G-band ( GHz) and W-band ( GHz) InP DHBT Power Amplifiers

Transcription

1 G-band (1--GHz) and W-band (7--GHz) InP DHBT Amplifiers Vamsi K. Paidi, Zach Griffith, Yun Wei, Mattias Dahlstrom, Miguel Urteaga, Navin Parthasarathy, Munkyo eo, Lorene amoska, Andy Fung, Mark J. W. Rodwell, fellow, IEEE Abstract We report common-base power amplifiers designed for G-band (1--GHz) and W-band (7-- GHz) in InP mesa double heterojunction bipolar transistor (DHBT) technology. The common-base topology is preferred over common-emitter and common-collector topologies due to its superior high frequency maximum stable gain (MG). Base feed inductance and collector emitter overlap capacitance, however, reduce the common-base MG. A single-sided collector contact reduces C ce and hence, improves the MG. A single-stage common-base tuned amplifier exhibited 7-dB small-signal gain at 17 GHz. This amplifier demonstrated.7 dbm output power with -db associated power gain at 17 GHz. A two-stage common-base amplifier exhibited.1 dbm output power with.3-db associated power gain at 17 GHz and demonstrated 9.1 dbm saturated output power. Another two-stage common-base amplifier exhibited. dbm output power with an associated power gain of.-db at 1 GHz. In W-band, different designs of single-stage common-base power amplifiers demonstrated saturated output power of 1.1 dbm at GHz and 13.7 dbm at 93 GHz. Index Terms InP heterojunction bipolar transistor, millimeter-wave amplifier, MMIC amplifiers. I. INTRODUCTION -band (7--GHz) and G-Band (1--GHz) W amplifiers have applications in wide-band communication systems, atmospheric sensing and Manuscript received April,. This work is supported in part by ONR under N and Caltech s President s Fund. The research described in this paper was carried out in part by the Jet Propulsion Laboratory, Caltech under a contract with the National Aeronautics and pace Administration. Vamsi K. Paidi, Zach Griffith, Yun Wei, Mattias Dahlstrom, Navin Parthasarathy, Munkyo eo, Mark J. W. Rodwell are University of California at anta Barbara, CA 931, UA.( phone-93-; fax: -93-7; paidi@ ece.ucsb.edu). Miguel Urteaga is with Rockwell cientific Corporation, Thousand Oaks, CA 913, UA. Lorene amoska, Andy Fung are with California Institute of technology Jet Propulsion Laboratories, Pasadena, CA 99, UA. automotive radar. The high mobility of InGaAs, high electron saturation velocity of InP and submicron scaling result in wide-bandwidth transistors with high available gain in this frequency band. In a transferred substrate InP HBT process,.3 db gain is reported at 17 GHz with a single stage amplifier [1]. tate-of-the-art results in InP HEMT technologies include a three stage amplifier with 3-dB gain at 1 GHz [], a three stage amplifier with 1-1-dB gain from 1-19 GHz [3], and a three-stage power amplifier with 1-dB gain from 1-17 GHz []. Recent work in scaled InP/InGaAs/InP mesa DHBT with 3 nm Carbon-doped InGaAs base with graded base doping and 1 nm of total depleted collector thickness achieved wide-bandwidth transistors with 37 GHz f t and 9 GHz f max [][]. In this paper, we describe how this technology is used to realize several power amplifiers in the 7--GHz frequency range. II. INP DHBT PROCE The transistors in the circuit are formed from a Molecular Beam Epitaxial (MBE) layer structure with a highly doped 3 nm InGaAs base and a 1 nm collector and are fabricated in a triple mesa process with both active junctions defined by selective wet etch chemistry. The increased collector thickness over [][] is intended to maintain high f max despite increases in device critical dimensions, motivated by the desire for improved transistor yield. Polyimide passivates and planarises the devices. One level of deposited metal forms circuit interconnects and electrical contacts to transistors and resistors. in metal-insulator-metal (MIM) capacitors and coplanar waveguide transmission lines are employed to synthesize the tuning elements. Plated airbridges bridge the ground planes and suppress the coplanar waveguide slot-line modes.

