"We look at the tiny devices that have reshaped the world of electronics."
Along with the solid-state diode, the point-contact transistor--invented in 1947 at Bell Labs--started the semiconductor
revolution and has gone on the become one of the rudimentary devices in today's electronic equipment. The transistor,
whether in discrete or IC form, is at the heart of most modern circuitry. Therefore, understanding how transistors
function will help you properly design circuits containing them, and in case of a failure, enable you to find and
correct the problem.
Bipolar-Transistor Composition:
A bipolar transistor is basically a two PN junctions connected back-to-back within the same piece of semiconductor
material and sharing a common P- or N-doped semiconductor region. There are two types of bipolar transistor, the NPN
and the PNP. Fig. 1A is a simplified illustration of the composition of the NPN type of
transistor. In our illustration, the NPN type unit is shown as P-doped semiconductor material sandwiched between two
layers of N-doped material. The composition of a PNP transistor is just the opposite of that, (i.e. the N- and P-doped
materials in the transistor are interchanged). It follows then that biasing considerations for NPN units are also
opposite from those for the PNP unit.
Note from Fig. 1A that a bipolar transistor is comprised of a center region called the
base surrounded by two other regions known as the collector and the emitter. The difference between them will be
discussed shortly. The two junctions are arranged so that they are very close together; that's done by making the
shared base region very thin and lightly doped. That causes the two junctions to interact with one another.
Conduction is the collector-base junction depends largely on what happens in the emitter-base junction.
Because the region is lightly doped, it has a relatively small number of free carriers (holes in a P-type base and
electronics in an N-type base) to conduct current. On the other hand, the emitter region is quite heavily doped,
containing a much larger amount of donor impurity (for the NPN type) or acceptor impurity (for the PNP type), so there
are many more free carriers available in the emitter region to conduct current than in the adjacent base region.
Because of that, the emitter-base junction, when forward biased, conducts much the same as a common PN junction diode.
The current that flows (composed of electrons for NPN units and holes, in the case of PNP transistors) is mainly from
the emitter to the base rather than vice versa. That is where the emitter derives its name--it emits or injects
current carriers in the other regions of the device.
The third region of a transistor, the collector, is lightly doped, much the same as the base, except with the opposite
type of doping impurity, so it (like the base region) has relatively few free carriers available to conduct current in
the normal way. The collector-base junction is normally reverse biased, so a depletion layer forms, spreading out on
either side of the junction. The depletion layer effectively removes the carriers that would otherwise balance out
the charges on the fixed impurity atoms of the crystals, setting up a potential barrier to match the applied reverse
voltage.
To the normal majority carriers in the base and emitter, that potential barrier is a big wall that must be overcome
before they can pass to the other side. So just as in the case of a normal diode, virtually no current flows across
the collector-base junction when left to its own devices. However, the junction is not left to its own devices.
Remember that the base region is deliberately made very thin and lightly doped, while the emitter is made much more
heavily doped. Because of that, applying a forward bias to the emitter-base junction causes vast majority carriers
to be injected into th the base, and straight into the reverse-biased collector-base junction. Those carriers are
actually minority carriers in the base region, because that region is of opposite semiconductor type to the emitter.
To those majority-turned-minority carriers, the collector-base junction depletion region is not a barrier at all but
an inviting, accelerating filed; so as soon as they reach the depletion layer, they are immediately swept into the
collector region.
Forward biasing the emitter-base junction causes two things to happen that might seem surprising at first: Only a
relatively small current actually flows between the emitter and the base. much smaller than would flow in a normal PN
diode despite the forward bias applied to the junction between them. A much larger current instead flows directly
between the emitter and the collector regions, in this case, despite the fact that the collector-base junction is
reversed biased.
That effect is illustrated in Fig. 1A, which (hopefully) will help you to understand what
is going on. The diagram shows a NPN transistor, but the action in a PNP unit is similar except for the opposite
region polarity and conduction mainly by holes rather than electrons.
From a practical point of view, the behavior of bipolar transistors means that, unlike the simple PN-junction diode, it
is capable of amplification. In effect, a small input current made to flow between the emitter and collector. Only a
small voltage--around 0.6 volts for a typical silicon transistor--is needed to produce the small input current
required.
In contrast, the reverse-bias voltage applied across the collector-base junction can be much larger; typically
anywhere from 6 to 90 volts or more. So in producing and being able to control a larger current in this much
higher-output circuit, the transistor's small input current and voltage can achieve considerable voltage, power, and
current, gains.
Bipolar transistors, therefore, work very well as both amplifiers and electronics switches. That is why they have
become the workhorses of modern electronics, virtually replacing the vacuum tube. The diagram in Fig. 1A
is designed to show how a bipolar transistor works, rather than its physical construction. The actual form of the
modern, planar, double-defuse epitaxial-junction transistor is shown in Fig. 1A.
The collector region is formed from a lightly doped layer grown epitaxially on the main substrate, which is made from
the same type (but more heavily doped) material to provide a low resistance connection. Here, both are N-doped material;
for a PNP transistor, they would be P-doped material.
The base region is formed by lightly diffusing the opposite type impurity into a medium-sized area of the chip surface
to reverse that type of area and create the base-collector unction. The emitter region is formed by a second and
heavier diffusion over the smaller area inside the first, but this time with the same kind of impurity as used for the
epitaxial collector region.
The second diffusion is very carefully controlled so that the emitter region that results extends almost--but not
quite--to the bottom of the base. That leaves the area of the base right below the emitter quite thin to ensure that
as many as possible of the carriers are injected from the emitter region will be swept through to the collector. The
thinner that active base region, the higher (in general) the gain of the transistor.
Note that although the collector and emitter regions are made of the same type of semiconductor material, the two are
physically quite different. The emitter is heavily doped (for a good carrier injection) and can be relatively small
since the emitter-base junction does not need to dissipate much power (heat). In contrast, the collector is lightly
doped (for a wide depletion area) and its junction is much larger since, being reversed biased, it must dissipate much
more power.
Connections to the emitter and base regions are made by way of aluminum electrodes deposited on the surface. Thin
wires are bonded to the electrodes for connection to the main device leads. The low-resistance substrate itself is
used to connect to the collector region.
That is the basic construction used for most modern bipolar transistors, whether they are discrete units or part of an
IC containing thousands of transistors. The main difference is size, although, in an IC, the collector region of the
transistor will generally be in an epitaxial layer grown on the opposite kind of substrate, and separated by diffused
walls (of the opposite type material) to separate the transistors from each other.
In an IC, the active part of an individual transistor might only be a couple micrometers square, while a very large
transistor (used to switch hundreds of amperes) might be on a single wafer of 10 mm or more in diameter. Typical
small-to-medium power, discrete transistors used in consumer and hobby electronics are grown on chips measuring from
1- to about 3-mm square--the rest of the component is protective packaging.
Transistor Operation:
Refer to Fig. 2, a PNP version of the illustration shown in Fig. 1A.
Note that both are essentially the same, except that in this instance, the collector is more negative than the base or
the emitter. That is an important characteristic to remember when it comes to the operation of bipolar transistors.
If a positive voltage is applied to the P-doped emitter (to the left), current will be swept through the base-emitter
junction--with the holes from the P-doped material moving to the right and the electrons form the N-doped material
moving to the left. Some of the holes moving into the N-doped base region will combine with the electrons and become
neutralized, while others will migrate to the base-collector junction.
Normally, if the base-collector junction is negatively biased, there would be no current flow in the circuit. However,
there would be additional holes in the junction to travel to the base-collector junction, and electrons can then
travel toward the base-emitter junction, so a current flows even through that section of the sandwich is biased (at
cutoff) to prevent conduction. Most of the current travels between the emitter and collector and does not flow out
through the base.
The amplitude of the collector current depends principally on the magnitude of emitter current (e.g., the collector
current). Note that between each PN junction, there is an area known as the depletion or transition region that is
similar in some characteristics to a dielectric layer. That layer varies in accordance with the operating voltage.
The semiconductor materials on either side of the depletion regions constitute the plates of a capacitor. The
base-collector capacitance is indicated in Fig. 2 as Cbc, and the base-emitter
capacitance is designated Cbe. A change in signal and operating voltages causes a non-linear change in
those junction capacitances.
There is also a base-emitter resistance (Rbe that must be considered. In practical transistors, emitter
resistance is on the order of a few ohms, while the collector resistance is many hundreds or even thousands of times
larger. The junction capacitance in combination with the base-emitter resistance determine the useful upper-frequency
limit of a transistor by establishing an RC time constant.
Because the collector is reversed biased, the collector-to-base resistance is high. On the other hand, the emitter and
collector currents are substantially equal, so the power in the collector circuit is larger than the power in the
emitter circuit.
(P = I2R, so the powers are proportional to the respective resistances, if the currents are the same.)
In practical transistors, emitter resistance is on the order of a few ohms, while the collector resistance is many
hundreds or thousands of times larger, so power gains of 20 to 40dB, or even more, are possible.
Figure 3 shows the schematic symbols for both the NPN and PNP version of the bipolar
transistor. The first two letters of the designators (NPN or PNP) indicate the polarities of the voltages applied to
the collector and emitter in normal operation. For example, in a PNP unit, the emitter is made more positive with
respect to the collector and the base, and the collector is made more negative with respect to the base. Another way
of saying that is: the collector is more negative than the base and the base is more negative than the emitter.
Transistor Amplifiers:
Transistors are among the most commonly used building blocks in electronics. While they can be used as electronically
controlled switches, they are widely configured for amplifier use. In fact, the vast majority of electronic circuits
contain one or more amplifiers of some type or another.
However, what exactly do we mean by the term amplifier? By definition an amplifier is a circuit that draws power from
a source other than the input signal and produces an output that is usually an enlarged reproduction of the input signal.
We say usually because not all amplifiers are used to magnify the input signal--buffer amplifiers (often called
unity-gain amplifiers) are not designed to magnify the input signal. When operated as a buffer, the transistor is
used to isolate one stage from the effects of one that follows. Since buffer amplifiers provide no increase in signal
level, a 10-millivolt (mV) signal applied to the input of a unity-gain amplifier produces an output signal at the same
10-mV level (a carbon copy of the input signal).
There are may types of amplifiers, however, and all fall into one of two broad categories: voltage amplifiers or
current (often referred to as a power) amplifiers. The term voltage amplifier implies to a circuit in which a low
voltage is applied to the input to produce a higher voltage at the output. The term power amplifier is generally
reserved for those that supply an appreciable power (or current) increase to the load.
Because a vast array of amplifier circuits in use in modern electronics, amplifier circuits are often subdivided by
application--AF, IF, RF, Instrumentation, op-amp, etc. Another way of categorizing amplifiers is by configuration:
common-emitter, common-collector, and common-base for example. The important parameters in such circuits are the
cutoff frequency and the input/output impedances. The cut-off frequency at which the gain of an amplifier falls below
0.707 times the maximum gain of the circuit. The input impedance is the output impedance of the transistor.
Amplifier Configurations:
An example of a common-base amplifier is shown in Fig. 4A. The optimum load impedance can
range from a few thousand ohm to 100,000 ohms, depending on the circuit's requirements. In this type of circuit, the
output signal (at the collector) is in phase with the input signal (applied at the emitter). THe current that flows
through the base resistance of the transistor is therefore in phase as well, so the circuit tends to be regenerative
and will oscillate if the current-amplification factor is greater than one.
A common-emitter (also called a "grounded-emitter") amplifier is shown in Fig. 4B. Base
current in this amplifier configuration small and the input impedance is therefore fairly high (several thousand ohms
on the average). Collector resistance on the other hand, can be tens of thousands of ohms, depending on the signal's
source impedance. The common-emitter amplifier has a lower cutoff frequency than does the common-base type, but gives
the highest power gain of the three configurations. Note that the output signal is 180° out-of-phase with (or the
opposite of) the input (base-current) signal, so the feedback that flows through the small emitter resistance is
negative (degenerative), keeping the circuit stable. The common-emitter amplifier is one of the most often seen
configurations for the bipolar transistor.
The common-collector amplifier (also referred to as an emitter follower), see Fig. 4C, has
a high input impedance and a low output impedance.
The impedance is approximately:
The fact that the input resistance is directly related to the load resistance is a disadvantage of this type of amplifier
if the load is one whose resistance or impedance varies with frequency.
The current transfer ratio of this type of circuit is:
and the cutoff frequency is the same as in the common-emitter amplifier circuit. The output and input currents of
this type of circuit are in phase.
Amplifier Classifications:
Amplifiers may be otherwise classified by their specific operational characteristics, in particular, the bias voltages
between the emitter-base and base-collector junctions. The relationship between the bias voltage and the cutoff
voltage of an amplifier is what classifies an amplifier as being class A, B, C, or AB. Each class has a specific
characteristic that makes it most suitable for a particular application.
In a class-A amplifier--which is the least efficient, but offers the least distortion--the transistor is biased so that
its quiescent operation point is in the middle of the power-supply extremes, i.e., the transistor is always turned
on and the resulting output varies around the bias voltage; see the output waveform in Fig. 5A.
Because of that, the input signal must be small enough so that its positive and negative swings do not drive the
amplifier near the non-linear cutoff and saturation regions.
Since a high-value resistor is used to change the output voltage to a current (I=V/R) in a class-A configuration,
the output current is small. That is important since current flows at all times in such amplifiers, with or without
an input signal. Power is wasted and efficiency (the ratio of output to total power consumed) is low--only about
20-25%--in call-A amplifiers. Class-A amplifiers can be configured for single-ended or push-pull operation and are
used in AF (audio frequency), IF (intermediate frequency), and RF (radio frequency) applications.
Class-A operation is suitable for voltage amplifiers. In a voltage amplifier, the emphasis is on the magnitude of the
output voltage. Figure 6 shows a single-ended class-A amplifier. Such an amplifier might
be used in a preamplifier stage, where input signals are typically small, and a faithful reproduction of the input
using a single transistor is needed. That configuration allows a small input current to control current drawn from a
power source, and thus produce a stronger replica of a weaker original signal.
In Class-B operation, the transistor is biased at cutoff (see Fig. 5B), so that output current flows during only half of
the input cycle. It is used where high efficiency and low distortion are required--for instance, in power-output
configurations. When the Class-B amplifier is used for audio applications, two such amplifiers connected in the
push-pull configuration are required, so that current can flow alternately through the two amplifiers. In other words,
on amplifier is turned on, while the other is turned off.
On the other hand, when the Class-B amplifier is used in RF applications, it can be configured for single-ended
operation. Since, in the absence of an input signal its current output is negligible, it is used where high efficiency
(60-70%) and low distortion are required, which is very important in high-power amplifiers.
Class-AB amplifiers (see Fig. 5C) are biased somewhere between Class-A and Class-B operation, and have efficiencies
(25-35%) and distortion characteristics that lie between those of Class-A and Class-B amplifiers. Class-AB amplifiers
require a somewhat larger input signal than do Class-A amplifiers. The class-AB amplifier is used in push-pull
configurations for both audio-and radio-frequency applications.
In Class-C operation--which has the highest efficiency (perhaps more than 90%), but offers the greatest
distortion--the transistor is biased beyond the cutoff region (see Fig. 5D). Because of that, output output current
flows during less than half (about a third) of the input cycle, making it unsuitable for amplifying signals of varying
amplitude, such as audio. That type of amplifier is normally used to amplify a signal of fixed amplitude; for
instance, it is often used in RF power output stages of a transmitter. Current in a Class-C amplifier flows in a
series of power pulses that excite an LC-tank circuit into oscillation. Because of that the output waveform is a
sinewave, that varies in amplitude if modulated. Class-C amplifiers can be configured for push-pull or single-ended
operation. Table 1 summarizes the conduction angles and efficiency ratings of the various classes of transistor
amplifier.
IC containing thousands of transistors. The main difference is size, although, in an IC, the collector region of the
transistor will generally be in an epitaxial layer grown on the opposite kind of substrate, and separated by diffused
walls (of the opposite type material) to separate the transistors from each other.
Figure 3 shows the schematic symbols for both the NPN and PNP version of the bipolar
transistor. The first two letters of the designators (NPN or PNP) indicate the polarities of the voltages applied to
the collector and emitter in normal operation. For example, in a PNP unit, the emitter is made more positive with
respect to the collector and the base, and the collector is made more negative with respect to the base. Another way
of saying that is: the collector is more negative than the base and the base is more negative than the emitter.
We say usually because not all amplifiers are used to magnify the input signal--buffer amplifiers (often called
unity-gain amplifiers) are not designed to magnify the input signal. When operated as a buffer, the transistor is
used to isolate one stage from the effects of one that follows. Since buffer amplifiers provide no increase in signal
level, a 10-millivolt (mV) signal applied to the input of a unity-gain amplifier produces an output signal at the same
10-mV level (a carbon copy of the input signal).
Class-A operation is suitable for voltage amplifiers. In a voltage amplifier, the emphasis is on the magnitude of the
output voltage. Figure 6 shows a single-ended class-A amplifier. Such an amplifier might
be used in a preamplifier stage, where input signals are typically small, and a faithful reproduction of the input
using a single transistor is needed. That configuration allows a small input current to control current drawn from a
power source, and thus produce a stronger replica of a weaker original signal.