Test and Measurement Equipment Instrument Selection Informational

How do I select a power meter and sensor for accurate measurements at millimeter wave frequencies?

Q: Can I use a sub-6 GHz power sensor at mmWave?

No. A sensor rated to 18 GHz (SMA connector) cannot measure at 28 GHz. The SMA connector is physically incapable of propagating signals above 18-26.5 GHz (depending on the specific SMA variant). The sensor response is undefined above its rated frequency. You must use a sensor with the correct connector type and frequency rating for mmWave work.

Q: How do I measure EIRP at mmWave?

EIRP cannot be measured with a power meter at the connector (because many mmWave devices have integrated antennas with no RF port). Instead: measure the radiated power in an anechoic chamber using a calibrated reference antenna. EIRP = received power + path loss - reference antenna gain. The path loss is calculated from the known distance and frequency. This is OTA (Over-the-Air) power measurement, which is the standard method for 5G FR2 devices.

Q: What about power measurement above 110 GHz?

For frequencies above 110 GHz (sub-THz, 6G research): Virginia Diodes (VDI) provides waveguide power sensors to 500 GHz. Erickson PM5B calorimeter: precision power measurement to 2 THz. Thomas Keating absolute power meter: traceable measurements to THz. These are specialized research instruments, costing $10,000-50,000 each. The measurement accuracy at sub-THz is ±0.3-1.0 dB (lower than mmWave due to calibration challenges).

How do I select a power meter and sensor for accurate measurements at millimeter wave frequencies? Accurate power measurement at mmWave (above 26 GHz) requires sensors specifically designed for these frequencies, with attention to connector type, frequency response flatness, and calibration traceability: (1) Power sensor types for mmWave: thermocouple sensors: frequency range: DC to 50-110 GHz. Power range: -30 to +20 dBm. Accuracy: ±0.02-0.05 dB (highest accuracy at mmWave). Response: slow (100 ms-5 seconds). Best for: calibration and reference measurements. Examples: Keysight 8487D (50 GHz, 2.4 mm), Keysight W8486A (75-110 GHz, waveguide WR-10). Diode sensors: frequency: 10 MHz to 50-67 GHz. Power range: -70 to +20 dBm (wide dynamic range). Accuracy: ±0.1-0.2 dB at mmWave (frequency response becomes less flat). Response: fast (< 10 ms). Examples: Keysight U2004A (50 GHz, 2.4 mm), R&S NRP-Z58 (40 GHz). Peak/Modulated sensors: needed for 5G FR2 TDD signals (pulsed, time-gated measurements). Must capture the peak envelope and average power independently. Examples: Keysight N1924A with N8487A sensor (50 GHz, 2.4 mm, peak + average). (2) Connector considerations: 2.4 mm connector: rated to 50 GHz. The standard connector for mmWave power sensors below 50 GHz. 1.85 mm connector: rated to 67 GHz. Used for V-band power measurements. 1.0 mm connector: rated to 110 GHz. Required for W-band measurements. Waveguide flanges: WR-15 (50-75 GHz), WR-10 (75-110 GHz). Required for above 67-70 GHz. Each connector requires mating cables and adapters of matching type. Cross-type adapters (e.g., 2.4 mm to 1.85 mm) must be high quality to avoid reflections and measurement error. (3) Accuracy at mmWave: the dominant error sources at mmWave are: frequency response flatness: the sensor response varies more with frequency at mmWave (±0.5-1.5 dB without correction). Calibration factors (CF): frequency-dependent correction factors provided by the sensor manufacturer reduce this to ±0.1-0.3 dB. Mismatch uncertainty: at mmWave, the SWR of both the source and sensor is typically higher (1.3-1.8:1 vs 1.1:1 at lower frequencies). Mismatch uncertainty: ±0.2-0.8 dB at mmWave (dominant error source). Connector repeatability: each connection at mmWave introduces ±0.02-0.05 dB of uncertainty. Minimize the number of connections and adapters. (4) Cost: mmWave thermocouple sensor (50 GHz): $3,000-6,000. mmWave waveguide sensor (110 GHz): $8,000-15,000. Power meter (2-channel): $5,000-12,000. USB power sensor (eliminates separate meter): $3,000-8,000.

Category: Test and Measurement Equipment

Updated: April 2026

Product Tie-In: VNAs, Spectrum Analyzers, Signal Generators

mmWave Power Measurement

Power measurement at mmWave is a fundamental but challenging task, as the measurement uncertainties increase significantly compared to sub-6 GHz work.

Calibration and Traceability

(1) mmWave power sensors must be calibrated with traceability to national standards (NIST, PTB, NPL). The calibration includes: frequency response correction factors at each calibration frequency (typically every 0.5-2 GHz). Linearity verification across the power range. Reflection coefficient (SWR) at each frequency. Calibration interval: typically 12 months. Calibration cost: $500-2,000 per sensor (mmWave calibration is more expensive than sub-6 GHz). (2) Zero and reference calibration: before each measurement session: zero the sensor (remove all RF input, allow the sensor to measure its own offset). Calibrate using the power meter built-in 50 MHz reference oscillator (typically +1.00 dBm). At mmWave: the reference oscillator does not verify the mmWave frequency response. For the highest accuracy: verify the sensor against a known power level at the test frequency using a calibrated signal generator.

mmWave Power Sensor Specs

Thermocouple: ±0.02-0.05 dB accuracy (best)
Diode: ±0.1-0.2 dB at mmWave
Mismatch uncertainty: ±0.2-0.8 dB (dominant error)
Connectors: 2.4mm (50 GHz), 1.0mm (110 GHz)
Cost: $3K-15K per mmWave sensor

Common Questions

Frequently Asked Questions

Can I use a sub-6 GHz power sensor at mmWave?

No. A sensor rated to 18 GHz (SMA connector) cannot measure at 28 GHz. The SMA connector is physically incapable of propagating signals above 18-26.5 GHz (depending on the specific SMA variant). The sensor response is undefined above its rated frequency. You must use a sensor with the correct connector type and frequency rating for mmWave work.

How do I measure EIRP at mmWave?

EIRP cannot be measured with a power meter at the connector (because many mmWave devices have integrated antennas with no RF port). Instead: measure the radiated power in an anechoic chamber using a calibrated reference antenna. EIRP = received power + path loss - reference antenna gain. The path loss is calculated from the known distance and frequency. This is OTA (Over-the-Air) power measurement, which is the standard method for 5G FR2 devices.

What about power measurement above 110 GHz?

For frequencies above 110 GHz (sub-THz, 6G research): Virginia Diodes (VDI) provides waveguide power sensors to 500 GHz. Erickson PM5B calorimeter: precision power measurement to 2 THz. Thomas Keating absolute power meter: traceable measurements to THz. These are specialized research instruments, costing $10,000-50,000 each. The measurement accuracy at sub-THz is ±0.3-1.0 dB (lower than mmWave due to calibration challenges).

Keep Reading