Download Power MOSFET Linearizer of a High

sensors
Article
Power MOSFET Linearizer of a High-Voltage
Power Amplifier for High-Frequency
Pulse-Echo Instrumentation
Hojong Choi 1 , Park Chul Woo 1 , Jung-Yeol Yeom 2, * and Changhan Yoon 3, *
1
2
3
*
Department of Medical IT Convergence Engineering, Kumoh National Institute of Technology,
Gumi 39253, Korea; [email protected] (H.C.); [email protected] (P.C.W.)
School of Biomedical Engineering, Korea University, Seoul 02841, Korea
Department of Biomedical Engineering, Inje University, Gimhae 50834, Korea
Correspondence: [email protected] (J.-Y.Y.); [email protected] (C.Y.); Tel.: +82-2-3290-5662 (J.-Y.Y.);
+82-55-320-3301 (C.Y.)
Academic Editors: Xiaoning Jiang and Chao Zhang
Received: 28 February 2017; Accepted: 31 March 2017; Published: 4 April 2017
Abstract: A power MOSFET linearizer is proposed for a high-voltage power amplifier (HVPA) used
in high-frequency pulse-echo instrumentation. The power MOSFET linearizer is composed of a DC
bias-controlled series power MOSFET shunt with parallel inductors and capacitors. The proposed
scheme is designed to improve the gain deviation characteristics of the HVPA at higher input
powers. By controlling the MOSFET bias voltage in the linearizer, the gain reduction into the HVPA
was compensated, thereby reducing the echo harmonic distortion components generated by the
ultrasonic transducers. In order to verify the performance improvement of the HVPA implementing
the power MOSFET linearizer, we measured and found that the gain deviation of the power MOSFET
linearizer integrated with HVPA under 10 V DC bias voltage was reduced (−1.8 and −0.96 dB,
respectively) compared to that of the HVPA without the power MOSFET linearizer (−2.95 and
−3.0 dB, respectively) when 70 and 80 MHz, three-cycle, and 26 dBm input pulse waveforms are
applied, respectively. The input 1-dB compression point (an index of linearity) of the HVPA with
power MOSFET linearizer (24.17 and 26.19 dBm at 70 and 80 MHz, respectively) at 10 V DC bias
voltage was increased compared to that of HVPA without the power MOSFET linearizer (22.03 and
22.13 dBm at 70 and 80 MHz, respectively). To further verify the reduction of the echo harmonic
distortion components generated by the ultrasonic transducers, the pulse-echo responses in the
pulse-echo instrumentation were compared when using HVPA with and without the power MOSFET
linearizer. When three-cycle 26 dBm input power was applied, the second, third, fourth, and
fifth harmonic distortion components of a 75 MHz transducer driven by the HVPA with power
MOSFET linearizer (−48.34, −44.21, −48.34, and −46.56 dB, respectively) were lower than that of the
HVPA without the power MOSFET linearizer (−45.61, −41.57, −45.01, and −45.51 dB, respectively).
When five-cycle 20 dBm input power was applied, the second, third, fourth, and fifth harmonic
distortions of the HVPA with the power MOSFET linearizer (−41.54, −41.80, −48.86, and −46.27 dB,
respectively) were also lower than that of the HVPA without the power MOSFET linearizer
(−25.85, −43.56, −49.04, and −49.24 dB, respectively). Therefore, we conclude that the power
MOSFET linearizer could reduce gain deviation of the HVPA, thus reducing the echo signal harmonic
distortions generated by the high-frequency ultrasonic transducers in pulse-echo instrumentation.
Keywords: high-voltage power amplifier; ultrasonic transducer; power MOSFET linearizer
Sensors 2017, 17, 764; doi:10.3390/s17040764
www.mdpi.com/journal/sensors
improvements of the HVPA, such as gain, bandwidth, and linearity have been one of the critical
technical issues. Especially, linearity is an important parameter for HVPA used in the ultrasound
transmitters because the non-linear HVPA leads to high harmonic distortions, thus deteriorating the
echo signal quality of the ultrasonic transducers [5].
For
high-frequency
ultrasonic transducers compared to low-frequency ultrasonic transducers,
Sensors
2017,
17, 764
2 of 12
the size of the piezoelectric materials (the main material of the ultrasonic transducers) is small, thus
reducing the capacitance values of the piezoelectric materials as the operating frequency of the
1. Introduction
ultrasonic
transducers are increased [6,7]. Therefore, the parasitic capacitances of the coaxial cables
can affect
the performances
of the high-frequency
transducers.
addition, the maximum
A pulse-echo
instrumentation
typically ultrasonic
consists of
ultrasonicIn transmitters,
receivers,
excitation
voltages
into
the
high-frequency
ultrasonic
transducers
are,
accordingly,
limited(HVPA)
due to
and transducers [1–3]. In the ultrasonic transmitters, the high voltage power amplifier
the
size
reduction
of the components
piezoelectricsince
materials
[8]. triggers
These external
interventions
affect
the
is one
of the
main electronic
the HVPA
the ultrasonic
transducers
through
performances
of
high-frequency
pulse-echo
instrumentation.
Therefore,
linearizing
techniques,
such
the expander devices, which consist of cross-coupled diodes [2,4]. Therefore, the performance
as
compensating
active
and passive
components
improvements
of for
the several
HVPA, non-linear
such as gain,
bandwidth,
and electronic
linearity have
been oneinofthe
theHVPA,
critical
could
help
improve
HVPA
linearity,
thus
improving
the
performance
of
high-frequency
pulse-echo
technical issues. Especially, linearity is an important parameter for HVPA used in the ultrasound
instrumentations.
Thethe
basic
operating
principle
linearizer
and
HVPA is described
in Figurethe
1.
transmitters because
non-linear
HVPA
leadsof
tothe
high
harmonic
distortions,
thus deteriorating
As
depicted
in
the
figure,
the
gain
of
the
HVPA
typically
decreases
as
the
input
power
increases
echo signal quality of the ultrasonic transducers [5].
due to
thehigh-frequency
sensitive parasitic
impedances
[9–12]. compared to low-frequency ultrasonic transducers,
For
ultrasonic
transducers
Linearizer
techniques
have
been
widely
researched
applications
[11].
the size of the piezoelectric materials (the main
materialinofRF
thecommunication
ultrasonic transducers)
is small,
However,
the linearizer
techniques
only
been recently
introduced
the ultrasound
research
thus reducing
the capacitance
valueshave
of the
piezoelectric
materials
as thein
operating
frequency
of the
[13,14].
The
systematic
approaches
use
the
analog-to-digital
converter,
digital-to-analog
converter,
ultrasonic transducers are increased [6,7]. Therefore, the parasitic capacitances of the coaxial cables
and
which lead
to high-frequency
increased cost, larger
size,transducers.
and higher complexity
can memory
affect thespaces,
performances
of the
ultrasonic
In addition,[13].
the Therefore,
maximum
itexcitation
is not preferable
for
implementation
into
ultrasound
array
transducer-based
systems.
voltages into the high-frequency ultrasonic transducers are, accordingly, limited due toThe
the
pre-linearizer
technique
provides
low cost,
small external
size, and
low complexity
compared
to the
size reduction of
the piezoelectric
materials
[8]. These
interventions
affect the
performances
of
systematic
approaches
[11].instrumentation.
However, the reported
pre-linearizer
requires complex
mathematical
high-frequency
pulse-echo
Therefore,
linearizing techniques,
such as compensating
model
analysis
to compensate
non-linearity
the HVPAinand
it is not
suitable
for the
for several
non-linear
active andthe
passive
electronic in
components
the HVPA,
could
help improve
ultrasound
HVPA
utilizing
input
and
output
DC
coupling
capacitors
[14].
Therefore,
we
develop
a
HVPA linearity, thus improving the performance of high-frequency pulse-echo instrumentations.
simple
linearizer
circuit
to beof
used
for HVPAand
ultrasound
As shown
Figure 1,
The basic
operating
principle
the linearizer
HVPA isapplications.
described in Figure
1. Asindepicted
in the
the
linearizers
need
to
be
designed
to
have
opposite
characteristics
with
the
HVPA.
Therefore,
the
figure, the gain of the HVPA typically decreases as the input power increases due to the sensitive
linearizer
gain deviation
should be positively increased to reduce the gain deviation of the HVPA.
parasitic impedances
[9–12].
Linearizer
Gain
deviation
HVPA + Linearizer
HVPA
Input power
Figure1.1.Gain
Gaindeviation
deviationof
ofHVPA
HVPAwith
withand
andwithout
without the
the linearizer
linearizer as
as aa function
function of
of input
input power.
power.
Figure
Linearizer
been widely
researched
in RF communication
applications
[11].
