24 January 2003

A Brief Review of Transistor Theory

by Paul Dito

Transistors are literally ubiquitous. The most advanced CPUs and telecommunication devices are ultimately a collection of transistor circuits. Figure 1 shows a wafer of silicon with circuits of transistors ready to be chopped up and put in plastic packages for mounting of circuit boards. In most cases the transistors are being used as extremely fast switches in digital circuits like CPUs and DSPs (digital signal processors), but sometimes they are still used as an analog amplifier, as in the transmitter of a cell phone. Both "flavors" of transistor circuits can be useful to the amateur scientist. Transistors can be used as an electronic switch, allowing a computer to switch a high current load, or as a signal amplifier, as in an infrared receiver.

Figure 1. Click image to enlarge.

While a complete understanding of nuclear physics isn’t necessary for proficiency with transistor circuits, some background and a review of fundamentals will help when it comes to circuit design and troubleshooting. One important concept to keep in mind is the idea of electron flow in electric circuits. For various reasons students are still taught that electricity flows from positive to negative. This is likely due to the idea that electricity flowed, much like water, from a high potential to a lower potential. Modern schematics still adhere to this convention (as illustrated by the diode symbol pointing from positive to negative). While this won’t cause confusion for most applications, when it comes to transistors it’s important to know that the flow of electrons is from negative to positive (or from a lower potential to a higher voltage potential). Another important concept that is deceptively simple, but holds the key to transistor operation: like charges repel, opposites attract.

First a review of the history behind the development of transistors. Modern electronics can trace it’s beginning to the light bulb. In 1883 Edison was experimenting with various ways to prevent blackening on the inside of the light bulb. At one point he tried an extra electrode inside the bulb along with the filament. He noted that a small current would flow to the extra electrode when a positive voltage was applied to it. This phenomena was eventually called the "Edison effect" and led Englishman John Ambrose Fleming to the development of the "diode" (‘di’ for two and ‘ode’ for the electrodes). When the extra electrode of the diode was positively biased with respect to the filament electrode, a current will flow between the two. If the potential voltages are reversed, no electrons are attracted from the filament, and therefore no current flows. Essentially the diode is a one way switch.

Figure 2. Click image to enlarge.

Around the turn of the century Lee DeForest discovered the triode vacuum tube (or valve) by placing another element between the filament (cathode) and extra electrode (plate or anode). This third electrode is called the grid, and it controls the amount of current flowing between the cathode and plate depending on it’s electrical bias. "Electrical bias" meaning if the grid was negative (or a lower potential) with respect to the cathode, it would repel the negative charge of the electrons flowing to the plate, thus choking off the current. This is the principal of amplification: one small voltage or current controlling a much larger voltage or current.

The transistor was made possible by the discovery of semiconductor elements. The transistor was developed at Bell Telephone Labs. The invention was announced in 1948 and in 1956 three key scientists received the Nobel Prize in Physics for their discovery. The three were Dr. John Bardeen, Dr. Walker H. Brattain, and Dr. William Shockley.

Semiconductors are made by taking elements that are generally insulators, like silicon (Si) or Germanium (Ge), to which impurities are added. Silicon, for example, has 4 valance electrons, which are very stable. Since these electrons are strongly bonded to the nucleus of the silicon, it is hard to dislodge them and current won’t flow through the material. When impurities, such as phosphorus (P), arsenic (As), or boron (B) are added to the lattice of silicon, free and mobile charges are created which allow current to flow. The process of introducing these impurities is known as "doping" the silicon.

Figure 3. Click image to enlarge.

There are two basic forms of semiconductor material: P type and N type. The N is for negative and denotes doped silicon that contains extra electrons. An example of N type semiconductor is silicon doped with arsenic. P type is positive and has less electrons, sometimes referred to as having "holes". An example of P type material is silicon with some boron added. Doping takes place by heating silicon in a crucible and introducing gasses under the correct conditions to control the diffusion process.

 

Manufacturers spend billions of dollars on semiconductor foundries to get these processes exactly right. The transistors in the CPU of the host you use to browse the web are based on sub-micron geometries. Just a few molecules wandering where they shouldn’t can ruin many thousands of dollars worth of processors. That is the significance of the "bunny suits" you see technicians wear in sterile semiconductor manufacturing environments. A human hair is much wider than the circuit traces connecting computer cores together. Figure two shows some transistors in an op-amp circuit.

Figure 4. Click image to enlarge.

Semiconductor diodes are formed by putting P type material in contact with N type. The attraction/repulsion of charges allow current to pass in one direction depending on external voltages. Figure three shows a circuit with a PN diode with a positive voltage applied to the N side, while a negative voltage is connected to the P side. Since opposite charges attract, all of the spare electrons in the N-type material are pulled away from the junction toward the voltage source, likewise with the "holes" on the P side. No current flows and the area devoid of carrier charges is called the depleted region. This is called reverse biasing the diode.

If the voltages are reversed everything changes. The free electrons on the N side move toward the junction where they combine with holes on the P side. These electrons are replaced from the voltage source, and current flows through the diode (Figure 4). In this case the diode if forward biased. This is an oversimplification as there are other forces at play, but this explains the basic operation of a semiconductor diode.

Figure 5. Click image to enlarge.

The transistor was developed at Bell Telephone Labs. The invention was announced in 1948 and in 1956 three key scientists received the Nobel Prize in Physics for their discovery. The three were Dr. John Bardeen, Dr. Walker H. Brattain, and Dr. William Shockley. J.R. Pierce of Bell Labs coined the term ‘transistor’ from transfer resistor because transistors use an input current to control an output voltage.

A transistor is formed by placing two diodes back to back in one piece of semiconductor material. This is done by layering N and P material in a sandwich (Figure 5). In reality a transistor etched into a die would look more like Figure 6. The external connections are now called emitter, base, and collector. The base layer is very thin compared to the other layers.

Figure 6. Click image to enlarge.

When the base is more positive than the emitter, and the collector more positive than the base, then electrons will move from the emitter (hence the name) and will move toward the base. Once in the base, most of the electrons will be swept up by the collector because the base is very thin and lightly doped and the collector is reversed biased, allowing the electrons to be ‘sucked’ (or collected) into the positive terminal of the voltage source (Figure 7).

Figure 7. Click image to enlarge.

The current between the emitter and collector will be much greater than the current through the emitter/base connection, so the transistor is very useful in amplification applications. It’s also useful as a switch, since a small signal (from a computer port, for example) can control a much larger current through a heavy load (a motor for example).

This is a drastic oversimplification of a very complex phenomena, but it should give you an idea of what goes on inside a transistor. Once you start experimenting the concepts of biasing should become more clear. Also, the failure modes of transistors may make more sense now. Now that you know the base is a very thin piece of impure silicon, it’s easy to see how excess current through the base/emitter can allow, because of the gain, enough current through the device to destroy itself.

Next time we’ll play with some basic DC circuits to get a feel for the currents through an active transistor….

 

 

 

References

Frederiksen, Thomas M., Intuitive IC Electronics, McGraw Hill 1989.

Carr, Joseph J., Elements of Electronic Communications, Prentice-Hall, 1978.