APPLICATION NOTES FOR TANTALUM CAPACITORS

APPLICATION NOTES FOR TANTALUM CAPACITORS
1. GENERAL APPLICATION CLASS
Solid tantalum capacitors are usually applied in circuits where the AC component is small compared to the
DC component. Typical uses known to KEMET Electronics
include blocking, by-passing, decoupling, and filtering.
They are also used in timing circuits. If two of these polar
capacitors are connected “back-to-back” (i.e., negative-tonegative or positive-to-positive), the pair may be used in
AC applications (as a non-polar device).
2. STORAGE CONDITIONS
Capacitors may be stored without applied voltage
over the operating temperature range specified in the catalogs for each Series. The range is from -55 to +125° C for
all Series.
Tantalum capacitors do not lose capacitance from the
“de-forming” effect as do liquid-electrolytic capacitors.
Storage at high temperature may cause a small, temporary increase in leakage current (measured under standard conditions), but the original value is usually restored
within a few minutes after application of rated voltage.
Series which are not hermetically sealed exhibit
reversible changes in parameters with respect to relative
humidity (RH). Capacitance increases with increasing
humidity. The limiting change, reached upon establishment of equilibrium with the environment, is approximately -5% to +12% over the range from 25% to 95% RH,
referred to the standard 50% RH. The amount of change
is dependent upon size (capacitance and voltage rating,
ie: CV product); small sizes might change no more than
±5%. Equilibrium at such extremes is seldom attained by
plastic-cased capacitors, and the change in capacitance is
consequently less. The rate of response to humidity
changes increases with increasing temperature.
Dissipation factor also increases with increasing RH. The
limiting change, at equilibrium with 95% RH, is approximately 50%.
DC leakage current may rise upon exposure to a
combination of high temperature and high humidity, but is
normally restored by voltage conditioning under standard
conditions. The increase will be greater than that experienced under temperature influence alone because of conduction through absorbed water.
Hermetically-sealed and non-hermetic Series may be
affected by absorption of water on external insulating surfaces. The water film may also attract a layer of dust from
the air, increasing the effect. The most sensitive parameter is leakage current.
3. POLARITY
These capacitors are inherently polar devices and
may be permanently damaged or destroyed if connected
with the wrong polarity. The positive terminal is identified
on the capacitor body by a polarity mark and the capacitor
body may include an obvious geometrical shape.
However, some Series contain two capacitors connected
(negative-to-negative) to form “non-polar” capacitors.
Rated voltage (see para. 8) may be applied to these
Series in either direction.
4. OPERATING ENVIRONMENT
Most of the discussion under “Storage Conditions”
will apply also when capacitors are operated within the
applicable electrical ratings described below. The temporary increase in leakage current (at standard conditions)
following elevated-temperature exposure is not observed,
however, if the capacitors are operated with adequate DC
voltage applied.
5. CAPACITANCE
Capacitance is measured at 120 Hz and 25° C with
up to 1 volt rms applied. Note that, in either operation,
peak AC plus DC bias must not exceed either rated voltage (normally polarized) or 15% of rated voltage in the
reverse direction at 25°C. Measurement circuits are of
high impedance, however, and under these conditions 1
volt rms may be applied even to 6 volt capacitors (23%
peak reversal) without a DC bias. DC bias is thus normally not used, except when rated voltage is below 6 volts and
the AC signal level exceeds 0.3 vrms. However,
MIL-C-39003 provides for up to 2.2 volts DC. DC bias is
not commonly used at room temperature, but is more commonly used at elevated temperatures.
DC bias causes a small reduction in capacitance, up
to about 2% when full rated voltage is applied as bias. DF
is also reduced by the presence of DC; rated voltage may
cause a decrease in DF of about 0.2% (e.g., a decrease
from 3.6 to 3.4% DF).
Capacitance changes very little below 1 kHz but
decreases more noticeably at higher frequencies. Larger
capacitance values decline more rapidly than small ratings. The effect of frequency upon capacitance is shown in
Figure 1.
1.0
Reference
1.0 at 120Hz
0.9
100
1K
10K
Frequency - Hertz
Figure 1.
Normal Effect of Frequency upon Capacitance
KEMET Electronics Corporation, P.O. Box 5928, Greenville, S.C. 29606 (864) 963-6300
75
Tantalum Application Notes
KEMET solid tantalum capacitors are identified by the
initial “T,” followed by a unique “Series” number; for example, T110, T322, T350, etc. Each Series denotes a general physical form and type of encapsulation, as well as limits on dimensions and certain electrical characteristics
under standard conditions of 25°C, 50% relative humidity,
and one atmosphere pressure. Specific requirements are
set forth in the respective Product Series in this catalog. All
Military products are 100% electrically screened for the
parameters shown in the respective product section. For
non-military product, all series are 100% screened for
leakage, capacitance and dissipation factor. All Series are
inspected to electrical limits using a minimum .1% AQL
sampling plans, according to the Military Standard
MIL-STD-105, even after 100% testing. This sampling
plan, to the best of KEMET Electronics’ knowledge, meets
or exceeds the generally accepted industry standard for
similar products. KEMET capacitors may also be supplied,
with prior agreement, to meet specifications with requirements differing from those of KEMET catalogs. Reference
ESR values are provided but are NOT 100% screened
These Notes apply generally to all KEMET solid tantalum capacitors and illustrate typical performance under
normal application conditions, except where noted.