The
HVPA techniques
in ultrasonichave
transmitters
usually
utilizes large-value
coupling capacitors
in order
However,
linearizer
techniques
have onlythe
been
introduced
the ultrasound
to
block DCthe
voltages
which
in turn deteriorate
echorecently
signal quality
of thein performances
of
research [13,14].pulse-echo
The systematic
approacheslike
usehigh
the analog-to-digital
converter,
digital-to-analog
high-frequency
instrumentations
harmonic distortion
components
of the echo
converter,
and memory
spaces,
which lead
to increased
cost,
and input
higherpowers
complexity
[13].
signals
because
the coupling
capacitors
reduce
the gain of
thelarger
HVPAsize,
at high
[2,15,16].
Therefore,
is not preferable
implementation
ultrasound
array
transducer-based
systems.
In
a typicalit HVPA
topology, for
large-value
couplinginto
capacitors
in the
input
and output ports
are
The pre-linearizer
technique
low
cost,the
small
and
low complexity
compared
to the
required
to block DC
voltages provides
as they can
affect
echosize,
signal
quality
of the ultrasonic
transducers.
systematic approaches
[11]. However,
pre-linearizer
requirescomponents
complex mathematical
Unfortunately,
these coupling
capacitorsthe
arereported
inherently
non-linear passive
such that it
model
analysisunwanted
to compensate
the non-linearity
in the HVPA
it is notdistortions
suitable forgenerated
the ultrasound
also
generates
harmonic
distortions. Therefore,
theand
harmonic
from
HVPA utilizing input and output DC coupling capacitors [14]. Therefore, we develop a simple
linearizer circuit to be used for HVPA ultrasound applications. As shown in Figure 1, the linearizers
need to be designed to have opposite characteristics with the HVPA. Therefore, the linearizer gain
deviation should be positively increased to reduce the gain deviation of the HVPA.
The HVPA in ultrasonic transmitters usually utilizes large-value coupling capacitors in order
to block DC voltages which in turn deteriorate the echo signal quality of the performances of
high-frequency pulse-echo instrumentations like high harmonic distortion components of the echo
signals because the coupling capacitors reduce the gain of the HVPA at high input powers [2,15,16].
In a typical HVPA topology, large-value coupling capacitors in the input and output ports are
required to block DC voltages as they can affect the echo signal quality of the ultrasonic transducers.
Sensors 2017, 17, 764
3 of 12
Sensors 2017, 17, 764
3 ofit
12
Unfortunately,
these coupling capacitors are inherently non-linear passive components such that
also generates unwanted harmonic distortions. Therefore, the harmonic distortions generated from
HVPAdirectly
directlyaffect
affectthe
theperformance
performanceofofultrasonic
ultrasonictransducers.
transducers.InInorder
ordertotoreduce
reducethe
theharmonic
harmonic
HVPA
distortion
components
of
the
echo
signal
generated
by
the
HVPA,
we
propose
employing
simple
distortion components of the echo signal generated by the HVPA, we propose employing simple
linearizertechniques
techniquesutilizing
utilizingvariable
variablepower
powerMOSFET
MOSFETdevices
deviceswith
withshunt
shuntinductors
inductorsand
andcapacitors.
capacitors.
linearizer
The
detailed
architecture,
operating
mechanism,
and
implementation
of
the
linearizer
an
The detailed architecture, operating mechanism, and implementation of the linearizer with anwith
HVPA
HVPA
are described
in 2.
Section
2. The capability
to the
reduce
the harmonic
distortions
of the
echo
are
described
in Section
The capability
to reduce
harmonic
distortions
of the echo
signals
signals
of
the
ultrasonic
transducers
are
verified
with
the
developed
linearizer
with
an
HVPA
in
of the ultrasonic transducers are verified with the developed linearizer with an HVPA in Section 3.
Section
3.
The
conclusion
of
the
paper
is
given
in
Section
4.
The conclusion of the paper is given in Section 4.
Materialsand
andMethods
Methods
2.2.Materials
Figure2a,b
2a,bshow
showschematic
schematicdiagrams
diagramsof
ofthe
theHVPA
HVPAand
andthe
thepower
powerMOSFET
MOSFETlinearizer.
linearizer.In
Inthe
the
Figure
HVPAarchitecture
architecture(Figure
(Figure2a),
2a),the
theHVPA
HVPAisisaaClass
ClassA-type
A-typepower
poweramplifier
amplifierused
usedininthe
theultrasound
ultrasound
HVPA
applications.The
Thelarge
large
value
pF coupling
DC coupling
capacitors
(C1C2and
2) are
DC
applications.
value
100100
pF DC
capacitors
(C1 and
) areCused
toused
blockto
DCblock
voltage.
voltage.
The two
variable
resistors
(R1 and
) are
to of
setthethe
bias voltage
of the
The
two variable
2 kΩ
resistors2(RkΩ
R2 ) are used
to setR2the
biasused
voltage
high-voltage
transistor
1 and
high-voltage
transistor
(T
2
).
The
3.3
μH
RF
choke
inductor
(L
1
)
is
used
to
maximize
the
voltage
(T
).
The
3.3
µH
RF
choke
inductor
(L
)
is
used
to
maximize
the
voltage
amplification
of
HVPA.
2
1
amplification
of the
The
2 μH inductor
(Lb1) is
used to prevent
bias
voltage
A 28
The
2 µH inductor
(Lb1HVPA.
) is used
to prevent
bias voltage
reduction.
A 28 V DC
voltage
(Vreduction.
DC ) is supplied
V DCa voltage
(VDC) is In
supplied
from
a power
supply.
In a typical
HVPA
topology,
large
value to
RF
from
power supply.
a typical
HVPA
topology,
a large
value RF
choke
inductora is
required
choke the
inductor
is required
to reduce
the DC
voltage
the power
supply
(VDC),
reduce
DC voltage
drop from
the power
supply
(VDCdrop
), thusfrom
allowing
the output
voltage
of thus
the
allowing
theamplified
output voltage
the HVPA
to be amplified
without
someoutput
suppression
the
HVPA
to be
withoutofsome
suppression
because the
maximum
voltagebecause
is limited
maximum
outputsupply
voltage[12].
is limited
by the
DCresistor
power issupply
[12]. If aoflarge
valuevalue
resistor
used
by
the DC power
If a large
value
used instead
the large
RF is
choke
instead of
largesupply
value DC
RF choke
theaffect
power
DC voltage
drop could
affect the
inductor,
thethe
power
voltageinductor,
drop could
thesupply
maximum
output voltage
amplification.
maximum
output
amplification.
Bias inductors
with
values are
also
set the
Bias
inductors
withvoltage
large values
are also needed
to set the
DClarge
bias voltage
after
theneeded
resistortodivider
DC bias
the resistor
divider circuit
(R1 bias
and voltage
R2). Therefore,
the maximum
DC gate
bias
circuit
(R1voltage
and R2 ).after
Therefore,
the maximum
DC gate
is generated
by the resistor
divider
voltage
is
generated
by
the
resistor
divider
circuits.
If
we
use
a
resistor
instead
of
the
inductor,
there
circuits. If we use a resistor instead of the inductor, there would be an additional DC voltage drop and
would
be limit
an additional
DC
voltage
drop and that
could limit the output voltage amplification [12].
that
could
the output
voltage
amplification
[12].
(a)
(b)
Figure2.2.Schematic
Schematicdiagrams
diagramsof
of(a)
(a)HVPA
HVPAand
and(b)
(b)power
powerMOSFET
MOSFETlinearizer.
linearizer.
Figure
As shown in Figure 2b, we designed a power MOSFET linearizer which consists of shunt
As shown in Figure 2b, we designed a power MOSFET linearizer which consists of shunt
capacitors and inductors with power MOSFET devices working with variable DC bias voltages. The
capacitors and inductors with power MOSFET devices working with variable DC bias voltages.
3.3 μH inductor value (LL3) is used to minimize the voltage drop from the power supply (VL1). The
The 3.3 µH inductor value (LL3 ) is used to minimize the voltage drop from the power supply (VL1 ).
100 pF capacitor (CL3) is used to apply stable DC bias voltage to the power MOSFET device (IRF620,
The 100 pF capacitor (CL3 ) is used to apply stable DC bias voltage to the power MOSFET device
Vishay Siliconix, Malvern, PA, USA) because DC power noise signals could go to ground. Two
(IRF620, Vishay Siliconix, Malvern, PA, USA) because DC power noise signals could go to ground.
capacitors (CL1 and CL2) are used to short possible unwanted AC noise from the power supply (VL1).
Two capacitors (CL1 and CL2 ) are used to short possible unwanted AC noise from the power supply
The 50-Ω coaxial cable is used to minimize the signal fluctuation between the power MOSFET
(VL1 ). The 50-Ω coaxial cable is used to minimize the signal fluctuation between the power MOSFET
linearizer and the HVPA.
linearizer and the HVPA.
