Transistor Amplifiers: Small-Signal, Multistage & Power Stages
Exhaustive analysis of small-signal amplifiers using hybrid h-parameters; high- and low-frequency cutoffs and Bode plots; multistage RC-coupled, transformer, and direct-coupled cascaded systems; Class A, B, AB, and C power amplifier topologies, push-pull configurations, theoretical efficiency limits, and crossover distortion remedies.
ยง4.1 Small-Signal Analysis & Hybrid h-Parameter Model of CE BJT
1. Two-Port Network Formalism and Hybrid h-Parameters
For small alternating signals superimposed on DC quiescent operating points, a bipolar junction transistor in Common Emitter (CE) configuration is modeled linearly as a two-port network relating input voltage $v_b$, input current $i_b$, output current $i_c$, and output voltage $v_c$:
The four parameters are precisely defined under AC short-circuit and open-circuit conditions:
- $h_{ie} = \left.\frac{v_b}{i_b}\right|_{v_c = 0}$ : Short-circuit input impedance $(\Omega)$
- $h_{re} = \left.\frac{v_b}{v_c}\right|_{i_b = 0}$ : Open-circuit reverse voltage ratio (dimensionless, typically $\sim 10^{-4}$)
- $h_{fe} = \left.\frac{i_c}{i_b}\right|_{v_c = 0}$ : Short-circuit forward current transfer ratio / AC current gain $(\beta_{ac})$
- $h_{oe} = \left.\frac{i_c}{v_c}\right|_{i_b = 0}$ : Open-circuit output admittance $(\text{S} = \Omega^{-1}$, typically $\sim 20\,\mu\text{S}$)
2. Simplified CE Hybrid Model & Gain Derivations
In standard practical audio and RF amplifier design, $h_{re} \approx 0$ (reverse feedback negligible) and $h_{oe} R_L \ll 0.1$ (output conductances much smaller than load conductance). The circuit simplifies to an input resistor $h_{ie}$ and a dependent current source $h_{fe} i_b$.
With an effective AC collector load $R_L' = R_C \parallel R_L$:
Hence, the Voltage Gain $A_v$ is:
The negative sign reflects the fundamental $180^\circ$ phase inversion between input base signal and output collector signal in a Common Emitter amplifier.
The Current Gain $A_i$, Input Impedance $Z_{in}$, and Output Impedance $Z_{out}$ are:
3. Frequency Response and Cutoff Frequencies
The frequency response curve of a CE amplifier exhibits three distinct regimes:
- Low-Frequency Range ($f < f_L$): The reactances of coupling capacitors $C_1, C_2$ and emitter bypass capacitor $C_E$ ($X_C = 1/(2\pi f C)$) grow large. This introduces voltage division and degenerative negative feedback across $R_E$, reducing voltage gain at a slope of $+20\,\text{dB/decade}$ per dominant pole.
- Midband Range ($f_L \le f \le f_H$): Coupling/bypass capacitors act as virtual short circuits ($X_C \approx 0$), while internal parasitic transistor capacitances act as virtual open circuits. The gain remains flat at maximum midband gain $A_{vm}$.
- High-Frequency Range ($f > f_H$): The internal depletion and diffusion capacitances (base-emitter capacitance $C_\pi$ and base-collector Miller capacitance $C_\mu$) introduce shunting paths to ground, rolling off gain at $-20\,\text{dB/decade}$.
The Bandwidth ($BW$) is defined between the half-power ($-3\,\text{dB}$) points:
ยง4.2 Multistage Amplifiers: RC-Coupled, Transformer & Direct Coupling
1. Cascading Amplifiers & Decibel Gain Formalism
A single transistor stage often cannot simultaneously provide adequate voltage gain, current drive, and impedance matching. Multiple amplifier stages are therefore cascaded in series, where the output of stage $n$ serves as the input to stage $n+1$.
The overall voltage gain is the multiplicative product of individual loaded gains:
Expressed logarithmically in Decibels (dB):
Decibels convert complicated cascading products into simple additions.
2. Comparative Analysis of Coupling Schemes
| Coupling Type | Frequency Response | Impedance Matching | Cost & Size | Primary Applications |
|---|---|---|---|---|
| RC Coupling | Excellent flat midband (audio 20 Hz โ 20 kHz); rolls off at DC and very high RF | Poor (collector $R_C$ shunts next stage $Z_{in}$) | Extremely low cost, compact, highly reliable | Audio preamplifiers, general-purpose voltage gain |
| Transformer Coupling | Poor audio flatness; resonant peaking; zero response at DC | Superior ($Z_p/Z_s = (N_p/N_s)^2$) for maximum power transfer | Bulky, heavy, expensive, magnetic hum pickup | RF tuned amplifiers, driver stages to low-impedance speakers |
| Direct Coupling | Extends down to DC ($0\,\text{Hz}$); no lower cutoff frequency | Moderate | Minimal parts, ideal for monolithic IC fabrication | Operational amplifiers, biosensors, DC instrumentation |
3. Loading Effect in RC-Coupled Cascades
When stage 1 is coupled to stage 2 through capacitor $C_c$, the effective AC load seen by collector 1 is not simply $R_{C1}$, but the parallel combination of $R_{C1}$ and the input impedance of stage 2:
Because $R_{L1}' < R_{C1}$, the loaded gain $A_{v1}$ is significantly smaller than the open-circuit gain of an isolated stage. Neglecting this loading effect produces massive design errors.
