Logo PTB

Radio-frequency field strength

Retraceability to SI basic units: Realization of electric and magnetic radio-frequency field strength

At PTB electromagnetic fields with a field strength as well defined as possible are used as default for the calibration of portable field strength sensors (known as ‘radiation monitors’). Naturally, these fields cannot be preserved as embodied measurement standards. Instead, these physical entities need to be available on demand (‘realized’) in a standard measurement equipment anytime. This task requires generation methods where the field strength is derived from the physical basic units (‘retraceability’). Here it is assumed that all parts of the test equipment are calibrated and hence retraced. In this case the field can be derived from this data and physical laws, only. In the frequency and field strength range of interest we assume classical electrodynamics (see Jackson, John David) represented by Maxwell’s equations to be valid.

In particular for the magnetic field strength (H, measured in A/m) such a retracement seems to be quite simple since only the basic entities length (in meters) and current (in Amperes) are involved. In practice the realization of the relevant RF field entities – electric and magnetic field strength (E, H) or energy flux density (S) – requires a multitude of intermediate steps and measurement units that are specially suited for RF technology. For the realization of fields the Working Group 2.21 uses methods based on RF real power, scattering parameters (see Michel, Hans J.) and, if applicable, mechanical dimensions, only. For these entities rf measurement equipment (power meters, vector network analyzers) with a small measurement uncertainty is available. It can be retraced to the respective calibration standards of the Opens internal link in current windowWorking Group 2.22 – High Frequency Measurement Techniques.

Opens internal link in current windowField Generators

Opens internal link in current windowOrder Processing

For calibration the generated fields need to have certain well defined and reproducible properties:

the energy transport will be in one direction only as linearly polarized transversal electromagnetic wave (so called ‘TEM’ wave)

the far-field conditions are fulfilled, which means:

  • electric and magnetic field exist simultaneously
  • field vectors are perpendicular to each other and to the propagation direction and oscillate in phase
  • their amplitudes are in a fixed ratio (E/H~377 Ohm).

Depending on frequency or wavelength, respectively, the electromagnetic waves and the technical equipment have different properties. Especially at lower frequencies emitter antennas are quite large and the performance of absorber cabins is rather poor. The determination of the antenna gain is problematic. In addition, reflections from the cabin walls overlap the calculated field and increase the overall uncertainty. Hence, ‘TEM waveguides’ are used in this range instead, but have a rather limited volume. At higher frequencies the performance of cabin absorbers improves and the gain of horn antennas can be determined with small uncertainty. For the calibration of radiation monitors it is suitable to split the frequency range at approximately 1 GHz. In the lower frequency region the Work Group 2.21 generates waves in special ‘TEM’ and ‘GTEM’ cells, whereas the radiation at higher frequencies is generated via horn antennas in absorber cabins.


Metrological Properties of RF Radiation Monitors

In contrast to most other metrological equipment the display value of a RF field monitor does not only depend on the physically defined quantity of the field strength but also on a number of further significant influences. A calibration should take such influences into consideration, especially problems due to

  • frequency dependence and non-linearity
  • anisotropy
  • additional error sources
  • technical imperfections of the radiation monitors

Frequency dependence and non-linearity

During the calibration in unmodulated fields at a given frequency the frequency dependent sensitivity and the linearity of the radiation monitor will be determined. After exposition of the radiation monitor at defined test values, the calibration factor is determined from the representation value and the display value of the field strength. Later this value can be used to correct frequency characteristics and non-linearity of the radiation monitor since the frequency of the field to be measured is usually known or can be determined easily. While the frequency dependence is usually construction-conditioned, sensitivity and non-linearity are sensor-specific and hence statistically spread. Problems arise if a superposition of fields containing different frequencies is measured with the sensor possessing different sensitivities. In this case it is not possible to correct the display values with calibration data.


Despite constructive optimization the display value of a radiation monitor is only approximately isotropic i. e. independent of direction of incidence and polarisation of the electromagnetic wave. This term is often misunderstood since in some publications and calibration certificates ‘anisotropy’ simply denotes the sensitivity variation of a calibration during which the sensor is rotated around an axis oriented under the so called ‘analytical angle’ in the lineary polarized field. Provided an appropriate construction of the sensor, all three antenna elements will be oriented in parallel/orthogonal to the field vector one after another. This procedure is therefore well suited for a functional test but is better named as ‘rotational asymmetrie’. In contrast to this the ‘anisotropy’ describes the maximum variation of sensitivity for arbitrary orientations of the whole field strength monitor (full device exposition including sensor, connection cable and display unit) in the field. This appears to be more appropriate for the practical application since during personal safety and EMC measurements usually the whole radiation monitor is exposed to a field with often unknown structure.

Practical experience shows that for most devices anisotropy is construction-conditioned and that it shows a distinctive frequency dependence (often including resonances). Therefore the determination of anisotropy during calibration is of cardinal importance but rather complex. First the field generator of the calibration laboratory needs to provide a sufficiently large volume to orient the complete device in arbitrary directions while the center of the sensor remains at a fixed point. Furthermore, a full frequency dependence has to be measured in each orientation. In order to keep the effort (and hence the calibration costs) in sustainable limits, the whole procedure has to be automatized. In addition to this the number of examined orientations has to be restricted for instance to three main positions of the device that are clear from the construction and for which can be assumed that the measured frequency dependences represent the extrema of the sensitivity.

Additional error sources

For personal safety or EMC measurements normative reference values are usually given for undisturbed, ‘empty’ field strengths which would require an undisturbed field without a person or a device under test in it. Since also the field strength monitor itself represents an unavoidable disturbance of the field, this measurement task can only be solved approximately. Using the undisturbed ‘empty’ field strength as reference value during a calibration one assumes that the disturbance due to the field strength monitor during calibration is comparable to its disturbance during the later measurement application. This includes the disturbance of the field strength monitor into the calibration but requires that the operation of the device during the calibration process is similar to its operation during the measurement including mounting and cables.

Technical imperfections of the radiation monitors

This includes typical influences for electronic measurement equipment as zero point stability, restricted resolution of digitalization and ambient temperature. These influences have to be considered in the measurement uncertainty budget either by own examinations during calibration (time intensive!), manufacturer’s data or estimates due to experience. Some radio frequency field sensors suffer from further unwanted influences. For instance magnetic field sensors might be sensitive to electric fields also present in the far-field.

Furthermore to be examined: cross-sensitivity of the sensor to orthogonal field components, amplitude- and pulse modulation, mixing products for multi-frequency signals, ambient temperature dependence, long term stability and mechanical sensitivity.