Editor’s Note: As additive manufacturing techniques evolve, these processes offer significant advantages and benefits to the electronics industry. Microwave Journal addresses aspects of this emerging area with a two-part article. This first part, published in the September Supplement of Microwave Journal, introduced dielectric measurement concepts and important challenges, along with a portion of a panel discussion with RF dielectric measurement experts from the industry. This second part continues the discussion with the panel of experts.

Discussing Dielectric Measurements With The Experts (Continued)

What methods are you currently using to determine the dielectric constant and/or the loss tangent of dielectric materials?

Marzena Olszewska-Placha, Ph.D., Vice President for Research & Development, QWED Sp. z o.o.:

Most common methods we use for determining complex permittivity are those which we also offer to our customers and which are known for their high accuracy. Those are split-post dielectric resonators (operating at single frequencies between 1 and 15 GHz) and Fabry-Perot open resonators (wideband single-device solution for a frequency band of 20 to 125 GHz) dedicated to laminar and other sheet dielectrics. Those two types of fixtures are excellent for dielectrics of Dk between 1 and 15 and loss tangent going down to 2 10^-5, with achievable accuracy of less than 0.5 percent and less than 2 percent, respectively. However, we also have and work on more specific designations, like ceramic resonators with losses going down to even 10^-7, low loss liquids and nonstandard materials like bituminous mixtures, foods, etc.

Jonathan Chisum, Ph.D., Associate Professor for Electrical Engineering, University of Notre Dame:

We ordinarily perform full-wave simulation of candidate unit cells because it is fast and accurate for complex unit cell designs. Simulations are validated using transmit/reflect measurements and the Nicholson-Ross-Weir (NRW) method in waveguides and free space.

Nico Garcia, Ph.D., CEO & Founder, Cheshir Industries:

If we are using a commercial material such as a Rogers laminate or resin for the first time, we will do an NRW measurement using a loaded rectangular waveguide. We have not had any issues with Rogers’ materials, but we characterize every new material we use as a matter of due diligence. We have also done free-space measurements. This is a great way to measure the dielectric performance of a stack of materials and accurately characterize a stepped impedance transform in free space, which you cannot do in a rectangular waveguide due to the non-TEM propagation.

John Schultz, Ph.D., Chief Scientist, Compass Technologies Group LLC:

We use and develop a variety of dielectric test measurement technology. Free-space focus-beam, both microwave and mmWave, is one such. Another is an Epsilometer that we developed alongside Copper Mountain Technologies. What is unique about the Epsilometer is that the dielectric extraction is actually performed from an interpretation of the response of the sample in the Epsilometer referenced from a database of computational electromagnetic (CEM) codes using a CEM inversion method. This is as opposed to traditional parameter extraction methods which typically rely on physics-based approximations. Other examples of CEM inversion dielectric measurements we have developed include a slotted stripline waveguide for in-situ measurements of continuous sheet materials and square coaxial transmission lines that can be loaded with a cubic dielectric sample.

What challenges are you having with these methods?

Marzena Olszewska-Placha

As for the challenges, material science and industry are continuously expanding towards new materials. Therefore, if they appear, the challenges typically come from our customers, either associated with using our fixtures for their new applications or while providing measurement services for our client’s materials. These also stimulate further development on our side.

Jonathan Chisum

The downsides to these methods differ.

Waveguide:

  • The method is limited to less than an octave of bandwidth.
  • It is difficult to achieve a very accurate estimate of loss tangent in low loss materials. We have found that if great care is taken to maintain phase stability, including phase-stable cables, throughout a TRL waveguide calibration and the sample is manually placed in the waveguide cross-section several times and average values are computed. The method can achieve highly accurate results for both permittivity and loss tangents.1 The TE10 waveguide mode is not the same as the operating mode. This is acceptable so long as the unit cell under investigation is much smaller than the waveguide cross-section, in which case, the field will be nearly uniform across the cell.

Free space:

  • This is a wideband method limited only by the antennas.
  • Spot-focus antennas are used. It is difficult to achieve a small spot size and therefore the material under test must present a large cross-section, typically several inches in diameter, to ensure the majority of the energy interacts with the material and the surrounding medium. This can be costly to fabricate to test structures.