Figure 3 shows the equivalent circuit model of the power MOSFET linearizer. As shown in
Figure 3, depending on the gate bias voltage, the MOSFET in the linearizer works as a variable
resistor (TR1) to provide variable performance when used in conjunction with two fixed 20 pF and
100 nH values of capacitors (CL1 and CL2) and inductors (LL1 and LL2). These capacitors are bypass
capacitors to block possible noise at the input port generated by the function generator and the
Sensors 2017, 17, 764
4 of 12
Figure 3 shows the equivalent circuit model of the power MOSFET linearizer. As shown in
Figure 3, depending on the gate bias voltage, the MOSFET in the linearizer works as a variable
resistor (TR1 ) to provide variable performance when used in conjunction with two fixed 20 pF and
100 nH values of capacitors (CL1 and CL2 ) and inductors (LL1 and LL2 ). These capacitors are bypass
capacitors to block possible noise at the input port generated by the function generator and the power
Sensors 2017, 17, 764
4 of 12
supply [12,17]. Low frequency ultrasound applications may not require bypass capacitors, however,
high-frequency
ultrasound
are very sensitive
to the performance
of thebypass
HVPAcapacitors,
due to the
power supply [12,17].
Lowapplications
frequency ultrasound
applications
may not require
limited
maximum
output power
of the ultrasonic
transducers
However,
as large
capacitance
however,
high-frequency
ultrasound
applications
are very [8].
sensitive
to the
performance
of can
the
affect
the
performance
of
the
circuit,
and
the
MOSFET
(T
)
inherently
has
large
parasitic
capacitance,
1
HVPA due to the limited maximum output power of the ultrasonic transducers [8]. However, as
smaller
value bypass
areperformance
preferred here.
theand
20 pF
value
selected tohas
be
large capacitance
cancapacitor
affect the
of Therefore,
the circuit,
thecapacitor
MOSFET
(T1) isinherently
much
less thancapacitance,
the parasiticsmaller
capacitances
the MOSFET
and output
capacitances
23020and
large parasitic
value of
bypass
capacitor(input
are preferred
here.
Therefore, =the
pF
70
pF).
The
total
output
of
the
power
MOSFET
linearizer
can
be
expressed
by:
capacitor value is selected to be much less than the parasitic capacitances of the MOSFET (input and
 power MOSFET linearizer can be

output capacitances = 230 and 70 pF). The total
output of the
expressed by:


1



 1
1 + jwCds
Vin1
1 jwLL2 

(1)
≈ 
+
Vin−1
≈ 
(1)
 

1
1

+

 1



1
11
+
jwCgd +
1 TR1
1 jwL L2 + jwCds
+
+
where w
w is
is the
the operating
operating frequency.
frequency.
where
LL3
CL3
TR1
In-1
LL1 Cgs
+CL1
Cgd
Vin1
CL2
+Cds
LL2
Figure
3. The
model of
of the
the power
power MOSFET
MOSFET linearizer.
linearizer.
Figure 3.
The equivalent
equivalent circuit
circuit model
As shown in Equation (1), the performance of the power MOSFET linearizer is determined by
As shown in Equation (1), the performance of the power MOSFET linearizer is determined
the variable resistor of the MOSFET (TR1) and inductor (LL2) because the MOSFET device has larger
by the variable resistor of the MOSFET (TR1 ) and inductor (LL2 ) because the MOSFET device has
pre-determined parasitic capacitances (Cgd and Cds) compared to the two capacitors (CL1 and CL2).
larger pre-determined parasitic capacitances (Cgd and Cds ) compared to the two capacitors (CL1 and
Therefore, the performance of the power MOSFET linearizer with different inductor values (LL2)
CL2 ). Therefore, the performance of the power MOSFET linearizer with different inductor values
were measured in order to find the appropriate operating frequencies of the power MOSFET
(LL2 ) were measured in order to find the appropriate operating frequencies of the power MOSFET
linearizer because the power MOSFET linearizer could reduce the gain of the HVPA at certain
linearizer because the power MOSFET linearizer could reduce the gain of the HVPA at certain operating
operating frequencies. The 100, 33, 22, and 12 nH inductors were selected to have minimum loss of
frequencies. The 100, 33, 22, and 12 nH inductors were selected to have minimum loss of the MOSFET
the MOSFET linearizer at 50, 70, 80, and 90 MHz operating frequencies. We selected a 33 nH
linearizer at 50, 70, 80, and 90 MHz operating frequencies. We selected a 33 nH inductance value to
inductance value to minimize the signal loss (lower than −3.8 dB) measured at 70 and 80 MHz. With
minimize the signal loss (lower than −3.8 dB) measured at 70 and 80 MHz. With the selected inductor
the selected inductor values, the performances of the MOSFET linearizer could essentially be
values, the performances of the MOSFET linearizer could essentially be controlled depending on
controlled depending on the MOSFET resistance (TR1). In Section 3, the power MOSFET linearizer is
the MOSFET resistance (TR1 ). In Section 3, the power MOSFET linearizer is characterized because
characterized because the variable resistances of the MOSFET can be varied under different DC
the variable resistances of the MOSFET can be varied under different DC values (VL1 ) as shown in
values (VL1) as shown in Figure 2b.
Figure 2b.
3. Results and Discussion
3.1. Linearizer Performance Measurement
The HVPA and power MOSFET linearizer were implemented on a printed circuit board with
two layers. In the HVPA, the heat-sink was attached on top of the high-voltage transistors as shown
Sensors 2017, 17, 764
5 of 12
3. Results and Discussion
3.1. Linearizer Performance Measurement
The HVPA and power MOSFET linearizer were implemented on a printed circuit board with
two layers. In the HVPA, the heat-sink was attached on top of the high-voltage transistors as
shown
in Figure
Sensors
17,
55 of
Sensors 2017,
2017,
17, 764
764 4. Our target ultrasonic transducer frequency for the pulse-echo instrumentations
of 12
12
is over 50 MHz, as it is known that in studies with the eye, intravascular, acoustic stimulation,
resolutions
compared
to low-frequency
than
15 MHz)applications
ultrasonic applications
[18–21].
and
cavitation,
high-frequency
(over 50(less
MHz)
ultrasonic
have reported
betterFigure
spatial5
shows
the
experimental
setup
to
measure
the
gain
deviation
at
1-dB
compression
point
of the
resolutions compared to low-frequency (less than 15 MHz) ultrasonic applications [18–21]. Figure
5
HVPA the
with
and withoutsetup
the power
MOSFET
linearizer
in order
to acquire
the proper
bias
shows
experimental
to measure
the gain
deviation
at 1-dB
compression
pointworking
of the HVPA
voltages
the power
MOSFET
linearizer.
The gate-threshold
voltage
of the MOSFET
the
power
with
and of
without
the power
MOSFET
linearizer
in order to acquire
the proper
working in
bias
voltages
MOSFET
linearizer
is typically
2 V and 4 V,voltage
depending
the drain-source
voltage
such
of
the power
MOSFET
linearizer.between
The gate-threshold
of theon
MOSFET
in the power
MOSFET
that the power
MOSFET
linearizer
performs
as a variable
over the
gate-threshold
voltages.
linearizer
is typically
between
2 V and
4 V, depending
on theresistor
drain-source
voltage
such that the
power
The three-cycle
70 or
80 MHzas
input
pulsesresistor
generated
an arbitrary signal
generator
(AFG3252,
MOSFET
linearizer
performs
a variable
overfrom
the gate-threshold
voltages.
The three-cycle
Tektronix
Inc.,input
Beaverton,
OR, USA) from
wereanapplied
to signal
the HVPA
without
and with
the power
70
or 80 MHz
pulses generated
arbitrary
generator
(AFG3252,
Tektronix
Inc.,
MOSFET linearizer
preset DC
biasing
a high-voltage
(E3631A,
Beaverton,
OR, USA)under
were applied
to the
HVPAwith
without
and with thepower
power supply
MOSFET
linearizerAgilent
under
Technologies,
Santa
CA, USA).
Thesupply
output
was then
fed Technologies,
through twoSanta
20 dB
power
preset
DC biasing
withClara,
a high-voltage
power
(E3631A,
Agilent
Clara,
CA,
attenuators
(BW-S20W20+,
Brooklyn,
NY, USA)
and acquired
with anMini-circuits,
oscilloscope
USA).
The output
was then Mini-circuits,
fed through two
20 dB power
attenuators
(BW-S20W20+,
(MSOX4154A,
Keysight
Technology,
Santa
Clara, CA,
USA) to show
the Technology,
gain deviations
the
Brooklyn,
NY, USA)
and acquired
with an
oscilloscope
(MSOX4154A,
Keysight
Santaof
Clara,
HVPA
with
without
thedeviations
power MOSFET
linearizer.
