Capacitors

5
5.1 Introduction
A capacitor consists, basically, of two conductors separated by a dielectric. The operation of a capacitor in a
circuit is dependent upon its ability to charge and discharge. When a capacitor charges, an excess of electrons
is accumulated on one plate, and a deficiency of electrons is created on the other plate. Capacitance is deter-
mined by the size of the conductive material (the plates) and their separation (determined by the type and
thickness of the dielectric material). Capacitance is directly proportional to plate size and inversely propor-
tional to the distance between the plates. The unit of capacitance is the farad (F). A capacitance of 1 F results
when a potential of 1 V causes an electric charge of 1 coulomb to accumulate on a capacitor.

5.2 Basic Principles

The value of a parallel-plate capacitor can be found from

x ε ⎡⎣( N − 1) A ⎤⎦
C= × 10−13 (5.1)
d

where
C = Capacitance (F)
ε = Dielectric constant of the insulation
d = Spacing between the plates
N = Number of plates
A = Area of the plates
x = 0.0885 when A and d are in centimeters
The work necessary to transport a unit charge from one plate to another is

e=kg (5.2)

where
e = Volts expressing energy per unit charge
k = Proportionality factor between the work necessary to carry a unit charge between the two plates
and the charge already transported
g = Coulombs of charge already transported
The latter quantity (k) is equal to 1/C, where C is the capacitance in F.
The value of a capacitor can now be calculated from

C = q
--- (5.3)
e

where q = charge (C) and e is found from Equation 5.2.

The energy stored in a capacitor is

81

© 2007 by Taylor & Francis Group, LLC

82 AC Power Systems Handbook

2
CV
W = ---------
- (5.4)
2

where
W = Energy (J)
C = Capacitance (F)
V = Applied voltage (V)
If a direct current is applied to a capacitor, the device will charge to the value of the applied voltage.
After the capacitor is fully charged, it will block the flow of direct current. However, if an ac voltage is
applied, the changing value of current will cause the device to alternately charge and discharge. In a
purely capacitive circuit, the situation shown in Figure 5.1 will exist. The greatest amount of current will
flow when the voltage changes most rapidly; that point occurs at the 0 and 180° positions in the sine wave
where the polarity reverses. At these positions, maximum current is developed in the circuit, as shown. It
is evident by studying the waveform that, in a purely capacitive circuit, voltage will lag current by 90°.
Because all practical circuits contain some resistance, a lag of 0 to 90° may be experienced in practice. Fig-
ure 5.2 illustrates a case in which voltage lags current by 30º.

AC Capacitive
load V
source

(a)

+
Power

Current
Voltage, current, power

90° 270°
0
Time 180° 360°

Voltage

_
(b)

90°

V
(c)

Figure 5.1 A purely capacitive circuit: (a) circuit diagram; (b) representative waveforms; (c) vector representation.

© 2007 by Taylor & Francis Group, LLC

Capacitors 83

R
AC
source V
C

(a)

+
Power
Voltage, current, power

Voltage

0
30° 90° 180° 210° 270° 360°
Time
Current
_
(b)

30°

V
(c)

Figure 5.2 A resistive-capacitive circuit: (a) circuit diagram; (b) representative waveforms; (c) vector representation.

Because of the electrostatic field developed around a capacitor, an opposition to the flow of alternat-
ing current exists. This opposition is known as capacitive reactance, defined as

1
Xc = (5.5)
2πfC
where
Xc = Capacitive reactance (Ω)
2 π R = The mathematical expression of one sine wave
f = Frequency (Hz)
C = Capacitance (F)
The dielectric used for a given capacitor varies, depending upon the application. Common dielec-
trics include air, gas, mica, glass, and ceramic. Each has a different dielectric constant, temperature range,
and thickness.
The dielectric constant of a material determines the electrostatic energy that may be stored in that
material per unit volume for a given voltage. The value of the dielectric constant expresses the ratio of a
capacitor whose dielectric is a vacuum to one using a given dielectric material. The dielectric constant of

© 2007 by Taylor & Francis Group, LLC

84 AC Power Systems Handbook

Table 5.1 Comparison of Capacitor Dielectric Constants (After [1].)