ยง4.3 Power Amplifiers: Classification (Class A, B, AB, C) & Efficiency Limits
1. Large-Signal Operation & Conduction Angle Classifications
Unlike small-signal voltage amplifiers that handle millivolt signals, Power Amplifiers (large-signal amplifiers) deliver significant power (watts to kilowatts) into low-impedance loads (speakers, antennas) with high efficiency while keeping device dissipation within thermal tolerances ($T_j < T_{j,max}$).
Amplifier classes are categorized strictly by the conduction angle $\theta_c$ of the collector current during one complete sinusoidal input cycle ($360^\circ$):
- Class A ($\theta_c = 360^\circ$ or $2\pi$ radians): The transistor conducts continuously for the entire input cycle. The Q-point is biased in the center of the active linear region. Distortions are minimal, but static DC power is continually wasted even with zero signal.
- Class B ($\theta_c = 180^\circ$ or $\pi$ radians): The transistor is biased exactly at cutoff ($I_{CQ} = 0$). Conduction occurs for exactly one half-cycle ($180^\circ$). Zero static DC power at idle.
- Class AB ($180^\circ < \theta_c < 360^\circ$): Biased slightly above cutoff with a small quiescent current ($I_{CQ} > 0$). Conduction occurs for slightly more than half a cycle, eliminating Class B crossover distortion.
- Class C ($\theta_c < 180^\circ$): Biased well beyond cutoff. Conduction occurs in brief pulses ($80^\circ - 120^\circ$). High harmonic distortion; used exclusively with tuned resonant LC tanks in high-efficiency RF transmitters.
2. Mathematical Derivation of Maximum Collector Efficiency
Collector Efficiency $\eta$ is defined as the ratio of average AC output power delivered to the load to the average DC power drawn from the power supply:
(a) Series-Fed Class A Amplifier
For a series-fed Class A amplifier with supply $V_{CC}$ and collector resistor $R_C$:
Under maximum unclipped sinusoidal swing, $V_m = V_{CC}/2$ and $I_m = I_{CQ} = V_{CC}/(2 R_C)$:
For a transformer-coupled Class A amplifier, the DC drop across the primary is near zero, allowing peak-to-peak voltage swing up to $2 V_{CC}$, doubling the theoretical limit to $\eta_{max} = 50\%$.
(b) Push-Pull Class B Amplifier
In a complementary-symmetry push-pull Class B configuration operating from a supply $V_{CC}$ (or dual $\pm V_{CC}$):
For peak load voltage $V_m$ across load $R_L$, peak collector current is $I_m = V_m / R_L$. The average current drawn from the DC supply during half-wave pulses by each rail is $I_{dc} = \frac{2}{\pi} I_m$. Thus:
The AC power delivered to the load is:
Collector efficiency as a function of output swing is:
Under maximum theoretical output voltage swing where $V_m = V_{CC}$:
ยง4.4 Push-Pull Configurations, Crossover Distortion & Class AB Biasing
1. Push-Pull Operation & Harmonic Distortion Cancellation
A push-pull amplifier utilizes two matched transistors ($Q_1$ and $Q_2$) operating in anti-phase: $Q_1$ conducts during the positive half-cycle of the input signal, pushing current into the load, while $Q_2$ conducts during the negative half-cycle, pulling current from the load.
Mathematically, expanding the nonlinear collector current transfer characteristics in a Taylor series:
In a balanced push-pull output transformer, net load current is proportional to the difference $i_{c1} - i_{c2}$:
Crucial Result: All even-order harmonic distortion terms ($a_2 v_{in}^2, a_4 v_{in}^4$) completely cancel out! Furthermore, DC core saturation in output transformers is eliminated because quiescent currents produce opposing magnetic fluxes.
2. Crossover Distortion in Pure Class B
In a pure Class B push-pull stage, transistors are biased at $V_{BE} = 0$. However, real silicon bipolar transistors require a threshold forward voltage $V_{BE} \approx 0.6\text{ to }0.7\,\text{V}$ before base-emitter conduction begins.
Consequently, whenever the input signal passes through the zero-crossing within the deadband $-0.7\,\text{V} < v_{in} < +0.7\,\text{V}$, neither transistor conducts ($i_{c1} = i_{c2} = 0$). The output waveform flattens to zero near every crossing, generating severe high-order harmonic distortion known as Crossover Distortion.
3. Class AB Biasing and Diode Thermal Tracking
To eliminate crossover distortion, the transistors are biased into Class AB by applying a slight forward bias ($V_{bias} \approx 2 V_D \approx 1.4\,\text{V}$) across the base terminals using two series silicon diodes ($D_1, D_2$) or an active $V_{BE}$-multiplier circuit.
