Dale E. Dawson

CW measurement of HBT thermal resistance

Measurements of the temperature dependence of beta and V/sub BE/ were made on AlGaAs-GaAs HBTs and used to determine device thermal resistance. The measurements were CW and not switched or pulsed in order to have a simpler procedure. With base doping greater than 10/sup 19/ cm/sup -3/, HBTs have negligible base-width modulation (i.e., flat I/sub C/ versus V/sub CE/ characteristics) which makes CW thermal resistance measurement especially direct and simple. >

A Closed-Form Expression for Representing the Distributed Nature of the Spiral Inductor

A closed form expression for the modeling of rectangular spiral inductors to twice the self-resonant frequency is derived and compared to experimental results. The mutual inductance effects of the ground plane, and phase shift effects are considered in this analysis.

Stability and improved circuit modeling considerations for high power MMIC amplifiers

The cluster-matching approach for large-periphery power FETs brings with it certain problems including more complex circuitry and additional modes of possible oscillations. Solutions to these problems are offered including an analysis of the modes along with methods of suppression and improved circuit modeling. These solutions were implemented in the design of a two-stage, 1.6-W monolithic power amplifier which is also discussed.<<ETX>>

Closed-Form Solutions for the Design of Optimum Matching Networks

An analytic solution of optimum matching networks up through order n = 4 is derived. Prior solutions use iterative methods. Examples are given that show direct solution of the matching network elements, and the examples also include insertion of an ideal transformer. Comments on n = 5 and 6 networks are also included.

A dual-varactor analog phase shifter operating at 6 to 18 GHz

A monolithic microwave integrated circuit (MMIC) analog reflection phase shifter is described that achieves 120 degrees of phase shift from 6 to 18 GHz using a dual-varactor reflection circuit. This circuit allows varactors with a 3:1 capacitance ratio to achieve the performance that normally requires 10:1 ratio diodes. The varactor diode used is a surface-oriented structure with a hyperabrupt doping profile selectively ion implanted to a depth of 0.70 mu m. Measured phase shift was typically 0 to 150+or-20 degrees with 2.7 dB typical insertion loss (4.0 dB worst case) for one coupler section over 6 to 18 GHz. >

A 1.8-W, 6-18-GHz HBT MMIC power amplifier with 10-dB gain and 37% peak power-added efficiency

A two-stage 6-18-GHz high-efficiency AlGaAs-GaAs HBT MMIC power amplifier has been designed and tested. At 7-V collector bias, this fully matched monolithic amplifier delivered an output power of 1.8+or-0.6 W over the band. The peak output power was 2.45 W at 13 GHz with an associated gain of 11.1 dB and 36.7% power-added efficiency (PAE). Amplifiers from five different wafers showed similar performance.<<ETX>>

A 2 watt, 8-14 GHz HBT power MMIC with 20 dB gain and >40% power-added efficiency

A two-stage, 8-14 GHz high-efficiency AlGaAs/GaAs HBT MMIC power amplifier has been designed and tested. At 7 V collector bias, this common-emitter monolithic amplifier has achieved 20 dB gain, 33 dBm (CW) output power, and >40% power-added efficiency over the 8-14 GHz band. The amplifier is designed for 25 /spl Omega/ input and output impedance, and all the matching networks, as well as biasing circuits, are contained within this HBT MMIC. To our knowledge, this is the highest efficiency, the highest gain, and the highest output power reported for any monolithic power amplifier covering a 6 GHz bandwidth in the X-Ku band. >

Thermal modeling, measurements and design considerations of GaAs microwave devices

This paper addresses the thermal management issues associated with fabricating microwave power devices. Thermal management of microwave devices includes: modeling thermal resistance with techniques such as Greens function/method of images, conformal map, finite element, spectral domain, and Fourier series; measurement techniques such as improved IR, self heating, laser probing, and direct thermocouple measurement; and layout and fabrication alternatives such as silicon substrates, shunt metal paths, and selective thinning that allow smaller form factors while maintaining or reducing thermal resistance. Examples of several device layout and fabrication approaches (MESFET, HBT, PHEMT) show how high frequency performance and low thermal resistance can be part of the unit cell design.

A 5-10 GHz, 1-Watt HBT amplifier with 58% peak power-added efficiency

Four 0.25-W GaAs Heterojunction Bipolar Transistors (HBTs) were combined in a single-stage hybrid microstrip amplifier. An output power of minimum 1 Watt (W) was achieved over the 5.5-9.5 GHz band with >48% power-added efficiency (PAE). The peak PAE was 58% at 7 and 9.5 GHz with an average efficiency of 52% over the 5-10 GHz band. This result was reproduced on two more units with a minimum efficiency of 48% and an average efficiency of 51%. To our knowledge, this is the highest efficiency obtained from any 1-W amplifier covering 5-10 GHz bandwidth. >

A 1-Watt, 8-14-GHz HBT amplifier with >45% peak power-added efficiency

Four 0.25-W gallium arsenide heterojunction bipolar transistors (HBTs) were combined in a single-stage hybrid microstrip amplifier. An output power of >1 W was achieved over the 8.5-13.5-GHz band with >35% power-added efficiency (PAE). The peak PAE was 45.4% at 12.5 GHz. This result was repeated on a second unit that was subsequently tuned for improved performance at the upper end of the band. The PAE at 14 GHz increased to >43% with 1-W output while, at 8 GHz, it remained at approximately 30%.<<ETX>>