Semiconductors - Practical Electronics for Inventors, Fourth Edition - Paul Scherz, Simon Monk

Practical Electronics for Inventors, Fourth Edition - Paul Scherz, Simon Monk (2016)

Chapter 4. Semiconductors

4.1 Semiconductor Technology

The most important and perhaps most exciting electrical devices used today are built from semiconductive materials. Electronic devices, such as diodes, transistors, thyristors, thermistors, photovoltaic cells, phototransistors, photoresistors, lasers, and integrated circuits, are all made from semiconductive materials, or semiconductors.



4.1.1 What Is a Semiconductor?

Materials are classified by their ability to conduct electricity. Substances that easily pass an electric current, such as silver and copper, are referred to as conductors. Materials that have a difficult time passing an electric current, such as rubber, glass, and Teflon, are called insulators. There is a third category of material whose conductivity lies between those of conductors and insulators. This third category of material is referred to as a semiconductor. A semiconductor has a kind of neutral conductivity when taken as a group. Technically speaking, semiconductors are defined as those materials that have a conductivity σ in the range between 10−7 and 103 mho/cm (see Fig. 4.2). Some semiconductors are pure-elemental structures (e.g., silicon, germanium), others are alloys (e.g., nichrome, brass), and still others are liquids.




Silicon is the most important semiconductor used in building electrical devices. Other materials such as germanium and selenium are sometimes used, too, but they are less popular. In pure form, silicon has a unique atomic structure with very important properties useful in making electrical devices.



Silicon is ranked second in the order of elements appearing in the earth’s crust, an average of 27 percent occurring in igneous rocks. It is estimated that a cubic mile of seawater contains about 15,000 tons of silicon. It is extremely rare to find silicon in its pure crystalline form in nature, and before it can be used in making electronic devices, it must be separated from its binding elements. After individuals—chemists, material scientists, etc.—perform the purification process, the silicon is melted and spun into a large “seed” crystal. This long crystal can then be cut up into slices or wafers that semiconductor-device designers use in making electrical contraptions.



For the semiconductor-device designer, a silicon wafer alone does not prove very useful. A designer would not use the silicon wafer in its pure form to build a device; it just does not have quite the right properties needed to be useful. A semiconductor-device designer is looking for a material that can be made to alter its conductive state, acting as a conductor at one moment and as an insulator at another moment. For the material to change states, it must be able to respond to some external force applied at will, such as an externally applied voltage. A silicon wafer alone is not going to do the trick. In fact, a pure silicon wafer acts more as an insulator than a conductor, and it does not have the ability to change conductive states when an external force is applied. Every designer today knows that a silicon wafer can be transformed and combined with other transformed silicon wafers to make devices that have the ability to alter their conductive states when an external force is applied. The transforming process is referred to as doping.


Doping refers to the process of “spicing up” or adding ingredients to a silicon wafer in such a way that it becomes useful to the semiconductor-device designer. Many ingredients can be added in the doping process, such as antimony, arsenic, aluminum, and gallium. These ingredients provide specialized characteristics such as frequency response to applied voltages, strength, and thermal integrity, to name a few. By far, however, the two most important ingredients that are of fundamental importance to the semiconductor-device designer are boron and phosphorus.

When a silicon wafer is doped with either boron or phosphorus, its electrical conductivity is altered dramatically. Normally, a pure silicon wafer contains no free electrons; all four of its valence electrons are locked up in covalent bonds with neighboring silicon atoms (see Fig. 4.5). Without any free electrons, an applied voltage will have little effect on producing an electron flow through the wafer.



A silicon wafer in pure form doesn’t contain any free charge carriers; all the electrons are locked up into covalent bonds between neighboring atoms.

However, if phosphorus is added to the silicon wafer, something very interesting occurs. Unlike silicon, phosphorus has five valence electrons instead of four. Four of its valence electrons will form covalent bonds with the valence electrons of four neighboring silicon atoms (see Fig. 4.6). However, the fifth valence electron will not have a “home” (binding site) and will be loosely floating about the atoms. If a voltage is applied across the silicon-phosphorus mixture, the unbound electron will migrate through the doped silicon toward the positive voltage end. By supplying more phosphorus to the mixture, a larger flow of electrons will result. Silicon that is doped with phosphorus is referred to as n-type silicon, or negative-charge-carrier-type silicon.



A phosphorus atom added to a silicon wafer provides an extra unbound electron that aids in conduction. Silicon doped with phosphorus is called n-type silicon.

Now, if you take pure silicon and add some boron, you will see a different kind of conduction effect. Boron, unlike silicon or phosphorus, contains only three valence electrons. When it is mixed with silicon, all three of its valence electrons will bind with neighboring silicon atoms (see Fig. 4.7). However, there will be a vacant spot—called a hole—within the covalent bond between one boron and one silicon atom. If a voltage is applied across the doped wafer, the hole will move toward the negative voltage end, while a neighboring electron will fill in its place. Holes are considered positive charge carriers even though they do not contain a physical charge per se. Instead, it only appears as if a hole has a positive charge because of the charge imbalance between the protons within the nucleus of the silicon atom that receives the hole and the electrons in the outer orbital. The net charge on a particular silicon atom with a hole will appear to be positive by an amount of charge equivalent to one proton (or a “negative electron”). Silicon that is doped with boron is referred to as p-type silicon, or positive-charge-carrier-type silicon.



When boron is added to silicon, a hole is formed. This hole acts like a positive charge (see text) that aids in conduction. Silicon doped with boron is called p-type silicon.

As you can see, both n-type and p-type silicon have the ability to conduct electricity; one does it with extra unbound electrons (n-type silicon), and the other does it with holes (p-type silicon).

A Note to Avoid Confusion

Boron atoms have three valence electrons, not four like silicon. This means that the combined lattice structure has fewer free valence electrons as a whole. However, this does not mean that a p-type silicon semiconductor has an overall positive charge; the missing electrons in the structure are counterbalanced by the missing protons in the nuclei of the boron atoms. The same idea goes for n-type silicon, but now the extra electrons within the semiconductor are counterbalanced by the extra protons within the phosphorus nuclei.

Another Note to Avoid Confusion (Charge Carriers)

What does it mean for a hole to flow? A hole is nothing, right? How can nothing flow? Well, it is perhaps misleading, but when you hear the phrase “hole flow” or “flow of positive charge carriers in p-type silicon,” electrons are in fact flowing. You may say, doesn’t that make this just like the electron flow in n-type silicon? No. Think about tipping a sealed bottle of water upside down and then right side up (see Fig. 4.8). The bubble trapped in the bottle will move in the opposite direction of the water. For the bubble to proceed, water has to move out of its way. In this analogy, the water represents the electrons in the p-type silicon, and the holes represent the bubble. When a voltage is applied across a p-type silicon semiconductor, the electrons around the boron atom will be forced toward the direction of the positive terminal. Now here is where it gets tricky. A hole next to a boron atom will be pointing toward the negative terminal. This hole is just waiting for an electron from a neighboring atom to fall into it, due in part to the lower energy configuration there. Once an electron, say, from a neighboring silicon atom, falls into the hole in the boron atom’s valence shell, a hole is briefly created in that silicon atom’s valence shell. The electrons in the silicon atom lean toward the positive terminal, and the newly created hole leans toward the negative terminal. The next silicon atom over will let go of one of its electrons, and the electron will fall into the hole, and the hole will move over again—the process continues, and it appears as if the hole flows in a continuous motion through the p-type semiconductor.



A Final Note to Avoid Confusion

And finally, why is a hole called a positive charge carrier? How can “nothing” carry a positive charge? Well, what’s going on here is this: A hole, as it moves through the mostly silicon-based crystal, causes a brief alteration in the electrical field strength around the silicon atom in the crystal where it happens to be situated. When an electron moves out of the way, thus creating a new hole, this silicon atom as a whole will be missing an electron, and hence the positive charge from the nucleus of the silicon atom will be felt (one of the protons is not counterbalanced). The “positive charge carrier” attributed to holes comes from this effective positive nuclear charge of the protons fixed within the nucleus.

4.1.2 Applications of Silicon

You may be asking yourself, why are these two new types of silicon (n-type and p-type) so useful and interesting? What good are they for semiconductor-device designers? Why is there such a fuss over them? These doped silicon crystals are now conductors. Big deal, right? Yes, we now have two new conductors, but the two new conductors have two unique ways of passing an electric current—one does it with holes, the other with electrons. This is very important.

The manners in which n-type and p-type silicon conduct electricity (electron flow and hole flow) are very important in designing electronic devices such as diodes, transistors, and solar cells. Some clever people figured out ways to arrange slabs, chucks, strings, etc. made of n-type and p-type silicon in such a way that when an external voltage or current is applied to these structures, unique and very useful features result. These unique features are made possible by the interplay between hole flow and electron flow between the n-type and p-type semiconductors. With these new n-type/p-type contraptions, designers began building one-way gates for current flow, opening and closing channels for current flow controlled by an external electrical voltage and/or current. Folks figured out that when an n-type and a p-type semiconductor were placed together and a particular voltage was applied across the slabs, light, or photons, could be produced as the electrons jumped across the junction between the interface. It was noticed that this process could work backward as well. That is, when light was exposed at the np junction, electrons were made to flow, thus resulting in an electric current. A number of clever contraptions have been built using n-type and p-type semiconductor combinations. The following chapters describe some of the major devices people came up with.



4.2 Diodes

A diode is a two-lead semiconductor device that acts as a one-way gate to electric current flow. When a diode’s anode lead is made more positive in voltage than its cathode lead—a condition referred to as forward biasing—current is permitted to flow through the device. However, if the polarities are reversed (the anode is made more negative in voltage than the cathode)—a condition referred to as reversed biasing—the diode acts to block current flow.



Diodes are used most commonly in circuits that convert ac voltages and current into dc voltages and currents (e.g., ac/dc power supply). Diodes are also used in voltage-multiplier circuits, voltage-shifting circuits, voltage-limiting circuits, and voltage-regulator circuits.

4.2.1 How p-n Junction Diodes Work

A p-n junction diode (rectifier diode) is formed by sandwiching together n-type and p-type silicon. In practice, manufacturers grow an n-type silicon crystal and then abruptly change it to a p-type crystal. Then either a glass or plastic coating is placed around the joined crystals. The n side is the cathode end, and the p side is the anode end.

The trick to making a one-way gate from these combined pieces of silicon is getting the charge carriers in both the n-type and p-type silicon to interact in such a way that when a voltage is applied across the device, current will flow in only one direction. Both n-type and p-type silicon conducts electric current; one does it with electrons (n-type), and the other does it with holes (p-type). Now the important feature to note here, which makes a diode work (act as a one-way gate), is the manner in which the two types of charge carriers interact with each other and how they interact with an applied electrical field supplied by an external voltage across its leads. Below is an explanation describing how the charge carriers interact with each other and with the electrical field to create an electrically controlled one-way gate.

Forward-Biased (“Open Door”)


When a diode is connected to a battery, as shown here, electrons from the n side and holes from the p side are forced toward the center (pn interface) by the electrical field supplied by the battery. The electrons and holes combine, and current passes through the diode. When a diode is arranged in this way, it is said to be forward-biased.

Reverse-Biased (“Closed Door”)


When a diode is connected to a battery, as shown here, holes in the n side are forced to the left, while electrons in the p side are forced to the right. This results in an empty zone around the p-n junction that is free of charge carriers, better known as the depletion region. This depletion region has an insulative quality that prevents current from flowing through the diode. When a diode is arranged in this way, it is said to be reverse-biased.


A diode’s one-way gate feature does not work all the time. That is, it takes a minimal voltage to turn it on when it is placed in forward-biased direction. Typically for silicon diodes, an applied voltage of 0.6 V or greater is needed; otherwise, the diode will not conduct. This feature of requiring a specific voltage to turn the diode on may seem like a drawback, but in fact, this feature becomes very useful in terms of acting as a voltage-sensitive switch. Germanium diodes, unlike silicon diodes, often require a forward-biasing voltage of only 0.2 V or greater for conduction to occur. Figure 4.12 shows how the current and voltage are related for silicon and germanium diodes.



Another fundamental difference between silicon diodes and germanium diodes, besides the forward-biasing voltages, is their ability to dissipate heat. Silicon diodes do a better job of dissipating heat than germanium diodes. When germanium diodes get hot—temperatures exceeding 85°C—the thermal vibrations affect the physics inside the crystalline structure to a point where normal diode operation becomes unreliable. Above 85°C, germanium diodes become worthless.

4.2.2 Diode Water Analogy

A diode (or rectifier) acts as a one-way gate to current flow—see the water analogy in Fig. 4.13. Current flows in the direction of the arrow, from anode (+) to cathode (-), provided the forward voltage VF across it exceeds what’s called the junction threshold voltage. As a general rule of thumb, silicon p-n junction diodes have about a 0.6-V threshold, germanium diodes a 0.2-V threshold, and Schottky diodes a 0.4-V threshold. However, don’t take this rule too seriously, because with real-life components, you’ll find these thresholds may be a few tenths of a volt off. For example, it’s entirely possible for a silicon p-n junction diode’s threshold to be anywhere between 0.6 and 1.7 V; for germanium, 0.2 to 0.4 V; and for Schottky diodes, 0.15 to 0.9 V.



Note that if you actually put 12 V across a forward-biased diode as shown in Fig. 4.13, a very large current would flow, probably destroying the diode. Also, the axes of Fig. 4.13 are not to scale.

In terms of limits, avoid supplying a diode with a forward current IF beyond its peak current rating I0(max). If you do, you’ll get internal junction meltdown. Likewise, avoid applying a reverse voltage VR any bigger than the diode’s peak inverse voltage (PIV) rating. This, too, can render a diode worthless. See the graph in Fig. 4.13.

4.2.3 Kinds of Rectifiers/Diodes

There are numerous types of diodes, each specifically designed to work better in one application or another. Diodes for high-power applications (switching, power supplies, etc.) which draw lots of current or rectify high voltages typically go by the name rectifier diodes. On the other hand, diodes that go by names such as signal, switching, fast recovery, or high speed are designed to provide a low internal capacitance (they store less charge but usually have weaker junctions for large currents). At high speeds, these diodes will reduce RC switching time constants, which means fewer time delays and lower signal losses.

Schottky diodes have a particularly low junction capacitance and faster switch-ing (∼10 ns) when compared to silicon p-n junction diodes due to their special metal-semiconductor-junction interface. They also have a lower junction threshold voltage—as low as 0.15 V, but usually a bit more (0.4 V average). Both these characteristics enable them to detect low-voltage, high-frequency signals that ordinary p-n junction diodes would not see. (A Schottky with a 0.3-V threshold can pass signals greater than 0.3 V, but a silicon p-n junction diode with a 0.7-V threshold can only pass signals greater than 0.7 V.) For this reason, Schottky diodes are very popular in low-voltage signal rectifiers in RF circuits, signal switching in telecommunication, small dc/dc converters, small low-voltage power supplies, protection circuits, and voltage clamping arrangements. Their high-current density and low-voltage drop also make them great in power supplies, since they generate less heat, requiring smaller heat sinks to be used in design. Therefore, you’ll find both rectifier and fast-switching Schottky diodes listed in the catalogs.

Germanium diodes are used mostly in RF signal detection and low-level logic design due to their small threshold voltage of about 0.2 V. You do not see them in high-current rectifying applications, since they are weaker and leak more than silicon diodes when temperatures rise. In many applications, a good Schottky signal diode can replace a germanium diode.

Common Diode/Rectifier Packages



4.2.4 Practical Considerations

Five major specs to consider when choosing a diode are peak inverse voltage, PIV; forward current-handling capacity, IO(max); response speed tR (time for diode to switch on and off); reverse-leakage current, IR(max); and maximum forward-voltage drop, VF(max). Each of these characteristics can be manipulated during the manufacturing process to produce the various special-purpose diodes. In rectification applications (e.g., power supplies, transient protection), the most important specs are PIV and current rating. The peak negative voltages that are stopped by the diode must be smaller in magnitude than the PIV, and the peak current through the diode must be less than IO(max). In fast and low-voltage applications, tR and VF become important characteristics to consider. In the applications section that follows, you’ll get a better sense of what all these specs mean.

TABLE 4.1 Selection of Popular Diodes









Signal (Ge)


8.5 mA

15 µA



Signal (Ge)


4.0 mA

5 µA



Signal (Ge)


5.0 mA



Fast Switch


75 mA

25 nA





10 mA

25 nA

450 mA





100 mA

50 nA





1 A

0.03 mA

30 A





1 A

0.03 mA

30 A





1 A

0.03 mA

30 A





1 A

0.03 mA

30 A





1 A

0.03 mA

30 A





3 A

500 µA

200 A




3 A

500 µA

200 A




3 A

500 µA

200 A




1 A

1 mA

25 A





1 A

25 A




1 A

25 A





3 A




15 mA

50 mA





1 mA

100 nA

10 mA


Diodes come in a variety of different packages. Some are standard two-lead deals; others are high-power packages with heat-sink attachments (e.g., TO-220, DO-5). Some come in surface-mount packages, and others contain diode arrays in IC form, used for switching applications. Dual-diode and diode-bridge rectifiers also come in a variety of package sizes and shapes for varying power levels.

4.2.5 Diode/Rectifier Applications

Voltage Dropper



When current passes through a diode, there is a voltage drop across it of about 0.6 V, for a silicon p-n junction diode. (Germanium diodes have around a 0.2-V drop; Schottky, around 0.4 V—all these values vary slightly, depending on the specific diode used.) By placing a number of diodes in series, the total voltage drop across the combination is the sum of the individual voltage drops across each diode. Voltage droppers are often used in circuits where a fixed small voltage difference between two sections of a circuit is needed. Unlike resistors that can be used to lower the voltage, the diode arrangement typically doesn’t waste as much power to resistive heating and will supply a stiffer regulated voltage that is less dependent on current variations. Later in this chapter, you’ll see that a single zener diode can often replace a multiple series diode arrangement like the one shown here.

Voltage Regulator



Here’s a spin-off of the last circuit, making use of the three diodes to create a simple regulated (steady) low-voltage output equal to the sum of the threshold voltages of the diodes: 0.6 V + 0.6 V + 0.6 V = 1.8 V. The series resistor is used to set the desired output current (I) and should be less than the value calculated using the following formula, but not so low that it exceeds the power ratings of itself and the diodes:

Diodes and the series resistor must have proper power ratings for the amount of current drawn. Use P = IV. Note that for higher-power critical voltage sources, you’ll typically use a zener diode regulator or, more commonly, a special regulator IC instead.

Reverse-Polarity Protection

Battery reversal or power polarity reversal can be fatal to portable equipment. The best design is to use a mechanical block to safeguard against reverse installation. However, even momentarily fumbling around and making contacts can lead to problems. This is especially true for one or more single-cell battery applications that use AA-alkaline, NiCad, and NiMH batteries. For these systems you must ensure that any flow of reverse current is low enough to avoid damaging the circuit or the battery.


FIGURE 4.17 Series diode: This is the simplest battery-reversal protection. It can be used with external power connections, too—plug-and-jack. The diode allows current from a correctly installed battery to flow to load, but blocks current flow from a backward-installed battery. The drawback with a series diode is that the diode must handle the full load current. Also, the forward voltage drop of the diode shortens the equipment’s operating time—cuts off about 0.6 V right away. Schottky diodes with low thresholds can do better. See Problem 1 at the end of Sec. 4.2 for another reverse-polarity protection circuit.

Parallel diode: In applications that call for alkaline or other batteries that have high output impedances, you can guard against reverse installations by using a parallel (shunt) diode, while eliminating the diode’s voltage drop. This approach protects the load but draws high current from the backward-installed battery. The diode must be properly rated for current and power. Another application of the parallel diode is in meter protection, where it acts to divert large currents entering the negative terminal of the meter.

Note: In more sophisticated battery-powered designs, special ICs or transistor arrangements are used to provide essentially zero voltage drop protection, while providing a number of other special features, such as reverse polarity protection, thermal shutdown, and voltage level monitor.

Transient Suppression with Fly-Back Diodes


When current flowing through an inductor is suddenly switched off, the collapsing magnetic field will generate a high-voltage spike in the inductor’s coils. This voltage spike or transient may have an amplitude of hundreds or even thousands of volts. This is particularly common within relay coils. A diode—referred to as a fly-back diode for this type of application—placed across the relay’s coil can protect neighboring circuitry by providing a short circuit for the high-voltage spike. It also protects the relay’s mechanical contacts, which often get viciously slapped shut during an inductive spike. Notice, however, that the diode is ineffective during turn-on time. Select a rectifier diode with sufficient power ratings (1N4001, 1N4002, etc.). Schottky rectifiers (e.g., 1N5818) work well, too.


