Pulse Load MLCC - KOA Speer Electronics

DC, AC and Pulse Load
of Multilayer Ceramic
Capacitors
Bolivar Drive * P.O. Box 547
Bradford, PA 16701 USA
Phone: 814-362-5536
Fax: 814-362-8883
www.koaspeer.com
DC, AC and Pulse Load of Multilayer Ceramic Capacitors
Introduction
General introduction
Multilayer Ceramic Capacitors (MLCCs) are increasingly
being used in applications in which the electrical load
becomes critical. This publication is particularly
concerned with the susceptibility of MLCCs to:
- DC electrical fields
- AC electrical fields at various frequencies, and
- pulse loads
The resistance to these types of load is important for
application areas such as automotive and lighting and in
switched-mode power supplies. The use at elevated
temperatures in some applications adds further
demands.
CONTENTS
Introduction
DC load
AC load
Pulse load
ESD pulses
Automotive pulses
Surge pulses
ANNEX Application Questionnaire
3
4
5
7
7
9
10
11
This application note presents typical data that may help
the application engineer in selecting the optimal
product.The aim is to present general rules concerning
the immunity to given electrical loads. The MLCC
insulation resistance was used as a criterion to check
the immunity against a given electrical load.
The data, however, is not relevant for all applications and
we recommend that customers use the application
questionnaire in the Annex, for special questions.
Before going into details of the various tests, we present
some general information important for a proper
understanding of the various failure mechanisms.
First we shall describe MLCC construction and present
structural parameters that determine resistance to
electrical breakdown.
Three types of breakdown mechanism are then
discussed: dielectric breakdown, electro-thermal breakdown
and electro-mechanical breakdown.
Construction
An MLCC comprises several layers of non-fired stacked
ceramic foils on which electrode material is printed.
These foils are pressed and sintered to obtain a
compact multilayer structure.
The capacitance of an MLCC depends on the capacitive
area of each electrode (Ae), on the number of inner
electrodes (Ne), the thickness of the ceramic dielectric
(d) and on the relative dielectric constant of the
ceramic material (Âr):
(1)
2
MLCCs in a series with a given rated voltage have a
related minimum dielectric thickness. The dielectric
thickness is greatest for high-voltage products and may
also be larger for low-capacitive MLCCs in general.
Figure 1 shows a typical construction.An electrical load
will give rise to an electric field across the dielectric as
well as across the various creepage paths (margins).The
three types of margin are: the end margin, the side
margin and the cover layer thickness (see Fig.1). The
magnitude of these margins is great compared with the
dielectric thickness. Hence only the dielectric layers
form a potential breakdown pathway.
end margin
dielectric
thickness
cover layer
thickness
side margin
inner electrodes
ceramic material
MSB651
terminations
Fig.1
Construction of a ceramic multilayer capacitor
Dielectric breakdown
The MLCC may show dielectric breakdown at very strong
field strengths. The component may fail because of the
limited intrinsic dielectric strength of the ceramic
material.
This failure mechanism is related not only to the quality
of the ceramic dielectric material but also to the
dielectric thickness d (the field strength is inversely
proportional to d) and the area of the dielectric.
With the actual MLCC constructions in a given voltage
series (as referred to above), there is no simple relation
between the dielectric breakdown and capacitance. At
lower capacitances, the larger dielectric thickness and
the higher total effective capacitive surface may have
opposing effects on the dielectric breakdown voltage.
Electro-thermal breakdown
High local temperatures caused by power dissipation
may result in electro-thermal breakdown. The ambient
temperature is an important factor here. Additionally,
electro-thermal breakdown is influenced by the heat
generated inside the MLCC and by the heat flow to its
surroundings. The heat generated inside the MLCC
depends on the dissipation factor (which is a function of
temperature, frequency, voltage and construction), on
the voltage amplitude, the voltage-time relation (e.g. the
frequency of an AC load) and the capacitance.
3
Bolivar Drive
■
P.O. Box 547
■
Bradford, PA 16701
■
USA
■
814-362-5536
■
Fax 814-362-8883
■
www.koaspeer.com
DC, AC and Pulse Load of Multilayer Ceramic Capacitors
Introduction
MLCCs in a series with a given rated voltage have a
related minimum dielectric thickness. The dielectric
thickness is greatest for high-voltage products and may
also be larger for low-capacitive MLCCs in general.
General introduction
Multilayer Ceramic Capacitors (MLCCs) are increasingly
being used in applications in which the electrical load
becomes critical. This publication is particularly
concerned with the susceptibility of MLCCs to:
- DC electrical fields
- AC electrical fields at various frequencies, and
- pulse loads
Figure 1 shows a typical construction.An electrical load
will give rise to an electric field across the dielectric as
well as across the various creepage paths (margins).The
three types of margin are: the end margin, the side
margin and the cover layer thickness (see Fig.1). The
magnitude of these margins is great compared with the
dielectric thickness. Hence only the dielectric layers
form a potential breakdown pathway.
The resistance to these types of load is important for
application areas such as automotive and lighting and in
switched-mode power supplies. The use at elevated
temperatures in some applications adds further
demands.
CONTENTS
Introduction
DC load
AC load
Pulse load
ESD pulses
Automotive pulses
Surge pulses
ANNEX Application Questionnaire
3
4
5
7
7
9
10
11
end margin
dielectric
thickness
cover layer
thickness
This application note presents typical data that may help
the application engineer in selecting the optimal
product.The aim is to present general rules concerning
the immunity to given electrical loads. The MLCC
insulation resistance was used as a criterion to check
the immunity against a given electrical load.
side margin
inner electrodes
ceramic material
MSB651
terminations
The data, however, is not relevant for all applications and
we recommend that customers use the application
questionnaire in the Annex, for special questions.
Fig.1
Construction of a ceramic multilayer capacitor
Dielectric breakdown
The MLCC may show dielectric breakdown at very strong
field strengths. The component may fail because of the
limited intrinsic dielectric strength of the ceramic
material.
Before going into details of the various tests, we present
some general information important for a proper
understanding of the various failure mechanisms.
First we shall describe MLCC construction and present
structural parameters that determine resistance to
electrical breakdown.
This failure mechanism is related not only to the quality
of the ceramic dielectric material but also to the
dielectric thickness d (the field strength is inversely
proportional to d) and the area of the dielectric.
Three types of breakdown mechanism are then
discussed: dielectric breakdown, electro-thermal breakdown
and electro-mechanical breakdown.
With the actual MLCC constructions in a given voltage
series (as referred to above), there is no simple relation
between the dielectric breakdown and capacitance. At
lower capacitances, the larger dielectric thickness and
the higher total effective capacitive surface may have
opposing effects on the dielectric breakdown voltage.
Construction
An MLCC comprises several layers of non-fired stacked
ceramic foils on which electrode material is printed.
These foils are pressed and sintered to obtain a
compact multilayer structure.
Electro-thermal breakdown
High local temperatures caused by power dissipation
may result in electro-thermal breakdown. The ambient
temperature is an important factor here. Additionally,
electro-thermal breakdown is influenced by the heat
generated inside the MLCC and by the heat flow to its
surroundings. The heat generated inside the MLCC
depends on the dissipation factor (which is a function of
temperature, frequency, voltage and construction), on
the voltage amplitude, the voltage-time relation (e.g. the
frequency of an AC load) and the capacitance.