2 III. AMPLIFIER DEIGN A. Models and imulations The transistor PICE model parameters used in the simulations are extracted from the measured two-port - parameters. The amplifiers are simulated using Agilent Technologies Advanced Design ystem software. A planar method-of-moments EM simulator (Momentum) modeled the coplanar waveguide structures and the MIM capacitors. B. Transistor Characteristics The DC common-base characteristics of a two finger.7µm µm InP common-base power DHBT is shown in Fig. 1. The common-base breakdown voltage (V br,cb ) is more than 7 V (Fig.). The feasible (I c, V ce ) loadline is however, constrained by the device safe operating area (Fig. 1), as determined by both breakdown voltage and thermal resistance. The devices have shown GHz f t and 9 GHz f max when biased at 3mA/µm current density and 1.7V V ce (Fig. 3). The degradation in f max relative to [] is due to a wider base mesa intended to improve yield and due to relatively poor base ohmic contacts in this process run. I c, ma 3 1 afe operating area - V CB, V Fig. 1. Common-base DC characteristics of a two finger.7µm µm common-base DHBT for I b =., 1, 1., and. ma. I c, ma V CB, V Fig.. Common-base DC characteristics of a two finger.7µm µm common-base DHBT for I b =, 1, 1, and µa. 3 U U, h 1 db 1 1 h f τ = GHz f max =9 GHz Fig. 3. hort circuit current gain and Mason s gain as a function of frequency of a single finger.7µm 7µm common-emitter DHBT. C. Transistor Layout Parasitics The common-base topology is chosen as it has higher Maximum table in this band when compared to the common-emitter and common-collector topologies (Fig. ). At present, however, we ignore both base feed inductance, L b and collector-emitter overlap capacitance, C ce. At 1 GHz, the common-base topology exhibits 1- db MG while the common-emitter and common-collector topologies exhibit db and 3 db respectively. Because power amplifiers use large signal load match, rather than a small-signal output match, the gain falls below the MG.

3 3 U base plug MG/MAG, db 1 1 Common emitter Common base base L b Common Collector 1 1 Fig.. Comparison of MG/MAG of common-base, common emitter and common collector configuration of an InP DHBT. L b and C ce are omitted. The above comparison between different configurations ignores the effect of L b and C ce. While these parasitics reduce the common-base MG, in G-band, the commonbase topology still provides the highest gain when compared to the common-emitter and common-collector configurations. If not modeled in the designs, L b and C ce could potentially cause instability. Base inductance is due to the long thin base contact metal stripes on either side of the emitter (Fig. (a)). Loop inductance depends upon the current return path; this is difficult to identify in the transistor geometry, hence L b is not readily modeled with accuracy. This creates uncertainty in the stability analysis. -parameter extractions indicate approximately 3 ph base feed inductance per 1 µm long emitter finger having. µm base contact width. The collector to emitter overlap capacitance (C ce ) also reduces MG. C ce is the capacitance between the emitter interconnect metal and the collector ohmic contact metal (Figs. (a), (b)). These metals are separated by ~- nm polyimide. This thickness varies in our process, rendering C ce variable. Degradation in MG/MAG of a common-base topology due to layout parasitics L b and C ce of an InP DHBT with double sided collector contacts is shown in Fig.. Potential instability in the small-signal characteristics due to L b and C ce was observed in the first-generation amplifiers fabricated. In second-generation designs, the collector to emitter overlap capacitance was significantly reduced by employing single-sided collector contacts as opposed to double-sided collector contacts (Fig. 7). In addition to reducing C ce, this also increases the collector resistance and thus, further improves circuit stability. NiCr resistors provide additional resistive stabilization in some designs. C ce emitter interconnect metal N- collector base N+ subcollector semi-insulating InP C ce collector Fig. a. Cross-section and top view of an InP mesa DHBT with double-sided collector contacts. L b emitter base interconnect metal N- collector N+ subcollector semi-insulating InP base plug base collector Fig. b. Cross-section and top view of an InP mesa DHBT with single-sided collector contacts.

