Two ways of guiding the wave
In a transmission line, such as a coaxial section or a waveguide, the field is confined and the sample occupies the whole cross-section. In free space, the wave travels between antennas, typically with lenses, and strikes a sheet. In both cases, scattering parameters are measured and inverted with a model to extract permittivity and, where applicable, permeability.
Transmission line
- Confinement gives a good signal-to-noise ratio and well-established models.
- The sample must fill the cross-section exactly. An air gap between sample and wall biases the result, especially in high-permittivity materials.
- Each line size covers one frequency band, so covering a wide range requires several samples and sections.
Free space
- There is no contact: the sample does not need to be machined to an exact shape; all that is required is a flat, homogeneous sheet that is large enough.
- It is suitable for high frequencies and for tests at elevated temperature, since only the sample goes into the furnace.
- Care must be taken with edge diffraction, alignment and reflections between antennas, which are mitigated by calibration with references and time-domain filtering.
How to choose
- If the sample can be machined precisely and the band is limited, a line usually gives lower uncertainty.
- If the sample is large or fragile, or must be measured hot, free space is more practical.
- If a material needs to be measured over a very wide band or at very high frequencies, free space avoids changing lines.
What the engineer sees
In free space, ripples from multiple reflections and a sensitivity to thickness and angle appear. In a line, the results are smoother but respond to any mechanical play of the sample.
When it does not apply
With very low-loss materials and a need for high accuracy, a resonator usually outperforms both.