4.1 Amplifier Operating Classes & Push-Pull Systems

4.1 Amplifier Operating Classes & Push-Pull Systems

In aircraft avionics and communication systems, power amplifiers must deliver high-fidelity signals to loads ranging from cockpit loudspeakers and interphone headsets to flight control servo actuators and VHF/UHF transmitting antennas. The operational performance of a transistor power amplifier—specifically its linearity, distortion figure, power conversion efficiency, and thermal dissipation—is fundamentally determined by its operating class.

Amplifier classification is established strictly by the location of the quiescent operating point (Q-point) along the transistor's dynamic transfer characteristic ($I_C$ versus $V_{BE}$ or $I_B$). This DC bias position establishes the conduction angle ($\theta_c$), defined as the portion of a complete $360^\circ$ AC input sinusoidal cycle during which collector current flows through the active device.

---

1. Operating Classification & The Dynamic Transfer Characteristic

The four primary amplifier classes recognized in EASA Part-66 Module 04 are Class A, Class B, Class AB, and Class C:

Conduction Angle (θc):
Class A:  ==================== 360° (Full cycle)
Class AB: ============ 185° - 200° (Slightly more than half cycle)
Class B:  ========== 180° (Exact half cycle)
Class C:  ===== 90° - 150° (Less than half cycle)

Mathematical Definition of Conversion Efficiency

Amplifier conversion efficiency ($\eta$) measures how effectively the stage converts DC input electrical power drawn from the aircraft power distribution rail ($P_{dc}$) into useful AC signal power delivered to the load ($P_{ac}$):

$\eta = \frac{P_{ac}}{P_{dc}} \times 100\% = \frac{P_{load}}{V_{CC} \cdot I_{dc}} \times 100\%$

Any electrical power not transferred to the AC load must be dissipated internally as heat within the transistor collector-base junctions ($P_D = P_{dc} - P_{ac}$). In high-power aircraft Line Replaceable Units (LRUs), managing this wasted thermal energy governs component size, heat sink architecture, and reliability.

---

2. Class A Amplifiers: Maximum Fidelity vs High Quiescent Loss

In a Class A amplifier, the DC bias network positions the Q-point near the center of the active region on the AC load line. Consequently, collector current flows continuously for the entire input cycle:

• Conduction Angle ($\theta_c$): $360^\circ$ ($2\pi\text{ radians}$).
• Linearity: Exceptional. Because the signal swings exclusively within the linear active region without entering cutoff or saturation, output waveform distortion is virtually zero for small signals.
• Theoretical Efficiency Limits:
  • Direct Resistive Collector Load ($R_C$): Maximum theoretical efficiency is $25%$.
  • Transformer-Coupled Load: Maximum theoretical efficiency reaches $50%$ because the transformer primary has negligible DC resistance, allowing the AC collector voltage to swing to twice the DC supply voltage ($2 V_{CC}$).

The Quiescent Thermal Penalty

The defining drawback of Class A operation is that full DC quiescent collector current ($I_{CQ}$) flows continuously, even when no AC input signal is present ($V_{in} = 0$). At zero signal input, the amplifier draws full DC power from the aircraft bus, converting $100%$ of it into waste heat inside the transistor casing:

$P_{D(quiescent)} = V_{CEQ} \cdot I_{CQ} = \frac{V_{CC}}{2} \cdot I_{CQ}$

Consequently, Class A operation is restricted in modern aviation to low-level signal processing—such as Cockpit Voice Recorder (CVR) preamplifiers, microphone inputs, and high-precision sensor signal conditioners where fidelity is paramount and power levels remain below a few hundred milliwatts.

[!NOTE] Transformer-Coupled Class A Efficiency: In a transformer-coupled Class A stage, the primary winding presents almost zero DC resistance ($R_{DC} \approx 0\ \Omega$), placing $V_{CEQ} \approx V_{CC}$. When an AC signal is applied, magnetic field collapse can drive the instantaneous collector voltage up to $2 V_{CC}$, doubling the peak-to-peak AC output voltage swing and raising maximum theoretical efficiency from $25%$ to $50%$.

---

3. Class B Amplifiers & The Push-Pull Topology

To overcome the poor efficiency and constant thermal dissipation of Class A, the Class B amplifier biases the transistor Q-point exactly at cutoff ($I_{CQ} = 0, V_{CEQ} = V_{CC}$):

• Conduction Angle ($\theta_c$): $180^\circ$ ($\pi\text{ radians}$, exactly half the input cycle).
• Theoretical Maximum Efficiency: $\eta_{max} = \frac{\pi}{4} \approx 78.5%$.

Because a single transistor biased at cutoff clips off the negative half of the sinusoidal waveform, linear audio amplification requires a complementary push-pull circuit. A push-pull stage pairs an NPN transistor with a matched PNP transistor:

1. Positive Input Half-Cycle: The positive-going base voltage forward-biases the NPN transistor ($Q_1$), driving current through the load into the speaker/servo terminal ("push"). The PNP transistor ($Q_2$) is reverse-biased and cut off.
2. Negative Input Half-Cycle: The negative-going input swings below ground, reverse-biasing $Q_1$ into cutoff while forward-biasing $Q_2$, which sinks current from the load back to the negative rail ("pull").

In the absence of an input signal ($V_{in} = 0$), both transistors are turned off, drawing zero quiescent current from the aircraft power supply ($I_{CQ} = 0, P_{dc} \approx 0$).

---

4. The Crossover Distortion Mechanism

Although theoretical Class B operation offers high efficiency, practical silicon bipolar transistors exhibit a non-linear threshold: the base-emitter barrier potential ($V_{BE} \approx 0.7\text{ V}$). Conduction cannot occur until the applied base-emitter voltage exceeds this threshold.

In a simple, unbiased Class B push-pull pair:

• When the input AC signal swings between $-0.7\text{ V}$ and $+0.7\text{ V}$, neither transistor receives sufficient forward bias to conduct.
• During this $1.4\text{ V}$ deadband, both $Q_1$ and $Q_2$ remain cut off.
• The output voltage remains at zero volts, producing a distinct flat notch or "step" at every zero-crossing transition of the waveform.

This phenomenon is known as crossover distortion. It introduces high-amplitude odd harmonics (3rd, 5th, 7th order) that severely degrade speech intelligibility in aircraft VHF communications and cockpit interphone systems, particularly at low signal amplitudes where the $1.4\text{ V}$ deadband represents a large fraction of the total signal.

---

5. Class AB Amplifiers: The Aviation Linear Standard

The universal engineering solution to crossover distortion is the Class AB amplifier. By introducing a small forward DC bias voltage ($V_{bias} \approx 1.4\text{ V}$ total, or $\approx 0.7\text{ V}$ per transistor) across the two base terminals, both transistors are held just at the threshold of conduction under quiescent conditions:

• Conduction Angle ($\theta_c$): Slightly greater than $180^\circ$ (typically $185^\circ\text{ to }200^\circ$).
• Practical Efficiency: $60%\text{ to }70%$ (approaching the $78.5%$ theoretical limit of pure Class B).
• Distortion: Crossover distortion is completely eliminated because each transistor begins conducting just before the opposing transistor turns off, ensuring smooth handover through the zero-crossing region.

Biasing Architectures

Two primary circuits provide temperature-compensated Class AB base bias in avionics power stages:

1. Series Diode Biasing: Two forward-biased silicon junction diodes ($D_1, D_2$) are connected in series between the bases of $Q_1$ and $Q_2$. Powered by pull-up and pull-down resistors, the diodes establish a constant $2 \times 0.7\text{ V} = 1.4\text{ V}$ potential difference.
2. $V_{BE}$ Multiplier (Rubber Diode): An auxiliary transistor and two-resistor voltage divider configure an adjustable bias voltage $V_{CE} = V_{BE} (1 + R_1/R_2)$, providing precision calibration.

[!WARNING] Thermal Runaway in Push-Pull Stages: As output power transistors heat up under load, their internal base-emitter voltage drops at approximately $-2\text{ mV}/^\circ\text{C}$, which increases quiescent base and collector currents. Higher current causes further heating, creating an unstable regenerative spiral known as thermal runaway. To prevent catastrophic destruction, the biasing diodes (or $V_{BE}$ multiplier transistor) must be thermally bonded to the same heat sink as the output power transistors. When the heat sink temperature rises, the diode voltage drop decreases concurrently, automatically reducing base bias and stabilizing quiescent current.

