Actually, the system of cloud and ground that produces lightning is itself a giant capacitor and the lightning is a failure of that capacitor. Like all capacitors, the system consists of two charged surfaces separated by an insulating material. In this case, the charged surfaces are the cloud bottom and the ground, and the insulating material is the air. During charging, vast amounts of separated electric charge accumulate on the two surfaces--the cloud bottom usually becomes negatively charged and the ground below it becomes positively charge. These opposite charges produce an intense electric field in the region between the cloud and the ground, and eventually the rising field causes charge to begin flowing through the air: a stroke of lightning.

In principle, you could tap into a cloud and the ground beneath and extract the capacitor's charge directly with wires. But this would be a heroic engineering project and unlikely to be worth the trouble. And catching a lightning strike in order to charge a second capacitor is not likely to be very efficient: most of the energy released during the strike would have to dissipate in the air and relatively little of it could be allowed to enter the capacitor. That's because no realistic capacitor can handle the voltage in lightning.

Here's the detailed analysis. The power released during the strike is equal to the strike's voltage times its current: the voltage between clouds and ground and the current flowing between the two during the strike. Voltage is the measure of how much energy each unit of electric charge has and current is the measure of how many units of electric charge are flowing each second. Their product is energy per second, which is power. Added up over time, this power gives you the total energy in the strike. If you want to capture all this energy in your equipment, it must handle all the current and all the voltage. If it can only handle 1% of the voltage, it can only capture 1% of the strike's total energy.

While the current flowing in a lightning strike is pretty large, the voltage involved is astonishing: millions and millions of volts. Devices that can handle the currents associated with lightning are common in the electric power industry but there's nothing reasonable that can handle lightning's voltage. Your equipment would have to let the air handle most of that voltage. The air would extract power from the flowing current in the lightning bolt and turning it into light, heat, and sound. Your equipment would then extract only a token fraction of the stroke's total energy. Finally, your equipment would have to prepare the energy properly for delivery on the AC power grid--its voltage would have to be lowered dramatically and a switching system would have to convert the static charge on the capacitors to an alternating flow of current in the power lines.

An ordinary wire will carry electric current in either direction, while a diode will only carry current in one direction. That's because the electric charges in a wire are free to drift in either direction in response to electric forces but the charges in a diode pass through a one-way structure known as a p-n junction. Charges can only approach the junction from one side and leave from the other. If they try to approach from the wrong side, they discover that there are no easily accessible quantum mechanical pathways or "states" in which they can travel. Sending the charges toward the p-n junction from the wrong side can only occur if something provides the extra energy needed to reach a class of less accessible quantum mechanical states. Light can provide that extra energy, which is why many diodes are light sensitive--they will conduct current in the wrong direction when exposed to light. That is the basis for many light sensitive electronic devices and for most photoelectric or "solar" cells.

Sound consists of small fluctuations in air pressure. We hear sound because these changes in air pressure produce fluctuating forces on various structures in our ears. Similarly, microphones respond to the changing forces on their components and produce electric currents that are effectively proportional to those forces.

Two of the most common types of microphones are capacitance microphones and electromagnetic microphones. In a capacitance microphone, opposite electric charges are placed on two closely spaced surfaces. One of those surfaces is extremely thin and moves easily in response to changes in air pressure. The other surface is rigid and fixed. As a sound enters the microphone, the thin surface vibrates with the pressure fluctuations. The electric charges on the two surfaces pull on one another with forces that depend on the spacing of the surfaces. Thus as the thin surface vibrates, the charges experience fluctuating forces that cause them to move. Since both surfaces are connected by wires to audio equipment, charges move back and forth between the surfaces and the audio equipment. The sound has caused electric currents to flow and the audio equipment uses these currents to record or process the sound information.

In an electromagnetic microphone, the fluctuating air pressure causes a coil of wire to move back and forth near a magnet. Since changing or moving magnetic fields produce electric fields, electric charges in the coil of wire begin to move as a current. This coil is connected to audio equipment and again uses these currents to represent sound.