Here’s a more practical example for driving a relay that contains an extra diode placed across a transistor driver in order to protect the transistor from damage due to inductive spikes generated from the relay’s coil when the transistor is turned off. This arrangement also deadens spikes during turn-on time. This dual diode arrangement is sometimes used in voltage regulator circuits, where one diode is wired from the output to the input and another is wired from ground to the output. This prevents any attached loads from sending damaging spikes back into the IC’s output.


Here’s another example of how inductive kickback from a motor that is running and then suddenly turned off can generate a voltage transient that can potentially damage connected electronics—in this case, a 2N2907 transistor. The diode reroutes or shorts the induced voltage to the opposite terminal of the motor. Here a 1N5818 Schottky diode is used—though you could use other p-n junction types, too. The Schottky diode happens to be a bit faster and will clip the transient voltage a bit lower down—at around 0.4 V.

Note: Devices such as TVs and varistors are specially designed for transient suppression. See the section on transient suppressors later in this chapter.


Diode Clamps

Diode clamps are used to clip signal levels, or they can shift an ac waveform up or down to create what’s called a pulsing dc waveform—one that doesn’t cross the 0-V reference.


In the adjustable waveform clipper circuit, the maximum output is clipped to a level determined by the resistance of the potentiometer. The idea is to set the negative end of the diode to about 0.6 V lower than the maximum desired output level, to account for the forward voltage drop of the diode. That’s what the potentiometer is intended to do. +V should be a volt or so higher than the peak input voltage.


The adjustable attenuator is similar to the last circuit, but the additional opposing diode allows for clipping on both positive and negative swings. You can use separate potentiometers if you want separate positive and negative clipping level control. +V should be a volt or so higher than the peak input voltage.


The diode voltage clamp provides dc restoration of a signal that has been ac-coupled (capacitively coupled). This is important for circuits whose inputs look like diodes (e.g., a transistor with grounded emitter); otherwise, an ac-couple signal would fade away.


In the diode switch circuit, an input waveform is ac-coupled to the diode through C1 at the input and C2 at the output. R2 provides a reference for the bias voltage. When the switch is thrown to the ON position, a positive dc voltage is added to the signal, forward-biasing the diode and allowing the signal to pass. When the switch is thrown to the OFF position, the negative dc voltage added to the signal reverse-biases the diode and the signal does not get through.



Half-wave rectifier: Used to transform an ac signal into pulsing dc by blocking the negative swings. A filter is usually added (especially in low-frequency applications) to the output to smooth out the pulses and provide a higher average dc voltage. The peak inverse voltage (PIV)—the voltage that the rectifier must withstand when it isn’t conducting—varies with load, and must be greater than the peak ac voltage (1.4 × Vrms). With a capacitor filter and a load drawing little or no current, it can rise to 2.8 × Vrms (capacitor voltage minus the peak negative swing of voltage from transformer secondary).


Full-wave center-tap rectifier: This commonly used circuit is basically two combined half-wave rectifiers that transform both halves of an ac wave into a pulsing dc signal. When designing power supplies, you need only two diodes, provided you use a center-tap transformer. The average output voltage is 0.9 Vrms of half the transformer secondary; this is the maximum that can be obtained with a suitable choke-input filter. The peak output voltage is 1.4 × Vrms of half the transformer secondary; this is the maximum voltage that can be obtained from a capacitor-input filter. The PIV impressed on each diode is independent of the type of load at the output. This is because the peak inverse voltage condition occurs when diode A conducts and diode B does not conduct. As the cathodes of diodes A and B reach a positive peak (1.4 Vrms), the anode of diode B is at the negative peak, also 1.4 Vrms, but in the opposite direction. The total peak inverse voltage is therefore 2.8 Vrms. The frequency of the output pulses is twice that of the half-wave rectifier, and thus comparatively less filtering is required. Since the diodes work alternately, each handles half of the load current. The current rating of each rectifier need be only half the total current drawn from the supply.


Full-wave bridge rectifier: This rectifier produces a similar output as the last full-wave rectifier, but doesn’t require a center-tap transformer. To understand how the device works, follow the current through the diode one-way gates. Note that there will be at least a 1.2-V drop from zero-to-peak input voltage to zero-to-peak output voltage (there are two 0.6-V drops across a pair of diodes during a half cycle). The average dc output voltage into a resistive load or choke-input filter is 0.9 × Vrms of the transformer’s secondary; with a capacitor filter and a light load, the maximum output voltage is 1.4 × Vrms. The inverse voltage across each diode is 1.4 Vrms; there the PIV of each diode is more than 1.4 Vrms.

See the following text for the pros and cons of the various rectifier configurations.


Voltage Multiplier Circuits



Half-wave voltage doubler: Takes an ac input voltage and outputs a dc voltage that is approximately equal to twice the input’s peak voltage (or 2.8 times the input’s RMS voltage). (The actual multiplication factor may differ slightly, depending on the capacitor, resistor, and load values.) In this circuit, we take VIN to mean the secondary voltage from the transformer. During the first negative half cycle, DA conducts, charging C1 to the peak rectified voltage VIN (peak), or 1.4 VIN (RMS). During the positive half cycle of the secondary voltage, DA is cut off and DB conducts, charging capacitor C2. The voltage delivered to C2 is the sum of the transformer peak secondary voltage, VIN (peak) plus the voltage stored in C1, which is the same, so the sum gives 2 VIN (peak), or 2.8 VIN (RMS). On the next negative cycle, DB is nonconducting and C2 will discharge into an attached load. If no load is present, the capacitors will remain charged—C1 to 1.4 VIN (RMS), C2 to 2.8 VIN (RMS). When a load is connected to the output, the voltage across C2 drops during the negative half cycle and is recharged up to 2.8 VIN (RMS) during the positive half cycle. The output waveform across C2 resembles that of a half-wave rectifier circuit because C2 is pulsed once every cycle. Figure 4.21 shows levels to which the two capacitors are charged throughout the cycle. In actual operation, the capacitor will not discharge all the way to zero, as shown.

Full-wave doubler: During the positive half cycle of the transformer secondary voltage, DA conducts charging C1 to VIN (peak) or 1.4 VIN (RMS). During the negative half cycle, DB conducts, charging C2 to the same value. The output voltage is the sum of the two capacitor voltages, which will be 2 VIN (peak) or 2.8 VIN (RMS) under no-load conditions. The graph shows each capacitor alternately receiving a charge once per cycle. The effective filter capacitance is that of C1 and C2 in series, which is less than the capacitance of either C1 or C2 alone. R1 and R2 are used to limit the surge current through the rectifiers. Their values are based on the transformer voltage and the rectifier surge current rating, since at the instant the power supply is turned on, the filter capacitors look like a short-circuited load. Provided the limiting resistors can withstand the surge current, their current-handling capacity is based on the maximum load current from the supply. The peak inverse voltage across each diode is 2.8 VIN (RMS).

Pros and Cons of the Rectifier Circuits

Comparing the full-wave center-tap rectifier and the full-wave bridge rectifier, you’ll notice both circuits have almost the same rectifier requirements. However, the center-tap version has half the number of diodes as the bridge. These diodes will require twice the maximum inverse voltage ratings of the bridge diodes (PIV > 2.8 Vrms, as opposed to >1.4 Vrms). The diode current ratings are identical for the two circuits. The bridge makes better use of the transformer’s secondary than the center-tap rectifier, since the transformer’s full winding supplies power during both half cycles, while each half of the center-tap circuit’s secondary provides power only during its positive half-cycle. This is often referred to as the transformer utilization factor, which is unity for the bridge configuration and 0.5 for the full-wave center-tap configuration.

The bridge rectifier is often not as popular as the full-wave center-tap rectifier in high-current, low-voltage applications. This is because the two forward-conducting series-diode voltage drops in the bridge introduce a volt or more of additional loss, and thus consume more power (heat loss) than a single diode would within a full-wave rectifier.

In regard to the half-wave configuration, it’s rarely used in 60-Hz rectification for other than bias supplies. It does see considerable use, however, in high-frequency switching power supplies in what are called forward converter and fly-back converter topologies.



Voltage tripler: On one half of the ac cycle, C1 and C3 are charged to VIN (peak) through D1, D2, and D3. On the opposite half of the cycle, D2 conducts and C2 is charged to twice VIN (peak), because it sees the transformer plus the charge in C1 as its source. (D1 is cut off during this half cycle.) At the same time, D3 conducts, and with the transformer and the charge in C2 as the source, C3 is charged to three times the transformer voltage.

Voltage quadrupler: Works in a similar manner as the previous one. In both these circuits, the output voltage will approach an exact multiple of the peak ac voltage when the output current drain is low and the capacitance values high.

Capacitance values are usually 20 to 50 µF, depending on the output current drain. Capacitor dc ratings are related to VIN (peak) by:

C1—Greater than VIN (peak) or 0.7 VIN (RMS)

C2—Greater than 2 VIN (peak) or 1.4 VIN (RMS)

C3—Greater than 3 VIN (peak) or 2.1 VIN (RMS)

C4—Greater than 4 VIN (peak) or 2.8 VIN (RMS)

Diode Logic Gates


FIGURE 4.23 These simple diode logic gates are useful for learning the basics of digital logic, and can also be adapted for non-logic-level electronics (e.g., higher-voltage and power analog-like circuits)—see the following battery-backup example (Fig. 4.24). When designing high-power circuits, make sure your diodes have the proper PIV and current ratings for the job. It’s also important to note that the recovery time of power diodes won’t be as fast as digital logic ICs or fast-switching diodes.

Battery Backup



Devices are powered by a wall adapter with battery backup, typically diode-OR for the battery and wall-adapter connections, as shown in Fig. 4.24. Normally if the switch is closed, power is delivered to the load from the 12-V wall adapter through D1; D2 is reverse-biased (off), since its negative end is 2.4V more positive than its positive end. If power is interrupted (switch opened), D1 stops conducting, and the battery kicks in, sending current through D2 into the load; D1 blocks current from flowing back into the wall adapter. There is a penalty for using diodes for battery backup, however, since the diode in series with the battery limits the minimum voltage at which the battery can supply power (around a 0.6-V drop for silicon p-n junction, 0.4 V for Schottky). Better battery-backup designs implement transistors or special ICs that contain an internal comparator which switches over battery power through a low-resistance transistor without the 0.6-V penalty. Check out MAXIM’s website for some example ICs.

AM Detector



Diodes are often used in the detection of amplitude modulated (AM) signals, as demonstrated in the simple AM radio in Fig. 4.25. Within an AM radio signal, an RF carrier signal of constant high frequency (550 to 1700 kHz) has been amplitude modulated with an audio signal (10 to 20,000 Hz). The audio information is located in duplicate in both upper and lower sidebands, or the envelope of the AM signal. Here, an antenna and LC-tuning circuit act to “resonate” in on the specific carrier frequency of interest (transform radio signal into corresponding electrical signal). A signal diode (e.g., 1N34) is then used to rectify out the negative portion of the incoming signal so it can be manipulated by the next dc stages. The rectified signal is then stripped of its high-frequency carrier by passing through a low-pass filter. The output signal is the audio signal. This signal can be used to drive a simple crystal earpiece, a modern sensitive headphone, or a telephone receiver earpiece. (Low-impedance earphones or speakers will require additional amplification via a coupling capacitor of 1 μF or so.)

Schottky Diode Termination



Schottky diode termination can be used to counteract the high-speed transmission line effects, which cause over/undershoots from signal reflections, reduce noise margins, and destroy timing. These types of distortions can cause false triggering in clock lines and erroneous data on address, data, and control lines, as well as contribute significantly to clock and signal jitter. For applications where transmission line impedance is variable or unknown, it’s not possible to specify a termination resistance value—an alternative is needed. The Schottky diode termination has the ability to maintain signal integrity, save significant power, and permit flexible system design. A Schottky diode termination consists of a diode series combination, where one diode clamps to VCC, or supply voltage, and the other to ground. The diodes at the end of the transmission line minimize the effect of reflection via a clamping operation. The top diode clamps voltages that exceed VCC by the forward-bias threshold limit. This clamping will minimize overshoots caused by reflections. For falling edge signals, a clamp diode to ground affects a similar termination. This clamping function does not depend on matching the transmission line characteristic impedance, making it useful in situations where the line impedance is unknown or variable.

Read-Only Memory (ROM)



This circuit is a simple read-only memory (ROM) made with diodes. Here, the ROM acts as a decimal-to-binary encoder. With no buttons pressed, all LEDs are lit. If 1 is pressed, current from the supply is diverted away from the 23, 22, and 21 lines via the diodes to ground, but is allowed to pass on the 20 line, thus presenting 0001 on the LED readout. In reality, using a PROM such as this for encoding—or anything else, for that matter—isn’t practical. Usually there is a special encoder IC you buy or you simply take care of the encoding—say, with a multiplexed keypad that’s interfaced with a microcontroller—the actual encoding is taken care of at the programming level. At any rate, it’s a fun circuit, and this gives you a basic idea of how read-only memory works.

4.2.6 Zener Diodes

A zener diode acts like a two-way gate to current flow. In the forward direction, it’s easy to push open; only about 0.6 V—just like a standard diode. In the reverse direction, it’s harder to push open; it requires a voltage equal to the zener’s breakdown voltage VZ. This breakdown voltage can be anywhere between 1.8 and 200 V, depending on the model (1N5225B = 3.0 V, 1N4733A = 5.1 V, 1N4739A = 9.1 V, etc.). Power ratings vary from around 0.25 to 50 W.


FIGURE 4.28 The reverse-bias direction is the standard configuration used in most applications, along with a series resistor. In this configuration, the zener diode acts like a pressure release value, passing as much current as necessary to keep the voltage across it constant, equal to VZ. In other words, it can act as a voltage regulator. See application in Fig. 4.29.

Zener Voltage Regulator

These circuits act as voltage regulators, preventing any supply voltage or load current variations from pulling down the voltage supplied to the load. The following explains how the zener diode compensates for both line and load variations.


FIGURE 4.29 Line regulation example: If the line voltage increases, it will cause an increase in line current. Since the load voltage is constant (maintained by the zener), the increase in line current will result in an increase in zener current, thus maintaining a constant load current. If the line voltage decreases, less line current results, and less current is passed by the zener. See graph in Fig. 4.29, top right.

Load regulation example: If the load voltage attempts to decrease as a result of decreased load resistance (increased load current), the increase in load current is offset by the decrease in zener current. The voltage across the load will remain fairly constant. If the load voltage attempts to increase due to an increase in load resistance (decrease in load current), the decrease in load current is offset by an increase in zener current. See graph in Fig. 4.29, bottom right.

The following formulas can be used when selecting the component values:

See Problem 3 at end of this section for a design example.

Note that zener regulators are somewhat temperature dependent and aren’t the best choice for critical applications. A linear regulator IC, though more expensive, is less dependent on temperature variations due to an internal error amplifier. They do typically use an internal zener to supply the reference, nonetheless.

Selection of Popular Zener Diodes








500 MW

1 W

5 W

200 MW

500 MW

1 W



BZX84C2V4, MMBZ5221B








BZX84C3V0, MMBZ52251B

BZT52C3V0, ZMM5225B





BZX84C3V3, MMBZ5226B

BZT52C3V3, ZMM5226B






BZX84C3V6, MMBZ5227B

BZT52C3V6, ZMM5227B





BZX84C3V9, MMBZ5228B

BZT52C3V9, ZMM5228B






BZT52C4V3, ZMM5229B






BZX84C4V7, MMBZ5230B

BZT52C4V7, ZMM5230B






BZX84C5V1, MMBZ5231B

BZT52C5V1, ZMM5231B

SMAZ5V1, ZM4733A





BZX84C5V6, MMBZ5232B

BZT52C5V6, ZMM5232B

SMAZ5V6, ZM4734A




BZT52C6V0, ZMM52330B





BZX84C6V2, MMBZ5234B

BZT52C6V2, ZMM5234B

SMAZ6V2, ZM4735A





BZX84C6V8, MMBZ5235B

BZT52C6V8, ZMM5235B

SMAZ6V8, ZM4736A





BZX84C7V5, MMBZ5236B

BZT52C7V5, ZMM5236B

SMAZ7V5, ZM4737A





BZX84C8V2, MMBZ5237B

BZT52C8V2, ZMM5237B

SMAZ8V2, ZM4738A




BZT52C8V7, ZMM5238B





BZX84C9V1, MMBZ5239B

BZT52C9V1, ZMM5239B

SMAZ9V1, ZM4739A






BZT52C10, ZMM5240B

SMAZ10, ZM4740A





BZX84C11, MMBZ5241B

BZT52C11, ZMM5241B






BZX84C12, MMBZ5242B

BZT52C12, ZMM5242B

SMAZ12, ZM4742A






BZT52C13, ZMM5243B





BZT52C14, ZMM5244B





BZX84C15, MMBZ5245B

BZT52C15, ZMM5245B

SMAZ15, ZM4744A





BZX84C16, MMBZ5246B

BZT52C16, ZMM5246B

SMAZ16, ZM4745A









BZX84C18, MMBZ5248B

BZT52C18, ZMM5248B

SMAZ18, ZM4746A









BZX84C20, MMBZ5250B

BZT52C20, ZMM5250B

SMAZ20, ZM4747A





BZX84C22, MMBZ5251B

BZT52C22, ZMM5251B

SMAZ22, ZM4748A





BZX84C24, MMBZ5252B

BZT52C24, ZMM5252B

SMAZ24, ZM4749A









BZX84C27, MMBZ5254B

BZT52C27, ZMM5254B

SMAZ27, ZM4750A











BZT52C30, ZMM5256B

SMAZ30, ZM4751A






BZT52C33, ZMM5257B

SMAZ33, ZM4752A





BZX84C36, MMBZ5258B

BZT52C36, ZMM5258B

SMAZ36, ZM4753A





BZX84C39, MMBZ5259B

BZT52C39, ZMM5259B

SMAZ39, ZM4754A





BZT52C43, ZMM5260B






BZT52C47, ZMM5261B






BZT52C51, ZMM5262B












































4.2.7 Zener Diode Applications

Split Supply from Single Transformer Winding



Here’s a method for obtaining a split supply from a non-center-tapped transformer using two zener diodes. Z1 and Z2 are selected of equal voltage and power rating for desired split voltage and load. As with the previous example, the temperature dependency of the zener diodes makes this arrangement less accurate than a supply that uses two separate regulator ICs. However, it’s a simple alternative for noncritical applications. See Chap. 11 on power supplies.

Waveform Modifier and Limiter



Two opposing zener diodes act to clip both halves of an input signal. Here a sine wave is converted to a near squarewave. Besides acting to reshape a waveform, this arrangement can also be placed across the output terminal of a dc power supply to prevent unwanted voltage transients from reaching an attached load. The breakdown voltages in that case must be greater than the supply voltage, but smaller than the maximum allowable transient voltage. A single bidirectional TVS does the same thing—see the section on transient suppressors.

Voltage Shifter



This circuit shifts the input voltage down by an amount equal to the breakdown voltage of the zener diode. As the input goes positive, the zener doesn’t go into breakdown until it reaches 5.1 V (for the 1N5281B). After that, the output follows the input, but shifted 5.1 V below it. When the input goes negative, the output will follow the input, but shifted by 0.6 V—the forward threshold voltage drop of the zener.

Voltage Regulator Booster



Zener diodes can be used to raise the level of a voltage regulator and obtain different regulated voltage outputs. Here 3-V and 6-V zener diodes are placed in series to push the reference ground of a 5 V regulator IC up 9 V to a total of 14 V. Note that in real designs, capacitors may be required at the input and output. See the section on voltage regulator ICs.

Overvoltage Protection



If excessive voltage is applied to the jack (say, via an incorrectly rated wall plug-in supply), the zener diode will conduct until the fuse is blown. The breakdown voltage of the zener should be slightly above the maximum tolerable voltage that the load can handle. Either a fast- or a slow-blow fuse can be used, depending on the sensitivity of the load. The current and voltage ratings of the fuse must be selected according to the expected limits of the application. Note that there are other, similar overvoltage protection designs that use special devices, such as TVSs and varistors. These devices are cheap and are very popular in design today.

Increasing Wattage Rating of Zener



Here’s a simple circuit that effectively increases the wattage rating (current-handling capacity) of a zener diode by letting a power transistor take care of the majority of the regulating current. The zener itself takes a small portion of the total current and creates a base voltage/current (with the help of the base-to-ground resistor) that changes the collector-to-emitter current flow according to any variations in line or load current.

Simple LED Voltmeter



Here’s a simple circuit voltmeter that uses the sequence of zener diodes with increasing breakdown voltages. LEDs glow in sequence as the input voltage rises. It’s okay to use different zener diodes so long as the series resistors limit current through LED to a safe level. Most LEDs are happiest around 20 mA or so. You can calculate the worst-case scenario to be at the 5 V LED leg when Vin = 16 V. If you’re looking for more sophistication, you can always use an analog-to-digital converter, along with a microcontroller and LCD or LED display.

