Power Supplies

https://electricalacademia.com/electronics/power-supply-definition-functions-components/

Definition: A power supply is an electronic circuit designed to provide various ac and dc

voltages for equipment operation.
=>Proper operation of electronic equipment requires a number of source voltages. Low dc
voltages are needed to operate ICs and transistors.
=>High voltages are needed to operate CRTs and other devices. Batteries can provide all of
these voltages.
=>However, electricity for electrical and electronic devices are commonly supplied by the local
power company. This power comes out of an outlet at 115-volt ac, with a frequency of 60 Hertz.
Different voltages are needed to operate some equipment.
Power Supply Functions
The complete power supply circuit can perform these functions:
1. Step voltages up or step voltages down, by transformer action, to the required ac line
voltage.
2. Provide some method of voltage division to meet equipment needs.
3. Change ac voltage to pulsating dc voltage by either half-wave or full-wave rectification.
4. Filter pulsating dc voltage to a pure dc steady voltage for equipment use.
5. Regulate power supply output in proportion to the applied load.
Power Supply Components
=>A block diagram illustrating these functions is shown in Figure 1. Note that certain functions
are not found in every power supply. See Figure 2 for a typical commercial power supply
components.

1
Figure 1. Block diagram for power supply components. Input is 117 volts ac. Processes used in a
typical power supply are shown below the blocks. The output of the power supply can be dc or
ac. The output of this supply is five volts dc.

Figure 2. Regulated dc power supply. (Knight Electronics)

Power Transformers Diodes
=>The first device in a power supply is the transformer. Its purpose is to step up or step down
alternating source voltage to values needed for radio, TV, computer, or other electronic circuit
use.
=>Most transformers do not have any electrical connection between the secondary and primary
windings. See Figure 3. This means that the transformer isolates the circuit connected to the
primary from the circuit connected in the secondary.
Isolation is a term that means there are no electrical connections between the primary and
secondary on the transformer.

Figure 3. Isolation in a transformer.

=>An isolation transformer is a transformer that has the specific purpose of isolating the primary
circuit from the secondary circuit.
=>Using an isolation transformer is a safety feature because it helps prevent shocks in the
secondary. Our body or hands must be joined across both leads of the secondary connections in
order to receive a shock.
=>The safety condition described above does not hold true in the primary with commercial ac
provided by the power company. One connection is hot, which means that the connection is
electrically energized. The other is grounded, or neutral. Standing on the ground while touching
the hot connection will result in a shock. Touching the ground connection alone will not result in
a shock.
=>Secondary windings can be tapped to provide different voltages. A tap placed midway
between the two ends of a secondary winding is called a center tap.
=>Many power supplies use a center tap secondary transformer winding. The tapped voltages,
Figure 4, are 180 degrees out of phase with respect to the center tap.
A variety of transformers can be found in nearly all electronic devices. You should understand
the basic theory and purpose of the transformer. Review Chapter 12 if necessary.
A Lesson in Safety
Transformers produce high voltages that can be very dangerous. Proper respect and extreme
caution must be used at all times when working with, or measuring, high voltages.

2
 Figure 4. A center tap transformer.
Half-Wave and Full-Wave Rectification
After a voltage has gone through a power supply’s transformer, the next step is rectification.
The process of changing an alternating current to a pulsating direct current is called
rectification.
=>When changing an ac signal to dc, there are two types of rectification: half-wave rectification
and full-wave rectification.
With the half-wave rectifier, only half of the input signal passes on through the rectifier. With the
full-wave rectifier, the entire input wave is passed through.
Half-Wave Rectification
=>In Figure 5, the output of a transformer is connected to a diode and a load resistor that are in
series. The input voltage to the transformer appears as a sine wave.
The polarity of the wave reverses at the frequency of the applied voltage. The output voltage of
the transformer secondary also appears as a sine wave. The magnitude of the wave depends on
the turns ratio of the transformer. The output is 180 degrees out of phase with the primary.
The top of the transformer (point A) is joined to the diode anode. Note that the B side of the
transformer is connected to ground.
=>During the first half cycle, point A is positive. The diode conducts, producing a voltage drop
across resistor R equal to IR. During the second half cycle, point A is negative. The diode anode
is also negative. No conduction takes place, and no IR drop appears across R.

3
 Figure 5. Basic diode rectifier schematic.
An oscilloscope connected across R produces the waveform shown to the right in Figure 6. The
output of this circuit consists of pulses of current flowing in only one direction and is at the same
frequency as the input voltage. The output is a pulsating direct current.

Figure 6. Input and output waveforms of a diode rectifier.

Only one half of the ac input wave is used to produce the output voltage. This type of rectifier is
called a half-wave rectifier.
Look at the polarity of the output voltage in Figure 6. One end of the resistor R is connected to
ground. The current flows from the ground to the cathode. This connection makes the end of R
connected to the cathode positive as shown in Figure 5.
A negative rectifier can be made by reversing the diode in the circuit, Figure 7. The diode
conducts when the cathode becomes negative causing the anode to become positive.
The current through R would be from the anode to ground making the anode end of R negative
and the ground end of R more positive.
Voltages taken from across R, the output, would be negative with respect to ground. This circuit
is called an inverted diode. It is used when a negative supply voltage is required.