Certain Series will respond differently to various environmental conditions. For example, hermetically sealed
series are relatively immune to humidity effects, while
plastic-encased series are not. The intent of these Notes
is not to delineate such differences but to provide generalized information concerning performance characteristics.
Multiplier of 120Hz
Capacitance
INTRODUCTION
APPLICATION NOTES FOR TANTALUM CAPACITORS
Capacitance Value
Capacitance typically changes with temperature
according to the curve of Figure 2.
percent in this calculation). Another expression, rarely
ESR
used, is the “power factor,” or Power factor is cos u,
.
while DF is ctn u.
+20
7. DC LEAKAGE (DCL)
DC leakage is affected by voltage to a much larger
extent, and this effect can frequently be used to advantage
in circuits where only very low leakage currents can be tolerated. Typical response of DCL to applied voltage is illustrated in Figure 4.
+10
0
-10
-20
Figure 2.
Typical Effect of Temperature upon Capacitance
6. DISSIPATION FACTOR (DF)
DF is measured at 120 Hz and 25° C with up to 1 volt
rms applied. Note that, in either operation, peak AC plus DC
bias must not exceed either rated voltage (normally polarized) or 15% of rated voltage in the reverse direction at
25°C. Measurement circuits are of high impedance, however, and under these conditions 1 volt rms may be applied
even to 6 volt capacitors (23% peak reversal) without a DC
bias. DC bias is thus normally not used, except when rated
voltage is below 6 volts and the AC signal level exceeds 0.3
vrms. However, MIL-C-39003 provides for up to 2.2 volts
DC. DC bias is not commonly used at room temperature,
but is more commonly used at elevated temperatures.
Dissipation Factor (DF) is a useful low-frequency
measure of the resistive component in capacitors. It is the
ratio of the unavoidable resistance to the capacitive reactance, usually expressed in percent. DF increases with
temperature above +25° C and may also increase at lower
temperatures. Unfortunately, one general limit for DF cannot be specified for all capacitance/voltage combinations,
nor can response to temperature be simply stated.
Catalogs for the respective series list DF limits under various conditions.
Dissipation factor increases with increasing frequency as would be expected from the decreasing capacitive
reactance. DF is not a very useful parameter above about
1 kHz. The DF of larger capacitance values increases
more rapidly than that of smaller ratings. Figure 3 shows
typical effect of frequency on DF.
Multiplier of 120 Hz D. F.
20.0
10.0
0.1
0.1
0.01
0.01
0.0001
0.001
00
10
10
20
20
30
30
40
40
50
50
6060
70
70
Temperature -°C
8080
100110
110
9090 100
Percentage of Rated Voltage
Figure 4.
Typical Range of DC Leakage as
a Function of Applied Voltage
DC leakage current (DCL) increases with increasing
temperature according to the typical curve of Figure 5.
10.0
Reference 1.0
at + 25°C
1.0
0.1
-60 -40 -20
1K
10K
Frequency - Hertz
Figure 3.
0
+20 +40 +60 +80 +100 +125
Operating Temperature –˚C
1.0
100
Normal Effect of Frequency upon
Dissipation Factor
DC bias causes a small reduction in capacitance, up
to about 2% when full rated voltage is applied, as bias. DF
is also reduced by the presence of DC bias. Rated voltage
may cause a decrease in DF of about 0.2% (e.g., a
decrease from 3.6 to 3.4% DF).
DF is defined as
ESR
c and is also referred to occa-
sionally, as tan d or “loss tangent.” The “Quality
Factor,” Q, is the reciprocal of DF (DF is not expressed in
76
Multiplier of DC Leakage Current
Operating Temperature °C
1.0
Multiplier of DC Leakage Current
0 +20 +40 +60 +80 +100 +120
Multiplier of DC Leakage Current
1.0
-80 -60 -40 -20
Figure 5.
Typical Effect of Temperature
upon DC Leakage Current
Leakage current is measured at a rated voltage
through +85°C and may also be measured at +125°C with
2/3 of rated voltage applied.
8. RATED VOLTAGE
This term refers to the maximum continuous DC
working voltage permissible at temperatures of +85° C or
below. The lower operating temperature is specified as
-55° C. Operation above +85° C is permissible, with
reduced working voltage. Typical working voltage reduction is to 2/3 of rated voltage at +125° C.
