A semiconductor is a material that is neither a conductor nor an insulator - it is somewhere in between. This sounds like a resistor, and indeed it is possible to make resistors from semiconductor material. However, in electronics, semiconductor has a more specific meaning.
If we take a crystal of some pure material, such as silicon, we find that each silicon atom has four bonds (this is called the valency of the atom), and each bond links it to another silicon atom. This bonding forms a crystal lattice and occupies all of the electrons in the silicon atoms. Because there are no free electrons, this material cannot pass a current, and is an insulator.
If we carefully add an additional element to the material, the situation changes. If we add an atom with a valency of five instead of four, we find that wherever an atom of the introduced element is found, four of its bonds will attach to nearby silicon atoms, but the fifth bond will be unattached, leaving a free electron available, which will carry a current. Such a material is said to be doped with the donor material, but because only a small amount of donor material is added, the number of free electrons is quite small and so it will conduct, but rather poorly.
N-type and P-type
If we dope silicon with material having a valency of five, extra free electrons are created in the crystal. These electrons carry charge in the normal manner, and are called majority carriers. Material doped in this way is referred to as N-type or "negative type" semiconductor (because the charge is carried by electrons, and such charge is actually negative with respect to conventional current).
We can also dope silicon with a donor having a valency of only three instead of four. This also creates a semiconductor material because the shortage of bonds leaves holes in the lattice. These holes are the absence of an electron, but they still carry a charge, equivalent to the charge on an electron, but of the opposite sign. These are called minority carriers. Holes carry charge in the opposite direction to normal current, because their charge has the opposite sign. In fact what happens is that current is carried by electrons as usual, but wherever an electron meets a hole, it falls into the hole, opening up a new hole behind it. So while the current is still really caused by electrons, it appears as if the holes are migrating in the opposite direction. This is exactly equivalent to a positive charge moving the opposite way.
P and N type materials are in themselves not very interesting, but remarkable properties emerge when a sandwich is formed from the two kinds of materials. Such sandwiches are called junctions and are the basis for all kinds of semiconductor devices.
A diode allows current to flow one way but not another. It is important to remember that when looking at a circuit diagram that conventional current flows in the direction of the arrow (positive to negative). The word "di-ode" means two-terminaled.
A diode is formed by bonding a single piece of N-type semiconductor to a single piece of P-type. This two-layer device consists of a single P-N junction, and is the simplest semiconductor device. When current is flowing through the device, it is said to be forward biased and has a fairly low resistance. If the voltage across the diode is reversed, the diode is said to be reverse biased and usually no current will flow; but see also Zener diodes.
How it works is as follows. When reverse biased, free electrons in the N-type material, being negative, are attracted to the positive potential applied to the diode. At the other side, positively charged holes are attracted to the negative terminal. The carriers (holes and electrons) are pulled away from the junction area, leaving it depleted of any charge carriers. This forms a depletion layer and it is an insulator, so no current can flow through the device, except in Zener diodes.
When the potential difference across the diode is reversed, the opposite happens - carriers are pushed away from the terminals towards the junction, until they reach the opposite type of material. Here, electrons are able to "fall into" holes, completing the circuit and allowing current to flow.
Diodes are used also to make DC out of AC. They are then called rectifiers.
Zener diodes are used as voltage sources.
A light-emitting diode or LED is a 2-layer semiconductor device just as any typical diode is, but it is doped using exotic materials to give it its light-emitting properties. When reverse-biased, an LED blocks the current just as any diode does, but when forward biased, each time an electron falls into a hole, it gives up its energy as a number of photons, which we see as visible light.
The exact amount of energy given up dictates the colour of the LED, and this in turn is controlled by the exact proportions and types of doping material used. Low energy photons are seen as red light, while higher energies gives us orange, green and blue. White LEDs are made by producing a wide range of colours simultaneously.
