VNA Reflections

Approaches to Conducting Material Measurements of High-frequency Circuits Using a VNA

October 30, 2020

Most high-frequency electronic circuits in today’s designs are built on dielectric materials and optimal operation depends on their dielectric properties. In general, these materials can be categorized into three segments – insulators, semiconductors, and conductors.

Figure 1: The three categories of materials.

The materials, as shown in figure 1, can be further described as:

• Insulators (dielectrics), in which the energy gap between the valence and conduction band is huge. The result is no free electrons are available.
• Semiconductors have a smaller energy band gap, so some electron movement from the valence band to the conduction band can occur.
• Conductors allow electrons to move freely between the valence and conduction bands because they overlap.

Engineers who design these circuits must gain a vital understanding of the properties of the dielectric materials. It is especially true with respect to the dielectric constant and loss tangent at the operating conditions. In today’s post, we will discuss how to best gain this understanding using a vector network analyzer (VNA).

Material Measurement Methods and Techniques

Dielectric measurement is becoming an increasingly important tool to understand material behavior, because of the dramatic growth in high-frequency designs. This type of measurement can provide the electrical or magnetic characteristics of the materials, which is a critical parameter when using them in emerging applications such as those operating in the millimeter wave (mmWave) frequencies. A number of methods have been developed to measure the complex permittivity/permeability of materials in the time/frequency domain based on reflection and reflection-transmission measurements.

For the dielectric measurement, engineers cannot rely on a single technique to characterize all the materials over a wide range of frequencies. Resonant and non-resonant methods are two prominent approaches.

Resonant Methods

Highly accurate and precise, resonant approaches generally include the resonator and the perturbation method. Five main families of resonant techniques - micro strip; cavity resonator; dielectric resonators; open resonators; and split post resonators (SPDR) – comprise the resonant method. Two measurement techniques fall under this category:

• Resonator method is founded on the fact that the resonant frequency and quality factor of a dielectric resonator with given dimensions are determined by its permittivity and permeability. This method is often used to measure low loss dielectrics whose permeability is μ0.

• Perturbation method is centered on the resonant perturbation theory. For a resonator with given electromagnetic boundaries, when part of the electromagnetic boundary condition is changed by introducing a sample, its resonant frequency and quality factor will also be altered. From the changes of the resonant frequency and Q, the properties of the sample can be derived. This method is suitable for lower and moderate loss samples.

Non-Resonant Methods

In non-resonant methods, the material properties are derived by the impedance and the wave velocities in the material. When a wave travels from one medium to another, both its impedance and velocity change. Useful information can be deduced from the reflections and reflection transmission of the wave in the medium. Both permittivity and permeability can be calculated using this procedure.

Non-resonant methods are mainly comprised of two processes:

• Reflection method directs electromagnetic waves to a sample under study and the properties of the material sample are deduced from the reflection coefficient. Usually a reflection method measures either permittivity or permeability. Two types of reflections are often used in materials – property characterization and open-reflection. As coaxial lines can cover broad frequency bands, coaxial lines are often used in developing measurement fixtures for reflection methods.

• Transmission/reflection method inserts a piece of transmission line/waveguide into the material under test (MUT). The material properties are deduced on the basis of the reflection from the material and the transmission through the MUT. This method is widely used in the measurement of the permittivity and permeability of low conductivity materials.
  
  The principle behind transmission/reflection method is that the characteristic impedance of the piece of transmission line loaded with the sample is different from that of the transmission line without the sample. Therefore, such difference results in special transmission and reflection properties at the interfaces. The permittivity and permeability of the sample are derived from the reflection and transmission coefficients of the sample-loaded cell.

Figure 2: Transmission/Reflection test system configuration using an Anritsu ShockLine™ VNA.

Free Space Material Measurement Technique

Free space material measurements are used for characterizing materials that remain in an unchanged state. There are many advantages associated with the free space material measurement technique, since the MUT can be placed under a variety of conditions (i.e. temperature, pressure, and tensile variations) and the response can be monitored using the same setup for a broad range of frequencies. It must be noted that the MUT must be large and flat for this technique to be used.

A test station for free space material measurements usually utilizes two antennas. One is connected to port 1 and other to port 2 of a VNA. They are aligned in the test setup so they face each other. Focus beam antennas are typically used, though simple horn antennas can also be integrated into the system in some test environments.  

VNAs are typically calibrated before making this type of measurement. The most common calibration techniques are the through-reflect-line (TRL), through-reflect-match (T/LRM), and the line-reflect-line (LRL). The LRL calibration method, having the highest calibration quality, is often chosen.

Figure 3 highlights the importance of time domain gating when this technique is used. It ensures that the multiple reflections from the DUT and edges of the antennas are not used by the VNA in the final measurement calculations. The dielectric properties are then determined by post-processing the measured reflection and transmission data using a program. Focal/lens antennas in the free space material measurement further enhance system accuracy, as they take care of the diffraction from the edges of the samples.

Figure 3: Display highlighting the significance of time domain gating.