4
 Figure 7. An inverted diode produces a negative voltage.
It is possible to have a power supply that provides half-wave rectification without the use of a
transformer. This circuit is not isolated. There is no step up or step down of current voltages.
This circuit is a simpler, less costly design, and since there is no transformer, it can be used in
smaller spaces, Figure 8.

Figure 8. Half-wave rectification without a transformer.

Full-Wave Rectification
=>The pulsating direct voltage output of a half-wave rectifier can be filtered to a pure dc voltage.
=>However, the half-wave rectifier uses only one half of the input ac wave.
=>A better filtering action can be obtained by using two diodes. With this setup, both half cycles
of the input wave can be used.
=>Both half cycles at the output have the same polarity in this full-wave rectifier. Figure
9 follows the first half cycle. Figure 10 follows the second half cycle.

5
 Figure 9. Arrows show current in full-wave rectifier during the first half cycle.

Figure 10. The direction of current during the second half cycle.
=>To produce this full-wave rectification, a center tap is made on the secondary winding. This
tap is attached to the ground.
=>In Figure 9, point A is positive and diode anode D1 is positive. Electron flow is shown by the
arrows. During the second half of the input cycle, point B is positive, diode anode D2 is positive,
and current flows as shown in Figure 10.
=>No matter which diode is conducting, the current through load resistor R is always in the same
direction. Both positive and negative half cycles of the input voltage cause the current through R
in the same direction.
=>The output voltage of this full-wave rectifier is taken from across R. It consists of direct
current pulses at twice the frequency of input voltage, Figure 11. To produce this full-wave
rectification in this circuit, the secondary voltage was cut in half by the center tap.

6
 Figure 11. The waveforms of input and output of full-wave diode rectifier.
=>The diodes, D1 and D2, used in Figures 9 and 10, are packaged both individually and in
pairs. Figure 12 shows a two rectifier package.
=>The center lead is used as the connection for the cathodes. The cathodes are wired together.

Figure 12. Dual diodes with a center tap.

Bridge Rectifiers
=>It is not always necessary to use a center-tapped transformer for full-wave rectification. Full
secondary voltage can be rectified by using four diodes in a circuit called a bridge rectifier,
Figure 13 and 14. Two circuits are shown so that the current can be observed in each half cycle.

7
 Figure 13. Current in bridge rectifier during the first half cycle.

Figure 14. Current in bridge rectifier during the second half cycle.
=>In Figure 13, point A of the transformer secondary is positive. Current flows in the direction
of the arrows. When point B is positive, current flows as in Figure 14.
=>Again, notice that the current through R is always in one direction. Both halves of the input
voltage are rectified and the full voltage of the transformer is used.
=>Bridge rectifiers can be used in circuits without transformers. Without transformers, the
voltage or current will not be stepped up or down. There will be no isolation. These circuits are
also called line-operated bridge circuits, Figure 15.
Caution
=>Connecting an oscilloscope directly to a line-operated bridge rectifier will result in a dead
ground when the oscilloscope ground is connected to the line voltage bridge. An isolation
transformer with a 1 to 1 ratio must be used to prevent the ground lead on the scope from being
connected to the hot conductor.

8
 Figure 15. Line-operated bridge rectifier circuit.
=>The output of either the half-wave or full-wave rectifier is a pulsating voltage. Before it can be
applied to other circuits, the pulsations must be reduced. A steadier dc is needed. It can be
obtained using a filter network.
=>In Figure 16, the line, Eavg, shows the average voltage of the pulsating dc wave. It is equal to
0.637 × peak voltage. The shaded portion of the wave above the average line is equal in area to
the shaded portion below the line.
Movement above and below the average voltage is called the ac ripple. It is this ripple that
requires filtering.
=>The percentage of ripple as compared to the output voltage must be kept to a small value. The
ripple percentage can be found using the formula:

Figure 16. Average value of full-wave rectifier output.

Smoothing
 Capacitor Filters
=>A capacitor connected across the rectifier output provides some filtering action, Figure 17.
The capacitor is able to store electrons.

9
=>When the diode or rectifier is conducting, the capacitor charges rapidly to near the peak
voltage of the wave. It is limited only by the resistance of the rectifier and the reactance of the
transformer windings.
=>Between the pulsations in the wave, voltage from the rectifier drops. The capacitor then
discharges through the resistance of the load.
=>The capacitor, in effect, is a storage chamber for electrons. It stores electrons at peak voltage
and then supplies electrons to the load when rectifier output is low. See Figure 18.

Figure 17. Filtering action of a capacitor.

Figure 18. Input and output of the capacitor filter showing the change in the waveform.
Capacitors used for this purpose are electrolytic types because large capacitances are needed in
a limited space. Common values for the capacitors range from 4 to 2000 microfarads. Working
voltages of capacitors should be in excess of the peak voltage from the rectifier.
 LC Filters
=>The filtering action can be improved by adding a choke in series with the load. This LC filter
circuit appears in Figure 19. The filter choke consists of many turns of wire wound on a
laminated iron core.

10
 Figure 19. Further filtering is produced by the choke in series with the load.
Recall that inductance was that property of a circuit that resisted a change in current. A rise in
current induced a counter emf that opposed the rise.
=>A decrease in current induced a counter emf that opposed the decrease. As a result, the choke
constantly opposes any change in current. Yet, it offers very little opposition to a direct current.
=>Chokes used in radios have values from 8 to 30 henrys. Current ratings range from 50 to 200
milliamperes.
=>Larger chokes can be used in transmitters and other electronic devices. Filtering action as a
result of the filter choke is shown in Figure 20.

