Practical Papers, Articles
and Application Notes
Kye Yak See, Technical Editor
T
ime flies, this is the fourth issue of the magazine for year
2012. I hope the revamped magazine design and interesting
articles of the past three issues have enhanced your read-
ing pleasure.
In this issue, Hongyu Li, Victor Khilkevich, Tianqi Li and David Pomme-
renke from Missouri University of Science and Technology; and Seong-
tae Kwon and Wesley Hackenberger from TRS Technologies share
their experience in ESD protection for electronic products. Their paper,
“Nonlinear Capacitors for ESD Protection”, investigates the nonlinear
behavior of capacitance versus applied voltage for ESD protection
capacitors. The comparison between the conventional multi-layer
ceramic capacitor (MLCC) and the proposed anti-ferroelectric (AFE)
capacitor shows that the capacitance of the AFE capacitor increases
with increasing applied voltage, which is a desirable characteristic for
ESD protection effectiveness.
The next paper, “Examination of Contaminant-Induced Faults in Con-
nectors”, authored by Hua Zeng from Hitachi Automotive and Todd
Hubing from Clemson University, studies the effects of salt water expo-
sure on cable connector impedances. Several tests are performed to
explore the shunting resistance across the pins of wiring harness con-
nectors. The test results show that salt-induced corrosion and moisture
may cause intermittent shunting resistances capable of affecting the
normal operation of various systems. It is also found that the shunting
Nonlinear Capacitors for ESD Protection
Hongyu Li, Victor Khilkevich, Tianqi Li, David Pommerenke, Missouri University of Science and Technology, USA;
hlfm4@mst.edu, khilkevichv@mst.edu, tlx6f@mail.mst.edu, davidjp@mst.edu and
Seongtae Kwon, Wesley Hackenberger, TRS Technologies, Inc., USA
seongtae@trstechnologies.com, wes@trstechnologies.com
Abstract—In order to protect electronic products from Electro-
static Discharge (ESD) damage, multi-layer ceramic capacitors
(MLCC) are often used to bypass the transient ESD energy. Most
dielectric materials used in MLCC are nonlinear, since the dielec-
tric constant decreases with increasing voltage, reducing the
capacitance value, thus degrading the ESD protection effect.
Using a large initial capacitance value will ensure sufficient ESD
protection; however, the shunt capacitors also limit the signal
bandwidth of the ESD-protected data channel, thus setting a maxi-
mal capacitance value at data voltage levels. This paper investi-
gates the nonlinearity of capacitors and suggests improved trade-
off between ESD protection and data bandwidth by using the Anti-
ferroelectric (AFE) capacitors as ESD protection. The dielectric
constant of AFE material increases with increasing voltage. The
voltage dependence of X7R and AFE capacitors are measured
38 ©2012 IEEE Electromagnetic Compatibility Magazine – Volume 1 – Quarter 4
resistances are non-linear with applied voltage, which will have direct
impact on ESD or radiated RF susceptibility of products that may be
exposed to a salt-water environment.
The last paper, “Test Methods for RF-Based Electronic Safety
Equipment: Part 1 – From Field Tests to Performance Metrics”, is
a two-part series of the RF-based electronic safety equipment
test methods contributed by Kate A. Remley and William F. Young
from the National Institute of Standards and Technology. The
two-part series describes in detail the development of free-field
test methods for wireless electronic safety equipment that repli-
cate field-test conditions in a laboratory environment. The test
methods are developed to support the National Fire Protection
Association (NFPA) in the revision of NFPA 1982: Standard on
Personal Alert Safety Systems (PASS), but will be applicable to
other types of RF-based equipment as well. In Part 1, they illus-
trate methods for extracting performance metrics from a series
of field tests conducted by NIST researchers. In Part 2, that will
appear in the next issue, they will replicate the key field test
conditions in the laboratory and verify device performance
under those conditions.
If you have any comments or suggestions on further improving
this column, please do not hesitate to write to me through email at
ekysee@ntu.edu.sg. I wish all readers a Happy New Year!
using static and nanosecond transient measurements. The ESD
protection effectiveness with different material capacitors are
compared by simulation. Due to very limited availability of suitable
AFE material samples only hand-made capacitors have tested
without investigating the long term stability of the material.
Index Terms— ESD protection, AFE material, nonlinear capacitor
I. Introduction
ESD is one of the most important reliability problems in an electronic
product. In order to provide ESD protection to electronic products,
decoupling capacitors and series resistors can be used as shown in
Fig. 1. The capacitors absorb the injected charge, and limit the maxi-
 While the world benefits from what’s new,
IEEE can focus you on what’s next.

Develop for tomorrow with

today’s most-cited research.
Over 3 million full-text technical documents
can power your R&D and speed time to market.

• IEEE Journals and Conference Proceedings

• IEEE Standards
• IEEE-Wiley eBooks Library
• IEEE eLearning Library
• Plus content from select publishing partners

IEEE Xplore® Digital Library

Discover a smarter research experience.

