极性的区域（例如“P”区和“N”区），所以晶体管被叫作双向器件，或双向晶体管。一个晶体管有三个区域，并从这三个区域引出三个管脚。要使工作电路运行，晶体管需与两个外部电压或极性连接。其中一个外部电压工作方式类似于二极管。事实上，保留这个外部电压并去掉上半部分，晶体管将会象二极管一样工作。例如在简易收音机中用晶体管代替二极管作为检波器。在这种情况下，其所起的作用和二极管所起的作用一模一样。 可以给二极管电路加正向偏置电压或反向偏置电压。