Most LEDs need to have a carefully limited amount of current, since they are easy to burn out if too much flows. Thus a series resistor is necessary to make sure the current can't be too high. As long as the current is kept within the LED's design limits, they are extremely robust and have very long lifetimes - far greater than ordinary lightbulbs.
A simple experiment:
Take the a LED and the appropriate resistor and battery to make it work and simply switch the two connections of the diode to the resistor and battery. If done correctly it should light in one direction and not the other.
A transistor is a semiconductor device with three terminals. There are numerous types of transistors, but in general they all do the same job - they control a current under the influence of another, usually much smaller current or voltage. The most basic type of transistor is the field effect transistor (FET), and as these are what is used in most kinds of integrated circuit, they are also the most common.
An FET consists of a rod of doped semiconductor material (say, N-type) which is surrounded by a metal plate. The metal plate is insulated from the semiconductor and there is no P-N junction here. The metal plate is connected to the control terminal of the device or gate. The two ends of the rod (called the channel) are connected to the source and the drain. A circuit is so arranged so that a current flows through the device from source to drain. Because the material is semiconducting, some current will flow even if the gate terminal is not connected.
If we connect up the gate so that we can change the voltage on it, the metal plate can accumulate charge, just like a capacitor. The charges on this plate can attract or repel the carriers in the channel. If carriers are repelled away from the gate, the channel increases in resistance, because fewer carriers can get through in the region of the repulsion. By changing the design of the gate, it will attract carriers, making available more carriers to carry the current, which will thus increase. So by changing the voltage on the gate, the resistance or conductance of the channel can be varied at will.
When a current is changed by the influence of a voltage, this is called transconductance. The word "transistor" was originally coined to mean a "transconducting resistor".
The type of FET we have just described is sometimes called an "insulated gate" FET or IGFET. It is the simplest kind. There are others, such as the Metal-Oxide Semiconductor FET (MOSFET), or the Junction FET or JFET, but all operate on similar principles.
Historically, the FET was not the first kind of transistor that was invented, though the FET as described above works in a very similar fashion to the much older thermionic valve or tube. When transistors were first invented, the type that was made is known as a bipolar junction transistor. This is the kind usually meant by the unqualified term "transistor". Junction transistors are more complicated than FETs however, which is why we explained the FET first.
- See also Semiconductors
Bipolar Junction Transistors
A Bipolar Junction Transistor or BJT consists of a three-layer sandwich of semiconductor material, arranged N-P-N or P-N-P. The central section is usually much narrower than the other two and is connected to a terminal called the base. The outer areas are connected to terminals called the collector and the emitter.
At first glance a BJT looks a bit like two diodes placed back-to-back, and indeed if one tests a BJT using a multimeter it may be treated as one when testing between the base and either of the other two terminals. However, its physical behaviour is not the same as two diodes back-to-back, because of the very close physical proximity of the two junctions. This is an important point to note to understand the working of a BJT.
In a circuit, a current flows from the collector to the emitter through the device. If the base terminal is not connected, in fact no current will flow because at least one of the internal junctions is reverse biased, so just like the diode, carriers are repelled away from the junction join to form an insulating depletion layer. Unlike the diode, this depletion layer is not altered by reversing the voltage across the device, but by injecting extra carriers via the base terminal.
If a small current is arranged to flow into the base terminal, it adds carriers to the depletion layer, making it narrower. This reduces its resistance and allows more current to flow across the junction from collector to emitter. This current can be very much greater than the current flowing the in the base, yet is proportional to it, thus the base current acts as a control for the collector-emitter current.
The ratio between the base current and the collector current is called the gain of the device (also called its beta), and could be 100 times or even more. Thus in one sense a transistor amplifies the signal - the much larger current changes in sympathy with the smaller base current. However this relationship is not linear - overall it follows a square law - the collector current varies in proportion to the square of the base current. However over limited ranges the output is more or less linear, and usually for amplification purposes a transistor is operated on this part of the curve. The fact that it is a curve is where distortion arises when amplifying music or other small signals.
See also Electronics/RAM and ROM