Request a Free Trial

www.ieee.org/tryieeexplore

Follow IEEE Xplore on

11-PIM-0544a_Xplore_WhatsNext_7x10_FINAL.indd 1 12/16/11 9:08 AM

 ectric
everal
s and R2 1MΩ
hofthe Fig. 3. Static measurement block diagram rise tt
rise
con.thin C2 R1 1MΩ (TLP
(TLP
how as C1 is the capacitor under test, i.e. the DUT. V1 is the DC micro
micro
S)
ousand voltage power supply which provides the DC bias. Resistors R1 is em
is em
ccess and R2LCR or isolate the
(1 MΩ) C1 impedance of the capacitor from the
V1
induc
induc
ied impedance VNAof the power DUT supply. C2 blocks DC to protect the meas
meas
AFE
ric instrument. Its capacitance should not be voltage dependent. trans
transm
terial
ral Two 8kV 20nF paper in oil (PIO) capacitors comprise C2 meas
meas
capac
endtwo placed in parallel. Tests showed R2 1MΩ
no noticeable voltage capac
ectric
the Fig.dependence. The measurement
3. Static measurement block diagram fixture is shown in Fig. 4. The
Fig. 3. Static measurement block diagram
ength
hin two cylindrical devices are the PIO capacitors.
as C1 is the capacitor under test, i.e. the DUT. V1 is the DC
TRS
nd voltage power supply which provides the DC bias. Resistors R1
plated
ess and R2 (1 MΩ) isolate the impedance of the capacitor from the
. 2. A impedance of the power supply. C2 blocks DC to protect the T
TL
nFEtwo instrument. Its capacitance should not be voltage dependent.
AFE
ial Two 8kV 20nF paper in oil (PIO) capacitors comprise C2 Fig. Fig. 5.
Fig. 5. Static
5. Static measurement
Static measurement results
measurementresults forfor
results
for four X7R
four
four capacitors
X7R
X7R from
capacitors
capacitors different
from
from different
wo vendors
placed in parallel. Tests showed no noticeable voltage different vendors vendors
ric dependence. The measurement fixture is shown in Fig. 4. The Fig. 7.
7.
Fig.
gth two cylindrical devices are the PIO capacitors.
Th
Th
RS [9]
[9]
ted
A Q
Q C
C
wo and
Fig.
Fig. 4.
4. Test
Test fixture
fixture for
for the
the static
staticmeasurement
measurementusing
usingnetwork
networkanalyser or or
LCR and
FE meter
analyser
LCR meter II 
CC

For
For
B. Static
showed Measurement
no noticeable Results
voltage for X7RThe
dependence. Capacitors
measurement
fixture beco
becom
Static measurement results for X7R capacitors are
is shown in Fig. 4. The two cylindrical devices are the PIO in
shown
capacitors.
Fig. 5. All four capacitors (4.7nF, 50V) showed similar Q C
C
Q
behaviour. The capacitance decreases from 4.7nF to about 1nF
and
and
at 400V DC bias. We observed no damage to the capacitors at
B. Static Measurement Results for X7R Capacitors
400V.
ic is Fig. 4. Test fixture for the static measurement using network analyser or LCR II  C
C
Static measurement results for X7R capacitors are shown in Fig.
ferent meter Fig. 6.
6. Static
Static measurement
measurement results
results for
for the
the AFE
AFE material
material capacitor
capacitor So th
th
Fig. So
were 5.C.
All four capacitors
Static Measurement(4.7nF, Results
50V) showed similar
for AFE behaviour. The
Capacitor Fig. 6. Static measurement results for the AFE material capacitor
capacitance
B. TheStatic
decreases
measurement
Measurement
fromfor
result 4.7nF to
Resultsanfor AFE about 1nF at 400V
X7Rcapacitor
Capacitors
DC bias.
is shown in Fig. CV
C V 
We observed no damage to the capacitors at 400V. IV. T TRANSIENT MEASUREMENT
Static measurement results for X7R capacitors are shown inand the capacitor is IV.
6. The capacitance is about 3nF without any DC bias, RANSIENT M EASUREMENT
limited by the capacitance value and the source
Fig.increases
5. All four with capacitors
increasing (4.7nF,DC bias 50V)voltage
showedbeforesimilarthe impedance
The nonlinearity
The nonlinearity of of the
(about 300 Ohm
the capacitors
capacitors can can be
for an IEC 61000-4-2
be seen
seen clearly
clearly from
ESD generator
from
at The
The
transformation from the AFE phase to the FE phase. For this the static measurement results. However, ESD is a transient volta
t, and behaviour.
C. StaticThe capacitanceResults
Measurement decreases for from 4.7nF to about 1nF the
AFE Capacitor lower static measurement
frequencies and around results.
100 However,
Ohm at higher ESD is a transient
frequencies). volta
AFE sample, the transformation occurs at about 325V, causing process of of nanosecond time time scale.
scale. ESD
ESD currents
currents typically
typically have
have curre
h DC at 400V DC bias. We observed no damage to the capacitors at process The voltagenanosecond
and current rise time at a capacitor with hundreds curre
a peak in its capacitance value of about 8.5nF at
The measurement result for an AFE capacitor is shown in Fig. 6. rise
325V. In the rise time
pF ortime
of less
ofnF
less than
is than
1ns. However,
1ns. However,
between a the
the voltage
fewvoltage
and
and current
of current
rise
rise noise
The 400V. a few generally and tens nanosec- noise
is FE phase the capacitor acts as normal capacitor, and the times
times at
at the
the capacitor
capacitor is
is limited
limited by
by the
the capacitance
capacitance value
value and
and the
the of sco
of sc
The capacitance is about 3nF without any DC bias, and increas- onds. Transient measurements investigate if the capacitance can
ent capacitance decreases with increasing DC
es with increasing DC bias voltage before the transformation bias voltage. The source
source impedance
impedance (about
(about 300
300 Ohm
Ohm for
for
react with sufficient speed to provide ESD protection. an
an IEC
IEC 61000-4-2
61000-4-2 ESD
ESD deriv
deriv
ere C. Static Measurement
capacitance Results for
at AFE Capacitor generator at at lower
lower frequencies
frequencies and and around
around 100 100 Ohm Ohm at at higher
higher
from the AFEdropsphasebackto thetoFE3.5nF
phase. 500V.
For this AFE sample, the generator
The measurement
transformation result
occurs at for an 325V,
about AFE capacitor
causing a is shown
peak frequencies). The voltage and current rise
in itsin Fig. frequencies). The voltage and current rise time at a capacitor time at a capacitor B.
with hundreds pF oror aa few
few nF nF is generally
generally between
between aa few few and
and B.
6. capacitance
The capacitance value is about 8.5nF
of about 3nF at without
325V. Inany DCphase
the FE bias,theand with A. hundreds
Transient pF
Measurement is
Method
tens of nanoseconds. Transient measurements investigate if the If t
If
increases
capacitorwithacts asincreasing DC bias
normal capacitor, voltage
and the before the tens of nanoseconds. Transient measurements investigate if the
capacitance to ch
transformation
decreases with from the AFEDCphase
increasing to the FE
bias voltage. Thephase. capacitance
For this capacitance
capacitance In the measurementcan react
can react with
setupwith sufficient
sufficient
illustrated speed
in Fig.speed to provide
to providethe
7, C represents ESD
ESD to ch
nd
AFE sample,
drops back the transformation
to 3.5nF at 500V. occurs at about 325V, causing protection. protection.
capacitor under test. A pulse with duration of 70ns and a rise time was rr
was
DC Resu
Resu
a peak in its capacitance value of about 8.5nF at 325V. In the of 150ps generated by the Transmission Line Pulser (TLP) is inject- the ca
ca
The the
FE phase the capacitor acts as normal capacitor, and the edA.intoTransient the capacitor under test through
Measurement Methoda microstrip transmission 9. Th
A. Transient Measurement Method 9. Th
IV. Transient
capacitance Measurement
decreases with increasing DC bias voltage. The line.InAthe loop formed by a trace and vias is embedded underneath the tr
tr
measurement setup setup illustrated
illustrated in in Fig.
Fig. 7,7, CC represents
represents the
capacitance drops back to 3.5nF at 500V. theIntransmission
the measurementline. The mutual inductance between the trans-
the capacitor under test. A pulse with
the capacitor under test. A pulse with duration of 70ns andduration of 70ns and aa
The nonlinearity of the capacitors can be seen clearly from the mission line and loop is used to measure the derivative of the cur-
static measurement results. However, ESD is a transient process rent flowing on the transmission line. The current is obtained by
of nanosecond time scale. ESD currents typically have rise time integrating the measured derivative of the current. The voltage
of less than 1ns. However, the voltage and current rise times at across the capacitor is measured with an oscilloscope.

