Electrical Engineering › Section P · Semiconductor theory, diodes, and transistors › Topic P.10
Biasing techniques, biasing calculations, and coupling
An amplifier is only as good as its operating point. This topic shows how a few resistors set that point and keep it steady while β varies threefold, and how signals are coupled in and out without disturbing it.
Why bias needs a circuit
"Common-emitter, common-collector, common-base, and cascode amplifiers" showed that a BJT amplifier needs a dc bias to sit in its active region, at a quiescent or Q point, before any signal arrives. A battery in series with the input works on paper, but a real amplifier sets the bias from its one supply with resistors. The hard part is stability. Small-signal transistors have β of about 50 to 300, and one type can spread threefold; β rises with temperature, and \(V_{BE}\) falls about 2 mV per degree Celsius of warming. A good bias network holds the collector current nearly constant despite all three.
Classes of operation
The Q point decides how much of each cycle the transistor conducts. Class A conducts all cycle and reproduces the waveform faithfully, but it draws current with no signal, so its efficiency is at most 25% with a resistive collector load, 50% with transformer coupling. Class B is biased at cutoff and conducts half the cycle; two transistors in push-pull share the halves, for at most 78.5%. Class AB conducts slightly more than half to smooth the hand-over. Class C conducts less than half into a tuned tank that rings through the rest, suiting constant-amplitude radio signals such as FM. Class D switches fully on and off, often above 90% efficient.
The dc load line
For a common-emitter stage with \(R_C\) and \(R_E\), and \(I_E \approx I_C\), Kirchhoff's voltage law around the collector loop gives
Plotted on the characteristic curves this is a straight line, the dc load line. Its ends are cutoff, \(I_C = 0\) and \(V_{CE} = V_{CC}\), and saturation, \(V_{CE} \approx 0\) and \(I_C = V_{CC}/(R_C + R_E)\). The Q point lies on it. For class A, a Q point near the middle allows the largest symmetrical swing; near either end, one peak clips first. A larger \(R_C\) lowers the saturation end, so the stage saturates at a smaller base current.
Worked example: base bias versus collector feedback
A stage on 12 V has \(R_C = 2.7\,\text{k}\Omega\) and should run at 2 mA with β = 100. 1. Base bias, one resistor from supply to base: \(R_B = 100 \times 11.3/2\,\text{mA} = 565\,\text{k}\Omega\); use 560 kΩ. Then \(I_C = 2.02\,\text{mA}\). 2. Swap in a β = 300 part: active mode would need 6.05 mA, beyond the 4.37 mA saturation limit, so the stage saturates and stops amplifying. 3. Collector feedback, \(R_B\) from collector to base:
With 270 kΩ, \(I_C = 2.09\,\text{mA}\) at β = 100 and 3.14 mA at β = 300. 4. Check: tripling β raised the current only 1.5 times, because more \(I_C\) lowers the collector voltage feeding the base.
Voltage-divider bias
The standard circuit adds an emitter resistor \(R_E\) and feeds the base from a divider, \(R_1\) from \(V_{CC}\) to the base and \(R_2\) from the base to ground. Replace the divider by its Thevenin equivalent, as in "Thevenin's theorem": \(V_{TH} = V_{CC}R_2/(R_1 + R_2)\) and \(R_{TH} = R_1 \parallel R_2\). Kirchhoff's voltage law around the base loop then gives
When \(R_E\) is much larger than \(R_{TH}/(\beta + 1)\), β barely matters. The emitter resistor is negative feedback: if the current rises, the emitter voltage rises and squeezes \(V_{BE}\). Common rules: make \(V_E\) about a tenth of \(V_{CC}\), and \(R_{TH}\) at most a tenth of \(\beta R_E\).
Worked example: a divider-biased stage
A stage on 12 V has \(R_1 = 56\,\text{k}\Omega\), \(R_2 = 12\,\text{k}\Omega\), \(R_E = 1.0\,\text{k}\Omega\) and \(R_C = 3.9\,\text{k}\Omega\). 1. Thevenin: \(V_{TH} = 12 \times 12/68 = 2.12\,\text{V}\) and \(R_{TH} = 9.88\,\text{k}\Omega\). 2. At β = 100: \(I_E = 1.418/(1000 + 97.8)\) A = 1.29 mA. 3. At β = 300: \(I_E = 1.418/(1000 + 32.8)\) A = 1.37 mA, a rise of only about 6%. 4. At β = 100, \(I_C = 1.28\,\text{mA}\) and
5. Check: the load line saturates at \(12/4.9\,\text{k}\Omega = 2.45\,\text{mA}\), so this Q point sits near the middle, good for class A.