The capacitance of an MLCC depends on the capacitive
area of each electrode (Ae), on the number of inner
electrodes (Ne), the thickness of the ceramic dielectric
(d) and on the relative dielectric constant of the
ceramic material (Âr):
(1)
2
3
■
P.O. Box 547
■
Bradford, PA 16701
■
USA
■
814-362-5536
■
Fax 814-362-8883
■
www.koaspeer.com
The heat flow to the surroundings of an MLCC may
take place by conduction, radiation and convection. It
depends on the MLCC geometry, the thermal
conductivities of the various materials (ceramic,
terminations, solder, print board material), air flow,
radiation, temperature gradients and heat-transfer
coefficients.
Power dissipation
The general equation for power dissipation upon
stressing the MLCC with AC fields is
P = ˆCV2RMStan‰
■
Bradford, PA 16701
0
10-1
4
200 V
2
0
10-1
Fig.2
■
www.koaspeer.com
103
104
105
1
102
10
C (nF)
103
DC instant breakdown voltage as a function of
capacitance, rated voltage and size for X7R MLCCs.
Curves through the cumulative defect points are given
MSC953
100
cumulative
survival
(%)
80
DC instant breakdown voltage as a function of
capacitance, rated voltage and size for NP0 MLCCs.
Curves through the 50% cumulative defect points
are given
A
60
B
C
D
E
40
MSC951
100
cumulative
survival
(%)
80
20
A
B
C
D
E
0
F
0
Fig.5
40
1
2
3
4
5
6
7
8
VBR (DC) (kV)
Typical DC instant breakdown voltage curves for
X7R MLCCs. A: 47 nF/1206/50 V,
B: 10 nF/1206/50 V, C: 330 pF/1206/50 V,
D: 470 pF/1206/500 V, E: 1 nF/1808/1 kV
20
AC load
0
0
Fig.3
Fax 814-362-8883
102
10
1
C (nF)
Comparable and sometimes improved DC breakdown
voltages are reached for larger product sizes (1812
compared with 1206), if the capacitance, dielectric
thickness and type of ceramic material are the same.
Additionally, to avoid corona effects it is advisable to use
MLCCs with larger termination separation lengths
(1812 = 1808 > 1206), especially when using
low-capacitance MLCCs.
■
Fig.4
500 V
50 V
Typical breakdown voltages of products with a rated
voltage ≥ 50 V are above 1.5 kV for NP0 and above 500 V
for X7R products. NP0 type products with a
construction comparable to X7R type products turn
out to have superior DC immunity. As expected, DC
immunity is better for products with large dielectric
thickness. This is the case for high-voltage products
(rated voltage 200 V, 500 V, 1 kV and 3 kV), and because
of their construction, it is also the case for low
capacitive products in general.
814-362-5536
200 V
50 V
60
■
500 V
3 kV
6
Figures 2 to 5 present experimental data on the DC
instant breakdown voltage. In these figures, exposure
time is about 5 s. Breakdown after prolonged exposure
will be somewhat lower. Corona effects may occur at
higher voltage levels, especially when testing smaller
sized products such as 0603 and 0805.
USA
4
8
An MLCC loaded at increasing voltages will finally
break down because of the limited dielectric strength.
This failure mechanism has been treated in the
preceding chapter.
■
1 kV
10
DC load
4
VBR (DC)
(kV)
2
VBR (DC)
(kV)
in which Rth is the thermal resistance for heat transfer
to the environment by conduction, radiation and
convection in K/W and ¢T is the temperature rise of
the MLCC in K. From equations (2) and (3) it follows
that the temperature rise of an MLCC caused by an AC
load of voltage V at a given frequency will, in theory, be
proportional to V2RMS. This has been found for type I
MLCCs, not for type II MLCCs, in which tan‰ is a
function not only of frequency but also of applied
voltage and temperature.
Electro-mechanical breakdown due to the ferro-electric
properties of the ceramic material does not occur in
MLCCs with type I ceramic material (NP0).
P.O. Box 547
(3)
MSC952
6
MSC950
12
P = ¢T
Rth.
Electro-mechanical breakdown
When piezoelectric materials are exposed to an
electric field they are deformed. This has become
known as the inverse piezoelectric effect. For
polycrystalline ferroelectric materials, as used in
Phycomp type II MLCCs (dielectric X7R and Y5V),
below the Curie point the crystallites take on
tetragonal symmetry. The + and - charge sites do not
coincide, resulting in electric dipoles.The material is said
to be composed of Weiss domains. Within a Weiss
domain, all the dipoles are aligned, giving a net dipole
moment to the domain. The directions of polarization
between neighbouring domains within a crystallite can
differ by 90° or 180°. Exposing the material to a strong
electric field below the Curie point will result in growth
of domains most nearly aligned with the field at the
expense of other domains. The material will also
lengthen in the direction of the field. If this change in
dimension takes place slowly, the resulting stresses in
the material may be relaxed. However, at fast field
changes, i.e. at high dV/dt or in other words at high
currents, the stresses may exceed a critical threshold
value and result in electro-mechanical breakdown.
■
(2)
At thermal equilibrium, the power generated inside the
MLCC equals the heat transferred to the environment.
This means that
A special type of electro-thermal breakdown may occur
when the voltage difference between the terminations
is high enough to cause air discharges (corona). These
discharges on the outer surface of the MLCC may lead
to high local currents that destroy regions in the MLCC
itself (‘burning spots’). Factors influencing this are:
- the form of the electrodes. Sharp points result in high
electrical field gradients which may be damaging. Sharp
points may be the result of bad soldering.
- air humidity and surface condition (presence of
conductive surface contamination, e.g. from human skin).
- distance between the two terminations. This distance
increases in the order 0402 < 0603 < 0805 < 1206 =
1210 < 1808 = 1812 .
Bolivar Drive
Note that the maximum DC load is not equivalent to
voltage rating, which is usually 10 to 20 times lower. In the
first case, DC breakdown levels refer to the behaviour at
ambient temperatures, while the voltage rating of an
MLCC is determined by its behaviour at elevated
temperatures (maximum specified temperature of 125 °C
for both NP0 and X7R type MLCCs) and over extended
periods (1000 hr at 1.5 times the rated voltage) in order
to meet the international CECC requirements.
2
4
6
8
10
12
VBR (DC) (kV)
Typical DC instant breakdown voltage curves for
NP0 MLCCs. A: 10 pF/0805/200 V,
B: 1 nF/1206/50 V, C: 10 nF/1812/50 V,
D: 15 pF/1206/500 V, E: 82 pF/1812/3 kV,
F: 10 pF/1808/3 kV
The AC immunity levels presented below are not
specified but represent typical values.
AC breakdown is observed as a strong drop in the
insulation resistance. The voltage level is used as an
indicator only, because V and I are related.
Figures 6 and 7 show AC breakdown curves at 50 Hz as
a function of the MLCC capacitance.There is a very close
correspondence between the data presented here and
the DC immunity data presented in the previous chapter.
