Search for notes by fellow students, in your own course and all over the country.
Browse our notes for titles which look like what you need, you can preview any of the notes via a sample of the contents. After you're happy these are the notes you're after simply pop them into your shopping cart.
Title: Power Supplies for LED Driving
Description: The essential guide to designing cost- and energy-efficient power systems for LED devices.
Description: The essential guide to designing cost- and energy-efficient power systems for LED devices.
Document Preview
Extracts from the notes are below, to see the PDF you'll receive please use the links above
Elsevier US
PLD
Prelims H8341
14 2 2008
Page:1
Trim:7
...
25 in
Power Supplies for LED Driving
Elsevier US
PLD
Prelims H8341
14 2 2008
Page:2
Trim:7
...
25 in
Elsevier US
PLD
Prelims H8341
14 2 2008
Page:3
Trim:7
...
25 in
Power Supplies for LED Driving
Steve Winder
AMSTERDAM • BOSTON • HEIDELBERG • LONDON
NEW YORK • OXFORD • PARIS • SAN DIEGO
SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO
Newnes is an imprint of Elsevier
Elsevier US
PLD
Prelims H8341
14 2 2008
Page:4
Trim:7
...
25 in
Newnes is an imprint of Elsevier
30 Corporate Drive, Suite 400, Burlington, MA 01803, USA
Linacre House, Jordan Hill, Oxford OX2 8DP, UK
Copyright Ó 2008 by Elsevier Inc
...
No part of this publication may be reproduced, stored in a retrieval system, or
transmitted in any form or by any means, electronic, mechanical, photocopying,
recording, or otherwise, without the prior written permission of the publisher
...
com
...
com), by selecting ‘‘Support & Contact’’
then ‘‘Copyright and Permission’’ and then ‘‘Obtaining Permissions
...
Library of Congress Cataloging-in-Publication Data
Application submitted
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
...
books
...
com
Printed in the United States of America
08 09 10 11 12 13
10 9 8 7 6
5 4 3
2 1
Working together to grow
libraries in developing countries
www
...
com | www
...
org | www
...
org
Elsevier US
PLD
Prelims H8341
14 2 2008
Page:5
Trim:7
...
25 in
Table of Contents
Preface
...
1
1
...
2
Objectives and General Approach
...
2
Chapter 2: Characteristics of LEDs
...
1
2
...
3
2
...
5
Applications for LEDs
...
12
Equivalent Circuit to an LED
...
14
Common Mistakes
...
17
3
...
2
3
...
4
3
...
17
Current Source
...
29
Common Mistakes
...
31
Chapter 4: Linear Power Supplies
...
1
4
...
3
4
...
33
Advantages and Disadvantages
...
37
Common Errors in Designing Linear LED Drivers
...
39
5
...
2
5
...
40
Buck Circuits for DC Applications
...
46
www
...
com
Elsevier US
PLD
Prelims H8341
14 2 2008
vi
Page:6
Trim:7
...
25 in
Table of Contents
5
...
5
5
...
7
Buck Circuits Powered by an AC Phase Dimmer
...
54
Double Buck
...
59
Chapter 6: Boost Converters
...
1
6
...
3
6
...
5
6
...
62
HV9912 Boost Controller
...
67
Design of a Discontinuous Conduction Mode Boost LED Driver
...
98
Conclusions
...
99
7
...
2
7
...
4
7
...
100
SEPIC Buck-Boost Converters
...
139
Common Mistakes in Boost-Buck Circuits
...
140
Chapter 8: LED Drivers with Power Factor Correction
...
1
8
...
3
8
...
5
Power Factor Correction
...
142
Buck-Boost-Buck (BBB)
...
147
Conclusions
...
149
9
...
2
9
...
150
Three Winding Fly-Back
...
158
Chapter 10: Essentials of Switching Power Supplies
...
1 Linear Regulators
...
2 Switching Regulators
...
175
11
...