Dielectric K (Dielectric Constant)
Air or vacuum 1.0
Paper 2.0 – 6.0
Plastic 2.1 – 6.0
Mineral oil 2.2 – 2.3
Silicone oil 2.7 – 2.8
Quartz 3.8 – 4.4
Glass 4.8 – 8.0
Porcelain 5.1 – 5.9
Mica 5.4 – 8.7
Aluminum oxide 8.4
Tantalum pentoxide 26.0
Ceramic 12.0 – 400,000.0

air is 1, the reference unit employed for characterizing this parameter. As the dielectric constant is
increased or decreased, the capacitance will increase or decrease, respectively. Table 5.1 lists the dielectric
constants of various common materials.
The dielectric constant of most materials is affected by both temperature and frequency, except for
quartz, Styrofoam, and Teflon, whose dielectric constants remain essentially constant.
In addition to capacitance, a practical
capacitor has inductance and resistance com-
ponents, as shown in Figure 5.3. The stray com- Rp
ponents are identified as follows:
L Rs
• Rs = Series resistance of wire leads, con-
tact terminations, and electrodes
• Rp = Shunt resistance resulting from the C
resistivity of the dielectric and case mate-
Figure 5.3 The equivalent circuit of a capacitor.
rial, and dielectric losses
• L = Stray inductance resulting from the
leads and the electrodes
The equivalent series resistance (ESR) of a capacitor is the ac resistance of the device, reflecting both
the series resistance (Rs) and the parallel resistance (Rp) at a given frequency. This parameter permits the loss
resulting from the foregoing elements to be expressed as a loss in a single resistor in the equivalent circuit.
The power factor (PF) of a capacitor defines the electrical losses in the device operating under an ac
voltage. In an ideal device, the current will lead the applied voltage by 90°. A practical capacitor, because
of its dielectric, electrode, and contact termination losses, exhibits a phase angle of less than 90°. The
power factor of a capacitor is defined as the ratio of the effective series resistance to the impedance of the
capacitor. PF usually is expressed as a percentage.
The quality factor (Q) of a capacitor is the ratio of the capacitor reactance to the resistance of the
device at a specified frequency. The Q is determined by the equations
1
Q= (5.6)
2πfCR
1
Q = ------ (5.7)
PF

where
Q = Quality factor
f = Frequency (Hz)
C = Value of capacitance (F)
R = Internal resistance (Ω)
PF = Power factor

© 2007 by Taylor & Francis Group, LLC

Capacitors 85

Other important specifications for capacitors include:

• Dielectric absorption (DA): The reluctance of the dielectric to give up stored electrons when the
capacitor is discharged. This is often called memory because if a capacitor is discharged through
a resistance and the resistance is removed, the electrons that remain in the dielectric will reconvene
on the electrode, causing a voltage to appear across the capacitor. DA is usually measured by
charging the capacitor for 5 min, discharging it for 5 s, then having an open circuit for 1 min,
after which the recovery voltage is read. The percentage of DA is defined as the ratio of recovery
voltage to charging voltage times 100.
• Dissipation factor (DF): The ratio of the effective series resistance to capacitive reactance. DF
normally is expressed as a percentage.
• Leakage current: The current flowing through the capacitor when a dc voltage is applied.
• Insulation resistance: The ratio of the applied voltage to the leakage current. Insulation resistance
is normally expressed in megohms.
• Ripple current/voltage: The rms value of the maximum allowable alternating current or voltage
(superimposed on any dc level) at a specific frequency at which the capacitor may be operated
continuously at a specified temperature.
• Surge voltage: The maximum operating voltage of the capacitor at any temperature.