4 MG/MAG Fig.. Comparison of common-base MG/MAG with and without layout parasitics. The InP DHBT has a double-sided collector contact. MG/MAG ingle sided collector Double sided collector Without C ce, L b With C ce 1 1 With C ce, L b formed by cascading two identical single-stage designs. The output of the first stage is large signal matched to the second stage input, avoiding first stage premature power gain compression. IV. MEAUREMENT A. mall-signal Measurements G-band amplifiers are measured on wafer using an HP 1C Vector Network Analyser with Oleson Microwave Labs Millimeter Wave VNA extensions. The test-set extensions are connected to GGB Industries coplanar wafer probes via WR- waveguides. The amplifier measurements are calibrated using off-wafer TRL calibration standards. W-band amplifier small-signal gains and return losses were measured on-wafer using using a W-band Agilent 1 Network Analyzer calibrated with an off-wafer calibration using TRL calibration standards. V eb, bias V in Ohms First tage Input Matching network Output Loadline Matching network f f amplifier designs 1 1 econd tage V cb, bias Fig. 7. Comparison of MG/MAG of common-base HBT with single-sided collector contact and double-sided collector contacts. The design values of the power amplifier gains are also shown. Input Matching network Ohms Output Loadline Matching network V out R L Input Matching Network Output Loadline Match Fig. 9. Two-stage common base amplifier. V in V out R L B GHz Measurements BWO ource Variable attenuator W-band amplifier Probe loss 17-1 GHz band ~. db WR- WR- Picoprobe Picoprobe chottky Calorimeter Doubler DUT Fig.. single-stage common base amplifier. D. Circuit Design Fig. shows a single-stage amplifier circuit schematic. hunt capacitors are either in x MIM capacitors or CPW open-circuit stubs. Two-stage amplifiers (Fig. 9) are W-band meter Fig GHz power measurement setup.

5 G-band power measurements were performed at JPL. The 17-1-GHz power measurement setup is shown in Fig. 1. W-band power from a BWO power source is amplified and is doubled in frequency using a chottky diode frequency doubler. The frequency doubler output drives the input of the device under test (DUT). The DUT output power is measured using a calorimeter. Because the input and output power are measured at separate times, the saturated power gain measurements are subject to approximately ±1-dB drift in gain. The saturated output power measurement is not subject to this drift, and we estimate the output power data is accurate to ±.-db. Data is corrected for measured probe attenuation. C. 1-1 GHz Measurements The 1-1 GHz measurement setup is shown in Fig.. A 1 GHz Gunn oscillator drives DUT. A variable attenuator adjusts the input power. 1 GHz Gunn ource Variable attenuator Probe loss 1 GHz band ~ 3. db WR- Wafer Probe DUT WR- Wafer Probe Fig GHz power measurement setup. Calorimeter D. 7--GHz Measurements The 7--GHz power measurement setup is shown in Fig. 1. The output of a DC--GHz frequency synthesizer is amplified and tripled in frequency to 7--GHz. This signal is further amplified to drive the DUT input. The DUT output power is measured using a W-band power sensor. DC-GHz ource DC-GHz Amplifier Frequency tripler W-band power amplifier WR-1 Picoprobe DUT Fig GHz power measurement setup. V. REULT WR-1 Picoprobe W-band meter Probe loss 7- GHz band ~ 1.1 db Fig. 13. Die photograph of the single stage common base MMIC amplifier centered at 17 GHz. This measures.3 mm.3 mm. This power amplifier is designed to obtain 1. dbm saturated output power at 1 GHz, when biased at I c = 3 ma and V cb =.V. The output power vs. input power characteristic is shown in fig. 1. The amplifier exhibited a saturated output power of.77 dbm with an associated power gain of -db at 17 GHz when biased at I c = ma and V cb =. V. This power amplifier demonstrated > dbm saturated output power between GHz (Fig. 1). The circuit exhibited 7.9 db uncompressed gain under the above conditions at 17 GHz. Measured -parameter data exhibits potential instablility in the 1-17-GHz range due to feedback parasitics L b and C ce. 1,,, db frequency, GHz Fig. 1. Measured -parameters of the 17 GHz single-stage amplifier. A. 17 GHz ingl-tage Amplifier A die photograph is shown in Fig. 13. The transistor has two separate. µm 1 µm fingers. The amplifier bandwidth is limited by the output tuning network. The transistor output is large-signal load-line matched for maximum saturated output power as opposed to a smallsignal match for maximum gain. This amplifier exhibited 7-dB small signal gain at 17 GHz when biased at I c = 3 ma and V cb = 1.V. (Fig. 1)