2
11
...
4
Discrete Semiconductors
...
182
The Printed Circuit Board (PCB)
...
193
www
...
com
Elsevier US
PLD
Prelims H8341
14 2 2008
Page:7
Trim:7
...
25 in
Table of Contents
vii
Chapter 12: Magnetic Materials for Inductors and Transformers
...
1
12
...
3
12
...
5
12
...
197
Iron Dust Cores
...
198
Core Shapes and Sizes
...
199
Copper Losses
...
203
13
...
2
13
...
4
EMI Standards
...
205
EMC Standards
...
215
Chapter 14: Thermal Considerations
...
1 Efficiency and Power Loss
...
2 Calculating Temperature
...
3 Handling Heat – Cooling Techniques
...
225
15
...
2
15
...
4
15
...
225
Circuit Breakers
...
226
Capacitor Ratings
...
227
Bibliography
...
231
Author Biography
...
newnespress
...
5Â9
...
5Â9
...
Later, yellow/amber, green and finally blue colored LEDs became available, which
triggered an explosion in applications
...
Recently blue colored LEDs have been
combined with yellow phosphor to create white light
...
Driving a single LED, or a long string of LEDs connected in series, has relatively few
problems when the current is low (may be 20 mA)
...
Of course, a simple linear regulator
could be used if power dissipation was not an issue, or a simple resistor if current
regulation is not critical
...
A switching regulator is essential if the LED string voltage is higher
than the supply voltage, or if the supply voltage has wide variation
...
This book describes a number of LED driving methods
...
However, the content is not
exhaustive and further reading in some peripheral topics will be necessary
...
Steve Winder, 2007
...
newnespress
...
The older type of LED requiring a 20 mA supply can be
abused to some extent
...
There are
several manufacturers that produce power levels up to 20 W, and more; these higher
powers use LED chip arrays
...
Power LEDs are being used in increasing numbers; in channel lighting (signage), traffic
lights, street lights, automotive, mood lighting (colour changing ‘wall wash’), theatre
lighting for steps and emergency exits
...
This book will cover all types of LED drivers, from low power to UB-LEDs and beyond
...
In a few cases a linear regulator can be
used, which is simple, but most cases require a switching power supply with a constant
current output
...
With a
switching supply, the main issues are EMI and efficiency, and of course cost
...
1
...
It is important to understand the
characteristics of components before they can be used effectively
...
newnespress
...
This section will
highlight errors that engineers have made, and how these can be avoided, with the
hope that readers will not make the same mistakes
...
Our
own mistakes are more memorable, but also more costly!
Usually the first problem for a designer is to choose between different topologies
...
Power supply design equations will be given and example designs of practical supplies
will be worked through
...
Power LEDs generate a lot of heat in a
small area, which makes thermal management difficult, so an adjacent power supply
should be efficient and not add too much heating effect
...
In many cases, customers want to use
standard off-the-shelf parts, because of ease of purchase and cost
...
In some cases the
difference is negligible
...
All component value changes will introduce some ‘error’ in the final result
...
It will also show how the
calculated component value compares with the actual value used, and will include a
description of why the choice was made
...
2 Description of Contents
In Chapter 2, the description of some LED applications will show the breadth of the
LED driving subject and how LEDs’ physical characteristics can be used to an
advantage
...
One of the characteristics is colour; an LED
emits a very narrow band of wavelengths so the colour is fairly pure
...
newnespress
...
But the current level determines the light
output level: higher currents give higher luminosity from a given LED
...
Chapter 3 will show that there are several ways to drive LEDs
...
The trouble is that this is not a good match
for the LED power requirement
...
So, if we have a constant voltage supply, we need to have some form of current
control in series with the LED
...
In fact, a short circuit in any part of the
circuit could lead to a catastrophic failure so we may have to provide some
protection
...
This could also be used to detect an open circuit
...
Chapter 3 continues describing features of constant current circuit
...
