See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/251920641

A variable output power, high efficiency, power amplifier for the 2.45 GHz ISM
band

Article · November 2009

DOI: 10.1109/IMOC.2009.5427463

CITATION READS
1 121

5 authors, including:

Ahmed Sayed Georg Boeck

Egyptian Atomic Energy Authority Technische Universität Berlin
27 PUBLICATIONS   238 CITATIONS    258 PUBLICATIONS   2,052 CITATIONS   

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Georg Boeck on 10 August 2015.

The user has requested enhancement of the downloaded file.

 A Variable Output Power, High Efficiency, Power
Amplifier for the 2.45 GHz ISM Band
Sajjad Ahmed, Ahmed Sayed, Henrique Portela, Olof Bengtsson*, Georg Boeck
Microwave Engineering Laboratory, Berlin University of Technology
Einsteinufer 25, 10587 Berlin, Germany
*Ferdinand-Braun-Institut für Hoechstfrequenztechnik
Gustav-Kirchhoff-Str. 4,12489 Berlin, Germany
sayed@mwt.ee.tu-berlin.de

Abstract— This paper, presents the design and evaluation of a technology for high power applications in the 10 W to 300 W
high efficiency, 30 W, two-stage, power amplifier for the 2.45 power ranges below 3 GHz. With its high breakdown field,
GHz ISM band. The designed amplifier targets an industrial GaN high-power devices can be made small thereby reducing
application where precise control over the output power with the parasitic losses and increasing the efficiency [4]. In [5] a
good efficiency is necessary. A saturated output power of
50 dBm output power with 15 dB power gain and 63 % drain
44.7 dBm (30 Watts) was reached at a maximum drain efficiency
of 67.8 %. A full 5 dB control of the output power was achieved efficiency (η) is presented where as in [6], drain efficiency (η)
by varying the gate bias of the driver stage. A minimum power of 50 %, output power of 250 W and 15 dB power gain has
added efficiency of 39.4% was reached over the full controllable been reported.
output power range at the design frequency. The saturated In this paper, a two stage, 30 W PA at the 2.45 GHz ISM
output power over a 150 MHz bandwidth around the design band is presented. GaN-HEMT technology is used for the high
frequency was more than 44 dBm with a power added efficiency efficiency power stage and Si-LDMOS technology is used for
of more than 60 %. improved gain controllability in the driver stage. Part II
highlights the design of the amplifier, concept,
I. INTRODUCTION
implementation and realization. In part III the amplifier is
Radio frequency (RF) commonly refers to high frequency evaluated, under small signal operation for model verification
radio communication but there is an entirely different category and under large signal conditions, simulations are compared to
where RF devices are used for non-communication purposes. measurements under all conditions. Part IV is a brief
The frequency band utilized for these applications is known as discussion of the possibilities with the design and in part V the
Industrial, Scientific and Medical (ISM) band and applications work is concluded.
in this band are centred at different frequencies ranging from a II. AMPLIFIER CONCEPT AND DESIGN
few MHz to the mm-wave range [1]. Applications at ISM-
In the application, the power amplifier is to deliver
band were for a long time considered as non-commercial but
microwave power up to 30 W (44.8 dBm) at 2.45 GHz with
recently these bands have seen a greater commercial interest,
precise control of output power with high efficiency at a
mainly because they are now also utilized for communication
constant available input power of 100 mW (20 dBm).
like WiFi and Bluetooth applications. Other applications
Considering these specifications a two-stage amplifier was
which fall under the ISM category include RF heating, plasma
designed with a Cree AlGaN/GaN HEMT (CGH40025H)
generators, laser exciters, RF lighting, medical telemetry,
transistor for the power stage (PS) and a Freescale Si LDMOS
magnetic resonance imaging and many others [1]. In some of
(MW6S004N) transistor for the driver stage (DS). A block
the applications, the RF output power (POUT) is a few mili-
diagram for the two-stage amplifier is shown in Fig. 1.
watts where as it might reach hundreds of watts in other Power Stage
Driv er stage
applications. Traditionally the vast amount of power used in Interstage
Input Output
heating applications have ruled out solid-state solutions but matching matchig matching
Load
recent technology development and application demands are source