4.2.8 Varactor Diodes (Variable Capacitance Diodes)

A varactor or variable capacitance diode (also called a varicap) is a diode whose junction capacitance can be altered with an applied reverse voltage. In this way, it acts as a variable capacitor. As the applied reverse voltage increases, the width of its junction increases, which decreases its capacitance. The typical capacitance range for varactors ranges from a few picofarads to over 100 pF, with a maximum reverse voltage range from a few volts to close to a hundred volts, depending on device. (Many standard diodes and zener diodes can be used as inexpensive varactor diodes, though the relationship between reverse voltage and capacitance isn’t always as reliable.)

The low capacitance levels provided by a varactor usually limit its use to high-frequency RF circuits, where the applied voltage is used to change the capacitance of an oscillator circuit. The reverse voltage may be applied via a tuning potentiometer, which acts to change the overall frequency of an oscillator, or it may be applied by a modulating signal (e.g., audio signal), which acts to FM-modulate the oscillator’s high-frequency carrier. See the examples that follow.

When designing with varactor diodes, the reverse-bias voltage must be abso-lutely free of noise, since any variation in the bias voltage will cause changes in capacitance. Unwanted frequency shifts or instability will result if the reverse-bias voltage is noisy. Filter capacitors are used to limit such noise.

Varactors come in both single and dual forms. The dual varactor configuration contains two varactors in series-opposing configuration, with common anodes and separate cathodes. In this configuration, the varactors acts as series capacitors that change capacitance levels together when a voltage is applied to the common anode lead. See Fig. 4.39.

FM Modulator



FM modulation: FM (frequency modulation) is produced when the frequency of a carrier is changed instantaneously according to the magnitude of an applied modulating signal. (The frequency of the carrier is usually in the megahertz, while the modulating signal is typically in the hertz to kilohertz range, e.g., audio modulating radio signal.) One way to produce FM is to use a voltage-controlled oscillator. The oscillator will have an output frequency proportional to the modulating signal’s amplitude. As the amplitude of the modulating signal increases, the frequency of the carrier increases. Here a Colpitts LC oscillator uses a varactor diode in place of one of its regulator capacitors that form the tuned circuit. The modulating voltage is applied across the diode and changes the diode’s capacitance in proportion. This causes the oscillator frequency to change, thus generating FM in the process. L2 (RFC) is a radiofrequency choke that prevents high-frequency signals from feeding back into the modulating source. C3 and C4 are ac-coupling capacitors. The rest of the components go into making the Colpitts oscillator.

Oscillator with Pot-Controlled Varactor Tuning



Unlike the preceding circuit, this circuit acts simply as a variable high-frequency oscillator, whose frequency is varied via a potentiometer (R1). The voltage from the pot is applied to a dual varactor diode D1 through a low-frequency filter (C1, R2) to ensure that the varactor bias is clean dc. This alters the effective capacitance of the D1-L1 tuned circuit, which changes the frequency of the entire oscillator. C2 and C6 are dc-blocking (ac-coupling) capacitors. Q1 is an N-channel JFET in common drain configuration with feedback to the gate through C3. R3 is the gate bias resistor. R4 is the drain voltage resistor with filter capacitor C5.

4.2.9 PIN Diodes

PIN diodes are used as RF and microwave switches. To high-frequency signals, the PIN diode acts like a variable resistor whose value is controlled by an applied dc forward-bias current. With a high dc forward bias, the resistance is often less than an ohm. But with a small forward bias, the resistance appears very large (kiloohms) to high-frequency signals. PIN diodes are constructed with a layer of intrinsic (undoped) semiconductor placed between very highly doped p-type and n-type material, creating a PIN junction.

In terms of application, PIN diodes are used primarily as RF and microwave switches—even at high power levels. A common application is their use as transmit/receive switches in transceivers operating from 100 MHz and up. They are also used as photodetectors in fiber-optic systems. For the most part, you’ll never need to use them, unless you are a graduate student in electrical engineering or physics, or are working for a high-tech firm.

RF Switching with PIN Diodes



At RF frequencies, switching is very finicky, requiring special design techniques to minimize signal contamination and degradation. Here are two switching circuits that make use of PIN diodes. In the SPST switch circuit, a signal from a RF generator (VG), can be allowed to pass, or can be prevented from passing to the load by applying a bias voltage to the PIN diode. The RFC is a high-frequency choke to prevent RF from entering bias supply, while the capacitor to ground is used to supply clean dc at the bias input. The SPDT switch circuit is similar to the first, but with, of course, two bias inputs.

4.2.10 Microwave Diodes (IMPATT, Gunn, Tunnel, etc.)

There are a number of diodes that you’ll probably never have to use, but they are around nevertheless. These diodes are used for very special purposes at the high-frequency end—microwave and millimeter wave (>20 GHz) range, often in microwave amplifiers and oscillators. Most standard diodes and bipolar transistors usually won’t cut it at such high speeds, due to the relatively slow diffusion or migration of charge carriers across semiconductor junctions. With the tunnel, Gunn, IMPATT, and other diodes, the variable effects that lead to useful alterations in, say, an amplifier’s gain or an oscillator’s resonant frequency involve entirely different physics—physics that allows for alterations at essentially the speed of light. The physics may be electron tunneling (through electrostatic barrier separating p-type and n-type regions, rather than being thermionically emitted over the barrier, as generally occurs in a diode)—tunnel diode. Or it may be due to a negative resistance at forward biasing because of an increase in effective mass (slowing down) of electrons due to complex conduction band symmetry—Gunn diodes. It may also be a negative resistance resulting in electrons moving to higher, less mobile bands, reducing current flow with applied forward bias—IMPATT diodes. Anyway, you get the idea—it’s hairy high-frequency stuff that should probably be left to the experts. (Note: TRAPATT and Baritt diodes are also used in microwave applications.)

4.2.11 Problems

Problem 1: What does this circuit do? What’s the final output voltage? What are the individual voltage drops across each diode with plug tip-positive and plug tip-negative? (Assume each diode has a 0.6 V forward voltage drop.) To prevent diode meltdown, what would be the minimum load resistance, assuming 1N4002 diodes?



Answer: Polarity protection circuit that will output the same polarity regardless of the polarity applied to input. The final output voltage is 11.4 V. Tip-positive: VD1 = 0.6 V, VD2 = 11.4 V, VD3 = 11.4 V, VD4 = 0.6 V; Tip-negative: VD1 = 11.4 V, VD2 = 0.6 V, VD3 = 0.6 V, VD4 = 11.4 V. Load resistance should not drop below 11.4 Ω, assuming 1N4002 diodes, since they have a maximum current rating of 1 A. It’s a good idea to keep the current to around 75 percent of the maximum value for safety, so 15 Ω would be a better limit.

Problem 2: What does the output look like for the circuit to the left in Fig. 4.42? What happens if a load of 2.2K is attached to the output?



Answer: Clamp circuit, where the output is shifted so that it’s practically pure alternating dc, for the exception of a 0.6 V negative dip due to the diode drop. This gives a maximum peak of 27.6 V and a minimum of −0.6 V. (Recall Vpeak = 1.41 × Vrms.) With the load attached, the output level decreases slightly—the capacitor/load resistor acts like a high-pass filter, with a cutoff frequency of 1/(2πRC). In simulation, the output goes to 8.90 V(RMS) or 24.5 peak, −0.6 V minimum.

Problem 3: A 10- to 50-mA load requires a regulated 8.2 V. With a 12 V ± 10 percent power supply and 8.2 V zener diode. What series resistance is required? Assume from the data sheets (or experimentation) that the zener diode’s minimum regulation current is 10 mA. Determine the power ratings for the resistor and zener diode.



Answer: Vin,max = 13.2 V, Vin,min = 10.8 V: RS = (10.8 V − 8.2 V)/(10 mA + 50 mA) = 43 Ω; PR = (13.2 V − 8.2 V)2/(43 Ω) = 0.58 W; PZ = 8.2 V(13.2 V − 8.2 V)/(43 Ω) = 0.95 W. See Fig. 4.29 for details.





p-n Junction


Acts as one-way gate to current-flow, from anode (A) to cathode (C). Comes in silicon and germanium types. Both require a forward-bias voltage to conduct; typically 0.6 to 1.7 V for silicon, and 0.2 to 0.4 V for germanium. Used in rectification, transient suppression, voltage multiplication, RF demodulation, analog logic, clamps, fast switches, and voltage regulation.



Similar in operation to p-n junction diode, but designed with special metal semiconductor junction instead of a p-n junction. This provides for extremely low junction capacitance that stores less charge. Results of this junction yield quicker switching times, useful in fast clamping and high-frequency applications approaching the gigahertz range. Also, generally has a lower forward-bias voltage of around 0.4 V (average)—but can be from 0.15 to 0.9 V or more. Used in similar applications as p-n junction diode, but offers better low-signal level detection, speed, and low-power loss in rectification due to low forward threshold.



Conducts from A to C like p-n junction diode, but will also conduct from C to A if the applied reverse voltage is greater than the zener’s breakdown voltage rating VZ. Acts like a voltage-sensitive control valve. Comes with various breakdown voltages—1.2 V, 3.0 V, 5.1 V, 6.3 V, 9 V, 12 V, etc., and power ratings. Applications include voltage regulation, waveform clipping, voltage shifting, and transient suppression.

LED & Laser


Light-emitting diode (LED) emits a near constant wavelength of light when forward-biased (A > C) by a voltage of about 1.7 V. Comes in various wavelengths (IR through visible), sizes, power ratings, etc. Used as indicator and emitting source in IR and light-wave communications. Laser diode is similar to LED, but provides a much narrower wavelength spectrum (about 1 nm compared to around 40 nm for LED), usually in the IR region. They have very fast response times (lns). These features provide clean signal characteristics useful in fiber-optic systems, where minimized dispersion effects, efficient coupling, and limited degradation over long distances are important. They are also used in laser pointers, CD/DVD players, bar-code readers, and in various surgical applications.



Generates a current when exposed to light, or can be used to alter current flow passing through it when the light intensity changes. Operates in reverse-bias direction (current flows from C to A) when exposed to light. Current increases with light intensity. Very fast response times (ns). Not as sensitive as phototransistors, but their linearity can make them useful in simple light meters.

Varactor (Varicap)


Acts like a voltage-sensitive variable capacitor, whose capacitance decreases as the reverse-bias voltage on the diode increases. Designed with a junction specifically formulated to have a relatively large range of capacitance values for a modest range of reverse-bias voltages. Capacitance range in the picofarad range, so they are usually limited to RF applications, such as tuning receivers and generating FM.

PIN, IMPATT Gunn, Tunnel, etc.


Most of these are resistance devices used in RF, microwave, and millimeter wave applications (e.g., amplifiers and oscillators). Unique conduction physics yields much faster response times when compared to standard diodes that use charge carrier dispersion across a p-n junction.

4.3 Transistors

Transistors are semiconductor devices that act as either electrically controlled switches or amplifier controls. The beauty of transistors is the way they can control electric current flow in a manner similar to the way a faucet controls the flow of water. With a faucet, the flow of water is controlled by a control knob. With a transistor, a small voltage and/or current applied to a control lead acts to control a larger electric flow through its other two leads.

Transistors are used in almost every electric circuit you can imagine. For example, you find transistors in switching circuits, amplifier circuits, oscillator circuits, current-source circuits, voltage-regulator circuits, power-supply circuits, digital logic ICs, and almost any circuit that uses small control signals to control larger currents.

4.3.1 Introduction to Transistors

Transistors come in a variety of designs and come with unique control and current-flow features. Most transistors have a variable current-control feature, but a few do not. Some transistors are normally off until a voltage is applied to the base or gate, whereas others are normally on until a voltage is applied. (Here, normally refers to the condition when the control lead is open circuited. Also, on can represent a variable amount of current flow.) Some transistors require both a small current and a small voltage applied to their control lead to function, whereas others only require a voltage. Some transistors require a negative voltage and/or output current at their base lead (relative to one of their other two leads) to function, whereas others require a positive voltage and/or input current at their base.

The two major families of transistors are bipolar transistors and field-effect transistors (FETs). The major difference between these two families is that bipolar transistors require a biasing input (or output) current at their control leads, whereas FETs require only a voltage—practically no current. [Physically speaking, bipolar transistors require both positive (holes) and negative (electrons) carriers to operate, whereas FETs only require one charge carrier.] Because FETs draw little or no current, they have high input impedances (∼1014 Ω). This high input impedance means that the FET’s control lead will not have much influence on the current dynamics within the circuit that controls the FET. With a bipolar transistor, the control lead may draw a small amount of current from the control circuit, which then combines with the main current flowing through its other two leads, thus altering the dynamics of the control circuit.

In reality, FETs are definitely more popular in circuit design today than bipolar transistors. Besides drawing essentially zero input-output current at their control leads, they are easier to manufacture, cheaper to make (require less silicon), and can be made extremely small—making them useful elements in integrated circuits. One drawback with FETs is in amplifier circuits, where their transconductance is much lower than that of bipolar transistors at the same current. This means that the voltage gain will not be as large. For simple amplifier circuits, FETs are seldom used unless extremely high input impedances and low input currents are required.

Table 4.4 provides an overview of some of the most popular transistors. Note that the term normally used in this chart refers to conditions where the control lead (e.g., base, gate) is shorted (is at the same potential) with one of its channel leads (e.g., emitter, source). Also, the terms on and off used in this chart are not to be taken too literally; the amount of current flow through a device is usually a variable quantity, set by the magnitude of the control voltage. The transistors described in this chart will be discussed in greater detail later on in this chapter.

TABLE 4.4 Overview of Transistors






Normally off, but a small input current and a small positive voltage at its base (B)—relative to its emitter (E)—turns it on (permits a large collector-emitter current). Operates with VC > VE. Used in switching and amplifying applications.


Normally off, but a small output current and negative voltage at its base (B)—relative to its emitter (E)—turns it on (permits a large emitter-collector current). Operates with VE > VC. Used in switching and amplifying applications.

Junction FET


Normally on, but a small negative voltage at its gate (G)—relative to its source (S)—turns it off (stops a large drain-source current). Operates with VD > VS. Does not require a gate current. Used in switching and amplifying applications.


Normally on, but a small positive voltage at its gate (G)—relative to its source (S)—turns it off (stops a large source-drain current). Operates with VS > VD. Does not require a gate current. Used in switching and amplifying applications.

Metal oxide semiconductor FET (MOSFET) (depletion)


Normally on, but a small negative voltage at its gate (G)—relative to its source (S)—turns it off (stops a large drain-source current). Operates with VD > VS. Does not require a gate current. Used in switching and amplifying applications.


Normally on, but a small positive voltage at its gate (G)—relative to its source (S)—turns it off (stops a large source-drain current). Operates with VS > VD. Does not require a gate current. Used in switching and amplifying applications.

Metal oxide semiconductor FET (MOSFET) (enhancement)


Normally off, but a small positive voltage at its gate (G)—relative to its source (S)—turns it on (permits a large drain-source current). Operates with VD > VS. Does not require a gate current. Used in switching and amplifying applications.


Normally off, but a small negative voltage at its gate (G)—relative to its source (S)—turns it on (permits a large source-drain current). Operates with VS > VD. Does not require a gate current. Used in switching and amplifying applications.

Unijunction FET (UJT)


Normally a very small current flows from base 2 (B2) to base 1 (B1), but a positive voltage at its emitter (E)—relative to B1 or B2—increases current flow. Operates with VB2 > VB1. Does not require a gate current. Only acts as a switch.

4.3.2 Bipolar Transistors

Bipolar transistors are three-terminal devices that act as electrically controlled switches or as amplifier controls. These devices come in either npn or pnp configurations, as shown in Fig. 4.44. An npn bipolar transistor uses a small input current and positive voltage at its base (relative to its emitter) to control a much larger collector-to-emitter current. Conversely, a pnp transistor uses a small output base current and negative base voltage (relative its emitter) to control a larger emitter-to-collector current.



Bipolar transistors are incredibly useful devices. Their ability to control current flow by means of applied control signals makes them essential elements in electrically controlled switching circuits, current-regulator circuits, voltage-regulator circuits, amplifier circuits, oscillator circuits, and memory circuits.

How Bipolar Transistors Work

Here is a simple model of how an npn bipolar transistor works. (For a pnp bipolar transistor, all ingredients, polarities, and currents are reversed.)


An npn bipolar transistor is made by sandwiching a thin slice of p semiconductor between two n-type semiconductors. When no voltage is applied at the transistor’s base, electrons in the emitter are prevented from passing to the collector side because of the p-n junction. (Remember that for electrons to flow across a p-n junction, a biasing voltage is needed to give the electrons enough energy to “escape” the atomic forces holding them to the n side.) Notice that if a negative voltage is applied to the base, things get even worse—the p-n junction between the base and emitter becomes reverse-biased. As a result, a depletion region forms and prevents current flow.


If a positive voltage (of at least 0.6 V) is applied to the base of an npn transistor, the pn junction between the base and emitter is forward-biased. During forward bias, escaping electrons are drawn to the positive base. Some electrons exit through the base, but—this is the trick—because the p-type base is so thin, the onslaught of electrons that leave the emitter get close enough to the collector side that they begin jumping into the collector. Increasing the base voltage increases this jumping effect and hence increases the emitter-to-collector electron flow. Remember that conventional currents are moving in the opposite direction to the electron flow. Thus, in terms of conventional currents, a positive voltage and input current applied at the base cause a “positive” current I to flow from the collector to the emitter.



Figure 4.46 shows a typical characteristic curve for a bipolar transistor. This characteristic curve describes the effects the base current IB and the emitter-to-collector voltage VEC have on the emitter/collector currents IE and IC. (As you will see in a second, IC is practically equal to IE.)



Some important terms used to describe a transistor’s operation include saturation region, cutoff region, active mode/region, bias, and quiescent point (Q-point). Saturation region refers to a region of operation where maximum collector current flows and the transistor acts much like a closed switch from collector to emitter. Cutoff region refers to the region of operation near the voltage axis of the collector characteristics graph, where the transistor acts like an open switch—only a very small leakage current flows in this mode of operation. Active mode/region describes transistor operation in the region to the right of saturation and above cutoff, where a near-linear relationship exists between terminal currents (IB, IC, IE). Bias refers to the specific dc terminal voltages and current of the transistor to set a desired point of active-mode operation, called the quiescent point, or Q-point.

The Formulas

The fundamental formula used to describe the behavior of a bipolar transistor (at least within the active region) is

IC = hFEIB = βIB


Rule 1 For an npn transistor, the voltage at the collector VC must be greater than the voltage at the emitter VE by at least a few tenths of a volt; otherwise, current will not flow through the collector-emitter junction, no matter what the applied voltage is at the base. For pnp transistors, the emitter voltage must be greater than the collector voltage by a similar amount.

Rule 2 For an npn transistor, there is a voltage drop from the base to the emitter of 0.6 V. For a pnp transistor, there is a 0.6-V rise from base to emitter. In terms of operation, this means that the base voltage VB of an npn transistor must be at least 0.6 V greater than the emitter voltage VE; otherwise, the transistor will not pass an collector-to-emitter current. For a pnp transistor, VB must be at least 0.6 V less than VE; otherwise, it will not pass a collector-to-emitter current.

where IB is the base current, IC is the collector current, and hFE (also referred to as β) is the current gain. Every transistor has its own unique hFE. The hFE of a transistor is often taken to be a constant, typically around 10 to 500, but it may change slightly with temperature and with changes in collector-to-emitter voltage. (A transistor’s hFE is given in transistor spec tables.) A simple explanation of what the current-gain formula tells you is this: If you take a bipolar transistor with, say, an hFE of 100 and then feed (npn) or sink (pnp) a 1-mA current into (npn) or out of (pnp) its base, a collector current of 100 mA will result. Now, it is important to note that the current-gain formula applies only if rules 1 and 2 are met, i.e., assuming the transistor is within the active region. Also, there is a limit to how much current can flow through a transistor’s terminals and a limit to the size of voltage that can be applied across them. We will discuss these limits later in the chapter (Fig. 4.47).

