Apego

History

The first electronic amplifier was produced in 1906 by the American inventor Lee De Forest, using the first version of one of its inventions: the Vacuum-tube. In 1908, Lee De Forest improved the vacuum-tube by adding an electrode to him, thus giving rise to the first Triode. The triode was quickly improved by the addition of one (for the Tétrode) then of two additional grids, mitigating certain undesirable effects, in particular the effect “dynatron” (zone where the tube presents a negative Résistance). This tube Pentode is then quickly adopted for the majority of the amplifiers with tubes, for its best output. The tube amplifiers are also known under the name of “lamp” amplifiers, because of the form of the tubes and the light which they emit when they function (see photo opposite).

Since the beginning of the years 1960, grace the appearance of the really reliable first Transistor S of power and at the reduced cost, the majority of the amplifiers uses transistors. One prefers the transistors with the tubes in the majority of the cases because they are less cumbersome, function with tensions weaker and immediately operational are once energized, contrary to the electron tubes which require ten seconds of heating.

The tubes are always used in specific applications like the audio amplifiers, especially those intended for the electric guitars, and the applications of “very” strong power or high frequency as for the microwawe ovens, the industrial heating by radio frequency, and the power gain for the television and radio transmitters.

In the field of space telecommunications asking of strong powers, one also uses amplifiers with Klystron and travelling wave tubes (ATOP). There exist moreover, embarked on satellite board S, of the amplifiers of the type SSPA ( Solid State Power Amplifier ).

Principle of operation and theory

An electronic amplifier uses one or more component credits (Transistor or electron Tube) in order to increase the electric output of the signal present in entry. The active components used in the electronic amplifiers make it possible to control their output current according to an electric quantity (running or tension), image of the signal to be amplified. The output current of the active components is directly drawn from the power supply of the amplifier. According to the way in which they are implemented in the amplifier, the active components thus make it possible to increase the tension and/or the current of the electrical signal of entry. The principle of operation of an amplifier is presented in the simplified diagram present opposite. This diagram uses a bipolar Transistor as amplifying component, but it can be replaced by a MOSFET or a electron Tube. The circuit of polarization ensuring the adjustment of the tension at rest was omitted for reasons of simplification. In this circuit, the current produced by the tension of entry will be amplified of β (with β >> 1) by the transistor. This amplified current then crosses the resistance of exit and one recovers at exit the tension

- \ beta. R.i_e

. With

i_e

the current of entry and

R

the value of resistance.

The amplifiers can be conceived to increase the tension (amplifier of tension), the current (amplifying follower) or both (amplifier of power) of a signal. The electronic amplifiers can be supplied with a simple tension (a food positive or negative, and the zero) or a symmetrical tension (a positive food, negative and the zero). The food can also carry the name of “bus” or “rail”. One speaks then about positive or negative bus and rail of positive or negative tension.

The amplifiers are often composed of several stages laid out in series in order to increase the total profit. Each stage of amplification is generally different from the others so that it corresponds to the specific needs of the stage considered. One can thus draw advantage from the strong points of each assembly while minimizing their weaknesses.

If it is considered that the amplifier supply is independent of the entry signal and exit of the amplifier, one can represent this amplifier by a Quadripôle. The formalism of the quadripoles makes it possible to obtain a matric relation between the currents and the tensions of entry and exit. It was introduced into the years 1920 by the German mathematician Franz Breisig. In the case of an amplifier of tension, the electric quantities are defined by four parameters: impedance of Ze entry, impedance Zs output, profit of Transconductance G and the parameter of G₁₂ reaction. One has then:

${V_1 \ choose V_2} = \ begin {pmatrix} Ze & G_ {12} \ \ G & Zs \ end {pmatrix} {I_1 \ choose I_2}$ .

For a perfect amplifier, G₁₂ is null (the exit does not influence the entry), Zs is also null (the output voltage does not depend on the output current), and the profit G is constant. There is then the profit of the amplifier:

$G= \ frac {v2} {v1} = \ frac {G} {Ze} =cte$ .

In practice these conditions are not completely observed, involving of this fact of the characteristics deteriorated concerning the band-width, the power gain, the noise due to the factor temperature, or the distortion of the signal. One evaluates the performances of an amplifier by studying his Rendement, his Linéarité, his Band-width and the Rapport signal on noise between the entry and the exit.