KEMET Electronics Corporation, P.O. Box 5928, Greenville, S.C. 29606 (864) 963-6300
APPLICATION NOTES FOR TANTALUM CAPACITORS
Percent of
Working Voltage
9. WORKING VOLTAGE
This is the maximum recommended peak DC operating voltage for continuous duty at or below 85°C without
DC voltage surges or AC ripple superimposed. No voltage
derating is required below 85°C. Capacitors may be operated to 125°C with working voltage linearly derated to 2/3
of the 85°C rating at 125°C as shown in Figure 6.
Total impedance of the capacitor is the vector sum of
capacitive reactance (XC) and ESR, below resonance;
above resonance total impedance is the vector sum of
inductive reactance (XL) and ESR.
ESR
θ
150
125
100
75
50
25
0
-75
XC =
-50
-25
0
+25
+50
+75
+100 +125
Temperature -°C
Working Voltage Change with Temperature
Figure 6.
10. SURGE VOLTAGE
Surge voltage is defined as the maximum voltage to
which the capacitor should be subjected under transient
conditions, including peak AC ripple and all DC transients.
δ
1 ohm
2πfC
Ζ
XC
where:
f = frequency, Hertz
C = capacity, Farad
Figure 7a
Total Impedance of the Capacitor
Below Resonance
DC Working
15/
Voltage @ 85°C 2 3 4 6 10 16 20 25 35 50 60 75 100 125
2.6 4 5.3 8 13 20 26 33 46 65 78 98 130 140
TABLE 1
11. REVERSE VOLTAGE
Although these are polar capacitors, some degree of
transient voltage reversal is permissible, as seen below.
The capacitors should not be operated continuously in
reverse mode, even within these limits.
Temperature, °C.
Percentage of
Rated Voltage
+25
+85
+125
15
5
1
TABLE 2
Ζ
Surge Voltage Ratings
A typical surge voltage test is performed at +85°C
with the applicable surge voltage per Table 1. The surge
voltage is applied for 1000 cycles of 30 seconds on voltage through a 33 ohm series resistor and 30 seconds off
voltage with the capacitor discharged through a 33 ohm
resistor. Upon completing the test, the capacitors are
allowed to stabilize at room temperature. Capacitance, DF,
and DCL are then tested:
1. The DCL should not exceed the initial 25°C limit.
2. The capacitance should be within ±10% of initial
value.
3. The DF should not exceed the initial 25°C limit.
XL = 2πfL
XL
where:
f = frequency, Hertz
L = inductance, Henries
δ
ESR
Figure 7b
Total Impedance of the Capacitor
Above Resonance
To understand the many elements of a capacitor, see
Figure 8.
L
RS
C
RL
Reverse Voltage Ratings
12. EQUIVALENT SERIES RESISTANCE (ESR)
Equivalent Series Resistance (ESR) is the preferred
high-frequency statement of the resistance unavoidably
appearing in these capacitors. ESR is not a pure resistance, and it decreases with increasing frequency. Typical
ESR limits are established in each specific product series.
However, the ESR limits provided are for reference only,
and are not necessarily the actual value that a particular
Series product will attain.
θ
Cd
Figure 8.
Rd
The Real Capacitor
A capacitor is a complex impedance consisting of
many series and parallel elements, each adding to the
complexity of the measurement system.
L — Represents lead wire and construction inductance. In most instances (especially in solid tantalum and
KEMET Electronics Corporation, P.O. Box 5928, Greenville, S.C. 29606 (864) 963-6300
77
Tantalum Application Notes
Surge Voltage
@ 85°C
APPLICATION NOTES FOR TANTALUM CAPACITORS
monolithic ceramic capacitors) it is insignificant at the
basic measurement frequencies of 120 and 1000 Hz.
RS — Represents the actual ohmic series resistance
in series with the capacitance. Lead wires and capacitor
electrodes are contributing sources.
RL — Capacitor Leakage Resistance. Typically it can
reach 50,000 megohms in a tantalum capacitor. It can
exceed 1012 ohms in monolithic ceramics and in film
capacitors.
Rd — The dielectric loss contributed by dielectric
absorption and molecular polarization. It becomes very
significant in high frequency measurements and applications. Its value varies with frequency.
Cd — The inherent dielectric absorption of the solid
tantalum capacitor which typically equates to 1-2% of the
applied voltage.
As frequency increases, Xc continues to decrease
according to its equation above. There is unavoidable
inductance as well as resistance in all capacitors, and at
some point in frequency, the reactance ceases to be
capacitive and becomes inductive. This frequency is called
the self-resonant point. In solid tantalum capacitors, the
resonance is damped by the ESR, and a smooth, rather
than abrupt, transition from capacitive to inductive reactance (XL = 2πfL) follows.