CA,
USA)
toand
show
the gain
of the HVPA
with and without the power MOSFET linearizer.
Figure
board.
Figure 4.
4. The
The implemented
implemented power
power MOSFET
MOSFET linearizer
linearizer and
and HVPA
HVPA on
on aa printed
printed circuit
circuit board.
board.
Figure
5. The
measurement setup
for
with
and
without
the
power
MOSFET
linearizer.
Figure
for HVPA
HVPA with
with and
and without
without the
the power
power MOSFET
MOSFET linearizer.
linearizer.
Figure 5.
5. The
The measurement
measurement setup
setup for
HVPA
In order to assess the power MOSFET linearizer characteristics, we measured the voltage gain
In
order
to assess
power
MOSFET
linearizer
welinearizer
measured
voltage
gain
deviation
graphs
of thethe
HVPA
with
and without
the characteristics,
power MOSFET
at the
70 and
80 MHz
deviation
graphs
of
the
HVPA
with
and
without
the
power
MOSFET
linearizer
at
70
and
80
MHz
under 5 V DC bias. The gain deviation graphs of (1) the power MOSFET linearizer; (2) the HVPA
under
5 V(3)
DCthe
bias.
The with
gain the
deviation
of (1)
the power
linearizer;
(2) the
HVPA
only, and
HVPA
power graphs
MOSFET
linearizer
at 70MOSFET
MHz input
power are
plotted
in
only,
HVPA with
power MOSFET
linearizer
at 70linearizer
MHz input
in
Figureand
6a. (3)
Thethe
measured
gainthe
deviations
of the power
MOSFET
andpower
HVPAare
at plotted
20 and 26
Figure
6a. The
measured
gain
deviations
of the
power
MOSFET
linearizer
and HVPA
at 20deviations
and 26 dBof
m
dBmm input
power
were 0.06
and
0.35 dB, and
−0.265
and
−2.95 dB,
respectively.
The gain
input
power
were
0.06
and
0.35
dB,
and
−
0.265
and
−
2.95
dB,
respectively.
The
gain
deviations
of
the
the HVPA with the power MOSFET linearizer reduced to −0.261 and −2.09 dB, respectively. The gain
HVPA
with
the power
MOSFETlinearizer,
linearizer HVPA,
reducedand
to −
0.261 with
and −
2.09
dB, respectively.
The gain
deviation
graph
of the MOSFET
HVPA
the
power
MOSFET linearizer
at
70 MHz input power is plotted in Figure 6b. The measured gain deviations of the linearizer at 20 and
26 dBmm input power were 0.25 and 0.64 dB, respectively, and those of the HVPA at 20 dBmm and 26 dBmm
input power were −0.33 and −3.00 dB, respectively. The gain deviations of the HVPA with a power
MOSFET linearizer improved to −0.16 and −1.30 dB, respectively. From these results, we can
conclude that the gain deviation of the HVPA with the MOSFET linearizer was suppressed as the
Sensors 2017, 17, 764
6 of 12
deviation graph of the MOSFET linearizer, HVPA, and HVPA with the power MOSFET linearizer at
70 MHz input power is plotted in Figure 6b. The measured gain deviations of the linearizer at 20 and
26 dBm input power were 0.25 and 0.64 dB, respectively, and those of the HVPA at 20 dBm and 26 dBm
input power were −0.33 and −3.00 dB, respectively. The gain deviations of the HVPA with a power
MOSFET linearizer improved to −0.16 and −1.30 dB, respectively. From these results, we can conclude
that the gain deviation of the HVPA with the MOSFET linearizer was suppressed as the input power
increased. The difference in results when 70 and 80 MHz input power were applied to the power
MOSFET linearizer is due to the non-linear frequency-dependent power MOSFET device parameters
(total parasitic capacitances and inductance in the equivalent circuit model of the power MOSFET
Sensors 2017, 17, 764
6 of 12
devices)
shown in Figure 3. As derived in Equation (1), the frequency-dependent gain characteristic
of
thecharacteristic
power MOSFET
is dependent
on the
inductanceon
(LL2
) and
two parasitic
capacitances
of thelinearizer
power MOSFET
linearizer
is dependent
the
inductance
(LL2) and
two
(Cds
and
C
).
parasitic gd
capacitances (Cds and Cgd).
Power MOSFET linearizer
HVPA
0.06 dB at 20 dBm
0.35 dB
HVPA
1 with power MOSFET linearizer
at 26 dBm
-1
-0.265 dB at 20 dBm
-0.261 dB
at 20 dBm
-2
-3
-4
-2.95 dB at 26 dBm
-10
0
10
20
Input power (dBm)
(a)
-2.09 dB
at 26 dBm
30
Gain deviation (dB)
Gain deviation (dB)
0
power MOSFET linearizer
HVPA
0.25 dB at 20 dBm
0.64 dB
HVPA
at 26 dBm
1 with power MOSFET linearizer
2
2
0
-1
-0.33 dB at 20 dBm
-2
-3
-4
-3.00 dB at 26 dBm
-10
0
10
20
Input power (dBm)
-0.16 dB
at 20 dBm
-1.30 dB
at 26 dBm
30
(b)
Figure 6. The gain deviation graphs of the power MOSFET linearizer, HVPA, and HVPA with
Figure 6. The gain deviation graphs of the power MOSFET linearizer, HVPA, and HVPA with power
power MOSFET linearizer when (a) 70 and (b) 80 MHz input power were applied.
MOSFET linearizer when (a) 70 and (b) 80 MHz input power were applied.
Figure 7 shows the gain deviation graphs of the HVPA with and without the MOSFET
linearizer
several
DC deviation
bias voltages
at 70ofand
MHz,with
respectively.
In Figure
7a, whenlinearizer
70
Figure 7under
shows
the gain
graphs
the 80
HVPA
and without
the MOSFET
MHz
and
26
dB
m
input
powers
were
applied
to
the
power
MOSFET
linearizer,
the
gain
deviation
under several DC bias voltages at 70 and 80 MHz, respectively. In Figure 7a, when 70 MHz and 26 dBm
values
underwere
6, 9, applied
and 10 V
conditions
(−2.01,
−1.85, the
and gain
−1.80deviation
dB, respectively)
were 6, 9,
input
powers
to DC
the bias
power
MOSFET
linearizer,
values under
reduced
compared
to
1
and
3
V
DC
bias
conditions
(−2.25
and
−2.16
dB,
respectively).
In
Figure
7b, to 1
and 10 V DC bias conditions (−2.01, −1.85, and −1.80 dB, respectively) were reduced compared
when
70
MHz
and
26
dB
m input powers were applied to the power MOSFET linearizer, the gain
and 3 V DC bias conditions (−2.25 and −2.16 dB, respectively). In Figure 7b, when 70 MHz and 26 dBm
deviation under 11, 12, 13, 14, and 15 V DC bias conditions (−1.80, −1.82, −1.80, −1.83, and −1.82 dB,
input powers were applied to the power MOSFET linearizer, the gain deviation under 11, 12, 13, 14,
respectively) were still lower compared to that of the HVPA (−2.95 dB). In Figure 7a,b, when 70
and 15 V DC bias conditions (−1.80, −1.82, −1.80, −1.83, and −1.82 dB, respectively) were still lower
MHz and 26 dBm input power were applied to the HVPA with and without the power MOSFET
compared
tothe
thatgain
of the
HVPA values
(−2.95of
dB).
Figurewith
7a,b,
when
70 MOSFET
MHz andlinearizer
26 dBm input
linearizer,
deviation
theInHVPA
the
power
were power
higher were
applied
to
the
HVPA
with
and
without
the
power
MOSFET
linearizer,
the
gain
deviation
values
than that of the HVPA without the MOSFET linearizer (−2.95 dB). In Figure 7b, when 80 MHz andof the
HVPA
the powers
power MOSFET
linearizer
higher
thanlinearizer,
that of thethe
HVPA
MOSFET
26 dBwith
m input
were applied
to the were
power
MOSFET
gain without
deviationthe
values
linearizer
Figure
when 80(−1.11,
MHz and
26and
dBm−0.96
inputdB,
powers
were applied
to the power
under 6,(−
9,2.95
anddB).
10 VInDC
bias7b,
conditions
−1.01,
respectively)
were reduced
MOSFET
linearizer,
gain
deviation
under
6, 9,and
and−1.31
10 VdB,
DCrespectively).
bias conditions
(−1.11,
when compared
tothe
1 and
3V
DC bias values
conditions
(−1.37
In Figure
7d,−1.01,
when
80 dB,
MHzrespectively)
and 26 dBm input
applied
to the to
power
the gain(−1.37
and
−0.96
were powers
reducedwere
when
compared
1 andMOSFET
3 V DClinearizer,
bias conditions
deviation
under
11, 12, 13, 14,Inand
15 V 7d,
DC when
bias conditions
−1.00,
−1.00,
−1.07,were
and applied
−1.01
and
−1.31 dB,
respectively).