The “Band-width to -3 dB” (Decibel) of an amplifier is the range of Fréquence S where the profit in tension of the amplifier is higher than the maximum profit minus three decibels. If one does not reason in decibel, that corresponds to the frequency band where the profit in tension higher than the maximum profit is divided by root of two, which corresponds to a division of the power provided to the load by two. The band-width is usually noted B or BP. Occasionally one meets broader band-widths, for example the band-width to -6 dB, frequency band where the profit in tension is higher than half of the maximum profit.

The Linéarité of an amplifier corresponds to its capacity to keep its constant profit whatever the entry. The greatest limitation of linearity comes from the power supply of the amplifier: the tension of entry as that of exit cannot exceed the supply voltage of the amplifier. When that arrives, one speaks about saturation of the amplifier. The linearity of an amplifier is also limited by its Scanning rate (or slew spleen ) which represents the speed of maximum variation that it can reproduce. When that the variation of the entry signal of an amplifier is higher at its scanning rate, its exit is a line of slope $\ mathrm {SR}$ .

$\ mathrm {SR} = \ max \ left (\ frac {dv_s (T)}{dt} \ right)$ .

The scanning rate is expressed in V/µs.

Noise in the electronic amplifiers

In electronics, the noise indicates the random and nondesired signals, even parasitic, being superimposed on the useful signals. In an amplifier these interfering signals can come from its environment or components the component. There exist five types of noise in electronics: the thermal Noise, the Bruit granulates, the Bruit flicker, the Bruit in crenels and the Bruit of avalanche. It is possible to reduce the noise in an amplifier while attacking directly its origins (see below) but also by as much as possible limiting the band-width of the amplifier, in order to eliminate the noise present apart from its work frequencies.

Thermal noise

The thermal Noise, also named noise of resistance , or noise Johnson or noise of Johnson-Nyquist is the noise produced by thermal agitation of the charge carriers, i.e. electron S in a electrical resistance in thermal balance. The thermal noise is a white Bruit of which the spectral concentration of power depends only on the value of resistance. The thermal noise can be modelled by a source of tension in series with the resistance which produces the noise.

The thermal noise was measured for the first time in 1927 by the physicist John Bertrand Johnson with the Bell Labs. Its article Thermal Agitation off Electricity in Conductors showed that statistical fluctuations occurred in all the electric drivers, producing a random variation of potential at the boundaries of this driver. This thermal noise was thus identical for all resistances of the same value and was thus not ascribable with a poor manufacture. Johnson described her observations with her colleague Harry Nyquist who was able to give a theoretical explanation of it.

The noise granulates

The Bruit granulates was highlighted in 1918 by Walter Schottky. This noise appears in the devices or the number of electrons is enough weak to give a detectable statistical fluctuation. In electronics, this noise appears in the devices containing Semi-conducteur (transistors, etc) and the electron tubes. The noise shot is a white Bruit of which the spectral concentration of power depends only on the median value of the current crossing the noisy component.

Note: the thermal noise and the noise shot are both had with quantum fluctutations, and certain formulations make it possible to gather them in only one and single concept.

The noise flicker

The Noise flicker, also named noise in 1/f , noise in excess or pink Bruit is a noise whose Spectral concentration of power is in 1/f. That means that the more the frequency increases, the more the amplitude of this noise decreases. This type of noise exists in all the active components and has very varied origins, like impurities in materials or of creations and parasitic recombinations due to the basic current of a transistor. The noise flicker is always relative to a D.C. current. The noise flicker can be reduced by improving the manufactoring processes of the semiconductors and decreasing the consumption of the amplifier. Unfortunately, the reduction of the consumption of an amplifier passes by an increase in the value of certain resistances what will increase the thermal noise. Noise in crenels, in an audio amplifier, produced “pops” which were worth to him the name of noise popcorn . The appearance of these “pops” is aléatoire : they can appear several times a second then disappearing during several minutes.

The origins of this noise are not currently known, but it seems that they are related on imperfections in the Semi-conducteur S and to the implant of heavy ions. The conditions most favorable to the appearance of this noise seem to be low temperatures and the presence of resistances of strong value.

