HFE0111 Firas

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

From January 2011 High Frequency Electronics

Copyright 2011 Summit Technical Media, LLC

High Frequency Design

CLASS-E AMPLIFIERS

Design of Input Matching


Networks for Class-E
RF Power Amplifiers
Firas Mohammed Ali Al-Raie
The Polytechnic Higher Institute

lass-E power amplifiers have high


efficiency, which
makes them attractive in
modern wireless mobile
communication systems.
There are several techniques and approaches
developed solely to design
the output load networks of such amplifiers to
shape the RF power devices output voltage
and current for minimum power loss.
However, little attention is brought to the
design of the input matching network and to
the device bias conditions, with their effects on
the overall circuit performance. This paper
attempts to discuss these topics through a systematic design and simulation approach for a
typical 5 watt class-E power amplifier operating at 150 MHz.

Class-E power amplifer


design usually emphasizes
output network design, but
this article examines amplifier performance when an
untuned input is replaced
with a matching network

Introduction
Several methods have been developed for
the design of the load network for class-E RF
power amplifier. Among those are the shunt
capacitance [1], shunt inductance [2], finite
DC feed inductance [3], and parallel circuit [4]
techniques. The most popular configuration is
the shunt capacitance technique due to its
simplicity and designability, which means
that when the amplifier is built as designed, it
works as expected [1].
The schematic diagram of the class-E
power amplifier with shunt capacitance configuration is presented in Figure 1. In this circuit LG and LD represent the gate and drain
bias RF chokes respectively, CB is a DC blocking capacitor, Cb1 and Cb2 are bypass capacitors, VGG is the gate bias voltage, VDD is the
40

High Frequency Electronics

Figure 1 Typical class-E power amplifier


with shunt capacitance configuration.

drain supply voltage, C is the capacitor shunting the active device Q1, Lo and Co constitute
a series resonant circuit tuned at the operating frequency, and R is the optimum resistance seen by the load network for the
required output power. The active device Q1
(power MOSFET in this case) operates as an
ON/OFF switch.
In class-E power amplifier circuit, efficiency is maximized by minimizing power dissipation in the active device, while providing the
desired output power. The circuit can be
arranged so that high drain voltage and high
drain current dont exist at the same time.
For idealized class-E power amplifier operation, it is necessary to provide the following
optimum conditions for the drain voltage vD(t)
across the power MOSFET just prior to the
start of the devices ON state at the moment
t = T, where T is the period of the input driving signal [5]:
VD ( t )

t =T

=0

(1)

High Frequency Design

CLASS-E AMPLIFIERS

Figure 2 ID versus VGS for the


MRF134 power MOSFET.

Figure 3 Simulated drain DC characteristics of the MRF134 power


MOSFET.

dvD ( t )
=0
dt t = T

(2)

Equations (1) and (2) state that the drain voltage


should be zero at the turn-on moment, and that the slope
of this waveform is zero at the same moment.

Load Network Design Equations


The load network of class-E power amplifier is not
intended to provide a conjugate match to the transistor
output impedance. Design equations for the load network
elements (C, C0, L0, and R) can be derived by writing time
domain equations for the voltage vD(t) at the drain of the
RF power MOSFET when it is OFF, and the current iD(t)
passing through the RF device when it is ON. A set of
simultaneous differential equations can be formed
according to the necessary conditions (1) and (2) and
solved to determine the network elements [6].
Nathan Sokal, the inventor of this amplifier, has
developed explicit form equations to calculate the values
of the network elements at any output power and loaded
quality factor QL.
These equations are formulated as [7]:
V 2
0.451759 0.402444
R = 0.5768 DD 1

P
QL
QL 2

out

(3)

C=

0.91424 1.03175
0 .6
1

1 +
+
5.44658R
QL
QL 2 2 . LD

(4)

Co =

1.01468 0.2
1
1

1 +

R QL 0.104823
QL 1.7879 2 LD

(5)

Lo =

QL . R

(6)

where Pout is the required output power, and is the


operating frequency.
42

High Frequency Electronics

Figure 4 Configuration of the output matching network.