Now, if you apply the law of conservation of current (follow the arrows in Fig. 4.47), you get the following useful expression relating the emitter, collector, and base currents:

IE = IC + IB



If you combine this equation with the current-gain equation, you can come up with an equation relating the emitter and base currents:

IE = (hFE + 1)IB

As you can see, this equation is almost identical to the current-gain equation (IC = hFEIB), with the exception of the +1 term. In practice, the +1 is insignificant as long as hFE is large (which is almost always the case). This means that you can make the following approximation:


Finally, the second equation below is simply rule 2 expressed in mathematical form:

VBE = VBVE = + 0.6 V (npn)

VBE = VBVE = − 0.6 V (pnp)

Figure 4.47 shows how all the terminal currents and voltages are related. In the figure, notice that the collector voltage has a question mark next to it. As it turns out, the value of VC cannot be determined directly by applying the formulas. Instead, VC’s value depends on the network that is connected to it. For example, if you consider the setup shown in Fig. 4.48, you must find the voltage drop across the resistor in order to find the collector voltage. By applying Ohm’s law and using the current-gain relation, you can calculate VC. The results are shown in the figure.



It is important to note that the equations here are idealistic in form. In reality, these equations may result in “unreal” answers. For instance, they tend to “screw up” when the currents and voltages are not within the bounds provided by the characteristic curves. If you apply the equations blindly, without considering the operating characteristics, you could end up with some wild results that are physically impossible.

One final note with regard to bipolar transistor theory involves what is called transresistance rtr. Transresistance represents a small resistance that is inherently present within the emitter junction region of a transistor. Two things that determine the transresistance of a transistor are temperature and emitter current flow. The following equation provides a rough approximation of the rtr:

In many cases, rtr is insignificantly small (usually well below 1000 Ω) and does not pose a major threat to the overall operation of a circuit. However, in certain types of circuits, treating rtr as being insignificant will not do. In fact, its presence may be the major factor determining the overall behavior of a circuit. We will take a closer look at transresistance later on in this chapter.

Here are a couple of problems that should help explain how the equations work. The first example deals with an npn transistor; the second deals with a pnp transistor.

EXAMPLE 1 Given VCC = +20 V, VB = 5.6 V, R1 = 4.7 kΩ, R2 = 3.3 kΩ, and hFE = 100, find VE, IE, IB, IC, and VC.



EXAMPLE 2 Given VCC = + 10 V, VB = 8.2 V, R1 = 560 Ω, R2 = 2.8 kΩ, and hFE = 100, find VE, IE, IB, IC, and VC.



Bipolar Transistor Water Analogy


The base of the npn water transistor is represented by the smaller tube entering the main device from the left side. The collector is represented by the upper portion of the vertical tube, while the emitter is represented by the lower portion of the vertical tube. When no pressure or current is applied through the “base” tube (analogous to an npn transistor’s base being open circuited), the lower lever arm remains vertical while the top of this arm holds the upper main door shut. This state is analogous to a real bipolar npn transistor off state. In the water analogy, when a small current and pressure are applied to the base tube, the vertical lever is pushed by the entering current and swings counterclockwise. When this lever arm swings, the upper main door is permitted to swing open a certain amount that is dependent on the amount of swing of the lever arm. In this state, water can make its way from the collector tube to the emitter tube, provided there is enough pressure to overcome the force of the spring holding the door shut. This spring force is analogous to the 0.6 V biasing voltage needed to allow current through the collector-emitter channel. Notice that in this analogy, the small base water current combines with the collector current.


The main feature to note here is the need for a lower pressure at the base for the pnp water transistor to turn on. By allowing current to flow out the base tube, the lever moves, allowing the emitter-collector door to open. The degree of openness varies with the amount of swing in the lever arm, which corresponds to the amount of current escaping through the base tube. Again, note the 0.6 V biasing spring.


Basic Operation




Here, an npn transistor is used to control current flow through a light bulb. When the switch is thrown to the on position, the transistor is properly biased, and the collector-to-emitter channel opens, allowing current to flow from VCC through the light bulb and into ground. The amount of current entering the base is determined by

To find the collector current, you can use the current-gain relation (IC = hFEIB), provided that there is not too large a voltage drop across the light bulb (it shouldn’t cause VC to drop below 0.6 V + VE). When the switch is thrown to the off position, the base is set to ground, and the transistor turns off, cutting current flow to the light bulb. R2 should be large (e.g., 10 kΩ) so that very little current flows to ground.

In the pnp circuit, everything is reversed; current must leave the base in order for a collector current to flow.




The circuit here shows how an npn transistor can be used to make a simple current source. By applying a small input voltage and current at the transistor’s base, a larger collector/load current can be controlled. The collector/load current is related to the base voltage by

The derivation of this equation is shown with the figure.




Two common methods for biasing a current source are to use either a voltage-divider circuit (shown in the leftmost circuit) or a zener diode regulator (shown in the rightmost circuit). In the voltage-divider circuit, the base voltage is set by R1 and R2 and is equal to

In the zener diode circuit, the base voltage is set by the zener diode’s breakdown voltage such that

VB = Vzener




The network shown here is called an emitter follower. In this circuit, the output voltage (tapped at the emitter) is almost a mirror image of the input (output “follows” input), with the exception of a 0.6 V drop in the output relative to the input (caused by base-emitter pn junction). Also, whenever VB ≤ 0.6 V (during negative swings in input), the transistor will turn off (the pn junction is reversed-biased). This effect results in clipping of the output (see graph). At first glance, it may appear that the emitter follower is useless—it has no voltage gain. However, if you look at the circuit more closely, you will see that it has a much larger input impedance than an output impedance, or more precisely, it has a much larger output current (IE) relative to an input current (IB). In other words, the emitter follower has current gain, a feature that is just as important in applications as voltage gain. This means that this circuit requires less power from the signal source (applied to Vin) to drive a load than would otherwise be required if the load were to be powered directly by the source. By manipulating the transistor gain equation and using Ohm’s law, the input resistance and output resistance are:




The circuit shown here is called a common-collector amplifier, which has current gain but no voltage gain. It makes use of the emitter-follower arrangement but is modified to avoid clipping during negative input swings. The voltage divider (R1 and R2) is used to give the input signal (after passing through the capacitor) a positive dc level or operating point (known as the quiescent point). Both the input and output capacitors are included so that an ac input-output signal can be added without disturbing the dc operating point. The capacitors, as you will see, also act as filtering elements.

To design a common-collector amplifier used to power a 3-kΩ load, which has a supply voltage VCC = +10 V, a transistor hFE of 100, and a desired f3dB point of 100 Hz, you

1. Choose a quiescent current IQ = IC. For this problem, pick IQ = 1 mA.

2. Next, select VE = ½VCC to allow for the largest possible symmetric output swing without clipping, which in this case, is 5 V. To set VE = 5 V and still get IQ = 1 mA, make use of RE, whose value you find by applying Ohm’s law:

3. Next, set the VB = VE + 0.6 V for quiescent conditions (to match up VE so as to avoid clipping). To set the base voltage, use the voltage divider (R1 and R2). The ratio between R1 and R2 is determined by rearranging the voltage-divider relation and substituting into it VB = VE + 0.6 V:

Fortunately, you can make an approximation and simply let R1 = R2. This approximation “forgets” the 0.6-V drop but usually isn’t too dramatic. The actual sizes of R2 and R1 should be such that their parallel resistance is less than or equal to one-tenth the dc (quiescent) input resistance at the base (this prevents the voltage divider’s output voltage from lowering under loading conditions):

(using the approximation R = R1 = R2)

Here, Rin(base),dc = hFERE, or specially, Rin(base),dc = 100(5 k) = 500 k. Using the approximation above, R1 and R2 are calculated to be 100 k each. (Here you did not have to worry about the ac coupler load; it did not influence the voltage divider because you assumed quiescent setup conditions; C2 acts as an open circuit, hence “eliminating” the presence of the load.)

4. Next, choose the ac coupling capacitors so as to block out dc levels and other undesired frequencies. C1 forms a high-pass filter with Rin (see diagram). To find Rin, treat the voltage divider and Rin(base),ac as being in parallel:

Notice that Rin(base),ac is used, not Rin(base),dc. This is so because you can no longer treat the load as being absent when fluctuating signals are applied to the input; the capacitor begins to pass a displacement current. This means that you must take RE and Rload in parallel and multiply by hFE to find Rin(base,ac):

Now you can find Rin:

Rin = 40 kΩ

Once you have found Rin, choose C1 to set the f3dB point (C1 and Rin form a high-pass filter.) The capacitor value C1 is found by using the following formula:

C2 forms a high-pass filter with the load. It is chosen by using




The transistor configuration here is referred to as the common-emitter configuration. Unlike the emitter follower, the common emitter has voltage gain. To figure out how this circuit works, first set VC = ½VCC to allow for maximum swing without clipping. Like the emitter follower, again pick a quiescent current IQ to start with. To set VC = ½VCC with a desired IQ, use RC, which is found by Ohm’s law:

For example, if VCC is 10 V and IQ is 0.5 mA, RC is then 10 k. The gain of this circuit is found by realizing that ΔVE = ΔVB (where Δ represents a small fluctuation). The emitter current is found using Ohm’s law:

Using VC = VCC ICRC and the last expression, you get

Since VC is Vout and VB is Vin, the gain is

But what about RE? According to the circuit, there’s no emitter resistor. If you use the gain formula, it would appear that RE = 0 Ω, making the gain infinite. However, as mentioned earlier, bipolar transistors have a transresistance (small internal resistance) in the emitter region, which is approximated by using

Applying this formula to the example, taking IQ = 0.5 mA = IC IE, the RE term in the gain equation, or rtr equals 52 Ω. This means the gain is actually equal to

Notice that the gain is negative (output is inverted). This results in the fact that as Vin increases, IC increases, while VC (Vout) decreases (Ohm’s law). Now, there is one problem with this circuit. The rtr term happens to be very unstable, which in effect makes the gain unstable. The instability stems from rtr dependence on temperature. As the temperature rises,

VE and IC increase, VBE decreases, but VB remains fixed. This means that the biasing-voltage range narrows, which in effect turns the transistor’s “valve” off. To eliminate this pinch, an emitter resistor is placed from emitter to ground (see second circuit). Treating RE and rtr as series resistors, the gain becomes

By adding RE, variations in the denominator are reduced, and therefore, the variations in gain are reduced as well. In practice, RE should be chosen to place VE around 1 V (for temperature stability and maximum swing in output). Using Ohm’s law (and applying it to the example), choose RE = VE/IE = VE/IQ = 1 V/1 mA = 1 k. One drawback that arises when RE is added to the circuit is a reduction in gain. However, there is a trick you can use to eliminate this reduction in voltage gain and at the same time maintain the temperature stability. If you bypass RE with a capacitor (see third circuit), you can make RE “disappear” when high-frequency input signals are applied. (Recall that a capacitor behaves like an infinitely large resistor to dc signals but becomes less “resistive,” or reactive to ac signals.) In terms of the gain equation, the RE term goes to zero because the capacitor diverts current away from it toward ground. The only resistance left in the gain equation is rtr.




The circuit shown here is known as a common-emitter amplifier. Unlike the common-collector amplifier, the common-emitter amplifier provides voltage gain. This amplifier makes use of the common-emitter arrangement and is modified to allow for ac coupling. To understand how the amplifier works, let’s go through the following example.

To design a common-emitter amplifier with a voltage gain of −100, an f3dB point of 100 Hz, and a quiescent current IQ = 1 mA, where hFE = 100 and VCC = 20 V:

1. Choose RC to center Vout (or VC) to ½VCC to allow for maximum symmetrical swings in the output. In this example, this means VC should be set to 10 V. Using Ohm’s law, you find RC:

2. Next we select RE to set VE = 1 V for temperature stability. Using Ohm’s law, and taking IQ = IE = 1 mA, we get RE = VE/IE = 1 V/1 mA = 1 kΩ.

3. Now, choose R1 and R2 to set the voltage divider to establish the quiescent base voltage of VB = VE + 0.6 V, or 1.6 V. To find the proper ratio between R1 and R2, use the voltage divider (rearranged a bit):

This means R1 = 11.5R2. The size of these resistors is found using the similar procedure you used for the common-collector amplifier; their parallel resistance should be less than or equal to 1/10Rin(base),dc.

After plugging R1 = 11.5R2 into this expression and using Rin(base),dc = hFERE, you find that R2 = 10 kΩ, which in turn means R1 = 115 kΩ (let’s say, 110 kΩ is close enough for R1).

4. Next, choose R3 for the desired gain, where

(The double line means to take RE and R3 in parallel.) To find rtr, use rtr = 0.026 V/IE = 0.026 V/IC = 0.026 V/1 mA = 26 Ω. Now you can simplify the gain expression by assuming RE “disappears” when ac signals are applied. This means the gain can be simplified to

Solving this equation for R3, you get R3 = 74 Ω.

5. Next, choose C1 for filtering purposes such that C1 = 1⁄(2πf3dBRin). Here, Rin is the combined parallel resistance of the voltage-divider resistors, and Rin(base),ac looking in from the left into the voltage divider:

Solving this equation, you get Rin = 5 k Ω. This means

6. To choose C2, treat C2 and rtr + R3 as a high-pass filter (again, treat RE as being negligible during ac conditions). C2 is given by






The zener diode circuit here can be used to make a simple voltage regulator. However, in many applications, the simple regulator has problems; Vout isn’t adjustable to a precise value, and the zener diode provides only moderate protection against ripple voltages. Also, the simple zener diode regulator does not work particularly well when the load impedance varies. Accommodating large load variations requires a zener diode with a large power rating—which can be costly.

The second circuit in the figure, unlike the first circuit, does a better job of regulating; it provides regulation with load variations and provides high-current output and somewhat better stability. This circuit closely resembles the preceding circuit, except that the zener diode is connected to the base of an npn transistor and is used to regulate the collect-to-emitter current. The transistor is configured in the emitter-follower configuration. This means that the emitter will follow the base (except there is the 0.6-V drop). Using a zener diode to regulate the base voltage results in a regulated emitter voltage. According to the transistor rules, the current required by the base is only 1/hFE times the emitter-to-collector current. Therefore, a low-power zener diode can regulate the base voltage of of a transistor that can pass a considerable amount of current. The capacitor is added to reduce the noise from the zener diode and also forms an RC filter with the resistor that is used to reduce ripple voltages.

In some instances, the preceding zener diode circuit may not be able to supply enough base current. One way to fix this problem is to add a second transistor, as shown in the third circuit. The extra transistor (the one whose base is connected to the zener diode) acts to amplify current sent to the base of the upper transistor.




By attaching two transistors together as shown here, a larger current-handling, larger hFE equivalent transistor circuit is formed. The combination is referred to as a Darlington pair. The equivalent hFE for the pair is equal to the product of the individual transistor’s hFE values (hFE = hFE1hFE2). Darlington pairs are used for large current applications and as input stages for amplifiers, where big input impedances are required. Unlike single transistors, however, Darlington pairs have slower response times (it takes longer for the top transistor to turn the lower transistor on and off) and have twice the base-to-emitter voltage drop (1.2 V instead of 0.6 V) as compared with single transistors. Darlington pairs can be purchased in single packages.

Types of Bipolar Transistors



This type of transistor is used to amplify low-level signals but also can be used as a switch. Typical hFE values range from 10 to 500, with maximum IC ratings from about 80 to 600 mA. They come in both npn and pnp forms. Maximum operating frequencies range from about 1 to 300 MHz.



These transistors are used primarily as switches but also can be used as amplifiers. Typical hFE values range from around 10 to 200, with maximum IC ratings from around 10 to 1000 mA. They come in both npn and pnp forms. Maximum switching rates range between 10 and 2000 MHz.



These transistors are used for small signals that run at high frequencies for high-speed switching applications. The base region is very thin, and the actual chip is very small. They are used in HF, VHF, UHF, CATV, and MATV amplifier and oscillator applications. They have a maximum frequency rating of around 2000 MHz and maximum IC currents from 10 to 600 mA. They come in both npn and pnp forms.



These transistors are used in high-power amplifiers and power supplies. The collector is connected to a metal base that acts as a heat sink. Typical power ratings range from around 10 to 300 W, with frequency ratings from about 1 to 100 MHz. Maximum IC values range between 1 to 100 A. They come in npn, pnp, and Darlington (npn or pnp) forms.



These are two transistors in one. They provide more stability at high current levels. The effective hFE for the device is much larger than that of a single transistor, hence allowing for a larger current gain. They come in npn (D-npn) and pnp (D-pnp) Darlington packages.



This transistor acts as a light-sensitive bipolar transistor (base is exposed to light). When light comes in contact with the base region, a base current results. Depending on the type of phototransistor, the light may act exclusively as a biasing agent (two-lead phototransistor) or may simply alter an already present base current (three-lead phototransistor). See Chap. 5 for more details.



This consists of a number of transistors combined into a single integrated package. For example, the transistor array shown here is made of three npn transistors and two pnp transistors.


Important Things to Know about Bipolar Transistors

The current gain of a transistor (hFE) is not a very good parameter to go by. It can vary from, say, 50 to 500 within same transistor group family and varies with changes in collector current, collector-to-emitter voltage, and temperature. Because hFE is somewhat unpredictable, one should avoid building circuits that depend specifically on hFE values.

All transistors have maximum collector-current ratings (IC,max), maximum collector-to-base (BVCBO), collector-to-emitter (BVCEO), and emitter-to-base (VEBO) breakdown voltages, and maximum collector power dissipation (PD) ratings. If these ratings are exceeded, the transistor may get zapped. One method to safeguard against BVEB is to place a diode from the emitter to the base, as shown in Fig. 4.62a. The diode prevents emitter-to-base conduction whenever the emitter becomes more positive than the base (e.g., input at base swings negative while emitter is grounded). To avoid exceeding BVCBO, a diode placed in series with the collector (Fig. 4.62b) can be used to prevent collector-base conduction from occurring when the base voltage becomes excessively larger than the collector voltage. To prevent exceeding BVCEO, which may be an issue if the collector holds an inductive load, a diode placed in parallel with the load (see Fig. 4.62c) will go into conduction before a collector-voltage spike, created by the inductive load, reaches the breakdown voltage.



Pinouts for Bipolar Transistors

Bipolar transistors come in a variety of different package types. Some transistors come with plastic housings; others come with metal can-like housings. When attempting to isolate the leads that correspond to the base, emitter, and collector terminals, first check to see if the package that housed the transistor has a pinout diagram. If no pinout diagram is provided, a good cross-reference catalog (e.g., NTE Cross-Reference Catalog for Semiconductors) can be used. However, as is often the case, simple switching transistors that come in bulk cannot be “looked up”—they may not have a label. Also, these bulk suppliers often will throw together a bunch of transistors that all look alike but may have entirely different pinout designations and may include both pnp and npn polarities. If you anticipate using transistors often, it may be in your best interest to purchase a digital multimeter that comes with a transistor tester. These multimeters are relatively inexpensive and are easy to use. Such a meter comes with a number of breadboard-like slots. To test a transistor, the pins of the transistor are placed into the slots. By simply pressing a button, the multimeter then tests the transistor and displays whether the device is an npn or pnp transistor, provides you with the pinout designations (e.g., “ebc,” “cbe,” etc.), and will give you the transistor’s hFE.





Here, an npn transistor is used to control a relay. When the transistor’s base receives a control voltage/current, the transistor will turn on, allowing current to flow through the relay coil and causing the relay to switch states. The diode is used to eliminate voltage spikes created by the relay’s coil. The relay must be chosen according to the proper voltage rating, etc.




The differential amplifier shown here is a device that compares two separate input signals, takes the difference between them, and then amplifies this difference. To understand how the circuit works, treat both transistors as identical, and then notice that both transistors are set in the common-emitter configuration. Now, if you apply identical input signals to both V1 and V2, identical currents flow through each transistor. This means that (by using VC = VCC ICRC) both transistors’ collector voltages are the same. Since the output terminals are simply the left and right transistors’ collector voltages, the output voltage (potential difference) is zero. Now, assume the signals applied to the inputs are different, say V1 is larger than V2. In this case, the current flow through the left transistor will be larger than the current flow through the right transistor. This means that the left transistor’s VC will decrease relative to the right transistor’s VC. Because the transistors are set in the common-emitter configuration, the effect is amplified. The relationship between the input and output voltages is given by

Rearranging this expression, you find that the gain is equal to RC/rtr.

Understanding what resistor values to choose can be explained by examining the circuit shown here. First, choose RC to center VC to ½VCC, or 5 V, to maximize the dynamic range. At the same time, you must choose a quiescent current (when no signals are applied), say, IQ = IC = 50 μA. By Ohm’s law RC = (10V − 5V)/50 μA = 100 kΩ. RE is chosen to set the transistor’s emitters as close to 0 V as possible. RE is found by adding both the right and left branch’s 50 μA and taking the sum to be the current flow through it, which is 100 μA. Now, apply Ohm’s law: RE = 0 V − 10 V/100 μA = 100 kΩ. Next, find the transresistance: rtr ≈ 0.026 V/IE = 0.026 V/50 μA = 520 Ω. The gain then is equal to 100 kΩ/520 Ω = 192.