Figure 20. Waveforms show the filtering action of the capacitor and choke together.
=>A second capacitor can be used in the filter section after the choke, to provide more filter
action. See Figure 21.
=>The action of this capacitor is similar to the first capacitor. The circuit configuration appears
as the Greek letter π. The filter is called a pi (π) section filter.

11
 Figure 21. Pi (π) section filter.
=>When the first filtering component is a capacitor, the circuit is called a capacitor input filter.
=>When the choke is the first filtering component, it is called a choke input filter, Figure 22.
=>The choke input filter looks like an inverted L, so it is also called an L section filter. Several
of these filter sections can be used in series to provide added filtering.

Figure 22. Choke input L filter

=>In the capacitor input filter, the capacitor charges to the peak voltage of the rectified wave.
=>In the choke input, the charging current for the capacitor is limited by the choke.
=>The capacitor does not charge to the peak voltage. As a result, the output voltage of the
power supply using the capacitor input filter is higher than one using the choke input filter.

Voltage regulation

=>A voltage regulator is a component of the power supply unit that ensures a steady constant
voltage supply through all operational conditions.

=>It regulates voltage during power fluctuations and variations in loads. It can regulate AC as
well as DC voltages.

12
How Does a Voltage Regulator Work?

=>A voltage regulator is a circuit that creates and maintains a fixed output voltage, irrespective
of changes to the input voltage or load conditions.

=>Voltage regulators (VRs) keep the voltages from a power supply within a range that is
compatible with the other electrical components. While voltage regulators are most commonly
used for DC/DC power conversion, some can perform AC/AC or AC/DC power conversion as
well. This article will focus on DC/DC voltage regulators.

Types of Voltage Regulators: Linear vs. Switching

=>There are two main types of voltage regulators: linear and switching. Both types regulate a
system’s voltage, but linear regulators operate with low efficiency and switching regulators
operate with high efficiency. In high-efficiency switching regulators, most of the input power is
transferred to the output without dissipation.
Linear Regulators
=>A linear voltage regulator utilizes an active pass device (such as a BJT or MOSFET), which is
controlled by a high-gain operational amplifier. To maintain a constant output voltage, the linear
regulator adjusts the pass device resistance by comparing the internal voltage reference to the
sampled output voltage, and then driving the error to zero.

=>Linear regulators are step-down converters, so by definition the output voltage is always
below the input voltage. However, these regulators offer a few advantages: they are generally
easy to design, dependable, cost-efficient, and offer low noise as well as a low output voltage
ripple.

=>Linear regulators, such as the MP2018, only require an input and output capacitor to
operate (see Figure 1). Their simplicity and reliability make them intuitive and simple devices
for engineers, and are often highly cost-effective.

13
Figure 1: MP2018 Linear Regulator

Switching Regulators
=>A switching regulator circuit is generally more complicated to design than a linear regulator,
and requires selecting external component values, tuning control loops for stability, and careful
layout design.

=>Switching regulators can be step-down converters, step-up converters, or a combination of the

two, which makes them more versatile than a linear regulator.

=>Advantages of switching regulators include that they are highly efficient, have better thermal
performance, and can support higher current and wider VIN / VOUT applications. They can achieve
greater than 95% efficiency depending on the application requirements. Unlike linear regulators,
a switching power supply system may require additional external components, such as inductors,
capacitors, FETs, or feedback resistors. The HF920 is an example of a switching regulator that
offers high reliability and efficient power regulation (see Figure 2).

14
Figure 2: HF920 Switching Regulator

Limitations of Voltage Regulators

=>One of the main disadvantages for linear regulators is that they can be inefficient, as they
dissipate large amounts of power in certain use cases. The voltage drop of a linear regulator is
comparable to a voltage drop across a resistor. For instance, with a 5V input voltage and a 3V
output voltage, there is a 2V drop between the terminals, and the efficiency is limited to 3V/5V
(60%). This means linear regulators are best suited for applications with lower VIN /
VOUT differentials.
=>It is important to consider the estimated power dissipation of a linear regulator in application,
since using larger input voltages results in high power dissipation that can overheat and damage
components.

=>Another limitation of linear voltage regulators is that they are only capable of buck (step-
down) conversion, in contrast to switching regulators, which also offer boost (step-up) and buck-
boost conversion.

=>Switching regulators are highly efficient, but some disadvantages include that they are
generally less cost-effective than linear regulators, larger in size, more complex, and can create
more noise if their external components are not carefully selected. Noise can be very important
for a given application, as noise can affect circuit operation and performance, as well as EMI
performance.

