Measurements, Testing, and Calibration Power and Signal Measurement Informational

How do I measure the spurious emissions of a transmitter to verify regulatory compliance?

Q: What instruments do I need for spurious emission testing?

Minimum: (1) Spectrum analyzer with frequency range from 9 kHz to at least 10× the carrier frequency. For a 3.5 GHz transmitter: an analyzer to 26.5 GHz (available from Keysight, R&S, Anritsu). For 77 GHz: an analyzer to 110+ GHz (typically a base analyzer plus external harmonic mixer for mmWave). (2) High-power attenuator rated for the transmitter output power. (3) Calibrated test cables and adapters. (4) If radiated: anechoic chamber, calibrated receiving antennas, and antenna positioner. Cost: bench-top conducted test: $50,000-$200,000 (analyzer + accessories). Full radiated test (with chamber): $200,000-$2,000,000+. Note: if you do not have your own test lab: use an accredited third-party test lab. Cost for spurious emission test: $2,000-$10,000 per product.

Q: What about spurious emissions from digital circuits?

Digital signals (clock, data, bus activity) generate broadband emissions at the clock frequency and harmonics. A 100 MHz clock: harmonics at 200, 300, 400, ... MHz (decreasing in amplitude at approximately -20 dB/decade for a trapezoidal waveform). At the 50th harmonic (5 GHz): the harmonic level may be -40 to -60 dBm (depending on the rise time and PCB layout). If -40 dBm at 5 GHz exceeds the spurious limit (-13 dBm: it does not): no action needed. If it does exceed the limit: shield the digital section (a metal can over the digital IC area). Add filtering on all digital lines that cross from the digital section to the RF section. Improve the PCB ground plane (eliminate gaps and slots that allow digital noise to radiate). In practice: digital spurious emissions rarely exceed the regulatory limits at RF frequencies unless the PCB layout is poor (e.g., digital clock trace routed near the RF output).

Q: What is the difference between OOB and spurious emissions?

Out-of-band (OOB) emissions: are caused by the modulation process itself (the modulated signal has spectral components extending beyond the nominal channel bandwidth). OOB emissions are evaluated starting from the edge of the channel bandwidth, extending to ±2.5× the necessary bandwidth. The regulatory limit for OOB is typically a spectrum mask (e.g., -45 dBc at ±10 MHz offset for LTE). Spurious emissions: are caused by non-ideal hardware effects (harmonics, mixer products, oscillator leakage, parasitic oscillation). Measured outside the OOB domain (i.e., at frequencies more than ±2.5× the necessary bandwidth from the carrier). The regulatory limit for spurious is an absolute level (e.g., -13 dBm for FCC, or -36 dBm for ETSI Category B). Both must be measured and both must comply. The OOB measurement is typically done in the immediate vicinity of the carrier (±50 MHz for 5G NR). The spurious measurement covers the full frequency range from 9 kHz to 10× the carrier.

Measuring the spurious emissions of a transmitter for regulatory compliance requires a systematic approach that covers all frequency bands where the transmitter might generate unwanted signals: (1) Regulatory requirements: FCC Part 2/Part 22/Part 24/Part 27 (US): spurious emission limits for transmitters. Typical limit: -13 dBm (for licensed devices) or -41.3 dBm/MHz (for unlicensed Part 15 devices). ETSI EN 301 893 / EN 302 217 (EU): spurious emission limits for specific radio equipment types. ITU recommendation SM.329: categorizes spurious emissions into: out-of-band emissions (within ±2.5× the necessary bandwidth of the carrier), and spurious emissions (all other frequencies outside the OOB domain). (2) Measurement setup: the transmitter is connected to a calibrated spectrum analyzer through an attenuator (to protect the analyzer from the transmitter power). Frequency range: from 9 kHz to at least 10× the carrier frequency (or to the maximum frequency specified by the regulation). For a 5G NR transmitter at 3.5 GHz: measure from 9 kHz to at least 26.5 GHz (or higher). For a 77 GHz radar: measure from 9 kHz to 200+ GHz. Resolution bandwidth (RBW): set according to the regulation. Typically: 1 kHz for emissions below 150 kHz, 10 kHz for 150 kHz to 30 MHz, 100 kHz for 30 MHz to 1 GHz, and 1 MHz for above 1 GHz. The RBW affects the measured level of broadband noise (noise scales with sqrt(BW)). (3) Attenuator selection: the attenuator must handle the transmitter power without distortion. For a +30 dBm (1 W) transmitter: use a 30 dB attenuator rated for 1 W. The attenuator accuracy affects the measured emission level (an error in the attenuation value directly translates to an error in the emission measurement). (4) Measurement procedure: set the transmitter to full power at the nominal carrier frequency. Sweep the spectrum analyzer across the full frequency range. Compare the measured emission levels to the limit lines (regulatory limits). Identify any emissions that exceed the limits. If emissions exceed limits: investigate the source (harmonic of the carrier, mixing product, oscillator leakage, clock spurious) and apply filtering or shielding.

Category: Measurements, Testing, and Calibration

Updated: April 2026

Product Tie-In: Power Meters, Spectrum Analyzers, Signal Generators

Transmitter Spurious Emission Measurement

Spurious emission measurement is a regulatory requirement for all transmitters. Failing the spurious emission test prevents the product from being certified and sold.

