What it really measures
The instrument measures the magnitude and phase of the impedance between its terminals. The sample is placed in a sample holder connected to those terminals. From the equivalent model of the holder, normally a capacitance with a parallel conductance, the real permittivity and the losses are obtained, or the inductance in the case of a magnetic material in a wound core.
Internal principle
Many instruments use the auto-balancing bridge technique: the current through the sample is passed through an amplifier with a feedback resistor, which holds a point at zero potential. The voltage across that resistor is proportional to the current, and it is compared with the applied voltage by a phase-sensitive detector. Other architectures, such as the RF method with reflection measurement, are used at higher frequencies.
What determines accuracy
- Impedance range: accuracy is best in an intermediate zone and worsens at very high or very low impedances. A low-capacitance dielectric at low frequency presents a very high impedance.
- Calibration and compensation: open circuit and short circuit, to eliminate parasitics from the sample holder and the cables.
- Measurement plane: it must be at the sample holder, not at the instrument.
- Signal level: in nonlinear materials or those with an amplitude-dependent response, the level influences the result.
What the engineer sees
As frequency increases, errors appear from cable inductance and sample holder resonances, which show up as an apparent jump in capacitance. If this happens, the measurement no longer corresponds to the material.
How to interpret the result
Verify the result with a reference component or a known sample before trusting the value for the material. Record the frequency, signal amplitude and equivalent model used, because the instrument can present the same data in series or in parallel with different loss values.