15
Switching Regulator Topologies: Step-Down, Step-Up, Linear, LDO, and Adjustable

=>There are various topologies for linear and switching regulators. Linear regulators often rely
on low-dropout (LDO) topologies. For switching regulators, there are three common topologies:
step-down converters, step-up converters, and buck-boost converters. Each topology is described
below:
LDO Regulators
=>One popular topology for linear regulators is a low-dropout (LDO) regulator. Linear
regulators typically require the input voltage to be at least 2V above the output voltage.
However, an LDO regulator is designed to operate with a very small voltage difference between
input and output terminals, sometimes as low as 100mV.
Step-Down and Step-Up Converters
=>Step-down converters (also called buck converters) take a larger input voltage and produce a
lower output voltage. Conversely, step-up converters (also called boost converters) take a lower
input voltage and produce a higher output voltage.
Buck-Boost Converters
=>A buck-boost converter is a single-stage converter that combines the functions of a buck and a
boost converter to regulate the output over a wide range of input voltages that can be greater or
less than the output voltage.

Voltage Regulator Control

=>The four fundamental components of a linear regulator are a pass transistor, error amplifier,
voltage reference, and resistor feedback network. One of the inputs to the error amplifier is set by
two resistors (R1 and R2) to monitor a percentage of the output voltage. The other input is a
stable voltage reference (VREF). If the sampled output voltage changes relative to VREF, the error
amplifier changes the pass transistor’s resistance to maintain a constant output voltage (V OUT).

=>Linear regulators typically only require an external input and output capacitor to operate,
making them easy to implement.

=>On the other hand, a switching regulator requires more components to create the circuit. The
power stage switches between VIN and ground to create charge packets to deliver to the output.
Similar to a linear regulator, there is an operational amplifier that samples the DC output voltage
from the feedback network and compares it to an internal voltage reference.

=>Then the error signal is amplified, compensated, and filtered. This signal is used to modulate
the PWM duty cycle to pull the output back into regulation. For example, if the load current
increases rapidly and causes an output voltage droop, the control loop increases the PWM duty
cycle to supply more charge to the load and bring the rail back into regulation.

16
Linear and Switching Regulator Applications
=>Linear regulators are often used in applications that are cost-sensitive, noise-sensitive, low-
current, or space constrained. Some examples include consumer electronics such as headphones,
wearables, and Internet-of-Things (IoT) devices. For instance, applications such as a hearing aid
could use a linear regulator because they don’t have a switching element that could create
unwanted noise and interfere with the device’s performance.

=>Moreover, if designers are mainly interested in creating a low-cost application, they need not
be as concerned with power dissipation, and can rely on a linear regulator.

=>Switching regulators are beneficial for more general applications, and are especially useful in
applications that need efficiency and performance, such as consumer, industrial, enterprise, and
automotive applications (see Figure 3). For example, if the application requires a large step-
down solution, a switching regulator is better suited, since a linear regulator could create high
power dissipation that would damage other electrical components.

Figure 3: MPQ4430-AEC1 Step-Down Regulator

What are the Basic Parameters for a Voltage Regulator IC?

=>Some of the basic parameters to consider when using a voltage regulator are the input voltage,
output voltage, and output current. These parameters are used to determine which VR topology is
compatible with a user’s IC.
=>Other parameters — including quiescent current, switching frequency, thermal resistance,
and feedback voltage — may be relevant depending on the application.

=>Quiescent current is important when efficiency during light-load or standby modes is a

priority. When considering switching frequency as a parameter, maximizing the switching
frequency leads to smaller system solutions.

17
=>Additionally, thermal resistance is critical to remove heat from the device and dissipate it
across the system. If the controller includes an internal MOSFET, then all losses (conductive and
dynamic) are dissipated in the package and must be considered when calculating the maximum
temperature of the IC.
=>Feedback voltage is another important parameter to examine because it determines the lowest
output voltage that the voltage regulator can support. It is standard to look at the voltage
reference parameters. This limits the lower output voltage, the accuracy of which impacts the
accuracy of the output voltage regulation.

How to Pick the Right Voltage Regulator

=>To select the proper voltage regulator, the designer must first understand their key parameters
such as VIN, VOUT, IOUT, system priorities (e.g. efficiency, performance, cost), and any additional
key features, such as power good (PG) indication or enable control.

=>Once the designer has defined these requirements, use a parametric search table to find the
best device to meet the desired requirements. The parametric search table is a valuable tool for
designers, as it offers different features and packages available to meet the required parameters
for your application.

=>Every MPS device comes with a datasheet that details what external parts are needed, and
how to calculate their values to achieve an efficient, stable, and high-performance design. The
datasheet can be used to calculate component values such as output capacitance, output
inductance, feedback resistance, and other key system components.

=>Furthermore, you can use simulation tools such as the DC/DC Designer or MPSmart software,
reference application notes, or contact your local FAE with questions.
=>MPS offers a variety of efficient, compact linear and switching voltage regulators, including
the HF500-x family, MP171x family, MP20056, MP28310, MPQ4572-AEC1, and MPQ2013-
AEC1.

18
Voltage multipliers

=>A voltage multiplier is a specialized rectifier circuit producing an output that is theoretically
an integer time the AC peak input, for example, 2, 3, or 4 times the AC peak input. Thus, it is
possible to get 200 VDC from a 100 Vpeak AC source using a doubler, and 400 VDC from a
quadrupler. Any load in a practical circuit will lower these voltages.
=>We’ll first go over several types of voltage multipliers—voltage doubler (half- and full-wave),
voltage tripler, and voltage quadrupler—then make some general notes about voltage multiplier
safety and finish up with the Cockcroft-Walton multiplier.