now starting to make transistor solutions a cost efficient

alternative [2],[3]. For the heating application this work is
targeting, moderate power but controllability of the output
Fig. 1 Block diagram for two stage power amplifier
power is of main concern. The amount of output power
directly relates to the temperature of the heated object. High The high power density of GaN makes the device size very
efficiency is needed over the full output range to reduce small which corresponds to high impedances at source and
operating cost, size and the cost of additional heat removing load compared to other devices which in-turn makes matching
hardware. easier [7]. The selection of Si LDMOS transistor for the driver
Lately, Gallium-Nitride (GaN) high electron mobility stage is primarily because of its high gain and high drain
transistor (HEMT) technology has emerged as a high voltage characteristics. To achieve the objective of high
efficiency alternative to the mainstream Silicon LDMOS

978-1-4244-5357-3/09/$26.00©2009IEEE 821
efficiency the transistors are biased towards class B operation. C. Design Realization
The amplifier was designed for a drain supply (VD) of 28V. The amplifier was built on Rogers RO4003 substrate with a
A. Design Procedure dielectric constant εr of 3.66 and thickness of 0.51 mm. Fig. 2
shows a prototype for the designed PA.
Large-signal models were available for both transistors
from the manufacturers. Although their validity could be
questioned they were later shown to well model the behaviour
under the conditions used in this design. Initially harmonic-
balance, load-pull simulations were conducted in Agilent-
ADS to establish a high power added efficiency (PAE) at the
requested output power for the power stage. The trade-off
between output power, gain, efficiency and quiescent point
was observed in the simulations and an optimal quiescent
current of around 10 % and 5 % of the maximum saturated Fig. 2 Power amplifier prototype
drain current for both power and driver stage respectively Via holes were manually fabricated. Thermal control was
were determined to achieve the desired objectives of output established using an aluminium heat sink with a cooling fan.
power and efficiency. Independent load and source-pull Bias was separately controllable for the two stages in the
simulations were then performed for both stages. Initially, initial tests but was later fixed for the power stage and
load and source-pull simulations were performed with ideal integrated with a temperature-gain control circuitry. The
components and ideal optimum load and source impedances matching networks were made on separate substrates so that
were found. Independently both transistors were stabilized to linear performance of the networks could be measured and
unconditional stability by fulfilling the Rollet’s stability verified individually.
criterion [8] (K > 1). This was achieved by adding a parallel
resistance with capacitance in series at the input, effectively III. AMPLIFIER EVALUATION
reducing the low frequency gain. λ/4 transmission lines were The design process was totally based on a model approach.
used in the biasing networks to improve isolation at RF Small signal evaluation was therefore used to verify the
frequency. Load and source pull simulation were performed correctness of the models used thereby validating the design
again to achieve optimum impedances for the two stages. The process.
optimum load and source impedances found are summarized
Table I. A maximum optimum driver stage quiescent current A. Small Signal Performance
(IDQ(DS)) was found to be 32 mA at an optimum power stage The small-signal s-parameters of the full amplifier are show
quiescent current (IDQ(PS)) of 123 mA. in Fig. 3.
TABLE I 40
OPTIMUM LOAD AND SOURCE IMPEDANCE AND QUIESSENT S21sim S21meas
S11sim S22meas
CURRENT S22sim S11meas

ZLOAD ZSOURCE 20
IDQ
S-Parameters [dB]

[] [] [mA]

Power stage
6.2 – j10.3 6.7 – j26.5 123 0
(PS)
Driver stage
6.0 – j6.6 5.0 – j30.0 32
(DS)
-20
B. Design Implementation
The matching networks were realized using micro-strip
lines together with open stubs in low pass form. Input
matching was designed to achieve good return loss. It was -40
2.0 2.2 2.4 2.6 2.8 3.0
implemented using a tapered line to compensate for some of ƒ [GHz]
the series capacitance introduces by the capacitor in the
stabilizing network. The inter-stage matching network was Fig. 3 Measured and simulated small signal s-parameters at VD = 28 V and
designed to minimize mismatch losses and for maximum maximum drain current for the driver stage.
transfer of power between the two stages. The output It can be observed that a small-signal gain of 27 dB with
matching network was designed for minimum loss to the load input and output reflections less than -15 dB in the operating
and a simulated insertion loss < 0.1 dB was achieved for a frequency is achieved. The K factor was measured to > 4 over
frequency range of 200 MHz with centre frequency of the frequency band from 1 kHz – 10 GHz. This is mainly
2.45 GHz. Momentum simulations were used to verify the because both the stages were individually stabilized before
impedances created by the linear matching networks. cascading. Compared to simulations, gain measurements