6 , db Output, dbm 1 - Output (%) Fig. 1. Output power, Added Efficiency () vs. input power of the 17 GHz single-stage amplifier at 17 GHz. Maximum Output, dbm 1 Maximum output power Associated power gain Fig GHz amplifier saturated output power as a function of frequemcy. B. 1 GHz single-stage amplifier A second single-stage common-base amplifier (Fig. 17) exhibited.-db small signal gain at 1 GHz (Fig. 1) when biased at I c =31 ma and V c =1.V. This amplifier s small signal gain is >3-dB between 1-1-GHz. The transistor has two separate. µm 1 µm fingers. Fig. 17. Die photograph of a 1 GHz amplifier. 1,,, db Fig. 1. -parameters of a 1 GHz single-stage amplifier. This power amplifier exhibited.3 dbm saturated output power with.-db associated power gain at 17 GHz (Fig. 19) when biased at I c =7 ma and V c =.1V., db, Output, dbm Output Fig. 19. Output power, vs. input power of the 1 GHz single-stage amplifier. C. 17 GHz two- stage amplifier A die photograph is shown in Fig.. This amplifer is a cascaded version of two individual amplifiers designed for Ώ input resistance and Ώ load. Each stage employs two separate. µm 1 µm HBT fingers. The smallsignal measurements are performed with the first stage biased at I c = ma and V cb =1. V and the second stage biased at I c =3 ma and V cb =1. V. mall-signal measurements indicate 7-dB gain at 17 GHz and 13-dB gain at 1 GHz. There is a potential instability in in 1-1 GHz range (Fig. 1). This amplifier exhibited.1 dbm output power with.3 db associated power gain at 17 GHz and demonstrated 9.1 dbm saturated output power (Fig. ). These measurements are performed with the first stage is biased 3 1 (%)

7 at I c = ma and V cb =. V and the second stage biased at I c =9 ma and V cb =1.V. At 1. GHz, the power amplifier exhibited 1.3 dbm output power with 3.-dB associated power gain (Fig. 3). The first stage is then, biased at I c = ma and V cb =. V and the second stage is biased at I c = 1 ma and V cb =.V. Uncompressed gain at 1. GHz is 9.-dB., db, Output, dbm 1 Output (%) Fig.. Die photograph of a 17 GHz two-stage MMIC amplifier. This measures 1mm X.7 mm. 1 1,, db Fig. 1. mall-signal measurements of the 17 GHz two stage amplifier. 1 Output 3, db, Output, dbm (%) Fig. 3. measurements of the 17 GHz two-stage amplifier at 1. GHz. D. 1 GHz two-stage amplifier A second two-stage amplifier (Fig. ) exhibited 1-dB gain at 1 GHz with the first stage is biased at I c =3 ma and V cb =1. V and the second stage biased at I c = ma and V cb =1. V (Fig. ). This two-stage amplifier demonstrated dbm output power at 1. GHz with. db associated power gain (Fig. ). At 1 GHz, a. dbm saturated output power is obtained with an associated power gain of.-db. These power measurements are performed with the first stage biased at I c =3 ma and V cb =. V, and the second stage biased at I c =9 ma and V cb =.7 V. Fig.. Die photograph of a 1 GHz two-stage MMIC amplifier. Fig.. measurements of the 17 GHz two-stage amplifier.