Voltage Doubler

=>A voltage doubler application is a DC power supply capable of using either a 240 VAC or 120
VAC source. The supply uses a switch-selected full-wave bridge to produce about 300 VDC
from a 240 VAC source.
=>The 120 V position of the switch rewires the bridge as a doubler, producing about 300 VDC
from the 120 VAC. In both cases, 300 VDC is produced. This is the input to a switching
regulator producing lower voltages for powering, say, a personal computer.

Half-Wave Voltage Doubler

=>The half-wave voltage doubler in Figure 1(a) is composed of two circuits: a clamper shown
in (b) and a peak detector (half-wave rectifier) in (c). C 2 has been added to this peak detector
(half-wave rectifier).

19
 Figure 1. Half-wave voltage doubler (a) is composed of (b) a clamper and (c) a half-wave
rectifier.

Half-wave Voltage Doubler Operation Circuit Analysis

=>Referring to Figure 1(b) above , C 2 charges to 5 V (4.3 V considering the diode drop) on the
negative half cycle of AC input. The right end is grounded by the conducting D 2. The left end is
charged at the negative peak of the AC input. This is the operation of the clamper.
=>During the positive half cycle, the half-wave rectifier is shown in Figure 1 (c) above . Diode
D2 is out of the circuit since it is reverse-biased. C 2 is now in series with the voltage source. Note
the polarities of the generator and C 2, series aiding.
=>Thus, rectifier D 1 sees a total of 10 V at the peak of the sinewave, 5 V from the generator, and
5 V from C2. D1 conducts waveform v(1), as illustrated in Figure 2, charging C 1 to the peak of
the sine wave riding on 5 V DC - Figure 2 v(2). Waveform v(2) is the output of the doubler,
which stabilizes at 10 V (8.6 V with diode drops) after a few cycles of sine wave input.

Figure 2. Voltage doubler: v(4) input. v(1) clamper stage. v(2) half-wave rectifier stage, which is
the doubler output.

Full-Wave Voltage Doubler

=>The full-wave voltage doubler is composed of a pair of series stacked half-wave

rectifiers (Figure 3). The corresponding netlist is in Figure below.

20
Full-Wave Voltage Doubler Operation Analysis

=>The bottom rectifier charges C 1 on the negative half cycle of input. The top rectifier charges
C2 on the positive half-cycle. Each capacitor takes on a charge of 5 V (4.3 V considering diode
drop). The output at node 5 is the series total of C 1 + C2 or 10 V (8.6 V with diode drops).

Figure 3. Full-wave voltage doubler consists of two half-wave rectifiers operating on alternating
polarities.

Note that the output v(5) in Figure 4 reaches full value within one cycle of the input v(2)
excursion.

21
Figure 4. Full-wave voltage doubler: v(2) input, v(3)voltage at mid point, v(5) voltage at output.

Deriving Full-wave Doublers from Half-wave Rectifiers

=>The figure below illustrates the derivation of the full-wave doubler from a pair of opposite
polarity half-wave rectifiers (a). The negative rectifier of the pair is redrawn for clarity (b). Both
are combined at (c), sharing the same ground. At (d), the negative rectifier is re-wired to share
one voltage source with the positive rectifier.
=>This yields a ±5 V (4.3 V with diode drop) power supply, though 10 V is measurable between
the two outputs. The ground reference point is moved so that +10 V is available with respect to
the ground.

Figure 5. Full-wave doubler: (a) Pair of doublers, (b) redrawn, (c) sharing the ground, (d)
sharing the same voltage source. (e) move the ground point.

Voltage Tripler

=>A voltage tripler (Figure 6 below) is built from a combination of a doubler and a half-wave
rectifier (C3, D3). The half-wave rectifier produces 5 V (4.3 V) at node 3.
=>The doubler provides another 10 V (8.4 V) between nodes 2 and 3, for a total of 15 V (12.9 V)
at the output node 2 with respect to ground.

22
 Figure 6. Voltage tripler composed of doubler stacked atop a single stage rectifier.

=>Note that V(3) in Figure 7 rises to 5 V (4.3 V) on the first negative half cycle. Input v(4) is
shifted upward by 5 V (4.3 V) due to 5 V from the half-wave rectifier.
=> And 5 V more at v(1) due to the clamper (C 2, D2). D1 charges C1 (waveform v(2)) to the peak
value of v(1). The net list for the schematic of Figure 6 is also included in Figure 7.

Figure 7. Voltage tripler: v(3) half-wave rectifier, v(4) input+ 5 V, v(1) clamper, v(2) final
output.

23
Voltage Quadrupler

=>A voltage quadrupler is a stacked combination of two doublers shown in Figure 8. Each
doubler provides 10 V (8.6 V) for a series total at node 2 with respect to ground of 20 V (17.2 V)

Figure 8. Voltage quadrupler, composed of two doublers stacked in series, with output at node 2.

=>The waveforms of the quadrupler are shown in Figure 9. Two DC outputs are available: v(3),
the doubler output, and v(2), the quadrupler's output.
=>Some of the intermediate voltages at clampers illustrate that the input sinewave (not shown),
which swings by 5 V, is successively clamped at higher levels: at v(5), v(4), and v(1). Strictly
v(4) is not a clamper output. It is simply the AC voltage source in series with the v(3) doubler's
output. Nonetheless, v(1) is a clamped version of v(4).