---

6. Class C Amplifiers & The Resonant Tank "Flywheel Effect"

In a Class C amplifier, the transistor is biased well beyond cutoff by applying a negative DC base voltage ($V_{BE} < 0$) or using signal self-biasing networks:

• Conduction Angle ($\theta_c$): Significantly less than $180^\circ$, typically $90^\circ\text{ to }150^\circ$.
• Theoretical Efficiency: $80%\text{ to }90%+$.
• Collector Current Waveform: Narrow, periodic pulses of high amplitude that flow only during the positive peak of the input cycle.

Because collector current flows in short pulses, a Class C amplifier produces extreme non-linear harmonic distortion. It is completely unusable for audio amplification. However, it is the premier choice for high-power Radio Frequency (RF) power transmitters—such as aircraft VHF COM transmitters (118–137 MHz), Distance Measuring Equipment (DME), and Mode-S ATC transponders.

The Parallel LC Tank Flywheel Effect

In an RF transmitter, the resistive collector load is replaced by a parallel LC resonant tank circuit tuned precisely to the input signal carrier frequency ($f_0 = \frac{1}{2\pi\sqrt{LC}}$):

1. When the transistor conducts its brief pulse, it injects a burst of energy into the tank, charging capacitor $C$ and establishing a magnetic field in inductor $L$.
2. When the transistor cuts off for the remaining $210^\circ\text{ to }270^\circ$ of the cycle, the tank circuit undergoes continuous electromagnetic oscillation: the magnetic field collapses, inducing current that charges the capacitor in reverse, which subsequently discharges back through the inductor.
3. This circulating reactive energy—termed the flywheel effect—reconstructs a pure, unclipped sinusoidal voltage waveform across the load antenna, filtering out all high-order harmonics.

---

Summary Comparison Matrix

---

Worked Engineering Calculation: Class B Push-Pull Output Stage

Problem

An avionics technician is analyzing a complementary push-pull power amplifier driving an $8.0\ \Omega$ cockpit warning loudspeaker from a dual $\pm 24.0\text{ V}$ DC split power supply rail. The amplifier is driven with a sinusoidal input producing a peak output voltage $V_p = 20.0\text{ V}$ across the load.

Calculate:

1. The AC signal power delivered to the speaker ($P_{ac}$).
2. The total DC power drawn from the power supplies ($P_{dc}$).
3. The amplifier conversion efficiency ($\eta$).
4. The total power dissipated as heat in both output transistors ($P_{D(total)}$).

Solution

1. AC Load Power ($P_{ac}$): $P_{ac} = \frac{V_{rms}^2}{R_L} = \frac{(V_p / \sqrt{2})^2}{R_L} = \frac{V_p^2}{2 R_L} = \frac{(20.0\text{ V})^2}{2 \times 8.0\ \Omega} = \frac{400}{16} = 25.0\text{ W}$

2. Total DC Power Drawn ($P_{dc}$): Peak load current is: $I_p = \frac{V_p}{R_L} = \frac{20.0\text{ V}}{8.0\ \Omega} = 2.50\text{ A}$ Each power supply rail delivers a half-wave rectified current pulse with average DC value: $I_{dc} = \frac{I_p}{\pi} = \frac{2.50\text{ A}}{\pi} \approx 0.7958\text{ A}$ Total DC power drawn from both $\pm 24.0\text{ V}$ rails ($V_{CC} = 24.0\text{ V}$ each) is: $P_{dc} = 2 \cdot V_{CC} \cdot I_{dc} = 2 \times 24.0\text{ V} \times 0.7958\text{ A} = 38.20\text{ W}$

3. Conversion Efficiency ($\eta$): $\eta = \frac{P_{ac}}{P_{dc}} \times 100\% = \frac{25.0\text{ W}}{38.20\text{ W}} \times 100\% \approx 65.45\%$ (Note: The maximum theoretical efficiency for this peak swing is $\frac{\pi}{4} \times \frac{V_p}{V_{CC}} = 0.7854 \times \frac{20}{24} = 65.45%$, confirming mathematical consistency).

4. Total Transistor Thermal Dissipation ($P_{D(total)}$): $P_{D(total)} = P_{dc} - P_{ac} = 38.20\text{ W} - 25.0\text{ W} = 13.20\text{ W}$ Each individual transistor dissipates: $P_{D(each)} = \frac{13.20\text{ W}}{2} = 6.60\text{ W}$ The aircraft LRU heat sink must safely remove $13.20\text{ W}$ of thermal dissipation.