©2012 IEEE Electromagnetic Compatibility Magazine – Volume 1 – Quarter 4 41

 inductance
measure the between the transmission
derivative ofcurrent. line andflowing
the current loop is used
on theto
the
measured
measured
transmission
measured derivative
derivative
line. of of
The
derivative the
of the
current.
current
the TheThe
is obtained
current. The voltage
voltage
by across
across thethe
integrating
voltage across the
measure the
transmission derivative
line. Thewith of
current the current flowing on
is obtained by integrating the
capacitor
capacitor
measured
capacitor isderivative
measured
is measured with
of thean
an oscilloscope.
oscilloscope.
current. The voltage across the
measured isderivative
transmission measured
line. The with
of theancurrent.
current oscilloscope.
is obtained by integrating
The voltage across the
capacitor isderivative
measured measured of with ancurrent.
the oscilloscope.
The voltage across the
capacitor is measured with an oscilloscope.
HH
capacitor is measuredH with an oscilloscope.
H
H Loop
H Loop
Loop
Loop
Loop CC
C
TLP
TLP Loop Scope
TLP Scope
Scope
Scope C Scope
Scope
TLP C Scope
mifferent
different TLP Scope C Scope
m different TLP Scope Scope
m different Scope
m different
Fig.Fig.
m different Fig. 7.
Fig.
7. Block
7. Block
7. Block
Block
diagram
diagram for
diagram
diagram
for transient
fortransient
for transient
transient
measurement
measurement setup
setup
measurement
measurement setup
setup
Fig. 7. Block diagram for transient measurement setup
Fig.The charge
7. Block and current
diagram formeasurement
for transient a linear capacitor
setup are described by
TheThe charge
The
charge
Fig. 7. and
charge current
and
anddiagram
Block current for
for afor
current a linear
for
linear
transient capacitor
acapacitor
linear
measurement areare
capacitor described
are described
setup by [9]by by
described
[9][9]
[9]The charge and current for a linear capacitor are described by
The charge and current for a linear capacitor are described by
[9]
QCThe C ,chargeV, and current for a linear capacitor are described (1) (1)(1) by
Q [9]Q  V C V , (1)
[9] (1)
Q  C V ,
andand
and Q  C V ,
and (1)
Q  C V , (1)
and
IC C
I andI  dV
 dV
C  dV t  dt
t  .dt .
t  dt . (2)
(2)(2)
(2)Fig. 8. Measured voltage cross an X7R capacitor. The peak around 0 ns is
and Fig.
Fig. 8.
caused8. Measured
Measured voltage
by the effectivevoltage cross
cross an
inductance anX7Rof thecapacitor.
X7R capacitorThe
capacitor. andpeak
The peak around
around00ns
its mounting. nsis
This
I
For  C a  dV
I a Cnonlinear
 t
nonlinear  dt .
t  dt .time-independent
time-independent capacitor the equations (2)caused
(2) by
inductance
is caused the
byis effective
about
the 1.3 nH.
effective inductance
inductance of the
of thecapacitor
capacitor and its
and mounting.
its mounting. This
ForFor a dV nonlinear time-independent capacitor
capacitor thethe equations
equations
For become C  dV t time-independent
Ia nonlinear dt . capacitorcapacitorthe equations (2)inductance is aboutis1.3
This inductance nH. 1.3 nH.
about
become
For
become a nonlinear time-independent the equations
For a nonlinear time-independent capacitor the equations
become
become
QCVCaV
For dV
nonlinear time-independent capacitor the equations
Q become
Q  C VdV ,dV,, (3)(3)
(3)
become
Q  C  V
Q  C V   dV ,
  dV , (3) (3)
andand
and (3)
Q  C V   dV , (3)
CV dV t  dt tt .dt
and
IICVC
and
I and VdV dV dt.. (4)(4)
(4)
and
I  C
C V
V   dV  t 
  dV capacitance dt .
t  capacitance
(4)
So So I the
the
So nonlinear
nonlinear dt . cancan be bebe calculated
calculated by by (4)
I the C Vnonlinear
  dV t  dt capacitance
. can calculated by (4) (4)
So the nonlinear I II
capacitance can be calculated by
CV
So
C VSo  nonlinear
the nonlinear . capacitance can be calculated by
C Vthe dt .
Itt dt
So the nonlinear capacitance
. capacitance cancan be calculated
be calculated by by
dV dV 
tdV
C V   I dt . (5)(5)
(5)
C V   dV It  dt .
arly
yarlyfromfrom
from TheThe The
C V  
measurementsdV
measurements t  dt . capture the current derivative (5) and (5)
the
transient dV t  dt capture
measurements capture thethe current
current derivative
derivative andand thethe
(5)
nsient
arly from voltage
transient voltage
The
voltage acrossacross
measurements
across the the
the capacitor.
capacitor.
capture
capacitor. The The
the
Thepost post
post processing
processing
current derivative
processing obtains
obtains and
obtains the (5)
the
the
arly havefrom Thecurrent
yally
have
transient
ally
arly from
The
have current
voltage
current andmeasurements
measurements andthethe
across
and voltage
the
capture
voltage
the
voltage
capture
the current
derivative.
derivative.
capacitor.
derivative.
the
Low
The
current
derivative
Low pass
post
Low pass derivative
and the
filtering
filtering
processing
pass filtering
and
volt-
is used
is obtains
used
is and
the
for
forthe
used for
transient
rrent rise age The
voltage
across
noise measurements
across
the
suppression capacitor.the capture
capacitor.
The post the
The current
post
processing derivative
processing
obtains the obtains
cur- the
the
nt
ally
urrent rise
have
transientrise noisenoise suppression
current
voltage and
suppression
across the andand
voltage
the and de-trending
de-trending
derivative.
de-trending
capacitor. The
is
is Low
used
is
post
used
usedto
pass to
remove
to remove
filtering
processing remove the isthe
used
the
obtains
effect
effect for
effect
the
ally
e and have
the current
rentof and theand the voltage derivative. Low pass filtering is used for
of scope’s imperfect andDC-offset on on the integration of
of the current
nd the voltage derivative. Low pass filtering is ofused for
uerrent
ally
urrentand the of scope’s
rise
have
rise noise scope’s
current imperfect
suppression
and imperfect
the and DC-offset
voltage de-trending
DC-offset
derivative. the
on integration
is
the
Low used tofiltering
integration
pass remove the current
the
the
is effect
current
used for
4-2 ESD noisenoise
derivative. suppression
suppression and de-trending
de-trending is usedis used
to to
remove remove
the the
effect effect
2-4-2
e ESD
andESD
rrent the derivative.
rise of scope’s
derivative.
noise suppression imperfect andDC-offset
de-trending on the integration
is used toofremove of the thecurrent
effectFig. 9. Capacitance versus time for X7R capacitors from different vendors
ueat and
higherthe of scope’s
of scope’s imperfect imperfect DC-offset
DC-offset on theon the
integrationintegration the of the
current current Fig. 9. Capacitance versus time for X7R capacitors from different vendors
higher
4-2
at
e ESD
higher
and the derivative.
of scope’s imperfect DC-offset on the integration of the currentFig. 9. Capacitance versus time for X7R capacitors from different vendors
-4-2
capacitorESD derivative.
derivative.
pacitor
at