Actually, you are asking about a current of electrons, which carry a negative charge. It's true that electrons can't be sent across the p-n junction from the p-type side to the n-type side. There are several things that prevent this reverse flow of electrons. First, there is an accumulation of negative charge on the p-type side of the p-n junction and this negative charge repels any electrons that approach the junction from the p-type end. Second, any electron you add to the p-type material will enter an empty valence level. As it approaches the p-n junction, it will find itself with no empty valence levels in which to travel the last distance to the junction. It will end up widening the depletion region--the region of effectively pure semiconductor around the p-n junction; a region that doesn't conduct electricity.

A microwave oven that's built properly and not damaged emits so little electromagnetic radiation that the speaker should never notice. The speaker might have some magnetic field leakage outside its cabinet, and that might have some effect on a microwave oven. However, most microwaves have steel cases and the steel will shield the inner workings of the microwave oven from any magnetic fields leaking from the speaker. The two devices should be independent.

A bipolar transistor is a sandwich consisting of three layers of doped semiconductor. A pure semiconductor such as silicon or germanium has no mobile electric charges and is effectively an insulator (at least at low temperatures). Dope semiconductor has impurities in it that give the semiconductor some mobile electric charges, either positive or negative. Because it contains mobile charges, doped semiconductor conducts electricity. Doped semiconductor containing mobile negative charges is called "n-type" and that with mobile positive charges is called "p-type." In a bipolar transistor, the two outer layers of the sandwich are of the same type and the middle layer is of the opposite type. Thus a typical bipolar transistor is an npn sandwich--the two end layers are n-type and the middle layer is p-type.

When an npn sandwich is constructed, the two junctions between layers experience a natural charge migration--mobile negative charges spill out of the n-type material on either end and into the p-type material in the middle. This flow of charge creates special "depletion regions" around the physical p-n junctions. In this depletion regions, there are no mobile electric charges any more--the mobile negative and positive charges have cancelled one another out!

Because of the two depletion regions, current cannot flow from one end of the sandwich to the other. But if you wire up the npn sandwich--actually an npn bipolar transistor--so that negative charges are injected into one end layer (the "emitter") and positive charges are injected into the middle layer (the "base"), the depletion region between those two layers shrinks and effectively goes away. Current begins to flow through that end of the sandwich, from the base to the emitter. But because the middle layer of the sandwich is very thin, the depletion region between the base and the second end of the sandwich (the "collector") also shrinks. If you wire the collector so that positive charges are injected into it, current will begin to flow through the entire sandwich, from the collector to the emitter. The amount of current flowing from the collector to the emitter is proportional to the amount of current flowing from the base to the emitter. Since a small amount of current flowing from the base to the emitter controls a much larger current flowing from the collector to the emitter, the transistor allows a small current to control a large current. This effect is the basis of electronic amplification--the synthesis of a larger copy of an electrical signal.

A light emitting diode (an LED) produces light when a current of electrons passes through the junction between its two pieces of semiconductor--from a n type semiconductor cathode to an p type semiconductor anode. The LED's light is actually produced in the anode when an electron that has just crossed the p-n junction and is orbiting a positively charged region (called a "hole") drops into the hole to fill it. In filling the hole, the electron releases energy and that energy becomes light through a process called fluorescence.

The energy in a particle of light (a photon) is related the color of that light--with blue photons having more energy than red photons. Here is where the difficulty in making blue LED's comes in: to produce a blue photon, the electron in an LED must give up lots of energy as it fills the hole in the anode. This need for a large energy release places a severe demand on the semiconductors from which the blue LED is made. These semiconductors need an unusually large band gap--the energy spacing between two types of paths that electrons can follow in the semiconductor. It wasn't until recently that good quality semiconductors with the appropriate electrical characteristics were available for this task.

When the two lamps are in parallel with one another, they share the current passing through the rest of the circuit. Current arriving at the two lamps can pass through either lamp before continuing its trip around the circuit. The two lamps operate independently and each one draws the current that it normally does when it experiences the voltage drop provided by the rest of the circuit. With both lamps providing a path for current, the current through the rest of the circuit is the sum of the currents through the two lamps.

