要求每天阅读一篇技术文档，不需要记下来，只是能看懂就好。。后发现，这就是专业英语的课程资料。
----------------------------------------------------------------------------------------------------------------------
At the most basic level, a signal amplifier does exactly what you expect – it makes a signal
bigger! However, the way in which it is done does vary with the design of the actual amplifier,
the type of signal, and the reason why we want to enlarge the signal. [1] We can illustrate this by
considering the common example of a “Hi-Fi” audio system.
In a typical modern Hi-Fi: system, the signals will come from a unit like a CD player, FM
tuner, or a Tape/Minidisk unit. The signals they produce have typical levels of the order of
100 mV or so when the music is moderately loud. This is a reasonably large voltage, easy to
detect with something like an oscilloscope or a voltmeter. However, the actual power levels of
these signals are quite modest. Typically, these sources can only provide currents of a few
milliamps, which by P＝VI means powers of just a few milliwatts. A typical loudspeaker will
require between a few Watts and perhaps over 100 Watts to produce loud sound. Hence we will
require some form of Power Amplifier (PA) to “boost” the signal power level from the source
and make it big enough to play the music.
Fig. 1.1 shows four examples of simple analog amplifier stages using various types of
device. In each case the a.c. voltage gain will usually be approximated by
provided that the actual device has an inherent gain large enough to be controlled by the resistor
values chosen.Note the negative sign in expression 1-1 which indicates that the examples all
invert the signal pattern when amplifying. [2] In practice, gains of the order of up to hundred are
possible from simple circuits like this, although it is usually a good idea to keep the voltage gain
below this. Note also that vacuum state devices tend to be called “valves” in the UK and “tubes”
in the USA.
Many practical amplifiers chain together a series of analog amplifier stages to obtain a high
overall voltage gain. For example, a PA system might start with voltages of the order of 0.1 mV
from microphones, and boost this to perhaps 10 to 100 V to drive loudspeakers. This requires an
overall voltage gain of 109, so a number of voltage gain stages will be required.
In many cases we wish to amplify the current signal level as well as the voltage. The
example we can consider here is the signal required to drive the loudspeakers in a “Hi-Fi” system.
These will tend to have a typical input impedance of the order of 8 Ohms. So to drive, say, 100
Watts into such a loudspeaker load we have to simultaneously provide a voltage of 28 Vrms and
3.5 Arms. Taking the example of a microphone as an initial source again a typical source
impedance will be around 100 Ohms. Hence the microphone will provide just 1 nA when
producing 0.1 mV. This means that to take this and drive 100 W into a loudspeaker the amplifier
system must amplify the signal current by a factor of over 109 at the same time as boosting the
voltage by a similar amount. [3] This means that the overall power gain required is 1018 – i.e. 180
dB!
This high overall power gain is one reason it is common to spread the amplifying function
into separately boxed pre- and power-amplifiers. The signal levels inside power amplifiers are so
much larger than these weak inputs that even the slightest ‘leakage’ from the output back to the
input may cause problems. By putting the high-power (high current) and low power sections in
different boxes we can help protect the input signals from harm.
In practice, many devices which require high currents and powers tend to work on the basis
that it is the signal voltage which determines the level of response, and they then draw the current
they need in order to work. [4] For example, it is the convention with loudspeakers that the
volume of the sound should be set by the voltage applied to the speaker. Despite this, most
loudspeakers have an efficiency (the effectiveness with which electrical power is converted into
acoustical power) which is highly frequency dependent. To a large extent this arises as a natural
consequence of the physical properties of loudspeakers. We won’t worry about the details here,
but as a result a loudspeaker’s input impedance usually varies in quite a complicated manner with
the frequency. (Sometimes also with the input level.)
Fig. 1.2 shows a typical example. In this case, the loudspeaker has an impedance of around
12 Ohms at 150 Hz and 5 Ohms at 1 kHz. So over twice the current will be required to play the
same output level at 1 kHz than is required at 150 Hz. The power amplifier has no way to “know
in advance” what kind of loudspeaker you will use, so simply adopts the convention of asserting
a voltage level to indicate the required signal level at each frequency in the signal and supplying
whatever current the loudspeaker then requires.
This kind of behavior is quite common in electronic systems. It means that, in information
terms, the signal pattern is determined by the way the voltage varies with time, and ideally the
current required is then drawn. Although the above is based on a high-power example, a similar
situation can arise when a sensor is able to generate a voltage in response to an input stimulus but
can only supply a very limited current. In these situations we require either a current amplifier or
a buffer. These devices are quite similar, and in each case we are using some form of gain device
and circuit to increase the signal current level. However, a current amplifier always tries to
multiply the current by a set amount. Hence it is similar in action to a voltage amplifier which
always tries to multiply the signal current by a set amount. The buffer differs from the current
amplifier as it sets out to provide whatever current level is demanded from it in order to maintain
the signal voltage told to assert. Hence it will have a higher current gain when connected to a
more demanding load.