24
 Figure 9. Voltage quadrupler: DC voltage available at v(3) and v(2). Intermediate waveforms:
Clampers: v(5), v(4), v(1).

Notes on Voltage Multipliers and Line-driven Power Supplies

=>Some notes on voltage multipliers are in order at this point. The circuit parameters used in the
examples (V = 5 V 1 kHz, C = 1000 pf) do not provide much current, micro amps. Furthermore,
load resistors have been omitted. Loading reduces the voltages from those shown.
=>If the circuits are to be driven by a kHz source at low voltage, as in the examples, the
capacitors are usually 0.1 to 1.0 µF, so milliamps of current are available at the output.
=>If the multipliers are driven from 50/60 Hz, the capacitor is a few hundred to a few thousand
microfarads to provide hundreds of milliamps of output current. If driven from line voltage, pay
attention to the polarity and voltage ratings of the capacitors.
=>Finally, any direct line-driven power supply (no transformer) is dangerous to the experimenter
and line-operated test equipment. Commercial direct-driven supplies are safe because the
hazardous circuitry is enclosed to protect the user.
=>When bread boarding these circuits with electrolytic capacitors of any voltage, the capacitors
will explode if the polarity is reversed. Such circuits should be powered up behind a safety
shield.

Cockcroft-Walton Multiplier

=>A voltage multiplier of cascaded half-wave doublers of arbitrary length is known as

a Cockcroft-Walton multiplier, as illustrated in Figure 10. This multiplier is used when a high
voltage at a low current is required.
=>The advantage over a conventional supply is that an expensive high-voltage transformer is not
required– at least not as high as the output.

25
 Figure 10. Cockcroft-Walton x8 voltage multiplier; output at v(8).

=>The pair of diodes and capacitors to the left of nodes 1 and 2 in Figure 10 constitute a half-
wave doubler. Rotating the diodes by 45 o counterclockwise and the bottom capacitor by
90o makes it look like Figure prior (a). Four of the doubler sections are cascaded to the right for a
theoretical x8 multiplication factor.
=>Node 1 has a clamper waveform (not shown), a sinewave shifted up by 1x (5 V). The other
odd-numbered nodes are sinewaves clamped to successively higher voltages. Node 2, the output
of the first doubler, is a 2x DC voltage v(2) in Figure 11. Successive even numbered nodes
charge to successively higher voltages: v(4), v(6), v(8).

Figure 11. Cockcroft-Walton (x8) waveforms. Output is v(8).

=>Without diode drops, each doubler yields 2Vin or 10 V, considering two diode drops (10-1.4)
= 8.6 V is realistic. For a total of 4 doublers, one expects 4·8.6 = 34.4 V out of 40 V.
=>Consulting Figure 10, v(2) is about right; however, v(8) is < 30 V instead of the anticipated
34.4 V. The bane of the Cockcroft-Walton multiplier is that each additional stage adds less than
the previous stage. Thus, a practical limit to the number of stages exists.
=>It is possible to overcome this limitation with a modification to the basic circuit. [ABR] Also
note the time scale of 40 ms compared with 5 ms for previous circuits. It required 40 msec for
the voltages to rise to a terminal value for this circuit.

26
 =>The netlist in Figure 11 above has a “.tran 0.010 m 50 m” command to extend the simulation
time to 50 ms, though only 40 ms is plotted.
The Cockcroft-Walton multiplier serves as a more efficient high voltage source for
photomultiplier tubes requiring up to 2000 V. [ABR] Moreover, the tube has numerous dynodes,
terminals requiring connection to the lower voltage “even numbered” nodes. The series string of
multiplier taps replaces a heat-generating resistive voltage divider of previous designs.
An AC line-operated Cockcroft-Walton multiplier provides high voltage to “ion generators” for
neutralizing electrostatic charge and for air purifiers.

Voltage Multiplier Review

 A voltage multiplier produces a DC multiple (2,3,4, etc) of the AC peak input voltage.
 The most basic multiplier is a half-wave doubler.
 The full-wave double is a superior circuit as a doubler.
 A tripler is a half-wave doubler and a conventional rectifier stage (peak detector).
 A quadrupler is a pair of half-wave doublers
 A long string of half-wave doublers is known as a Cockcroft-Walton multiplier.

Methods of protection

 Fuses (and Circuit Breakers) ...

 RCD, RCCB, GFCI or ELCB. ...
 Zener Diodes and TVS Diodes. ...
 Metal Oxide Varistors (MOVs) ...
 Crowbar Protection - Gas Discharge Tubes (GDTs) ...
 Other Types of Crowbar Protection. ...
 Ground Isolation Techniques. ...
 Capacitors.

=>Protecting a power supply and its load from each other’s faults requires components and
functions such as the fuse, undervoltage lockout, crowbars, and clamps.

=>Power protection is like insurance: You pay for it, yet hope you don’t need it. But it’s not a
simple “purchase.” The first protection question is, “What am I seeking to protect and against
what event(s)?” The answer is two-fold: the supply and its components need protection from
load faults, while the load needs protection against supply faults.

=>The second question is, “What kinds of faults?” These can be excessive current or voltage,
ranging from a short circuit and the associated high currents, to transients and voltage spikes due
to ESD (electrostatic discharge) or even lightning. Some faults are due to component failure,

27
where others can be due to a wiring mistake. Finally, in some cases, the fault mode can even be a
supply voltage that’s too low.