But when the two lamps are in series with one another, each lamp carries the entire current passing through the circuit. Current arriving at the two lamps must pass first through one lamp and then through the other lamp before continuing its trip around the circuit. There is no need to add the currents passing through the lamps because it is the same current in each lamp. Moreover, the voltage drop provided by the rest of the circuit is being shared by the two lamps so that each lamp experiences roughly half the overall voltage drop. Since lamps draw less current as the voltage drop they experience decreases, these lamps draw less current when they must share the voltage drop. Thus the current passing through the circuit is much less when the two lamps are inserted into the circuit in series than in parallel.

An operational amplifier is an extremely high gain differential voltage amplifier--a device that compares the voltages of two inputs and produces an output voltage that's many times the difference between their voltages. How the operational amplifier performs this subtraction and multiplication process depends on the type of operational amplifier, but in most cases two input voltages control how current is shared between two paths of a parallel circuit. Even a tiny difference between the input voltages produces a large current difference in the two paths--the path that's controlled by the higher voltage input carries a much larger current than the other path. The imbalance in currents between the two paths produces significant voltage differences in their components and these voltage differences are again compared in a second stage of differential voltage amplification. Eventually the differences in currents and voltage become quite large and a final amplifier stage is used to produce either a large positive output voltage or a large negative output voltage, depending on which input has the higher voltage. In a typical application, feedback is used to keep the two input voltages very close to one another, so that the output voltage actually falls in between its two extremes. At that operating point, the operational amplifier is exquisitely sensitive to even the tiniest changes in its input voltages and makes a wonderful amplifier for small electric signals.

VU and dB meters both measure the audio power involved in recording and they both use logarithmic scales to report that power. Because of these logarithmic scales, a factor of 10 increase in power produces an increase of 10 in both the VU reading and the dB reading. For example, -20 dB is 10 times the power of -30 dB. In both measures, the zero is chosen as the highest acceptable power--the highest power for which distortion is acceptable.

Where VU and dB differ is in how they measure audio power. VU is short for "volume units" and it is a measure of average audio power. A VU meter responds relatively slowly and considers the sound volume over a period of time. Its zero is set to the level at which there is 1% total harmonic distortion in the recorded signal. dB is short for "decibels" and it is a measure of instantaneous audio power. A dB meter responds very rapidly and considers the audio power at each instant. Its zero is set to the level at which there is 3% total harmonic distortion. Because of these differences in zero definitions, the dB meter's zero is roughly at the VU meter's +8. Nonetheless, both meters are important and both should be kept at or below zero to avoid significant distortion in a recording. In certain situations, such as when there are sudden loud sounds or with instruments that are very rich in harmonics, it's possible to have the dB meter read above zero even though the VU meter remains below zero.

A stereo contains a power supply that converts 110-volt alternating current into lower-voltage direct current. This direct current is ultimately when powers the speakers. The stereo's power supply first lowers the voltage with the help of a transformer. Alternating current from the power line flows back and forth through a coil of wire in this transformer, the primary coil, and causes that coil to become magnetic. Since the coil's magnetism reverses 120 times a second (60 full cycles of reversal each second), along with the alternating current, it produces an electric field--changing magnetic fields always produce electric fields. This electric field pushes current through a second coil of wire in the transformer, the secondary coil, and transfers power to that current. There are fewer turns of wire in the secondary coil than in the primary coil, so charges flowing in the secondary coil never reach the full 120 volts of the primary coil. Instead, more current flows in the secondary coil than in the primary coil, but that secondary current involves less energy per charge--less voltage. In this manner, power is transferred from a modest current of high voltage charges in the primary coil to a large current of low voltage charges in the secondary coil.

Having used the transformer to produce lower voltage alternating current, the power supply than converts this alternating current into direct current with the help of four diodes and some capacitors. Diodes are one-way devices for electric current and, with four of them, it's possible to arrange it so that the alternating current leaving the transformer always flows in the same direction through the circuit beyond the diodes. The diodes act as switches, always directing the current in the same direction around the rest of the circuit. The capacitors are added to this circuit to store separated electric charge for the times while the alternating current is reversing and the diodes receive no current from the transformer. The capacitors store separated charge while there is plenty of it coming from the transformer and provide current while the alternating current is reversing. Overall, the stereo's power supply is a steady source of direct current.

