RF Guide
RF Receiver Dynamic Range: Blockers, Compression, and Intermodulation
A receiver can detect a weak signal in a quiet lab and fail in the presence of a strong neighbor. Design both ends of the input range, with explicit reference planes and a measured performance criterion.
18 min · Advanced · Updated October 16, 2026
Define the Interference Test First
Record wanted frequency, modulation, bandwidth, target error rate or detection probability, input sensitivity, antenna or cable reference plane, blocker frequency offset, waveform, and gain-control state. Two receivers can have the same thermal noise floor and different blocker tolerance. The RF Noise Figure and Sensitivity Guide establishes the quiet-channel floor; this guide starts where interference and large-signal behavior take over.
The RF Link Budget Calculator compares received wanted power with a sensitivity input. That margin is necessary but not sufficient when a blocker is present. A -100 dBm sensitivity and -70 dBm wanted signal imply 30 dB quiet-channel margin, yet a much stronger adjacent signal may still desensitize the receiver.
Keep Power and Gain References Consistent
Formula reference
Input-referred screening relationships
Pout,small(dBm) = Pin(dBm) + G(dB)IIP3(dBm) = OIP3(dBm) - G(dB)IP1dB ≈ OP1dB - Gsmall (specified test only)PIM3,in(dBm) ≈ 3Ptone,in(dBm) - 2IIP3(dBm)Variable definitions
- G is small-signal power gain between stated reference planes
- IIP3/OIP3 are input/output-referred extrapolated intercepts
- P1dB is a measured compression point, not a universal safe maximum
- The IM3 relation assumes two equal tones and an asymptotic third-order region
The RF Power Converter & Calculator converts dBm, watts, and dB ratios. Use it to avoid mixing absolute power with gain. Do not sum two dBm levels directly; convert to linear power first when total incoherent power is the relevant quantity.
A cable before the LNA lowers both wanted and blocker levels at the LNA input, while also raising the receiver input-referred noise figure through loss. A filter after an overloaded LNA cannot undo the distortion already generated. Compare all stages at their own input planes, including switches, duplexers, filters, connectors, and ADC drivers.
Compression Is Not an Intercept Point
A device P1dB is observed on a specified single-tone power sweep, where output gain has fallen by 1 dB relative to the extrapolated linear line. It is not the input level at which every receiver function stops working. Aggregate power, modulation crest factor, frequency, impedance match, bias, temperature, and gain-control mode can move the effective headroom.
Suppose input P1dB is -5 dBm and a single blocker reaches -10 dBm at that same reference plane. The nominal difference is 5 dB, not a guarantee of error-free reception. A second blocker or a high-crest-factor waveform can produce larger peaks; device tolerances and the wanted signal still matter. Test the specified blocker waveform across frequency and gain states.
| Failure mode | Mechanism | Required check |
|---|---|---|
| Gain compression | Large blocker drives a stage outside small-signal gain | Single-tone P1dB is only a screening reference |
| Third-order intermodulation | Two blockers create products near a wanted channel | Two-tone IIP3 estimate applies in its asymptotic region |
| Reciprocal mixing | LO phase noise spreads blocker energy into channel | Needs oscillator and mixer phase-noise data |
| ADC overload | Composite input exceeds analog or converter range | Check peak waveform and AGC behavior |
| Spur or alias | Clock, mixer, or sampling products overlap wanted band | Frequency planning and physical test are needed |
| Front-end loss | Preselector improves blocking but worsens NF | Evaluate insertion loss before LNA |
Two-Tone Intermodulation Example
For an equal-tone test with each input tone at -20 dBm and input IP3 of 0 dBm, the third-order input-referred product extrapolates to 3(-20) - 2(0) = -60 dBm, or 40 dB below each input tone. With 20 dB small-signal gain, each output tone is near 0 dBm and the estimated IM3 product is near -40 dBm. This is a slope-based estimate while the device remains in the appropriate weakly nonlinear region. Do not use it near compression or for arbitrary wideband modulation without a suitable model.
Whether an IM3 product harms reception depends on its frequency. Two blocker frequencies can produce products at 2f1-f2 and 2f2-f1; only products in or near the wanted channel under the actual filtering and modulation conditions threaten that channel. A single IIP3 number does not describe every frequency pair, gain state, or temperature.