=>The components that must be added to a circuit or system to provide circuit protection receive
little appreciation. They don't enhance the functionality, nor do they add to the glamour, appeal,
or performance of the product.

=>They take up space, add cost, complicate the bill of materials (BOM), and usually sit quietly
without doing anything. That's the situation until they’re needed, when they’re expected to
quickly react and protect other components in the circuit from malfunction or even destruction.

=>Protecting against any and all possible power problems is complex, costly, and generally
unneeded. It’s the role of the design engineer to assess whether fault protection is needed; after
all, there’s little reason to protect a smartphone against power-rail spikes caused by lightning.

=>There are many protection-related components and techniques to choose from. Most are
passive, but some are active. This article will focus only on the passive or mostly passive types.

=>As with most design issues, overlapping perspectives exist on the same basic topic. For power
protection, you can look at it first in terms of potential fault conditions and then the options for
dealing with them, or in terms of the various protection components and then the faults for which
they’re used.

=>A circuit or system may use one or multiple levels and types of protection. Many of these
protection features are built into the power supply, whether it’s a dc-dc converter IC or a larger
ac-dc unit. In other cases, such as when an engineer is designing a supply from individual
components, some of them may need to be added.

It Begins with Overcurrent and Fuses

=>Whether a result of a short circuit external to the supply or within it, overcurrent is a major
concern. It can initiate a cascade of additional failures, put users at risk, and even start a fire.

=>The oldest solution is a fuse (also called a fusible link) (Fig. 1) with apparently simple
operation: when the current flow exceeds the fuse’s current threshold, the current causes the
special wire within the fuse to overheat (I2R heating), melt, and open, thus cutting the current to
zero.

28
 1. Many schematic-diagram symbols exist for the fuse; these are a few of them. (Source:
Slideplayer.com)

=>Once the fuse blows open, the current flow is completely cut off, and can be restored only by
replacing the fuse itself, which is either a benefit or a negative, depending on the application.
=>The more-complex circuit breaker is an alternative to the fuse which doesn’t need replacement
after activation. Some breakers are thermally activated, some are magnetically activated; either
way, like the fuse, the breaker is a current-triggered device.

=>Although the fuse is “ancient,” it’s inexpensive, reliable, easy to design in, and effective.
Basic fuses are available with ratings under 1 A to hundreds of amps (Fig. 2). While fuses do
have a voltage rating, that’s primarily for contact rating and physical spacing, as the fuse itself is
triggered only by the current through it and not the voltage.

=>For some devices, the fuse isn’t a good choice (think of a smartphone’s limited-energy
internal power circuits), while it’s the best choice in others, and often used in conjunction with
other protection techniques. The fuse is frequently added to help a product meet regulatory safety
requirements, due to the directness of its functionality.

=>Note that despite their simple principle, they’re offered in many variations and subtleties, such
as how long it takes for it to react and open the circuit (which is a function of both the current
and elapsed time). Fuse datasheets have many charts showing performance under various
conditions, and specialty fuses are available for unique situations.

Undervoltage Lockout (UVLO)

=>UVLO ensures that a power-supply (or dc-dc) converter doesn’t attempt to operate when its
own input voltage is too low (Fig. 3). This is done for two reasons. First, circuitry within the
supply or converter may malfunction or act in an indeterminate way if the input dc voltage is too

29
 low, and some higher-power components may actually be damaged. Second, it prevents the
supply/converter from drawing on primary power if it can’t produce valid output power.

3. A power supply doesn’t “instantly” come up to full output, but instead has turn-on and turn-
off transitional ranges and time. UVLO ensures the supply doesn’t attempt to provide a full
output when its input voltage is below the minimum needed for proper operation. (Source: Texas
Instruments)

=>To implement UVLO, a small, low-power comparison circuit within the supply/converter
compares the input voltage to a preset threshold and puts the unit into quiescent mode until the
threshold is crossed. To ensure that the UVLO doesn’t “chatter” around the threshold, a small
amount of hysteresis is added.

Overvoltage Protection (OVP)

=>Although a supply or power converter is designed to normally produce a fixed dc-output

voltage, an internal failure in the supply may cause this voltage to rise, and possibly damage the
load to which the supply is connected. OVP is a function that monitors the supply/converter
output versus an internal reference and short-circuits that output if the voltage rises above the
threshold. The OVP must do several things:

 Obviously, prevent any excessive voltage from appearing at the protected components.
 Not interfere with normal operation, but instead be “invisible” to the power supply.
 Distinguish between normal transient voltage fluctuations and excessive overvoltage.
 Be fast, and respond before the load is damaged when a genuine overvoltage situation does
occur.
 Not have false positives (false trips), which are a nuisance, and not fail to respond to real
overvoltage conditions.

The Crowbar

=>One widely used OVP function is the “crowbar,” supposedly so named because it has the
same effect as placing a metal crowbar across the output and thus shorting the output

30
voltage. There are two kinds of crowbars: one where the crowbar, once tripped, will only be reset
if the power is turned off; and one where it will reset itself once the fault is cleared.

=>The second one is useful when the condition that tripped the crowbar is due to some sort of
transient rather than a hard failure in the supply. While most supplies now come with a built-in
crowbar, many vendors offer a small, separate crowbar circuit that can be added to an existing
supply if needed.