10VDC Rated
Impedance
ESR
(Ohms)
100
d
µF Fd
µ
d
10 33
µF d
0
F
10 0 µ
22
10
10
1
35VDC Rated
100
2.
4.
7
22
10
47
µ
2.7 Fd
µF
d
1 4.7
µF
d
7
Impedance
ESR
(Ohms)
µF
d
µF
d
µF
d
47 µFd
22 µFd
0.1
100
1000
10K
100K
1M
10M
Frequency (Hz)
Figure 11a.
ESR and Impedance vs. Frequency
Typical ESR and Z performance is given for representative capacitor ratings in Figures 9 through 11.
Measured impedance will be affected by the length of lead
wire included. Data for the curves were taken by including
1/2" of each lead wire in the measuring circuit.
Despite the fact that the reactance is entirely inductive above the self-resonance, these capacitors find use
as decoupling devices above 10 MHz. Special designs
have been developed for minimum inductance and are
used above 100 MHz.
ESR and Z are also affected by temperature. At 100
kHz, ESR decreases with increasing temperature. The
amount of change is influenced by the size of the capacitor and is generally more pronounced on smaller ratings.
Multiplier of 100kHz ESR
10
µF
d
33 µFd
100 µFd
0.1
1
220 µFd
100
1000
10K
100K
1M
10M
Frequency (Hz)
Figure 9.
0.1
ESR and Impedance vs. Frequency
-50
-25
0
25
50
75
100
125
Temperature - Degrees Centigrade
Figure 11b Typical Effect of Temperature
on 100 kHz ESR
20VDC Rated
30
10
47
3
d
µF Fd Fd
µ
68 00 µ
1
Impedance
ESR
(Ohms)
1
47
0.3
µF
d
68 µFd
0.1
100
100 µ
Fd
1000
10K
100K
1M
10M
13. POWER DISSIPATION
Permissible power dissipation has been empirically
established for all Series and is listed in each respective
product section.
See pages 6-41 for herm seal, 42-50 for axial and
radial molded, and 61-70 for tantalum dipped.
It is usually most convenient to translate the permissible power into an AC voltage rating. Assuming a sinusoidal waveform, the “ripple voltage” permissible may be
calculated from the impedance and ESR data shown in the
respective product section. However, three criteria must
be observed:
Frequency (Hz)
Figure 10.
78
ESR and Impedance vs. Frequency
KEMET Electronics Corporation, P.O. Box 5928, Greenville, S.C. 29606 (864) 963-6300
APPLICATION NOTES FOR TANTALUM CAPACITORS
E2R
P = I2R= Z2
where: I =
E =
P =
Z =
rms ripple current (amperes)
rms ripple voltage (volts)
power (watts)
impedance at specified frequency
(ohms)
R = equivalent series resistance at
specified frequency (ohms)
Maximum allowable rms ripple voltage may be
determined as follows:
E(max) @
25°C=Z
√ P(max)
R
E(max) =
E(max) @
P(max) =
85°C=0.9 E(max) @ 25°C
125°C=0.4 E(max) @ 25°C
maximum watts shown on
Performance Characteristic pages 5,
42, 49, 58 and 61.
Permissible AC ripple current can be determined by
the following:
P(max)
I rms =
R
√
If two polar capacitors are connected back-to-back,
the pair may be operated on AC without need for DC
bias. The first two criteria above must be observed. If DC
is applied, the sum of DC and peak AC must not exceed,
in either direction, the maximum working voltage specified
for the ambient temperature.
(1) Some KEMET Series provide convenient assemblies of
non-polar pairs. The two negative terminals are connected
internally. It is also permissible to connect the two positive
terminals to form a non-polar pair.
14. LONG-TERM STABILITY
Within the general class of electrolytic capacitors,
solid tantalum capacitors offer unusual stability of the three
important parameters: capacitance, dissipation factor, and
leakage current. These solid-state devices are not subject
to the effects of electrolysis, deforming or drying-out associated with liquid-electrolyte capacitors.
When stabilized for measurement at standard conditions, capacitance will typically change less than ±3% during a 10,000 hour life test +85° C. The same comparative
change has been observed in shelf tests at +25° C extending for 50,000 hours. (Some of this change may stem from
instrument or fixture error.)
Dissipation factor exhibits no typical trend. Data from
10,000 hour life tests at +85° C show that initial limits (at
standard conditions) are not exceeded at the conclusion of
these tests.
Leakage current is more variable than capacitance or
DF; in fact, leakage current typically exhibits a logarithmic
dependence in several respects. MIL-C-39003/1 permits
leakage current (measured at standard conditions) to rise
(1)
by a factor of four over 10,000 hour life tests. Typical
behavior shows a lower rate of change, which may be
negative or positive. Initial leakage currents are frequently
so low (less than 0.1 nanoampere in the smallest CV
capacitors, to about 10 microampere in the largest CV
types) that changes of several orders of magnitude have
no discernable effect on the usual circuit designs.