In terms of applications, differential amplifiers can be used to extract a signal that has become weak and which has picked up considerable noise during transmission through a cable (differential amplifier is placed at the receiving end). Unlike a filter circuit, which can only extract a signal from noise if the noise frequency and signal frequency are different, a differential amplifier does not require this condition. The only requirement is that the noise be common in both wires.

When dealing with differential amplifiers, the term common-mode rejection ratio (CMRR) is frequently used to describe the quality of the amplifier. A good differential amplifier has a high CMRR (theoretically infinite). CMRR is the ratio of the voltage that must be applied at the two inputs in parallel (V1 and V2) to the difference voltage (V1 V2) for the output to have the same magnitude.




Recall that an npn emitter follower acts to clip the output during negative swings in the input (the transistor turns off when VB VE + 0.6 V). Likewise, a pnp follower will clip the output during positive input swings. But now, if you combine an npn and pnp transistor, as shown in the circuit shown here, you get what is called a push-pull follower, or complementary-symmetry amplifier, an amplifier that provides current gain and that is capable of conducting during both positive and negative input swings. For Vin = 0 V, both transistors are biased to cutoff (IB = 0). For Vin > 0 V, the upper transistor conducts and behaves like an emitter follower, while the lower transistor is cut off. For Vin < 0 V the lower transistor conducts, while the upper transistor is cut off. In addition to being useful as a dc amplifier, this circuit also conserves power because the operating point for both transistors is near IC = 0. However, at IC = 0, the characteristics of hFE and rtr are not very constant, so the circuit is not very linear for small signals or for near-zero crossing points of large signals (crossover distortion occurs).




Here, two matched pnp transistors can be used to make what is called a current mirror. In this circuit, the load current is a “mirror image” of the control current that is sunk out of the leftmost transistor’s collector. Since the same amount of biasing current leaves both transistors’ bases, it follows that both transistors’ collector-to-emitter currents should be the same. The control current can be set by, say, a resistor connected from the collector to a lower potential. Current mirrors can be made with npn transistors, too. However, you must flip this circuit upside down, replace the pnp transistors with npn transistors, reverse current directions, and swap the supply voltage with ground.




The circuit here is an expanded version of the previous circuit, which is used to supply a “mirror image” of control current to a number of different loads. (Again, you can design such a circuit with npn transistor, too, taking into consideration what was mentioned in the last example.) Note the addition of an extra transistor in the control side of the circuit. This transistor is included to help prevent one transistor that saturates (e.g., its load is removed) from stealing current from the common base line and hence reducing the other output currents.




A. A bistable multivibrator is a circuit that is designed to remain in either of two states indefinitely until a control signal is applied that causes it to change states. After the circuit switches states, another signal is required to switch it back to its previous state. To understand how this circuit works, initially assume that V1 = 0V. This means that the transistor Q2 has no base current and hence no collector current. Therefore, current that flows through R4 and R3flows into the base of transistor Q1, driving it into saturation. In the saturation state, V1 = 0, as assumed initially. Now, because the circuit is symmetric, you can say it is equally stable with V2 = 0 and Q1 saturated. The bistable multivibrator can be made to switch from one state to another by simply grounding either V1 or V2 as needed. This is accomplished with switch S1. Bistable multivibrators can be used as memory devices or as frequency dividers, since alternate pulses restore the circuit to its initial state.

B. A monostable multivibrator is a circuit that is stable in only one state, in this case Vout = 0V. It can be thrown into its unstable state (Vout = VCC) by applying an external trigger signal, but it will automatically return to its stable state after a duration set by the RTCT network. Here when a negative trigger pulse is applied at the input, the fast decaying edge of the pulse will pass through capacitor C1 to the base of Q1 via blocking diode D1 turning Q1 ON. The collector of Q1 which was previously at VCC quickly drops to below zero voltage, effectively giving CT a reverse charge of −0.6V across its plates. Transistor Q2 now has a minus base voltage at its base, holding the transistor fully OFF. This represents the circuit’s unstable state (VOUT = VCC). CT begins to discharge this −0.6V through RT, attempting to charge up to VCC. This negative voltage at Q2’s base begins to decrease gradually at a rate set by the time constant of RTCT. As Q2’s base voltage increases up to VCC, it begins to conduct, causing Q1 to turn off again. The system returns to its original stable state.

C. An asatable multivibrator is a circuit that is not stable in either of two possible output states and it acts like an oscillator. It also requires no external trigger pulse, but uses positive feedback network and a RC timer network to create built-in triggering that switches the output between VCC and 0V. The result is a square wave frequency generator. In the circuit to the left, Q1 and Q2 are switching transistors connected in cross-coupled feedback network, along with two time-delay capacitors. The transistors are biased for linear operation and are operated as common emitter amplifiers with 100% positive feedback. When Q1 is OFF, its collector voltage rises toward VCC, while Q2 if ON. As this occurs, plate A of capacitor C1 rises towards VCC. Capacitor C1’s other plate B, which is connected to the base of Q2 is at 0.6V since Q2 is in conducting state, thus the voltage across C1 is 6.0 − 0.6V = 5.4V. (It’s high value of charge). The instant Q1 switches ON, plate A of C1 falls to 0.6V, causing an equal and instantaneous fall in voltage on plate B of C1. C1 is pulled down to −5.4 (reverse charge) and this negative voltage turns transistor Q2 hard OFF (one unstable state). C1 now begins to charge in the opposite direction via R3 toward the +6V supply rail, and Q2’s base voltage increases toward VCC with a time constant C1 R3. However, when Q2’s base voltage reaches 0.6V, Q2 turns fully ON, starting the whole process over again, but now with C2 taking the base of Q1 to −5.4V while charging up via resistor R2 and entering the second unstable state. This process repeats over and over again as long as the supply voltage is present. The amplitude of the output is approximately VCC with the time period between states determined by the RC network connected across the base terminals of the transistors. It’s possible to drive low impedance loads (or current loads) such as LEDs, speakers, etc. without affecting the operation of the astable multivibrator by introducing another transistor into the circuit—as shown in figure C.



FIGURE 4.69 The two circuits here form logic gates. The OR circuit allows the output (C) to swing to a high voltage when either A or B or both A and B are high. In other words, as long as at least one of the transistors is biased (turned on), a high voltage will appear at the output. In the AND gate circuit, both A and B must be high in order for C to go high. In other words, both transistors must be biased (turned on) for a high voltage to appear at the output.

4.3.3 Junction Field-Effect Transistors

Junction field-effect transistors (JFETs) are three-lead semiconductive devices that are used as electrically controlled switches, amplifier controls, and voltage-controlled resistors. Unlike bipolar transistors, JFETs are exclusively voltage-controlled—they do not require a biasing current. Another unique trait of a JFET is that it is normally on when there is no voltage difference between its gate and source leads. However, if a voltage difference forms between these leads, the JFET becomes more resistive to current flow (less current will flow through the drain-source leads). For this reason, JFETs are referred to as depletion devices, unlike bipolar transistors, which are en-hancement devices (bipolar transistors become less resistive when a current/voltage is applied to their base leads).



JFETs come in either n-channel or p-channel configurations. With an n-channel JFET, a negative voltage applied to its gate (relative to its source lead) reduces current flow from its drain to source lead. (It operates with VD > VS.) With a p-channel JFET, a positive voltage applied to its gate reduces current flow from its source to drain lead. (It operates with VS > VD.) The symbols for both types of JFETs are shown at the left.

An important characteristic of a JFET that is useful in terms of applications is its extremely large input impedance (typically around 1010 Ω). This high input impedance means that the JFET draws little or no input current (lower pA range) and therefore has little or no effect on external components or circuits connected to its gate—no current is drawn away from the control circuit, and no unwanted current enters the control circuit. The ability for a JFET to control current flow while maintaining an extremely high input impedance makes it a useful device used in the construction of bidirectional analog switching circuits, input stages for amplifiers, simple two-terminal current sources, amplifier circuits, oscillators circuits, electronic gain-control logic switches, audio mixing circuits, etc.

How a JFET Works

An n-channel JFET is made with an n-type silicon channel that contains two p-type silicon “bumps” placed on either side. The gate lead is connected to the p-type bumps, while the drain and source leads are connected to either end of the n-type channel (see Fig. 4.71).



When no voltage is applied to the gate of an n-channel JFET, current flows freely through the central n-channel—electrons have no problem going through an n-channel; there are a lot of negative charger carriers already in there just waiting to help out with conduction. However, if the gate is set to a negative voltage—relative to the source—the area in between the p-type semiconductor bumps and the center of the n-channel will form two reverse-biased junctions (one about the upper bump, another about the lower bump). This reverse-biased condition forms a depletion region that extends into the channel. The more negative the gate voltage, the larger is the depletion region, and hence the harder it is for electrons to make it through the channel. For a p-channel JFET, everything is reversed, meaning you replace the negative gate voltage with a positive voltage, replace the n-channel with a p-channel semiconductor, replace the p-type semiconductor bumps with n-type semiconductors, and replace negative charge carriers (electrons) with positive charge carriers (holes).

JFET Water Analogies

Here are water analogies for an n-channel and p-channel JFET. Pretend water flow is conventional current flow and water pressure is voltage.



When no pressure exists between the gate and source of the n-channel water JFET, the device is fully on; water can flow from the drain pipe to the source pipe. To account for a real JFET’s high input impedance, the JFET water analogy uses a plunger mechanism attached to a moving flood gate. (The plunger prevents current from entering the drain source channel, while at the same time it allows a pressure to control the flood gate.) When the gate of the n-channel JFET is made more negative in pressure relative to the source tube, the plunger is forced to the left. This in turn pulls the accordion-like flood gate across the drain-source channel, thus decreasing the current flow.



The p-channel water JFET is similar to the n-channel water JFET, except that all currents and pressures are reversed. The p-channel JFET is fully on until a positive pressure, relative to the source, is applied to the gate tube. The positive pressure forces the accordion across the drain-source channel, hence reducing the current flow.


Technical Stuff

The following graph describes how a typical n-channel JFET works. In particular, the graph describes how the drain current (ID) is influenced by the gate-to-source voltage (VGS) and the drain-to-source voltage (VDS). The graph for a p-channel JFET is similar to that of the n-channel graph, except that ID decreases with an increasing positive VGS. In other words, VGS is positive in voltage, and VDS is negative in voltage.



When the gate voltage VG is set to the same voltage as the source (VGS = VGVS = 0 V), maximum current flows through the JFET. Technically speaking, people call this current (when VGS = 0 V) the drain current for zero bias, or IDSS. IDSS is a constant and varies from JFET to JFET. Now notice how the ID current depends on the drain-source voltage (VDS = VDVS). When VDS is small, the drain current ID varies nearly linearly with VDS (looking at a particular curve for fixed VGS). The region of the graph in which this occurs is called the ohmic region, or linear region. In this region, the JFET behaves like a voltage-controlled resistor.

Now notice the section of the graph were the curves flatten out. This region is called the active region, and here the drain current ID is strongly influenced by the gate-source voltage VGS but hardly at all influenced by the drain-to-source voltage VDS (you have to move up and down between curves to see it).

Another thing to note is the value of VGS that causes the JFET to turn off (point where practically no current flows through device). The particular VGS voltage that causes the JFET to turn off is called the cutoff voltage (sometimes called the pinch-off voltage VP), and it is expressed as VGS,off.

Moving on with the graph analysis, you can see that when VDS increases, there is a point where ID skyrockets. At this point, the JFET loses its ability to resist current because too much voltage is applied across its drain-source terminals. In JFET lingo, this effect is referred to as drain-source breakdown, and the breakdown voltage is expressed as BVDS.

For a typical JFET, IDSS values range from about 1 mA to 1 A, VGS,off values range from around −0.5 to −10 V for an n-channel JFET (or from +0.5 to +10 V for a p-channel JFET), and BVDS values range from about 6 to 50 V.

Like bipolar transistors, JFETs have internal resistance within their channels that varies with drain current and temperature. The reciprocal of this resistance is referred to as the transconductance gm. A typical JFET transconductance is around a few thousand Ω−1, where Ω−1 = 1/Ω or ℧.

Another one of the JFET’s built-in parameters is its on resistance, or RDS,on. This resistance represents the internal resistance of a JFET when in its fully conducting state (when VGS = 0). The RDS,on of a JFET is provided in the specification tables and typically ranges from 10 to 1000 Ω.

Useful Formulas




OHMIC REGION JFET is just beginning to resist. It acts like a variable resistor.

SATURATION REGION JFET is most strongly influenced by gate-source voltage, hardly at all influenced by the drain-source voltage.

CUTOFF VOLTAGE (VGS,OFF) Particular gate-source voltage where JFET acts like an open circuit (channel resistance is at its maximum).

BREAKDOWN VOLTAGE (BVDS) The voltage across the drain and source that causes current to “break through” the JFET’s resistive channel.

DRAIN-CURRENT FOR ZERO BIAS (IDSS) Represents the drain current when gate-source voltage is zero volts (or gate is connected to source, VGS = 0 V).

TRANSCONDUCTANCE (gm) Represents the rate of change in the drain current with the gate-source voltage when the drain-to-source voltage is fixed for a particular VDS. It is analogous to the transconductance (1/Rtr) for bipolar transistors.

DRAIN-CURRENT FOR ZERO BIAS (IDSS) Represents the drain current when gate-source voltage is zero volts (or gate is connected to source, VGS = 0 V).

TRANSCONDUCTANCE (gm) Represents the rate of change in the drain current with the gate-source voltage when the drain-to-source voltage is fixed for a particular VDS. It is analogous to the transconductance (1/Rtr) for bipolar transistors.


Sample Problems




If an n-channel JFET has a IDSS = 8 mA and VGS,off = −4 V, what will be the drain current ID if R = 1 kΩ and VDD = +18 V? Assume that the JFET is in the active region.

In the active region, the drain current is given by

Unfortunately, there is one equation and two unknowns. This means that you have to come up with another equation. Here’s how you get the other equation. First, you can assume the gate voltage is 0 V because it’s grounded. This means that

VGS = VG VS = 0 V − VS = −VS

From this, you can come up with another equation for the drain current by using Ohm’s law and treating ID = IS:

This equation is then combined with the first equation to yield

which simplifies to

VGS2 + 10VGS + 16 = 0

The solutions to this equation are VGS = −2 V and VGS = − 8 V. But a VGS of −8V would be below the cutoff voltage, so we can ignore it, which leaves VGS = −2V as the correct solution. This means that VGS = −2 V is the correct solution, so you disregard the − 8-V solution. Now you substitute VGS back into one of the ID(active) equations to get




If VGS,off = −4 V and IDSS = 12 mA, find the values of ID and gm and RDS when VGS = −2 V and when VGS = +1 V. Assume that the JFET is in the active region.

When VGS = −2 V,

To find gm, you first must find gm0 (transconductance for shorted gate):

Now you can find gm:

To find the drain-source resistance (RDS), use

By applying the same formulas as above, you can find that when VGS = +1 V, ID = 15.6 mA, gm = 0.0075 = 7500 μmhos, and RDS = 133 Ω.

Basic Operations




The two circuits here demonstrate how a JFET acts like a voltage-controlled light dimmer. In the n-channel circuit, a more negative gate voltage causes a larger drain-to-source resistance, hence causing the light bulb to receive less current. In the p-channel circuit, a more positive gate voltage causes a greater source-to-drain resistance.



FIGURE 4.78 A simple current source can be constructed by shorting the source and gate terminals together (this is referred to as self-biasing), as shown in the left-most circuit. This means that VGS = VG VS = 0 V, which means the drain current is simply equal to IDDS. One obvious drawback of this circuit is that the IDDS for a particular JFET is unpredictable (each JFET has its own unique IDDS that is acquired during manufacturing). Also, this source is not adjustable. However, if you place a resistor between the source and ground, as shown in the right-most circuit, you can make the current source adjustable. By increasing RS, you can decrease ID, and vice versa (see Problem 2). Besides being adjustable, this circuit’s ID current will not vary as much as the left circuit for changes in VDS. Though these simple JFET current sources are simple to construct, they are not as stable as a good bipolar or op amp current source.



FIGURE 4.79 The JFET circuit here is called a source follower, which is analogous to the bipolar emitter follower; it provides current gain but not voltage gain. The amplitude of the output signal is found by applying Ohm’s law: VS = RSID, where ID = gmVGS = gm(VG VS). Using these equations, you get

Since VS = Vout and VG = Vin, the gain is simply RSgm/(1 + RSgm). The output impedance, as you saw in Problem 2, is 1/gm. Unlike the emitter follower, the source follower has an extremely larger input impedance and therefore draws practically no input current. However, at the same time, the JFET’s transconductance happens to be smaller than that of a bipolar transistor, meaning the output will be more attenuated. This makes sense if you treat the 1/gm term as being a small internal resistance within the drain-source channel (see rightmost circuit). Also, as the drain current changes due to an applied waveform, gm and therefore the output impedance will vary, resulting in output distortion. Another problem with this follower circuit is that VGS is a poorly controlled parameter (a result of manufacturing), which gives it an unpredictable dc offset.




The source follower circuit from the preceding example had poor linearity and an unpredictable dc offset. However, you can eliminate these problems by using one of the two arrangements shown here. In the far-left circuit, you replace the source resistor with a bipolar current source. The bipolar source acts to fix VGS to a constant value, which in turn eliminates the nonlinearities. To set the dc offset, you adjust R1. (R2 acts like RS in the preceding circuit; it sets the gain.) The near-left circuit uses a JFET current source instead of a bipolar source. Unlike the bipolar circuit, this circuit requires no adjusting and has better temperature stability. The two JFETs used here are matched (matched JFETs can be found in pairs, assembled together within a single package). The lower transistor sinks as much current as needed to make VGS = 0 (shorted gate). This means that both JFETs’ VGS values are zero, making the upper transistor a follower with zero dc offset. Also, since the lower JFET responds directly to the upper JFET, any temperature variations will be compensated. When R1 and R2 are set equal, Vout = Vin. The resistors help give the circuit better ID linearity, allow you to set the drain current to some value other than IDSS, and help to improve the linearity. In terms of applications, JFET followers are often used as input stages to amplifiers, test instruments, or other equipment that is connected to sources with high source impedance.



Recall the emitter-follower and common-emitter bipolar transistor amplifiers from the last chapter. These two amplifiers have JFET counterparts, namely, the source-follower and the common-source amplifier shown here. (The source-follower amplifier provides current gain; the common-source amplifier provides voltage gain.) If you were to set up the equations and do the math, you would find that the gain for the amplifiers would be

(source-follower amp.)

(common-source amp.)

where the transconductance is given by

As with bipolar amplifiers, the resistors are used to set the gate voltages and set the quiescent currents, while the capacitors act as ac couplers/high-pass filters. Notice, however, that both JFET amplifiers only require one self-biasing resistor.


Now, an important question to ask at this point is, Why would you choose a JFET amplifier over a bipolar amplifier? The answer is that a JFET provides increased input impedance and low input current. However, if extremely high input impedances are not required, it is better to use a simple bipolar amplifier or op amp. In fact, bipolar amplifiers have fewer nonlinearity problems, and they tend to have higher gains when compared with JFET amplifiers. This stems from the fact that a JFET has a lower transconductance than a bipolar transistor for the same current. The difference between a bipolar’s transconductance and JFET’s transconductance may be as large as a factor of 100. In turn, this means that a JFET amplifier will have a significantly smaller gain.





According to the graphs you saw earlier, if VDS drops low enough, the JFET will operate within the linear (ohmic) region. In this region, the ID versus VDS curves follow approximate straight lines for VDS smaller than VGS VGS,off. This means that the JFET behaves like a voltage-controlled resistor for small signals of either polarity. For example, if you take a voltage-divider network and replace one of the resistors with a JFET, you get a voltage-controlled voltage divider (see upper left-hand figure). The range over which a JFET behaves like a traditional resistor depends on the particular JFET and is roughly proportional to the amount by which the gate voltage exceeds VGS,off. For a JFET to be effective as a linearly responding resistor, it is important to limit VDS to a value that is small compared with VGS,off, and it is important to keep |VGS| below |VGS,off|. JFETs that are used in this manner are frequently used in electronic gain-control circuits, electronic attenuators, electronically variable filters, and oscillator amplitude-control circuits. A simple electronic gain-control circuit is shown here. The voltage gain for this circuit is given by gain = 1 + RF/RDS(on), where RDS is the drain-source channel resistance. If RF = 29 kΩ and RDS(on) = 1 kΩ, the maximum gain will be 30. As VGS approaches VGS,off, RDS will increase and become very large such that RDS >> RF, causing the gain to decrease to a minimum value close to unity. As you can see, the gain for this circuit can be varied over a 30:1 ratio margin.