Measurement Categories

(1) Harmonic emissions: the PA generates harmonics of the carrier frequency. Second harmonic (2f0): typically -20 to -40 dBc (relative to the carrier) at the PA output. After the harmonic filter: -50 to -70 dBc. Third harmonic (3f0): typically -30 to -50 dBc at the PA output. After filter: -60 to -80 dBc. The regulatory limit is typically -13 dBm absolute (or -43 dBm/MHz for ETSI). For a +30 dBm carrier: the harmonic must be < -13 dBm, i.e., -43 dBc or better. This requires a harmonic filter with at least -43 dB rejection at 2f0 (which is easily achievable with a low-pass filter). (2) Spurious mixer products: if the transmitter uses a frequency conversion (upconversion from IF to RF): the mixer produces spurious products at m×f_LO ± n×f_IF. These products can fall anywhere in the frequency spectrum. The mixer spurs are suppressed by: IF and RF bandpass filters, mixer balance (double-balanced mixers suppress even-order products), and image rejection. (3) Oscillator leakage: the local oscillator signal can leak through the mixer and appear at the antenna output. LO leakage is typically -30 to -60 dBc at the mixer output. Additional filtering and isolation are needed to meet the spurious limit. (4) Clock and digital spurious: digital clocks (reference oscillator, processor clocks) can couple to the RF output through: PCB crosstalk, ground bounce, and power supply coupling. These appear as narrowband spurs at the clock frequency and its harmonics. Mitigation: proper PCB layout (separate digital and RF sections), shielding, and filtering on the DC supply.

Antenna Port Measurement

(1) Conducted measurement: the transmitter output is connected via cable to the spectrum analyzer (through an attenuator). This measures the conducted emissions at the antenna port. This is the standard regulatory measurement for the transmitter module. (2) Radiated measurement: the transmitter (with antenna) is placed in an anechoic chamber. The radiated emissions are measured using a calibrated receiving antenna and spectrum analyzer. This measures the total radiated power at each frequency (including the antenna pattern). Radiated measurement is required for: devices with integrated antennas (where the conducted measurement is not possible), and for final product certification (the enclosure and cable routing can create additional spurious radiations). (3) Far-field distance: for mmWave: the far-field distance is short (2D²/λ). For a 10 cm antenna at 77 GHz: far field starts at 2 × (0.1)² / 0.0039 = 5.1 m. The anechoic chamber must be at least this large (or use a compact antenna test range (CATR) for near-field to far-field transformation).

Spurious Emission Limits

FCC spurious limit: < -13 dBm (licensed TX)
Harmonic at PA: -20 to -40 dBc, after filter: -50 to -70 dBc
Required rejection: carrier_dBm - limit = filter dB
RBW: 1MHz above 1 GHz (standard)
Mismatch correction: same as power measurement

Common Questions

Frequently Asked Questions

What instruments do I need for spurious emission testing?

Minimum: (1) Spectrum analyzer with frequency range from 9 kHz to at least 10× the carrier frequency. For a 3.5 GHz transmitter: an analyzer to 26.5 GHz (available from Keysight, R&S, Anritsu). For 77 GHz: an analyzer to 110+ GHz (typically a base analyzer plus external harmonic mixer for mmWave). (2) High-power attenuator rated for the transmitter output power. (3) Calibrated test cables and adapters. (4) If radiated: anechoic chamber, calibrated receiving antennas, and antenna positioner. Cost: bench-top conducted test: $50,000-$200,000 (analyzer + accessories). Full radiated test (with chamber): $200,000-$2,000,000+. Note: if you do not have your own test lab: use an accredited third-party test lab. Cost for spurious emission test: $2,000-$10,000 per product.

What about spurious emissions from digital circuits?

Digital signals (clock, data, bus activity) generate broadband emissions at the clock frequency and harmonics. A 100 MHz clock: harmonics at 200, 300, 400, ... MHz (decreasing in amplitude at approximately -20 dB/decade for a trapezoidal waveform). At the 50th harmonic (5 GHz): the harmonic level may be -40 to -60 dBm (depending on the rise time and PCB layout). If -40 dBm at 5 GHz exceeds the spurious limit (-13 dBm: it does not): no action needed. If it does exceed the limit: shield the digital section (a metal can over the digital IC area). Add filtering on all digital lines that cross from the digital section to the RF section. Improve the PCB ground plane (eliminate gaps and slots that allow digital noise to radiate). In practice: digital spurious emissions rarely exceed the regulatory limits at RF frequencies unless the PCB layout is poor (e.g., digital clock trace routed near the RF output).

What is the difference between OOB and spurious emissions?

Out-of-band (OOB) emissions: are caused by the modulation process itself (the modulated signal has spectral components extending beyond the nominal channel bandwidth). OOB emissions are evaluated starting from the edge of the channel bandwidth, extending to ±2.5× the necessary bandwidth. The regulatory limit for OOB is typically a spectrum mask (e.g., -45 dBc at ±10 MHz offset for LTE). Spurious emissions: are caused by non-ideal hardware effects (harmonics, mixer products, oscillator leakage, parasitic oscillation). Measured outside the OOB domain (i.e., at frequencies more than ±2.5× the necessary bandwidth from the carrier). The regulatory limit for spurious is an absolute level (e.g., -13 dBm for FCC, or -36 dBm for ETSI Category B). Both must be measured and both must comply. The OOB measurement is typically done in the immediate vicinity of the carrier (±50 MHz for 5G NR). The spurious measurement covers the full frequency range from 9 kHz to 10× the carrier.

Keep Reading