I. INTRODUCTION
Gallium nitride (GaN) hybrids and monolithic microwave integrated circuits (MMICs) find various applications for energy-efficient high-power microwave amplification [Reference Mishra, Shen, Kazior and Wu1, Reference Quay2]. Hybrid powerbar transistors achieve very high power-added efficiency (PAE) values of >60% for large powerbars with output power levels of >100 W in the frequency range from 1–4 GHz, e.g. [Reference Mitani, Haematsu, Yokogawa, Junichiro and Yasunori3, Reference Palmour4]. HPA MMIC results with output power levels of 50 W at X-band have been demonstrated [Reference Piotrowicz5]. A significant number of outstanding MMIC results are not fully visible [Reference Mishra, Shen, Kazior and Wu1]. Further developments, especially, in the US and Japan, stretch the application frequencies to the mm-wave up to 110 GHz [Reference Micovic6]. This paper gives an overview on the results obtained on packaged powerbar and on MMIC level.
II. EVOLUTION OF DEVICE TECHNOLOGY
The critical step toward the realization of state-of-the art of III-N hybrids and MMIC amplifiers are achievements with respect to the optimization of the active device technologies. With the strong thermal constraints based on the high-power densities achieved with GaN [Reference Mishra, Shen, Kazior and Wu1], the improvement of PAE and gain became much more important than the simple increase of the power densities. This is true for all frequencies, however, for the particular gate length, for the upper ranges of the targeted frequencies of operation. For powerbars based on gate lengths in the range of 0.5 µm, this critical area is the frequency range of 4–6 GHz. MMIC-related frequencies in this paper range from 6 to 30 GHz. They are realized using gate lengths of 0.25 and 0.15 µm which are appropriate for this frequency range. As a reference, Fig. 1 gives the overview of loadpull data measured at 2, 10 and 28 GHz in CW-operation achieved at Fraunhofer in order to visualize the status in device technology for gate lengths of 0.5, 0.25 and 0.15 µm on MMIC level. We also see that the operation bias for the particular technology reduces with increasing frequency. The breakdown voltages taken at an operation temperature of 150°C are typically a factor of three higher for the technologies given. The results in Fig. 1 are given for gate peripheries which are relevant for large gate width power bars and MMICs.
Device CW-power and PAE for MMIC applications for three gate lengths over frequency.
III. POWERBAR TECHNOLOGY (1–6 GHz)
Hybrid powerbars are particularly attractive for the frequency range between 0.9 and 6 GHz. The key issues are the increase of efficiency at package level, gain improvement, oscillation control for unmatched and prematched devices, and ruggedness for large VSWR ratios. These factors are of course interdependent. As an example for the results of optimization, Fig. 2 gives the CW-power results of a GaN power bar in microstrip transmission line technology measured at 2 GHz and V DS = 50 V. We observe a linear gain of more than 20.7 dB at 2 GHz, a maximum power of 50 W, and an associated power density of 4.2 W/mm at an efficiency level of 60%. The microstrip transmission line technology ensures good linear gain for the device in package. Figure 3 gives the overview of the PAE and DE of various powerbar designs in package as a function of output power achieved for the frequencies 2 and 4 GHz when operated in CW between V DS = 40 and 50 V. The drain efficiencies are also given in order to highlight the impact of the gain on PAE at the two frequencies.
Device CW-power, gain, DE, and PAE at 2 GHz for a 12 mm powerbar in microstrip transmission line technology, measured in package.
PAE and DE in CW over output power for packaged power bars measured at 2 and 4 GHz.
The PAE and DE levels at 2 GHz are in the range of >65% for devices up to 50 W and roll-off slightly to the 120 W level. At 4 GHz, the typical efficiency level is in the range of 50% up to the power range of 50 W. Further work on the packaging and layout side shows that the roll of PAE versus power roll of can be reduced to the level of the smaller devices.
IV. EVOLUTION OF MMIC LIBARIES AND MODELING
A) Passive development and library extraction
For the MMIC design library-specific III-N passive components require development in order to allow the same design flow and flexibility as in the GaAs MMIC world. Both coplanar and microstrip transmission line components have been made available, including all technology-specific elements like metal–insulator–metal (MIM) capacitors, resistors, and inductances. The issues in GaN libraries include the realization of high-power transmission lines, MIM capacitors with high-breakdown voltage per area, high-power resistors, and area-efficient high-current inductors. As the MMIC processes are maturing, spread analysis techniques and advanced passive components gain importance. Specific advanced passive components such as MIM-on-top-of via and individual source-vias require more specific development. Modeling of the passive components is mostly based on conventional ADS models, which are conventionally extracted and used.