4-2higher
capacitor B.
ESD B.derivative. Transient
Transient Measurement
Measurement Results
Results forfor X7R X7R Capacitors
Capacitors Fig. 10 compares transient and static measurement results for
at
fewhigher
and B. Transient Measurement Results for X7R Capacitors rate of 10
Fig. current
compareschange at the beginning
transient and static and end of the TLP
measurement pulse.for
results
wfew
at andand
acapacitor
higher If the TLP is set to 1200V charge voltage itit takes about 60ns X7R capacitors andindicate
the trend of the nonlinearity matches well.
capacitor
ate if the If B.
the
If Transient
TLP
the TLP is set
is Measurement
to
set 1200V
to 1200V chargeResults
charge voltage for it
voltage X7R takes Capacitors
about
takes about 60ns 60ns The
X7R measurements
capacitors and the trendan inductance
of the of about
nonlinearity 5.5nH
matches for the
well.
gateif the
few if
capacitor and
the to B.
charge Transientthe Measurement
capacitor to 450V, Results
as for
shown X7R
in Capacitors
Fig. 8. The stress
aide fewESD and B.
to to B.
charge
Transient
If the
charge Transient
theTLP thecapacitor
Measurementis Measurement
set
capacitor to to 450V,
1200V
to Results
450V, Results
as
chargeshown
for
as for
X7R
voltage
shown inX7R
Fig.
Capacitors
init Capacitors
8.
takes
Fig. The
8.about
Thestress60ns
stress AFE capacitor test setup. Only 1.3nH had been measured for the
ESD
ate
ide
few ifESD
the wasIf repeated the TLP is set totimes1200V charge the voltage it takes about 60ns
gate ifand
the wastowas repeated
charge
If repeated
the TLPthe many is
many
settimes
capacitor
many totimes
1200V
to
to 450V, assure
assure
to charge
assure theshown
as the
voltage
phenomena
phenomena initFig.
phenomena takes are are
are stable.
stable.
8.about
The stress
60nsX7R
stable. C. capacitor
Transient (fig. 8). Its 0603 package
Measurement Resultsallowed
for AFE a lower
Capacitorinduc-
ide
ate ESD
if the to charge
Results from the capacitor
three repeats to 450V,
are as
plotted shownin in
Fig. Fig.
8. The 8. The
change stress
of C. Transient Measurement Results for AFE Capacitor
ide ESD IfResultsthe
was TLP
Results
to fromis
repeated
charge set
from threeto
the capacitormany repeats
1200V
three repeats charge
times areto plotted
voltage
assure
are plotted
to 450V, in
it
the
as shown Fig.
takes 8.
about
phenomena
in Fig. The
in Fig.8. The change
60ns are
8. are to of
stable.
change
The tance placement within the test setup.
of If the TLP is charged to 2500V, it takes about 60ns to
stress The change of the capaci-
charge
ide ESD charge was repeated many times tofor
assure the phenomena stable.
thethe
Results
the
was
capacitance
capacitance
the capacitor
capacitance
repeatedfrom versus
three
many
versus
to
versus time
450V,time
repeats
time
times for
asare
to X7R
for X7R
shown
plotted
X7R
assure
capacitors
capacitors
in Fig.
the in 8.
Fig.
capacitors is
The
phenomena 8. is
shown
is shown
stress
Theshown in
was
change
are
in
Fig.
in Fig.
of
Fig.
stable. theIf
tance the
AFE TLP
versus is
time
capacitor charged
for
to AFE to
400V, 2500V,
capacitor
as it
is
shown takes
shown
in about
in
Fig. Fig.
11. 60ns
12. The
Resultsto charge
result
from
Results
9. The from
results three
show repeats
that the are
X7R plotted
capacitors in Fig. react 8. The
fast change to
enough ofshows that the AFE capacitor reacts fast enough to the transient
9. The
repeated
the
9. The
Results results
many
capacitance
results
from show
times show
three that
to
versus the
assure
that time
repeats X7R
the the
for
X7R
are capacitors
phenomena
X7R
plotted react
capacitors
capacitors in are
Fig.react fast
stable.
8. is enough
shown
fast
The changein
enough to
Fig.the
to
ofthreeAFE capacitor
pulses are to 400V,
plotted inas shown
Fig. 11. in Fig.
Those 11. Results
indicate from
good
epresents the transient
the capacitance signal versus are timeplotted
for X7R capacitors is shown in Fig.signal, allowing this beneficial property being used for ESD pro-
esents
epresents the the
Results
9.transient
the The from signal
three
results
transient
capacitance repeats
show
signal versus that the X7R in Fig.
capacitors 8. Thereact change
fast of
enough three
to pulses are plotted in Fig. 11. Those indicate good
ns 9. The resultsversus show thattimethe for
X7R X7R capacitors
capacitors react is fast
shown in Fig.
enough repeatability. The peak and the dip at the beginning and the end
totection.
andand
epresents
0ns a aa the9.
and capacitance
the transient
The results signal
show time
that for X7R
the X7R capacitors
capacitors isreact
shown fast inenough
Fig. to Fig. 13compares
repeatability.
of the pulse are The duepeakthe and
the transient
the dip
parasitic atand
thestatic
inductance
measurement
beginning and the end
in the measurement
epresents 9. The theresults
transient show signal that the X7R capacitors react fast enough to results for the AFE capacitor and the trend of the nonlinearity
ns and
epresents a the transient signal of the pulse are due the parasitic inductance
setup and the large rate of current change at the beginning and in the measurement
0ns and a the transient signal matches
setup well. The deviation above 300V may at bethe
caused by the and
ns and a end ofand the the
TLPlargepulse. rateThe ofmeasurements
current change indicate beginning
an inductance
limitation
end of
of the5.5nH the transient
TLP pulse. measurement
TheAFE measurements method. Due
indicate to its
anOnly large
inductance
Fig. 10 compares transient and static measurement results for
of about
capacitance it was
forbarely
the possible
capacitor
to reach
test
400V
setup.
at the
1.3nH
highest
of
had about
been 5.5nH
measured for the for AFE
the capacitor
X7R test setup.
capacitor (fig. Only
8). Its 1.3nH
0603
X7R capacitors and the trend of the nonlinearity matches well. TLP charge
had been line settingfor
measured of 2500V
the X7Ras shown in Fig.(fig.
capacitor 11, causing
8). Its 0603
package allowed
uncertainties a lower
in thea capacitance inductance
estimationplacement within
as the capacitance the test
package allowed lower inductance placement within the test
values are derived just at the beginning of the falling voltage edge.
C. Transient Measurement Results for AFE Capacitor
Fig. 10
If the TLP is charged to 2500V, it takes about 60ns to charge the V. Comparision of the Capacitors for ESD Protection Fig. 10
X7R ca
X7R ca
AFE capacitor to 400V, as shown in Fig. 11. Results from three puls-
es are plotted in Fig. 11. Those indicate good repeatability. The The effectiveness of the capacitors for ESD protection can be
peak and the dip at the beginning and the end of the pulse are due compared by simulation. The nonlinear capacitor models are
the parasitic inductance in the measurement setup and the large based on the static and transient measurements.