15. FAILURE MODE
Capacitor failure may be induced by exceeding the
rated conditions of forward DC voltage, reverse DC voltage, surge voltage, surge current, power dissipation, or
temperature. As with any practical device, these capacitors also possess an inherent, although low, failure rate
when operated within the rated condition.
The dominant failure mode is by short-circuit. Minor
parametric drifts (see Section 14 “Long-Term Stability”) are
of no consequence in circuits suitable for solid tantalum
capacitors. Catastrophic failure occurs as an avalanche in
DC leakage current over a short (millisecond) time span.
The failed capacitor, while called “short-circuited”, may
exhibit a DC resistance of 10 to 104 ohm.
If a failed capacitor is in an unprotected low-impedance circuit, continued flow of current through the capacitor may obviously produce severe overheating. This heat
may melt the internal solder (all Series) and the sealing
solder used in hermetic Series. The short-circuit failure
may thereby be converted to an open-circuit failure. If the
circuit does not open promptly, the over-heated capacitor
may damage the circuit board or nearby components.
Protection against such occurrence is obtained by currentlimiting devices or fuses provided by the circuit design.
Fortunately, the inherent failure rate of KEMET solid
tantalum capacitors is low, and this failure rate may be further improved by circuit design. Statistical failure rates are
provided for those capacitors with characters other than
“A” in the next-to-last position of the part number. Relating
circuit conditions to failure rate is aided by the guides in
the section following.
16. RELIABILITY PREDICTION
Three important application conditions largely control
failure rate: DC voltage, temperature, and circuit impedance. Estimates of the respective effects are provided by
the nomograph in Figure 12 and Table 3 following. The
nomograph related failure rate to voltage and temperature
while the table relates failure rate to impedance. These
estimates apply to steady-state DC conditions, and they
assume usage within all other rated conditions.
Standard conditions, which produce a unity failure
rate factor, are rated voltage, +85° C, and 0.1 ohm-per-volt
circuit impedance. While voltage and temperature are
straightforward there is sometimes difficulty in determining
impedance. What is required is the circuit impedance seen
by the capacitor. If several capacitors are connected in
parallel, the impedance seen by each is lowered by the
source of energy stored in the other capacitors. Energy is
similarly stored in series inductors.
Failure rate is conventionally expressed in units of
percent per thousand hours. As a sample calculation, suppose a particular batch of capacitors has a failure rate of
0.5% Khr under standard conditions. What would be predicted failure rate at 0.7 times rated voltage, +60° C and
0.8Ω/V? The nomograph gives a factor of 7 x 10-4, and the
table gives a factor of 0.3. The failure rate estimate is then:
0.5 x 7 x10-4 x 0.3 = 1.05 x 10-4, or 0.0001% Khr
KEMET Electronics Corporation, P.O. Box 5928, Greenville, S.C. 29606 (864) 963-6300
79
Tantalum Application Notes
1. Dissipated power must not exceed the limits
specified for the Series.
2. The positive peak AC voltage plus the DC voltage
must not exceed the maximum working voltage permitted
at the ambient temperature.
3. The negative peak AC voltage, in combination
with the DC voltage, must not exceed the permissible
reverse voltage at the ambient temperature.
The rms ripple voltage limitation imposed by power
dissipation is given by:
APPLICATION NOTES FOR TANTALUM CAPACITORS
125
1.2
10 3
120
1.1
10 2
110
85°C
Rating
1.0
100
90
10 -1
85
10 -2
80
70
60
Connect the temperature and
applied voltage ratio of
interest with a straight edge.
The multiplier of failure rate is
given at the intersection of this
line with the model scale.
10 -3
10 -4
0.8
0.7
0.67
125°C
Rating
0.6
0.5
Applied Voltage Ratio, V
0.9
1.0
Failure Rate Multiplier, F
Operating Temperature-Degrees Celsius
10 1
10 -5
50
40
Given T1 & V1 Read Failure
Rate Multiplier F1
Given T, & F2
Read Reguired Voltage V2
Given F3 & V3
Read Allowable Temp T3
0.4
10 -6
0.3
10 -7
0.2
10 -8
30
25
T
Figure 12.
F
Reliability Nomograph
Circuit Impedance
(ohms/volt)
0.1
0.2
0.4
0.6
0.8
1.0
2.0
3 or greater
TABLE 3
V
0.1
Failure Rate Improvement
(multiplying factors)
1.0
.8
.6
.4
.3
.2
.1
.07
Relationship of Failure Rate to Impedance
Voltage “de-rating” is a common and useful approach
to improved reliability. It can be pursued too far, however,
when it leads to installation of higher voltage capacitors of
much larger size. Inherent failure rate is roughly proportional to CV1.6, where C is capacitance and V is rated voltage. The effect becomes particularly noticeable above 50volt ratings. It is possible to lose more via higher inherent
failure rate than is gained by voltage derating.