5.2.1 Series and Parallel Connections

The formulas for series and parallel connection of capacitors can be obtained from the general consider-
ation of series and parallel connection of impedances [2]. For series connection

1- 1 1 - + … + --------
1-
----- = --------- + -------- (5.8)
sC sC 1 sC 2 sC n

where C1, C2, ... , Cn are the capacitances of the capacitors connected in series. Then, the equivalent capac-
itance C can be found as

1-⎞ –1
1- + … + -----
1- + -----
C = ⎛ ----- (5.9)
⎝ C1 C2 Cn ⎠

and is always less than the value of the smallest capacitance. This result can be used when the conditions
for using the impedance concept are valid, that is, the capacitors are discharged.
The series connection of two capacitors has the equivalent capacitance

C1 C2
C = -----------------
- (5.10)
C1 + C2

In a similar way one can obtain the equivalent capacitance of parallel connection as

C = C1 + C2 + … + Cn (5.11)

and is always larger than the largest capacitance.

Furthermore, the following two results are useful (for initially discharged capacitors):
• The voltage applied to two capacitors connected in series is divided between these two capacitors
inversely proportionally to their capacitance (thus, the larger part of the voltage is on the smaller
capacitance).

© 2007 by Taylor & Francis Group, LLC

86 AC Power Systems Handbook

• The current applied to two capacitors connected in parallel is divided between these two capacitors
proportionally to their capacitance (thus, the larger part of the current goes through the larger
capacitance).

5.2.2 Practical Capacitors

Practical capacitors can be generally divided capacitors in four broad classes [2]:
• The first class includes low-loss capacitors with good capacitance stability. These are (in accordance
with the dielectric) mica, glass, some ceramic, and low-loss plastic film (polypropylene, polysty-
rene) capacitors.
• The second class are capacitors of medium loss and medium stability designed to operate over a
wide range of dc and ac voltages. These are paper (oil and wax impregnated), plastic film, and
some ceramic capacitors. The dc applications include coupling, decoupling, bypass, smoothing
and power separating filters, and energy storage. The ac applications include motor start, lightning,
interference suppression, and power-line applications (switching and measurement equipment).
• The third class are aluminum and tantalum electrolytic capacitors providing high capacitance in
a small size. For the same rated voltage tantalum capacitors have as much as three times larger
capacitance per volume unit as aluminum, they are more reliable and have longer service life. The
electrolytics are designed for dc and polarized voltage applications. Less-expensive aluminum
capacitors are used in radio and television equipment; more expensive tantalum capacitors are
destined for military and harsh environmental applications.
• The fourth class are fixed value capacitors (mica, glass, oil, gas, and vacuum) designed for high-
voltage (up to 35 kV peak), high-current (up to 200 A) transmitter and power control applications.
The typical construction of common discrete capacitors is illustrated in Figure 5.4. Table 5.2 lists the
characteristics of common capacitor types.

5.3 Capacitor Failure Modes

Experience has shown that capacitor failures are second only to semiconductors and vacuum tubes in
components prone to malfunction in electronic equipment. Capacitors for ac applications range from
high-voltage oil-filled devices, such as the one shown in Figure 5.5, to low voltage, high capacitance
devices of the type typically found in power supplies (both linear and switching). Of all the various types
of capacitors used today, it is estimated that electrolytics present the greatest potential for problems to
equipment users.

5.3.1 Electrolytic Capacitors

Electrolytic capacitors are popular because they offer a large amount of capacitance in a small physical
size. They are widely used as filters in low-voltage power supplies and as coupling devices in audio and RF
stages. An aluminum electrolytic capacitor consists of two aluminum foil plates separated by a porous
strip of paper (or other material) soaked with a conductive electrolyte solution. Construction of a typical
device is illustrated in Figure 5.6. The separating material between the capacitor plates does not form the
dielectric but, instead, serves as a spacer to prevent the plates from mechanically short-circuiting. The
dielectric consists of a thin layer of aluminum oxide that is electrochemically formed on the positive foil
plate. The electrolyte conducts the charge applied to the capacitor from the negative plate, through the
paper spacer, and into direct contact with the dielectric. This sandwich arrangement of foil-spacer-foil is
then rolled up and encapsulated.
Problems with electrolytic capacitors fall into two basic categories: mechanical failure and failure of
electrolyte.