42 ©2012 IEEE Electromagnetic Compatibility Magazine – Volume 1 – Quarter 4

 itline
wassetting
barelyofpossible
2500V asto shown
reach 400V
in Fig.at11,
thecausing
highestuncertainties
TLP charge
line
in thesetting of 2500Vestimation
capacitance as shown in asFig.
the 11, causing uncertainties
capacitance values are
in the capacitance
derived estimation
just at the beginning as the
of the capacitance
falling values are
voltage edge.
derived just at the beginning of the falling voltage edge.

Fig. 13
capacito
Fig. 1
capaci

A. S
A.
The
Th
capaci
capac
contro
iscontr
defi
is def
model
mode
functio
functi
specifi
specit
This
This
points
0 ns is point
mappe
. This Fig. 11. Measured voltage cross an AFE capacitor mapp
values
Fig.11.
Fig. 11.Measured
Measuredvoltage
voltagecross
crossanan AFE
AFE capacitor
capacitor value

Fig. 14.
Fig. 14

B. E
rs Fig. 12. Capacitance versus time for AFE capacitor B.
Fig.
Fig.12.
12.Capacitance
Capacitanceversus
versustime
timeforfor
AFE capacitor
AFE capacitor ESD
ES
shown
ts for show
imped
well. setup. The change of the capacitance versus time for AFE V. COMPARISON OF THE CAPACITORS FOR ESD PROTECTION imped
V. COMPARISON OF THE CAPACITORS FOR ESD PROTECTION charge
capacitor is shown in Fig. 12. The result shows that the AFE The effectiveness of the capacitors for ESD protection can be charg
The effectiveness of the capacitors for ESD protection can be and R
capacitor reacts fast enough to the transient signal, allowing compared by simulation. The nonlinear capacitor models are and R
compared by simulation. The nonlinear capacitor models are interac
intera
hargethis beneficial property being used for ESD protection. Fig. based on the static and transient measurements.
based on the static and transient measurements. The m
The m
from13compares the transient and static measurement results for the R3+R4
R3+R
goodAFE capacitor and the trend of the nonlinearity matches well.
e endThe deviation above 300V may be caused by the limitation of
mentthe transient measurement method. Due to its large capacitance
g andit was barely possible to reach 400V at the highest TLP charge
tanceline setting of 2500V as shown in Fig. 11, causing uncertainties
.3nHin the capacitance estimation as the capacitance values are
0603derived just at the beginning of the falling voltage edge.
test