Figure
80 MHz were
and 26
dBm−1.06,
input
powers
to
dB,
respectively.
In
Figure
7c,d,
when
80
MHz
and
26
dB
m input power were applied to the HVPA
the power MOSFET linearizer, the gain deviation under 11, 12, 13, 14, and 15 V DC bias conditions
with
and without
power
MOSFET
linearizer,
the gain deviation
values
the HVPA
with
were
−1.00,
−1.06, −the
1.00,
−1.07,
and −1.01
dB, respectively.
In Figure
7c,d,ofwhen
80 MHz
andthe
26 dBm
power MOSFET linearizer were higher than that of the HVPA without the power MOSFET
input power were applied to the HVPA with and without the power MOSFET linearizer, the gain
linearizer (−3.00 dB). Figure 7e,f show the normalized gain deviation of the HVPA with and without
deviation values of the HVPA with the power MOSFET linearizer were higher than that of the HVPA
the MOSFET linearizer vs. the output power under 1, 3, 6, 9, and 10 V DC bias conditions, when 70
without
the power MOSFET linearizer (−3.00 dB). Figure 7e,f show the normalized gain deviation of
and 80 MHz input were applied to the power MOSFET linearizer respectively. As shown in Figure
the7,HVPA
with
and without
the MOSFET
vs. the output
power
3, 6, 9,MOSFET
and 10 V DC
the gain
deviation
performances
of thelinearizer
HVPA improved
with the
helpunder
of the 1,power
bias
conditions,
when
70
and
80
MHz
input
were
applied
to
the
power
MOSFET
linearizer
linearizer over high input power ranges. These experimental results illustrate that therespectively.
gain
Asdeviation
shown inofFigure
7, the
gain
performances
the
with
the
the HVPA
with
thedeviation
power MOSFET
linearizerof
can
beHVPA
alteredimproved
by controlling
the
DChelp
biasof the
feed voltages of the MOSFET. This fact is also meaningful because high-frequency ultrasonic
transducers requires a relatively higher input power from the HVPA to achieve reasonable echo
sensitivity compared to low-frequency ultrasonic transducers [1,4,8].
Sensors 2017, 17, 764
7 of 12
0.5
0.5
0.0
0.0
-0.5
HVPA only
1V
3V
6V
- 2.16(3V)
- 2.25 (1V)
9V
10 V
-1.0
-1.5
-2.0
-2.5
-3.0
-10
0
10
20
Input power (dBm)
- 1.8 (10V)
- 1.85(9V)
- 2.01(6V)
- 2.95
Gain deviation (dB)
Gain deviation (dB)
power MOSFET linearizer over high input power ranges. These experimental results illustrate that
the gain deviation of the HVPA with the power MOSFET linearizer can be altered by controlling the
DC bias feed voltages of the MOSFET. This fact is also meaningful because high-frequency ultrasonic
transducers requires a relatively higher input power from the HVPA to achieve reasonable echo
sensitivity
to low-frequency ultrasonic transducers [1,4,8].
Sensors 2017,compared
17, 764
7 of 12
-0.5
HVPA only
11 V
12 V
- 1.80 (11V)
13 V
- 1.82(12V)
14 V
15 V
-1.0
-1.5
-2.0
-2.5
-3.0
30
-10
(a)
-0.5
- 0.96 (10V)
HVPA only
1V
3V
6V
9V
10 V
-1.0
-1.5
-2.0
-2.5
-10
- 1.31 (3V)
- 1.37 (1V)
- 1.01(9V)
- 1.11(6V)
- 3.00
0
10
20
Input power (dBm)
Gain deviation (dB)
Gain deviation (dB)
30
(b)
0.0
-3.0
0.0
-0.5
- 1.00(13V)
HVPA only
11 V
12 V
13 V
14 V
15 V
-1.0
-1.5
-2.0
-2.5
-3.0
30
-10
(c)
- 1.00 (11V)
- 1.06(12V)
0
10
20
Input power (dBm)
- 1.01(15V)
- 1.07(14V)
- 3.00
30
(d)
0.5
0.0
-0.5
1V
HVPA only
3V
6V
9V
10 V
-1.0
-1.5
-2.0
-2.5
0
- 2.16(3V)
- 2.25 (1V)
10
20
30
Output power (dBm)
(e)
- 1.8 (10V)
- 1.85(9V)
- 2.01(6V)
- 2.95
40
Gain deviation (dB)
0.5
Gain deviation (dB)
- 2.95
0.5
0.5
-3.0
0
10
20
Input power (dBm)
- 1.80 (13V)
-1.82(15V)
- 1.83(14V)
0.0
-0.5
- 0.96 (10V)
-1.0
HVPA only
- 1.31 (3V)
1V
- 1.37 (1V)
3V
6V
9V
10 V
-1.5
-2.0
-2.5
-3.0
0
10
20
30
Output power (dBm)
- 1.01(9V)
- 1.11(6V)
- 3.00
40
(f)
Figure 7. Normalized gain deviation of the HVPA with and without a power MOSFET linearizer vs.
Figure 7. Normalized gain deviation of the HVPA with and without a power MOSFET linearizer vs.
input power at (a) 70 MHz under DC bias voltages (1, 3, 6, 9, and 10 V DC, respectively), (b) under
input power at (a) 70 MHz under DC bias voltages (1, 3, 6, 9, and 10 V DC, respectively), (b) under DC
DC bias voltages (11, 12, 13, 14, and 15 V DC, respectively), at (c) 80 MHz under DC bias voltages (1,
bias voltages (11, 12, 13, 14, and 15 V DC, respectively), at (c) 80 MHz under DC bias voltages (1, 3,
3, 6, 9, and 10 V DC, respectively), at (d) 80 MHz under DC bias voltages (11, 12, 13, 14, and 15 V DC,
6, 9, and 10 V DC, respectively), at (d) 80 MHz under DC bias voltages (11, 12, 13, 14, and 15 V DC,
respectively). Normalized gain deviation of the HVPA with and without a power MOSFET
respectively). Normalized gain deviation of the HVPA with and without a power MOSFET linearizer
linearizer vs. output power at (e) 70 MHz under DC bias voltages (1, 3, 6, 9, and 10 V DC,
vs. output power at (e) 70 MHz under DC bias voltages (1, 3, 6, 9, and 10 V DC, respectively), and (f) at
respectively), and (f) at 80 MHz under DC bias voltages (1, 3, 6, 9, and 10 V DC, respectively).
80 MHz under DC bias voltages (1, 3, 6, 9, and 10 V DC, respectively).
The input 1-dB compression point (IP1dB) provides an index of the linearity of the power
amplifier
components
[22]. In fact,
HVPA
higheran
IPindex
1dB hasof
lower
gain suppression
than
that
The input
1-dB compression
point
(IP1dBwith
) provides
the linearity
of the power
amplifier
with lower IP
1dB. In
Figure
summarizes
the measured
IP1dB
of the
HVPA
with and without
power
components
[22].
fact,8HVPA
with higher
IP1dB has
lower
gain
suppression
than thatthe
with
lower
linearizer
at
70
and
80
MHz,
respectively.
As
shown
in
Figure
8a,
when
70
MHz,
26
dB
m
IPMOSFET
.
Figure
8
summarizes
the
measured
IP
of
the
HVPA
with
and
without
the
power
MOSFET
1dB
1dB
and
3-cycle
pulsed
powers
were
sent
to
the
HVPA
with
and
without
the
power
MOSFET
linearizer,
linearizer at 70 and 80 MHz, respectively. As shown in Figure 8a, when 70 MHz, 26 dBm and 3-cycle
the measured
IP1dB of
thetoHVPA
without
the
power
MOSFET
linearizer
(22.03linearizer,
dBm) was lower
than
pulsed
powers were
sent
the HVPA
with
and
without
the power
MOSFET
the measured
that with the power MOSFET linearizer under 10V DC bias voltage (24.17 dBm). In Figure 8b, when
the 80 MHz, 26 dBm, and three-cycle pulsed powers were sent to the HVPA with and without the
power MOSFET linearizer, the measured IP1dB of the HVPA without the power MOSFET linearizer
(22.13 dBm) was lower than that with power MOSFET linearizer under 10 V DC bias voltage (26.19
dBm). Therefore, we confirm that the power MOSFET linearizer increased the IP1dB of the HVPA.
Sensors 2017, 17, 764
8 of 12
IP1dB of the HVPA without the power MOSFET linearizer (22.03 dBm ) was lower than that with the
power MOSFET linearizer under 10 V DC bias voltage (24.17 dBm ). In Figure 8b, when the 80 MHz,
26 dBm , and three-cycle pulsed powers were sent to the HVPA with and without the power MOSFET
linearizer, the measured IP1dB of the HVPA without the power MOSFET linearizer (22.13 dBm ) was
lower than that with power MOSFET linearizer under 10 V DC bias voltage (26.19 dBm ). Therefore,
we confirm
that
the power MOSFET linearizer increased the IP1dB of the HVPA.
