Multibias and temperature dependence of the current‐gain peak in GaN HEMT

Abstract

Thermal and multibias behavior of the peak in the short‐circuit current‐gain (h21) has been investigated for a GaN HEMT, aiming to contribute to an extensive knowledge on it. To obtain a simple and complete insight of this phenomenon and its influence in device performance over operating conditions, high‐frequency multibias scattering (S‐) parameter measurements have been analyzed from low to high temperature. It has been observed that the current‐gain peak might get to be more or less serious depending on the working circumstances. The peak affecting h21 has been successfully reproduced by using an equivalent‐circuit model. Moreover, a novel procedure has been developed to interpret this kind of phenomenon by quantifying the area of the current‐gain peak (ACGP), which is denoted as the area corresponding to h21 curves with and without the peak. It is found that the ACGP is strongly dependent on bias and less dependent on temperature. The relevance of a comprehensive evaluation of the peak in h21 lies in its usefulness for empowering RF engineers to efficiently consider it for both device modeling and circuit design.

Scalability of Multifinger HEMT Performance

“…The CGP appears to be less pronounced in larger devices, as can be mathematically predicted by using the definition of the damping factor (ζ ) in terms of the equivalent-circuit elements [21]. To quantify the size of the CGP, two methods have been proposed in prior works, which are based on calculating the second order derivative of h21 in dB vs the frequency [22] or on estimating the area between the two h21 curves with and without peak [23]. Hence, depending on the given application, a proper selection of W can enable achievement of a peak in the current-gain at the frequency of interest.…”

Experimental insight into the temperature effects on DC and microwave characteristics for a GaAs pHEMT in multilayer 3‐D MMIC technology

“…It is worth mentioning that the data at the higher frequencies in Figure 14 seem to show that, by going beyond the upper frequency limit of the experiments, h 21 of the tested pHEMT will deviate from the ideal behavior and will be affected by the current-gain peak (CGP), arising from the resonance between intrinsic capacitances and extrinsic inductances. 60,61 Furthermore, Figure 15 reports the frequency dependence of the measured maximum available/stable gain (MAG/MSG) under the four different temperature conditions. It is highlighted in the inset of Figure 15 that the maximum oscillation frequency (f max ) extracted from the measured small-signal power gain is lowered by heating the transistor.…”

An Experimental and Systematic Insight into the Temperature Sensitivity for a 0.15-µm Gate-Length HEMT Based on the GaN Technology

“…Throughout the years, many studies have been dedicated to the investigation of how the temperature impacts the performance of GaN-based HEMT devices. To this end, both electro-thermal simulations [ 1 , 2 , 3 , 4 , 5 , 6 ] and measurement-based analysis [ 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 ] have been developed. Although the electro-thermal device simulation is undoubtedly a very powerful and costless tool to deeply understand the underlying physics behind the operation of the transistor in order to improve the device fabrication, the measurement-based investigation is a step of crucial importance for achieving a reliable validation of a transistor technology prior to its use in real applications.…”