VNA Reflections

Modeling Microwave Devices Using a Broadband VNA

October 31, 2012

Predicting the performance of nonlinear devices under varying conditions is far more challenging than forecasting how their linear counterparts will perform. Inaccurate predictions often result in multiple turns during the design cycle, slowing time to market and increasing development costs. Vector Network Analyzers (VNAs) are traditionally used to improve nonlinear model accuracy, however, with many designs now reaching microwave and millimeter wave frequencies, modeling is a challenge that cannot be met by all VNAs.

One example is when Modelithics developed novel non-linear diode models for W Band Single Anode, and W Band ZBD flip ship Schottky diodes from Virginia Diodes, Inc. (VDI). Renowned for developing models that reliably track how component performance will change with various input parameters over a specified frequency range, Modelithics required a VNA with wide frequency
coverage to help it develop models with features such as temperature and substrate scaling, and can validate performance over varied frequency, bias and RF input power.

Modelithics used the Anritsu VectorStar™, as well as on-board probing with calibration reference at the component pad-stacks, to conduct two-port S-parameter measurements and validate the
model’s performance from DC to 125 GHz.  The S-parameter measurements, along with precision DC I-V and capacitance, were used to extract the models. With these models, designers can have improved simulation accuracy and more rapid design success using the VDI diodes.

Learn more about Modelithics, Anritsu and VDI.

Signal Integrity: Using a VNA to Find the Ideal Emphasis Setting

October 23, 2012

As transmission speeds increase, there are greater demands for product size reductions, lower power consumption and reduced cost. Maintaining signal integrity (SI) in this environment is difficult, to say the least. Emphasis technology is one method to correct frequency-dependent transmission loss – a key detriment to SI.

Emphasis setting parameters become more complex as transmission rates become faster. It can be extremely difficult to find the ideal settings to maintain SI. One method is to search for the ideal settings while verifying the output waveform, but it is extremely time-consuming and hard to come up with a rational explanation of why those settings are ideal. Finding the ideal emphasis setting automatically can simplify and shorten the emphasis setting procedure.

A way to achieve this automatic setting is to add the inverse characteristics of the DUT frequency characteristics measured using a vector network analyzer (VNA) to the input signal as emphasis settings. In other words, letting the DUT transfer function be H(f) and the ideal emphasis transfer function is:

G( f ) =H^-1 ( f )

 This allows engineers to obtain the ideal eye opening. Accordingly, ideal emphasis settings can be calculated.

By integrating Anritsu’s VectorStar™MS4640A VNA, with a BERT, such as the MP1800A, MP1825B 4 Tap Emphasis, and software, it’s possible to make S-parameter measurements and use a simulation package, such as AWR’s  associated Microwave Office, to predict the non-equalized eye pattern. (Note: VectorStar has the broadest coverage in the industry – from 70 kHz to 125 GHz. To acquire the best S-parameter data, the VNA should have broad frequency coverage. Learn why here.)  It is then possible to use the measured S-parameter data to program the equalizer with optimum tap settings to obtain an open eye at the receive end. The result is an automatic procedure that helps cut measurement and design verification times.

To learn more about this method, feel free to read a new white paper, which can be found here.