The world has been a very different place since the invention of the humble transistor. There are many pages where you can read about the invention and history but I’m not prepared to give a history lesson. A transistor has two uses: amplification and switching. They also come in many varieties, sizes and flavours (Bipolars, FETs, MOSFETs). A transistor is made from semi-conductive materials such as germanium and silicon.
Out of all the elements in the periodic table, why are germanium and silicon semi-conductive? The answer lies in the chemistry and physics (electronics is applied physics!). The two elements I have mentioned belong in Group IV of the periodic table. This means that they have a valency of 4, meaning they can form 4 bonds due to a half-filled outer electron shell. Their outer electron shells can only hold 8 electrons. So, in order for them to be “happy”,they want that gap filled. This gap can be filled in the process called doping. This is where we introduce either an N or P type element. An N type is the result of doping the semiconductor with a Group V element, whereas a P type is doping with a Group III element. The effect of this is that the N type has additional negative carriers (electrons) making it conductive whereas P type has holes (a gap in the electron shell) which we also treat as particles, oddly enough, which are positive carriers.
If we were to join a section of P type with a section of N type, we would get a diode (see badly drawn figure 3). Figure 1 is a typical diode; Figure 2 is the electrical symbol for completeness’ sake. I’m talking loosely here— it is not as simple as taking a lump of both materials and slapping them together. A diode is a passive component where electrical current is allowed to flow one way and not the other. A one-way street for electrons, if you like. What actually happens is that, in the realm of physics, depending on whether the diode is forward-biased (when the positive voltage is connected to the anode (the red +)) or reverse-biased (when the positive voltage is connected to the cathode (the blue -)). With forward bias, you simply get current flowing through (which has a typical voltage drop of 0.7V); in the reverse bias, you get a depletion region in the PN junction, which prevents current flow.
Now, if we were to get two diodes and slap them back to back, we would get a transistor (Fig 4). Fig 5 shows the NPN layout, and, unsurprisingly, this is called an NPN transistor. PNP transistors also exist, with a P region either side of the N region. The process for producing these semiconductor materials is very complex which is why no two transistors are exactly the same—thus, though we have to design circuits around this property, it can be exploited in many useful ways. I have also greatly exaggerated the P region in Fig 5. as it is very very very thin (that is, measurable in nanometers to microns) in the real thing. The PN junction is the basic building block for pretty much all semiconductor components such as LEDs, all transistors, and diodes, Solar Cells and ICs (Integrated Circuits).
Fig. 6 shows the electrical symbol for a NPN bipolar transistor. Bipolar transistors are used mainly for amplification, due to the fact they are relatively slow at switching. It takes a typical bipolar 300ns to switch on and 1300ns to swtich off whereas FETs take ~50ns to switch on and off making them much more faster. The transistor has three connections labelled B, C and E. These are the base, collector and emitter. Depending on how these three are wired up, determines which one of three main types of amplifiers you have: common collector, emitter follower or common base.
The reason bipolars make great amplifiers is because a small base current will cause a much larger emitter-collector current to flow. This amplification is known as the amplifier’s gain, which has the symbol β (beta) or hfe. The beta of a typical transistor such as the BC337 ranges from 60 to 630. The large variance is due to the complex manufacturing process. If we had a transistor with a beta of 200, and we fed 10mA through the base, we would get 200x10mA through the collector, which is 2A. This is particularly useful when we want large currents to be controlled by integrated circuits which can only pass a small amount of current. DC or AC (and even both superimposed) can be fed through the base.
Even though a gain of 600 is pretty large, what if we require a higher gain? First of all, there are some applications where we deal with very small signals— for example the radio receivers in our mobiles. Typically the received signal is less than a microvolt. To put this into perspective, a microvolt is 1*10^-6V (for those who do not understand standard engineering notation or scientific notation, it is 0.000001V). This is why poorly-designed antennas such as the one in the iPhone 4 will drop your call if you hold the device “incorrectly”. A microvolt is a tiny amount of voltage and too small to be used for anything in general elctronics so it must be amplified. Usually a signal will go through several stages of amplification to get the desired required result, but the same result can be gained with a super-beta transistor which is also known as a Darlington configuration.
The gain of this configuration is beta^2. So in the previous example the beta value was 200, in this new configuration the beta value is 200^2 which equals 40000. If you think about it, that’s absolutely insane—mind you don’t blow the transistor by applying too much base current!
Transistors can be found in almost anything that runs on electricity, quietly doing their jobs so we can live much easier lives. Did you know that each pixel on your TFT (Thin Film Transistor) monitor has a transistor on it? Incredible, if you ask me.