Fig. 10. Comparison of static and transient measurement results for 4.7nF 50V Fig. 13. Comparison of static and transient measurement result for AFE
X7R capacitors capacitor
Fig. 10. Comparison of static and transient measurement results for 4.7nF Fig. 13. Comparison of static and transient measurement result for AFE
50V X7R capacitors capacitor
A. SPICE Model for Nonlinear Capacitor
The Analog Behavioral Model is used to model the nonlinear
capacitor,
©2012 IEEE Electromagnetic Compatibility Magazineas– Volume
shown1in Fig. 14.
– Quarter 4 The capacitor is modeled by a 43
controlled current source, GVALUE in PSpice, whose current
 This table contains voltage-capacitance pairs picked from
points on the measured curve. The voltage input is nonlinearly
Fig. 15. Equivalent circuit model for ESD generator
mapped from the voltage values in the table to the capacitance
values. Linear interpolation is used between table values [9].

Fig. 1

Fig.14.
Fig. 14.SPICE
SPICEmodel
modelforfornonlinear
nonlinear capacitor
capacitor

B.SPICE
A. ESD Model
CurrentforSource Model
Nonlinear Capacitor Fig. 16.
Fig. 16.Current
Currentgenerated
generatedbybythethe equivalent
equivalent circuit
circuit model
model for ESD
for the the ESD
generator when it is charged to 2000V
ESD generator is modeled using the equivalent circuit as generator when it is charged to 2000V
The Analog
shown Behavioral
in Fig. 15. This Model is used
circuit to model
models thethe nonlinear
current and the
capacitor, as
impedance ofshown
the ESD in Fig. 14. The capacitor
generator. Initially,isthe
modeled by a are
capacitors C. Capacitors
shown in Fig. 16. Intothis
Compare
example, the ESD generator is charged
ION controlled current source, initiates
GVALUE in
charged until the switch thePSpice, whose current
breakdown. C4, L2,isR4 to 2000V initially.
Sample1 from the X7R capacitors is selected to compare to
an be defined
and R5 by setequation (2). The
the initial risetime derivative
time, R1 and of the
C2voltage is the
represent the AFE capacitor. Fig. 17 shows the capacitance normalized to
s are modeled bybetween
interaction using thethe discrete
body of derivative
the ESDofgenerator
time (DDT)and function
ground. their value at 0V. In the simulation both X7R and AFE G
in PSpice.
The A voltage dependent
main discharge constant (330 capacitance is specified
Ohm, 150pF) by
is formed by C. Capacitors
capacitors aretode-normalized
Compare to 1nF at 0V; therefore, both
using a look-up table based on the measurement.
R3+R4+R5 and by C1+C2+C4. R6 represents the ESD target This table con- protection circuits have the same frequency response, but
tains
and thevoltage-capacitance
current flowing through pairs picked from points
it is shown on16.
in Fig. the In
mea-
this Sample1
body, from
thus thetotal
their X7R capacitors
charge andispulse
selected to compare
length to theThe
are similar.
different protection behavior.
example, the ESD generator is charged to 2000V initially.the
sured curve. The voltage input is nonlinearly mapped from AFE capacitor. Fig. 17 shows the capacitance
source impedance for the HBM testing is 1500Ω. Passing normalized to their
the
voltage values in the table to the capacitance values. Linear value at 0V. In the simulation both X7R and AFE
HBM test ensures a robustness of the IC input for currents capacitors are
up to
D. ESD Protection
de-normalized to 1nF at Effectiveness Comparison
interpolation is used between table values [9].
about 1.3A which we 0V; therefore, both
assume protection circuits
as the failure threshold of the
have thesimulation
The same frequencycircuit response,
is but different
explained in protection
Fig.usage
18. The
simulation. Another consideration leads to the of a device
series
behavior.
under protection is assumed as an IC. Diode ESD protection is
resistor. Without such a resistor the internal ESD protection of Fig. 1
B. ESD Current Source Model commonly used in IC design. Here only these diodes are
the IC would compete with the PCB based protection, possibly
modeledto and
leading the ICthe protecting
IC’s internal structureESD
an external is protection.
omitted. The A
ESD generator is modeled using the equivalent circuit as shown D. ESD Protection Effectiveness Comparison Th
series resistor allows for a sufficient voltage dropbody
destruction threshold of the commonly used human model
separating Usin
in Fig. 15. This circuit models the current and the impedance of (HBM) test levelmethods
is 2000V. The IC level HBM test and the IEC
both protection electrically. A resistance of 200Ω is destr
the ESD generator. Initially, the capacitors are charged until the The simulation
61000-4-2 circuit
testing areisboth
explained
based in Fig. 18. The device aunder
selected in this simulation. ESR andonESL
discharges
representfrom human
the effective
switch initiates the breakdown. C4, L2, R4 and R5 set the initial protection is assumed as an IC. Diode ESD protection is com-
series resistance and inductance of the capacitor respectively,
rise time, R1 and C2 represent the interaction between the body monly used in IC design. Here only these diodes are modeled
whose values are set to 100mΩ and 1nH. The circuit is excited
of the ESD generator and ground. The main discharge constant and the IC’s internal structure is omitted. The destruction
and by the ESD generator
totalshown in human
Fig.pulse
15. The ESD (HBM)
generator is
(330 the
Ohm,current
150pF) flowing
is formedthrough it is shown
by R3+R4+R5 and by in Fig. 16. R6
C1+C2+C4. In this body,
threshold thus their
of the commonly charge
used and length
body modelare similar.
test The
example,
representsthetheESD
ESD generator
target and is
thecharged
currentto 2000Vthrough
flowing initially.
it is
charged
source
level
to 2000V
impedance
is 2000V.
initially.
for the
The IC level HBM
HBM testtesting
and theisIEC
1500Ω. Passing
61000-4-2 test- the
HBM test ensures a robustness of the IC input for currents up to
about 1.3A which we assume as the failure threshold of the
simulation. Another consideration leads to the usage of a series
Fig. 15. Equivalent circuit model for ESD generator resistor. Without such a resistor the internal ESD protection of
the IC would compete with the PCB based protection, possibly
leading to the IC protecting an external ESD protection. A
series resistor allows for a sufficient voltage drop separating
both protection methods electrically. A resistance of 200Ω is
selected in this simulation. ESR and ESL represent the effective
series resistance and inductance of the capacitor respectively,
whose values are set to 100mΩ and 1nH. The circuit is excited
by the ESD generator shown in Fig. 15. The ESD generator is
charged to 2000V initially.