© 2007 by Taylor & Francis Group, LLC

 Capacitors
Table 5.2 Parameters and Characteristics of Discrete Capacitors (After [2].)
TC Insulation
Rated Voltage, ppm/ Tolerance, Resistance, Dissipation Dielectric Temperature Comments,
Capacitor Type Range VR °C ±% MΩ μF Factor,% Absorption,% Range, °C Applications Cost
Polycarbonate 100 pF – 30 μF 50 – 800 ±50 10 5 × 105 0.200 0.10 –55/+125 High quality, small, low High
TC
Polyester/Mylar 1000 pF – 50 μF 50 – 600 +400 10 105 0.750 0.30 –55/+125 Good, popular Medium
Polypropylene 100 pF – 50 μF 100 – 800 –200 10 105 0.200 0.10 –55/+105 High quality, low High
absorption
Polystyrene 10 pF – 2.7 μF 100 – 600 –100 10 106 0.050 0.04 –55/+85 High quality, large, low Medium
TC, signal filters
Polysulfone 1000 pF – 1 μF — +80 5 105 0.300 0.20 –55/+150 High temperature High
Parylene 5000 pF – 1 μF — ±100 10 105 0.100 0.10 –55/+125 High
Kapton 1000 pF – 1 μF — +I00 10 105 0.300 0.30 –55/+220 High temperature High
Teflon 1000 pF – 2 μF 50 – 200 –200 10 5 × 106 0.040 0.04 –70/+250 High temperature, High
lowest absorption
Mica 5 pF – 0.01 μF 100 – 600 –50 5 2.5 × 104 0.001 0.75 –55/+125 Good at RF, low TC High
Glass 5 pF – 1000 μF 100 – 600 +I40 5 106 0.001 –55/+125 Excellent long-term High
stability
Porcelain 100 pF – 0.1 μF 50 – 400 +I20 5 5 × 105 0.100 4.20 –55/+125 Good long-term High
stability
Ceramic (NPO) 100 pF – 1 μF 50 – 400 ±30 10 5 × 103 0.020 0.75 –55/+125 Active filters, low TC Medium
Ceramic 10 pF – 1 μF 50 – 30,000 — — — — — –55/+125 Small, very popular, Low
selectable TC
Paper 0.01 – 10 μF 200 – 1,600 ±800 10 5 × 103 1.000 2.50 –55/+125 Motor capacitors Low
Aluminum 0.1 – 1.6 F 3 – 600 +2500 –10/+100 100 10.000 8.00 –40/+85 Power supply filters High
short life
Tantalum (Foil) 0.1 – 1000 μF 6 – 100 +800 –10/+100 20 4.000 8.50 –55/+85 High capacitance, small High
size, low inductance
Thin-film 10 – 200 pF 6 – 30 +I00 10 106 0.010 — –55/+125 High
Oil 0.1 – 20 μF 200 – 10,000 — — — 0.500 — — High voltage filters, —
large, long life
Vacuum 1 – 1000 pF 2,000 – 3,600 — — — — — — Transmitters —

87
© 2007 by Taylor & Francis Group, LLC
88 AC Power Systems Handbook

Plastic core
Extended foil construction Plastic case

Lead

Silver
electrode

Ceramic
Parallel plate
End disk dielectric
construction
(a) (b)
Solder
coated lead
O-ring
Butt weld

Teflon
bushing

Anode lead
Electrolyte
Porous anode
Teflon spacer

Silver case
Cathode lead
(c)

Figure 5.4 Construction of discrete capacitors. (From [2]. Used with permission.)

5.3.1.1 Mechanical Failure

Mechanical failures relate to poor bonding of the leads to
the outside world, contamination during manufacture,
and shock-induced short-circuiting of the aluminum foil
plates. Typical failure modes include short circuits caused
by foil impurities, manufacturing defects (such as burrs
on the foil edges or tab connections), breaks or tears in
the foil, and breaks or tears in the separator paper.
Short circuits are the most frequent failure mode
during the useful life period of an electrolytic capacitor.
Such failures are the result of random breakdown of the
dielectric oxide film under normal stress. Proper capaci-
tor design and processing will minimize such failures.
Short circuits also can be caused by excessive stress, where
voltage, temperature, or ripple conditions exceed speci-
fied maximum levels.
Open circuits, although infrequent during normal
life, can be caused by failure of the internal connections
joining the capacitor terminals to the aluminum foil.
Mechanical connections can develop an oxide film at the Figure 5.5 An oil-filled, high-voltage capacitor.
contact interface, increasing contact resistance and even-
tually producing an open circuit. Defective weld connections also can cause open circuits. Excessive
mechanical stress will accelerate weld-related failures.