Practical Considerations

JFETs typically are grouped into the following categories: small-signal and switching JFETs, high-frequency JFETs, and dual JFETs. Small-signal and switching JFETs are frequently used to couple a high-impedance source with an amplifier or other device such as an oscilloscope. These devices are also used as voltage-controlled switches. High-frequency JFETs are used primarily to amplify high-frequency signals (with the RF range) or are used as high-frequency switches. Dual JFETs contain two matched JFETs in one package. As you saw earlier, dual JFETs can be used to improve source-follower circuit performance.



Like bipolar transistors, JFETs also can be destroyed with excess current and voltage. Make sure that you do not exceed maximum current and breakdown voltages. Table 4.5 is a sample of a JFET specification table designed to give you a feel for what to expect when you start searching for parts.

TABLE 4.5 Portion of a JFET Specification Table






GM TYPICAL (μmhos)



MIN (mA)

MAX (mA)
























Matched-pair n-ch













Here, an n-channel JFET is used to switch a relay. When the switch is set to position A, the JFET is on (gate isn’t properly biased for a depletion effect to occur). Current then passes through the JFET’s drain-source region and through the relay’s coil, causing the relay to switch states. When the switch is thrown to position B, a negative voltage—relative to the source—is set at the gate. This in turn causes the JFET to block current flow from reaching the relay’s coil, thus forcing the relay to switch states.




This circuit uses a JFET—set in the common-source arrangement—to combine (mix) signals from a number of different sources, such as microphones, preamplifiers, etc. All inputs are applied through ac coupling capacitors/filters. The source and drain resistors are used to set the overall amplification, while the 1-MΩ potentiometers are used to control the individual gains of the input signals.




Here, a JFET is used to construct a simple static electricity detector. When the antenna (simple wire) is placed near a charged object, the electrons in the antenna will be drawn toward or away from the JFET’s gate, depending on whether the object is positively or negatively charged. The repositioning of the electrons sets up a gate voltage that is proportional to the charge placed on the object. In turn, the JFET will either begin to resist or allow current to flow through its drain-source channel, hence resulting in ammeter needle deflection. R1 is used to protect the ammeter, and R2 is used to calibrate it.

4.3.4 Metal Oxide Semiconductor Field-Effect Transistors

Metal oxide semiconductor field-effect transistors (MOSFETs) are incredibly popular transistors that in some ways resemble JFETs. For instance, when a small voltage is applied at its gate lead, the current flow through its drain-source channel is altered. However, unlike JFETS, MOSFETs have larger gate lead input impedances (≥1014 Ω, as compared with ∼109 Ω for JFETs), which means that they draw almost no gate current whatsoever. This increased input impedance is made possible by placing a metal oxide insulator between the gate and the drain-source channel. There is a price to pay for this increased amount of input impedance, which amounts to a very low gate-to-channel capacitance (a few pF). If too much static electricity builds up on the gate of certain types of MOSFETs during handling, the accumulated charge may break through the gate and destroy the MOSFET. (Some MOSFETs are designed with safeguards against this breakdown—but not all.)



The two major kinds of MOSFETs are enhancement-type MOSFETs and depletion-type MOSFETs (see Fig. 4.88). A depletion-type MOSFET is normally on (maximum current flows from drain to source) when no difference in voltage exists between the gate and source terminals (VGS = VGVS = 0 V). However, if a voltage is applied to its gate lead, the drain-source channel becomes more resistive—a behavior similar to a JFET. An enhancement-type MOSFET is normally off (minimum current flows from drain to source) when VGS = 0 V. However, if a voltage is applied to its gate lead, the drain-source channel becomes less resistive.

Both enhancement-type and depletion-type MOSFETs come in either n-channel or p-channel forms. For an n-channel depletion-type MOSFET, a negative gate-source voltage (VG < VS) increases the drain-source channel resistance, whereas for a p-channel depletion-type MOSFET, a positive gate-source voltage (VG > VS) increases the channel resistance. For an n-channel enhancement-type MOSFET, a positive gate-source voltage (VG > VS) decreases the drain-source channel resistance, whereas for a p-channel enhancement-type MOSFET, a negative gate-source voltage (VG < VS) decreases the channel resistance.

MOSFETs are perhaps the most popular transistors used today; they draw very little input current, are easy to make (require few ingredients), can be made extremely small, and consume very little power. In terms of applications, MOSFETs are used in ultrahigh input impedance amplifier circuits, voltage-controlled “resistor” circuits, switching circuits, and found with large-scale integrated digital ICs.

Like JFETs, MOSFETs have small transconductance values when compared with bipolar transistors. In terms of amplifier applications, this can lead to decreased gain values. For this reason, you will rarely see MOSFETs in simple amplifier circuits, unless there is a need for ultrahigh input impedance and low input current features.

How MOSFETs Work



Both depletion and enhancement MOSFETs use an electrical field—produced by a gate voltage—to alter the flow of charge carriers through the semiconductive drain-source channel. With depletion-type MOSFETs, the drain-source channel is inherently conductive; charge carriers such as electrons (n-channel) or holes (p-channel) are already present within the n-type or p-type channel. If a negative gate-source voltage is applied to an n-channel depletion-type MOSFET, the resulting electrical field acts to “pinch off” the flow of electrons through the channel (see Fig. 4.88a). A p-channel depletion-type MOSFET uses a positive gate-source voltage to “pinch off” the flow of holes through its channel (see Fig. 4.88b). (The pinching-off effect results from depletion regions forming about the upper and lower gate contacts.) Enhancement MOSFETs, unlike depletion MOSFETs, have a normally resistive channel; there are few charge carriers within it. If a positive gate-source voltage is applied to an n-channel enhancement-type MOSFET, electrons within the p-type semiconductor region migrate into the channel and thereby increase the conductance of the channel (see Fig. 4.88c). For a p-channel enhancement MOSFET, a negative gate-source voltage draws holes into the channel to increase the conductivity (see Fig. 4.88d).

Basic Operation



The circuits shown here demonstrate how MOSFETs can be used to control current flow through a light bulb. The desired dimming effects produced by the gate voltages may vary depending on the specific MOSFET you are working with.


In terms of theory, you can treat depletion-type MOSFETs like JFETs, except you must give them larger input impedances. The following graphs, definitions, and formulas summarize the theory.



OHMIC REGION MOSFET is just beginning to resist. In this region, the MOSFET behaves like a variable resistor.

ACTIVE REGION MOSFET is most strongly influenced by gate-source voltage (VGS) but hardly at all influenced by drain-source voltage (VDS).

CUTOFF VOLTAGE (VGS,OFF) Often referred to as the pinch-off voltage (Vp). Represents the particular gate-source voltage that causes the MOSFET to block almost all drain-source current flow.

BREAKDOWN VOLTAGE (BVDS) The drain-source voltage (VDS) that causes current to “break through” MOSFET’s resistive channel.

DRAIN CURRENT FOR ZERO BIAS (IDSS) Represents the drain current when gate-source voltage is zero volts (or when gate is shorted to source).

TRANSCONDUCTANCE (gm) Represents the rate of change in the drain current with change in gate-source voltage when drain-source voltage is fixed for a particular VDS. It is analogous to the transconductance I/Rtr for bipolar transistors.

Useful Formulas for Depletion-Type MOSFETs


Technical Info and Formulas for Enhancement-Type MOSFETs

Predicting how enhancement-type MOSFETs will behave requires learning some new concepts and formulas. Here’s an overview of the theory.



OHMIC REGION MOSFET is just beginning to conduct. Acts like a variable resistor.

ACTIVE REGION MOSFET is most strongly influenced by gate-source voltage VGS but hardly at all influenced by drain-source voltage VDS.

THRESHOLD VOLTAGE (VGS,th) Particular gate-source voltage where MOSFET is just beginning to conduct.

BREAKDOWN VOLTAGE (BVDS) The voltage across drain source (VDS) that causes current to “break through” MOSFET’s resistance channel.

DRAIN-CURRENT FOR GIVEN BIAS (ID,on) Represents the amount of current ID at a particular VGS, which is given on data sheets, etc.

TRANSCONDUCTANCE (gm) Represents the rate of change in the drain current with the change in gate-source voltage when drain-source voltage is fixed. It is analogous to the transconductance I/Rtr for bipolar transistors.


Sample Problems


An n-channel depletion-type MOSFET has an IDDS = 10 mA and a VGS,off = −4 V. Find the values of ID, gm, and RDS when VGS = −2 V and when VGS = +1 V. Assume that the MOSFET is in the active region.



When considering VGS = −2 V, use the following:

Before you can find gm, you must find gm0—here’s what you use:

Now you can substitute gm0 into the following expression to find gm:

The drain-source resistance is then found by using RDS = 1/gm = 400 Ω.

If you do the same calculations for VGS = + 1 V, you get ID = 15.6 mA, gm = 6250 μmhos, and RDS = 160 Ω.


An n-channel enhancement-type MOSFET has a VGS,th = +2 V and an ID = 12 mA. When VGS = +4 V, find parameters k, gm, and RDS. Assume that the MOSFET is in the active region.



To find k, use the expression for the drain current in the active region:

ID = k (VGS VGS,th)2

Rearranging this equation and plugging in the knowns, you get

To find gm, use the following:

The drain-source resistance is then found by using RDS = 1/gm = 83 Ω.


In the following n-channel depletion-type MOSFET circuit, IDSS = 10 mA, VGS,off = −4 V, RD = 1 kΩ, and VDD = +20 V, find VD and the gain Vout/Vin.



Applying Ohm’s and Kirchhoff’s laws, you can come up with the following expressions:



where the last expression takes into account the grounded source terminal. (Note the 1-MΩ resistor. It is a self-biasing resistor and is used to compensate leakage currents and other parameters that can lead to MOSFET instability. The voltage drop across this resistor can be neglected because the gate leakage current is very small, typically in the nA or pA range.) Now, if you assume that there is no input voltage, you can say that ID = IDSS. This means that


= 20 V − (10 mA)(1 kΩ) = 10 V

To find the gain, use the following formula:


Substituting gm0 back into the gain formula, you get a resulting gain of 5.


In the following n-channel enhancement-type MOSFET circuit, if k = 1000 µmhos/V, VDD = 20 V, VGS,th = 2 V, and VGS = 5 V, what should the resistance of RD be to center VD to 10 V? Also, what is the gain for this circuit?



First, determine the drain current:

ID = k (VGS VGS,th)2

= (1000 μmhos/V)(5 V − 2 V)2 = 9 mA

Next, to determine the size of RD that is needed to set VD to 10 V, use Ohm’s law:

(The 1-MΩ resistor has the same role as the 1-MΩ resistor in the last example.)

To find the gain, first find the transconductance:

gm = 2k (VGS VGS,th) = 2(1000 μmhos/V)(5 V − 2 V) = 6000 μmhos

Next, substitute gm into the gain expression:

Important Things to Know about MOSFETs

MOSFETs may come with a fourth lead, called the body terminal. This terminal forms a diode junction with the drain-source channel. It must be held at a nonconducting voltage [say, to the source or to a point in a circuit that is more negative than the source (n-channel devices) or more positive than the source (p-channel devices)]. If the base is taken away from the source (for enhancement-type MOSFETs) and set to a different voltage than that of the source, the effect shifts the threshold voltage VGS,th by an amount equal to ½VBS1/2 in the direction that tends to decrease drain current for a given VGS. Some instances when shifting the threshold voltage becomes important are when leakage effects, capacitance effects, and signal polarities must be counterbalanced. The body terminal of a MOSFET is often used to determine the operating point of a MOSFET by applying an incremental ac signal to its gate.



Damaging a MOSFET Is Easy

MOSFETs are extremely fragile. Their delicate gate-channel oxide insulators are subject to electron bombardment from statically charged objects. For example, it is possible for you to blow a hole through one of these insulators simply by walking across a carpet and then touching the gate of the MOSFET. The charge you pick up during your walk may be large enough to set yourself at a potential of a few thousand volts. Although the amount of current discharged during an interaction is not incredibly large, it does not matter; the oxide insulator is so thin (the gate-channel capacitance is so small, typically a few pF) that a small current can be fatal to a MOSFET. When installing MOSFETs, it is essential to eliminate all static electricity from your work area. In Chap. 7 you’ll find guidelines for working with components subject to electrostatic discharge.

Kinds of MOSFETs

Like the other transistors, MOSFETs come in either metal can-like containers or plastic packages. High-power MOSFETs come with metal tabs that can be fastened to heat sinks. High/low MOSFET driver ICs also are available. These drivers (typically DIP in form) come with a number of independent MOSFETs built in and operate with logic signals.



Things to consider when purchasing a MOSFET include breakdown voltages, ID,max, RDS(on),max, power dissipation, switching speed, and electrostatic discharge protection.





Here, an n-channel enhancement-type power MOSFET is used to control the current flow through a lamp. The voltage-divider resistor R2 sets the gate voltage, which in turn sets the drain current through the lamp.




In the circuit shown here, an op amp is combined with an n-channel depletion-type MOSFET to make a highly reliable current source (less than 1 percent error). The MOSFET passes the load current, while the inverting input of the op amp samples the voltage across RS and then compares it with the voltage applied to the noninverting input. If the drain current attempts to increase or decrease, the op amp will respond by decreasing or increasing its output, hence altering the MOSFETs gate voltage in the process. This in turn controls the load current. This op amp/MOSFET current source is more reliable than a simple bipolar transistor-driven source. The amount of leakage current is extremely small. The load current for this circuit is determined by applying Ohm’s law (and applying the rules for op amps discussed in Chap. 8):

Iload = Vin/RS



FIGURE 4.100

Common-source and source-follower amplifiers can be constructed using both depletion- and enhancement-type MOSFETs. The depletion-type amplifiers are similar to the JFET amplifiers discussed earlier, except that they have higher input impedances. The enhancement-type MOSFET amplifiers essentially perform the same operations as the depletion-type MOSFET amplifiers, but they require a voltage divider (as compared with a single resistor) to set the quiescent gate voltage. Also, the output for the enhancement-type common-source MOSFET amplifier is inverted. The role of the resistors and capacitors within these circuits can be better understood by referring to the amplifier circuits discussed earlier.



FIGURE 4.101

In this circuit, an n-channel enhancement-type MOSFET is used to amplify an audio signal generated by a high-impedance microphone and then uses the amplified signal to drive a speaker. C1 acts as an ac coupling capacitor, and the R2 voltage divider resistor acts to control the gain (the volume).



FIGURE 4.102

The circuit shown here uses an n-channel depletion-type MOSFET as an interface between a logic circuit and an analog circuit. In this example, an AND gate is used to drive a MOSFET into conduction, which in turn activates the relay. If inputs A and B are both high, the relay is switched to position 2. Any other combination (high/low, low/high, low/low) will put the relay into position 1. The MOSFET is a good choice to use as a digital-to-analog interface; its extremely high input resistance and low input current make it a good choice for powering high-voltage or high-current analog circuits without worrying about drawing current from the driving logic.



FIGURE 4.103

A permanent magnet DC motor rotates either clockwise or counterclockwise, depending on polarity of the applied voltage across its terminals. A simple circuit that can be used to control the on/off state of a motor as well as direction of rotation is shown to the left. This circuit is an H-bridge motor-control circuit built with power MOSFETS. Transistors Q1 and Q2 are Nchannel MOSFETs, and Q3 and Q4 are P-channel MOSFETS. To turn on the motor as well as control the direction of rotation, switches SW1 and SW2 are used. Both switches are push button switches that are of the normal open variety. When SW1 is pressed, the voltage on gates Q1 and Q3 goes to zero, turning off Q1 and turning on Q3. This creates a current path from Q3, through the motor and through Q2. This causes the motor to turn clockwise. When SW1 is released, the motor turns off. When SW2 is pressed, Q2 turns off and Q4 turns on, creating a reverse current path through Q1, the motor and Q4, and the motor turns counterclockwise. To provide digital control, SW1 and SW2 can be replaced by transistors (or similar switching devices) that can be turned on and off using a microcontroller.

4.3.5 Insulated Gate Bipolar Transistors (IGBTs)

IGBTs are a hybrid of a MOSFET and a bipolar transistor. This is reflected in the IGBTs electronic symbol (see Fig. 4.104). This has the Gate terminal of a MOSFET and the Collector and Emitter terminals of a bipolar transistor. As you might have expect from this, the transistor is commonly used as a switch often at very high currents and voltages. The switch is voltage controlled like a MOSFET but has high current capabilities of a bipolar transistor.


FIGURE 4.104

IGBTs have found a niche in very high power applications such as electric vehicles, where modules are constructed from a number of IGBTs in parallel to achieve switching powers in the hundreds of amperes at high voltages. The very forgiving pulse capabilities have found application amongst hobbyists in solid-state Tesla coils.

4.3.6 Unijunction Transistors

Unijunction transistors (UJTs) are three-lead devices that act exclusively as electrically controlled switches (they are not used as amplifier controls). The basic operation of a UJT is relatively simple. When no potential difference exists between its emitter and either of its base leads (B1 or B2), only a very small current flows from B2 to B1. However, if a sufficiently large positive trigger voltage—relative to its base leads—is applied to the emitter, a larger current flows from the emitter and combines with the small B2-to-B1 current, thus giving rise to a larger B1 output current. Unlike the other transistors covered earlier—where the control leads (e.g., emitter, gate) provided little or no additional current—the UJT is just the opposite; its emitter current is the primary source of additional current.


FIGURE 4.105

How UJTs Work

A simple model of a UJT is shown in Fig. 4.105. It consists of a single bar of n-type semiconductor material with a p-type semiconductor “bump” in the middle. One end of the bar makes up the base 1 terminals, the other end the base 2 terminal, and the “bump” represents the emitter terminal. Below is a simple “how it works” explanation.


FIGURE 4.106

With no voltage applied to the emitter, only a relatively small number of electrons makes it through the n-region between base 1 and base 2. Normally, both connectors to bases 1 and 2 are resistive (each around a few thousand ohms).

When a sufficiently large voltage is applied to the emitter, the emitter-channel p-n junction is forward-biased (similar to forward-biasing a diode). This in turn allows a large number of base 1 electrons to exit through the emitter. Now, since conventional currents are defined to be flowing in the opposite direction of electron flow, you would say that a positive current flows from the emitter and combines with channel current to produce a larger base 1 output current.

Technical Info


FIGURE 4.107

Figure 4.107 shows a typical VE versus IE graph of an UJT, as well as a UJT equivalent circuit. In terms of the UJT theory, if B1 is grounded, a voltage applied to the emitter will have no effect (does not increase conductance from one base to another) until it exceeds a critical voltage, known as the triggering voltage. The triggering voltage is given by the following expression:

In this equation, RB1 and RB2 represent the inherent resistance within the region between each base terminal and the n-channel. When the emitter is open-circuited, the combined channel resistance is typically around a few thousand ohms, where RB1 is somewhat larger than RB2. Once the trigger voltage is reached, the p-n junction is forward-biased (the diode in the equivalent circuit begins to conduct), and current then flows from the emitter into the channel. But how do we determine RB1 and RB2? Will the manufacturers give you these resistances? They most likely will not. Instead, they typically give you a parameter called the intrinsic standoff ratio η. This intrinsic standoff ratio is equal to the RB1/(RB1 + RB2) term in the preceding expression, provided the emitter is not conducting. The value of η is between 0 to 1, but typically it hangs out at a value around 0.5.



FIGURE 4.108

Most frequently, UJTs are used in oscillator circuits. Here, a UJT, along with some resistors and a capacitor, makes up a relaxation oscillator that is capable of generating three different output waveforms. During operation, at one instant in time, CE charges through RE until the voltage present on the emitter reaches the triggering voltage. Once the triggering voltage is exceeded, the E-to-B1 conductivity increases sharply, which allows current to pass from the capacitor-emitter region, through the emitter-base 1 region, and then to ground. CE suddenly loses its charges, and the emitter voltage suddenly falls below the triggering voltage. The cycle then repeats itself. The resulting waveforms generated during this process are shown in the figure. The frequency of oscillation is determined by the RC charge-discharge period and is given by

For example, if RE = 100 kΩ, CE = 0.1 μF, and η = 0.61, then f = 106 Hz.

Kinds of UJTs



FIGURE 4.109

These UJTs are used in oscillatory, timing, and level-detecting circuits. Typical maximum ratings include 50 mA for IE, 35 to 55 V for the interbase voltage (VBB), and 300 to 500 mW for power dissipation.



FIGURE 4.110

A PUT is similar to a UJT, except that RBB, IV (valley current level), IP (peak current level), and g (intrinsic standoff ratio) can be programmed by means of an external voltage divider. Being able to alter these parameters is often essential in order to eliminate circuit instability. The electronic symbol for a PUT looks radically different when compared with a UJT (see figure). The lead names are also different; there is a gate lead, a cathode lead, and an anode lead. PUTs are used to construct timer circuits, high-gain phase-control circuits, and oscillator circuits. We have included a simple PUT application in the applications section that follows.