■
P.O. Box 547
■
■
■ Bradford,
■ USA ■■ 814-362-5536 ■ Fax
■ www.koaspeer.com
P.O.
Box 547
PA■16701
■ USA
Bradford,
PA 16701
814-362-5536
Fax 814-362-8883 ■ 814-362-8883
www.koaspeer.com
5
The heat flow to the surroundings of an MLCC may
take place by conduction, radiation and convection. It
depends on the MLCC geometry, the thermal
conductivities of the various materials (ceramic,
terminations, solder, print board material), air flow,
radiation, temperature gradients and heat-transfer
coefficients.
A special type of electro-thermal breakdown may occur
when the voltage difference between the terminations
is high enough to cause air discharges (corona). These
discharges on the outer surface of the MLCC may lead
to high local currents that destroy regions in the MLCC
itself (‘burning spots’). Factors influencing this are:
- the form of the electrodes. Sharp points result in high
electrical field gradients which may be damaging. Sharp
points may be the result of bad soldering.
- air humidity and surface condition (presence of
conductive surface contamination, e.g. from human skin).
- distance between the two terminations. This distance
increases in the order 0402 < 0603 < 0805 < 1206 =
1210 < 1808 = 1812 .
Electro-mechanical breakdown
When piezoelectric materials are exposed to an
electric field they are deformed. This has become
known as the inverse piezoelectric effect. For
polycrystalline ferroelectric materials, as used in
Phycomp type II MLCCs (dielectric X7R and Y5V),
below the Curie point the crystallites take on
tetragonal symmetry. The + and - charge sites do not
coincide, resulting in electric dipoles.The material is said
to be composed of Weiss domains
Power dissipation
The general equation for power dissipation upon
stressing the MLCC with AC fields is
P = ˆCV2RMStan‰
(2)
At thermal equilibrium, the power generated inside the
MLCC equals the heat transferred to the environment.
This means that
P = ¢T
Rth.
Note that the maximum DC load is not equivalent to
voltage rating, which is usually 10 to 20 times lower. In the
first case, DC breakdown levels refer to the behaviour at
ambient temperatures, while the voltage rating of an
MLCC is determined by its behaviour at elevated
temperatures (maximum specified temperature of 125 °C
for both NP0 and X7R type MLCCs) and over extended
periods (1000 hr at 1.5 times the rated voltage) in order
to meet the international CECC requirements.
VBR (DC)
(kV)
1 kV
4
500 V
200 V
MSC950
12
(3)
MSC952
6
2
VBR (DC)
(kV)
3 kV
10
in which Rth is the thermal resistance for heat transfer
to the environment by conduction, radiation and
convection in K/W and ¢T is the temperature rise of
the MLCC in K. From equations (2) and (3) it follows
that the temperature rise of an MLCC caused by an AC
load of voltage V at a given frequency will, in theory, be
proportional to V2RMS. This has been found for type I
MLCCs, not for type II MLCCs, in which tan‰ is a
function not only of frequency but also of applied
voltage and temperature.
50 V
8
0
10-1
6
Fig.4
500 V
50 V
4
200 V
2
DC load
0
10-1
102
10
1
103
104
105
DC instant breakdown voltage as a function of
capacitance, rated voltage and size for NP0 MLCCs.
Curves through the 50% cumulative defect points
are given
C (nF)
103
DC instant breakdown voltage as a function of
capacitance, rated voltage and size for X7R MLCCs.
Curves through the cumulative defect points are given
MSC953
A
60
B
C
D
E
40
MSC951
100
cumulative
survival
(%)
80
102
10
100
cumulative
survival
(%)
80
C (nF)
Fig.2
1
20
A
60
B
C
D
E
0
F
0
Fig.5
40
1
2
3
4
5
6
7
8
VBR (DC) (kV)
Typical DC instant breakdown voltage curves for
X7R MLCCs. A: 47 nF/1206/50 V,
B: 10 nF/1206/50 V, C: 330 pF/1206/50 V,
D: 470 pF/1206/500 V, E: 1 nF/1808/1 kV
20
AC load
0
0
Fig.3
2
4
6
8
The AC immunity levels presented below are not
specified but represent typical values.
10
12
VBR (DC) (kV)
Typical DC instant breakdown voltage curves for
NP0 MLCCs. A: 10 pF/0805/200 V,
B: 1 nF/1206/50 V, C: 10 nF/1812/50 V,
D: 15 pF/1206/500 V, E: 82 pF/1812/3 kV,
F: 10 pF/1808/3 kV
AC breakdown is observed as a strong drop in the
insulation resistance. The voltage level is used as an
indicator only, because V and I are related.
Figures 6 and 7 show AC breakdown curves at 50 Hz as
a function of the MLCC capacitance.There is a very close
correspondence between the data presented here and
the DC immunity data presented in the previous chapter.
4
Bolivar Drive
■
P.O. Box 547
■
Bradford, PA 16701
■
USA
■
814-362-5536
■
Fax 814-362-8883
■
www.koaspeer.com
5
Also here, increased immunity is found for lower
capacitance MLCCs at higher rated voltages. NP0 type
MLCCs of comparable construction are superior to
X7R type MLCCs.
Vbr(RMS) (kV)
at 50 Hz AC
MSC958
1.2
VBR,rms
(kV)
1 nF 50 V
200 V
10 nF 50 V
50 V
100 nF 50 V
100 nF 16 V
1
102
10
C (nF)
0
10
103
102
Fig. 7
AC instant breakdown voltage as a function of
capacitance, rated voltage and size for X7R MLCCs.
AC frequency: 50 Hz. Curves through the 50%
cumulative defect points are given
1.2
0.8
2.2 nF 50 V
ESD pulses
Electrostatic Discharge (ESD) of persons or electricallyloaded objects is a well known threat to the optimal
behaviour of electronic equipment.
100 nF 50 V
0.4
100 pF 50 V
0
10
102
10
(b)
Fig. 9
102
103
104
105
f (Hz)
106
(a)
103
104
f (Hz)
AC instant breakdown voltage as a function of AC
frequency for X7R MLCCs. Curves through the 50%
cumulative defect points are given.
(a) 0805, (b) 1206
MSC960
80
temperature
rise (K)
60
100 kHz
1 nF 500 V
1.2 nF 200 V
1.2
8.2 pF 50 V
1 nF 50 V
50 kHz
1.2 nF 50 V
0.8
40
20 kHz
500 kHz
0.4
50 V
10 kHz
20
0
10
2
160 kHz
102
103
104
105
500 V
(b)
Fig. 8
200 V
1
102
10
103
C (pF)
104
Nowadays, all equipment brought onto the market, put
into operation or already in production in countries
within the European Union must comply with EMC
(ElectroMagnetic Compatibility) requirements on both
emission and immunity. One of the immunity
requirements concerns ESD. Although there is no such
requirement for components, their behaviour will
influence the behaviour of the equipment in which they
operate.
MSC957
2.0
VBR,rms
(kV)
1.6
Vbr(RMS) (kV)
at 50 Hz AC
surge tests
2.2 nF 500 V
0.8
8.2 pF 50 V
0.4
relevant test
ESD
automotive pulses
The various pulses cover the whole spectrum from fast
and low-energy pulses to slow and high-energy pulses.