Fig. 15. Equivalent circuit model for ESD generator Fig. 17.
17. Normalized
Normalized capacitance
Fig. 15. Equivalent circuit model for ESD generator Fig. capacitanceofofX7R
X7Rand
andAFE
AFEcapacitors
capacitors

44 ©2012 IEEE Electromagnetic Compatibility Magazine – Volume 1 – Quarter 4

R
 about 0.3, while it reaches about 3 for the AFE capacitor Eu
20
investigated in this research. This larger ratio allows an [3] Be
Fig. 17. Normalized capacitance of X7R and AFE capacitors
improved trade-off between ESD protection and bandwidth. Ce
CA
[4] Pr
Ca
R 19
[5] Pi
200Ω Ce
the ESD
3.3V No
[6] De
20
C 25
[7] Ki
pare to Pb
lized to ESD 2,
Generator ESL [8] S.
d AFE “N
e, both 1nF Tu
se, but 43
ESR [9] Hi
100mΩ Ap

device
ction is
Fig.18.
18.Simulation
Simulation circuit,
circuit, ESD
ESD generator
generatorisischarged toto
2000V initially
des are Fig. charged 2000V initially
Fig. 19.
19. Comparison
Comparisonofofthe
theinput
inputcurrents of the protected IC IC
d. The Fig. currents of the protected
Theboth
ing are critical current
based flowing into
on discharges fromthe IC is shown
a human body, in Fig. 19.
thus
y model
Using the X7R capacitor, it reaches 1.6A which is
their total charge and pulse length are similar. The source above the ment methods. Similar capacitance changes have been
he IEC VI. CONCLUSION
destruction threshold ensured by 2000V HBM testing
impedance for the HBM testing is 1500Ω. Passing the HBM test of 1.3A. observed for static voltages and transient voltage changes. The
human
ensures a robustness of the IC input for currents up to about X7RX7R capacitors
capacitors lose are
mostoften used
of their as ESD protection.
capacitance while the AFE In this
1.3A which we assume as the failure threshold of the simula- applicationincrease
capacitors the voltage
their across the capacitor
capacitance values aswill surpass the
the voltage
rated voltage,
tion. Another consideration leads to the usage of a series resis- increases up to often reaching
a certain 400V
point as shownon ina Fig.
50V13,capacitor. The
for example.
tor. Without such a resistor the internal ESD protection of the voltage
This dependence
increasing of capacitors
nonlinearity with two improves
of AFE capacitor different ESD
dielectric
pro-
IC would compete with the PCB based protection, possibly materials,
tection at aX7R
givenand AFE,
signal are measured
bandwidth. using both ofstatic
The improvement and
the ESD
leading to the IC protecting an external ESD protection. A transient measurement
protection methods.
has been quantified withSimilar capacitance
simulation. changes
The temperature
series resistor allows for a sufficient voltage drop separating have been observed
dependence of the AFE’sforcapacitance
static voltages
and and transient
the long voltage
term reliabili-
both protection methods electrically. A resistance of 200Ω is changes. The X7R capacitors lose most
ty of AFE capacitors have not been investigated. of their capacitance
selected in this simulation. ESR and ESL represent the effective while the AFE capacitors increase their capacitance values as
series resistance and inductance of the capacitor respectively, the voltage increases up to a certain point as shown in Fig. 13,
whose values are set to 100mΩ and 1nH. The circuit is excited for example. This increasing nonlinearity of AFE capacitor
Acknowledgement
by the ESD generator shown in Fig. 15. The ESD generator is improves ESD protection at a given signal bandwidth. The
charged to 2000V initially. improvement
This material isofbased
the ESD
upon protection has been
work supported by the quantified with
National Sci-
simulation.
ence Theunder
Foundation temperature dependence
Grant No. 0855878. EMC of the AFE’s
The critical current flowing into the IC is shown in Fig. 19. Using capacitance and the long term reliability of AFE capacitors
the X7R capacitor, it reaches 1.6A which is above the destruction have not been investigated.
threshold ensured by 2000V HBM testing of 1.3A. In contrast the References
increasing capacitance of the AFE capacitor limits the peak cur-
rent below 0.9A. [1] Rostamzadeh, C.; Dadgostar, H.; Canavero, F.;, "Electrostatic Discharge analy-
sis of Multi Layer CeramicACKNOWLEDGMENT
capacitors," Electromagnetic Compatibility, 2009.
EMC 2009. IEEE International Symposium on , vol., no., pp.35-40, 17-21 Aug.
In designing a protection circuit, the 0V capacitance is deter- This
2009 material is based upon work supported by the National
mined by the required signal bandwidth. The protection effect is Science
[2] Foundation
Tenbohlen, S.; Streibl,under Grant No.
F.; Hartmann, 0855878.
J.; Zerrer, M.;, "Derating of ceramic
determined by the capacitance ratio between the capacitance at capacitors under ESD stress," Electromagnetic Compatibility - EMC Europe,
2008 International Symposium on , vol., no., pp.1-4, 8-12 Sept. 2008
the highest voltage reached during ESD and its 0V capacitance.
[3] Bergenthal, J.; Prymark, J.;, “Electrostatic Discharge (ESD) Concerns for
The larger the ratio is, the better the ESD protection will be. For Ceramic Capacitors,” Capacitor and Resistor Technology Symposium, CARTS
the X7R capacitor this ratio is usually about 0.3, while it reaches 99, 1999
about 3 for the AFE capacitor investigated in this research. This [4] Prymak, J. D.; Piper, J.; “ESD Susceptibility of ceramic Multilayer Capacitors,”
Capacitor and Resistor Technology Symposium, CARTS 96, 1996
larger ratio allows an improved trade-off between ESD protec- [5] Piper, J.; Prymak, J. D.;, “ESD Withstand Capability of Multilayer Ceramic
tion and bandwidth. Capacitors (MLCs)” Tech Topics Published by KEMET, Vol. 5, No. 3
[6] Demcko, Ron, and Ward, Brian, "MLCC ESD characterization," CARTS 2007
Symposium Proceedings, Albuquerque, NM, Mar. 26-29, 2007, pp. 257-262.
[7] Kim, I.; Bae, S.; Kim, K.; Kim, H.;, “Characteristics of Antiferroelectric PbZrO3
VI. Conclusion Thin Films,” Journal of the Korean Physical Society, Vol. 33, No. 2, August
1998, pp. 180~183
X7R capacitors are often used as ESD protection. In this applica- [8] S. Kwon, W. Hackenberger, E. Alberta, E. Furman, and M. Lanagan;, “Nonlin-
ear Dielectric Ceramics and Their Applications to Capacitors and Tunable
tion the voltage across the capacitor will surpass the rated volt- Dielectrics,” IEEE Electr. Insulation Mag. Vol 27, No 2 pp. 43-55 (2011).
age, often reaching 400V on a 50V capacitor. The voltage depen- [9] Hirasuna, B.;, “A Nonlinear Capacitor Model for use in PSpice”, Application
dence of capacitors with two different dielectric materials, X7R Note, Cadence 1999
and AFE, are measured using both static and transient measure-