8 ,,, db - 1 1,,, db Fig.. mall-signal measurements of the 1 GHz twostage amplifier 1, db, Output, dbm Output Fig.. measurements of the 1 GHz two-stage amplifier. E. GHz single-stage amplifier A single-stage amplifier (Fig. 7) exhibited. db small signal gain at GHz (Fig. ) when biased at I c =37 ma and V cb =1. V. The transistor has four separate. µm 1 µm fingers. This circuit demonstrated 1.1 dbm saturated output power at GHz with >-db associated power gain (Fig. 9). (%) Fig.. mall-signal measurements of the GHz amplifier., db, Output, dbm P out 1 1 Fig. 9. measurements of a GHz amplifier. F. 9 GHz single-stage amplifier A second common base amplifier (Fig. 3) exhibited - db small signal gain at 9 GHz when biased at I c =39 ma and V cb =1. V (Fig. 31). This amplifier demonstrated 13.7 dbm saturated output power at 93 GHz (Fig. 3) when biased at I c = ma and V cb =. V. 1 1 (%) Fig. 7. Die photograph of a GHz single stage amplifier. Fig. 3. Die photograph of a 9 GHz amplifier.

9 , db, Output, dbm 1,,, db Fig. 31. mall-signal measurements of 9 GHz amplifier Output Fig. 3. measurements of the 9 GHz amplifier at 93 GHz. VI. CONCLUION Common-base high gain G-band and W-band power amplifiers in InP mesa DHBT technology are presented. A single-stage common-base tuned amplifier with 7-dB small-signal gain at 17 GHz exhibited.7 dbm output power with -db associated power gain at 17 GHz. The common base topology provides the largest maximum stable gain. This configuration requires careful layout to minimize C ce and L b, or the maximum stable gain will be reduced. Despite large signal loadline matching, the design values of gain remain high at 1 GHz, and close to the MG. levels, efficiency and center frequency are below design values, an effect we attribute to modeling errors. Recent DHBTs have been reported at 9 GHz f max [], suggesting feasibility of power amplifiers at GHz. Increasing the number of HBT fingers should result in power amplifiers with output power more than 1 mw. The results presented here demonstrate the potential of InP DHBT technology for high performance ultra-highfrequency millimeter-wave circuit applications. ACKNOWLEDGEMENT This work is supported in part by ONR under N and Caltech s President s Fund. The research described in this paper was carried out in part by the Jet Propulsion Laboratory, Caltech under a contract with the National Aeronautics and pace Administration. REFERENCE [1] M. Urteaga, et al., G-Band(1--GHz) InP-Based HBT Amplifiers, IEEE Journal of olid tate Circuits, Vol. 3, No. 9, pp. -1, ept. 3. [] C. Pobanz, et al., A high-gain monolithic D-band InP HEMT amplifier, IEEE Journal of olid tate Circuits, Vol. 3, pp. 9-1, ept [3] R. Lai, et al., InP HEMT amplifier development for G- band (1- GHz) applications, in Int. Electron Devices Meeting Tech. Dig., an Francisco, CA, Dec., pp [] L. amoska, et al., "A mw, 1 GHz InP HEMT MMIC Amplifier Module," IEEE Microwave and Wireless Components Letters, Vol 1, No., pp. -, Feb.. [] Z. Griffith, et al., InGaAs/InP mesa DHBTs with simultaneously high f t and f max and low C cb /I c ratio, Electron Device Letters, submitted for publication. [] M. Dahlstrom, et al., InGaAs/InP DHBTs with > 37 GHz f τ and f max using a Graded Carbon-Doped base, 3 IEEE Device Research Conference, alt Lake City, UT, June 3-, 3, Late news submission. Vamsi K. Paidi received his B. Tech degree in electrical engineering from the Indian Institute of Technology (IIT), Madras, India, in, and is currently working toward the Ph. D degree at the University of Calfornia at anta Barbara (UCB). His research includes design and fabrication of high frequency power amplifiers for wireless applications using GaN HEMTs and InP DHBTs. Zach Griffith received his B. and M. degrees in electrical engineering from the University of California, anta Barbara, in 1999 and 1, respectively, where he is pursuing the Ph. D degree. His primary research includes design and fabrication of InP high speed digital ICs. Yun Wei received his B.. degree from the Fudan University, hanghai, China, the M.. degree from the Oregon Graduate Institute of Technology, Portland, and the Ph. D. degree from the University of California at anta Barbara in 1991, 199, and, respectively, all in electrical and computer engineering. From 1991 to 1997, he was a Researcher with the Institute of emiconductors, Chinese Academy of ciences, Beijing. His research was on high speed Bi-CMO circuits and devices. From 199 to the present, he has been with UCB working on ultra-high-speed compound semiconductor power HBTs and MMICs. Mattias Dahlstrom obtained his M. c in Eng. Physics and his Ph. D in Photonics from the Royal Institute of Technology (KTH), weden. He is currently at UCB further pursuing ultra high speed InP HBTs. His main focus is on device design, process development and high frequency measurements. Miguel Urteaga received his B. A. c. Degree in engineering physics from imon Frasier University, Vancouver, BC, Canada, and the M..

