Class AB Bias Point
Understanding the Class AB Bias Point
In high-power RF transmitter design, engineers are constantly fighting a war between linearity (clean signals) and efficiency (battery life and heat generation). The Class AB Bias Point is the ultimate compromise, making it the dominant topology for 95% of modern linear communications systems, including 4G/5G base stations, satellite downlinks, and Wi-Fi routers.
To operate in Class AB, the transistor's DC gate/base voltage is biased just slightly above the threshold of conduction (pinch-off). Unlike Class A (which rests at 50% max current) or Class B (which rests exactly at 0% current), a Class AB amplifier rests at a very low "trickle" current—typically 5% to 15% of the transistor's maximum rated current (Imax). Because it starts slightly "on," it smoothly conducts the entire positive half of the RF sine wave and a small portion of the negative half before pinching off. This results in a conduction angle between 180 and 360 degrees.
The Benefit of the Trickle Current
If an amplifier is biased exactly at zero (Class B), turning the transistor on takes a finite amount of time and voltage, creating a harsh distortion at the zero-crossing point (crossover distortion). By keeping a small trickle current flowing (Class AB bias), the transistor stays "awake." This completely eliminates crossover distortion, drastically improving the linearity and Intermodulation Distortion (IMD) performance while still achieving real-world power efficiencies of 50% to 65%.
Conduction Angle (θ): π < θ < 2π (180° to 360°)
Efficiency Range:
Lower bound: 50% (Approaching Class A)
Upper bound: 78.5% (Approaching Class B)
Practical RF Efficiency: 55% - 65%
Comparison
| Metric | Class A | Class AB (The Sweet Spot) | Class B |
|---|---|---|---|
| Quiescent Current | Massive (50%) | Low (5-15%) | Zero |
| Linearity (IMD3) | Superior (-45 dBc) | Very Good (-30 dBc) | Poor (Crossover distortion) |
| Power Efficiency | Terrible (20-30%) | Good (55-65%) | High (70%) |
| Idle Heat Generation | Extreme | Minimal | None |
Frequently Asked Questions
How do I set a Class AB bias point in the lab?
You start with the RF input turned completely off. You apply the required drain/collector voltage (e.g., 28V). Then, you slowly increase the gate/base voltage from zero until the DC power supply reads the specific 'trickle' quiescent current (Idq) recommended by the transistor datasheet (e.g., 200 mA).
Why does the DC current increase when RF is applied to a Class AB amp?
Because a Class AB amplifier pinches off during the negative half of the RF cycle, it acts like a rectifier. The harder you drive the RF input, the higher the average DC current drawn from the power supply. (In contrast, a Class A amplifier draws the exact same DC current whether RF is present or not).
Can Class AB handle complex modulations like 256-QAM?
Yes, but with caveats. 256-QAM has a very high Peak-to-Average Power Ratio (PAPR). To prevent the high peaks from compressing, the Class AB amplifier must be 'backed off' (run at a lower average power). Modern systems use Digital Predistortion (DPD) to fix the remaining non-linearities, allowing Class AB to perfectly transmit 5G waveforms.