This maintains both transistors barely conducting at an idle quiescent current $I_{CQ} \approx 10\text{ to }50\,\text{mA}$. When the input swings through zero, the conduction smoothly transfers from $Q_1$ to $Q_2$ with zero deadband. Mounting the biasing diodes on the same physical heatsink as the power output transistors ensures thermal tracking: as junction temperature rises, diode voltage drops at $-2\,\text{mV}/^\circ\text{C}$, matching the transistor $V_{BE}$ drop and preventing thermal runaway.
A Common Emitter BJT amplifier operates with a load resistance $R_L = 10\text{ k}\Omega$ and collector bias resistor $R_C = 4.7\text{ k}\Omega$. Transistor $h$-parameters are $h_{ie} = 2.0\text{ k}\Omega$, $h_{fe} = 100$, $h_{re} = 2.5 \times 10^{-4}$, and $h_{oe} = 25\ \mu\text{S}$. (a) Calculate the exact current gain $A_i = i_L / i_b$. (b) Calculate the input impedance $R_{\text{in}}$ seen at the base terminal. (c) Determine the overall voltage gain $A_v = v_o / v_{\text{in}}$ and output impedance $R_{\text{out}}$.
Step 1: Calculate Effective AC Collector Load $R_L'$
The AC collector resistance is the parallel combination of $R_C$ and external load $R_L$:
Step 2: Check Simplified Model Validity
Since $h_{oe} R_L' < 0.1$, the simplified model yields high accuracy within $5\%$.
Step 3: Calculate Voltage Gain $A_v$
Using the simplified formula:
Using the exact formula including $h_{re}$ and $h_{oe}$:
Remarkably, $\Delta h \approx 0$ here, so the exact gain equals precisely the simplified gain: $A_v = -319.7$.
Step 4: Calculate Current Gain $A_i$ and Input Impedance $Z_{in}$
A two-stage RC-coupled BJT amplifier consists of identical Common Emitter stages. For each transistor, $h_{ie} = 1.5\text{ k}\Omega$, $h_{fe} = 80$, and $h_{oe} \approx 0$. The collector resistors are $R_{C1} = R_{C2} = 3.3\text{ k}\Omega$, biasing resistors are $R_1 = 47\text{ k}\Omega$, $R_2 = 10\text{ k}\Omega$, and the load is $R_L = 4.7\text{ k}\Omega$. (a) Determine the effective AC load of the first stage $R_{L1}'$. (b) Calculate the individual voltage gains $A_{v1}$ and $A_{v2}$. (c) Compute the total overall voltage gain $A_v = A_{v1} \times A_{v2}$ in decibels.
Step 1: Calculate Input Impedance of Stage 2 ($Z_{in2}$)
The input impedance of Stage 2 includes the bias resistors in parallel with the transistor base input:
Step 2: Calculate Loaded Voltage Gain of Stage 2 ($A_{v2}$)
The effective AC load on Stage 2 is $R_{L2}' = R_{C2} \parallel R_L$:
Step 3: Calculate Loaded Voltage Gain of Stage 1 ($A_{v1}$)
Stage 1 is loaded by its own collector resistor $R_{C1}$ in parallel with the entire input impedance $Z_{in2}$ of Stage 2:
Step 4: Calculate Overall Voltage Gain $A_v$ and Decibel Gain
Note that the double phase inversion produces an overall positive gain (in-phase output).
A complementary-symmetry Class B push-pull amplifier operates from dual power supplies of $\pm V_{CC} = \pm 18\text{ V}$ and drives an $8.0\ \Omega$ loudspeaker load. (a) Calculate the maximum unclipped output signal swing and maximum AC output power $P_{L,\text{max}}$ (assuming ideal zero saturation voltage). (b) Determine the DC power supplied $P_{\text{dc}}$ under maximum output conditions. (c) Calculate the conversion efficiency $\eta$ and the maximum thermal power dissipation per transistor $P_{D,\text{max}}$.
Step 1: Calculate Maximum AC Power Delivered to Load $P_{ac,max}$
With peak output voltage $V_m = V_{CC} = 18\,\text{V}$ and load $R_L = 8\,\Omega$:
Step 2: Calculate DC Power Drawn from Supply $P_{dc}$
Peak output current is $I_m = V_m / R_L = 18 / 8 = 2.25\,\text{A}$. Average supply current is $I_{dc} = \frac{2}{\pi} I_m$:
Step 3: Collector Efficiency $\eta$ at Maximum Swing
Step 4: Power Dissipated by Transistors at Maximum Swing
The total heat power dissipated across both transistors is:
For each individual transistor:
Step 5: Worst-Case Transistor Thermal Dissipation Peak
In Class B amplifiers, maximum transistor dissipation does NOT occur at maximum signal swing! Total dissipation is $P_D(V_m) = \frac{2 V_{CC} V_m}{\pi R_L} - \frac{V_m^2}{2 R_L}$. Differentiating with respect to $V_m$ and setting to zero:
At this specific output voltage swing, the worst-case transistor dissipation is:
Heatsink sizing must be designed to withstand this $4.10\,\text{W}$ peak dissipation rather than the $2.765\,\text{W}$ full-power condition.
Solved University Examination Problems
Step-by-step mathematical solutions to classic university honors examination questions.