1 nF 50 V
VBR,rms
(kV)
MSC956
1 nF 50 V
1.5 nF 50 V
slow
(0.1-10 µs)
energy level
low (µJ-mJ)
medium
(1-10 s of mJ)
high
(1-100 s of J)
MSC959
1.2
0
MSC954
1
104
f (Hz)
10 nF 50 V
2.0
VBR,rms
(kV)
1.6
Note: for both AC and DC loads, the use of higher rated
voltage MLCCs (200 and 500 V) is generally recommended instead of the standard 50 V products.
0
10-1
103
(a)
Figure 10 shows temperature rise as a function of
voltage at higher frequencies for 1 nF 0805 50 V NP0
and X7R MLCCs.The temperature rise of NP0 MLCCs
is less than that of X7R MLCCs under the same
conditions. This is due to the lower loss factor of NP0
capacitors. Note that for NP0 types, temperature rise is
more or less proportional to frequency and the square
of the voltage as given in equations (2) and (3). The
temperature rise of X7R types deviates from this
behaviour. The data suggests that tan‰ decreases with
temperature and applied field which has indeed been
found for X7R types.
3
rise-time
fast (1 ns)
slow (1 µs)
0.4
0
10-1
Pulse load
This section discusses the immunity of MLCCs to
various types of transient. Three different groups of
transient can be considered:
220 pF 50 V
0.8
500 V
0.5
- At low frequencies (below roughly 50 kHz), the
breakdown level is nearly independent of frequency.
Breakdown occurs because of the limited dielectric
strength of the ceramic material and possibly because
of electro-mechanical effects.
The AC RMS instant breakdown level is about 35% of
the DC instant breakdown level.
- At frequencies roughly above 50 kHz, breakdown
occurs at decreasing voltage levels because of energy
dissipation.
Fig.6
TN-100
1.0
Figures 8 and 9 show AC breakdown levels as a function
of frequency. The AC instant breakdown curve as a
function of frequency can be divided into two regions:
4
MSC955
1.5
f (Hz)
106
50 kHz
10 kHz
0
AC instant breakdown voltage as a function of AC
frequency for NP0 MLCCs. Curves through the 50%
cumulative defect points are given.
(a) 0805, (b) 1206
AC instant breakdown voltage as a function of
capacitance, rated voltage and size for NP0 MLCCs.
AC frequency: 50 Hz. Curves through the 50%
cumulative defect points are given
0
100
200
1 nF 0805 50 V NPO
1 nF 0805 50 V X7R
300
400
500
V (Vrms)
Fig.10 Example of MLCC temperature rise as a function of
AC voltage for 1 nF 0805 50 V NP0 and X7R MLCCs
The immunity of MLCCs to ESD pulses is not well
characterized. Tests were therefore performed to
analyze the effect of these fast and low-energy pulses.
Pulse tests were performed according to two standards:
MIL-STD 883C (Human Body Model) and UZW-BO/
FQ-B302 (Machine Model) developed by Philips. These
standards were originally developed for the ESD testing
of semiconductors and we have adapted the test
methods to make them suitable for MLCCs. The tests
were performed on an adapted Verifier test apparatus.
MLCCs soldered onto IC substrates were subjected to
six positive pulses per test run. These pulses were
applied by direct contact rather than by air discharge.A
discharging step, not specified in the original standards,
was added between each pulse. (Negative pulses were
not applied owing to the apolarity of the MLCCs).
When subjected to ESD pulses, low-capacitance MLCCs
sometimes exhibited corona effects without internal
damage. In these cases, the products were immersed in
oil (Dow Corning fluid 550) to obtain the immunity data.
6
7
Bolivar Drive
■
P.O. Box 547
■
Bradford, PA 16701
■
USA
■
814-362-5536
■
Fax 814-362-8883
■
www.koaspeer.com
Also here, increased immunity is found for lower
capacitance MLCCs at higher rated voltages. NP0 type
MLCCs of comparable construction are superior to
X7R type MLCCs.
Vbr(RMS) (kV)
at 50 Hz AC
Figure 10 shows temperature rise as a function of
voltage at higher frequencies for 1 nF 0805 50 V NP0
and X7R MLCCs.The temperature rise of NP0 MLCCs
is less than that of X7R MLCCs under the same
conditions. This is due to the lower loss factor of NP0
capacitors. Note that for NP0 types, temperature rise is
more or less proportional to frequency and the square
of the voltage as given in equations (2) and (3). The
temperature rise of X7R types deviates from this
behaviour. The data suggests that tan‰ decreases with
temperature and applied field which has indeed been
found for X7R types.
0
10-1
1
10
102
C (nF)
0
10
103
102
103
104
f (Hz)
Fig. 7
AC instant breakdown voltage as a function of
capacitance, rated voltage and size for X7R MLCCs.
AC frequency: 50 Hz. Curves through the 50%
cumulative defect points are given
1.2
2.2 nF 500 V
0.8
2.2 nF 50 V
ESD pulses
Electrostatic Discharge (ESD) of persons or electricallyloaded objects is a well known threat to the optimal
behaviour of electronic equipment.
MSC956
100 nF 50 V
0.4
8.2 pF 50 V
102
102
10
(b)
Fig. 9
0.4
0
10
Nowadays, all equipment brought onto the market, put
into operation or already in production in countries
within the European Union must comply with EMC
(ElectroMagnetic Compatibility) requirements on both
emission and immunity. One of the immunity
requirements concerns ESD. Although there is no such
requirement for components, their behaviour will
influence the behaviour of the equipment in which they
operate.
100 pF 50 V
1 nF 50 V
1.5 nF 50 V
103
104
105
f (Hz)
106
(a)
103
104
f (Hz)
AC instant breakdown voltage as a function of AC
frequency for X7R MLCCs. Curves through the 50%
cumulative defect points are given.
(a) 0805, (b) 1206
MSC960
80
MSC957
2.0
VBR,rms
(kV)
1.6
temperature
rise (K)
The immunity of MLCCs to ESD pulses is not well
characterized. Tests were therefore performed to
analyze the effect of these fast and low-energy pulses.
Pulse tests were performed according to two standards:
MIL-STD 883C (Human Body Model) and UZW-BO/
FQ-B302 (Machine Model) developed by Philips. These
standards were originally developed for the ESD testing
of semiconductors and we have adapted the test
methods to make them suitable for MLCCs. The tests
were performed on an adapted Verifier test apparatus.
MLCCs soldered onto IC substrates were subjected to
six positive pulses per test run. These pulses were
applied by direct contact rather than by air discharge.A
discharging step, not specified in the original standards,
was added between each pulse. (Negative pulses were
not applied owing to the apolarity of the MLCCs).
60
100 kHz
1 nF 500 V
1.2 nF 200 V
1.2
8.2 pF 50 V
1 nF 50 V
50 kHz
1.2 nF 50 V
0.8
40
20 kHz
500 kHz
0.4
10 kHz
20
500 V
(b)
Fig. 8
200 V
1
10
102
103
C (pF)
104
surge tests
The various pulses cover the whole spectrum from fast
and low-energy pulses to slow and high-energy pulses.