The value of QL can be selected based on a trade-off


between operating bandwidth and harmonic distortion of
the output signal.

Typical Class-E Power Amplifier Design


For the clarification of the goals of this paper, a power
amplifier circuit has been designed and simulated using
a commercial microwave CAD program. Design specifications of the amplifier are to achieve an output RF power
of 5 W from an input driving level of 0.5 W, and drain efficiency of more than 65% at an operating frequency of
150 MHz.
The following sections describe a step-by-step design
procedure with the simulated results obtained from the
Agilents ADS microwave circuit analysis program.

RF Power Device Selection and Characterization


The first step of the amplifier design procedure is the
selection of the RF power transistor. For this design, the
Motorolas power MOSFET MRF134 has been chosen.
This device is capable of delivering 5 W at 400 MHz with
a typical power gain of more than 10 dB. It operates from
a 28 VDC supply and has a typical drain-to-source breakdown voltage of 65 V. The RF transistor library of the
computer program ADS contains a SPICE model for this
transistor which simplifies the simulation process. The
simulated input DC characteristic (ID versus VGS) of the
power MOSFET is shown in Figure 2 with VDS = 28 V. It
can be shown from this curve that the gate threshold voltage VGS(th) = 3 V. On the other hand, Figure 3 presents the
simulated output DC characteristic (ID versus VDS) at
several gate voltages. The drain ON resistance RD(on) can
be estimated from Figure 3 as 12.5 . This relatively
large value of RD(on) will cause a reduction in amplifiers
efficiency due to the dissipated power at the drain during
the ON period of the power device.

Calculation of the Load Network Elements


The design procedure begins by calculating the component values of the load network using equations (3) to
(6). For output power Pout = 5 W, operating frequency f =

High Frequency Design

CLASS-E AMPLIFIERS

Figure 5 Input return loss of the


output matching network.

Figure 6 Insertion loss of the output matching network.

Figure 8 Simulated drain voltage


and current waveforms of the
power MOSFET.

Figure 6 presents its insertion loss


versus frequency.

Design of the Biasing Network


The biasing network consists of
the drain and gate RF chokes, bypass
capacitors, DC blocking capacitors, in
addition to the gate and drain bias
voltages. For 50% duty cycle operation, the transistor is biased at the threshold point, which
means that VGG = VGS(th) = 3 V. This bias point actually
corresponds to class-B mode.
Based upon the previous calculations, the schematic
diagram of the amplifier circuit is shown in Figure 7.

Figure 7 The designed class-E power amplifier without the input matching network.

150 MHz, loaded quality factor QL = 5, and drain power


supply VDD = 28 V, the calculated values of the elements
of the load network are:
R = 80 , C = 3 pF, C0 = 3 pF, and L0 = 430 nH.
The output capacitance of the RF device, Cout, is measured as 9.7 pF from the device data sheet. This means
that it is greater than the required value of the shunt
capacitance. The excess value of 6.3 pF should be tuned
out by part of the drain bias RF choke.

Design of the Output Matching Network


An output matching network is needed to transform
the 50 amplifier impedance into the required load resistance, which is set to be 80 . This network has been
designed with the aid of an immittance Smith chart, and
is implemented in a T-section configuration, as shown in
Figure 4.
In addition to the transformation function of the output matching network, it also can be used to reduce the
harmonic content of the output RF signal. Figure 5 shows
the simulated input return loss of this network, while
44

High Frequency Electronics

Amplifier Performance Simulation


The designed amplifier circuit has been simulated
using ADS 2006A. With a single tone input signal of 0.5 W
power level and an operating frequency of 150 MHz, the
RF devices drain voltage and current are sketched as
depicted in Figure 8. As shown from this plot, the peak
values of drain voltage and current dont exist simultaneously which minimizes the devices power loss. However,
at the ON time of the RF signal, the drain voltage is about
3.5V due to the ON resistance at the drain, RD(on). This
may degrade the overall efficiency of the circuit. During
the OFF interval of the RF signal, a negative current
flows through the power MOSFETs output capacitance
Cout.
In Figure 9 the output signal waveform is plotted,
while its spectrum is displayed in Figure 10. It is obvious
that harmonics are reduced to acceptable levels due to the

Figure 9 Simulated waveform of


the load voltage.