Signal report/ratio on noise

The signal-to-noise ratio is a term used in engineering, Treatment of the signal or information theory to indicate the relationship between the size of a signal (useful information, significant) and that of the noise (useless information, nonsignificant). As many signals have a raised dynamic scale, the signal-to-noise ratios are often expressed in decibels. The signal report/ratio on noise indicates the quality of a transmission of information compared to the parasites. One thus defines the quality of a Amplificateur, whatever his standard and the category of signals which it treats. The more the report/ratio is raised, the less the apparatus denatures the signal of origin.

Classification of the systems and stages amplifying

There exists a great quantity of classifications, they often rise from the various characteristics of the diagram of an amplifier. All these characteristics have an influence on the parameters and the performances of the amplifier. The design of an amplifier is always a compromise between several factors like the cost, energy consumption, the imperfections of the components and, the need to make the amplifier compatible with the generator of the entry signal and the load at exit. In order to describe an amplifier, one generally speaks about his class, the method of coupling which was used between these various stages as well as the frequency band for which it is planned.

Classification by angle of conduction: classes of amplifiers

Classification by method of coupling

The amplifiers are sometimes classified by their method of coupling between the entry and the exit or the various stages of the amplifier. These various methods include the couplings capacitive, inductive (transformer) and the direct Couplage. The inductive coupling makes it possible to carry out an adaptation of impedance between the stages or to carry out a resonant circuit. It should be noted that the majority of the integrated amplifiers use a direct coupling between their stages. Its preceding work on the reduction of the distortions in the amplifiers had already enabled him to discover the amplifiers “ a priori ” ( feedforward in English) which modify the signal to be amplified in order to compensate for the distortions due to the components of power. Although having remade surface in the years 1970 to compensate for the distortions of the amplifiers BLU, in the years 1920 the practical realization of the amplifiers “ a priori ” proves to be difficult and they do not function very well. In 1927, the Black patent application for the negative feedback was accommodated like a request for invention of Perpetual motion. It was finally accepted nine years later, in December 1931, after Black and other members of the Bell laboratories developed the theory relating to the negative feedback.

An amplifier of looked after design, having all its stages in buckles open (without negative feedback), can arrive at a rate of distortion about the “percent”. Using the negative feedback, a rate of 0,001 % is current. The noise, including the distortions of crossing, can be practically eliminated.

It is the application which dictates the rate of distortion that one can tolerate. For the applications of the amplifying type hi-fi or of instrumentation, the rate of distortion must be minimal, often less 1 %.

Whereas the negative feedback seems to be the remedy for all the evils of an amplifier, much think that it is a bad thing. As it uses a loop, it takes him a time finished to react to an entry signal and for this short period, the amplifier is “out of control”. A musical transient of which the duration is of the same order of magnitude that this period coarsely will thus be distorted. And that, even if the amplifier has a low rate of distortion in permanent mode. It is primarily that which explains the existence of the “transitory distortions of intermodulation” in the amplifiers. This subject was largely discussed at the end of the years 1970 and most of the years 1980 , .

These arguments were sources of controversies during years, and brought to take into account these phenomena when designing amplifier in order to eliminate them . In the facts, the majority of the modern amplifiers use strong negative feedbacks, whereas the diagrams used for the top-of-the-range audio amplifiers seek to minimize it.

Whatever the merits of these arguments on the way in which it modifies the distortion, the negative feedback modifies the impedance of exit of the amplifier and consequently, its ratio damping. While simplifying, the damping ratio characterizes the skill of an amplifier to control an enclosure. If all occurs well, more the negative feedback is strong, more the output impedance is low and more the damping ratio is large. That has an effect on the performances in low frequencies of much of enclosures which returned one of low irregular if the damping ratio of the amplifier is too weak.

The concept of negative feedback is used with the operational amplifier to precisely define the profit, the band-width and much of other parameters.

An example of amplifying assembly

At ends of illustration, one will use this practical example of amplifier. It can be used as a basis for an audio amplifier of moderate power. Its diagram, although appreciably simplified, is typical of what one finds in a modern amplifier thanks to his push-pull of class AB at exit and to the use of a negative feedback. It uses bipolar transistors, but it can just as easily be carried out with field-effect transistors or tubes.