1 nF 50 V
VBR,rms
(kV)
0
0.8
relevant test
ESD
automotive pulses
MSC959
1.2
10 nF 50 V
2.0
VBR,rms
(kV)
1.6
energy level
low (µJ-mJ)
medium
(1-10 s of mJ)
high
(1-100 s of J)
slow
(0.1-10 µs)
(a)
50 V
1
rise-time
fast (1 ns)
slow (1 µs)
100 nF 50 V
100 nF 16 V
MSC954
2
Fig.6
10 nF 50 V
0.4
50 V
0
10
0
10-1
This section discusses the immunity of MLCCs to
various types of transient. Three different groups of
transient can be considered:
1 nF 50 V
200 V
Vbr(RMS) (kV)
at 50 Hz AC
3
Pulse load
220 pF 50 V
0.8
500 V
0.5
Note: for both AC and DC loads, the use of higher rated
voltage MLCCs (200 and 500 V) is generally recommended instead of the standard 50 V products.
4
MSC958
1.2
VBR,rms
(kV)
1.0
Figures 8 and 9 show AC breakdown levels as a function
of frequency. The AC instant breakdown curve as a
function of frequency can be divided into two regions:
- At low frequencies (below roughly 50 kHz), the
breakdown level is nearly independent of frequency.
Breakdown occurs because of the limited dielectric
strength of the ceramic material and possibly because
of electro-mechanical effects.
The AC RMS instant breakdown level is about 35% of
the DC instant breakdown level.
- At frequencies roughly above 50 kHz, breakdown
occurs at decreasing voltage levels because of energy
dissipation.
MSC955
1.5
160 kHz
102
103
104
105
f (Hz)
106
50 kHz
10 kHz
0
AC instant breakdown voltage as a function of AC
frequency for NP0 MLCCs. Curves through the 50%
cumulative defect points are given.
(a) 0805, (b) 1206
0
100
200
300
1 nF 0805 50 V NPO
1 nF 0805 50 V X7R
400
500
V (Vrms)
Fig.10 Example of MLCC temperature rise as a function of
AC voltage for 1 nF 0805 50 V NP0 and X7R MLCCs
AC instant breakdown voltage as a function of
capacitance, rated voltage and size for NP0 MLCCs.
AC frequency: 50 Hz. Curves through the 50%
cumulative defect points are given
When subjected to ESD pulses, low-capacitance MLCCs
sometimes exhibited corona effects without internal
damage. In these cases, the products were immersed in
oil (Dow Corning fluid 550) to obtain the immunity data.
6
7
Bolivar Drive
■
P.O. Box 547
■
Bradford, PA 16701
■
USA
■
814-362-5536
■
Fax 814-362-8883
■
www.koaspeer.com
MSC962
2.0
Human Body Model pulses according to
MIL-STD 883C.
These pulses simulate the discharge of an electricallyloaded human body by using a discharge circuit with a
capacitor of 100 pF, a resistance of 1.5 kΩ and an
inductance of 7.5 µH (not specified in the standard but
derived from pulse characteristics).
Machine Model pulses according to the Philips
standard UZW-BO/FQ-B302.
These pulses simulate the discharge of an electricallyloaded machine by using a discharge circuit with a
capacitor of 200 pF, a resistance of only 25 Ω and an
inductance of 2.5 µH (not specified in the standard, but
derived from pulse characteristics).
Figure 11 shows a typical pulse form. The pulse rise
time tri is less than 10 ns, the delay time td is 150 ns, the
peak current at 2 kV is 1.25 A (short circuit) and at 4 kV
it is 2.5 A (short circuit). Ir caused by ringing must be
less than 5% of the peak current. The voltage level is
0 to 10 kV.
A typical pulse form is presented in Fig.13. It is the
calibration-current waveform for a charged voltage of
400 V.The first peak current Ip1 is 3 A.The damping Ip1/Ip2
is 1.4. The resonant frequency is 7.1 MHz. The voltage
level is 0 to 2 kV.
0
Ip2
4
0
36.8
t
MSC961
X7R
500 V
ESD voltage
(kV)
NPO
500 V
8
X7R
25/200 V
4
NPO
50/200 V
160
200
t (ns)
The experimental data presented in Figs 12 and 14 show
that energy dissipation is highest (and the ESD
breakdown voltage level is lowest) when the tested
MLCC has a capacitance close to the capacitance of the
discharge circuit. This is in line with our model
calculations. Moreover, the figures show that the X7R
data seems to be shifted somewhat to higher
capacitance values compared with the NP0 data. This
may be due to the voltage-dependent capacitance of
X7R products, which causes a lowering of capacitance at
higher DC voltage levels.Additionally, the ESD immunity
of X7R products is found to be lower than that of NP0
products with the same capacitance and rated voltage.
2
1
10
102
103
104
105
106
C (pF)
Fig.12 Experimental data on ESD immunity of NP0 and
X7R MLCCs of size 0805 and 1206 based on
Human Body Model pulses according to MIL-STD
883C
8
Bolivar Drive
120
The advantage of the high-voltage products (rated
voltage 500 V) is clear when comparing the 50 V data
with the 500 V data since higher immunity levels are
obtained for the high-voltage products.
6
0
10-1
80
Figure 14 gives data on ESD immunity for NP0 and X7R
type MLCCs tested according to the Machine Model
standard. Several types of 25, 50, 200 and 500 V products
in NP0 and X7R ceramic were measured. All products
were resistant to at least 2.5 kV HBM and 1 kV MM
pulses. 0805 and 1206 products showed the same
behaviour.
Fig.11 Typical ESD pulse based on the Human Body Model
according to MIL-STD 883 C
10
40
Fig.13 Typical ESD pulse based on the Machine Model
according to standard UZW-BO/FQ-B302
developed by Philips
10
td
■
P.O. Box 547
■
Bradford, PA 16701
NPO
50 V
1.2
■
USA
■
1
10
102
103
104
105
106
C (pF)
814-362-5536
■
Fax 814-362-8883
Pulse train, pulse 3a and 3b. A typical pulse form is
given in Fig.15(b). The vehicle power supply voltage Up
is 12 or 24 V. The rise time tr is 5 ns. The pulse voltage
Us is +100 V (pulse 3b, Up = 12 V), +200 V (pulse 3b, Up
= 24 V), -150 V (pulse 3a, Up = 12 V) or -200 V (pulse 3a,
Up = 24 V).The pulse duration td is 0.1 µs.The cycle time
t1 is 100 µs. The pulse train duration t4 is 10 ms (100
pulses). The delay time t5 is 90 ms. The duration of the
test is 1 hour or longer (36000 pulse trains).
MSB664
t1
U
(%)
td
tr
100
90
2
t r1
X7R
200 V
Fig.14 Experimental data on ESD immunity of NP0 and
X7R MLCCs based on Machine Model pulses
according to Philips standard UZW-BO/FQ-B302.
Size 0805 and 1206
2
Ir
0
X7R
25/50 V
1.6
0.8
10-1
Ip1
I
(A)
MSB660
100
90
NPO
200 V
Single pulses, pulse 1 and 2. A typical pulse form is
given in Fig.15(a).The vehicle power supply voltage Up is
12 or 24 V.The rise time tr is in general 1 µs.The pulse
voltage Us is -100 V (pulse 1, Up = 12 V) or -200 V (pulse
1, Up = 24 V) or +100 V (pulse 2, Up = 12 V, 24 V).The pulse
duration td is 50 µs. The cycle time t1 is
0.5 - 5 s.The number of pulses is 5000 or higher.