Bypassing the emitter resistor
The emitter resistor that steadies the bias also cuts the ac gain to \(R_C/(r_e + R_E)\), only about 3.8 in the stage above (\(r_e\) = 20.1 Ω). A bypass capacitor \(C_E\) across \(R_E\) shorts it for signals while dc still flows through \(R_E\), so the bias stays stable and the gain rises towards \(R_C/r_e\), about 194. \(C_E\) must be large, because its reactance must be small beside the low resistance seen from the emitter, about \(r_e\) plus the base-side resistance over \(\beta + 1\). With a 600 Ω source that is about 25 Ω here, so a 20 Hz corner needs about 320 µF; use 330 µF. Bypassing only part of \(R_E\) gives a fixed, moderate gain.
Capacitor coupling
A coupling capacitor in series with the input passes the signal but blocks dc, so source and divider bias cannot upset each other. At the output it keeps the collector's dc off the load, and between stages it lets each be biased independently. The cost is a high-pass filter, as in "R-C low-pass and high-pass filters": with total resistance \(R\) on either side of the capacitor,
Below \(f_c\) the gain falls, the amplifier cannot pass dc at all, and slow waveforms such as low-frequency square waves droop. An oscilloscope's ac-coupled input does the same.
Worked example: sizing a coupling capacitor
The divider-biased stage above, with \(R_E\) bypassed and β = 100, is driven from a 600 Ω source and must pass 20 Hz. 1. Emitter resistance: \(r_e = 26/1.29 = 20.1\,\Omega\). 2. Resistance into the base: \(101 \times 20.1 = 2.03\,\text{k}\Omega\). 3. Input resistance, base in parallel with the divider: \(9.88 \parallel 2.03 = 1.69\,\text{k}\Omega\). 4. The capacitor sees \(R = 600 + 1690 = 2290\,\Omega\), so \(C = 1/(2\pi \times 20 \times 2290) = 3.48\,\mu\text{F}\). 5. Choose 4.7 µF: \(f_c = 14.8\,\text{Hz}\), safely below 20 Hz, as a larger capacitor should give. Input resistance is treated fully in "Feedback, amplifier impedances, current mirrors, ratings, and quirks".
Direct and transformer coupling
Direct coupling joins stages with wires or resistors and no capacitors, so it amplifies dc and has no low-frequency limit. But bias drift in an early stage is amplified by every later one, so it needs careful design or feedback. Integrated circuits use it almost everywhere, because large capacitors cannot be built on a chip. Transformer coupling passes only changes in current, so the load gets pure ac, and its turns ratio can match a low-resistance load. Transformers are bulky, struggle to cover a wide audio band, and primary dc pushes the core towards saturation; they survive mainly in tuned radio-frequency stages.
Common mistakes to avoid
First, designing base bias for one β; a different transistor can move the Q point into saturation. Second, ignoring the base current's loading of the divider: use \(R_{TH}/(\beta + 1)\), not the unloaded divider voltage alone. Third, forgetting \(V_{BE}\) when finding \(I_E\). Fourth, leaving \(R_E\) unbypassed and wondering why the gain is small. Fifth, sizing a coupling capacitor from the source resistance alone; the amplifier's input resistance is in the loop too. Sixth, omitting \(2\pi\) from \(f_c\).
Key takeaways
Bias sets the Q point, which must survive threefold β spreads and temperature. Classes A, B, AB, C and D trade fidelity against efficiency. The dc load line runs from \(V_{CC}\) at cutoff to \(V_{CC}/(R_C + R_E)\) at saturation. Base bias tracks β directly, collector feedback eases the problem, and voltage-divider bias nearly removes it. A bypass capacitor restores ac gain. Capacitor coupling blocks dc but sets a corner at \(1/(2\pi RC)\); direct coupling passes dc; transformer coupling isolates and matches.
Flashcards
Flashcard
The dc collector current and V_CE with no signal
Test your understanding
Question 1 of 5
An amplifier stage conducts for less than half of each cycle and drives a tuned LC tank that rings through the rest. Which class is it?
Question 2 of 5
A common-emitter stage has V_CC = 15 V, R_C = 3.3 kΩ and R_E = 1.5 kΩ. Where are the ends of its dc load line?
Question 3 of 5
A divider-biased stage on 10 V has R1 = 47 kΩ, R2 = 10 kΩ, R_E = 470 Ω and β = 100. What is its emitter current?
Question 4 of 5
A 1.0 kΩ source drives an amplifier whose input resistance is 4.0 kΩ through a coupling capacitor. Which capacitor puts the low-frequency corner at exactly 50 Hz?
Question 5 of 5
Transistors with β from 100 to 300 may be fitted. Which bias circuit keeps the collector current most nearly constant?
