Electrical Engineering › Section P · Semiconductor theory, diodes, and transistors › Topic P.9
Common-emitter, common-collector, common-base, and cascode amplifiers
A single BJT can amplify in three basic ways, depending on which terminal the input and output share. Each gives a different mix of voltage gain, current gain and impedance, and two of them stack into the cascode, an amplifier built for speed.
Three ways to connect one transistor
An amplifier has an input port and an output port, so with three transistor terminals one must be shared. The configuration is named after that terminal, judged for the signal alone, with the dc supply treated as ac ground. In the common-emitter (CE) amplifier the signal enters the base and leaves the collector. In the common-collector (CC) amplifier, also called the emitter follower, it enters the base and leaves the emitter. In the common-base (CB) amplifier it enters the emitter and leaves the collector. NPN and PNP versions behave the same way, with every polarity reversed.
Bias first, then signal
A BJT conducts one way only, so it cannot amplify a bare ac signal. Apply a sine wave straight to the base and the transistor cuts off for every half-cycle that fails to forward-bias the base-emitter junction; the output is a train of half-waves. The cure is a dc bias that holds the transistor in its active region with no signal, at a quiescent or Q point, so the signal swings the currents around it. Capacitors pass the signal while blocking the dc. Setting that bias is the subject of "Biasing techniques, biasing calculations, and coupling"; here the bias current is given.
The common-emitter amplifier
For small signals the base-emitter junction behaves like a small resistance in the emitter,
at about room temperature, where \(I_E\) is the dc emitter current. A base signal appears across \(r_e\) plus any unbypassed emitter resistor \(R_E\), and almost the same signal current flows in the collector resistor \(R_C\), so
The minus sign means inversion: as the base rises, the collector falls, a 180° shift. A load \(R_L\) replaces \(R_C\) with \(R_C \parallel R_L\). With large voltage and current gain, the CE stage gives the most power gain of the three.
Worked example: a common-emitter stage
A CE stage on 12 V has \(I_E = 1.2\,\text{mA}\), \(R_C = 3.9\,\text{k}\Omega\), unbypassed \(R_E = 220\,\Omega\) and β = 150. Find the gain and output for 50 mV peak in. 1. Emitter resistance: \(r_e = 26/1.2 = 21.7\,\Omega\). 2. Gain: \(A_v = -3900/(21.7 + 220) = -16.1\). 3. Output: \(16.1 \times 50\,\text{mV} = 0.807\,\text{V}\) peak, inverted. 4. Headroom: the collector sits at \(12 - 1.19 \times 3.9 = 7.35\,\text{V}\), so a 0.8 V swing fits easily. 5. Contrast: with \(R_E\) bypassed the gain would be \(3900/21.7 = 180\), but it would then hang on \(r_e\), which shifts with the current. A 10 kΩ load would cut the gain to \(2806/241.7 = 11.6\).
The common-collector amplifier
In the emitter follower the collector goes straight to the supply and the load sits in the emitter. Because the base-emitter junction holds a nearly constant 0.7 V, the emitter follows the base about 0.7 V lower:
just under 1 and non-inverting. The gain is in current: the load carries the whole emitter current, so the current gain is \(\beta + 1\). The input resistance at the base is high, about \((\beta + 1)(r_e + R_E)\), and the output resistance low, about \(r_e\) plus the source resistance over \(\beta + 1\). So the follower buffers a weak source from a heavy load.
Worked example: a follower as a buffer
A sensor with 10 kΩ internal resistance must drive a 1.0 kΩ load. 1. Directly, the divider passes only \(1/(1 + 10) = 0.0909\). 2. Insert a follower, β = 100, with the load as its emitter resistor and \(I_E = 5.0\,\text{mA}\): \(r_e = 5.2\,\Omega\). 3. Input resistance: \(101 \times 1005 = 101.5\,\text{k}\Omega\), so the input divider passes \(101.5/111.5 = 0.910\). 4. Follower gain: \(1000/1005.2 = 0.995\). 5. Overall: \(0.910 \times 0.995 = 0.906\), ten times better than the direct link. 6. Check: the output resistance the load sees is about \(5.2 + 10{,}000/101 = 104\,\Omega\), small beside 1 kΩ.