Sensors 2017,
17, 764
8 of 12
Input 1-dB compression points
Input 1-dB compression points
0
23.88(6V)
-1
22.03(HVPA Only)
23.43 (1V)
HVPA Only 23.56 (3V)
1V
3V
6V
9V
10 V
-2
-3
0
10
20
Input power (dBm)
23.97 (9V)
24.17(10V)
30
Gain deviation (dB)
Gain deviation (dB)
0
25.89(6V)
-1
-2
-3
22.13(HVPA Only)
HVPA Only 24.43 (1V)
24.85 (3V)
1V
3V
6V
9V
10 V
0
10
20
Input power (dBm)
(a)
26.16 (9V)
26.19(10V)
30
(b)
Figure 8. Measured 1-dB compression point graphs of the HVPA with and without power MOSFET
Figure 8. Measured 1-dB compression point graphs of the HVPA with and without power MOSFET
linearizer at (a) 70 MHz and (b) 80 MHz under DC bias voltages (1, 3, 6, 9, and 10 V DC,
linearizer at (a) 70 MHz and (b) 80 MHz under DC bias voltages (1, 3, 6, 9, and 10 V DC, respectively).
respectively).
Table
1 summarizesthe
theperformances
performances ofofthe
with
andand
without
the linearizer
at 70 and
Table
1 summarizes
theHVPA
HVPA
with
without
the linearizer
at 70 and
80
MHz
under
different
DC
bias
voltages.
80 MHz under different DC bias voltages.
Table 1. Summary of the performances of the HVPA with and without the power MOSFET
Table 1. Summary of the performances of the HVPA with and without the power MOSFET linearizer
linearizer under various DC bias voltages.
under various DC bias voltages.
HVPA with the Power MOSFET Linearizer
HVPA without
(DC Bias
Voltages)
Power MOSFET
HVPA with
the Power
MOSFET Linearizer
HVPA without 1 V
Linearizer
3V
6
V
9V
10 V
(DC
Bias
Voltages)
Power
MOSFET−2.25
Gain deviation at 70 MHz (dB)
−2.95
−2.16
−2.01
−1.85
−1.80
Linearizer −1.37 1 V −1.31 3 V −1.11 6 V −1.01 9 V −0.9610 V
Gain deviation at 80 MHz (dB)
−3.00
1dB at 70 MHz
m)
22.03−2.95
23.43 −2.2523.56 −2.1623.88 −2.0123.97 −1.85 24.17−1.80
GainIPdeviation
at 70(dB
MHz
(dB)
1dB at 80 MHz
m)
22.13−3.00
24.43 −1.3724.85 −1.3125.89 −1.1126.16 −1.01 26.19−0.96
GainIPdeviation
at 80(dB
MHz
(dB)
IP1dB at 70 MHz (dBm )
22.03
at
80
MHz
(dB
)
22.13
3.2.IPHigh-Frequency
Pulse-Echo
Instrumentation
m
1dB
23.43
24.43
23.56
24.85
23.88
25.89
23.97
26.16
24.17
26.19
Figure 9 shows the measurement setup for high-frequency pulse-echo instrumentation using
3.2. the
High-Frequency
Pulse-Echo
Instrumentation
HVPA with the
power MOSFET
linearizer. The DC voltages generated by a high-voltage power
supply (E3631A, Agilent Technologies, San Jose, CA, USA) were applied to the power MOSFET
Figure 9 shows the measurement setup for high-frequency pulse-echo instrumentation using
linearizer and HVPA and the pulse signals were generated by an arbitrary signal generator
the (AFG3252C,
HVPA with Tektronix
the powerInc.,
MOSFET
linearizer.
voltages
generated
by a The
high-voltage
Beaverton,
OR, The
USA)DC
were
fed to
the HVPA.
amplifiedpower
supply
(E3631A,
Agilent
Technologies,
San
Jose,
CA,
USA)
were
applied
to
the
power
MOSFET
high-voltage signals from the HVPA are transmitted through an expander into the ultrasonic
linearizer
and
HVPA
and
the
pulse
signals
were
generated
by
an
arbitrary
signal
generator
(AFG3252C,
transducer. The acoustic echo signals reflected from a quartz target were converted to the electrical
Tektronix
Inc., Beaverton,
OR, USA)
were fedMeanwhile,
to the HVPA.
The amplified
high-voltage
signals
echo signals
by the ultrasonic
transducers.
the discharged
high-voltage
signals
sent from
HVPA
are blocked by
the limiter
before the into
electrical
echo signalstransducer.
pass throughThe
the acoustic
limiter. echo
the from
HVPA
are transmitted
through
an expander
the ultrasonic
The echo
signals
were
amplified
by a were
preamplifier
(AU-1114,
Technologies
Newby
York,
NY,
signals
reflected
from
a quartz
target
converted
to theL3
electrical
echo Inc.,
signals
the ultrasonic
USA)
with
a
DC
power
supply
(E3630A,
Agilent
Technologies)
and
displayed
the
echo
signals
andby the
transducers. Meanwhile, the discharged high-voltage signals sent from HVPA are blocked
their spectra on the oscilloscope (MSOX4154A, Keysight Technology, Santa Clara, CA, USA).
limiter before the electrical echo signals pass through the limiter. The echo signals were amplified
Figure 10 shows the measured pulse-echo responses of the HVPA with and without the power
by a preamplifier (AU-1114, L3 Technologies Inc., New York, NY, USA) with a DC power supply
MOSFET linearizer in the high-frequency pulse-echo instrumentation using a 75 MHz
(E3630A,
Agilent Technologies)
and displayed
theOlympus
echo signals
and theirTokyo,
spectra
on the
oscilloscope
immersion-focused
ultrasonic transducer
(V3349,
Inc., Shinjuku,
Japan)
when
75
(MSOX4154A,
Keysight
Technology,
Santa
Clara,
CA,
USA).
MHz three-cycle 26 dBm input power was applied to the function generated. In Figure 10a,b, the echo
Figure 10
thewith
measured
pulse-echo
responses
of the
with
the
amplitude
of shows
the HVPA
the power
MOSFET linearizer
(13.92
mV)HVPA
is higher
thanand
that without
of the
HVPA
without the
power MOSFET
linearizer (12.35 mV).
In Figure instrumentation
10c,d, the second, third,
power
MOSFET
linearizer
in the high-frequency
pulse-echo
usingfourth,
a 75 MHz
and fifth harmonic distortion components of a 75 MHz ultrasonic transducer (V3349) applied by the
HVPA with the power MOSFET linearizer (−48.34, −44.21, −48.34, and −46.56 dB, respectively) are
lower than those of the HVPA (−45.61, −41.57, −45.01, and −45.51 dB, respectively). The −6 dB
Sensors 2017, 17, 764
9 of 12
immersion-focused ultrasonic transducer (V3349, Olympus Inc., Shinjuku, Tokyo, Japan) when 75 MHz
three-cycle 26 dBm input power was applied to the function generated. In Figure 10a,b, the echo
amplitude of the HVPA with the power MOSFET linearizer (13.92 mV) is higher than that of the HVPA
without the power MOSFET linearizer (12.35 mV). In Figure 10c,d, the second, third, fourth, and fifth
harmonic distortion components of a 75 MHz ultrasonic transducer (V3349) applied by the HVPA with
theSensors
power
MOSFET
linearizer (−48.34, −44.21, −48.34, and −46.56 dB, respectively) are lower
2017,
17,
99of
12
Sensors
2017,
17,764
764
ofthan
12
those of the HVPA (−45.61, −41.57, −45.01, and −45.51 dB, respectively). The −6 dB bandwidth of
the
signal
using
HVPA
with
power
MOSFET
is
bandwidth
ofusing
the echo
echo
signalwith
using
the
HVPA
with the
the
power (36.48%)
MOSFETislinearizer
linearizer
(36.48%)
is to
thebandwidth
echo signalof
the HVPA
thethe
power
MOSFET
linearizer
increased(36.48%)
compared
increased
compared
to
that
using
the
HVPA
without
the
power
MOSFET
linearizer
(33.46%).
increased
compared
to
that
using
the
HVPA
without
the
power
MOSFET
linearizer
(33.46%).
that using the HVPA without the power MOSFET linearizer (33.46%).
Figure
The
measurement
setup
Figure
9.The
Themeasurement
measurementsetup
setup for
for the
the high
frequency
pulse-echo
instrumentation.
Figure
9.9.
the
highfrequency
frequencypulse-echo
pulse-echoinstrumentation.
instrumentation.
