Unlike small-signal amplifiers that primarily focus on voltage or current gain, power amplifiers are designed to deliver significant power to a load (e.g., a loudspeaker). Their primary concerns are power efficiency, output power capability, and thermal management. Power amplifier classes are defined by the conduction angle of the active device (transistor) during one cycle of the input signal.

● Operating Principle: In a Class A amplifier, the transistor is biased such that it conducts current for the entire 360 degrees of the input AC cycle. The Q-point is typically set near the center of the DC load line. This ensures that the transistor is always in the active region, never cutting off or saturating for the full signal swing.
● Efficiency: Class A amplifiers are known for their linear operation and low distortion. However, they are highly inefficient. Even with no input signal, the transistor continuously draws quiescent current, dissipating power.
○ Maximum Theoretical Efficiency: For a capacitively coupled (or transformer-coupled) Class A amplifier, the maximum theoretical efficiency is 25% (for resistive load). For an ideal transformer-coupled Class A, it can reach 50%.
○ Reasons for Low Efficiency: Power is continuously dissipated in the collector resistor (R_C) and the transistor itself, even when no signal is applied. When an AC signal is present, power is transferred to the load, but significant power is still wasted as heat.
● Distortion: Generally low if operated within the linear region. However, as the input signal amplitude increases, the amplifier can enter saturation or cutoff, leading to significant clipping distortion.

● Operating Principle: In a Class B amplifier, each transistor is biased at cutoff. This means that a transistor only conducts for approximately 180 degrees (half) of the input AC cycle. A push-pull configuration uses two transistors (one NPN, one PNP, or two NPNs with a phase splitter) where one transistor handles the positive half of the output waveform, and the other handles the negative half.
● Efficiency: Class B amplifiers are much more efficient than Class A, with a maximum theoretical efficiency of 78.5%. This is because current is drawn from the power supply only when there is an input signal, reducing quiescent power dissipation.
● Distortion (Crossover Distortion): The major drawback of Class B is crossover distortion. Because each transistor is biased at cutoff, there is a small voltage region around 0V where neither transistor is fully turned on. This creates a "dead zone" in the output waveform, resulting in a distorted (not perfectly smooth) output near the zero-crossing points.

● Operating Principle: Class AB amplifiers are a compromise between Class A and Class B. Each transistor is biased slightly above cutoff, allowing a small quiescent current to flow even with no input signal. This ensures that both transistors are conducting for slightly more than 180 degrees (e.g., 190-200 degrees), overlapping their conduction regions slightly.
● Efficiency: Efficiency is lower than Class B but significantly higher than Class A (typically 50-70%).
● Distortion: The small quiescent current effectively eliminates crossover distortion, resulting in much cleaner output waveforms compared to Class B. This makes Class AB the most common class for audio power amplifiers.
● Biasing for Class AB: Small biasing voltages (e.g., using two forward-biased diodes in series with the base circuit of the push-pull transistors) are used to provide the necessary small quiescent current.

Feedback involves feeding a portion of the output signal back to the input. If the fed-back signal is out of phase with the input signal, it's called negative feedback (or degenerative feedback). Negative feedback is widely used to improve amplifier characteristics.
● Principle: A fraction of the output voltage or current is sampled and fed back to the input, where it is summed (subtracted for negative feedback) with the original input signal. This effectively reduces the overall gain but offers significant performance improvements.
● Types of Negative Feedback: There are four basic types based on how the output is sampled (voltage or current) and how it's mixed at the input (series or shunt):
○ Voltage-Series Feedback: Output voltage is sampled, fed back in series with the input.
○ Voltage-Shunt Feedback: Output voltage is sampled, fed back in shunt (parallel) with the input.
○ Current-Series Feedback: Output current is sampled, fed back in series with the input.
○ Current-Shunt Feedback: Output current is sampled, fed back in shunt with the input.

Let A be the open-loop gain (gain without feedback) and beta be the feedback factor (fraction of output fed back).
○ Closed-Loop Gain (A_f): The gain with feedback.
A_f=frac{A}{1+A eta}
For large A eta (i.e., A eta gg 1), A_f approximately equals frac{1}{eta}. This means the gain becomes almost entirely dependent on the feedback network, making it very stable and predictable.
○ Input Resistance with Feedback (R_in(f)):
■ For Series Input Feedback (like Voltage-Series or Current-Series):
Input resistance increases.
R_in(f)=R_in(1+A eta)
■ For Shunt Input Feedback (like Voltage-Shunt or Current-Shunt):
Input resistance decreases.
R_in(f)=frac{R_in}{1+A eta}
○ Output Resistance with Feedback (R_out(f)):
■ For Voltage Output Feedback (like Voltage-Series or Voltage-Shunt): Output resistance decreases.
R_out(f)=frac{R_out}{1+A eta}
■ For Current Output Feedback (like Current-Series or Current-Shunt): Output resistance increases.
R_out(f)=R_out(1+A eta)
○ Bandwidth with Feedback (BW_f): Negative feedback generally increases the bandwidth.
BW_f=BW(1+A eta)
○ Distortion and Noise Reduction: Negative feedback significantly reduces non-linear distortion (e.g., harmonic distortion) and noise generated within the amplifier.
Distortion with feedback = Distortion without feedback / (1+A eta).
Noise with feedback = Noise without feedback / (1+A eta).