© 2007 by Taylor & Francis Group, LLC

Capacitors 89

5.3.1.2 Temperature Cycling

Capacitors are subject to failures induced by thermal – +
cycling. Experience has shown that thermal stress is
a major contributor to failure in aluminum electro- – +
lytic capacitors. Dimensional changes between plas- Negative Positive
– +
tic and metal materials can result in microscopic lead lead
ruptures at termination joints, possible electrode
oxidation, and unstable device termination (chang- – + Welded
ing series resistance). The highest-quality capacitor Welded connection
connection – +
will fail if its voltage or current ratings are exceeded.
Appreciable heat rise (20°C during a 2-hour period Aluminum – + Aluminum
of applied sinusoidal voltage) is considered abnor- negative plate positive plate
mal and may be a sign of incorrect application of the Aluminum
component or impending failure of the device. oxide dielectric
Figure 5.7 illustrates the effects of high ambient Paper soaked in electrolyte
temperature on capacitor life. Note that operation at
Figure 5.6 The basic design of an aluminum electro-
33% duty cycle is rated at 10 years when the ambient
lytic capacitor.
temperature is 35°C, but the life expectancy drops to
just 4 years when the same device is operated at
55°C. A common rule of thumb is this: Within the range of + 75°C through the full-rated temperature,
stress and failure rates double for each 10°C increase in operating temperature. Conversely, the failure
rate is reduced by half for every 10°C decrease in operating temperature.
5.3.1.3 Electrolyte Failures
Failure of the electrolyte can be the result of application of a reverse bias to the component or of a drying
of the electrolyte itself. Electrolyte vapor transmission through the end seals occurs on a continuous basis
throughout the useful life of the capacitor. This loss has no appreciable effect on reliability during the
useful life period of the product cycle. When the electrolyte loss approaches 40% of the initial electrolyte
content of the capacitor, however, the electrical parameters deteriorate and the capacitor is considered to
be worn out.

85
Ambient degrees centigrade circulating air

75

65

55

45

35
0 2 4 6 8 10
Years operation 33% duty cycle

Figure 5.7 Life expectancy of an electrolytic capacitor as a function of operating temperature.

© 2007 by Taylor & Francis Group, LLC

90 AC Power Systems Handbook

As a capacitor dries out, three failure modes may be Aluminum

experienced: leakage, a downward change in value, or oxide
dielectric absorption. Any one of these can cause a system
to operate out of tolerance or fail altogether.
The most severe failure mode for an electrolytic is
increased leakage, illustrated in Figure 5.8. Leakage can
cause loading of the power supply or upset the dc bias of
an amplifier. Loading of a supply line often causes addi-
Negative Positive
tional current to flow through the capacitor, possibly lead lead
resulting in dangerous overheating and catastrophic fail-
ure.
A change of device operating value has a less devas-
tating effect on system performance. An aluminum elec-
trolytic has a typical tolerance range of about ± 20%. A
capacitor suffering from drying of the electrolyte can
experience a drastic drop in value (to just 50% of its rated
value, or less). The reason for this phenomenon is that
after the electrolyte has dried to an appreciable extent, the
charge on the negative foil plate has no way of coming in Paper soaked in
contact with the aluminum-oxide dielectric. This failure electrolyte Aluminum
oxide dielectric
mode is illustrated in Figure 5.9. Remember, it is the alu-
minum-oxide layer on the positive plate that gives the Figure 5.8 Failure mechanism of a leaky alu-
electrolytic capacitor its large rating. The dried-out paper minum electrolytic capacitor. As the device
ages, the aluminum oxide dissolves into the
spacer, in effect, becomes a second dielectric, which signif-
electrolyte, causing the capacitor to become
icantly reduces the capacitance of the device.
leaky at high voltages