2009 SBMO/IEEE MTT-S International Microwave & Optoelectronics Conference (IMOC 2009) 822
exhibit some degradation although the main behaviour is very C. Power Performance over Bandwidth
well tracked by the simulations. The small discrepancy is a Although the amplifier was designed for a fixed frequency
combination of many factors, transistor model in-accuracy, use at 2.45 GHz the ISM band allows some frequency
tolerance of the used devices and tolerance in the fabricated modulation of the signals. The amplifier was therefore
boards. measured over a limited bandwidth around the design
B. Power Performance at Maximum Bias frequency. Fig. 6 shows the power performance at saturation
over a 150 MHz frequency range.
A signal generator together with a pre-amplifier was used
80
as power source in order to generate desired maximum
available input power (PAVS) of 20 dBm during the power
measurements. Initially the maximum operating performance
was tested at maximum driver stage biasing, IDQ(DS) = 32 mA. 60
The measured results of output power, transducer power gain

PA Performance
(Gain), drain efficiency and power added efficiency are shown
in Fig. 4 and Fig. 5. 40
50

20
40 Gainsim [dB] Gainmeas [dB]
POUT[dBm], Gain [dB]

PAE sim [%] PAEmeas [%]

Pout sim [dBm] Pout meas [dBm]
0
2.35 2.40 2.45 2.50
30
ƒ [GHz]

Fig. 6 Output power and PAE for a frequency range of 150 MHz at VD = 28
20 V, IDQ(DS) = 23 mA and IDQ(PS) = 123 mA.

An output power of more than 29 W and a PAE of more

Gainsim [dB] Gainmeas[dB]
than 60 % could be measured over the full 150 MHz
Poutsim [dBm] Poutmeas [dBm]
10 bandwidth. Very good agreement was also found over the
5 10 15 20 25 band between simulated and measured results.
Fig. 4 Measured and simulated gain and output power performance at VD = D. Power Performance over Gain Control
28 V, IDQ(DS) = 23 mA and IDQ(PS) = 123 mA.
The gain of the amplifier is controlled by setting the gate
80 voltage of the driver stage LDMOS thereby adjusting IDQ(DS).
Fig. 7 shows the output power controllability of the amplifier.
46

60
Efficiency [%]

44

40
POUT[dBm]

42

20

 sim [%] PAE sim [%] 40

 meas [%] PAE meas [%]
0 POUT(meas)
5 10 15 20 25 POUT(sim)
Input Power dBm  38
0 10 20 30 40
Fig. 5 Measured and simulated drain efficiency and PAE at VD = 28 V, IDQ(DS) IDS_Driver [mA]
= 23 mA and IDQ(PS) = 123 mA.
Fig. 7 Measured and simulated power control at f = 2.45 GHz. VD = 28 V,
A saturated output power of 44.8 dBm (30 W) and a gain of and IDQ(PS) = 123 mA.
25 dB was achieved. The maximum drain efficiency is 67.8 %
at saturated power along with an almost similar PAE due to As shown the controllability tracks very well the simulated
the high gain of the amplifier. Very good agreement between values. By varying the current of the driver stage between
simulation and measurements were observed. 3 mA and 33 mA at a constant input power (PIN) at 100 mW
(20 dBm) a measured tuneable range of 5 dB was achieved

2009 SBMO/IEEE MTT-S International Microwave & Optoelectronics Conference (IMOC 2009) 823
 (39 dBm to 44 dBm). This was somewhat less than the 6 dB V. CONCLUSIONS
expected from simulations. The achieved PAE during power In this paper the design of a two stage 30 W power
control is shown in Fig. 8 amplifier with controllable gain for the ISM band is presented.
70 Power control and high efficiency over the controllable output
Gain (meas) Gain (sim) power range were the main design objectives. A model based
PAE (meas) PAE (sim)
60 design approach was used and verified in the work. High gain
requirement forced it to be a two-stage amplifier where as
Gain [dB], PAE [%]