Open circuit protection can take many forms
...
If the circuit failed open the voltage would rise up
to the level of the open circuit protection limit, which could also be detected
...
Advantages include no EMI generation, so
no filtering is required
...
Chapter 5 describes the most basic of switching LED drivers: the buck converter
...
The reader will be taken through the design process, followed by
an example design
...
These are used in many applications including
LCD backlights for television, and computer and satellite navigation display screens
...
newnespress
...
The reader will be taken through the design process, followed
by an example design, for both continuous mode and discontinuous mode drivers
...
These have the ability to drive a load that
is either higher or lower voltage compared to the input
...
Chapter 8 describes specialist converters: buck-boost and buck (BBB), and Bi-Bred
...
They combine power factor correction with constant
current output, but in many cases can be designed without electrolytic capacitors and
so are useful for high reliability applications
...
Chapter 9 describes fly-back converters
...
The use of
two windings or more in an inductor permits isolation of the output
...
Chapter 10 covers topics that are essential when considering a switch mode power
supply
...
The
advantages, disadvantages and limitations of each type will be analyzed in terms of
supply voltage range and the ability to perform PFC (power factor correction)
...
Chapter 11 describes electronic components for power supplies
...
There are so many different types of switching
elements: MOSFETs, power bipolar transistors and diodes, each with characteristics
that affect overall power supply performance
...
Magnetic components are often a mystery for many electronic engineers, and these
will be briefly described in Chapter 12
...
Then there are different core shapes
www
...
com
Introduction
5
and sizes
...
EMI and EMC issues are the subjects of Chapter 13
...
Good EMI design
techniques can reduce the need for filtering and shielding, so it makes sense to
carefully consider this in order to reduce the cost and size of the power supply
...
It is no use having
an otherwise excellent circuit that is destroyed by externally produced interference
...
Chapter 14 discusses thermal issues for both the LEDs and the LED driver
...
The LED itself dissipates most of the
energy it receives (volts times amps) as heat: very little energy is radiated as light,
although manufacturers are improving products all the time
...
Calculating the temperature is
important because there are operating temperature limits for all semiconductors
...
The product
must not injure people when it is operating
...
When the equipment is powered from the AC mains supply, the issues of
isolation, circuit breakers and creepage distances must be considered
...
newnespress
...
The fixed atoms have
positive and negative charge, respectively
...
This ‘intrinsic region’ now has the positive and negative charge of the fixed atoms,
which opposes any further free charge combination
...
In order for a P-N junction to conduct, we must make the P-type material
more positive than the N-type
...
Conduction
takes place when (in silicon) there is about 0
...
This potential difference gives electrons enough energy to
conduct
...
The first LEDs were made from gallium arsenide
(GaAs) and produced infrared light at about 905 nm
...
When a voltage is applied across the LED,
electrons are given enough energy to jump into the conduction band and
current flows
...
See Figure 2
...
www
...
com
8
Chapter 2
Electron migration
P
N
Hole Migration
Radiative transitions
Non-radiative Transitions
Figure 2
...
2
...
The energy gap in GaAsP material is higher than
GaAs, so the light wavelength is shorter
...
The most typical application was to show that
equipment was powered, or that some feature such as ‘stereo’ was active in a radio
...
When yellow and green LEDs became available, the number of applications
increased
...
For example, green = OK, yellow = requires attention,
red = faulty
...
One characteristic of the light from an LED is that it occupies a narrow spectrum
about 20 nm wide; the color is fairly pure
...
The very narrow
spectrum of a laser is important because, when used with optical fiber systems, the
narrow spectral width allows a wide system bandwidth
...
Another important characteristic of LED light is that current is converted into light
(photons)
...
So
www
...
com
Characteristics of LEDs
9
dimming lights by lowering the current is possible
...
Dimming by
pulse width modulation (PWM) is a viable alternative used by many people
...
The pulse width is reduced to dim the light, or increased to brighten the light
...
