Emitter Follower
An emitter follower is a transistor buffer circuit in which the output, taken from the emitter, follows the input applied to the base, lower by roughly one base-emitter drop (about 0.6 volt for silicon). It has no voltage gain to speak of but has a large current gain, so it converts a signal from a high-impedance source into the same signal at a low impedance, able to drive heavy loads. As a circuit it is identical to the common-collector configuration; the name "emitter follower" simply emphasises the job it does in practice, which is buffering and driving.
The emitter follower works because the base-emitter junction acts like a fixed voltage step, so whatever is at the base reappears at the emitter, 0.6 volt lower. If the emitter is pulled below base voltage minus 0.6 volt, the transistor conducts more and lifts it; if it rises above, the transistor conducts less and lets the emitter resistor or load pull it back down. The result is a self-correcting loop with a gain of RL / (RL + re), where re is about 26 millivolts divided by emitter current — about 0.97 to 0.99 in typical use. The offset is a design nuisance: a 0 to 5 volt signal into the base emerges as roughly −0.6 to 4.4 volts at the emitter, so the follower is often paired with a bias that places the base about 0.6 volt above the desired output. Its stage is also asymmetric: a single npn follower can source current to the load through the transistor, but can sink current only through the emitter resistor.
The impedance transformation is the follower's reason for existing. The impedance seen at the base is about (β + 1) times the resistance at the emitter, while the impedance seen looking into the emitter is the source resistance divided by (β + 1) plus the tiny internal re. A sensor with an output resistance of 100 kΩ can drive a 100 Ω load through a follower with only modest loss, which a direct connection would reduce to almost nothing. Typical practical uses include buffering an oscillator or a filter so that its characteristics are not changed by what follows it, driving cables and headphones, isolating a sensitive circuit from a load, and acting as the pass element of a simple series regulator. Two transistors in the Darlington arrangement, or a Sziklai (complementary) pair, multiply beta further for very high input impedance, at the cost of a larger offset of about 1.2 volts for the Darlington.
The push-pull follower is the classic power output stage, and it introduces the well-known problem of crossover distortion. A single follower conducts on only one polarity, so an output stage pairs an npn transistor, which sources current on positive swings, with a pnp transistor, which sinks current on negative ones. If both bases are simply tied together, neither transistor conducts until the input has moved about 0.6 volt away from zero, so the output has a flat spot at the zero crossing and the waveform is visibly distorted, especially at low levels where the dead zone is a large fraction of the signal. The remedy is class AB biasing: a pair of diodes or a transistor VBE multiplier between the two bases supplies about 1.2 volts so that each device conducts a small idle current. The diode or multiplier is thermally coupled to the output transistors so that the idle current does not run away as they warm up, and small emitter resistors add further stability. The same arrangement, with MOSFETs in place of bipolar devices, forms the source-follower output of many modern amplifiers.
The follower brings its own pitfalls, and the main one is that "no voltage gain" does not mean "no gain". The stage has power gain, which is why it can oscillate: a capacitive load such as a long cable combined with a high-impedance source can create a resonance, and a small series resistor at the base is the usual cure. Another is the thermal one: because the base-emitter voltage falls about 2 millivolts per degree Celsius, an output stage biased too aggressively with poor thermal coupling will increase its idle current as it warms, which raises dissipation and can end in thermal runaway. Followers are also limited in how far the output can swing toward the supply, since the transistor needs about 0.6 volt plus saturation headroom, which is a concern in low-voltage designs. Rail-to-rail op-amp output stages are in part an attempt to escape that limit by using common-emitter or common-source outputs instead.
Read the emitter follower as an impedance transformer and current booster with a built-in 0.6-volt offset: it adds no voltage, only the muscle to deliver it, and in push-pull form it is the output stage on which most audio power amplifiers and integrated-circuit op-amp outputs have been built.