Figure 10 Power spectrum of the


output signal.

Figure 11 Amplifier output power


versus input power.

Figure 12 Power gain versus input


power.

Figure 13 Simulated efficiency


versus input power.

Figure 14 Variation of the large


signal MOSFET input impedance
with input signal level.

filtering effect of the load and matching networks.


In order to display the power amplifiers performance
characteristics, a sweep of the input power level has been
carried out from 10 to 30 dBm at the operating frequency.
Figure 11 shows a sketch of the output power versus
input power. The output power is obtained from:
Pout = real ( 0.5VL . I L* )

(7)

where VL and IL are the peak values of the fundamental


components of load voltage and current respectively. The
output power is about 36 dBm at an input level of
27 dBm.
Figure 12 presents the operating power gain of the
amplifier, Gp, versus input power. The power gain is calculated from:
G p ( dB ) = Pout ( dBm ) Pin ( dBm )

(8)

Notice that the power gain is about 9.0 dB at an input


power level of 27 dBm.
Finally, Figure 13 displays a plot of the amplifiers DC
to RF efficiency with input power. The amplifier efficiency is about 71.1% at an input power of 27 dBm.
The efficiency of the amplifier circuit has been evaluated from:
=

Pout
Pdc

(9)

where Pdc is the DC power consumed by the RF device


and is obtained from:

High Frequency Design

CLASS-E AMPLIFIERS

Figure 15 The designed class E power amplifier after adding the input matching network.

Figure 16 The simulated output


power versus input power for the
final amplifier.

Pdc = VDD . I DD

(10)

where IDD is the DC component of the


drain current.

Input Matching Network Design


The input matching network can
be designed to match the large signal
input impedance of the RF power
device with the 50 source
impedance. Therefore, the large signal input impedance of the RF transistor should be estimated at the
nominal input power, operating frequency, and bias voltages with the
existence of the load and output
matching networks.
The large signal input impedance
of the power transistor consists of two
parts, resistance Rin and reactance
Xin:
46

High Frequency Electronics

Figure 17 Operating power gain


versus input power.

Zin = Rin + jX in

(11)

Rin and Xin can be estimated from:


Rin = real ( Vin / Iin )

(12)

X in = imag ( Vin / Iin )

(13)

where Vin is the fundamental component of the input voltage at the gate
of the MOSFET, and Iin is the fundamental component of the current
entering the gate of the transistor. Iin
can be estimated using a current
probe with the aid of ADS simulation
capabilities. Figure 14 shows a sketch
of the large signal input impedance of
the power device versus input power.
As shown from the plot in Figure
14, the input impedance is capacitive.

Figure 18 Amplifier efficiency versus input power.

At an input power of 27 dBm (0.5 W),


the input impedance is approximately 12 j 45 . The input matching network can thus be designed to match
this value with the 50 source
impedance. An immittance Smith
chart has been used to construct an
L-section matching network graphically. Figure 15 presents the final
power amplifier circuit after incorporating the input matching network.
There is no doubt that the input
matching network improves the net
input power delivered to the RF
device. The amplifier circuit was simulated again after adding the input
matching circuit using ADS. The output power of the circuit is displayed
in Figure 16 with a sweep of input
power from 10 to 30 dBm. As shown
from Figure 16, there is a slight
increase in output power being

Engineers:

Stay Informed!
Subscribe to
High Frequency
Electronics !
Be on top of your game:

Timely technical articles


New product information
Industry news and events
Knowledgeable editors

Subscribe online at our Web site:


just click on the Subscriptions
button on our main page, www.
highfrequencyelectronics.com

High Frequency Design

CLASS-E AMPLIFIERS
38.8 dBm for an input power of
27 dBm. The power gain is plotted in
Figure 17, and becomes equal to
9.8 dB at the nominal input power.
The DC to RF efficiency is sketched
in Figure 18 versus input power. The
efficiency becomes 71.8% at an input
power level of 27 dBm. However, no
attempts have been made to optimize
the component values of the input
matching network for better performance characteristics.
Table 1 summarizes the performance of the amplifier before and
after adding the input network.

Conclusion
The performance of Class E RF
power amplifier with a traditional
shunt capacitance load network has
been studied thoroughly. It was
shown that the high efficiency operation of such amplifiers is determined
mainly by the output load network.
However, with an accurate and proper design of the input matching network, the performance characteristics of the amplifier can be improved.
This article has presented and discussed the main guidelines for synthesizing the input matching circuits
for this type of RF amplifier to
achieve the improved performance.

Acknowledgement

Print & Online


Editions!

Both editions are identical


Online = no mail delay!
Link to advertisers
Find articles quickly

SUBSCRIBE or
RENEW TODAY!
www.
highfrequencyelectronics
.com
48

High Frequency Electronics

I would like to express my deep


gratitude to Dr. Andrei Grebennikov
for his continuous advice and useful
suggestions in high efficiency power
amplifier analysis, and also for providing me with some technical
papers and e-books in the field of RF
power amplifiers.

References
1. N.O. Sokal and A.D. Sokal,
Class E- A New Class of HighEfficiency Tuned Single-Ended
Switching Power Amplifiers, IEEE
J. Solid-State Circuits, Vol. SC-10,
June 1975, pp. 168-176.
2. M.K. Kazimierczuk, Class E
Tuned Power Amplifier with Shunt
Inductor, IEEE J. Solid-State

Before

After

Output Power

36 dBm

36.8 dBm

Power Gain

9.0 dB

9.8 dB

Efficiency

71.1%

71.8%

Pin = 0.5W (27 dBm), f0 = 150 MHz

Table 1 Performance comparison


before and after adding the input
matching network.

Circuit, Vol. SC-16, February 1981,


pp. 2-7.
3. R.E. Zulinski and J.W.
Steadman, Class E Power Amplifiers
and Frequency Multipliers with
Finite DC-Feed Inductance, IEEE
Trans. Circuits and Systems, Vol.
CAS-34, September 1987, pp. 10741087.
4. A. Grebennikov, SwitchedMode RF and Microwave ParallelCircuit Class E Power Amplifiers,
Int. J. RF and Microwave ComputerAided Eng., Vol. 14, Jan/Feb 2004, pp.
21-35.
5. A. Grebennikov, Load Network
Design Techniques for Class E RF
and Microwave Amplifiers, High
Frequency Electronics, Vol. 3, July
2004, pp. 18-32.
6. A. Grebennikov and N.O. Sokal,
Switchmode RF Power Amplifiers,
Elsevier Inc., 2007, Chapter 5.
7. N.O. Sokal, Class-E RF Power
Amplifiers, QEX, Jan/Feb 2001, pp.
9-19.

Author Information
Firas Mohammed Ali Al-Raie is
an RF engineer in the Department of
Electronic Engineer at The Polytechnic Higher Institute of Yefren, Libya.
He has also worked as a consultant
engineer and as lecturer at the
University of Al-Jabal Al-Gharbi,
Jado, Libya. He has a BSc Degree in
Electrical Engineering and an MSc
Degree in Electrical Engineering/
Electronics and Communication from
the University of Baghdad.
He can be reached by e-mail at
[email protected]

You might also like