When adding colored light, red, green and blue make light that appears
white to the human eye
...
There are big gaps in the color
spectrum, but the eye does not notice
...
The yellow phosphor
creates a wide spectrum and, when combined with the blue, appears white
...
Illuminating modern
resins used in tooth filling materials with blue light will harden the resin
...
Some interesting applications rely on the purity of the LED color
...
Photographic dark rooms can use colors where film is insensitive – traditionally dark
rooms have been illuminated by red colored incandescent lamps
...
It should be noted that the color of an LED would change as the LED’s temperature
changes
...
The only way to control ambient temperature
is to add a cooling fan, or by placing the LED away from the source of heat
...
The early LEDs were all rated at 20 mA and the forward voltage drop was about 2 V
for red, higher for other colors; later low current LEDs were created that operated
from a 2 mA source
...
Lumileds was created by HP and
Philips in 1999 and produced the first 350 mA LED
...
Power LEDs
www
...
com
10
Chapter 2
are being used in increasing numbers; in channel lighting (signage), traffic lights,
street lights, automotive, mood lighting (color changing ‘wall wash’), and also in
theaters for lighting steps and emergency exits
...
2
...
In the past, channel lighting used cold-cathode or fluorescent tubes, but
these had reliability problems
...
So, to cope with the shapes and environmental conditions, the most viable
technology is LED lighting
...
2: Channel Lighting
...
newnespress
...
One problem with traffic lights is controlling
the wavelength of the yellow (amber) light
...
Another problem is making them fail-safe – authorities
permit some degree of failure, but if more than 20% of the LEDs fail, the entire lamp
must be shut down and a fault reported to maintenance teams
...
This is particularly true if the LED driver circuit contains electrolytic capacitors,
which vent when hot and eventually lose their capacitance
...
Failing LED drivers can give LED lights a bad
name – why have LEDs that can work for over 100,000 hours if the LED driver fails
after 10,000 hours’ operation?
Street lights have been built using medium and high power LEDs
...
In some cases, white and yellow LEDs
are used together to create a ‘warm-white’ light
...
Automotive lighting has many applications; internal lights, headlights, stoplights,
daylight running lights (DRL), rear fog lights, reversing lights, etc
...
Linear drivers are sometimes used if the efficiency is not a critical requirement
...
Automotive stoplights using LEDs have a significant safety advantage over those
using filament lamps
...
In a filament lamp the response time is about 300 ms
...
6 km) per minute, or 88 feet per second
...
Stopping 300 ms sooner, having
seen the previous car’s brake lights earlier, could avoid death or injury
...
www
...
com
12
Chapter 2
Mood lighting is an effect caused by changing the color of a surface and uses human
psychology to control people’s feelings
...
Generally mood lighting systems
use red, green and blue (RGB) LEDs in a ‘wall wash’ projector to create any color in
the spectrum
...
In this case the color changes little – ideally not at all
...
Cold cathode tubes are sometimes used to
backlight computer screens, but here the exact color is not important
...
2 Light Measure
The total light flux is measured in units of lumens
...
At 555 nm, in
the green-yellow part of the spectrum where the eye is most responsive, 1 W = 683 lumen
...
This is the light produced by a lamp, radiating in all
directions equally, to produce 1 lumen per steradian
...
1 meter radius
1 cd
1 lux
or
1 lm/m2
1 m2
Figure 2
...
www
...
com
Characteristics of LEDs
13
A steradian has a projected area of 1 square meter, at a distance of 1 meter from the light
source
...
3
...
There is some competition between LED manufacturers to get the
highest luminous efficacy, but when comparing results it is important to make a note
of the electrical power levels used
...
2
...
The voltage drop depends on
the internal energy barrier required for the photons of light to be emitted, as
described earlier
...
Will every red LED have the same voltage drop? No, because
production variations will mean that the wavelength (color) will not be the same, and
thus the voltage drop will have differences
...
