Measurements, Testing, and Calibration Advanced Measurement Topics Informational

What is the hot S22 measurement of a power amplifier and why does it differ from small signal S22?

The hot S22 measurement of a power amplifier measures the output reflection coefficient (S22) while the amplifier is operating at its intended large-signal drive level (near compression or saturation), which differs significantly from the small-signal S22 measured with a low-power VNA stimulus. The difference arises because: a power amplifier operating near compression behaves as a nonlinear device where the output impedance changes with the signal level. At small signal: the transistor operates in its linear region, and the output impedance is determined by the small-signal model parameters (gds, Cds, matching network). At large signal: the transistor swings through its IV curves, clipping at the voltage rails and current limits, and the effective output impedance is a time-averaged value that depends on the signal amplitude, bias point, and load impedance. Hot S22 is measured using: the active measurement approach (a VNA is connected to the output of the PA while a separate signal source drives the input at the operational power level; the VNA injects a small probe signal at the output port and measures the reflected signal; the large-signal operating point is maintained during the measurement, giving the true large-signal output match). The hot S22 is critically important for: designing the output matching network of cascaded stages (the next stage's input sees the hot S22, not the small-signal S22), calculating the PA's stability under operational conditions, and designing the PA's output load for optimal power transfer and efficiency.

Category: Measurements, Testing, and Calibration

Updated: April 2026

Product Tie-In: VNAs, Probes, Chambers, Signal Generators

Power Amplifier Hot S22 Characterization

The distinction between small-signal S22 and hot S22 is crucial for power amplifier system design. Using small-signal S22 for system-level calculations can lead to: impedance mismatch and power loss between the PA and the next stage, unexpected oscillation due to incorrect stability analysis, and degraded power-added efficiency from sub-optimal load matching.

Measurement Techniques

VNA with external drive: Connect the VNA port 2 to the PA output. Drive the PA input with an external signal generator at the operational frequency and power level. The VNA measures S22 at the output port while the PA operates at full power. The VNA's source power at port 2 must be much lower than the PA's output power (typically 20-30 dB below) so the probe signal does not perturb the PA's operating point
Hot S-parameter extraction from load-pull data: During load-pull measurements, the PA's output impedance can be extracted from the variation of output power with load impedance. The hot S22 is the conjugate of the impedance that gives maximum power transfer
Nonlinear VNA (NVNA) extraction: A nonlinear VNA measures the complete large-signal waveforms at the DUT ports and extracts the hot S-parameters from the waveform data

Hot S22 Measurement Parameters

Small-signal S22: measured at low power (VNA stimulus ~ -20 dBm)
Hot S22: measured at operational drive level (PA output at P_sat)
Typical difference: |S22_hot - S22_small| = 0.1-0.5 (significant mismatch)
System mismatch loss: ML = (1-|Gamma_L|²) / |1-S22_hot × Gamma_L|²
PA output impedance: Z_out = Z_0 × (1+S22_hot)/(1-S22_hot)

Common Questions

Frequently Asked Questions

How much does hot S22 differ from small-signal S22?

Significantly. For a GaN PA at P_sat: small-signal S22 might be -15 dB (well matched), while hot S22 might be -3 to -6 dB (poorly matched). The output impedance typically shifts toward lower impedance at high power (the transistor's effective output capacitance increases as the voltage swing compresses). The phase of S22 also changes, sometimes by 30-90 degrees. This is why PA output matching networks designed for small-signal S22 often do not deliver optimal power in the system.

Do I need hot S22 for small-signal amplifiers?

No. For small-signal amplifiers (LNAs, IF amplifiers) that operate well below their compression point, the output impedance does not change significantly with signal level. Small-signal S22 from a standard VNA measurement is sufficient. Hot S22 is only needed when the amplifier operates in compression or near its maximum output power.

Can I simulate hot S22?

Yes, with a nonlinear device model (such as a foundry-provided GaN or GaAs FET model) in a circuit simulator (Keysight ADS, NI AWR). Perform a harmonic balance simulation at the operational drive level and extract the large-signal output reflection coefficient. Large-signal output match simulation is a standard step in PA design. The simulated hot S22 should be verified by measurement to validate the model accuracy.

Keep Reading