FIGURE 4.111

The circuit here causes a relay to throw its switch from one position to another in a repetitive manner. The positive supply voltage charges up the capacitor. When the voltage across the capacitor reaches the UJT’s triggering voltage, the UJT goes into conduction. This causes the relay to throw its switch to position 2. When the capacitor’s charge runs out, the voltage falls below the triggering voltage, and the UJT turns off. The relay then switches to position 1. R1 controls the charging rate of the capacitor, and the size of the capacitor determines the amount of voltage used to trigger the UJT. C also determines the charge rate.



FIGURE 4.112 Here, a UJT is combined with a few resistors, a bipolar transistor, and a capacitor to make an oscillatory sawtooth generator that has controlled amplification (set by the bipolar transistor). Like the preceding oscillators, C1 and R3 set the frequency. The bipolar transistor samples the voltage on the capacitor and outputs a ramp or sawtooth waveform.



FIGURE 4.113


PUT begins to conduct when

VA = VG + 0.7 V

where VG is set by the voltage divider:

When VA is reached, the anode current becomes:

Here, a PUT is programmed by R1 and R2 to set the desired triggering voltage and anode current. These two resistors form a voltage divider that sets the gate voltage VG (terminal used to turn PUT on or off). For the PUT to conduct, the anode voltage must exceed the gate voltage by at least 0.7 V. At a moment when the capacitor is discharged, the gate is reverse-biased, and the PUT is turned off. Over time, the capacitor begins charging through R4, and when enough charge is collected, a large enough voltage will be present to forward-bias the gate. This then turns the PUT on (i.e., if the anode current IA exceeds the peak current IP). Next, the capacitor discharges through the PUT and through R3. (Note: When a PUT is conducting, the voltage from the anode to the cathode is about 1 V.) As the capacitor reaches full discharge, the anode current decreases and finally stops when the gate no longer has a sufficient voltage applied to it. After that, the charging begins again, and the cycle repeats itself, over and over again. By tapping the circuit at the gate and source terminals, you can output both a spiked and sawtooth wave pattern.

4.4 Thyristors

4.4.1 Introduction

Thyristors are two- to four-lead semiconductor devices that act exclusively as switches—they are not used to amplify signals, like transistors. A three-lead thyristor uses a small current/voltage applied to one of its leads to control a much larger current flow through its other two leads. A two-lead thyristor, on the other hand, does not use a control lead but instead is designed to switch on when the voltage across its leads reaches a specific level, known as the breakdown voltage. Below this breakdown voltage, the two-lead thyristor remains off.

You may be wondering at this point, Why not simply use a transistor instead of a thyristor for switching applications? Well, you could—often transistors are indeed used as switches—but compared with thyristors, they are trickier to use because they require exacting control currents/voltages to operate properly. If the control current/voltage is not exact, the transistor may lay in between on and off states. And according to common sense, a switch that lies in between states is not a good switch. Thyristors, on the other hand, are not designed to operate in between states. For these devices, it is all or nothing—they are either on or off.

In terms of applications, thyristors are used in speed-control circuits, power-switching circuits, relay-replacement circuits, low-cost timer circuits, oscillator circuits, level-detector circuits, phase-control circuits, inverter circuits, chopper circuits, logic circuits, light-dimming circuits, motor speed-control circuits, etc.

TABLE 4.6 Major Kinds of Thyristors




Silicon-controlled rectifier (SCR)


Normally off, but when a small current enters its gate (G), it turns on. Even when the gate current is removed, the SCR remains on. To turn it off, the anode-to-cathode current flow must be removed, or the anode must be set to a more negative voltage than the cathode. Current flows in only one direction, from anode (A) to cathode (C).

Silicon-controlled switch (SCS)


Similar to an SCR, but it can be made to turn off by applying a positive voltage pulse to a four-lead, called the anode gate. This device also can be made to trigger on when a negative voltage is applied to the anode-gate lead. Current flows in one direction, from anode (A) to cathode (C).



Similar to an SCR, but it can switch in both directions, meaning it can switch ac as well as dc currents. A triac remains on only when the gate is receiving current, and it turns off when the gate current is removed. Current flows in both directions, through MT1 and MT2.

Four-layer diode


It has only two leads. When placed between two points in a circuit, it acts as a voltage-sensitive switch. As long as the voltage difference across its leads is below a specific breakdown voltage, it remains off. However, when the voltage difference exceeds the breakdown point, it turns on. Conducts in one direction, from anode (A) to cathode (C).



Similar to the four-layer diode but can conduct in both directions. Designed to switch either ac or dc.

Table 4.6 provides an overview of the major kinds of thyristors. When you see the phrase turns it on, this means a conductive path is made between the two conducting leads [e.g., anode (A) to cathode (C), MT1 to MT2]. Normally off refers to the condition when no voltage is applied to the gate (the gate is open-circuited). We will present a closer look at these thyristors in the subsections that follow.

4.4.2 Silicon-Controlled Rectifiers

SCRs are three-lead semiconductor devices that act as electrically controlled switches. When a specific positive trigger voltage/current is applied to the SCR’s gate lead (G), a conductive channel forms between the anode (A) and the cathode (C) leads. Current flows in only one direction through the SCR, from anode to cathode (like a diode).


FIGURE 4.114

Another unique feature of an SCR, besides its current-controlled switching, has to do with its conduction state after the gate current is removed. After an SCR is triggered into conduction, removing the gate current has no effect. That is, the SCR will remain on even when the gate current/voltage is removed. The only way to turn the device off is to remove the anode-to-cathode current or to reverse the anode and cathode polarities.

In terms of applications, SCRs are used in switching circuits, phase-control circuits, inverting circuits, clipper circuits, and relay-control circuits, to name a few.

How SCRs Work

An SCR is essentially just an npn and a pnp bipolar transistor sandwiched together, as shown in Fig. 4.115. The bipolar transistor equivalent circuit works well in describing how the SCR works.


FIGURE 4.115


Using the bipolar equivalent circuit, if the gate is not set to a specific positive voltage needed to turn the npn transistor on, the pnp transistor will not be able to “sink” current from its own base. This means that neither transistor will conduct, and hence current will not flow from anode to cathode.


If a positive voltage is applied to the gate, the npn transistor’s base is properly biased, and it turns on. Once on, the pnp transistor’s base can now “sink” current though the npn transistor’s collector—which is what a pnp transistor needs in order to turn on. Since both transistors are on, current flows freely between anode and cathode. Notice that the SCR will remain on even after the gate current is removed. This—according to the bipolar equivalent circuit—results from the fact that both transistors are in a state of conduction when the gate current is removed. Because current is already in motion through the pnp transistors base, there is no reason for the transistors to turn off.

Basic SCR Applications



FIGURE 4.116

Here, an SCR is used to construct a simple latching circuit. S1 is a momentary contact, normally open pushbutton switch, while S2 is a momentary contact, normally closed pushbutton switch. When S1 is pushed in and released, a small pulse of current enters the gate of the SCR, thus turning it on. Current will then flow through the load. The load will continue to receive current until the moment S2 is pushed, at which time the SCR turns off. The gate resistor acts to set the SCR’s triggering voltage/current. We’ll take a closer look at the triggering specifications in a second.


FIGURE 4.117 Here, an SCR is used to rectify a sinusoidal signal that is to be used to power a load. When a sinusoidal waveform is applied to the gate, the SCR turns on when the anode and gate receive the positive going portion of the waveform (provided the triggering voltage is exceeded). Once the SCR is on, the waveform passes through the anode and cathode, powering the load in the process. During the negative going portion of the waveform, the SCR acts like a reverse-biased diode; the SCR turns off. Increasing R1 has the effect of lowering the current/voltage supplied to the SCR’s gate. This in turn causes a lag in anode-to-cathode conduction time. As a result, the fraction of the cycle over which the device conducts can be controlled (see graph), which means that the average power dissipated by Rload can be adjusted. The advantage of using an SCR over a simple series variable resistor to control current flow is that essentially no power is lost to resistive heating.



FIGURE 4.118

An SCR along with a few resistors, a capacitor, and a UJT can be connected together to make a variable-speed control circuit used to run a dc motor. The UJT, the capacitor, and the resistors make up an oscillator that supplies an ac voltage to the SCR’s gate. When the voltage at the gate exceeds the SCR’s triggering voltage, the SCR turns on, thus allowing current to flow through the motor. Changing the resistance of R1 changes the frequency of the oscillator and hence determines the number of times the SCR’s gate is triggered over time, which in turn controls the speed of the motor. (The motor appears to turn continuously, even though it is receiving a series of on/off pulses. The number of on cycles averaged over time determines the speed of the motor.) Using such a circuit over a simple series variable resistor to control the speed of the motor wastes less energy.

Kinds of SCRs

Some SCRs are designed specifically for phase-control applications, while others are designed for high-speed switching applications. Perhaps the most distinguishing feature of SCRs is the amount of current they can handle. Low-current SCRs typically come with maximum current/voltage ratings approximately no bigger than 1 A/100 V. Medium-current SCRs, on the other hand, come with maximum current/voltage ratings typically no bigger than 10 A/100 V. The maximum ratings for high-current SCRs may be several thousand amps at several thousand volts. Low-current SCRs come in plastic or metal can-like packages, while medium and high-current SCRs come with heat sinks built in.

Technical Stuff


FIGURE 4.119

Here are some common terms used by the manufacturers to describe their SCRs:


On state-voltage. The anode-to-cathode voltage present when the SCR is on.


Gate trigger current. The minimum gate current needed to switch the SCR on.


Gate trigger voltage. The minimum gate voltage required to trigger the gate trigger current.


Holding current. The minimum current through the anode-to-cathode terminal required to maintain the SCR’s on state.


Peak gate power dissipation. The maximum power that may be dissipated between the gate and the cathode region.


Repetitive peak off-state voltage. The maximum instantaneous value of the off-state voltage that occurs across an SCR, including all repetitive transient voltages but excluding all nonrepetitive transient voltages.


Repetitive peak off-state current. The maximum instantaneous value of the off-state current that results from the application of repetitive peak off-state voltage.


Repetitive peak reverse voltage. The maximum instantaneous value of the reverse voltage that occurs across an SCR, including all repetitive transient voltages but excluding all nonrepetitive transient voltages.


Repetitive peak reverse current. Maximum instantaneous value of the reverse current that results from the application of repetitive peak reverse voltage.

Here’s a sample section of an SCR specifications table to give you an idea of what to expect (Table 4.7).

TABLE 4.7 Sample Section of an SCR Specifications Table





VT (V)














4.4.3 Silicon-Controlled Switches

A silicon-controller switch (SCS) is a device similar to an SCR, but is designed for situations requiring increased control, triggering sensitivity, and firing predictability. For example, the typical turn-off time for an SCS is from 1 to 10 microseconds as opposed to 5 to 30 microseconds for an SCR. Unlike an SCR, an SCS has lower power, current and voltage ratings, typically with a max anode current from 100 mA to 300 mA and a power dissipation from 100 to 500 mW. Unlike an SCR, a SCS can also switch OFF when a positive voltage/input current is applied to an extra anode gate lead. The SCS can also be triggered into conduction when a negative voltage/output current is applied to that same lead. The figure below shows the schematic symbol for an SCS.


FIGURE 4.120

SCS are used in practically any circuit that needs a switch that turns on and off through two distinct control pulses. They are found in power-switching circuits, logic circuits, lamp drivers, voltage sensors, pulse generators, etc.

How an SCS Works

Figure 4.121a shows a basic four-layer, three-junction P-N-P-N silicon model of an SCS with four electrodes, namely the cathode (C), cathode gate (G1), anode gate (G2), and anode (A). An equivalent circuit of the SCS can be modeled by the back-to-back bipolar transistor network shown in Fig. 4.121c. Using the two-transistor equivalent circuit, when a negative pulse is applied to the anode gate (G2), transistor Q1 switches ON. Q1 supplies base current to transistor Q2, and both transistors switch ON. Likewise, a positive pulse at the cathode gate G1 can switch the device on. Because the SCS uses only small currents, it can be switched off by an appropriate polarity pulse at one of the gates. At the cathode gate, a negative pulse is required to switch the device off, while at the anode gate a positive pulse is needed.


FIGURE 4.121


When buying an SCS, make sure to select a device that has the proper breakdown voltage, current, and power-dissipation ratings. A typical specification table will provide the following ratings: BVCB, BVEB, BVCE, IE, IC, IH (holding current), and PD (power dissipation). Here we have assumed the alternate lead name designations.

4.4.4 Triacs

Triacs are devices similar to SCRs—they act as electrically controlled switches—but unlike SCRs, they are designed to pass current in both directions, therefore making them suitable for ac applications. Triacs come with three leads, a gate lead and two conducting leads called MT1 and MT2. When no current/voltage is applied to the gate, the triac remains off. However, if a specific trigger voltage is applied to the gate, the device turns on. To turn the triac off, the gate current/voltage is removed.


FIGURE 4.122

Triacs are used in ac motor control circuits, light-dimming circuits, phase-control circuits, and other ac power-switching circuits. They are often used as substitutes for mechanical relays.

How a Triac Works

Figure 4.123 shows a simple n-type/p-type silicon model of a triac. This device resembles two SCRs placed in reverse parallel with each other. The equivalent circuit describes how the triac works.


FIGURE 4.123


Using the SCR equivalent circuit, when no current/voltage is applied to the gate lead, neither of the SCRs’ gates receives a triggering voltage; hence current cannot flow in either direction through MT1 and MT2.


When a specific positive triggering current/voltage is applied to the gate, both SCRs receive sufficient voltage to trigger on. Once both SCRs are on, current can flow in either direction through MT1 to MT2 or from MT2 to MT1. If the gate voltage is removed, both SCRs will turn off when the ac waveform applied across MT1 and MT2 crosses zero volts.

Basic Applications



FIGURE 4.124

Here is a simple circuit showing how a triac acts to permit or prevent current from reaching a load. When the mechanical switch is open, no current enters the triac’s gate; the triac remains off, and no current passes through the load. When the switch is closed, a small current slips through RG, triggering the triac into conduction (provided the gate current and voltage exceed the triggering requirements of the triac). The alternating current can now flow through the triac and power the load. If the switch is open again, the triac turns off, and current is prevented from flowing through the load.



FIGURE 4.125 A triac along with a variable resistor and a capacitor can be used to construct an adjustable full-wave rectifier. The resistance R of the variable resistor sets the time at which the triac will trigger on. Increasing R causes the triac to trigger at a later time and therefore results in a larger amount of clipping (see graph). The size of C also determines the amount of clipping that will take place. (The capacitor acts to store charge until the voltage across its terminals reaches the triac’s triggering voltage. At that time, the capacitor will dump its charge.) The reason why the capacitor can introduce additional clipping results from the fact that the capacitor may cause the voltage at the gate to lag the MT2-to-MT1 voltage (e.g., even if the gate receives sufficient triggering voltage, the MT2-to-MT1 voltage may be crossing zero volts). Overall, more clipping results in less power supplied to the load. Using this circuit over a simple series variable resistor connected to a load saves power. A simple series variable resistor gobbles up energy. This circuit, however, supplies energy-efficient pulses of current.



FIGURE 4.126

This circuit is used in many household dimmer switches. The diac—described in the next section—acts to ensure accurate triac triggering. (The diac acts as a switch that passes current when the voltage across its leads reaches a set breakdown value. Once the breakdown voltage is reached, the diac releases a pulse of current.) In this circuit, at one moment the diac is off. However, when enough current passes through the resistors and charges up the capacitor to a voltage that exceeds the diac’s triggering voltage, the diac suddenly passes all the capacitor’s charge into the triac’s gate. This in turn causes the triac to turn on and thus turns the lamp on. After the capacitor is discharged to a voltage below the breakdown voltage of the diac, the diac turns off, the triac turns off, and the lamp turns off. Then the cycle repeats itself, over and over again. Now, it appears that the lamp is on (or dimmed to some degree) because the on/off cycles are occurring very quickly. The lamp’s brightness is controlled by R2.



FIGURE 4.127

This circuit has the same basic structure as the light dimmer circuit, with the exception of the transient suppressor section (R2C2). The speed of the motor is adjusted by varying R1.

Kinds of Triacs

Triacs come in low-current and medium-current forms. Low-current triacs typically come with maximum current/voltage ratings no bigger than 1 A/(several hundred volts). Medium-current triacs typically come with maximum current/voltage rating of up to 40 A/(few thousand volts). Triacs cannot switch as much current as high-current SCRs.


FIGURE 4.128

Technical Stuff

Here are some common terms used by the manufacturers to describe their triacs:


RMS on-state current. The maximum allowable MT1-to-MT2 current


DC gate trigger current. The minimum dc gate current needed to switch the triac on


DC gate trigger voltage. The minimum dc gate voltage required to trigger the gate trigger current


DC holding current. The minimum MT1-to-MT2 dc current needed to keep the triac in its on state


Peak gate power dissipation. The maximum gate-to-MT1 power dissipation


Surge current. Maximum allowable surge current

Here’s a sample section of a triac specifications table to give you an idea of what to expect (Table 4.8).

TABLE 4.8 Sample Section of a Triac Specifications Table






IH (mA)









4.4.5 Four-Layer Diodes and Diacs

Four-layer diodes and diacs are two-lead thyristors that switch current without the need of a gate signal. Instead, these devices turn on when the voltage across their leads reaches a particular breakdown voltage (or breakover voltage). A four-layer diode resembles an SCR without a gate lead, and it is designed to switch only dc. A diac resembles a pnp transistor without a base lead, and it is designed to switch only ac.


FIGURE 4.129

Four-layer diodes and diacs are used most frequently to help SCRs and triacs trigger properly. For example, by using a diac to trigger a triac’s gate, as shown in Fig. 4.105a, you can avoid unreliable triac triggering caused by device instability resulting from temperature variations, etc. When the voltage across the diac reaches the breakdown voltage, the diac will suddenly release a “convincing” pulse of current into the triac’s gate.


FIGURE 4.130

The circuit in Fig. 4.130 right is used to measure diac characteristics. The 100-kΩ variable resistor is adjusted until the diac fires once for every half-cycle.


Here’s a typical portion of a specifications table for a diac (Table 4.9).

TABLE 4.9 Sample Section of a Diac Specifications Table






PD (mW)







Here, VBO is the breakover voltage, IBO is the breakover current, Ipulse is the maximum peak pulse current, Vswitch is the maximum switching voltage, and PD is the maximum power dissipation.

4.5 Transient Voltage Suppressors

There are numerous devices that can be used to stomp out unwanted transients. Earlier on, we saw how a decoupling capacitor could absorb supply line fluctuations, and we also saw how diodes could clip transient spikes caused by inductive switching action. These devices work fine for such low-power applications, but there are times when transients get so large and energetic that a more robust device is required. Here we’ll take a look at various transient suppressor devices, such as TVSs, varistors, multilayer varistors, Surgectors, and PolySwitches. But before we do that, here’s a little lecture on transients.

4.5.1 Lecture on Transients

Transients are momentary surges or spikes in voltage or current that can wreak havoc within circuits. The peak voltage of a transient can be as small as a few millivolts or as large as several thousand volts, with a duration lasting from a few nanoseconds to more than 100 ms, depending on their origin. In some cases, the transients are repetitive, recurring in a cyclic manner, as in the case of an inductive ringing transient caused by faulty wiring of a motor.

Transients are generated both internally within a circuit and externally—where they enter the circuit via power input lines, signal input/output lines, data lines, and other wires entering and leaving the circuit’s chassis. Internal transients, the predominant of the two, can result from inductive load switching, transistor/logic IC switching, arcing effect, and faulty wiring, to name a few.

In the case of inductive loads, such as motors, relay coils, solenoid coils, and transformers, the sudden switching off of these devices will cause the inductive component within the device to suddenly dump its stored energy into the supply line, creating a voltage spike—recall the inductor equation V = LdI/dt. In many cases, these induced voltages can exceed 1000 V, lasting anywhere from 50 ns to more than 100 ms. Any transistor or logic driver ICs as well as circuits that use the same supply line will suffer, either by getting zapped with the transient spike or suffering from erratic behavior due to propagation of the transient along the power line. (Power lines, or rails, are not perfect conductors and don’t have zero output impedance.)

Switching of TTL and CMOS circuits can also result in transient current spikes of a much smaller threat, yet enough to cause erratic behavior. For example, when the output transistors of a TTL gate switch on, a sudden surge in current is drawn from the supply line. This surge is often quick enough that the supply rail or PCB trace will dip in voltage (due to the fact that a conductor has built-in impedance). All circuits connected to the rail will feel this voltage dip, and the resulting consequences lead to oscillation or some sort of instability that can cause distortion or garble digital logic levels.