Trade Noise Against Preselection
A preselector attenuates out-of-band blockers before the first active device, but insertion loss ahead of the LNA worsens input-referred sensitivity. Compare the reduction in blocker level against loss of wanted signal and added noise figure. Use measured filter response at blocker frequency, not only its nominal passband loss; leakage around PCB, enclosure, and switch paths can bypass a carefully calculated filter.
Gain distribution also matters. High early gain can protect overall NF but push later mixer or ADC stages toward overload. Lower gain may improve blocker handling while making later-stage noise more visible. Analyze the complete chain with measured stage gains, NF, P1dB or other compression data, and IP3 where relevant. These quantities are operating-point dependent.
Bench Validation Matrix
| Sweep | Observe | Keep fixed |
|---|---|---|
| Wanted near sensitivity | BER/PER or detection criterion with no blocker | Bandwidth, modulation, coding, gain state |
| Single blocker power | Sensitivity shift and onset of compression | Reference plane and blocker offset |
| Blocker frequency | Desensitization versus offset and preselector response | Wanted level and waveform |
| Two equal tones | Third-order products versus tone power | Tone spacing and linear-region condition |
| Gain state | AGC transitions and ADC clipping | Wanted and blocker settings |
| Temperature and supply | Worst-case NF, gain, and linearity | Fixture calibration and device configuration |
Calibrate signal-generator power at the receiver input plane, accounting for combiners, attenuators, cables, and connector loss. Isolate generator intermodulation from receiver intermodulation. Record actual gain mode; an AGC transition can look like compression or hide it. Repeat with the intended antenna, board, enclosure, and supply when the laboratory front-end is not representative of the final product.
Common Mistakes
- Calling quiet-channel sensitivity the complete dynamic range.
- Adding dBm levels as if they were ordinary numbers.
- Applying output IP3 directly at the receiver input plane.
- Using two-tone IP3 as a hard maximum-power rating.
- Extrapolating the IM3 formula into compression.
- Ignoring pre-LNA insertion loss when adding a filter.
- Expecting a post-LNA filter to remove distortion created earlier.
- Checking only one blocker frequency or gain state.
- Comparing P1dB values measured with different bias and match.
- Ignoring generator and combiner intermodulation in a bench test.
Support reference
FAQ
What is receiver dynamic range?
It is the span between a defined minimum detectable wanted signal and a defined maximum signal or interference condition under which the receiver still meets its specified performance criterion.
Is sensitivity enough to prove blocker tolerance?
No. Sensitivity describes a specified weak-signal test. A strong nearby signal can compress a stage, create intermodulation, overload an ADC, or mix reciprocal noise into the wanted channel.
What is input P1dB?
Input P1dB is an approximate input level at which actual output is 1 dB below the extrapolated small-signal gain line under the specified single-tone test conditions.
How is IIP3 used?
IIP3 is the input-referred third-order intercept obtained from a two-tone test in the small-signal region. It supports an asymptotic IM3 estimate but is not a physical operating power limit.
Does an IIP3 calculation predict every blocker failure?
No. It does not cover compression, reciprocal mixing, filters, spurs, ADC overload, large-signal nonlinearities, or device behavior outside the extrapolation range.
Why does gain before a filter matter?
An unfiltered blocker may overload the first active stage before later selectivity can attenuate it. Moving selectivity earlier can improve blocker handling but may add insertion loss before the LNA.
Can I add two dBm values to combine blockers?
No. Convert each absolute dBm level to linear power, sum the powers where physically appropriate, then convert the sum back to dBm. Coherent signals require phase-aware analysis.
What does desensitization mean?
A receiver is desensitized when a blocker degrades the wanted signal's detection margin, error rate, or measured sensitivity relative to the no-blocker condition.
How should blocker performance be verified?
Test wanted signal near its specified sensitivity while sweeping blocker frequency and power, with documented modulation, bandwidth, BER/PER or detection target, reference plane, and gain settings.
Summary
Start with a defined weak-signal criterion, then test realistic strong-signal conditions at a consistent input plane. Use P1dB and IIP3 as bounded screening tools, balance preselection against noise, and measure desensitization on the complete receiver.