If there are temperature differences between two LEDs, this will give a color change
and hence differences in voltage drop
...
Thus the voltage drop reduces by approximately
2 mV per degree as the temperature rises
...
4
...
The ESR (equivalent series resistance) of a low power 20 mA LED
is about 20 ohms, but a 1 W 350 mA LED has an ESR of about 1–2 ohm (depending on
the semiconductor material used)
...
The ESR will have production variations too
...
For example, if the forward voltage drop increases by from
3
...
55 V (a 50 mV increase) when the forward current goes from 10 mA to
20 mA (a 10 mA increase), the ESR will be 50 mV/10 mA = 5 ohms
...
4, the Zener diode is shown as a perfect device
...
For initial testing of an
LED driver, a 5 W, 3
...
If the
www
...
com
14
Chapter 2
LED
Equivalent Circuit
(Perfect Zener diode)
Figure 2
...
driver is not working as planned the Zener diode may be destroyed, but this is far less
costly than destroying a power LED
...
2
...
5 shows how the forward voltage drop depends on the light
color and on the LED current
...
5 V
Figure 2
...
www
...
com
Characteristics of LEDs
15
voltage drop, Vf, is about 2 V for a red LED and about 3
...
The
exact voltage drop depends on the manufacturer, because of different dopant
materials and wavelengths
...
2
...
This includes connecting strings of LEDs in parallel, with the
assumption that the forward voltage drops are equal and the current will share
equally between the two or more strings
...
For example, a 1 W white Luxeon Star has a typical Vf = 3
...
79 V and the maximum is 3
...
This is over Æ15%
tolerance on the forward voltage drop!
www
...
com
CHAPTER 3
Driving LEDs
3
...
This behavior is like a Zener diode – in fact Zener
diodes make a good test load, rather than using expensive high power LEDs!
Driving a constant voltage load from a constant voltage supply is very difficult,
because it is only the difference between the supply voltage and the load voltage that
is dropped across the ESR
...
A slight variation in the supply voltage, or the load voltage, will cause
a very large change in current; see curve A in Figure 3
...
If the variation in supply voltage and forward knee voltage (Vf ) is known, the
variation is current can be calculated
...
1: LED Current Versus Supply Voltage
...
newnespress
...
Most
supply voltages from a regulated supply have a 5% tolerance, but from unregulated
supplies like automotive power, the tolerance is far greater
...
In practice, the Vf and voltage drop
across ESR are combined, since manufacturers quote the voltage drop at a certain
forward current
...
If there is a large difference between the source and load voltage, and a high ESR,
there is very little difference between the maximum and minimum LED current
...
However, in high
power LED circuits, a large voltage drop across a series resistor will be inefficient and
may cause heat dissipation problems
...
A standard 20 mA LED may have an ESR of 20 ohms, but a 350 mA
LED will have an ESR of 1–2 ohms typically
...
Even in low current LEDs,
the proportional change in current can be high
...
1
...
Low current loads can have a relatively high value
resistance added in series, in order to reduce the slope of the current versus voltage
graph; see curve B in Figure 3
...
With a series resistor added we are able to calculate the variation in current, provided
that the variation in supply voltage and load voltage is known
...
IMIN ¼
IMAX ¼
www
...
com
VSOURCE
VSOURCE
À VLOAD
REXT
MAX À VLOAD
REXT
MIN
MAX
MIN
Driving LEDs
19
As an example, let us drive from an automotive supply; this is a nominal 13
...
Let us select a red LED for
tail-lights (Lumileds Superflux HPWA-DDOO), with a forward voltage drop of
2
...
03 V at 70 mA forward current
...
So the typical voltage
drop at 70 mA needs to be 8
...
43 ohms
...