Arcing is another transient generator that comes from a number of sources, such as faulty contacts in breakers, switches, and connectors, where arcs jump between the gaps. When electrons jump the gap, the voltage suddenly rises, usually resulting in an oscillatory ringing transient. Faulty connections and poor grounding can also result in transients. For example, motors with faulty windings or insulation can generate a continuous stream of transients exceeding a few hundred volts. Poor electrical wiring practices can also aggravate load-switching transients.

Transients can also attack circuits from external sources through power input lines, signal input and output lines, data lines, and any other wire coming into or going out of a chassis containing the electronics. One cause of external transients is a result of induced voltages onto lines (power, telephone, distributed computer systems, etc.) due to lightning strikes near the lines or the switching of loads, capacitor banks, and so on, at the power utility. External transients may also enter the power line to a circuit due to inductive switching that occurs within a home, such as turning on a hair dryer, microwave, or washing machine. Usually the transient is consumed by other parallel loads, so the effects aren’t as pronounced. For valuable electrical equipment, such as computers, monitors, printers, fax machines, phones, and modems, it is a good idea to use a transient power surge/battery backup protector, with a phone line, too, which will handle the surges and dips in the power and signal lines.

Electrostatic discharge (ESD) is another common form of external transient that can do damage to sensitive equipment and ICs. It usually enters a system through the touch of a fingertip or handheld metal tool. Static electricity that is humidity-dependent can generate low-current transients up to 40,000 V. Systems that are interconnected with long wires, such as telephones and distributed computer systems are efficient collectors of radiated lightning energy. Close-proximity strikes can induce voltages of 300 V or more on signal lines.

Transients are to be avoided; they can cause electronics to operate erratically, perhaps locking up or producing garbled results. They can zap sensitive integrated circuits, causing them to fail immediately or sometime down the road. Today’s microchips are denser than older chips and a transient voltage can literally melt, weld, pit, and burn them, causing temporary or permanent malfunctions to occur. They might also be the cause for decreased efficiency—say, a motor running at higher temperatures due to transients, which interrupt normal timing of the motor and result in microjogging. This produces motor vibration, noise, and excessive heat.

4.5.2 Devices Used to Suppress Transients

There are several devices that can be used when designing circuits to limit the harmful effects of transients. Table 4.10 provides an overview of the most popular devices.

TABLE 4.10






Bypass capacitor

Logic: 0.01-0.22 µF Power: 0.1 µF and up


Used for low-power applications, such as RC snubbers and decoupling of digital logic rails to provide clean power

Low cost, available, simple to apply, fast action, bipolar

Uneven suppression, may fail unpredictably, high capacity

Zener diodes


Diversion/clamping in low-energy circuits running at high frequencies (e.g., high-speed data lines)

Low cost, fast, calibrated clamping voltage, easy to use, standard ratings, bidirectional

Low energy handling, tend to fail open (which can hurt circuit); actually used more for regulation than transients

Transient voltage suppressor diodes (TVS)


Diversion/clamping in low-voltage, low-energy systems, modest frequency

Fast, calibrated low clamping voltage, available, easy to use; fails short-circuited

High capacitance limits frequency, low energy, more expensive than zeners or MOVs

Metal oxide varistors (MOVs)


Diversion/clamping in most low to moderate frequency circuits at all voltage and current levels

Low cost, fast, available, calibrated clamping voltage, easy to use, standard ratings, and bidirectional; handles more total power than TVS; fails short-circuited

Moderate to high capacitance limits high frequency performance

Multilayer varistor (MLTV)


Diversion/clamping in low voltage (3-70 V) systems with modest frequencies

Fast, compact, high energy, low calibrated voltage bidirectional, surface mount

More expensive than zeners or MOVs, high capacitance limits frequency



Diversion (crowbar) for moderate to high energy and frequency circuits and data lines

High speed/moderate energy, sharp clamp voltage, moderate cost

Cost more than other methods, exhibit follow-on current

Avalanche diode


Low-voltage, high-speed logic protection

Very fast (sub-nanosecond response), low shunt capacitance (50 pF)

Low surge capability

Gas discharge and spark gap TVSs

Diversion (crowbar) for very high-energy/voltage applications

Very high-energy capability—upward of 20,000 A in some cases; leakage current is almost nonexistent (within the pA range)

Cost more than other methods, slow response time



Overcurrent protection for speakers, motors, power supplies, battery packs, etc.

Low cost, easy to use, overcurrent protection

Requires a cooling-down period to reset

Transient Voltage Suppressor Diodes (TVSs)

Transient voltage suppressor diodes (TVSs) are popular semiconductor devices used to instantly clamp transient voltages and currents (electrostatic discharge, inductive switching kickback, induced lightning surges, etc.) to safe levels before they can do damage to a circuit. Earlier, in the section on diodes, you saw how standard diodes and zener diodes could be used for transient suppression. Though standard diodes and zener diodes can be used for transient protection, they are actually designed for rectification and voltage regulation and are not as reliable or robust as TVSs.

TVSs come in both unipolar (unidirectional) and bipolar (bidirectional) types. The unipolar TVS breaks down in one direction (current flows in the opposite direction of the arrow—like a zener diode) when its specified breakdown voltage VBR is exceeded. The bipolar TVS, unlike the unipolar TVS, can handle transients in either direction, when the applied voltage across it exceeds its breakdown voltage. See Fig. 4.131.

In terms of design, the TVS should be invisible until a transient occurs. Electrical parameters such as breakdown voltage, standby current (leakage current), and capacitance should have no effect on normal circuit performance. The TVS’s VBR is typically 10 percent above VRWM, which approximates the circuit operating voltage to limit standby current and to allow for variations in VBR caused by the temperature coefficient of the TVS. (In catalogs, they give you both—VBR/VRWM: 12.4 V/11.1 V, 15.2 V/13.6 V, 190 V/171 V, etc.) VRWM should be equal to, or slightly greater than, the normal operating voltage of the protected circuit. When a transient occurs, the TVS clamps instantly to limit the spike voltage to a safe voltage level, while diverting current from the protected circuit. VC should be, of course, less than the maximum voltage the protected circuit can handle. Note that in ac circuits, you should use the peak voltage (Vpeak) values, not the RMS values for selecting VRWM and VBR (Vpeak = 1.4 VRMSVRWM). Also, make sure to choose a TVS that can handle the maximum expected transient pulse current. Figure 4.132 shows various TVS applications.


FIGURE 4.131


Reverse stand-off voltage (VRWM): Also called working voltage, represents the maximum rated dc operating voltage of TVS. At this point, the device will appear as a high impedance to the protected circuit. Discrete devices are available with VRWMs from 2.8 to 440 V.

Maximum breakdown voltage (VBR): The point where TVS begins to conduct and become a low-impedance path for a transient. Discrete devices are available with VBR from 5.3 to 484 V. The breakdown voltage is measured at a test current IT, typically 1 mA or 10 mA. VBR is about 10 percent greater than VRWM.

Maximum peak current (IPP): The maximum permissible surge current that the device can withstand before frying.

Leakage current (IR): The maximum leakage current measured at the working voltage.

Maximum clamping voltage (VC): The maximum clamping voltage during the specified peak impulse current IPP. Typically 35 to 40 percent higher than VBR (or 60 percent higher than VRWM).

Capacitance (CJ): Internal capacitance of TVS, which may become a factor in high-speed data circuits.


FIGURE 4.132

Fig. 1-3: Very high transient voltages are generated when an inductive load is disconnected, such as motors, relay coils, and solenoids. Here, TVSs provide protection to the driving circuitry, as well as limiting damage to relay and solenoid metal contacts.

Fig. 4-7: Typical power sources employing TVS for transient protection. The TVS is chosen for the breakdown voltage that is equal to or greater than the dc output voltage. In most applications, a fuse in the line is desirable.

Fig. 8-9: Input states are vulnerable to low-current, high-voltage static discharges or crosstalk transmitted on the signal wires. Usually an op amp or other IC will have an internal clamp diode, but this provides limited protection for high currents and voltage. Here, an external TVS diode is used to provide additional protection. The second circuit has a TVS on the output of an op amp to prevent a voltage transient due to a short circuit or an inductive load from being transmitted into the output stage.

Fig. 10: Transients generated on the line can vary from a few microseconds’ to several milliseconds’ duration and up to 10,000 V. This threat has given rise to high-noise-immunity ICs. However, the input diodes to these devices, again, have limited internal diode protection, and IC damage is still possible, resulting in either an open circuit or slow degradation of the circuit’s performance with time. Here, a TVS located on the signal line can absorb this excess energy and prevent damage.

Fig. 11: A selection of transient suppressor arrangements for RF coupling.

Metal Oxide Varistors (MOVs) and Multilayer Varistors

A metal oxide varistor (MOV) is a bidirectional semiconductor transient suppressor that acts like a voltage-sensitive variable resistor. Internally, it consists of a complex ceramic crystal structure with various multidirectional metal oxide p-n junction boundaries between crystal grains, all sandwiched between two electrodes. Each individual p-n junction is highly resistive, up until a voltage across the grain boundary in excess of around 3.6 V, where it then is bias on—has a very small resistance. The voltage at which the MOV itself switches is dependent on the average number of grains between its electrode leads. During the manufacturing process, this value can be varied to create any desired breakdown threshold. Due to the random orientation of the boundaries within the MOV, there is no directional, so the MOV acts as a bipolar device—it can be used for ac or dc applications.

In terms of applications, an MOV is usually connected across the mains input of the equipment or the circuit it’s protecting, with a series filter inductor and/or fuse thrown in to protect the MOV itself. In the presence of a transient, the MOV’s resistance switches from high resistance (several megaohms) to very low resistance (a few ohms), transforming itself into a high-current shunt for the transient current. MOVs are made with various clamping voltages, peak current ratings, and maximum energy ratings—reflecting the fact that an MOV can absorb a very large amount of power for a very brief time or smaller amounts over a longer time. For example, an MOV rated at 60 J can absorb 60 W for 1 s, or 600 W for 0.1 s, or 6 kW for 10 ms, or 60 kW for 1 ms, and so on.


FIGURE 4.133

In many regards, MOVs resemble back-to-back zener diodes. However, unlike diodes, MOVs can handle higher-energy transients than zener diodes, since there is no single p-n junction, but rather numerous p-n junctions throughout its structure. The highly conductive ZnO grains act as heat sinks, ensuring a rapid and even distribution of thermal energy throughout the device and minimizing temperature rise. (Note that MOVs can dissipate only a relatively small amount of average power and are unsuited to applications that demand continuous power dissipation.) They are about as fast as zeners and clamp surge voltage to safe levels. Leakage is very low, which means little power is stolen from the circuit. Unlike the zener and other devices, the varistor fails shorted. A zener will also fail in an open-circuit condition that leaves equipment unprotected during a subsequent surge. This helps protect the circuit against subsequent surges; a shorted varistor across an ac or other line might fracture if the energy is high. MOVs should be fused or located where they won’t effect other components should this happen.

In comparison to TVS diodes, MOVs can handle more total power/energy, while leaving a smaller footprint. TVS diodes, however, exhibit much better clamp ratios (better-quality protection) and faster response times (1 to 5 ns compared to about 5 to 200 ns for MOVs). The speed limitation of MOVs, however, is a result of parasitic inductance in the package and leads, and can be minimized with short lead design. MOVs also exhibit an inherent wear-out mechanism within their structure. As the device absorbs transient energy, the electrical characteristics (e.g., leakage, breakdown voltage) tend to drift. On the other hand, TVS diodes have no inherent wear-out mechanism. MOVs have an effective capacitance range from 75 pF for small MOVs to as high as 20,000 pF for large ones. This, combined with the lead inductance, makes practical MOVs slower than TVS diodes, but still fast—in the range of 5 to 200 ns, depending on the device. However, transients that they are designed to remove are usually much longer, so they are usually perfect for the job.

MOVs are found in power supplies of computers and other sensitive equipment, and in mains filters and stabilizers to prevent damage from mains-borne transients due to switching or lightning. They are used in telecommunication and data systems (power supply units, switching equipment, etc.), industrial equipment (control, proximity switches, transformers, motors, traffic lighting), consumer electronics (televisions and video sets, washing machines, etc.), and automotive products (all motor and electronic systems).

One variation of the MOV found in surface-mount form is the multilayer varistor, or MLTV. By having surface-mount contacts, lead self-inductance and series resistance are minimized, allowing for much quicker response time—less than 1 ns. A decrease in series resistance also translates into a massive increase in peak current capability per component unit volume. Even though this is the case, the energy ratings of MLTVs are rather conservative when compared to those of other varistors. One of MLTVs’ strong points is their ability to survive many thousands of strikes, at full rated peak current, without degradation. MLTVs have a characteristic similar to capacitors, having an effective dielectric constant of around 800—much lower than conventional capacitors. Because of this feature, MLTVs are also used in filter circuits. MLTVs come with an operating voltage from 3.5 V to around 68 V, and they are used extensively for transient voltage protection for ICs and transistors, as well as for many ESD and I/O protection schemes.

The following are specifications for MOV and MLTVs:

· Maximum continuous dc voltage (VM(DC)): The maximum continuous dc voltage that may be applied up to the maximum operating temperature of the device. The rated dc operating voltage (working voltage) is also used as the reference point for leakage current. This voltage is always less than the breakdown voltage of the device.

· Maximum continuous ac voltage (VM(AC)): The maximum continuous sinusoidal RMS voltage that may be applied at any temperature up to the maximum operating temperature of the device. It’s related to the previous dc rating by VM(DC) = 1.4 × VM(AC). This means that if a nonsinusoidal waveform is applied, the recurrent peak voltage should be limited to 1.4 × VM(AC).

· Transient energy rating (WTM): Energy is given in joules (watt-seconds). This represents the maximum allowable energy for a single 10/1000-µs impulse current waveform with continuous voltage applied.

· Peak current rating (IPK): The maximum current rating for a given maximum clamping voltage VC.

· Varistor voltage (VB(DC) or VNOM): The voltage at which the device changes from the off state to the on state and enters its conduction mode of operation. The voltage is usually characterized at the 1-mA point and has a specified minimum and maximum voltage listed.

· Clamping voltage (VC): The clamping voltage across an MOV at a peak current IPK.

· Leakage at rated dc voltage (IL): The leakage current when the device is in nonconducting mode, when a specified voltage is applied.

· Capacitance (Cp): The capacitance of the device, typically specified at a frequency of 1 MHz at a bias of 1 Vpp. This capacitance is usually 100 pF or lower for smaller devices, and up to a few thousand for larger ones.

In terms of design, a varistor must operate under both a continuous operating (standby) mode and the predicted transient (normal) mode. Determine the necessary steady-state voltage rating (working voltage), and then establish the transient energy absorbed by the varistor. Calculate the peak transient current through the varistor and determine the power dissipation requirements. Select a model to provide the required voltage-clamping characteristics.


FIGURE 4.134


There are other transient voltage suppression devices out there, such as the Surgector, gas discharge, and spark gap TVSs. The Surgector utilizes silicon thyristor technology to provide bidirectional “crowbar” clamping action for transients of either polarity. This is accomplished with a five-layer p-n junction structure. Surgectors remain in a low-leakage, reverse-bias state, presenting effectively no load to the circuit as long as the applied voltage is at or below its VDRM rating. A transient voltage exceeding this value will cause the device to avalanche (breakdown), beginning the clamping action across the line to which it is connected. As the leading edge of the transient voltage attempts to rise higher, the Surgector current will increase through the circuit’s source impedance until the VBO, or breakover voltage mode, is reached. Thyristor action is then rapidly triggered, and the Surgector switches to its “on,” or latched state. This very low impedance state crowbars the line with effectively the characteristics of a forward p-n junction, thereby short-circuiting the transient voltage.


FIGURE 4.135


A PolySwitch (also known as polyfuse, multiswitch, and generically resettable fuse) is a special positive temperature coefficient resistor that is constructed from a conductive polymer mix. It resembles a varistor and PTC thermistor in one. At normal temperatures, the conductive particles within the polymer form densely packed low-resistance chains, allowing current to flow easily. However, if the current flow through the PolySwitch increases to a point where its temperature rises above a critical level, the crystalline structure of the polymer suddenly changes into an expanded amorphous state. At this point, the device’s resistance dramatically increases, causing a sudden drop in current flow. The point at which this occurs is referred to as the trip current. If the voltage level is maintained after tripping, enough holding current will generally flow, keeping the device in a tripped state. The PolySwitch will reset itself only if the voltage is reduced and the device is allowed to cool, at which point the polymer particles rapidly return to their densely packed state, and the resistance drops.

PolySwitches can be used in numerous applications wherever you need a low-cost, self-resetting solid-state circuit breaker. They are used to limit over-current in speakers, power supplies, battery packs, motors, etc. For example, Fig. 4.136 shows how a PolySwitch used to protect a speaker from excessive current sourced by an amplifier. The PolySwitch is rated with a trip current that is slightly higher than that rated for the power level the speaker can handle. For example, an 8-Ω, 5-W speaker has a maximum current rating determined by the generalized power law.


FIGURE 4.136

Avalanche Diodes

Avalanche diodes are designed to break down and conduct at a specified reverse-bias voltage. This behavior is similar to that of a zener diode, but its operation is caused by a different mechanism, called the avalanche effect (a reverse electric field applied across a p-n junction causes a wave of ionization, reminiscent of an avalanche, leading to a large current). However, unlike zener diodes that are rather restricted in maximum breakdown voltage, avalanche diodes are available with breakdown voltage of over 4000 V. Avalanche diodes are used in circuits to guard against damaging high-voltage transients. They are connected to a circuit so that they’re reverse-biased (the cathode is set positive with respect to the anode). In this configuration, the avalanche diode is nonconducting and doesn’t interfere with the circuit. However, if the voltage rises beyond a safe design limit, the diode goes into avalanche breakdown, eliminating the harmful voltage by shunting current to ground. Avalanche diodes are specified with a clamping voltage VBR and a maximum-size transient that it can absorb, specified either in terms of joules of energy or as I2t. The avalanche breakdown event is not destructive, provided the diode isn’t overheated. One side effect that occurs in avalanche diodes is RF noise generation.

4.6 Integrated Circuits

An integrated circuit (IC) is a miniaturized circuit that contains a number of resistors, capacitors, diodes, and transistors stuffed together on a single chip of silicon no bigger than your fingernail. The number of resistors, capacitors, diodes, and transistors within an IC may vary from just a few to millions.

The trick to cramming everything into such a small package is to make all the components out of tiny n-type and p-type silicon structures that are embedded into the silicon chip during the production phase. To connect the little transistors, resistors, capacitors, and diodes together, aluminum plating is applied along the surface of the chip. Figure 4.137 shows a magnified cross-sectional view of an IC showing how the various components are embedded and linked together.


FIGURE 4.137 The structure of an IC

ICs come in analog, digital, or analog/digital form:

· Analog (or linear) ICs produce, amplify, or respond to varying voltages. Some common analog ICs include voltage regulators, operational amplifiers, comparators, timers, and oscillators.

· Digital (or logic) ICs respond to or produce signals having only high and low voltage states. Common digital ICs include logic gates (such as AND, OR, or NOR), microcontrollers, memories, binary counters, shift registers, multiplexers, encoders, and decoders.

· Analog/digital ICs share properties common with both analog and digital ICs. Analog/digital ICs may take a number of different forms. For example, the IC may be designed primarily as an analog timer but may contain a digital counter. Alternatively, the IC may be designed to read in digital information and then use this information to produce a linear output that can be used to drive, say, a stepper motor or LED display.

ICs are so pervasive that you are likely to use them in any project that you will undertake. You will find them used in many of the chapters that follow.

4.6.1 IC Packages

ICs come in many and various packages (see Fig. 4.138). The determining factors for the package type are the number of pins and the power dissipation. For example, a high-power voltage regulator IC may have three pins and look just like a high-power transistor.

However, the majority of ICs have many more pins and are arranged in a dual in-line (DIL) package (see Fig. 4.138) of 8, 14, 16, 20, 24, or 40 pins. There are also surface-mount versions of the DIL packages, as well as packages arranged as a square with pins on all sides. Some of the surface-mount packages have extremely small spacing between pins—sometimes as small as 0.5 mm, which is two pins every millimeter and not really intended for hand soldering.


FIGURE 4.138 IC packages

Some of the most common packages are listed in Table 4.11.

You will often find that the same ICs are available in multiple packages, making it possible to prototype in something easy to solder like DIL or SO, and then switch to a smaller package for the final product.

TABLE 4.11



PITCH (mm)



Dual in-line



Small outline IC package



Mini/shrink small outline package



Small outline transistor



Thin quad flat pack


Pins on four sides


Thin quad flat no leads


No pins or pads underneath body