Ce-Jun Wei

Direct extraction of equivalent circuit parameters for heterojunction bipolar transistors

A new method is presented for the direct extraction of hybrid-T equivalent circuits for heterojunction bipolar transistors. The method differs from previous ones by extracting the equivalent circuit without using test structures or numerical optimization techniques. Instead, all equivalent circuit parameters are calculated analytically from small-signal S-parameters measured under different bias conditions. The analysis includes the distributed nature of the HBT base. The calculated parameters are essentially frequency-independent and they exhibit systematic bias dependence over the typical operating range of the transistor. Thus, the present method ensures unique and physically meaningful parameters for transistor design improvement and large-signal circuit simulation. In addition, the present method is much faster than the numerical optimization method. >

An accurate large-signal model of GaAs MESFET which accounts for charge conservation, dispersion, and self-heating

A comprehensive large-signal model of a GaAs MESFET is presented to account for charge conservation and effects of dispersion and self-heating. An improved set of capacitance and charge equations, along with an enhanced Triquint Own model (TOM)-like drain current model, is used for consistent small- and large-signal simulations. Charge conservation is satisfied by deriving the capacitance part of the model from charge equations. Transconductance and output conductance dispersion is modeled by combination of a feedback network and a subcircuit, which describes the self-heating effect. An improved description of the near-pinchoff characteristics, high-voltage breakdown, and gain compression in the high-current region in the TOM is introduced. The new model accurately predicts the I-V, CV, bias-dependent S-parameter, waveform, power, and linearity characteristics of the MESFET.

Waveform-based modeling and characterization of microwave power heterojunction bipolar transistors

A waveform measurement technique has been successfully used to extract the large-signal nonlinear characteristics of microwave power heterojunction bipolar transistors. The extracted model parameters were compared to those extracted from dc and small-signal parameters in a conventional manner. It was found that, for high input drive conditions, the present model predicts a much longer collector transit time than the conventional model. Therefore, the present model is more consistent with the physical structure of the transistors and more suitable for evaluating future design improvement.

Large-signal modeling of self-heating, collector transit-time, and RF-breakdown effects in power HBTs

A large-signal heterojunction bipolar transistor (HBT) model has been developed which includes self-heating, collector transit-time, and RF-breakdown effects. The model has a compact form which is based on a compromise between accuracy and utility. As such, the model can be readily extracted and verified with the aid of RF waveform measurements. Using the model in simulations, it was found that RF breakdown was dependent on base biasing and loading conditions. Therefore, with proper circuit design, the maximum output power of the HBT can significantly exceed the limit of open-base breakdown voltage.

Hot-electron-induced degradation of pseudomorphic high-electron mobility transistors

Pseudomorphic high-electron mobility transistors have been found to undergo hot-electron-induced degradation. Due to the negative temperature dependence of hot-electron effects, it will be necessary to conduct electrical and temperature stress tests separately, in order to ascertain the reliability of these transistors under normal operating conditions.<<ETX>>

Non-invasive waveform probing for nonlinear network analysis

A noninvasive technique has been developed for the measurement of fundamental and harmonic S-parameters as functions of input frequency and power. This technique was demonstrated on a GaAs FET by plotting its drain current versus voltage trajectory at 5 GHz. By superimposing the RF trajectory on the DC characteristics, the origin of the device nonlinearity was clearly shown. The results make it possible to characterize the nonlinear behavior and to verify the large-signal model of the device in both the time and the frequency domains.<<ETX>>

Internal-node waveform analysis of MMIC power amplifiers

A novel internal-node waveform probing technique has been demonstrated on a C-band monolithic microwave integrated circuit (MMIC) power amplifier. The error of the measurement and its perturbance to circuit operation was estimated and verified to be within /spl plusmn/10%. Valuable insight was obtained from the variation of waveforms as a function of frequency, drive and location. The potential impact of this technique includes MMIC design verification, in-situ device model extraction, process diagnosis, and reliability assessment.

Hot-electron-induced degradation of metal-semiconductor field-effect transistors

Hot-electron trapping in the SiN passivation was found to be a cause for gradual degradation during RF operation of metal-semiconductor field-effect transistors. The time dependence and threshold energy for trap formation was determined by dc and electroluminescence tests. The spatial distribution of trapped electrons was directly observed by a novel high-voltage electron-beam-induced-current imaging technique. Argument was also made for trapping in the SiN instead of at the GaAs/SiN interface.

Temperature-dependent large signal model of heterojunction bipolar transistors

A large signal model of the heterojunction bipolar transistor has been developed which simulates forward active, saturation, and cutoff operation, with both transient and steady-state thermal effects included. This model is useful for simulating devise characteristics with different heatsinking configurations, allowing the use of on-wafer measurements to predict packaged device performance. All transport parameters have been extracted from bias-dependent S-parameter measurements, while parameters describing the thermal effects have been extracted from DC and pulsed I-V characterization. Excellent agreement have been found between the model and DC, pulsed, and RF measurements.<<ETX>>

Table-based dynamic FET model assembled from small-signal models

A data-table-based large-signal metal-semiconductor field-effect transistor model is presented based on an ensemble of bias-dependent small-signal equivalent circuits. The model is capable of accurate simulation of small-signal S-parameters as well as large-signal performance over the data-acquisition bias range. In addition to the dc current sources, the model contains two quasi-static charge sources and a dynamic radio frequency (RF) current source, which depends on the temperature rise. By introducing the dynamic RF current source, the problem of path dependence, which occurs in modeling large-size devices and devices with dispersion, is resolved. All equivalent elements are obtained by cubic spline interpolation. Extrapolation of the model beyond the measurement range is taken into account. The model is extracted by an in-house software without involving optimization. The validity of the model is demonstrated by comparing the simulation of small-signal S-parameters over a wide bias range, as well as power, linearity, and waveform characteristics to the measured data.