The relationships shown are more useful when the
failure rate has been statistically determined for a given
group of capacitors.
Failure rate is statistically determined for each production batch of KEMET High Reliability capacitors, as
described in Specification GR500 Catalog F2956. As
noted above, not all capacitance/voltage rate values are
inherently equal in failure rate. GR500 capacitors are
processed and subjected to 100% reliability testing as a
homogeneous group of one capacitance/voltage value.
Failure rate under standard conditions is available from 1
to 0.001% Khr, depending upon the capacitance/voltage
value.
Several Series are qualified under U.S. military specification MIL-C-39003. Failure rates as low as 0.001%/Khr
are available for all capacitance/voltage values in given
groups under this test program. The specifications and
80
their accompanying Qualified Products Lists should be
consulted for details.
For Series not covered by military specifications, and
internal sampling program is operated by KEMET Quality
Assurance. The confidence level chosen for reporting the
data is 60%. However, the cost of sampling each batch
produced is overwhelmingly prohibitive, and no claim is
made concerning knowledge of failure rate for any particular lot shipped. It is demonstrated that average failure
rate for all commercial Series is between .1 and 1%/Khr at
standard conditions and 60% confidence after 2,000
hours’ testing, +85°C, and rated voltage and ≤ 1 ohm total
series resistance.
17. SURGE CURRENT
All conventional reliability testing is conducted under
steady-state DC voltage. Experience indicates that AC ripple, within the limits prescribed, has little effect on failure
rate. Heavy surge currents are possible in some applications, however. Circuit impedance may be very low (below
the standard 0.1 ohm/volt) or there may be driving inductance to cause voltage “ringing.” Surge current may
appear during turn-on of equipment, for example.
Failure rate under current-surge conditions may not
be predictable from conventional life test data. A surge current test is utilized to ensure against a high frequency of
such failures, and a description is available free of charge.
The test has been adopted for all capacitors under
MIL-C-39003/06/09/10 and KEMET’s GR500 specifications.
18. ENVIRONMENTAL CONSIDERATION
It is not possible to foresee all the conditions to which
capacitors may be subjected. Following is a list of standard tests which every Series will survive. Data may be
available (upon request) under more severe stresses for
certain Series.
• LIFE TEST 85°C OR 125°C, 2000 Hours: When subjected to 2000 hours at 85°C at full rated DC voltage, or
125°C at 2/3 of 85°C voltage, the capacitor shall meet
the following requirements when tested at 25°C:
The DCL shall be within 1.25 times the initial DCL
limit.
Capacitance shall be within ±10% of the initial measured value.
The DF shall not exceed the initial limit.
• SHELF LIFE +85°, 2000 hours. Post test of capacitor
shall meet the following requirements when tested at
25°C:
The DCL shall be within 1.5 times the initial DCL
limit.
Capacitance shall be within ±10% of the initial measured value.
The DF shall not exceed 1.5 times the initial limit.
• LEAD STRENGTH MIL-STD-202 Method 211: Pull test
will be performed as in MIL-STD-202, Method 211. The
following details and exceptions shall apply.
a. Test condition letter—A
b. The body of the capacitor will be securely clamped
during test.
KEMET Electronics Corporation, P.O. Box 5928, Greenville, S.C. 29606 (864) 963-6300
APPLICATION NOTES FOR TANTALUM CAPACITORS
• VIBRATION; HIGH FREQUENCY: Per MIL-STD-202,
Method 204, Condition D, 10 Hz to 2000 Hz.
a. Mounting—Capacitors shall be mounted on a fixture
by the body. Leads shall be supported by rigidly supported terminals.
b. Electrical load conditions—During the test, the specified DC rated voltage shall be applied to the capacitors.
c. Test condition letter—D (20 G).
d. Duration and direction of motion—4 hours in each of
two mutually perpendicular directions (total of 8
hours), one parallel and the other perpendicular to
the axis.
e. Measurements during vibration—During the last
cycle, an electrical measurement shall be made to
determine intermittent operation or open- or short-circuiting. Observations shall also be made to determine intermittent contact or arcing or open- or shortcircuiting. Detecting equipment shall be sufficiently
sensitive to detect any interruption with a duration of
0.5 ms, or greater.
f. Examination after test—Capacitors shall be visually
examined for evidence of mechanical damage.