A speaker produces sound by using magnetic forces to push or pull a thin surface--the speaker cone--toward or away from the listener. As the cone moves forward, it compresses the air in front of it and as the cone moves backward, it rarefies the air in front of it. These compressions and rarefactions are what produce sound. But if you try to drive the cone into motions that are too extreme by turning up the volume of an amplifier too high, the cone will reach the limits of its motion. At that point, the cone may tear away from the electromagnetic coil that pushes it back and forth or it may tear away from the supports at its outer edge. The electromagnetic coil may also burn up because of overheating. All of these failures are lumped together as "blowing a speaker."

My understanding is that a Zobel network consists of a resistor in series with a capacitor and that the capacitor is normally connected to ground. When you attach the free end of this network to a wire carrying an audio signal, the network acts like a frequency-dependent load. At very low frequencies, the capacitor has plenty of time to charge through the resistor and the network has little effect on the audio signal--it acts as though it weren't there. At very high frequencies, the capacitor has no time to charge through the resistor and behaves like a wire. As a result, the network acts as though it were just the resistor connecting the audio signal wire to ground. So the impedance of the Zobel network varies from infinite at low frequencies to become equal to the resistance of the resistor at high frequencies. The crossover between these two behaviors is related to the RC time constant. I think that Zobel networks are used in audio amplifiers to dampen out high frequency oscillations that might occur in the absence of loads at high frequencies.

While some modern car horns are actually specialized computer audio systems, the old-fashioned electromagnetic car horns are still common. An electromagnetic horn uses an electromagnet to attract a steel diaphragm and turns that electromagnet on and off rhythmically so that the diaphragm vibrates. In fact, it uses the diaphragm's position to control the power to the electromagnet. Whenever the diaphragm is in its resting position or even farther from the electromagnet, a switch closes to deliver electric current to the electromagnet. The electromagnet then attracts the diaphragm's center. But when the diaphragm moves closer to the electromagnet, as the result of this attraction, the switch opens and current stops flowing to the electromagnet. Because of this arrangement, the diaphragm moves in and out and turns the electromagnet off and on as it does. The diaphragm's tone is determined by the natural resonances of its surface.

The answer depends a little on which type of transistor is used, so I'll consider only an audio amplifier based on MOSFETs (Metal-Oxide-Semiconductor Field Effect Transistors). One of these three-electrode devices allows a tiny electric charge on its gate electrode to control a substantial current flowing between its source and drain electrodes. In a typical amplifier, the current flowing in the input circuit is allowed to deposit or remove electric charge from the gate electrode(s) of one or more MOSFETs. This action dramatically changes how much current flows in a second circuit. This second circuit is ultimately responsible for the current that passes out of the amplifier and through the speakers that reproduce sound. As the current in the input circuit fluctuates to represent a particular musical passage, the charges on the gates of the MOSFETs also fluctuate and the MOSFETs vary the current through the output circuit and the speakers. Because MOSFETs are so sensitive to even a tiny amount of charge, it doesn't take much current in the input circuit to cause large changes in the current of the output circuit.

In analog audio, the air pressure fluctuations of sound at the microphone are represented by a continuously variable physical quantity such as an electric current, a voltage, or a magnetization. Thus as the air pressure at a tape recorder's microphone rises during one moment of a song, an electric current in the recorder will rise and a region of a magnetic tape surface will become particularly strongly magnetized in a particular direction. Overall, each value of air pressure is converted to a particular value of the physical quantity.

The problem with analog recording is that when the sound is recreated, any defect in the physical quantity representing air pressure will lead to an imperfection in the reproduced sound. For example, if the magnetization of the recording tape has changed slightly due to how it was stored, the sound that the tape recorder produces won't be exactly the same as the sound that the microphone heard. Digital recording avoids this problem by recording the information as bits. The physical quantity such as magnetization is representing bits (which take only two possible values) rather than the air pressure itself (which can take a broad range of values). Minor changes in the physical quantity representing these bits won't change the bits. Thus imperfections in the recording or playback process won't affect the sound quality.

Acoustic "white noise" is a collection of random sounds that together have the same volume at every frequency or pitch. It's defined more accurately as having the same amount of power in each unit of its bandwidth, so that the acoustic power between 20 and 21 cycles per second is the same as the acoustic power between 500 and 501 cycles per second.

Yes. I wouldn't try to detect mechanical contact, because you'd have trouble differentiating between forces exerted on the car by a hand and those exerted on it by sound waves. But you can tell whether a conducting object (such as a person) is near the car by looking at the car's electric properties. If you were to send electric charge on and off the car rapidly with a source of high-frequency alternating current, you would find that the amount of charge that flowed on or off the car during each cycle would change as the person's hand approached the car. That's because the charges on the car would push or pull on charges in the person's hand and the charges in the person's hand would move. In effect, the person's hand would make the car "larger" and it would draw more charge from your current source. Even if the person didn't touch the car, the nearness of the hand and car would change the way current flowed on and off the car. Such a change would be easy to detect with laboratory equipment and could probably be made by cheap consumer equipment, too. The only complications would be in not detecting everything--passing cars for example--and in not damaging the device with static discharges. Still, I think all of that could be done.

The same touch sensors that are used in "touch" lamps or some elevator buttons could be used to sense when you touch a car. A car is essentially insulated from the ground by its rubber wheels, so that when you touch it there is a tendency for electric charge to be transferred between the earth and the car through you. That's why you may receive a shock when you touch a car on a cold winter day. Many electronic devices are capable of detecting this charge transfer (in fact, many of them would be damaged by such sudden and large charge transfers). So building a car touch sensor would be easy. Whether there is a commercial product that does this is another matter, and I am not sure of the answer.

Electrostatic speakers uses the forces between electric charges (so called "electrostatic forces") to move a thin metal diaphragm back and forth rapidly. The motions of this diaphragm compress and rarefy the air in front of it, producing sound. On each side of the diaphragm is a rigid metallic grill that can hold electric charges. When the speaker is silent, the diaphragm has a large positive electric charge on it and both the metal grills have large negative charges on them (it could be the other way around, depending the speaker's exact design). The diaphragm is then attracted equally toward both grills and the electrostatic forces cancel perfectly. The diaphragm doesn't undergo any acceleration. To make the speaker produce sound, the electric charges on the two grills are changed so that the electrostatic forces on the diaphragm don't cancel. Instead, the diaphragm is pulled strongly toward whichever grill has more negative charge on it (or less positive charge). The charges on the grills fluctuate as the music plays and the diaphragm accelerates back and forth between the grills. It pushes on the air as it does and produces sound. You'll notice that the diaphragm is a moving part, so the claim that the speaker has "no moving parts" is misleading. The speaker cone of a conventional speaker only moves back and forth, too, so it has an equal claim to having "no moving parts." The relative expense of an electrostatic speaker comes from the requirement of careful construction and the need for a high voltage adapter to match an amplifier to the speaker.

Some audio amplifiers provide several different outputs, each characterized by the impedance of its expected load (e.g., the impedance of the speaker that you should attach to that output). This impedance measures the relationship between voltage and current that the load needs to function optimally. The higher the impedance, the more voltage the amplifier must provide to propel a particular electric current through the speaker. If the speaker that you attach to the amplifier has the wrong impedance, the amplifier won't be able to deliver its maximum audio power to the speaker and you may damage the amplifier, speaker, or both.

Since a typical household speaker has an impedance of 8 ohms, you should connect it to an amplifier's 8 ohm output. However, if you connect more than one speaker to the same output, you should be careful to determine the combined impedance. For example, two 8-ohm speakers in series have a combined impedance of 16 ohms while two 8-ohm speakers in parallel have a combined impedance of 4 ohms. Many amplifiers are designed to accommodate these arrangements.

When a distribution amplifier must send current long distances through thin wires, it will often use higher voltages and lower currents to minimize power losses in the wires. Such an amplifier expects its load to have an unusually large impedance. In this situation, the speaker that is used must either have a large impedance, so that it can use this high voltage/low current power directly, or there must be an impedance matching transformer between the amplifier and the speaker.

A typical amplifier examines the current flowing in its input circuit and produces a current in its output circuit that's proportional to but much larger than this input current. The factor by which the amplifier multiplies the input current to produce the output current is sometimes called the amplifier's "current gain." The tiny currents produced by a microphone attached to an audio amplifier's input circuit are boosted into huge currents that flow through speakers attached to the amplifier's output circuit. Since your voice is controlling these large currents, the speakers reproduce the sound of your voice.

While there are many techniques used to amplify currents, most modern audio amplifiers use transistors to do the amplification. A transistor is a device that permits a small current or electric charge to control the flow of a much larger current. The transistors inside the amplifier examine the current in the amplifier's input circuit and these transistors control the current passing through the amplifier's output circuit. Because the current in the output circuit needs electric power to continue flowing, a power supply inside the amplifier provides that current with power. As you talk into the microphone, the transistors adjust the current flowing through the output circuit so that that current is proportional to the current flowing through the input circuit.

You can calculate the impedance of a collection of speakers the same way you would calculate the resistance of a collection of resistors. Each time two speakers are connected in series, so that the electric current must pass through one and then the other to get to its destination, their impedances add. Thus two 4-ohm speakers in series are equivalent to one 8-ohm speaker (4 ohm + 4 ohm = 8 ohm). Each time two speakers are connected in parallel, so that the electric current can pass through one or the other to get to its destination, the reciprocals of their impedances add to give the reciprocal of their overall impedance. Thus two 4-ohm speakers in parallel are equivalent to one 2-ohm speaker (1/4 ohm + 1/4 ohm = 1/2 ohm). Once you have figured out the impedance of a pair of speakers, you can treat it as though it were one speaker and proceed to figure out the impedance of a larger group of speakers. For example, four 4-ohm speakers in series have an overall impedance of 16 ohms and four 4-ohm speakers in parallel have an overall impedance of 1 ohm.

Light emitting diodes are diodes that have been specially designed to emit light rather than heat during their operations. Whenever current is flowing through a diode, electrons are moving from the n-type semiconductor on one side of the diode's p-n junction to the p-type semiconductor on the other side of the junction. Once an electron (which is negatively charged) arrives in the p-type semiconductor, it's attracted toward an electron hole (which is positively charged) and the two move together. The electron soon fills the hole and it releases a small amount of energy when it does. In a normal diode, electrons lose energy at a rate of 0.6 joules of energy per coulomb of charge as they recombine with the electron holes. That means that the current flowing through the normal diode loses 0.6 volts as it flows through the diode. The missing energy becomes thermal energy or heat.

But in a light emitting diode (an LED), each electron that arrives in the p-type semiconductor after crossing the p-n junction recombines with an electron hole in a remarkable way. It gives up its extra energy as light! Each time an electron and an electron hole recombine, they emit one particle of light, a photon, and the frequency, wavelength, and color of that light depends on the amount of energy given up by the electron as it falls into the electron hole. The semiconductor material from which an LED is made has a characteristic called its band gap. This band gap measures the energy needed to pull an electron away from an electron hole in the material. If this band gap is small, the LED will emit infrared light. If this band gap is larger, the LED will emit red, orange, yellow, green, or even blue light (the farther to the right in that list, the more energy is required). Because each electron loses more energy in recombining with an electron hole in an LED than it would in a normal diode, the current flowing through an LED loses more voltage (typically 2 volts for red LEDs and as much as 4 volts for blue LEDs) than does the current flowing through a regular diode (typically 0.6 volts).

Physicists, chemists, materials scientists, and engineers have been working for years to perfect the materials used in LEDs, making them more and more efficient at turning the electrons' energies into light. Until recently, there were no suitable materials from which to build blue LEDs, but recent developments of large band gap semiconductors have made blue LEDs possible. In fact, even blue laser diodes are now being made. A laser diode is a specially designed LED in which all of the photons are copies of one another rather than being emitted independently by the individual electrons as they drop into their respective electron holes.

One final note: it's now possible to obtain a "white" LED! This device is actually a blue LED, combined with a fluorescent phosphor that converts the blue light into white light.