Anritsu has published an application note, Material Measurements with Vector Network Analyzers. The downloadable document provides more details on testing this aspect of high-speed designs.

How to Select a VNA Architecture for Efficient and Accurate Characterization of mmWave Devices

July 21, 2020

There are several key applications pioneering the use of millimeter wave (mmWave) communications technology, ranging from 5G New Radio (5G NR) and automotive radar to tactical communications and unmanned aerial vehicle (UAV) imaging and communications. Device characterization testing on devices operating at these higher frequencies poses challenges for design engineers in need of generating accurate models for circuit simulation.  

To aid these engineers in gaining design confidence, advanced vector network analyzer (VNA) technology (figure 1) has been developed that provides a superior method than traditional concatenated waveguide bands. These new architectures utilize broadband vector network analyzer (VNA) and mmWave connector technology to enable broadband testing to 220 GHz with a single test unit and setup for improved device modeling.

Figure 1

Component Testing Importance

testing of amplifiers, mixers, isolators/circulators, filters, attenuators, transmission lines (interconnect), transformers, and other mmWave components is necessary to refine the design and manufacturing process, pass certification testing, and ensure production quality. It must be conducted in a manner to ensure the highest accuracy over the frequency range of the device under test (DUT).

Design engineers rely on characterization testing to determine device performance parameters and create device models for electronic computer-aided design (ECAD) systems during the development process. How well the characterization testing is performed can often influence how many refining iterations are needed during development. It impacts the length of the design cycle and time-to-market.

Traditional Test Methods

Until recently, there have been two main methods of performing testing for wafer and semiconductor die probing – concatenated waveguide band approach and broadband measurement techniques. The concatenated waveguide band approach requires test instruments with frequency-extending hardware that uses waveguide interconnects that typically exhibit much less loss than coaxial options.

As waveguides are intrinsically banded interconnects, several frequency extender units – and possibly several test instruments – are needed to conduct tests to hundreds of gigahertz. One method is to use a VNA that operates to 110 GHz with a 1-mm coaxial interconnect, then integrate a frequency extender with D-band waveguide, followed by another frequency extender using WR5 (140 GHz to 220 GHz) waveguide.

Another factor with this approach is that back-end-of-line (BEOL) parasitic circuit elements and probe interconnect resistances and capacitances can adversely affect measurement accuracy. Figure 2 shows how it can impact on-wafer measurements certainty and device model accuracy.

Figure 2

Calibration must also be performed for each setup. Due to the wear on the landing pads, calibration structures, and probes, repeatability of each test is diminished and uncertainty increases. Different calibration standards may be used, increasing the challenge of successfully de-embedding the test setup. Conversely, with a true broadband measurement system (DC to the maximum frequency), very accurate and consistent de-embedding can be performed to the device level and with fewer touchdowns.

NLTL-based VNAs

Anritsu has developed a VNA architecture that overcomes the inherent limitations of VNA systems based on down conversion technologies, such as step-recovery diode (SRD), in high-frequency designs. The Anritsu non-linear transmission line (NLTL) sampling VNA technology is coupled with monolithic reflectometer designs and components to enhance broadband testing.

NLTL-based broadband VNAs (figure 3), such as the VectorStar™ family, provide signal quality, reduce cost, enable extreme bandwidth, mitigate isolation calibration issues, and enable more compact VNAs that can be placed closer to DUTs. NLTL-based VNA architectures rely on continuous wave signals that then undergo fall-time compression via non-uniform NLTLs to create narrow gating pulses for sampling receivers. Distributed harmonic generation is enabled by using the “harmonic growth” characteristics of NLTLs.

Figure 3

The result is extremely high RF sampler bandwidth with a single sampler covering a much broader frequency range with lower noise floor than an SRD-based sampler or many classical harmonic mixers. Utilization of a high LO range also minimizes noise introduction and image response issues.

NLTL VNA technology also results in a coupling structure with optimum raw directivity. This optimized raw directivity aids with the modeling process by improving measurement stability and lengthening the time between calibrations. Other broadband VNA technologies exhibit dramatic negative raw directivity, making them much more susceptible to environmental effects. Reduced calibration and measurement stability is the result.

By utilizing NLTL technology, the VectorStar VNAs benefit from RF and LO frequency scalability and high channel-to-channel isolation without suffering degradation from poor isolation experienced by other VNAs. Additionally, the monolithic design and compact nature of NLTL-based VNAs reduce the temperature variations between the reflectometers and constituent components.

Engineers reap benefits from this VNA architecture when characterizing their devices used in mmWave designs. The VNAs have enhanced calibration integrity, measurement accuracy, and measurement repeatability. Because of the reduction in size of the mmWave NLTL module, on-wafer probes may attach directly to the module, thereby eliminating the need for interconnect cables. The broadband VectorStar VNA thereby minimizes interconnection losses due to long and lossy transmission lines associated with typical mmWave testing environments.

You can learn much more about VNA technologies best suited for mmWave device characterization by downloading a new application note entitled New VNA Technologies Enable Millimeter-Wave Broadband  Testing to 220 GHz.