IMIN ¼
IMAX ¼
VSOURCE
MIN
VSOURCE
MAX
À VLOAD
REXT
À VLOAD
REXT
MAX
MIN
12 À 6:06
¼ 49:5 mA
120
16 À 4:38
¼
¼ 96:83 mA
120
¼
At the high limit of source voltage, the LED is overdriven by 38%
...
78 mA minimum
...
672 V) was included in the
minimum and maximum load voltage values, so we ignored ESR
...
6 ohms
...
Using
the same example, but operating with a typical LED current of 50 mA, we must
modify the results
...
518 V to 2
...
With a typical 13
...
828 V
...
88 ohms, but we already have 9
...
An external
resistor value of 180 ohms is the nearest preferred value for a current of 50 mA
...
6:1
ratio
...
Unless the LEDs are matched (or ‘binned’) to ensure the same forward voltage drop,
the current through one string could be considerably different from the current
through another
...
newnespress
...
Since the LED strings are in parallel, the voltage source for all strings is the
same
...
The forward voltage
also depends on the ambient temperature
...
The traditional way is to connect a current limiting resistor in series with each string
and power all the strings using a single voltage source
...
This method
is inexpensive, but suffers from power inefficiency and heat dissipation
...
A better way of powering the LED array is to regulate the total current through all
the strings and devise a means to divide that total current equally among the LED
strings
...
3
...
2 Active Current Control
Since a series resistor is not a good current control method, especially when the supply
voltage has a wide tolerance, we will now look at active current control
...
Here we will only
consider limiting LED current when the energy is supplied from a voltage source;
driving LEDs using energy from current sources will be discussed in Section 3
...
A current limiter has certain functional elements: a regulating device such as a
MOSFET or bipolar transistor; a current sensor such as a low value resistor; and
some feedback (with or without gain) from the current sensor to the regulating
device
...
2 shows these functions
...
Conduction of the drain-source channel is controlled from the
gate-source voltage, like any other MOSFET
...
As the gate voltage becomes negative with respect to the
source, the device turns off, see Figure 3
...
A typical pinch-off voltage is –2
...
www
...
com
Driving LEDs
21
V+
Load
Current
Limiter
Feedback
Current
Sense
V–
Figure 3
...
Id
+V
I
d
g
s
V
0V
VTH
Vgs
Figure 3
...
A current limiting circuit with a depletion mode MOSFET uses a resistor in series
with the source to sense the current (see Figure 3
...
The gate is connected to the
negative supply (0 V)
...
newnespress
...
4: Depletion MOSFET Current Limiter
...
The voltage at the MOSFET source increases in potential compared to the
0 V rail and the MOSFET gate
...
At a certain point, when the voltage drop
approaches the MOSFET pinch-off voltage, the MOSFET will tend to turn off and
thus regulate the current
...
A device with a typical Vth of –2
...
5 V to –3
...
However, the advantage is that high drain-source
breakdown voltages are possible and so a limiter designed using a depletion-mode
MOSFET can protect against short transients (longer periods of high voltage would
tend to overheat the MOSFET)
...
5
...
The disadvantage is that there is a
minimum dropout voltage of about 3 V
...
The LM317 has a feedback pin called ‘REF’, and this controls the regulation of the
current
...
25 V, the current
through the LM317 is reduced until the output terminal (OUT) is reduced below 1
...
If accurate current limiters are used, parallel strings of LEDs can be connected to the
same voltage source and then each string will have approximately the same current
...
newnespress
...
5: Linear Regulator as Current Limiter
...
The current limiters described here are purely to show how LEDs can be driven from
a constant voltage supply
...
Switching regulators are described in Chapters 5–10
...
1
...
If the LED goes short circuit, a higher voltage will be placed
across the current limiter
...
If the power dissipation cannot be tolerated when the load goes short circuit, a
voltage monitoring circuit will be needed
...
In the LM317 circuit previously described, the regulator itself has thermal shutdown
...
1
...
We can use this change to detect a failure
...
newnespress
...
6: Shorted Load Indication
...
6, a 10 V Zener diode is used in series with the base of an NPN transistor
...