Light emitting diodes (LEDs) that emit more than one color are actually two different LEDs connected to a single circuit in opposite directions. When current flows in one direction around that circuit, one of the LEDs emits light. When the current reverses directions, the other LED emits light. And when the current reverses directions rapidly, both LEDs emit light alternately. If one LED emits red light and the other green light, then the overall device will appear yellow or orange when they are both operating alternately in rapid sequence. The amount of light that an LED emits depends on the current flowing through it--the more electrons that are falling into holes in the p-type semiconductor, the more light that's being emitted. However, many devices that use LEDs just turn them on or off because that's easier than controlling the current flowing through them. Some day, flat panel displays may use three colors of LEDs--red, green, and blue--in order to present full color images like those on a current television screen. For that scheme to work, the LEDs must be able to emit different brightnesses, so the current flowing through each one must be adjustable.

As the steel strings of an electric guitar vibrate, they move back and forth across electromagnetic pickups on the guitar's surface. Each of these pickups consists of a coil of wire with a permanent magnet passing through its center. This permanent magnet has a north magnetic pole at one end and a south magnetic pole at the other end. Surrounding the permanent magnet are lines of magnetic flux that arc gracefully through space from the magnet's north pole to its south pole. These magnetic flux lines are associated with the forces that magnets exert on one another. Some of these flux lines pass very near the permanent magnet on their way from the north pole to the south pole and thus pass inside the coil of wire around the magnet. Other flux lines arc far outward and pass outside the coil of wire around the magnet. And a few of the flux lines pass through the steel string that lies just above one pole of the permanent magnet. Steel is a ferromagnetic metal, meaning that it easily develops strong north and south poles of its own when exposed to another magnet. This ferromagnetism is the result of a remarkable ordering process that takes place among the electrons inside the steel. The steel string is magnetized by its proximity to the permanent magnet in the pickup and it interacts strongly with the magnetic flux lines that pass near it. Some flux lines leaving the north pole of the permanent magnet connect to the south pole of the magnetized string and an equal number of flux lines leaving the north pole of the magnetized string connect to the south pole of the permanent magnet. Thus when the steel string vibrates back and forth, it pulls some of the flux lines with it. The paths that these flux lines take shift back and forth rhythmically as the string vibrates.

Whenever magnetic flux lines move, they create electric fields. An electric field is a phenomenon that exerts forces on charged particles, such as the mobile electrons in the coil of wire around the permanent magnet. As the string vibrates and the magnetic flux lines shift back and forth with it, electric fields appear in the wire coil and begin to push electrons through that coil. These electrons flow back and forth in the wire as the string vibrates. Wires connecting the pickup's coil to an electronic audio amplifier carry these moving electrons (actually an electric current) to the amplifier, where they are detected and used to control a much larger electric current. When this amplified current is sent through a speaker, the speaker produces a very loud sound that's an amplified version of the sound that the string itself is making as it pushes weakly on the air.

A diode is normal built by touching two different pieces of semiconductor together to form what is called a "p-n junction." Semiconductors are materials that are in between good conductors and good insulators. A pure semiconductor is a very poor conductor of electricity. With careful chemical processing, a semiconductor can be made into n-type semiconductor--a semiconductor that contains a small number of mobile electrons that permit it to carry electric current. With different processing, a semiconductor can also be made into p-type semiconductor--a semiconductor that contains a small number of mobile holes for electrons that permit it to carry electric current. It may seem strange that a hole for an electron can allow electricity to flow, but imagine a highway packed with cars (electrons) bumper to bumper. If there are a couple of empty places (holes) in the bumper-to-bumper traffic, then cars (electrons) can rearrange enough that the traffic can flow. Both mobile electrons and mobile holes allow these two chemically treated semiconductors to carry current.