X7R
500 V
ESD
voltage (kV)
MSB662
4
In Fig.12 the ESD immunity has been given for NP0 and
X7R type MLCCs tested according to the MIL-STD
883C standard.
I
(%)
NPO
500 V
■
www.koaspeer.com
Automotive pulses
Other pulses relevant for MLCC applications are the
so-called automotive pulses. Compared with ESD pulses,
treated in the previous section, these pulses are
characterized by slower rise times but higher energies.
Automotive pulses are generally characterized in the
international standards DIN 40839, ISO/TR 7637/1 and
SAE J1113. The immunity tests in these standards are
aimed at determining the ability of various electrical
devices to withstand transients that normally occur in
motor vehicles.Transients can be added to the standard
electrical voltage of 12 V or 24 V, caused, for example, by
the release of stored energy during start and turn off of
vehicles.These are general tests, not for MLCCs only.
Us
Up 10
0
t
t2
(a)
U
(%)
MSB665
100
Us
UB
0
t
t1
No defects were found after testing NP0 and X7R
products with a minimum rated voltage of 50 V and a
size of 1206 and larger.This concerns test pulses no. 1,
2, 3a and 3b, mentioned in the various standards.
With the exception of the DC offset (Up in Fig.15),
these pulses were produced with a Schaffner NSG
500/B14 pulse generator.
t4
t5
(b)
MSB666
100
90
U
(%)
Us
The maximum absolute value of the peak voltage
mentioned in the standard documents was 200 V for
single pulses and pulse trains. In our tests we were able
to over stress samples to a 350 to 500 V level and an
average dV/dt of 500 V/µs, without failure.
10
0
The automotive pulses according to DIN 40839 part I,
ISO/TR 7637/1 and SAE J1113 are characterized as
follows:
BolivarDrive
Drive
Bolivar
Bolivar
Drive
t
tr
td
Fig.15 Automotive pulses. (a) single pulses, pulses 1 and 2;
(b) pulse train, pulses 3a and 3b
9
MSC962
2.0
Human Body Model pulses according to
MIL-STD 883C.
These pulses simulate the discharge of an electricallyloaded human body by using a discharge circuit with a
capacitor of 100 pF, a resistance of 1.5 kΩ and an
inductance of 7.5 µH (not specified in the standard but
derived from pulse characteristics).
Machine Model pulses according to the Philips
standard UZW-BO/FQ-B302.
These pulses simulate the discharge of an electricallyloaded machine by using a discharge circuit with a
capacitor of 200 pF, a resistance of only 25 Ω and an
inductance of 2.5 µH (not specified in the standard, but
derived from pulse characteristics).
Figure 11 shows a typical pulse form. The pulse rise
time tri is less than 10 ns, the delay time td is 150 ns, the
peak current at 2 kV is 1.25 A (short circuit) and at 4 kV
it is 2.5 A (short circuit). Ir caused by ringing must be
less than 5% of the peak current. The voltage level is
0 to 10 kV.
A typical pulse form is presented in Fig.13. It is the
calibration-current waveform for a charged voltage of
400 V.The first peak current Ip1 is 3 A.The damping Ip1/Ip2
is 1.4. The resonant frequency is 7.1 MHz. The voltage
level is 0 to 2 kV.
In Fig.12 the ESD immunity has been given for NP0 and
X7R type MLCCs tested according to the MIL-STD
883C standard.
I
(%)
MSB660
I
(A)
0
t
td
t r1
Fig.11 Typical ESD pulse based on the Human Body Model
according to MIL-STD 883 C
MSC961
10
X7R
500 V
ESD voltage
(kV)
0
8
6
X7R
25/200 V
4
NPO
50/200 V
2
0
10-1
1
10
102
103
Figure 14 gives data on ESD immunity for NP0 and X7R
type MLCCs
tested according to the Machine Model
4
0
40
160 500
200
standard. Several
types
of 8025, 50,120200 and
t (ns) V products
in NP0 and X7R ceramic were measured. All products
Fig.13 Typical ESD pulse based on the Machine Model
were resistant to at least 2.5 kV HBM and 1 kV MM
according to standard UZW-BO/FQ-B302
pulses. 0805 and 1206 products showed the same
developed by Philips
behaviour.
The advantage of the high-voltage products (rated
voltage 500 V) is clear when comparing the 50 V data
with the 500 V data since higher immunity levels are
obtained for the high-voltage products.
The experimental data presented in Figs 12 and 14 show
that energy dissipation is highest (and the ESD
breakdown voltage level is lowest) when the tested
MLCC has a capacitance close to the capacitance of the
discharge circuit. This is in line with our model
calculations. Moreover, the figures show that the X7R
data seems to be shifted somewhat to higher
capacitance values compared with the NP0 data. This
may be due to the voltage-dependent capacitance of
X7R products, which causes a lowering of capacitance at
higher DC voltage levels.Additionally, the ESD immunity
of X7R products is found to be lower than that of NP0
products with the same capacitance and rated voltage.
NPO
500 V
104
105
106
C (pF)
Fig.12 Experimental data on ESD immunity of NP0 and
X7R MLCCs of size 0805 and 1206 based on
Human Body Model pulses according to MIL-STD
883C
X7R
200 V
NPO
50 V
Pulse train, pulse 3a and 3b. A typical pulse form is
given in Fig.15(b). The vehicle power supply voltage Up
is 12 or 24 V. The rise time tr is 5 ns. The pulse voltage
Us is +100 V (pulse 3b, Up = 12 V), +200 V (pulse 3b, Up
= 24 V), -150 V (pulse 3a, Up = 12 V) or -200 V (pulse 3a,
Up = 24 V).The pulse duration td is 0.1 µs.The cycle time
t1 is 100 µs. The pulse train duration t4 is 10 ms (100
pulses). The delay time t5 is 90 ms. The duration of the
test is 1 hour or longer (36000 pulse trains).
1.2
1
10
102
103
104
105
106
C (pF)
Fig.14 Experimental data on ESD immunity of NP0 and
X7R MLCCs based on Machine Model pulses
according to Philips standard UZW-BO/FQ-B302.
Size 0805 and 1206
2
Ip2
10
X7R
25/50 V
1.6
0.8
10-1
Ip1
2
36.8
NPO
200 V
Single pulses, pulse 1 and 2. A typical pulse form is
given in Fig.15(a).The vehicle power supply voltage Up is
12 or 24 V.The rise time tr is in general 1 µs.The pulse
voltage Us is -100 V (pulse 1, Up = 12 V) or -200 V (pulse
1, Up = 24 V) or +100 V (pulse 2, Up = 12 V, 24 V).The pulse
duration td is 50 µs. The cycle time t1 is
0.5 - 5 s.The number of pulses is 5000 or higher.
X7R
500 V
ESD
voltage (kV)
MSB662
4
Ir
100
90
NPO
500 V
MSB664
t1
U
(%)
td
tr
100
90
Automotive pulses
Other pulses relevant for MLCC applications are the
so-called automotive pulses. Compared with ESD pulses,
treated in the previous section, these pulses are
characterized by slower rise times but higher energies.