0.004
0.004
33Cycle
Cycle26
26dB
dBmmHVPA
HVPA
33Cycle
Cycle26
26dBm
dBmHVPA
HVPA++Power
PowerMOSFET
MOSFETLinearizer
Linearizer
0.008
0.008
12.35
12.35mV
mV
Amplitude (V)
(V)
Amplitude
Amplitude (V)
(V)
Amplitude
0.008
0.008
0.000
0.000
0.000
0.000
36.2
36.2
36.4
36.4 36.6
36.6
Time
Time(us)
(us)
36.8
36.8
37.0
37.0
-0.008
-0.008
36.0
36.0
33Cycle
Cycle26
26dB
dBmmHVPA
HVPA
00
--66dB
dBBW
BW==33
33.46
.46%
%
HD3
HD3==
HD2
HD2== --41.57
HD4==
41.57dB
dB HD4
-20
-20
--45.61
45.61dB
dB
--45.01
45.01dB
dB
100
200
300
400
100
200
300
400
Frequency
Frequency(MHz)
(MHz)
(c)
(c)
36.4
36.4 36.6
36.6
Time
Time(us)
(us)
36.8
36.8
37.0
37.0
33Cycle
Cycle26
26dBm
dBmHVPA
HVPA++Power
PowerMOSFET
MOSFETLinearizer
Linearizer
00
-20
-20
HD5
HD5==
--45.51
45.51dB
dB
-40
-40
36.2
36.2
(b)
(b)
Normalized Spectrum
Spectrum (dB)
(dB)
Normalized
Normalized Spectrum
Spectrum (dB)
(dB)
Normalized
(a)
(a)
-60
-60
00
13.92
13.92mV
mV
-0.004
-0.004
-0.004
-0.004
-0.008
-0.008
36.0
36.0
0.004
0.004
500
500
-40
-40
-60
-60
00
--66dB
dBBW
BW==36.48
36.48%
%
HD3
HD3==
--44.21
44.21dB
dB
HD2
HD2==
HD5
HD5==
HD4
HD4==
--48.34
dB
48.34 dB - 48.34 dB --46.56
46.56dB
dB
- 48.34 dB
100
100 200
200 300
300 400
400
Frequency
Frequency(MHz)
(MHz)
500
500
(d)
(d)
Figure
10. The
echo
amplitude
of
the
HVPA
(a)
without
(b) with
the power
MOSFET linearizer
Figure
The
echo
amplitudeof
ofthe
theHVPA
HVPA(a)
(a) without
without and
and
linearizer
Figure
10.10.
The
echo
amplitude
and (b)
(b) with
withthe
thepower
powerMOSFET
MOSFET
linearizer
and
and the
the spectrum
spectrum of
of the
the HVPA
HVPA (c)
(c) without
without and
and (d)
(d) with
with the
the power
power MOSFET
MOSFET linearizer
linearizer when
when
and the spectrum of the HVPA (c) without and (d) with the power MOSFET linearizer when three-cycle
three-cycle
three-cycle 26
26dB
dBmm input
input power
power was
was applied.
applied.
26 dBm input power was applied.
Figure
Figure 11
11 shows
shows the
the measured
measured pulse-echo
pulse-echo responses
responses of
of the
the HVPA
HVPA with
with and
and without
without the
the power
power
MOSFET
MOSFET linearizer
linearizer using
using aa 75
75 MHz
MHz immersion-focused
immersion-focused ultrasonic
ultrasonic transducer
transducer (V3349)
(V3349) when
when 75
75 MHz
MHz
five-cycle
20
dB
m
input
power
was
applied.
In
Figure
11a,b,
the
echo
amplitude
of
the
HVPA
five-cycle 20 dBm input power was applied. In Figure 11a,b, the echo amplitude of the HVPA with
with
the
the power
power MOSFET
MOSFET linearizer
linearizer (10.22
(10.22 mV)
mV) is
is higher
higher than
than that
that of
of the
the HVPA
HVPA (9.95
(9.95 mV).
mV). In
In Figure
Figure 11c,d,
11c,d,
the
the second,
second, third,
third, fourth,
fourth, and
and fifth
fifth harmonic
harmonic distortions
distortions of
of the
the transducer
transducer applied
applied by
by the
the HVPA
HVPA with
with
the
the power
power MOSFET
MOSFET linearizer
linearizer (−41.54,
(−41.54, −41.80,
−41.80, −48.86,
−48.86, and
and −46.27
−46.27 dB,
dB, respectively)
respectively) are
are lower
lower than
than
Sensors 2017, 17, 764
10 of 12
Figure 11 shows the measured pulse-echo responses of the HVPA with and without the power
MOSFET linearizer using a 75 MHz immersion-focused ultrasonic transducer (V3349) when 75 MHz
five-cycle 20 dBm input power was applied. In Figure 11a,b, the echo amplitude of the HVPA with
the power MOSFET linearizer (10.22 mV) is higher than that of the HVPA (9.95 mV). In Figure 11c,d,
the second, third, fourth, and fifth harmonic distortions of the transducer applied by the HVPA with
the power
MOSFET
linearizer (−41.54, −41.80, −48.86, and −46.27 dB, respectively) are lower
Sensors
2017, 17,
764
10than
of 12
those of the HVPA without the power MOSFET linearizer (−25.85, −43.56, −49.04, and −49.24 dB,
respectively).
respectively). The
The −6
−6dB
dBbandwidth
bandwidthofofthe
theecho
echosignal
signalusing
usingthe
theHVPA
HVPA with
with the
the power
power MOSFET
linearizer (30.01%)
(30.01%)increased
increased
compared
to that
the HVPA
theMOSFET
power MOSFET
compared
to that
usingusing
the HVPA
withoutwithout
the power
linearizer
linearizer
in the measurement
of the pulse-echo
power
(17.35%). (17.35%).
As shownAs
in shown
the measurement
results of results
the pulse-echo
responses,responses,
the powerthe
MOSFET
MOSFET
could
the harmonic
unwanteddistortion
harmonic
distortion and
components
andamplitudes
increase the
linearizer linearizer
could reduce
the reduce
unwanted
components
increase the
of
amplitudes
of the
echo signal
generated
by 75 MHz
ultrasonic transducers.
the echo signal
generated
by 75
MHz ultrasonic
transducers.
0.008
5 Cycle 20 dBm HVPA + Power MOSFET Linearizer
5 Cycle 20 dBm HVPA
0.008
Amplitude (V)
Amplitude (V)
0.004
9.95 mV
0.000
0.004
0.000
-0.004
-0.004
-0.008
36.0
10.22 mV
36.2
36.4 36.6
Time (us)
36.8
37.0
-0.008
36.0
36.2
(a)
- 49.04 dB
HD5 =
- 49.24 dB
-40
100
200
300
Frequency (MHz)
400
Normalized Spectrum (dB)
Normalized Spectrum (dB)
- 6 dB BW =17.35 %
HD2 =
- 25.85 dB HD3 =
- 43.56 dB HD4 =
0
37.0
(b)
0
-60
36.8
5 Cycle 20 dBm HVPA + Power MOSFET Linearizer
5 Cycle 20 dBm HVPA
-20
36.4 36.6
Time (us)
500
(c)
0
- 6 dB BW =30.01 %
-10
HD3 =
HD2 = - 41.80 dB
- 41.54 dB
HD4 =
-20
-30
HD5 =
- 48.86 dB - 46.27 dB
-40
-50
-60
0
100
200
300
400
Frequency (MHz)
500
(d)
Figure
11. The
Figure 11.
The echo
echo signal
signal amplitude
amplitude when
when using
using the
the HVPA
HVPA (a)
(a) without
without and
and (b)
(b) with
with the
the power
power
MOSFET
MOSFET linearizer
linearizer and
and the
the spectrum
spectrum of
of the
the HVPA
HVPA (c)
(c) without
without and
and (d)
(d) with
with the
the power
power MOSFET
MOSFET
linearizer
linearizer when
when five-cycle
five-cycle 20
20 dB
dBm input
input power
power is
is applied.
applied.
m
Finally, while circuit topologies and operation frequencies are different, making direct
Finally, while circuit topologies and operation frequencies are different, making direct comparison
comparison difficult, Table 2 summarize the performances of our linearizer with that of selected
difficult, Table 2 summarize the performances of our linearizer with that of selected previously
previously reported linearizers.
reported linearizers.
Table 2. Performance comparison of our linearizer and a previously-reported design, including
pulse-echo performances with ultrasound transducers.