When an n-type semiconductor touches a p-type semiconductor, a diode is formed. The mobile electrons at the edge of the n-type semiconductor flow over the boundary (a p-n junction) and fill the mobile holes at the edge of the p-type semiconductor. This rearrangement creates a depletion region--a region near the p-n junction in which there are neither mobile electrons nor mobile holes. This depletion region normally won't carry electricity at all. But if you push electrons onto the n-type semiconductor, they will flow toward the p-n junction and replenish the missing mobile electrons. As these mobile electrons approach the p-n junction, they will repel the electrons that are filling the mobile holes on the p-type side of the junction and reopen the mobile holes. Electrons will begin to cross the p-n junction and current will flow through the diode. However, if you push electrons onto the p-type semiconductor, they will fill even more of the mobile holes there and the depletion region near the p-n junction will grow larger and more uncrossable. No current will flow through the diode. Thus a diode (a p-n junction) only carries current in one direction--electrons can only flow from the n-type semiconductor side to the p-type semiconductor side.

There are many types of transistors, so I will only describe an n-channel Metal-Oxide-Semiconductor Field Effect Transistor, or n-channel MOSFET. In this device, three layers of semiconductors are sandwiched together: an n-type piece (the source), a long, thin p-type piece (the channel), and another n-type piece (the drain). Two p-n junctions form between these three components and, since the junctions are arranged in opposite directions, they completely block current flow from the source through the channel to the drain. But a metal surface (the gate) that's separated from the channel by an extremely thin layer of oxide insulator can control the number of electrons on the channel material. If you put even a tiny bit of positive charge on the gate, it will attract electrons onto the channel and turn it from p-type semiconductor to n-type semiconductor. When that happens, both p-n junctions vanish and current can flow from the source to the drain. The MOSFET goes from being an insulating device when there is no charge on the gate to a conductor when there is charge on the gate! This property allows MOSFETs to amplify signals and control the movements of electric charge, which is why MOSFETs are so useful in electronic devices such as stereos, televisions, and computers.

The only difference between a well-designed tube amplifier and a well-designed solid-state amplifier is the device doing the amplification. In fact, a vacuum tube and an metal-oxide-semiconductor field-effect-transistor or MOSFET are extremely similar in behavior, so that amplifiers built with the two devices can be extremely similar. If these amplifying devices are used properly in a good amplifier, that amplifier should only boost the power of its input signal and shouldn't add anything that wasn't present in the input signal. As a result, you shouldn't be able to tell whether the audio amplifier you are listening to is based on tubes or on solid-state components.

During the silent passages of the music, the amplifier does not vary the amount of current passing through the speaker so that the speaker doesn't move and doesn't produce sound. To conserve energy and to avoid heating up the speaker, a good amplifier doesn't send any current through the speaker during a quiet passage. Whether or not the amplifier actually consumes power during the quiet passage depends on the exact design of the amplifier. Some stereo experts claim that they can hear the differences between amplifiers the do or do not consume power with their output transistors during the quiet times and claim that the power wasting amplifiers sound better.

As the current passing through the speaker's coil changes, the speaker cone moves back and forth toward or away from the speaker's permanent magnet. This moving cone pushes or pulls on the air, creating compressions and rarefactions that propagate through the air as sound.

The coil in a microphone is attached to a movable surface that is pushed back and forth by the sound. Near the coil is a magnet so that, as the coil moves, the magnet induces electrical currents in it. Whenever a magnet moves past a coil of wire or a coil of wire moves past a magnet, a current is induced in that coil of wire.

A decade or two ago, it was important to match the power transistors used to control currents leaving an audio amplifier. If the transistor that controlled current flowing one direction through the speaker was significantly different from the transistor that controlled current flowing in the opposite direction, then the sound reproduction would be poor. That's because the current flows would be asymmetric and asymmetric currents lead to distorted sounds from the speaker. The most common measure of this sort of error is called "total harmonic distortion," an indication of how much power the amplifier puts into unwanted high frequency currents. Without carefully matched power transistors, an amplifier might put several percent of its power into these harmonic frequencies.

However, modern audio amplifiers generally use feedback techniques to correct for their own internal imperfections. They can compensate so well for mismatches in their components that total harmonic distortion has virtually disappeared from amplifiers. Amplifiers are still rated according to total harmonic distortion, but now it is rarely more than a few thousandths of a percent and depends more on the feedback techniques used than on the perfection of the power switching components. In short, the power transistors in modern amplifiers don't have to be matched well any more.