RF measurement instruments and calibration
How to choose RF exposure measurement instruments and document calibration, traceability, and field verification defensibly.
Offline pages
Saved on this device only. Connect at least once every 30 days to confirm Premium. Sign-out or a change of account removes saved copies.
On this page
You stopped at:
ContinueA meter’s displayed number is an indication. Turning it into a defensible RF field result requires a defined quantity, a suitable sensor, the correct calibration relationship and the actual measurement chain. A self-test can pass while the probe, frequency range or conversion model is wrong for the job.
Build on RF field quantities and near- and far-field conditions. First write the measurand: for example, RMS electric field at a specified point, over stated bands and time, under a stated source condition. Transmitter output power, received antenna power and local field strength are different quantities.
Match the sensor to the question
A broadband field probe can map candidate maxima quickly, while a frequency-selective system can resolve band contributions. An E-field sensor does not automatically measure H, absorbed power or SAR. A feeder power sensor measures guided power; a personal monitor’s mounting and response may suit an alarm or trend without making its reading a complete exposure assessment.
Check usable frequency coverage, dynamic range, detector response, isotropy or polarisation, environmental conditions and spatial resolution. The probe should not substantially disturb the field it is meant to sample. A conventional receiving antenna’s far-field antenna-factor calibration cannot automatically be applied beside a source with a nonuniform near field. Select the method for the required quantity and field conditions before interpreting a convenient display unit.
Keep calibration, traceability and verification distinct
Calibration relates reference quantities and instrument indications, with uncertainties, under specified conditions, and provides a way to obtain results from indications. It need not adjust the instrument or declare it conforming to a requirement. See VIM 2.39.
Metrological traceability concerns the measurement result: a documented calibration chain connects it to a reference, with uncertainty contributions. A certificate or accreditation logo alone does not establish that your result used the right probe, conditions and conversion. See VIM 2.41.
Verification uses evidence to test stated requirements. A field functional check is one example; verification is not restricted to a quick self-test. Record what requirement was tested and how. Passing that test does not extend a calibration to a different probe or frequency. See VIM 2.44.
Read the record for the actual sensor and meter or characterised chain: identity, frequency points or range, correction convention, conditions, uncertainty and coverage factor. Distinguish as-found from as-left results and any adjustment. Calibration intervals follow the applicable instrument and quality-system policy, use and drift history; suspected damage or unexplained check-source change needs investigation rather than an assumption that an unexpired date proves valid operation.
Interpret one certificate contribution
Worked example 1
Recover standard uncertainty
Problem
A calibration certificate gives relative expanded uncertainty U = 10% with coverage factor k = 2. Find the corresponding relative standard uncertainty u.
Show full solution
Undo the coverage factor
Method
Divide expanded uncertainty by the stated coverage factor.Reason
Expanded uncertainty is defined as U = ku, so division recovers standard uncertainty.Working
Answer
u = U/k = 10%/2 = 5%.State what the result covers
Working
Answer
The relative standard uncertainty is 5% for this calibration contribution. It is not the whole survey uncertainty; positioning, source variation and other method contributions must also be considered.
Reconstruct a selective measurement chain
For a characterised, linear 50 Ω receiver chain, convert received power to loaded RMS voltage with V = square root of (P(50 Ω)). In decibel units, L_(V,receiver) = L_(P,dBm) + 106.9897… dB, where the voltage level uses the 1 μV reference. The constant depends on 50 Ω; it is not a universal power-to-voltage conversion.
If cable loss Lc and external attenuation A have not already been corrected, and a preamplifier has gain G, then L_(V,antenna) = L_(V,receiver) + Lc + A-G. Losses reduced what reached the receiver, so reconstructing the antenna voltage adds them. Gain increased it, so remove the gain. An antenna factor Kf relates field to the specified loaded antenna voltage: E = KfVₐₙₜₑₙₙₐ. Its level is 20 log ₁₀[Kf/(1 m⁻¹)] in dB/m. Add that level to obtain a field level referenced to 1 μV/m, then convert to V/m.
Some instruments already apply antenna, cable or gain corrections. Document the enabled correction tables and reference plane; adding them again would double-correct the result. Keysight’s field-strength documentation illustrates instrument-specific correction tables and their signs. Use the actual equipment’s convention and calibration, not a menu name alone.
Guided practice 2
Recover the field from a receiver indication
Problem
A receiver reports raw, uncorrected integrated power −20.0 dBm into 50 Ω for a stable narrowband signal. At the measured frequency, the characterised chain has 3.0 dB cable loss, 6.0 dB external attenuation and 20.0 dB preamplifier gain. A suitably aligned receiving antenna has factor 20.0 dB/m, defined using its 50 Ω loaded RMS voltage. The chain is linear, corrections are constant over the captured bandwidth, the full signal is captured, and the far-field calibration conditions apply. Find receiver voltage level in dBµV, antenna RMS voltage in mV and the corresponding electric-field component in V/m. Explain the correction signs and why this is not automatically an isotropic total-field result.
