Abstract-Dynamic temperature distributions in GaAs HBT are numerically analyzed in frequency domain as a function of power dissipation, frequency and space. Complete thermal characteristics, including frequency-dependent thermal impedance and phase lag behavior, are presented. The analysis is also extended for arbitrary periodic or aperiodic pulse heating operation to predict junction temperature of a Power Amplifier (PA) with non-constant envelope input signal. Dynamic junction temperatures of a single finger 2 µm×20 µm GaAs HBT are predicted for square pulse envelope signal input with power levels varying with up to 10 dB above a nominal average level of 40 mW and with pulse widths ranging from 10 ns to 100 µs. With the input envelope signal amplitude of 10 dB above the average, the analytical results show that junction temperature rises from room temperature of 27 • C to 39 • C when heated by 10 ns pulse and increases to 63 • C by 100 ns pulse, 105 • C by 1 µs pulse and to 198 • C by 100 µs pulse. A novel setup is developed for nano-second pulsed measurements, and the analysis is validated through time domain on wafer pulsed measurements at three different power levels: 0 dB, 3 dB, and 6 dB above the average level. Results show that analytical results track well with measured junction temperature within the accuracy of ±5 • C over the entire measurement set.

Abstract-Traditionally, the transmitter (TX) IQ imbalances distortion and power amplifier (PA) distortion are separately modeled. In this paper, the behavior of the two distortions are unified, and characterized by a single model. Rectangular structured Focused Time-Delay Neural Network (RSFTDNN) is proposed to uniformly model IQ imbalances and PA distortions. As a result, the physical distortions in the analog circuits are further abstracted. It also saves computation resources in simulation. Unlike the polynomial based model, which suffers from condition number effects and inaccuracy for deeply nonlinear system, the proposed RSFTDNN shows high accuracy. Two cases of real experiments are carried out, where RSFTDNN model shows much better performance than the polynomial based model in the sense of model accuracy.

A new sub-nanosecond high power impulse generator is presented in this paper for ultra-wideband communication and radar systems using InGaP/GaAs HBT technology. The topology is simple, compact and can be readily integrated into MMIC. It works on two principles: a faster transition can be obtained from a digital logic input into the HBT and a differentiation of the very fast transition current using the inductor to generate a large voltage at the output. Measurement results show the generated Gaussian pulse has a tunable peak voltage between 1 V to 7 V with full-width-halfmaximum (FWHM) between 200 ps to 250 ps. The impulse generator is biased at 3 V battery supply and consumes 27 mW to generate a 7 V high 250 ps wide Gaussian impulse at a 1 MHz pulse repetition rate (PRF).Index Terms -heterojunctions bipolar transistors, UWB impulse generator, high power impulse circuits, battery supply.

Abstract-A unilateral circuit model, which precisely predicts small signal response over a wide range of frequencies and bias points, is quantitatively analyzed and presented. The shortfall of current unilateral assumption and transformation technique is presented. A complete and explicit analysis is provided to develop a compact unilateral circuit model. The model is intended to predict input reflection, forward transmission and output reflection coefficients over wide range of frequencies. The technique is validated by transforming bilateral a small signal model of 3 × 3 µm × 40 µm, InGaP/GaAs HBT into its unilateral equivalent over the frequency range of 250 MHz to 30 GHz. The accuracy of the technique is corroborated at various bias conditions; collector current from 3 mA to 150 mA and collectoremitter voltage from 1 V to 5 V. Simulated results show very good agreement between small signal responses of transformed unilateral and bilateral circuit models.

No abstract