• SHOCK TEST: Per MIL-STD-202, Method 213. The following details shall apply:
a. Special mounting means—Capacitors shall be rigidly
mounted on a mounting fixture by the body. When
securing leads, care shall be taken to avoid pinching
the heads.
b. Test-condition letter—I (100 G peak). 6 ms. (sawtooth)
c. Measurements and electrical loading during shock—
During the test, observations shall be made to determine intermittent contact or arcing or open- or shortcircuiting. Detecting equipment shall be sufficiently
sensitive to detect any interruption with a duration of
0.5 ms. The DC rated voltage shall be applied to the
capacitors during the test.
d. Examinations after test—Capacitors shall be visually
examined for evidence of arcing, breakdown, and
mechanical damage.
• HUMIDITY LIFE TEST: Capacitors shall be capable of
withstanding 1000 hours at 55°C with an ambient
humidity of 90-95% RH with rated DC voltage applied.
After the capacitors have stabilized for a period of 24
hours at 25°C, they shall meet the following limits:
DCL shall not exceed 5 times the initial limit.
Capacitance shall be within ±10% of the initial value.
DF shall not exceed 2 times the initial limit.
• THERMAL SHOCK—MIL-STD-202, Method 107:
Capacitors shall be subjected to thermal shock in accordance with MIL-STD-202, Method 107, Test Condition A.
M39003 Components tested to MIL-STD-202, Method
107, Condition B. Measurements before and after
cycling are required. Conditioning prior to the first cycle
will be 15 minutes at the following standard inspection
conditions:
a. Relative Humidity—Less than 50%.
b. Ambient Temperature—25°C ±5°C.
c. Final measurements are made after stabilization at
room temperature.
• MOISTURE RESISTANCE—MIL-STD-202, Method 106:
Capacitors shall be tested in accordance with
MIL-STD-202, Method 106 including the following
details:
a. Mounting—The capacitors shall be mounted by normal mounting means
b. Initial Measurements
c. Polarizing and Load Voltage—Not applicable
d. Final measurements—After the final cycle and within
2 to 6 hours after removal of the capacitors from the
humidity chamber, capacitance, dissipation factor,
and DC leakage will be measured.
DCL should not exceed the initial 25°C limit.
Capacitance should be within ±10% of the initial measured value.
DF should not exceed the initial 25°C limit.
• RESISTANCE TO SOLVENTS — MIL-STD-202,
Method 215:
Brushing required after test.
DCL meets limit shown in respective Part Number
Tables.
Capacitance meets applicable tolerance.
DF meets limits shown in respective Part Number
Tables.
No visible damage to case or marking.
• RESISTANCE TO SOLDERING HEAT — MIL-STD-202,
Method 210, Test Condition.
Letter B. (260° for 10 Sec.)
Leads shall be immersed to within 1⁄4 inch of the capacitor body. Capacitance, DF, and DCL should meet original limits shown in respective Part Number Tables.
• SOLDERABILITY — MIL-STD-202, Method 208;
Number of terminations on each capacitor tested: 2.
Depth of insertion in flux and solder to within .125" of
capacitor body.
• FLAMMABILITY — The encapsulant for Molded and
Conformal Coated Product meets or exceeds the following requirements:
Underwriters Lab. UL 94V-0
Oxygen Index per ASTM-D-2863
28% min.
• STABILITY AT LOW AND HIGH TEMPERATURE
-55°C to 125°C: Capacitors will be capable of
withstanding extreme temperature testing at a
succession of continuous steps at +25°C, -55°C, +25°C,
+85°C, +125°C, +25°C, in the order stated. Capacitors
shall be brought to thermal stability at each test
KEMET Electronics Corporation, P.O. Box 5928, Greenville, S.C. 29606 (864) 963-6300
81
Tantalum Application Notes
c. Applied force—3 pounds (1.4 kg).
d. Test Condition letter—C (M39003 Test Condition letter—D)
e. Applied Force—1 pound, 3 bends.
APPLICATION NOTES FOR TANTALUM CAPACITORS
temperature. Capacitance, DF, and DCL are measured
at each test temperature except that DCL is not
measured at -55°C. DC bias of 2.0 ±0.5 vdc is
recommended for the capacitance and DF
measurements.
When measurements are made at the various steps, the
electrical limits for each temperature shall not exceed
the following limits.
Step 1, +25°C, DCL as indicated in original limit;
capacitance within tolerance specified; DF as indicated in original limit
shown in Part Number Tables.
Step 2, -55°C, Capacitance within ±10% of initial
value; ESR, DF within limit shown in
Part Number Tables.
Step 3, +25°C, DCL as indicated in original limit;
capacitance within ±5% of initial
value; ESR, DF within limit shown in
Part Number Tables.
Step 4, +85°C, DCL shall not exceed 10 times original
DCL limit at 25°C. Capacitance shall
be within ±10% of the initial value. DF
shall be within 125% of limits shown in
Part Number Tables. ESR shall be
within limits shown in Part Number
Tables.