Darlingtons and follower regulators
When one follower's current gain is not enough, a Darlington pair feeds the emitter of one transistor into the base of a second, so the gains multiply:
Two transistors with β = 100 give 10,201, but the output sits two base-emitter drops, about 1.4 V, below the input. Followers also strengthen the Zener regulator of "Inductor commutating and diode switching circuits, and Zener diodes": the Zener holds the base steady and supplies only base current, while the transistor carries the load. A 6.2 V Zener gives about 5.5 V out, so choose a Zener about 0.7 V above the output wanted.
The common-base amplifier
In the common-base stage the base is held at ac ground, the signal drives the emitter, and the output is taken at the collector. The source must now carry the whole emitter current, so the current gain is the ratio
just under 1: 0.993 for β = 150. The voltage gain is high and non-inverting, about \(R_C/r_e\): 181 for 4.7 kΩ at 1 mA. The input resistance is very low, about \(r_e\), which suits a 50 Ω radio-frequency cable but loads most sources. Its strength is speed: the grounded base screens output from input, so it works to higher frequencies than a CE stage.
The cascode and the Miller effect
The CE stage is slowed by the small collector-base capacitance \(C_{cb}\). With an inverting gain \(A_v\), the voltage across \(C_{cb}\) swings \(1 + |A_v|\) times the input, so the input sees a much larger Miller capacitance:
A 4 pF \(C_{cb}\) with a gain of −150 looks like 604 pF. The cascode stacks a CB transistor on a CE transistor. The CE stage's load is now the CB emitter, about \(r_e\), so its gain is about −1 and its Miller capacitance about \(2C_{cb}\), 8 pF. The CB stage supplies the gain, about \(-R_C/r_e\) overall, so the cascode keeps the CE input resistance and gains CB bandwidth.
Choosing a configuration
Typical small-signal behaviour, with no load and the emitter resistor bypassed where one is used:
The CE stage is the general gain block, the CC stage the buffer, and the CB stage the wideband, low-impedance input. Impedances including bias resistors are worked out in "Feedback, amplifier impedances, current mirrors, ratings, and quirks".
Common mistakes to avoid
First, dropping the minus sign of the CE gain. Second, expecting voltage gain from an emitter follower; its value is current gain and low output resistance. Third, ignoring \(r_e\) when \(R_E\) is small or bypassed, which overstates gain badly. Fourth, using the unloaded gain with a load connected: use \(R_C \parallel R_L\). Fifth, forgetting a Darlington's 1.4 V drop. Sixth, multiplying \(C_{cb}\) by β instead of \(1 + |A_v|\).
Key takeaways
Configurations are named for the terminal input and output share, and all need dc bias. With \(r_e \approx 26\,\text{mV}/I_E\), the CE gain is \(-R_C/(r_e + R_E)\), inverting. The CC stage follows its input about 0.7 V lower: gain just under 1, current gain β + 1, high input and low output resistance. A Darlington multiplies gains but loses 1.4 V. The CB stage has current gain α = β/(β + 1), high non-inverting voltage gain and input resistance near \(r_e\). The Miller effect multiplies \(C_{cb}\) by \(1 + |A_v|\); the cascode avoids it.
Flashcards
Flashcard
Input at base, output at collector; inverting voltage gain
Test your understanding
Question 1 of 5
A common-emitter stage has I_E = 2.0 mA, R_C = 4.7 kΩ and an unbypassed emitter resistor of 100 Ω. Including r_e, what is its voltage gain?
Question 2 of 5
An emitter follower's base carries a 1 V peak sine wave on a 4.0 V dc bias. What appears at the emitter?
Question 3 of 5
A Darlington pair uses transistors with β₁ = 80 and β₂ = 50. What is its overall current gain as a follower?
Question 4 of 5
Which set of properties describes a common-base amplifier?
Question 5 of 5
A CE stage with voltage gain −99 uses a transistor whose collector-base capacitance is 3 pF. What Miller capacitance does the input see?
