Why does a Class A amplifier draw the same DC current with or without an RF signal?

The bias point is at 50 percent of IDSS, the center of the transfer curve. When RF is applied, the drain current swings symmetrically above and below this point. The time-average current remains exactly IDQ because the positive and negative swings are equal. This means the DC supply delivers constant power PDC = VDD times IDQ regardless of RF output level. At zero RF output, all of PDC is dissipated as heat. At maximum RF output, heat dissipation drops to half of PDC (since half is converted to RF), reaching the 50% theoretical efficiency limit.

How much worse is Class A efficiency with backed-off signals?

Class A efficiency scales linearly with output power relative to P1dB. At 1 dB compression, efficiency is approximately 40 percent. Back off 6 dB (typical for single-carrier QAM) and efficiency drops to 10 percent. Back off 10 dB (typical for multi-carrier OFDM) and efficiency is 5 percent. For a 20 W PA at 5 percent efficiency, the DC input is 400 W and the heat sink must dissipate 380 W, which is why Class A is impractical for high-power base station applications.

Active Components

Class A Amplifier

Q: Where is Class A still the right choice?

Class A remains preferred in several niche applications: (1) driver stages in multi-stage PAs where linearity matters more than efficiency and the power level is low enough that the efficiency penalty is acceptable, (2) laboratory test equipment amplifiers where spurious-free dynamic range must be maximized, (3) LNA stages where the transistor is already biased near 50 percent IDSS for minimum noise figure, and (4) small-signal gain blocks where total DC power is under 100 mW and efficiency is irrelevant.

Remove the RF signal from a Class A power amplifier and its DC current does not change. The transistor sits at 50% of I_DSS, drawing full bias current whether it is amplifying a signal or not. Every watt of DC power that is not converted to RF becomes heat. At the theoretical maximum of 50% drain efficiency, half the DC power is wasted. At the 10 dB backoff that modern wideband signals demand, only 5% reaches the antenna; the rest heats the room. Despite this, Class A remains the bias of choice whenever linearity outweighs efficiency: in driver stages, test equipment, and the LNA stages where noise figure, not power, is the design priority.

Category: Active Components

Conduction Angle: 360° (full cycle)

Max Drain Efficiency: 50%

The Price of Perfect Linearity

DC power (constant regardless of RF):
P_DC = V_DD × I_DQ

Maximum RF output power:
P_RF,max = (V_DD − V_knee)² / (2 × R_opt)
where R_opt = (V_DD − V_knee) / I_DQ

Maximum drain efficiency:
η_max = P_RF,max / P_DC = (V_DD − V_knee) / (2×V_DD)
For V_knee → 0: η_max = 50%

Efficiency at backoff:
At 6 dB backoff: η ≈ 12.5% | At 10 dB backoff: η ≈ 5%

Where Class A Excels Despite the Waste

Application	Why Class A	Typical Power	Eff. Penalty Acceptable?
Driver stages	Linearity for pre-driver; power is low	10 to 100 mW	Yes (total waste <200 mW)
Test equipment amps	Maximum SFDR, no crossover distortion	100 mW to 1 W	Yes (plugged into wall)
LNA bias	50% I_DSS is near NF_min bias	<50 mW	Yes (noise matters, not efficiency)
Small-signal gain blocks	Simple bias, flat gain, no spurious	10 to 50 mW	Yes (negligible total power)

The Thermal Reality

For a hypothetical 20 W Class A PA at V_DD = 28 V:

I_DQ = P_DC/V_DD = 40 W / 28 V = 1.43 A (constant)
At full power (20 W RF out): dissipation = 40 − 20 = 20 W heat
At 10 dB backoff (2 W RF out): dissipation = 40 − 2 = 38 W heat
With no signal: dissipation = 40 W heat

The thermal design must handle the worst case (no signal), which is the maximum heat dissipation. This is the opposite of most amplifier classes, where maximum heat occurs near compression.

Common Questions

Frequently Asked Questions

Why does DC current stay constant with or without RF?

The bias is at 50% I_DSS. RF swings the current symmetrically above and below this point; the time-average is unchanged. At zero RF, all DC power becomes heat. At max RF, heat drops to half (50% efficiency limit).

Where is Class A still the right choice?

Driver stages (low power, linearity critical), test equipment (SFDR matters), LNAs (already biased near 50% I_DSS for NF_min), and small-signal gain blocks where total DC is under 100 mW.

How bad is efficiency at backoff?

Efficiency scales linearly with output power. At 6 dB backoff: 12.5%. At 10 dB backoff (OFDM): 5%. A 20 W PA at 5% efficiency dissipates 380 W of heat. That is why Class A is impractical for base station finals.

Related Terms

← Circuit Envelope Class A Bias Point →

Amplifier Tutorials

Load-Line Analysis Interactive Tool

Plot IV curves, set the bias point, and visualize drain current waveforms for Class A, AB, B, and C in real time. Calculates efficiency and output power automatically.

Try the Simulator