©2012 IEEE Electromagnetic Compatibility Magazine – Volume 1 – Quarter 4 45

 Biographies
Hongyu Li (S’09) received the B.S. degree in
Automation from Shenyang Jianzhu University,
Shenyang, China, in 2001, the M.S. degree in
System Engineering from the Tongji University,
Shanghai, China, in 2005, and is currently work-
ing toward the Ph.D. degree in Electrical Engi-
neering with the EMC Laboratory at the Mis-
souri University of Science and Technology. He
was an Application Engineer from 2005 to 2007 and a Senior
Application Engineer from 2007 to 2008 with the STMicroelec-
tronics, Shanghai, China. He is working with Broadcom Corpora-
tion in California as an SI engineer since 2011. His research
interests include signal integrity (SI), electromagnetic compati-
bility (EMC), and electrostatic discharge (ESD).
Tianqi Li received the B.S. degree in Communi-
cations Engineering from Electronic Engineer-
ing Department of Tianjin University, Tianjin,
China, in 2004. He then joined Huawei Technol-
ogies as a hardware designer. He joined the
Electromagnetic Compatibility Laboratory at the
Missouri University of Science and Technology
in 2009 and received his M.S. degree in Elec-
tronic Engineering in 2011. He is currently working toward the
Ph.D. degree. His research interests include electromagnetic
interference and compatibility, electrostatic discharge, circuit
modeling and signal integrity.
Victor V. Khilkevich received the Ph. D. degree
in electrical engineering from Moscow Power
Engineering Institute (Technical University),
Moscow, Russia in 2001. He is currently work-
ing as a research associate professor at the
Missouri University of Science and Technology.
His primary research interests are computa-
tional electrodynamics, automotive EMC model-
ing, and high-frequency measurement techniques.
Dr. David Pommerenke’s interest within EMC is
ESD, immunity and electronics often using
experimental approaches and numerical simu-
lations. He is professor at the EMC laboratory
at the Missouri University of Science and Tech-
nology. His Ph.D. is from the Technical Univer-
sity Berlin. He has published more than 100
papers and is an inventor on 11 patents. He has
contributed to a variety of EMC test methods and standards,
such as the IEC 61000-4-2 ESD test standard which he helped to
re-write as a US representative in the standard writing body.
10-EA-0303c-Tryengineering-3.25x4.75-Final.indd 1
46 ©2012 IEEE Electromagnetic Compatibility Magazine – Volume 1 – Quarter 4
Seongtae Kwon was born in Andong, South
Korea in1962. He is a Senior Scientist at TRS
Ceramics. He received a B.S. and a M.S.
degree, both in Inorganic Material Science and
Engineering in 1984 and 1986, respectively, from
Seoul National University in Korea and a Ph.D.
in Materials Science in 1997, from the Pennsyl-
vania State University. From 1986 to 1992 he
served as a research engineer in the Central Research Center of
Ssangyong Cement Co. in Korea. From 1997 to 1999 he worked as
the R&D and QC Manager in Ssangyong Materials. Before join-
ing TRS Ceramics in 2002, he was a Research Associate in the
Pennsylvania State University from 1999, and worked on templat-
ed grain growth of PMN-PT piezoelectric ceramics. His current
research interests include processing and developing of special
type capacitors such as antiferroelectric ceramics for high ener-
gy density capacitors, high temperature capacitors and cryogen-
ic capacitors and other ceramic processing.
Wesley Hackenberger has a Bachelor of Sci-
ence degree in Nuclear Engineering and a
Ph.D. in Materials from the Pennsylvania State
University. In 1995 Dr. Hackenberger joined TRS
Technologies, Inc. as Director of R&D. He was
named President of the company in 2003 and
continues to serve in that capacity to the pres-
ent day. At TRS Dr. Hackenberger has been
responsible for the development and commercialization of several
advanced materials technologies including single crystal piezo-
electrics, nonlinear dielectrics, high energy density capacitors,
andExplore
ferroelectrics
the for pulsed power
amazing worldapplications. Dr. Hackenberg-
of engineers—
er is an author on more than 50 scientific publications and is listed
all inventor
as an in oneonweb site…with over 10 pending applications.
five patents
TryEngineering.org
■ See the exciting work that engineers do
Learn how they make a difference n g i n e er
yE
■
Tr
in
■ Play online games and challenges
g
■ Find accredited engineering
programs, summer camps,
lesson plans, and more
Visit www.tryengineering.org today!
Brought to you by IEEE, IBM and TryScience
9/10/10 2:58 PM