Why is Class D theoretically 100% efficient?

Power dissipated (heat) in a transistor equals Voltage multiplied by Current (P = V*I). In a Class D amplifier, the transistor is operated purely as a switch. When the switch is 'ON', massive current flows, but the voltage across the transistor drops to zero (0 * I = 0 Watts of heat). When the switch is 'OFF', the voltage is high, but the current is zero (V * 0 = 0 Watts of heat). Because voltage and current never overlap, the transistor theoretically generates no heat, delivering 100% of the DC power to the RF load.

Why don't we use Class D for 5G microwave frequencies?

Switching losses. While a transistor is 100% efficient when fully ON or fully OFF, it takes a finite amount of time to transition between those states. During that brief transition, both voltage and current exist simultaneously, generating a spike of heat. At low frequencies (like audio or AM radio), this transition is negligible. But at 3 GHz, the transistor must switch on and off 3 billion times a second. The transition time becomes a massive percentage of the total cycle, and the parasitic output capacitance (Cds) must be constantly charged and discharged, causing the amplifier to burn up.

How do you amplify a varying signal with a switch?

Because a switch only outputs a square wave (full power or zero power), it cannot directly amplify an amplitude-modulated signal. To solve this, designers use Pulse Width Modulation (PWM) or Outphasing (LINC). In PWM, the varying amplitude of the input signal is converted into the *width* of the digital pulses. The switch outputs wide pulses for high power and narrow pulses for low power. A low-pass filter at the output integrates these pulses back into a smooth analog waveform.

Active Components

Class D Amplifier

To maximize power efficiency in an HF transmitter, an engineer Abandons linear amplification entirely. Instead of using the transistor as a variable resistor (like in Class AB), they use two transistors configured in a push-pull arrangement and drive them violently with a massive input signal. The transistors no longer amplify a sine wave; they act as brutal digital switches, alternating between fully ON (saturated) and fully OFF (pinched-off). Because a perfect switch drops no voltage when closed and draws no current when open, the power dissipation (V × I) is theoretically zero, yielding 100% efficiency. The output is a harsh square wave, which is passed through a resonant LC filter to strip away the massive harmonic content and reconstruct the fundamental sine wave. While dominant in audio and low-frequency RF, Class D struggles at microwave frequencies due to the finite time it takes to charge and discharge parasitic transistor capacitances during high-speed switching.

Category: Active Components

Topology: Push-Pull Switching (Voltage or Current Mode)

Theoretical Efficiency: 100%

Linear vs. Switch-Mode Architectures

Feature	Class AB (Linear)	Class D (Switch-Mode)
Transistor Operation	Variable Resistor (Current Source)	Digital Switch (On/Off)
Voltage/Current Overlap	High (Generates massive heat)	Zero (Theoretically no heat)
Output Waveform	Sine Wave (Clean)	Square Wave (Harmonic rich)
High-Frequency Limit	Excellent (Works at mmWave)	Poor (Limited by switching loss / Cds)

Voltage-Mode Class D Efficiency Loss:
P_loss = f · C_ds · V_DD²
Where f is the switching frequency and C_ds is the parasitic drain-to-source capacitance.

The Microwave Limitation:
Every time the transistor switches OFF, it must dump the energy stored in its parasitic capacitance (C_ds) as heat. At 1 MHz, this happens 1 million times a second, which is negligible. At 2 GHz, this happens 2 billion times a second. The resulting P_loss becomes so massive that the transistor instantly melts. This is why Class D is generally restricted to HF/VHF frequencies, while Class E is used for microwaves.

Common Questions

Frequently Asked Questions

What is the difference between Voltage-Mode and Current-Mode?

In Voltage-Mode Class D (VMCD), the push-pull switches generate a square voltage wave, and the series-resonant output filter forces the current to be a sine wave. In Current-Mode Class D (CMCD), an RF choke ensures the DC supply acts as a constant current source. The switches generate a square current wave, and a parallel-resonant tank forces the voltage to be a sine wave. CMCD is generally preferred at higher RF frequencies because it absorbs the transistor's parasitic capacitance into the parallel tank.

Why do both transistors sometimes turn on at the same time?

This is called "shoot-through" current. In a push-pull setup, if the top transistor hasn't fully turned off before the bottom transistor turns on, a direct short circuit is created between the power supply and ground. This results in massive destructive current. Designers must carefully tune the input drive signals to ensure a slight "dead-time" between switching events to prevent the amplifier from destroying itself.

Is Class D the same as Class F?

No. Class D uses the physical transistors as digital switches to violently chop the signal into a square wave. Class F uses a linear transistor (operating as a current source) but surrounds it with highly complex harmonic resonant stubs that organically "shape" the sine wave into a square wave at the drain. Class D is topological switching; Class F is harmonic impedance tuning.

Related Terms

← Class C Amplifier Class F Amplifier →

Switch-Mode PA Design

Switching Loss Calculator

Input your target operating frequency, VDD, and the parasitic Cds of your GaN transistor. Calculate the exact amount of power lost to parasitic discharge and determine if your Class D architecture will survive the thermal load.

Calculate Power Loss