Work through the model, then compare your reasoning
Hints
Hint 1
Check the model
Write the receiver-to-antenna correction separately from the antenna-factor conversion. All numbers refer to the same measured frequency.Show solution step by step
Convert the receiver power
Method
Use the dBm power reference and the stated resistance.Reason
A voltage conversion requires a load; dBm alone is a power level.Working
Answer
P = 10⁻³10^(-20/10) = 10⁻⁵ W, so V_receiver = square root of 50(10⁻⁵) = 0.0223607… V. Thus its level is 20 log ₁₀[V/(1 μV)] = 86.9897… dBμ V.
Move the reference plane and apply the factor
Method
Add losses, subtract gain, then use the loaded antenna factor.Reason
The receiver sees a voltage changed by the chain; antenna factor relates a different reference plane to the field.Working
Answer
Net chain correction is 3 + 6-20 = -11 dB. Antenna voltage level is 75.9897… dBμ V, giving Vₐₙₜₑₙₙₐ = 6.30210… mV. The factor is 10^(20/20) = 10 m⁻¹, so E = 0.06302096… V/m. This is the component for the stated antenna alignment and model. Other polarisations, bands, uncertainty and the exposure method remain separate questions; the numerical check does not establish survey validity.
Update a correction after changing a cable
Check your understanding 1
A frequency-selective instrument displays 0.120 V/m for a stable narrowband signal after applying its enabled antenna, gain, attenuation and cable tables. Its cable table still contains 3.0 dB loss, but the cable actually used for this reading has characterised loss 6.0 dB at that frequency. All other correction relationships remain applicable; the chain is linear and the stated calibration geometry is satisfied. Find the corrected field component by accounting for this table mismatch. Explain which correction is missing and what should change in the measurement record. This exercise supplies nominal corrections, not a complete uncertainty budget or exposure decision.
Show answer
The missing loss correction is 6.0-3.0 = 3.0 dB. The display already includes the other corrections and 3 dB of cable loss, so E_corrected = (0.120)10^((6.0-3.0)/20) = 0.1695045… V/m ≈ 0.170 V/m. Greater cable loss hid more of the antenna signal, so the reconstructed field is larger. Adding all 6 dB to this already corrected display would count the original 3 dB twice.
Update the applicable cable table for subsequent readings and record cable identity, characterised loss, frequency and whether a correction was applied by the display or afterwards. Retain the indication and correction history so this result can be reconstructed; check the chain and uncertainty before using it as a survey result. Compare your reasoning with the sign, the 3 dB increment and the amplitude conversion. The scalar check verifies only this supplied conversion, not the field setup or conformity.
Verify the chain you will actually use
Before collecting final readings, check identity and condition, permitted settings and range, the supported reference-source or functional-check procedure, and repeatable geometry. A rotation check can reveal a response or handling problem; it cannot by itself diagnose damage or establish full isotropy calibration. Investigation after a drop, overload warning or check-source change should follow the instrument procedure.
Keep meter and sensor identities, calibration/correction records, cable and accessory configuration, detector and integration settings, geometry, operator position, source state, included bands and uncertainty with the result. If a chain component changes, reconsider its corrections and verification before treating later readings as comparable. The report should allow another person to reconstruct the measurand and conversion, including limitations, rather than merely see the word “calibrated”.
Decide what a certificate and self-test establish
Check your understanding 2
A report applies the correction and 5% standard uncertainty from probe A’s serial-specific 1–3 GHz calibration to probe B, which is the same model. Probe B was used at 2 GHz and passed the meter’s self-test, but no calibration or other characterisation record covering B is supplied. Which claims do the supplied records support? What evidence is missing before claiming B’s result has that calibration relationship and uncertainty?
Show answer
The record characterises A under its stated conditions; the self-test supports only the requirements that test actually checks. Being the same model and operating inside A’s frequency range do not establish B’s correction, uncertainty or the traceability of B’s result. Obtain the calibration or valid characterisation evidence covering B and its actual chain and conditions, then apply the relevant relationship and uncertainty budget. Do not silently transfer A’s serial-specific values or relabel the self-test as calibration. This is a record-scope finding, not proof that B is faulty.
Review the measurement decisions
Check your understanding 3
What’s the difference between calibration and verification?
Show answer
Calibration supplies the indication-to-result relationship under stated conditions with uncertainty. Verification tests specified requirements using evidence; a functional field check is one kind. Its scope must be stated, and it must not be relabelled as calibration.
Check your understanding 4
In a selective system, why do settings belong in the method description?
Show answer
Because detector choice, RBW/VBW, sweep time, averaging, and band integration affect what energy is captured and how it is converted into RMS field or exposure ratio.
Check your understanding 5
Why should probe serial numbers be in your report?
Show answer
Because the probe (or antenna factor) is typically the calibrated sensor. Swapping probes changes the calibrated measurement relationship.
What to keep
Define the quantity, match the sensor and apply the actual chain corrections at the right reference plane. Calibration, traceability and verification answer different questions; a valid result needs records and uncertainty covering its real instrument, conditions and method.
Spotted an error? Report a correction
Syllabus and review details
No official syllabus alignment is listed for this lesson.
Last reviewed: