RF power amplifier linearity describes how accurately an amplifier reproduces the amplitude and phase characteristics of an input signal. When an amplifier is driven toward its nonlinear region, its output may no longer be a scaled copy of the input. This can cause gain variation, phase distortion and unwanted energy in adjacent frequency channels.
Engineers commonly use AM-AM, AM-PM, adjacent channel leakage ratio (ACLR) and error vector magnitude (EVM) to understand different aspects of RF signal quality. These measurements are related, but they are not interchangeable. Each describes a different behavior and must be interpreted using its own test conditions.
1. What Does RF Amplifier Linearity Mean?
An ideal linear amplifier produces an output that is proportional to its input. If the input amplitude changes by a given ratio, the output amplitude changes by the same ratio, while the phase relationship remains predictable. In a practical RF power amplifier, this approximation is valid only over a defined operating region.
As output power increases, device characteristics and circuit limits cause the amplifier response to depart from ideal linear behavior. Gain may compress or expand, and the output phase may shift as the input amplitude changes. With modulated signals, these effects can alter the transmitted waveform and create spectral components outside the intended channel.
Linearity is therefore not represented by one universal number. It is evaluated through several measurements, including gain compression, AM-AM and AM-PM characteristics, intermodulation products, ACLR or ACPR, and EVM. Which metrics matter most depends on the signal waveform, bandwidth, applicable standard and system-level requirements.
2. What Is AM-AM Distortion?
AM-AM means amplitude modulation to amplitude modulation. It describes how the output amplitude of an amplifier changes in response to changes in input amplitude. For an ideal linear amplifier, the relationship is proportional and the gain remains constant over the operating range.
In a real power amplifier, the output amplitude may increase more slowly as the input drive rises. This is gain compression. Some circuits may also show gain expansion over part of their operating range. An AM-AM curve plots output amplitude or output power against input amplitude or input power, making these changes visible.
AM-AM distortion changes the envelope of an amplitude-varying signal. It is related to the gain-compression behavior discussed in RF power amplifier datasheets and to the output operating point represented by P1dB. However, a single P1dB value does not fully describe the complete AM-AM curve.
3. What Is AM-PM Distortion?
AM-PM means amplitude modulation to phase modulation. It describes a change in output phase as the input signal amplitude changes. An ideal amplifier would preserve a constant phase relationship, apart from its fixed phase shift and propagation delay. A practical amplifier can introduce an amplitude-dependent phase shift.
AM-PM distortion matters because many modulated waveforms carry information in both amplitude and phase. Even if the output amplitude appears acceptable, an amplitude-dependent phase change can distort the constellation or alter the waveform’s spectral characteristics. AM-PM behavior is often plotted as phase shift versus input power or output power.
AM-AM and AM-PM are separate descriptions of nonlinear behavior. An amplifier may exhibit significant amplitude distortion, phase distortion or both. A datasheet or characterization report that provides both curves gives a more complete view than a single gain or output-power figure.
4. How Nonlinearity Causes Spectral Regrowth
A nonlinear amplifier can generate frequency components that were not present in the original input signal. With a modulated waveform, this may appear as spectral regrowth: energy spreads beyond the intended channel bandwidth and into neighboring frequency regions.
A common way to examine nonlinear behavior is a two-tone test. Two closely spaced input tones pass through the amplifier, and the output spectrum is examined for intermodulation products. Third-order products are of particular interest because they can fall close to the wanted tones and may be difficult to remove with ordinary filtering.
Two-tone intermodulation testing is useful for characterizing an amplifier, but it does not replace a modulated-signal test. Real waveforms have their own bandwidth, envelope statistics and modulation characteristics. For a system-level assessment, the test signal should represent the intended waveform as closely as practical.
5. What Is ACLR?
ACLR stands for adjacent channel leakage ratio. It compares the average power in the assigned or wanted channel with the average power measured in an adjacent channel, using defined measurement bandwidths and filters. It is used to assess how much signal energy leaks into neighboring channels.
ACLR is commonly expressed in decibels. Under the usual ratio convention, a larger positive ACLR value indicates less adjacent-channel leakage relative to the wanted-channel power. Some instruments and reports display adjacent-channel power relative to the main channel using a negative dBc value instead. Always check the instrument’s definition, sign convention, integration bandwidth and channel offsets before comparing results.
ACLR is a system-relevant metric because spectral leakage can affect other users or services operating in nearby channels. The applicable limit depends on the radio standard, frequency allocation, waveform and regulatory requirements. An amplifier’s ACLR result must therefore be interpreted in the context of the complete test setup and applicable specification.
6. What Is EVM?
EVM stands for error vector magnitude. It measures the difference between a measured modulation symbol and its ideal reference position in the signal constellation. EVM is commonly reported as a root-mean-square (RMS) value, expressed as a percentage or in decibels relative to the reference magnitude.
A lower EVM generally indicates that the measured symbols are closer to their ideal positions. Amplifier nonlinearity can worsen EVM through AM-AM and AM-PM distortion, particularly when the signal peaks drive the amplifier into compression. However, EVM is not an amplifier-only measurement: noise, oscillator errors, IQ imbalance, phase noise, filtering and other transmitter or receiver impairments can also contribute.
For a meaningful comparison, record the modulation format, signal bandwidth, subcarrier or symbol configuration where relevant, measurement filter, equalization settings, output level and test equipment configuration. Compare EVM results only when the measurement methods and conditions are sufficiently similar.
7. AM-AM, AM-PM, ACLR and EVM: How They Differ
| Metric | What it describes | Typical view | Key interpretation |
|---|---|---|---|
| AM-AM | Output amplitude versus input amplitude | Amplitude transfer curve | Shows gain compression or expansion |
| AM-PM | Output phase change versus input amplitude | Phase shift curve | Shows amplitude-dependent phase distortion |
| ACLR | Wanted-channel power relative to adjacent-channel power | Modulated output spectrum and channel-power measurement | Shows adjacent-channel leakage under defined conditions |
| EVM | Measured symbols’ error relative to ideal reference symbols | Constellation or demodulation measurement | Shows modulation accuracy, including multiple possible impairments |
These metrics complement one another. AM-AM and AM-PM help describe the amplifier’s nonlinear transfer behavior. ACLR and EVM evaluate the effect on a modulated signal using defined measurement procedures. A good result in one metric does not automatically guarantee a good result in the others.
8. Why Output Back-Off Matters
Output back-off is the amount by which an amplifier is operated below a stated reference output level, such as its saturation level or another defined compression point. Operating with back-off can reduce the time or extent to which signal peaks enter a strongly nonlinear region. This can improve linearity, although the available output power and efficiency trade-off must also be considered.
The required back-off depends on the waveform’s peak-to-average power ratio, amplifier characteristics, bandwidth, linearity target and any linearization methods used. There is no single back-off value that is correct for every RF power amplifier or modulation scheme. Use measured performance under representative signal conditions rather than assuming a fixed relationship.
Some systems use digital predistortion (DPD) or other linearization techniques to compensate for amplifier nonlinearities. Their effectiveness depends on the amplifier, signal bandwidth, observation and feedback paths, calibration and implementation. Do not assume that a module supports DPD or that linearization removes the need to verify output limits and spectral performance.
9. How to Evaluate RF Amplifier Linearity in a Datasheet
When comparing RF power amplifier modules, check the following information and request additional test data where the published specification is not sufficient:
- Operating frequency or full frequency range for each stated linearity result.
- Output power at which AM-AM and AM-PM curves were measured.
- Whether P1dB is input-referred or output-referred, and how Psat is defined.
- Signal waveform, modulation, bandwidth and peak-to-average power ratio.
- ACLR measurement offsets, integration bandwidth, filters and sign convention.
- EVM reference, measurement bandwidth, demodulation settings and reported statistic.
- Supply voltage, temperature, load impedance, duty cycle and cooling conditions.
- Whether values are typical, guaranteed limits or results from a specific sample.
- Test equipment, calibration and measurement reference plane where relevant.
Do not compare a CW two-tone result with a modulated ACLR or EVM result as if they were the same type of measurement. Each test answers a different question. For procurement or design approval, use the current controlled datasheet and test documentation for the exact model and revision.
10. RF Amplifier Linearity Evaluation Checklist
- Define the waveform, modulation, channel bandwidth and required output level.
- Identify the applicable ACLR, EVM or other signal-quality requirements.
- Review AM-AM and AM-PM behavior across the intended operating range.
- Check P1dB, Psat and the proposed operating back-off using consistent definitions.
- Confirm the supply, temperature, load, duty-cycle and thermal test conditions.
- Request representative modulated-signal test data if signal quality is critical.
- Verify whether DPD or another linearization feature is actually supported by the selected system.
- Use authorized test facilities and comply with applicable spectrum and safety requirements.
Frequently Asked Questions
Are AM-AM and AM-PM the same as gain compression?
No. AM-AM describes the amplitude transfer characteristic and includes gain compression or expansion. AM-PM describes amplitude-dependent phase shift. Gain compression is one aspect of AM-AM behavior, while AM-PM captures a different form of distortion.
Does a good ACLR result guarantee good EVM?
No. ACLR measures adjacent-channel leakage, while EVM measures modulation error relative to ideal symbols. They are related to signal quality but respond differently to impairments and use different measurement procedures.
Can EVM be used to evaluate an RF power amplifier by itself?
EVM can be measured at an amplifier output as part of a defined test setup, but the result may include impairments from the signal source, test equipment and other components. A carefully documented setup is needed to isolate or characterize the amplifier’s contribution.
Does operating below P1dB guarantee linear performance?
No. P1dB is a useful gain-compression reference, not a universal guarantee of modulation quality. The required operating point depends on the waveform, peak-to-average ratio, phase distortion, bandwidth and applicable signal-quality limits.
Related RF Power Amplifier Resources
For amplifier specifications, read How to Read an RF Power Amplifier Datasheet and How to Choose an RF Power Amplifier. For large-signal output ratings, see the discussion of 100 W RF power amplifier module selection.
You can also review GaN vs. LDMOS RF power amplifier modules, RF power amplifier thermal design and RF system architecture and integration. Browse the QHXK RF power amplifier module category or contact QHXK to request the latest datasheet and model-specific test information.
Related RF Power Amplifier Guides: P1dB vs. Psat; RF power amplifier testing; operating classes and biasing.
RF equipment must be selected, tested and operated only in accordance with applicable spectrum authorizations, safety rules and export-control requirements. Product availability does not itself authorize radio transmission or deployment.