V. EVOLUTION OF CIRCUIT DESIGN FOR GaN MMIC POWER AMPLIFIERS
A) Large-signal modeling
The models used for this work are based on an in-house two-dimensional voltage-lag model after an extensive phase of development. Thermal effects and low-frequency dispersion, and their impact on gain as well as PAE are well described. Following the general theory presented in [Reference Seelmann-Eggebert, Merkle, van Raay, Quay and Schlechtweg7], long-term memory effects are described by internal states. The parameters of this model are extracted from pulsed-DC, pulsed and CW-small-signal S-parameters, and time domain measurements including the harmonics over wide bandwidth. This model allows a state-dependent description of the intrinsic drain current and the gate-source and gate-drain space charges. The state quantities are the low-pass filtered voltages V GS and V DS. This approach facilitates a correct description of retarded responses as well as the instantaneous responses relevant for RF power performance.
B) Discussion of specific design issues
In several ways, GaN design differs from the MMIC design performed for GaAs and silicon. GaN offers very high intrinsic levels suitable for broadband matching to 50 Ω loads. Typical power cells for X-band design operated in the range of 30–40 V offer real impedance very close to 50 Ω. At the same time transformation networks lead to strong changes for these theoretical impedances at the reference planes. The complicated compression of the power gain due to dispersion and thermal effects over input power are major challenges for the design of amplifiers with high PAE at low gain compression. The introduction of field plates, formerly only used in silicon LDMOS technology at a few GHz into GaN technology up to at least 30 GHz, has led to the delicate balance of improving power compression and reliability while reducing the effective parasitic capacitances to a minimum for the individual process and frequency range.
C) GaN X-band (8–12 GHz) MMICs
The realization of X-band MMICs with high power and high bandwidth has been one of the major drivers for the development of GaN technology [Reference Moon8]. Bandwidth, center frequency, and PAE are key for the development of MMICs [Reference van Raay9, Reference Kühn10, Reference van Raay12]. The ultimate goal for such frequencies is to convert more energy into microwaves than into heat over a reasonable bandwidth for real multi-stage MMICs operated at a bias of up to 40 V. The realization of dual-stage designs is further key due to the gain requirements of TRX-chains. Figure 4 gives the evolution of linear gain, PAE, and output power of dual-stage X-band MMICs over time as developed at Fraunhofer IAF over the last five years. The values are consistently given for long-pulse operation (100 µs) for a duty cycle of 10%, all measured between 9 and 10 GHz. The minimum bandwidth for each design is 2 GHz. Figure 4 suggests a continuous development on MMIC level over time with significant improvements, while the realization of higher PAE and gain has found more attention than the simple increase of output power. This is caused by lack of primary power in most systems. In Fig. 4 we see a very consistent increase of the linear gain level up to >27 dB at the center frequency, which needs to be carefully balanced versus bandwidth and the maximum PAE at the upper band edge at X-band frequencies, e.g. >11 GHz.
Evolution of PAE, output power, and linear gain at Fraunhofer IAF over time for dual-stage MMICs with a bandwidth of >2 GHz for a duty cycle of 10%, all with center frequency between 9 and 10 GHz.
VI. BROADBAND POWER AMPLIFIERS 1–18 GHz
GaN is particularly attractive to the realization of broadband amplifiers. GaN is probably the first semiconductor technology, which can provide broadband operation at similar power levels and in frequency ranges as traveling-wave tube amplifiers and replace tubes at least for the lower power ranges. For very principal reasons, solid-state power amplifiers will always be less efficient than tubes, providing broadband matching losses, while they yield advantages such as compactness, size, omission of high-voltage supplies, and overall relevant lifetime. Thus, again PAE over bandwidth is a key for the application.
A) HPAs for the frequency range 1–6 Hz
Broadband power amplifiers in the frequency range 1–6 GHz are attractive for multi-band communication amplifiers, e.g. for base stations as well as for military radio and electronic warfare applications. Both cost and PAE on device and module level are key and typically lead to hybrid approaches. At the same time, the loss mechanisms have to be identified. For example, to obtain good PAE at the upper frequency range of the C-band, advanced MMIC approaches have been used to provide reference designs.
Figure 5 gives the pulsed output power measurement of a single stage MMIC at 6 GHz with an output periphery of 6.4 mm for a gate length of 0.25 µm. The MMIC delivers more than 40% PAE in a broadband matching situation with a linear gain of 10 dB and an output power of more than 45.5 dBm or 35 W when operated at V DS = 40 V. This translates into a power density of 5.5 W/mm under pulsed operation. The MMIC situation allows very precise matching and to consider the phase differences for the individual power transistors at 6 GHz. This shows the enormous power potential of GaN broadband designs. Hybrid designs based on gate lengths of 0.25 or 0.5 µm with their packaging may be more cost effective, however, should provide similar PAE values as given in Fig. 5.