5.3.2 Capacitor Life Span

The life expectancy of a capacitor — operating in an ideal
circuit and environment — will vary greatly, depending + –
upon the grade of device selected. Typical operating life,
according to capacitor manufacturer data sheets, range + –
from a low of 3 to 5 years for inexpensive electrolytic
Positive Negative
devices, to a high of greater than 10 years for computer- + –
lead lead
grade products. Catastrophic failures aside, expected life is
a function of the rate of electrolyte loss by means of vapor
+ –
transmission through the end seals, and the operating or
storage temperature. Properly matching the capacitor to
+ –
the application is a key component in extending the life of
an electrolytic capacitor. The primary operating parame-
+ –
ters include:
Dried paper spacer
• Rated voltage — the sum of the dc voltage and peak
Figure 5.9 Failure mechanism of an elec-
ac voltage that can be applied continuously to the
trolytic capacitor exhibiting a loss of
capacitor. Derating of the applied voltage will
capacitance. After the electrolyte dries,
decrease the failure rate of the device. the plates can no longer come in contact
• Ripple current — the rms value of the maximum with the aluminum oxide. The result is
allowable ac current, specified by product type at a decrease in capacitor value.
120 Hz and +85°C (unless otherwise noted). The
ripple current may be increased when the compo-
nent is operated at higher frequencies or lower
ambient temperatures.

© 2007 by Taylor & Francis Group, LLC

Capacitors 91

Anode lead
Tantalum
riser wire Butt weld
TIG (tungsten inert gas) weld

Tantalum tube
Closure weld

Tantalum-to-glass-
to-tantalum seal
Top vibration spacer

Tantalum anode
Tantalum case
Sintered tantalum
cathode surface
Gelled sulfuric acid
electrolyte

Bottom vibration
spacer

Cathode lead

Figure 5.10 Basic construction of a tantalum capacitor.

• Reverse voltage — the maximum voltage that can be applied to an electrolytic capacitor without
damage. Electrolytic capacitors are polarized and must be used accordingly.

5.3.3 Tantalum Capacitor

Tantalum electrolytic capacitors have become the preferred type of device where high reliability and long
service life are primary considerations. The tantalum pentoxide compound possesses high dielectric
strength and a high dielectric constant. As the components are being manufactured, a film of tantalum
pentoxide is applied to the electrodes by means of an electrolytic process. The film is applied in various
thicknesses. Figure 5.10 shows the internal construction of a typical tantalum capacitor. Because of the
superior properties of tantalum pentoxide, tantalum capacitors tend to have as much as three times
higher capacitance per volume efficiency as an aluminum electrolytic capacitor. This, coupled with the
fact that extremely thin films can be deposited during the electrolytic process, makes tantalum capacitors
efficient with respect to the number of microfarads per unit volume.
The capacitance of any device is determined by the surface area of the conducting plates, the distance
between the plates, and the dielectric constant of the insulating material between the plates. In the tanta-
lum capacitor, the distance between the plates is small; it is just the thickness of the tantalum pentoxide
film. Tantalum capacitors contain either liquid or solid electrolytes.

5.4 References
1. Ballou, G., “Capacitors and Inductors,” in The Electrical Engineering Handbook, R. C. Dorf (Ed.),
p. 15, CRC Press, Boca Raton, FL, 1993.
2. Filanovsky, I. M., “Capacitance and Capacitors,” in The Electronics Handbook, 2nd ed., J. C. Whitaker
(Ed.), p. 371, CRC Press, Boca Raton, FL, 2005.

© 2007 by Taylor & Francis Group, LLC

92 AC Power Systems Handbook

5.5 Bibliography
Meeldijk, V., “Why Do Components Fail?” Electronic Servicing & Technology, Intertec Publishing, Overland
Park, KS, November 1986.
Technical staff, “Aluminum Electrolytic Capacitors: Reliability, Expected Life, and Shelf Capability,”
Sprague Applications Guide, Sprague Electric Company, Lansing, NC, 1989.
Technical staff, “Introduction to Tantalum Capacitors,” Sprague Applications Guide, Sprague Electric
Company, Lansing, NC, 1989.

© 2007 by Taylor & Francis Group, LLC