The entry signal is coupled with the base of the Q1 transistor through the condensing of C1 connection. The condenser makes it possible the signal Alternatif to pass, but it blocks the tension continues due to the polarization of Q1 by the dividing bridge R1-R2. Thanks to C1, no preceding circuit is affected by the Biasing of Q1. Q1 and Q2 form a Amplificateur differential (a differential amplifier multiplies by a constant (called profit in tension) the difference between its two entries). The diagram used here to make a differential amplifier is also known under the name of Paire differential. This configuration is used to easily implement the negative feedback, which is provided to Q2 thanks to R7 and R8. The negative feedback in the differential amplifier makes it possible the amplifier to compare the entry with the current exit. The signal amplified by Q1 is sent directly on the second floor, Q3, which amplifies the signal more and provides the power continues polarization of the stage of exit (Q4 and Q5). R6 is used as load in Q3. A more advanced assembly would probably use an active load, a constant power source for example. Until now, the amplifier works in class A. the pair of exit east cabled into push-pull of class AB, also called complementary pair. They provide the majority of the current of the application and directly control the load through the condenser of C2 connection which blocks the component continues. The Diode S D1 and D2 provide a small power continues in order to polarize the pair of exit, so that the distortion of exit is minimized.

This diagram is simple, but it is a good base for the realization of a genuine amplifier because it automatically stabilizes its point of operation thanks to its loop of negative feedback, which functions of continuous until - beyond audio band. A genuine amplifier would probably use an additional circuit cause a drop in the profit beyond the useful waveband in order to avoid the possibility of Oscillation S nondesired. Moreover, the use of fixed diodes for polarization can pose problems if the diodes are not thermically and electrically matched with the transistors of exit. Indeed, if the transistors become too busy, they are likely to be destroyed by thermal runaway. The traditional solution to stabilize the components of exit east to add resistances of one ohm or more in series with the transmitters. The calculation of resistances and the condensers of the circuit is done according to the active components used and of the future use of the amplifier.

Integrated amplifiers

One calls amplifier integrated an amplifier being appeared as a Integrated circuit. A circuit integrated (Ci or electronic chip) is itself a type of component made up of several electronics components in miniaturized form. The integrated circuit makes it possible to reproduce one or more or less complex electronic functions, facilitating its implementation. They contains mainly Transistor S, Diode S, resistances, condensing S, more rarely of the Inductance S because they are able to be miniaturized with more difficulty.

The first integrated circuit was invented by Jack Kilby in 1958, thus providing the foundations of modern data processing. For the little story Jack Kilby, which had just joined the company, made this discovery whereas the majority of his/her colleagues benefitted from holidays organized by Texas Instruments. At the time, Kilby had quite simply connected between them various transistors by cabling them to the hand. It took thereafter only a few months to pass from the prototype stage of to the mass production of silicon chips containing several transistors. This discovery was worth in Kilby a Nobel Prize of Physique in 2000, whereas this last always sat at the directory of Texas Instruments and held more than 60 Brevet S with its name. In practice, the operational amplifiers consist of transistors, electron tubes or of any other amplifying components and they are implemented in discrete circuits or integrated.

The operational amplifiers were initially developed at the era of the electron tubes, they were then used in the analog computers. Currently, the operational amplifiers are available in the form of integrated circuits, although versions in the form of discrete components are used for specific applications.

The first AOP integrated available in great quantity, at the end of the years 1960, was the bipolar AOP Fairchild μA709, creates by Bob Widlar in 1965; it was quickly replaced by the μA741 which offered better performances while being more stable and simpler to implement. The μA741 is still nowadays manufactured, and became omnipresent in electronics. Several manufacturers produce a version improved of this AOP, recognizable thanks to the “741” present in their denomination. Since, more powerful circuits were developed, certain based on JFET (end of the year 1970), or on MOSFET (beginning of the year 1980). The majority of these modern AOP can replace a μA741, in a circuit of old design, in order to improve the performances of them.

The operational amplifiers are available under formats, stitchings, and levels of voltages supply standardized. With some external components, they can carry out a large variety of useful functionalities in Treatment of the signal. The majority of the standard AOP cost only a few tens of centimes of euros, but a AOP discrete or integrated with characteristics not-standard and of low volume of production can cost more than 100 euros part.

The principal manufacturers of operational amplifiers are: Analog Devices, Linear Technology, Maxim, National Semiconductor, STMicroelectronics and Texas Instruments.