Output power results of a single-stage MMIC at 6 GHz measured at V DS = 40 V in pulsed operation.
B) Broadband power amplifiers 6–18 GHz
For the higher frequency band, e.g., 6–18 GHz a broadband solution using a MMIC approach is typical due to the stringent needs to control the phase accurately with very low losses. Figure 6 gives the micrograph of a nonlinear amplifier realized already in 2004 based on a gate length of 150 nm and III-N cascodes FET. The MMIC provides a linear gain of >9 dB between 2 and 20 GHz, which makes such devices very suitable for driver stages in broadband amplifier chains. It provides output power levels of about 1 W in this case. A similar MMIC design was recently published by Triquint [Reference Campbell11] at the power level of 17 W.
Early European CPW-broadband MMIC with a gate length of 0.15 µm.
VII. ROBUST LNA MMIC
Next to power amplifier robust low-noise amplifiers have been realized as MMICs with the particular target to achieve very high survivability at reasonable noise figure level [Reference Suijker13]. Typical targets for the survivability are in the 5–10 W range of input power to a particular MMIC. Figure 7 provides an overview of typical noise figures NF50 and gain per MMIC stage, which are measured in low-noise operation for the frequency range of 1–30 GHz. This comparison is done on MMIC level with particular emphasis of reaching the robustness mentioned. We see a consistent noise level of less than 2 dB is obtained up to 14 GHz of operation with a reasonable associated gain per stage (on MMIC level) for low-noise operation. A third gate length of 0.15 µm is added leading to a shrinked process of GaN15. The higher frequency range up to 30 GHz can be addressed. The MMICs given in Fig. 7 all support a CW-input power level beyond 40 dBm (10 W), which is even increased in pulsed operation. This work shows the enormous potential to realize robust LNAs with good noise figure up to Ka-band frequencies.
MMIC LNA results at frequencies between 1 and 30 GHz.
VIII. CONCLUSIONS
The design, realization, and characterization of highly efficient powerbars and MMIC HPAs with AlGaN/GaN high electronic mobility transistors (HEMTs) has been presented for the frequency range between 1 and 30 GHz. Powerbars for the frequency range between 0.9 and 6 GHz have been developed based on a process GAN50 which deliver PAE values beyond 60% for 50 V operation with a linear gain of >20 dB for the 50 W class at 2 GHz. Based on a GaN25 MMIC process several high-power MMIC amplifiers with power levels >30 W are presented starting from 6 GHz up to advanced X-band amplifiers in microstrip transmission line technology. Low-noise amplifiers provide noise levels below 2 dB up to 14 GHz of operation with a robustness of >10 W input power.
ACKNOWLEDGEMENTS
The authors acknowledge the continuing support of the Federal Ministry of Defense (BMVg), the support of the Federal Ministry of Education and Research (BMBF), the Federal Office of Defense Technology and Procurement (BWB), and the Bundeswehr Technical Center for Information Technology and Electronics (WTD 81). Continuous collaboration with United Monolithic Semiconductors (UMS) is highly appreciated.
Jutta Kühn was born in Freiburg, Germany, in 1978. She received the Dipl.-Ing. degree in electrical engineering from the University of Karlsruhe, Germany, in 2005. Since 2006, she has been with the Fraunhofer Institute for Applied Solid State Physics (IAF) in Freiburg, Germany, where she is working toward her doctoral degree. Her research interests include the design of AlGaN/GaN HEMT power amplifiers with optimized power-added efficiency. Jutta Kühn is a recipient of the Amelia Earhart Fellowship 2008 awarded by Zonta International.
Markus Mußer was born in Villingen, Germany on April 24, 1974. He received the diploma in electrical engineering in 2007 from the university of applied sciences Nuremberg, Germany. In 2007 he joined the RF Devices and Circuits Department of the Fraunhofer Institute of Applied Solid-State Physics in Freiburg, Germany, where he is currently working on GaN RF power bars and RF high power ampplifiers.
Friedbert van Raay was born in Hilden, Germany, in 1960. He received an M.Sc. degree in electrical engineering from Technical University of Aachen, Germany, in 1984, and a Ph.D. degree from the University of Kassel in 1990. From 1992 to 1995, he was with the SICAN GmbH, Hannover, Germany, working on RF system development and measurement techniques. In 1995, he returned to the University of Kassel as a senior engineer. He supervised the Microwave Group within the Institute of High Frequency Engineering and worked on design of ultra-broadband GaAs MMICs, device modeling, and large-signal measurement techniques. In November 2001, he joined the Fraunhofer Institute of Applied Solid-State Physics in Freiburg, Germany. His research interests include MMIC design, device modeling, and measurement techniques.
