Short Topical Videos
- Horowitz & Hill, The Art of Electronics, 2nd Ed., Ch. 2
There are two broad classes of transistors: Bipolar Junction Transistors (BJTs), which are often used in discrete analog circuits and can typically provide higher gain over wider bandwidths, and Field Effect Transistors (FETs), which are sometimes used in front-end amplifiers because of their lower noise figures (even though they typically provide less gain over narrower bandwidths than BJTs), and are often used to act like a voltage-controlled switch, such as they do in almost all digital processors today.
Bipolar Junction Transistors
BJT transistors have 3 terminals: the emitter, the base, and the collector. Broadly speaking, a current to/from the base (for NPN/PNP-type transistors, respectively, as described below) is used to control the flow of charge from the collector to the emitter. As long as a couple of rules are followed, the behavior of BJTs is pretty straight-forward. These rules are different for NPN and PNP transistors.
For an NPN transistor, the rules are:
An NPN bipolar junction transistor
- In order to conduct, the voltage difference between the collector and the emitter () must be above a certain threshold (say, ).
- In order to conduct, the voltage difference between the base and emitter () must be above a certain threshold (say, ).
- Once it is conducting, the current flowing from the base () causes a current to flow from collector to emitter (), amplified by a factor .
is often in the range of 60-100, but can vary a lot from part to part.
For PNP transistors, the rules are similar, except that the emitter is now must be at a higher voltage than the collector, and the base must be at a lower voltage than the emitter:
A PNP bipolar junction transistor
- must be above, say, 0.2 V, to conduct.
- I) must be above, say 0.7 V.
The trick to designing circuits using BJTs is to use these 3 rules to effectively regulate when and how much the transistor conducts, based on a signal applied to the base. Because varies so much between transistors, it is considered poor practice to rely on being a particular value. Rather, as is discussed in more detail with regard to amplifiers, it is better to use resistors and capacitors in circuits that regulate gain to a value less than on the basis of the first two rules.
Field Effect Transistors (FETs)
All FETs have gate, drain, and source terminals that correspond roughly to the base, collector, and emitter of BJTs. FETs have a very high input resistance, on the order of 100M or more. This makes it effectively a voltage-controlled device, with a high degree of isolation between input and output. FETs have n-channel and p-channel varieties, which are analogous to the NPN and PNP types of bipolar junction transistors.
as a function of , for various (linearly spaced) values
FETs have essentially three operating modes that relate to a threshold voltage :
- Cutoff: causes no charge to flow from drain to source.
- Ohmic: and causes the FET to behave like a resistor whose value is controlled by .
- Active: and causes the FET to conduct, with the current being independent of , but sensitive to .
BJT and FET Applications
Both BJTs and FETs are commonly used to amplify signals. The details of amplifier design are discussed in another section, but the principle is to use a low-amplitude signal to control to a “valve” through which a lot of charge is able to flow.
Loading (when connecting a low resistance to an output line causes the signal on that line to droop) can be overcome by the same principle as amplifiers. When the signal determining the output voltage level is insulated from the current that flows through the circuit, the output resistance of the circuit is effectively divided by the inherent gain of the transistor.
By setting biasing to a fixed level, one can make a current flow through the transistor that is independent of the impedance of the load that is attached. This, of course, only works within the rules we established above for the operation of BJTs and FETs.
Switching Digital Signals
A NAND logic gate (“not and”) built from FETs
Modern digital electronics are built from billions of transistors. All logical operations can be synthesized by connecting various numbers of NAND gates, which (as illustrated above) can be built from FET switches.