This pulls the ‘FAILURE’ line to 0 V and
indicates a short circuit across the LEDs
...
2 Current Source
Since an LED behaves like a constant voltage load, it can be directly connected to a
current source
...
A pure current source will not limit the voltage,
so care must be taken to provide some limit; this will be covered in more detail in the
next subsection
...
The simplest of these is a current mirror,
which shares the current equally between strings based on the current flowing
through the primary string
...
7 shows a simple current mirror
...
By connecting all the bases and all the emitters
www
...
com
Driving LEDs
25
V+
Primary
String
–
Maximum
Voltage
Drop
Q1
Q2
Q3
Qn
V–
MATCHED NPN
Figure 3
...
together, every base-emitter junction voltage must be equal and therefore every
collector current must be equal
...
Since the collector and base of transistor Q1 are connected, the transistor
will be fully conducting until the collector voltage falls low enough for the baseemitter current to limit
...
The total current through Q1 to Qn will equal the current
source limit
...
In the slave strings,
some voltage will be dropped across the collector-emitter junction of the transistors
Q2 – Qn
...
3
...
1
Self-Adjusting Current Sharing Circuit
As an alternative, the current sharing circuit shown in Figure 3
...
www
...
com
26
Chapter 3
BIAS BUS
D1
TRANSISTOR BASE BUS
1K
1K
51 R
1K
51 R
1K
51 R
51 R
D1 HEADROOM ADJUST one or more diodes in series
Figure 3
...
Assuming that the LED array is driven from a current source, there will be equal
current division among all connected branches
...
Unlike the simple current mirror, this one automatically
adjusts for the maximum expected voltage difference between strings of LEDs, which
is a function of the number of LEDs in the string and the type of LED used
...
In high reliability applications, the failure of a single LED should not materially
affect the total light output
...
When an LED fails short, the voltage of the string containing the shorted LED will
have less voltage
...
When the failed LED string opens, the current
divider will automatically redistribute the total current among the remaining strings,
thus maintaining the light output
...
www
...
com
Driving LEDs
27
Equality of current division among the branches is dependent on the close matching
of the transistors, which are in close vicinity (ideally a single package with several
matched transistors)
...
Diodes connected to each collector detect the voltage of each branch
...
The cathode of each diode
is connected to a common ‘bias bus’
...
More than one external diode can be used to accommodate large voltage
variations
...
When a branch is not connected, there will be higher base current flowing in the
associated regulating transistor
...
3
...
2
Voltage Limiting
In theory, the output voltage of a constant current driver is not limited
...
In the
case of an LED load, the voltage limit will depend on the number of LEDs in a string
...
Limiting the LED string voltage is
necessary to prevent circuit damage and the voltage level will depend on the
particular circuit
...
4 V AC, so equipment designed to meet these requirements should
consider both mains supply isolation (if applicable) and output voltage limiting
...
www
...
com
28
Chapter 3
3
...
3 Open Circuit Protection
Some constant current drivers, especially switching boost converters, will produce
a sufficiently high voltage to destroy the driver circuit
...
Using a Zener diode to give feedback when the
output voltage exceeds a certain limit is the standard method
...
Other circuits will auto-restart when the open circuit condition is removed (i
...
when
the LEDs are reconnected)
...
A potential
divider comprising two resistors is usually used to scale down the output voltage to
the reference voltage level
...
2
...
In the case of an open circuit LED, the load is removed and so the output voltage from
the current source rises
...
In circuits where over-voltage protection is fitted, this can be used to indicate a failure
...
In a
basic current mirror, as shown in Figure 3
...
Detection of the rise in output
voltage would be a solution
...
The voltage
at the transistor collector of the broken string would fall to zero since there is no
connection to the positive supply, and this could be detected
...
A basic opto-coupler has an LED and a phototransistor in the same package
...
