Device modeling with NVNAs and X-parameters

Abstract

This paper reviews and contrasts two complementary device modeling approaches based on data readily obtainable from a nonlinear vector network analyzer (NVNA) [1]. The first approach extends the application of waveform data to improve the characterization, parameter extraction, and validation methodologies for "compact" transistor models. NVNA data is used to train artificial neural network -based constitutive relations depending on multiple coupled dynamic variables, including temperature and trap states for an advanced compact model suitable for GaAs and GaN transistors. The second approach is based on loaddependent X-parameters* [2], [3], [5], [6], measured using an output tuner working with the NVNA. It is demonstrated that Xparameters measured versus load at the fundamental frequency predict well the independent effects of harmonic load tuning on a 10W GaN packaged transistor without having to independently control harmonic loads during characterization. A comparison of the respective merits of the two approaches is presented.

Extraction and Validation of Power Amplifiers' Harmonics Dynamic <formula formulatype="inline"><tex Notation="TeX">$X$</tex></formula>-Parameters From Pulsed Envelope Domain Measurements

“…Subsequently, (8) is rewritten by changing the order of integration and the series summation as (10) For the sake of simplicity, a multivariate function is introduced in (10). The function is a combination of the impacts from all the related variables, and it is defined as (11) Substituting the dynamic part by (11) yields (12) The obtained result in (12) is identical to (2). Therefore, the derivation of harmonics dynamic -parameters model is obtained using a similar approach mentioned in [12]; hence the behavioral model is valid for the response prediction of harmonic.…”

“…Generally speaking, the basic principle of the -parameters model is to calculate the sum of the small-signal and the largesignal responses [11], where the large-signal refers to the fundamental amplitude on the input port and the small-signals (perturbations) represent the harmonic components on all the other ports. In other words, the -parameters represent a linear approximation in the frequency domain of a nonlinear response, which is due to the small variations around a large-signal operating condition.…”

A novel quadratic polynomial expansion of nonlinear scattering function for RF power amplifier modeling

“…Recently, high‐frequency nonlinear measurement systems have gained increasing market with the development of large signal network analyzers, like the Keysight nonlinear vector network analyzer (NVNA) and Rohde&Schwarz ZVA . Such instruments are typically capable of measuring both the amplitude and the phase of all significant harmonics of both input and output traveling waves of the active device under nonlinear operation.…”

“…[5][6][7] Recently, high-frequency nonlinear measurement systems have gained increasing market with the development of large signal network analyzers, like the Keysight nonlinear vector network analyzer (NVNA) and Rohde&Schwarz ZVA. [8][9][10]…”

Microwave FET model identification based on vector intermodulation measurements

“…Lately, look-up table (LUT) models [1], [2] have been very successful for their accurate predictions and easy extraction techniques, mainly based on gathering data from measurements. Unfortunately, to cover a wide range of operating conditions, they need a huge amount of data achieved with a large number of measurements.…”