Step 5, +125°C, DCL shall not exceed 12.5 times the
original limit at 25°C. Capacitance
shall be within ±12% of initial value.
DF shall be within 150% of limits
shown in Part Number Tables. ESR
shall be within limits shown in Part
Number Tables.
Step 6, +25°C, DCL as indicated in original limit;
capacitance within ±5% of initial
value; ESR, DF as indicated in original limit shown in Part Number
Tables.
Note: M39003 specifies ∆'s and limits by individual slash
sheet.
• DAMP HEAT, STEADY STATE: Meets requirements of
IEC Publication 384-15, method IEC 68-2-3. Climatic
category 55/125/56.
19. MOUNTING
All encapsulated Series fall into two general classes.
The first is provided with leads extending from opposite
ends of the body, generally along the principle axis of the
body (“axial leads”). The second is provided with parallel
leads extending from one side or face of the body (“radial
leads”). With either type, mounting points are normally
provided by the leads themselves.
Axial leads may be used for point-to-point wiring, but
usually, the wires are bent at 90° from the capacitor axis
for insertion through printed circuit (PC) boards. Axial
capacitors supplied on reels for machine insertion will
82
withstand the mechanical stresses of bending and inserting by all popular machines known to KEMET at this time.
Most KEMET axial Series may be supplied on reels to feed
such machines. Radial leads are intended to plug directly
into holes of PC boards. Auto-machines will insert combatible radial capacitor designs, and most KEMET capacitors may be supplied in appropriate reeled forms (ARIS).
With either axial or radial types, attention should be
paid to treatment of the capacitors during mounting and
afterward under service conditions. Difficulty during
mounting usually arises from lead damage or from overheating. Hand soldering technique or, more often, wavesolder machines cause the overheating. The internal cathode connection on most Series is made between solder
and a silver-pigmented paint. If too much heat is applied,
this solder may remelt and degrade the silver-solder interface or cause a direct short-circuit.
KEMET’s hermetically-sealed series has an internal
space into which molten cathode solder may run, depriving the cathode connection and possibly flowing across
the terminals to short-circuit the capacitor from the inside.
It is also possible to remelt the solder which bonds the rim
of the glass-metal seal, causing loss of hermeticity and
possibly a short-circuit. Finally, solder at the exit point of
the positive wire may be remelted with similar effect. This
solder is a high-temperature alloy, however, and it is much
less likely to be melted. (Re-dipping of lead wires is practiced by some users, introducing another hazard of remelting this solder).
Plastic-encased Series have only one site of solder,
the internal cathode connection. The rate of heat transfer
through the plastic is lower than through the metal can of
hermetic Series, but conduction along the negative leadwire to remelt this solder is very similar. There is little internal void within plastic cases, so remelted solder tends to
remain in its original location and solidify when heat is
removed. Short-circuiting is very unlikely, but reliability of
the internal connection may be compromised by leaching
of silver from the paint into the molten solder. The latter
effect degrades the cathode connection in hermetic parts
as well.
All encased capacitors will pass the Resistance to
Soldering Heat Test of MIL-STD-202, Method 210,
Condition B. This test dips each leadwire into molten solder at +260° C for 10 seconds while the capacitor body is
held vertically above the solder. KEMET capacitors will
pass this test when the depth of immersion brings the
capacitor body (or closest external solder joint, if it is closer as in some hermetic Series) to a minimum distance of
0.100 inches from the solder surface. This demonstration
of resistance to solder heat is in accordance with what is
believed to be the industry standard. More severe treatment must be considered reflective of an improper soldering process.
Shown in Figure 13 is a recommended solder wave
profile for both axial and radial leaded solid tantalum
capacitors.
KEMET Electronics Corporation, P.O. Box 5928, Greenville, S.C. 29606 (864) 963-6300
APPLICATION NOTES FOR TANTALUM CAPACITORS
Optimum Solder Wave Profile
Solder Wave Peak
Temperature 260°C
Entrance to Solder Wave
Flux Zone
Exit from Solder Wave
(Time in Wave – 2 to 4 Secs.)
Preheat Zone
Hot Air Debridging
250
225
80°C
to 120°C
Degrees – C°
200
Free
Free
Air
Air
Cool
Cool
175
Exit from
Solder
Machine
Entrance to
In-Line Cleaner
Exit from
In-Line Cleaner
(time in cleaner
may be less)
Bottom Side
Entrance Temperature
Range
to Solder
Machine
150
125
100
Immersion in
Cleaning
Vapor
150°C
Max.
75
Top Side
Nominal
50
25
0
1
2
3
4
5
6
Time (Minutes)
Figure 13.
KEMET Electronics Corporation, P.O. Box 5928, Greenville, S.C. 29606 (864) 963-6300
83