A Physics-Based Frequency Dispersion Model of GaN MESFETs

Abstract

Public reporting burden for this colleclion of information Is estimated to average 1 hour per response, induding the time for reviewing mstniction^ searching existng data sources, fl?«^«""af "f '"«"^""^ date nelde" and completing and reviewing this collection of infomiation. Send comments regarding this burden esUmate or any o*er aspect of m,scollect,on<rf »l°;;"^" ''"^" REPORT DATE (DD-MM-YYYY) SPONSOR/MONITOR'S ACRONYM(S)APRL/SNHA SPONSOR/MONITOR'S REPORT NUMBER(S) DISTRIBUTION / AVAILABILITY STATEMENTApproved for Public Release; Distribution Unlimited. SUPPLEMENTARY NOTES ABSTRACTA physics based model for GaN MESFETs is developed to determine the frequency dispersion of output resistance and transconductance due to traps. The equivalent circuit parameters are obtained by considering the physical mechanisms for current collapse and the associated trap dynamics. Detrapping time extracted from drain-lag measurements are 1.55 seconds and 58.42 seconds indicating trap levels at 0.69eV and 0.79eV, respectively. The dispersion frequency is in the range of MHz at elevated temperature, where a typical GaN power device may operate, although at room temperature it may be few Hz. For a 1.5 nm. 150 nm GaN MESFET with drain and gate biases of lOV and -IV respectively, 5% decrease in transconductance and 62% decrease in output resistance at RF frequencies from their DC values are observed. The dispersion characteristics are found to be bias dependent. A significant decrease in transconductance is observed when the device operates in the region where defrapping is significant. As gate bias approaches toward cutoff, the difference between output resistance at DC and that at RF increases. For drain and gate biases of lOV and -5V, output resistance decreases from 60.2 k. at DC to 7.5 k. at RF for a 1.5 urn. 150 nm GaN MESFET. This article was published in

Drain current transient and low-frequency dispersion characterizations in AlGaN/GaN HEMTs

“…The Y 22 real part (Real( Y 22 )) corresponding, in a first approximation, to the output conductance g d (or the reciprocal of output resistance) shows one inflexion point. So g d ( f ) dispersion presents one trap that corresponds to trap “E2” found by g m ( f ) dispersion according to [18]. Moreover, for the process “E2” we found the same time constants by g d ( f ) and g m ( f ) dispersions.…”

Characterising thermal resistances and capacitances of GaN high‐electron‐mobility transistors through dynamic electrothermal measurements

“…As a result, as the signal frequency increases, the transconductance of the transistors decreases. When the period of the signal is sufficiently smaller than the trapping time constants, the drain current cannot recover at all and the transconductance reaches its minimum value and is invariant to further frequency increase [10,13]. In this work, a fixed bias condition (class-B) is used.…”

Dynamic Transconductance Dispersion Characterization of Channel Hot-Carrier Stressed 0.25- $\mu\hbox{m}$ AlGaN/GaN HEMTs

“…Conductance techniques are often used to electrically evaluate trap characteristics in large area MOS capacitors [5]- [7], however in HEMTs these become inaccurate in the case of short channel devices due to the weak capacitance signal offered by the small gate area. The use of transconductance (g m ) frequency dispersion to characterize traps has been proposed in several papers for both GaN and GaAs MESFETs [8], [9]. However, all the previous analyses focused on the real part of the measured g m , not exploiting its imaginary part.…”