50
high efficiency over the full controllable power range was
achieved using GaN HEMT technology in the power stage
and Si-LDMOS technology in the driver stage. A gain of
40
25 dB was measured for the two stage amplifier at the
maximum output power of 30 W. In compression a drain
30
efficiency of 67.8 % and PAE of 67.2 % was achieved. The
amplifier performs well over a full bandwidth of 150 MHz
20 with a saturated output power of more than 44 dBm at a PAE
> 60% for the full band.
10 The output power is controllable within a 5 dB range by
39 40 41 42 43 44
setting the bias of the driver stage. More than 40 % PAE is
P OUT  dBm  maintained over the full controllable output power range
which makes the amplifier well suited for the heating
Fig. 8 Measured and simulated gain and PAE during power control at
f = 2.45 GHz. VD = 28 V and IDQ(PS) = 123 mA.
application it was designed for.

From this figure it can be seen that a PAE of almost 40 % is ACKNOWLEDGMENTS

maintained for the two-stage amplifier over the full tuneable The authors would like to thank Cree for providing the
range. model for the GaN-HEMT transistor used and Freescale for
providing the model for the Si-LDMOS used.
IV. DISCUSSION
Although this amplifier was designed for heating REFERENCES
applications at ISM band where power generated at other [1] “The ISM Revolution: The Next Big Thing”, By Iboun Taimiya Sylla,
frequencies is of little concern but the topology used may be Texas Instruments, 9th Feburary, 2009, www.rfdesignline.com
very suitable for more frequency sensitive applications [2] [2] Wojciech Wojtasiak, Daniel Gryglewski and Wojciech Gwarek,' “A
lOOW ISM-2.45GHz BAND POWER TEST SYSTEM,” *Institute of
within the ISM band. This covers e.g. applications within Radioelectronics, Warsaw University of Technology, Poland. Journal
communication and microwave enhanced chemical reactions. of telecommunications and information technology, 2005.
Fig. 9 represents the two-tone IMD performance of the [3] Isao Takenaka, Hidemasa Takahashi, Kazunori Asano, Junko
designed amplifier. Morikawa, Kohji Ishikura, Mikio Kanamori", Masaaki Kuzuhara and
Hiroaki Tsutsui, “High efficiency S-band 30W power GaAs FETs”
80 ULSI Device Development Laboratories, NEC Corporation. 9-1
,Seiran 2-Chome70tsu,Shiga 520,Japan.
P3(meas) P3(sim)
[4] A. H. Jardnal, “Large signal modelling for GaN devices for high power
60 P2(meas) P2(sim)
P1(meas) P1(sim) amplifier”, PhD., University of Kassel, Germany, Nov. 2006.
Output Power 2-tone [dBm]

[5] Nitronex Corporation, “GaN Power HEMT for 2.5 GHz WiMAX
40 Applications”, NPT25100, 2305 Presidential Drive. Durham, NC
27703.
20 [6] K. Krishnamurthy, M. J. Poulton, J. Martin, R. Vetury, J. Brown, and J.
IMD3 ~65 dBc
B. Shealy, “A 250 W S-Band GaN HEMT Amplifier”, Infrastructure
0 Product Line, RF Micro Devices, Inc., Charlotte, NC 28269, USA.
[7] J. Shumaker, M. Ohoka, and N. Ui, “Design of Power Amplifiers
-20 Using High Breakdown GaN HEMT Devices”, Eudyna Devices USA
Inc, and Eudyna Devices Inc Japan.
-40 [8] S. Cripps, “Advanced Techniques in RF Power Amplifier Design”, 1st
Ed., Artech House, 2002.
-60

-20 -10 0 10 20 30
Input Power 2-tone [dBm]

Fig. 9 Two-tone IMD performance of the amplifier for a tone-spacing of 100

MHx at the deigns frequency 2.45 GHz at VD = 28 V, IDQ(DS) = 23 mA and
IDQ(PS) = 123 mA.

2009 SBMO/IEEE MTT-S International Microwave & Optoelectronics Conference (IMOC 2009) 824

View publication stats