10 degree and Ph. D in electrical and computer engineering from the University of California at anta Barbara in 199, 1 and 3 respectively. He is currently with Rockwell cientific Corporation, Thousand Oaks, CA. His reaserch included device design and fabrication of high-speed InP HBTs, as well as the design of ultra-high-frequency integrated circuits. Navin Parthasarathy obtained his M.c (Hons) degree in Physics and BE (Hons.) degree in Electrical & Electronic Engineering from the Birla Institute of Technology and cience, Pilani, India in. He is currently pursuing his PhD at the University of California, anta Barbara under the supervision of Prof. Mark Rodwell. His research work includes the design and fabrication of InP based high speed transistors and circuits. Munkyo eo received the BEE and MEE degree in electronic engineering from eoul National University, Korea, in 199 and 199. From 1997 to, he was a research engineer at LG Electronics Inc., and designed RF sybsystem for wireless communication. He is currently working toward the Ph.D. degree at the University of California at anta Barbara. His research interest includes microwave/mixed-signal circuit design and digital signal processing. Lorene amoska received the B.. degree in Engineering Physics from the University of Illinois in 199, and the Ph.D. degree in Materials Engineering from the University of California, anta Barbara, in 199. he subsequently worked as a post-doctoral researcher in the Electrical and Computer Engineering Department at UC anta Barbara, where she engaged in the design and fabrication of state of the art microwave digital circuits based on InP HBTs. he joined the Jet Propulsion Laboratory in 199, where she is currently a enior Engineer involved in the design and testing of 7-3 GHz HEMT and HBT MMIC power amplifiers for local oscillator sources and transmitters in future space missions. Andy Fung received a BEE, MEE and Ph.D. in Electrical Engineering from the University of Minnesota in 1993, 199 and 1999, respectively. He is presently a member of the technical staff at the Jet Propulsion Laboratory, California Institute of Technology. His research involves the development of high speed InP HBTs and GaAs chottky diodes. Mark J. W. Rodwell received the B. degree from the University of Tennessee, Knoxville, in 19, and the M. and Ph. D degrees from tanford university, tanford, CA, in 19 and 19, respectively. He is Professor and Director of the Compound emiconductor Research Laboratories at the University of California at anta Barbara. He was with AT&T Bell Labs during His research focuses on very high-bandwidth bipolar transistors, high-speed bipolar IC design, and GHz mixed-signal ICs. His group has worked extensively in the area of GaAs chottky diode ICs for subpicosecond/millimeter wave instrumentation. Dr. Rodwell is the recipient of a 199 National cience Foundation Presidential Young Investigator award, and his work on submillimeterwave diode ICs was awarded the 1997 IEEE Microwave Prize and he was elected IEEE fellow in 3.