In Part 2, George discussed devices built with one P-N junction, appropriately named diodes. In this article, he considers devices with more junctions. He starts with two and looks at the ubiquitous, three-terminal bipolar junction transistor (BJT). George looks at the math, science and circuitry of these devices.
Last month we dived deeper into diodes. Now it’s time to consider devices with more junctions. Let’s start with the three-terminal bipolar junction transistor (BJT). Its terminals are called collector (C), emitter (E) and base (B). Figure 1 explains its construction, which is essentially a combination of two junction diodes. Based on their connections, we have two basic structures: PNP (Figure 1a) and NPN (Figure 1b), with the transistors designated accordingly. PNP and NPN devices with similar characteristics are called complementary pairs. In the early days, germanium (Ge) was used as the base semiconductor material and although germanium NPN transistors existed, the PNP variety ruled. The first NPN transistor I got to use was the silicon (Si) type.
The first point-contact transistor was the result of the work of American physicists John Bardeen, Walter Brattain and William Shockley in 1947—an achievement for which they received the Nobel Prize in Physics. As it usually happens with many inventions, the transistor was developed independently in Europe in 1948 by Germans Herbert Mataré and Heinrich Welker. The bipolar junction transistor was then developed and patented by William Shockley in 1950 at Bell Labs.
Germanium transistors manufactured in the ‘50s by the diffusion method had many growing pains. One was a poor frequency response due to large capacitance of the diffused electrodes. But the manufacturing technology improved and by the ‘60s we had ultra-high frequency (UHF) devices, for example the amazing (for their time) AF139 and AF239 PNP transistors. Germanium devices suffered from high leakage, relatively low operating voltage and, above all, significant temperature dependency. But all that changed with the arrival of the silicon planar NPN transistor.
One issue you have to keep in mind is that transistors are current amplifiers. They’re not voltage amplifiers like vacuum tubes before them and field effect transistors (FET) today. BJTs operate in distinct modes. The first is linear, where the collector current IC = β x IB. β stands for the transistor’s current gain—generally greater than 100 in modern devices.
The second mode of operation is saturation, which is used in digital circuitry. The transistor becomes a switch. It is fully turned on with the saturated collector current determined by its type and collector-emitter voltage VCE close to zero. Or it is turned off (cut off) with IB = 0. The collector current IC would ideally be zero too, but there is always some leakage.
THREE amplifier configurations
Let’s consider the three fundamental transistor amplifier circuit configurations as seen in Figure 2. Because of the present-day prevalence of NPN transistors, I shall use them in my examples whenever possible. We begin with the common base topology (Figure 2a). The base is grounded and, therefore, common to both the input and output. To amplify an AC signal, you’ll need to bias the base to overcome the base-emitter diode’s forward voltage—around 0.65 V for silicon transistors—and cause a base current IB to flow. The current gain of the common base topology is less than 1 because the collector current IC flows through the emitter as well. The input impedance of the common base topology is very low. The voltage gain, however, is high—provided the load resistance RL is also high. It is defined by Equation 1:
Common base configuration is used primarily in radio frequency (RF) circuits because it minimizes frequency-limiting collector-base capacitance. Common (grounded) emitter configuration is commonly used in amplifiers as well as switching circuits because it has the highest power gain. Similar to common base, the input impedance is somewhat low, but can be increased, at the cost of gain, by a small resistor between the emitter and ground. Here, the emitter current IE = IC+IB. The ratio IC/IE is called α and is always less than one. The relationship of the transistor currents can be expressed mathematically as:
Common collector topology is better known as emitter follower. It is frequently used for transformation of high impedance input signals to low impedance output. The current gain equals approximately β of the transistor and the input resistance, as a rule of thumb, β x RL. The gains are expressed mathematically as:
Characteristics of the three transistor topologies are summarized in Table 1. Sometime in the past, the common emitter DC gain symbol β was replaced with hFE. This is the parameter you will find nowadays in transistor specification sheets. It is an abbreviation that stands for “hybrid parameter forward current gain, common emitter.” Figure 3 is an example of a common emitter, NPN transistor collector’s I-V (current-voltage) characteristic where the base current IB is a parameter. The I-V characteristic of the base current versus base voltage is that of a diode as presented in Part 2 of this article series, available online here and the fill issue here: Circuit Cellar 351, October 2019.
|Common Base||Common Emitter||Common Collector|
|Output Impedance||Very High||High||Low|
|Phase Shift||0 degrees||180 degrees||0 degrees|
|Power Gain||Low||Very High||Medium|
Table 1 – Summary of characteristics of transistor amplifier topologies
Notice that the collector current IC dependency on the collector voltage VC is quite large for small collector voltages. Once the collector current saturation is reached the current changes very little. You can analyze transistor amplifiers’ low frequency response by utilizing the transistor’s equivalent circuit. More often than not parasitic characteristics are considered negligible for the given application and, therefore, ignored. Figure 4a is the common emitter equivalent circuit. For practical reasons it is often converted into a so-called T-circuit equivalent Figure 4b.
For the purpose of electrical analysis, you can consider the transistor to be a black box—a four terminal linear network as is shown in Figure 5. A transistor is a three-terminal device, but one terminal, the emitter for the common emitter configuration is, obviously, common. You can analyze this network under different conditions, each rendering a different set of parameters. With an open circuit, for example, impedance z-parameters will result. Short circuit conditions will produce admittance y-parameters. But, because transistors in common emitter connections have low input and high output impedances, it is advantageous to use hybrid parameters, called h-parameters. But this is not the end of it! There are also parallel-series m-parameters, cascade-forward a-parameters and cascade-backwards b-parameters.
So which parameters do you select and what can you do with them? For once, knowing one set of the parameters, you can convert them mathematically to any other set with just a small error. Then, inserting them into a matrix as shown by Equation (5), the performance data of the black box can be calculated. Equation (5) uses the h-parameters:
All this theory is very interesting, but unless you have access to a curve tracer and measure those parameters yourself, you’re out of luck. I’ve not found the black box parameters in any specification published by a transistor manufacturer.
But that’s is not a showstopper. Present day manufacturers provide a number of graphs with their flagship devices, allowing you to design any transistor circuit you can imagine. But you can also find many inexpensive transistors—on e-Bay for instance—whose specifications, if you’re lucky to get any, may provide you with perhaps only the following (and nothing else): maximum voltage and current ratings, perhaps the pin-out and maybe the hFE. And yet, even that isn’t a showstopper either. You can still build a circuit satisfying simple requirements, especially for low frequency operation.
Relying on feedback, you can set the DC operating point even without knowing the accurate hFE. Quite often it is safe to assume the hFE will be greater than 100. In such a case, a single stage, common emitter amplifier with a small emitter resistor RE will provide voltage gain of approximately RL/RE—where RL is the load resistance, comprising the collector resistor in parallel with whatever the additional load may be. For relatively slow switching digital circuits, the design is even simpler.
Next month, we continue this article series. In Part 4, I’ll show you some useful discrete transistor circuits and then we’ll zero in on the field effect transistors: Junction FETs and MOS.
For detailed article references and additional resources go to: www.circuitcellar.com/article-materials
PUBLISHED IN CIRCUIT CELLAR MAGAZINE • NOVEMBER 2019 #352 – Get a PDF of the issueSponsor this Article
George Novacek was a retired president of an aerospace company. He was a professional engineer with degrees in Automation and Cybernetics. George’s dissertation project was a design of a portable ECG (electrocardiograph) with wireless interface. George has contributed articles to Circuit Cellar since 1999, penning over 120 articles over the years. George passed away in January 2019. But we are grateful to be able to share with you several articles he left with us to be published.