Us
Up 10
0
Automotive pulses are generally characterized in the
international standards DIN 40839, ISO/TR 7637/1 and
SAE J1113. The immunity tests in these standards are
aimed at determining the ability of various electrical
devices to withstand transients that normally occur in
motor vehicles.Transients can be added to the standard
electrical voltage of 12 V or 24 V, caused, for example, by
the release of stored energy during start and turn off of
vehicles.These are general tests, not for MLCCs only.
t
t2
(a)
U
(%)
MSB665
100
Us
UB
0
t
t1
No defects were found after testing NP0 and X7R
products with a minimum rated voltage of 50 V and a
size of 1206 and larger.This concerns test pulses no. 1,
2, 3a and 3b, mentioned in the various standards.
t4
t5
(b)
MSB666
100
90
U
(%)
With the exception of the DC offset (Up in Fig.15),
these pulses were produced with a Schaffner NSG
500/B14 pulse generator.
Us
The maximum absolute value of the peak voltage
mentioned in the standard documents was 200 V for
single pulses and pulse trains. In our tests we were able
to over stress samples to a 350 to 500 V level and an
average dV/dt of 500 V/µs, without failure.
10
0
The automotive pulses according to DIN 40839 part I,
ISO/TR 7637/1 and SAE J1113 are characterized as
follows:
td
Fig.15 Automotive pulses. (a) single pulses, pulses 1 and 2;
(b) pulse train, pulses 3a and 3b
8
Bolivar Drive
■
P.O. Box 547
■
Bradford, PA 16701
■
t
tr
USA
■
814-362-5536
■
Fax 814-362-8883
■
www.koaspeer.com
9
ANNEX Application Questionnaire for DC, AC & Pulse Testing of MLCCs
Surge pulses
Surge pulses are high-energy pulses caused by the
switching of inductive or capacitive loads and (of less
importance to multilayer capacitors) lightning. A
standardized pulse is the so-called 1.2/50 µs surge
pulse.The pulse rise time is roughly 1 µs, comparable to
the single automotive pulses treated above. The pulse
duration is 50 µs. The voltage level, up to 4 kV, is much
higher than the voltage level used in automotive pulses.
The 1.2/50 µs surge pulses are standardized according
to IEC 1000-4-5.
We have found a clear relationship between the
breakdown peak voltage and the dielectric thickness of
the measured products.
MSC963
4
surge
breakdown
level (kV)
3
1
Maximum peak voltage:
Maximum peak-to-peak voltage:
Maximum RMS voltage:
Maximum peak current:
Maximum peak-to-peak current:
Frequency (main period)
X7R
10-1
1
102
10
C (nF)
surge
breakdown
level (kV)
.Vp
.Vp-p
.VRMS RMS
.Ip
.Ip-p
.Hz
3
NPO
2
1
U
0
0
20
0.5
t2
40
60
80
100
dielectric thickness (µm)
Fig.18 Instant breakdown peak voltage as a function of
capacitance for NP0 and X7R products subjected to
1.2/50 µs surge pulses.The straight line shows the
result of a linear fit of the combined NP0 and X7R
data
0.3
0.1
0
Fig.16 1.2/50 µs surge pulses
10
USA
■
814-362-5536
■
Fax 814-362-8883
■
www.koaspeer.com
. . . .V
. . . .V
. . . .V
. . . .V
. . . .V/µs
. . . .A/µs
4. Climatic requirements
Ambient temperature: . . . . . . . .minimum . . . . . . . . .°C
Average . . . . . . . . .°C
maximum . . . . . . . .°C
5. Remarks
..................................................................................
..................................................................................
6. Name: . .
Company:
Address: .
.........
Tel.: . . . . .
Fax: . . . . .
Email: . . .
t
■
t
t
3. Pulse application
Automotive pulses according to DIN 40839
pulse number:
1/2/3a/3b . .voltage level: . . . . . . .
ESD pulses according to MIL-STD 883C Human Body Model
number of pulses: . . . . . . . . . .voltage level: . . . . . . .
ESD pulses according to Machine Model
number of pulses: . . . . . . . . . .voltage level: . . . . . . .
Surge pulses according to IEC 1000-4-5, type 1.2/50 µs
number of pulses: . . . . . . . . . .voltage level: . . . . . . .
General
number of pulses: . . . . . . . . . .maximum dV/dt: . . . .
maximum current: . . . . . . . . . .A maximum dI/dt: . . .
X7R
Bradford, PA 16701
.Vp
.Vp-p
.VRMS
.Ip
.Ip-p
.Hz
Current waveform
MSC964
4
0.9
■
Continuous
operation
.........
.........
.........
.........
.........
.........
103
1.0
P.O. Box 547
Starting up/
intermittent
.........
.........
.........
.........
.........
.........
Fig.17 Instant breakdown peak voltage as a function of
capacitance for NP0 and X7R products treated with
1.2/50 µs surge pulses. NP0 products: rated voltage
50, 200 and 500 V, size 1206; X7R products: rated
voltage 16, 25, 200 and 500 V, sizes 0805, 1206
and 1812. The straight line shows the result of a
linear fit of the combined NP0 and X7R data
MSB667
■
AgPd/NiSn
NP0/X7R
0402/0603/0805/1206/1210/1808/1812/2220
Voltage waveform
0
10-2
Figures 17 and 18 give the surge immunity level of the
products referred to as a function of capacitance and
dielectric thickness respectively. Long-term treatment is
expected to result in somewhat lower immunity levels.
The usual picture arises from the initial breakdown
experiments, also seen after DC and AC load i.e. higher
immunity levels are obtained
- for lower capacitance products,
- for higher rated voltage MLCCs (greater dielectric
thickness),
- for NP0 products compared with X7R products with
identical dielectric thickness.
t1
Termination:
Dielectric material:
Size:
2. Application
2
NP0 products of rated voltage 50, 200 and 500 V and
size 1206, and X7R products of rated voltage 16, 25, 50,
200 and 500 V and size 0805, 1206 and 1812 were
treated with a 1.2/50 µs surge pulse. The pulses were
generated by an EM TEST USC 500 immunity-test
generator.
30%
max
1. General
Capacitance: . . . . . . . . . . . . . . . . .pF
Tolerance: . . . . . . . . . . . . . . . . . . .%
Rated voltage . . . . . . . . . . . . . . . .V
NPO
So called 1.2/50 µs pulses are given with the typical
pulse form shown in Fig.16.The front time t1 equals 1.67
x t3 = 1.2 µs, the time to half value t2 is 50 µs, the
voltage U ranges from 0 to 4 kV and the pulse
repetition frequency is 160 ms.
t3
Please complete in as much detail as you can.
..
.
..
..
..
..
..
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
11
Surge pulses
Surge pulses are high-energy pulses caused by the
switching of inductive or capacitive loads and (of less
importance to multilayer capacitors) lightning. A
standardized pulse is the so-called 1.2/50 µs surge
pulse.The pulse rise time is roughly 1 µs, comparable to
the single automotive pulses treated above. The pulse
duration is 50 µs. The voltage level, up to 4 kV, is much
higher than the voltage level used in automotive pulses.