=>The crowbar is a normally high-impedance circuit across the supply output (or input of the
load to be protected) (Fig. 4).

=>It transforms into a low-impedance circuit when an overvoltage situation occurs and triggers
it, and it stays in low-impedance mode until the current decreases below the “holding current.”
Subsequently, it returns to the high-impedance, normal-operation state.

=>The crowbar must be able to handle the current flowing through it during the time the supply
is in overvoltage state.

4. This crowbar circuit operates from an 8-V supply. The Zener diode sets overprotection at 9.1
V at that voltage; the diode starts to conduct, causing a trigger signal to switch on the thyristor
Q1 (note that the fuse is for protection against excessive current).

=>Other common crowbars are based on thyristor surge protectors (TSPs). These are silicon-
based PNPN devices with a breakdown voltage that can be set precisely by their manufacturer.
TSPs are offered in many package types and can dissipate various levels of surges.

=>There’s also the gas discharge tube (GDT), which is a miniature spark gap usually housed in a
ceramic enclosure and PCB-compatible. When triggered by a high voltage, the spark gap
conducts and all current flow is diverted.

31
=>Spark gaps can be manufactured so that they protect from modest voltages (around 100 V) to
thousands of volts. When the overvoltage situation clears, the TSP or GDT go back to normal,
high-impedance mode.

The Clamp
=>A complement to the crowbar is the clamp, which prevents the voltage from exceeding a
preset level. Clamps are often referred to as transient voltage suppressors (TVSs), since they may
be protecting against a startup transient or inductive transient rather than an actual failure (Fig.
5). For most clamps, the clamp function releases when the overvoltage condition clears.

5. The TVS, which is simple to apply, is placed between the voltage source and the load without
any interfering components that might affect its performance or impede the current path.
(Source: Enthusiast Wiring Diagrams/http://rasalibre.co/)

=>A clamp conducts just enough current to maintain the voltage across it at a safe, desired value
when the transient is above the clamp’s conduction voltage. It must be rated for the power it will
have to dissipate for a specific time, usually a relatively short transient event.

=>The TVS clamp—a silicon bipolar junction device similar to a basic rectifier diode but
designed to survive reverse breakdown-voltage situations—is available with breakdown voltages
from 4 to 500 V, and in various power ratings to provide different surge-protection capabilities.
A TVS is a bipolar junction device.

=>Compared to a clamp, the crowbar’s low holding voltage lets it carry higher fault current
without dissipating much power, so that it can handle higher currents and do so for longer
periods (Fig. 6). It’s also easier to configure the circuit so that the crowbar also causes a fuse to
blow (and thus stop current flow completely), if that’s desired.

32
 6. The basic response of a crowbar and a clamp to a short-lived surge shows how the crowbar
goes to a near short-circuit while the clamp limits to voltage increase. (Source: Bourns)

=>A clamp can also be built using a metal-oxide varistor (MOV), a bidirectional semiconductor
voltage-transient suppressor device. It conducts (i.e., switches) at a voltage related to the size and
number of special grains between its leads. MOV breakdown voltages range from about 14 V to
over a 1,000 V, with the larger ones intended to handle several kilovolt-amps (kVA), such as
from a lightning surge.

=>MOVs are low cost, fast acting, easy to use, and offered in many voltage ratings, and their
own failure mode is to short circuit (which is preferred in most fail-safe designs). However, they
can only dissipate small amounts of power, so they’re suited only for short-term and transient
OVP situations

=>In general, crowbars are better for long-term faults, while clamps are best suited for transient
events rather than outright supply failures. Many commercial power supplies incorporate both a
crowbar and a clamp.

=> If the concern is outright failure and associated high-current flow, which would soon
overwhelm the dissipation rating of the crowbar or clamp, the design should also include a fuse
or circuit breaker. The fuse/breaker will eventually blow from the overcurrent related to the
excess voltage and thus provide multi-factor protection.

Don’t Forget Thermal Protection

=>Finally, there’s the issue of thermal-overload protection. By its nature, any power supply
generates heat because it’s less than 100% efficient, and even an efficient supply generates a
potentially troublesome amount. For example, a 100-W supply that’s 90% efficient still

33
dissipates 10 W, which is very capable of warming up a small, sealed enclosure. For this reason,
the supply must be designed with sufficient active cooling (e.g., via a fan) or passive cooling
(achieved by convection air flow and conductive cooling paths).

=>But what happens when the fan fails, the air-flow path is blocked, or another heat source is
introduced into the enclosure? The supply may exceed its temperature rating, which shortens its
life and may even cause immediate malfunction.

=>The solution is a sensor within the supply (as a discrete device or incorporated within an IC)
that senses the ambient temperature and puts the supply into a quiescent mode if it exceeds a
preset limit. Some implementations allow the supply to resume operation if the temperature
drops, while others do not.

=>Power-supply protection is, not surprisingly, a nuanced topic. There are issues of current,
voltage, and power handling, dissipation by the protection circuit or components, and fault
duration, as well as protection component placement, cost, and footprint.

=>But protection is also good engineering practice and often mandated by regulatory standards.
Again, it’s like insurance: It comes in many forms and covers many types of bad events. You
hope you don’t need it, but there’s a chance you will for a variety of possible reasons.

34