A louder voice needs an energy source
A microphone produces a weak electrical signal. To make that signal drive a loudspeaker, a circuit must deliver more alternating electrical power than the microphone supplies. The extra energy comes from a battery or another power supply. The transistor’s achievement was to let a small electrical change regulate that larger flow inside a piece of semiconductor. A changed resistance or a rectifying contact was interesting, but neither observation alone established useful amplification.
In December 1947, John Bardeen and Walter Brattain obtained amplification with two closely spaced contacts on germanium at Bell Laboratories. The experiment connected an input disturbance to an output in a separately biased circuit. It therefore asked a sharper question than whether germanium conducts: could one contact control what another contact collects, with enough gain to make a signal useful? Understanding that question also explains why measuring the circuit mattered as much as assembling the famous small device.
Sources: [1][2]
A failed field-effect idea exposed the surface
A semiconductor contains mobile electrons and holes. A hole is an unoccupied electronic state that moves through the crystal as neighboring electrons fill it; treating it as a positive carrier greatly simplifies the description. An external electrode might seem able to attract more carriers into a region and change its conductivity, much as a capacitor changes the charge on a plate. That was the attraction of the early field-effect proposals.
The expected modulation was much weaker than the simple picture suggested. Bardeen explained that electronic states associated with the surface could capture much of the added charge. Changing the external field then changed the occupation of those surface states instead of efficiently changing the mobile charge in the crystal. Experiments on surface preparation and contact behavior were consequently central. The failure supplied a physical clue: the boundary of the semiconductor had its own electronic behavior, and a successful device had to work with that behavior.
Sources: [1][2]
Two nearby contacts made injection useful
In the point-contact device, metal points contacted the same germanium body, with a larger contact providing the common connection. One point was forward biased and another was used as the collector. In the relevant n-type germanium experiments, electrons were the majority carriers in the bulk, but injection of holes could strongly alter conduction near the collector. The polarity of those holes mattered: the effect could not be understood by assuming that every mobile charge was simply another conduction electron.
The contacts had to be close enough for the disturbance from injection to influence collection before the excess carriers disappeared. Surface conditions and the preparation of the collector contact also mattered. That sensitivity made a working amplifier a remarkable result, but it made the point-contact transistor difficult to describe with a single ideal law. In particular, one should not identify its common-contact current transfer directly with the base-current gain of a later junction transistor; the geometries and operating mechanisms have important differences.
Sources: [1][2]
Measure changes around a chosen operating point
To test amplification, first establish steady bias currents and voltages. Superimpose a small input signal and measure how the output changes under a specified load. The incremental current gain compares changes about that operating point, not arbitrary ratios of total currents. Voltage gain additionally depends on the impedances that turn those current changes into voltages. A large voltage ratio into almost no load does not by itself establish a large delivered power.
For a sinusoidal signal applied to a resistive load, the output power follows from the mean-square load voltage divided by resistance. In the expression, the subscript “rms” denotes root-mean-square voltage, and the two resistances describe the input and output loads used in that comparison. Every quantity must refer to the same signal frequency and a range where distortion is acceptably small. This procedure distinguishes an amplifier from a device that merely shifts a meter reading, and it identifies the bias supply as the source of the additional signal energy.
Sources: [1][2]
A thin base gave the junction transistor a clearer mechanism
Shockley’s junction-transistor theory organized carrier control into adjoining regions with different doping. In an NPN transistor, an electron-rich emitter and collector flank a thin, hole-rich base. Forward bias at the emitter-base junction injects electrons into the base. There they are minority carriers. Their concentration varies with position, driving diffusion, and a reverse-biased collector junction collects many that survive the journey.
The base must be thin compared with the distance over which excess carriers typically disappear by recombination. That distance is described by a diffusion length: the square root of diffusion coefficient times carrier lifetime. In this description, the coefficient has units of area per time and lifetime has units of time, giving a length. A thinner base or a longer lifetime can improve collection, although actual devices introduce tradeoffs involving resistance, capacitance and speed. The mechanism explains how a small base current can accompany a much larger collector current without requiring the base source to supply the output power.
Sources: [2][1]
Read the knee before trusting the gain law
An output-characteristic measurement holds the base current fixed and sweeps collector-emitter voltage. In the forward-active region the collector current is approximately proportional to base current and only weakly dependent on output voltage. Near low output voltage the curve bends: the collector junction is no longer in the required reverse-biased state, and the transistor enters saturation. At still larger voltages, real devices eventually encounter heating and breakdown limits.
Consider a deliberately simplified circuit with a six-volt supply, a one-kilohm collector resistor and an assumed forward gain of one hundred. A base current of twenty microamperes predicts about two milliamperes of collector current and about four volts across the transistor. Increasing the base current to eighty microamperes would naively predict eight milliamperes, but the resistor and supply can furnish less than six. The operating point moves toward saturation instead. The supply voltage equals the resistor drop plus the collector-emitter voltage. That circuit constraint must be satisfied together with the device characteristic.
Sources: [2]
The discovery changed how semiconductor evidence was assembled
The history joined several kinds of evidence. Surface experiments explained why an attractive field-effect proposal underperformed. The point-contact amplifier demonstrated useful control in a solid-state device. Studies of injected minority carriers and junction theory then provided a more systematic route to designing other structures. These were successive answers to different questions, not interchangeable versions of a single demonstration.
The 1956 Nobel Prize recognized Bardeen, Brattain and Shockley for semiconductor research and the transistor effect. Its significance reaches beyond making equipment smaller. It established that deliberately prepared regions, interfaces and nonequilibrium carrier populations could govern a circuit’s behavior. Modern transistor families use different forms of that control. Their success still depends on relating materials and geometry to measured electrical characteristics, rather than identifying a promising material and assuming that amplification follows automatically.
Sources: [1][2][3]
Turning carrier control into a circuit function
By organizing injection, diffusion and collection, a transistor lets a weak signal regulate a larger flow of energy from a supply. Bias and load determine whether that control provides amplification or drives the device toward cutoff and saturation. From surface-state experiments to junction devices, carrier distributions became something engineers could deliberately design, preparing the way for integrated circuits.
Sources: [2][3]