The 1.2/50 µs surge pulses are standardized according
to IEC 1000-4-5.
So called 1.2/50 µs pulses are given with the typical
pulse form shown in Fig.16.The front time t1 equals 1.67
x t3 = 1.2 µs, the time to half value t2 is 50 µs, the
voltage U ranges from 0 to 4 kV and the pulse
repetition frequency is 160 ms.
NP0 products of rated voltage 50, 200 and 500 V and
size 1206, and X7R products of rated voltage 16, 25, 50,
200 and 500 V and size 0805, 1206 and 1812 were
treated with a 1.2/50 µs surge pulse. The pulses were
generated by an EM TEST USC 500 immunity-test
generator.
Figures 17 and 18 give the surge immunity level of the
products referred to as a function of capacitance and
dielectric thickness respectively. Long-term treatment is
expected to result in somewhat lower immunity levels.
The usual picture arises from the initial breakdown
experiments, also seen after DC and AC load i.e. higher
immunity levels are obtained
- for lower capacitance products,
- for higher rated voltage MLCCs (greater dielectric
thickness),
- for NP0 products compared with X7R products with
identical dielectric thickness.
ANNEX Application Questionnaire for DC, AC & Pulse Testing of MLCCs
We have found a clear relationship between the
breakdown peak voltage and the dielectric thickness of
the measured products.
MSC963
4
surge
breakdown
level (kV)
3
NPO
Maximum peak voltage:
Maximum peak-to-peak voltage:
Maximum RMS voltage:
Maximum peak current:
Maximum peak-to-peak current:
Frequency (main period)
2
1
X7R
0
10-2
1
102
10
C (nF)
surge
breakdown
level (kV)
3
NPO
2
0
0
t3
t1
Fig.16 1.2/50 µs surge pulses
10
20
40
60
80
100
dielectric thickness (µm)
Fig.18 Instant breakdown peak voltage as a function of
capacitance for NP0 and X7R products subjected to
1.2/50 µs surge pulses.The straight line shows the
result of a linear fit of the combined NP0 and X7R
data
0
30%
max
t
Continuous
operation
.........
.........
.........
.........
.........
.........
.Vp
.Vp-p
.VRMS
.Ip
.Ip-p
.Hz
.Vp
.Vp-p
.VRMS RMS
.Ip
.Ip-p
.Hz
Current waveform
t
t
MSC964
4
U
0.1
Starting up/
intermittent
.........
.........
.........
.........
.........
.........
Fig.17 Instant breakdown peak voltage as a function of
capacitance for NP0 and X7R products treated with
1.2/50 µs surge pulses. NP0 products: rated voltage
50, 200 and 500 V, size 1206; X7R products: rated
voltage 16, 25, 200 and 500 V, sizes 0805, 1206
and 1812. The straight line shows the result of a
linear fit of the combined NP0 and X7R data
1
0.3
AgPd/NiSn
NP0/X7R
0402/0603/0805/1206/1210/1808/1812/2220
103
1.0
t2
Termination:
Dielectric material:
Size:
Voltage waveform
10-1
X7R
0.5
1. General
Capacitance: . . . . . . . . . . . . . . . . .pF
Tolerance: . . . . . . . . . . . . . . . . . . .%
Rated voltage . . . . . . . . . . . . . . . .V
2. Application
MSB667
0.9
Please complete in as much detail as you can.
3. Pulse application
Automotive pulses according to DIN 40839
pulse number:
1/2/3a/3b . .voltage level: . . . . . . .
ESD pulses according to MIL-STD 883C Human Body Model
number of pulses: . . . . . . . . . .voltage level: . . . . . . .
ESD pulses according to Machine Model
number of pulses: . . . . . . . . . .voltage level: . . . . . . .
Surge pulses according to IEC 1000-4-5, type 1.2/50 µs
number of pulses: . . . . . . . . . .voltage level: . . . . . . .
General
number of pulses: . . . . . . . . . .maximum dV/dt: . . . .
maximum current: . . . . . . . . . .A maximum dI/dt: . . .
. . . .V
. . . .V
. . . .V
. . . .V
. . . .V/µs
. . . .A/µs
4. Climatic requirements
Ambient temperature: . . . . . . . .minimum . . . . . . . . .°C
Average . . . . . . . . .°C
maximum . . . . . . . .°C
5. Remarks
..................................................................................
..................................................................................
6. Name: . .
Company:
Address: .
.........
Tel.: . . . . .
Fax: . . . . .
Email: . . .
Bolivar Drive
..
.
..
..
..
..
..
■
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
P.O. Box 547
.
.
.
.
.
.
.
.
.
.
.
.
.
.
■
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Bradford, PA 16701
.
.
.
.
.
.
.
.
.
.
.
.
.
.
■
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
USA
.
.
.
.
.
.
.
.
.
.
.
.
.
.
■
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
814-362-5536
.
.
.
.
.
.
.
.
.
.
.
.
.
.
■
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Fax 814-362-8883
.
.
.
.
.
.
.
.
.
.
.
.
.
.
■
.
.
.
.
.
.
.
11
www.koaspeer.com
YAGEO - A GLOBAL COMPANY
ASIA
China, Dongguan
Tel. +86 769 791 0053
Fax. +86 769 772 0295
China, Suzhou
Tel. +86 512 825 5568
Fax. +86 512 825 5386
Korea, Seoul
Tel. +82 2 515 0783
Fax. +82 2 3444 3979
Malaysia, Prai Penang
Tel: +60 4 397 3317
Fax: +60 4 397 3272
Malaysia, Puchong
Tel: +60 3 5882 2864
Fax: +60 3 5882 8700
Hong Kong
Tel. +852 2793 3130
Fax. +852 2763 6501
Japan, Tokyo
Tel. +81 3 5833 3331
Fax. +81 3 5833 3116
Singapore
Tel. +65 6244 7800
Fax. +65 6244 4943
Taiwan, Taipei
Tel. +886 2 2917 7555
Fax. +886 2 2917 0148
EUROPE
Benelux, Roermond
Tel. +31 475 385 357
Fax. +31 475 385 589
Finland, Espoo
Tel. +358 9 2707 5851
Fax. +358 9 2707 5852
Hungary, Budapest
Tel. +36 30 3777 441
Fax. +36 94 517 701
Italy, Milan
Tel. +39 02 2411 3055
Fax. +39 02 2411 3051
France, Paris
Tel. +33 1 55 51 84 00
Fax. +33 1 55 51 84 24
Germany, Hamburg
Tel. +49 4121 870-0
Fax. +49 4121 870-297
UK, Leatherhead
Tel. +44 1372 364500
Fax. +44 1372 364567
Spain, Barcelona
Tel. +34 93 317 2503
Fax. +34 93 302 3387
Sweden, Stockholm
Tel. +46 8514 933 55
Fax. +46 8514 933 51
Russia, Moscow
NORTH AMERICA
U.S.A., Addison TX
Tel. +1 214 561 2020
Fax. +1 214 561 2019
www.yageo.com
Printed in The Netherlands
Document order number: 9398 084 34011
Date of release: October 2002
www.yageo.com