Parameters
Output power
Operating frequency
Second harmonic distortion (HD2)
Third harmonic distortion (HD3)
[13]
55.1 dBm
5 MHz
−50 dB
-
[14]
41.59 dBm
140 MHz
−24.8 dB
-
Our Design
37.6 dBm
80 MHz
−41. 54 dB
−41.80 dB
4. Conclusions
Conventional HVPA used in the high-frequency pulse-echo instrumentation includes DC
coupling capacitors to minimize the high-voltage DC effects on the high-frequency ultrasound
Sensors 2017, 17, 764
11 of 12
Table 2. Performance comparison of our linearizer and a previously-reported design, including
pulse-echo performances with ultrasound transducers.
Parameters
[13]
[14]
Our Design
Output power
Operating frequency
Second harmonic distortion (HD2)
Third harmonic distortion (HD3)
55.1 dBm
5 MHz
−50 dB
-
41.59 dBm
140 MHz
−24.8 dB
-
37.6 dBm
80 MHz
−41. 54 dB
−41.80 dB
4. Conclusions
Conventional HVPA used in the high-frequency pulse-echo instrumentation includes DC
coupling capacitors to minimize the high-voltage DC effects on the high-frequency ultrasound
transducer. However, this structure could reduce the gain of the HVPA with higher input powers.
Harmonic distortions of the echo signals can severely limit the performance of high-frequency
pulse-echo instrumentation. Therefore, a linear HVPA to trigger the ultrasonic transducer is highly
desirable for such high-frequency pulse-echo instrumentation. Here, we proposed a novel power
MOSFET linearizer in order to increase the linearity (reducing gain deviations) over a wide range of
input power. The gain deviation of the power MOSFET linearizer has shown to have the opposite effect
to that of the HVPA, thus increasing gain flatness over a wide input power range. Therefore, the power
MOSFET linearizer can decrease the harmonic distortions of the HVPA, thereby reducing the unwanted
harmonic distortions of the ultrasonic transducer in high-frequency pulse-echo instrumentation.
The structure of the power MOSFET linearizer is simple to implement because, in the power
MOSFET linearizer structure, the MOSFET acts as a variable resistor to primarily affect the gain
deviation performances. Therefore, the operation of the MOSFET under varying DC bias conditions in
the power MOSFET linearizer was characterized for the HVPA. At 70 and 80 MHz, the measured gain
deviation values of the HVPA with power MOSFET linearizer under 10 V DC bias voltage (−2.95 dB
and −3.00 dB, respectively) were lower than that of the HVPA without power MOSFET linearizer
(−1.80 dB and −1.11 dB, respectively). At 70 and 80 MHz, the measured IP1dB point of the HVPA
with power MOSFET linearizer under 10 V DC bias voltage (24.17 and 26.19 dBm , respectively) were
higher than that of the HVPA without power MOSFET linearizer (22.03 and 22.13 dBm , respectively).
In order to verify the capability of the power MOSFET linearizer, the echo signal harmonic distortions
of the 75 MHz ultrasonic transducer were measured. Using 75 MHz 5-cycle 20 dBm input power,
the echo signal harmonic distortions using HVPA with power MOSFET linearizer (−41.54, −41.80,
−48.86, and −46.27 dB, respectively) were measured to be lower than those of HVPA without the
power MOSFET linearizer (−25.85, −43.56, −49.04, and −49.24 dB, respectively). We conclude that the
power MOSFET linearizer has proven to be a simple, yet effective, technique for the HVPA to reduce
unwanted harmonic distortions in the echo signals while providing higher echo signal amplitude and
even wider echo signal bandwidth.
Acknowledgments: This work has been supported in part by National Research Foundation of Korea
(NRF-2016M2B2A9A02945226) and by the MSIP (Ministry of Science, ICT and Future Planning), Korea, under the
ITRC (Information Technology Research Center) support program (IITP-2016-R2718-16-0015) supervised by the
IITP (National It Industry Promotion Agency).
Author Contributions: H.C., J.-Y.Y. and C.Y. conceived the idea. J.-Y.Y. and P.C.W. performed the experiments;
H.C., J.-Y.Y. and C.Y. wrote the paper together.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Shung, K.K. Diagnostic Ultrasound: Imaging and Blood Flow Measurements; Taylor & Francis: Boca Raton, FL,
USA, 2015.
Sensors 2017, 17, 764
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
12 of 12
Qiu, W.; Yu, Y.; Tsang, F.K.; Sun, L. A multifunctional, reconfigurable pulse generator for high-frequency
ultrasound imaging. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 2012, 59, 1558–1567. [PubMed]
Qiu, W.; Wang, C.; Li, Y.; Zhou, J.; Yang, G.; Xiao, Y.; Feng, G.; Jin, Q.; Mu, P.; Qian, M.; et al. A scanning-mode
2D shear wave imaging (s2D-SWI) system for ultrasound elastography. Ultrasonics 2015, 62, 89–96. [CrossRef]
[PubMed]
Szabo, T.L. Diagnostic Ultrasound Imaging: Inside Out; Elsevier Academic Press: London, UK, 2004.
Song, J.; Kim, S.; Sohn, H.-Y.; Song, T.-K.; Yoo, Y.M. Coded excitation for ultrasound tissue harmonic imaging.
Ultrasonics 2010, 50, 613–619. [CrossRef] [PubMed]
Zhu, B.; Xu, J.; Li, Y.; Wang, T.; Xiong, K.; Lee, C.; Yang, X.; Shiiba, M.; Takeuchi, S.; Zhou, Q. Micro-particle
manipulation by single beam acoustic tweezers based on hydrothermal PZT thick film. AIP Adv. 2016, 6,
035102. [CrossRef] [PubMed]
Zhu, X.; Guo, J.; He, C.; Geng, H.; Yu, G.; Li, J.; Zheng, H.; Ji, X.; Yan, F. Ultrasound triggered image-guided
drug delivery to inhibit vascular reconstruction via paclitaxel-loaded microbubbles. Sci. Rep. 2016, 6, 21683.
[CrossRef] [PubMed]
Zhou, Q.; Lam, K.H.; Zheng, H.; Qiu, W.; Shung, K.K. Piezoelectric single crystal ultrasonic transducers for
biomedical applications. Prog. Mater. Sci. 2014, 66, 87–111. [CrossRef] [PubMed]
Vuolevi, J.; Rahkonen, T. Distortion in RF Power Amplifiers; Artech House: London, UK, 2003.
Kazimierczuk, M.K. RF Power Amplifier; John Wiley & Sons: Hoboken, NJ, USA, 2014.
Cripps, S.C. Advanced Techniques in RF Power Amplifier Design; Artech House: Norwood, MA, USA, 2002.
Lee, T.H. The Design of CMOS Radio-Frequency Integrated Circuits; Cambridge University Press: Cambridge,
UK, 2004.
Gao, Z.; Gui, P.; Jordanger, R. An integrated high-voltage low-distortion current-feedback linear power
amplifier for ultrasound transmitters using digital predistortion and dynamic current biasing techniques.
IEEE Trans. Circuits Syst. II Express Briefs 2014, 61, 373–377. [CrossRef]
Choi, H.; Jung, H.; Shung, K.K. Power Amplifier Linearizer for High Frequency Medical Ultrasound
Applications. J. Med. Biol. Eng. 2015, 35, 226–235. [CrossRef] [PubMed]
Park, J.; Hu, C.; Li, X.; Zhou, Q.; Shung, K.K. Wideband linear power amplifier for high-frequency ultrasonic
coded excitation imaging. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 2012, 59, 825–832. [CrossRef]
[PubMed]
Svilainis, L.; Motiejunas, G. Power amplifier for ultrasonic transducer excitation. Ultragarsas 2006, 1, 30–36.
Razavi, B. Design of Analog CMOS Integrated Circuits, 1st ed.; McGraw-Hill Science: New York, NY, USA, 2000.
Shung, K.K. Diagnostic Ultrasound: Past, Present, and Future. J. Med. Biol. Eng. 2011, 31, 371–374. [CrossRef]
Li, X.; Wei, W.; Zhou, Q.; Shung, K.K.; Chen, Z. Intravascular photoacoustic imaging at 35 and 80 MHz.
J. Biomed. Opt. 2012, 17, 106005. [CrossRef] [PubMed]
Li, X.; Li, J.; Jing, J.; Ma, T.; Liang, S.; Zhang, J.; Mohar, D.; Raney, A.; Mahon, S.; Brenner, M. Integrated
IVUS-OCT imaging for atherosclerotic plaque characterization. IEEE J. Sel. Top. Quantum Electron. 2014, 20,
196–203.
Qiu, W.; Wang, X.; Chen, Y.; Fu, Q.; Su, M.; Zhang, L.; Xia, J.; Dai, J.; Zhang, Y.; Zheng, H. A Modulated
Excitation Imaging System for Intravascular Ultrasound. IEEE Trans. Biomed. Eng. 2016. [CrossRef]
[PubMed]
Katz, A. Linearization: Reducing distortion in power amplifiers. IEEE Microw. Mag. 2001, 2, 37–49. [CrossRef]
© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).