Matthias Seelmann-Eggebert received his diploma and a Ph.D. degree in physics from the University of Tübingen in 1980 and 1986, respectively. From 1980 to 1996 he was involved in R&D related to infrared detectors based on HgCdTe and developed electrochemical and surface physical methods for the characterization of compound semiconductor surfaces. From 1990 to 1991 he was a visiting scientist at Stanford University. From 1997 to 2000 he was engaged in the growth of CVD diamond. From 2001 he is a member of the Department of High Frequency Electronics of the IAF in Freiburg, Germany, and is concerned with the preparation and development of simulation models for active and passive III–V devices.
Rudolf Kiefer received his Dipl. Phys. degree in physics from the University of Freiburg, in 1979. He joined the Fraunhofer Institute of Applied Solid State Physics (IAF) in 1980 starting work on liquid crystal displays. For work on this topic he received a Ph.D. degree in physics in 1984 from the University of Freiburg. From 1984 to 1991 he investigated electro-optic effects in ferroelectric, liquid crystalline polymeric materials for optical storage applications. Afterwards he focused his work on studying a new switching effect of nematic liquid crystals to improve the viewing angle characteristics of thin film addressed liquid crystal displays (TFT-LCD). For this work, he received a SID Award in 1998. In 1994 he moved to the field of III–V optoelectronic semiconductor technology. He managed different technology projects developing processes for fabrication and laser facet deposition of InP-based high-speed laser and GaAs as well as GaSb-based high-power laser. From 2000 his work is focused on the development of GaN semiconductor technology for MMICs.
Michael Mikulla received his diploma degree in electrical engineering from the Technical University of Braunschweig, Germany, in 1989 and a Ph.D. degree from the University of Ulm, Germany, in 1994. From 1994 he has been with Fraunhofer IAF in Freiburg, Germany, working on both high-power semiconductor lasers and GaN-based high-power transistors and MMICs. He has authored and has published over 100 conference/journal papers. From 2002 he is the head of the III/V-Technology department of the institute and is also in charge of the GaN RF Power Electronics business unit.
Rüdiger Quay received the Diplom degree in physics from Rheinisch-Westfälische Technische Hochschule (RWTH) Aachen, Germany, in 1997, and a second Diplom in economics in 2003. He received his doctoral degree in technical sciences (with honors) from the Technische Universität Wien, Vienna, Austria. In 2009 he received the venia legendi (habilitation) in microelectronics, again from the Technische Universität Wien. He is currently a research engineer with the Fraunhofer Institute of Applied Solid-State Physics, Freiburg, Germany, heading the RF-devices and characterization group. He has authored and coauthored over 100 refereed publications and three monographs. He is a member of IEEE, MTT, and chairman of MTT-6.
Thomas Rödle was born in Friedrichshafen, Germany on 26 January 1967. He received his Ph.D. from the University of Göttingen (Germany) in 1996, working on inelastic light scattering on AlAs/GaAs-quantum wells. After having worked for the semiconductor branch of Siemens on cost reduction programs for front-end and back-end operations, he joined Philips Semiconductors in the Netherlands (now NXP Semiconductors) in 1997. He started his career at Philips in process and device development for radio frequency power amplifiers used in base stations. Thomas is currently acting as a project manager for wide bandgap semiconductor technologies at NXP Semiconductors.
Oliver Ambacher received his Dipl.-Phys. and Dr. degrees with honors from the Ludwig-Maximilians and Technical University Munich, in 1989 and 1993, respectively, where he was involved in the deposition and characterization of amorphous silicon for solar cells. In 1992 he received a German Science Foundation Graduate Research Fellowship. In 1993, he joined the Walter Schottky Institute of the TU-Munich to investigate the epitaxial growth of group-III nitrides-based heterostructures. Since 1995 the research of his group is focused on fabrication of GaN-based devices like UV detectors, surface acoustic wave devices or microwave amplifiers as well as on the understanding of polarization induced effects in group-III nitride heterostructures and quantum wells. During 1998/1999, he spent one year at Cornell University, Ithaca, NY, as an Alexander von Humboldt fellow, where he was involved in the optimization of polarization induced AlGaN/GaN HEMTs for high-frequency and high-power applications. He became a Professor of Nanotechnology and head of the Institute for Solid-State Electronics located at the Technical University of Ilmenau in 2002. In 2004 he was elected as head of the new Center of Micro- and Nanotechnologies. Since 2007 he is the head of the Fraunhofer Institute of Applied Solid State Physics and Professor for Compound Microsystems in Freiburg, Germany.