Thus when current is flowing through the LED string
www
...
com
Driving LEDs
29
and the opto-coupler’s internal LED, the photo-transistor is conducting
...
3
...
There are two main reasons for this: (1) cost of an LED, especially
high power devices, can be greater than the driver circuit; and (2) operating high
brightness LEDs for a long time under test conditions can cause eye strain and temporary
sight impairment (if LEDs viewed at close range)
...
3
...
1
Zener Diodes as a Dummy Load
Figure 3
...
This is the simplest
and cheapest load
...
6 V, 5 W Zener diode (3
...
This is not the perfect dummy load
...
42 V of a Lumileds ‘Luxeon Star’ 1 W LED
...
5 ohms, which is higher than the Luxeon Star’s 1 ohm
impedance
...
For simple buck circuits, the impedance only has a small effect
...
9: Zener Diode Dummy Load
...
A constant voltage load will have (in theory at least)
zero impedance, so simply adding a small value series resistor will give the correct
www
...
com
30
Chapter 3
impedance
...
A constant voltage load built using a low cost discrete solution is shown in Figure 3
...
This is a self-powered load and so can be isolated from ground
...
7 V for the emitter-base junction of the
transistor)
...
R
Vz
2N3055
Figure 3
...
The circuit is Figure 3
...
Although the Zener diode does have
a few ohms impedance, the current through it is very small and the effect of the
transistor is to reduce the impedance by a factor equal to the gain HFE
...
Changing
the collector current from 500 mA to 1 A will cause the base current to rise from
10 mA to 20 mA
...
This change at the transistor collector is equivalent to an
impedance of 30 mV/0
...
06 ohm
...
An impedance of 0
...
Because of the potentially high load
current, both the transistor and series resistor should be rated for high power
...
3
...
Instead, 3
...
Only once the circuit has been tested under all conditions should LEDs be used
...
newnespress
...
5 Conclusions
A voltage regulated LED driver is preferred when there are a number of LED
modules that can be connected in parallel
...
An example would be channel lighting, as used in shop name
boards
...
A series connection ensures that all the LEDs have the
same current flowing through them and the light output will be approximately equal
...
An efficiency of 75–90% can be achieved
...
www
...
com
CHAPTER 4
Linear Power Supplies
4
...
The
complete absence of any EMI radiation is one important technical reason
...
However, they also have disadvantages: in
some applications they have low efficiency and hence the introduction of thermal
problems; in other applications, such as when powered from the AC mains supply,
they have the disadvantage of large size
...
1
...
Inside
the LM317 are: (1) a power switch, which is an NPN transistor; (2) a voltage
reference set to produce 1
...
1
...
To produce a certain output voltage, a feedback resistor is connected from the output
(OUT) to the ADJ pin and a sink resistor is connected from the ADJ pin to ground,
thus creating a potential divider
...
www
...
com
34
Chapter 4
LM317
IN
OUT
+
–
ADJ
R1
R2
Figure 4
...
A capacitor on the output terminal also helps with stability
...
Variations of the LM317 regulator include fixed positive voltage versions (LM78xx)
and negative voltage versions (LM79xx), where ‘xx’ indicates the voltage; i
...
LM7805 is a +5 V 1 A regulator
...
This is typically in the range 1 V to 3 V, depending on the
current through the regulator (higher current requires a higher voltage differential)
...
Low dropout voltage regulators use a PNP transistor as the power switch, with the
emitter connected to the IN terminal and the collector connected to the OUT
terminal, see Figure 4
...
They also have a ground pin that enables an internal
reference voltage to be generated independent of the input to output voltage
differential
...
www
...
com
Linear Power Supplies
35
LP2950
INPUT
OUTPUT
+
–
GROUND
Figure 4
...
4
...
2
Voltage Regulators as Current Source or Sink
In Figure 4
...
V+
IN
ADJ
Load
LM317
OUT
IN
ADJ
LM317
R1
OUT
R1
Load
V–
CURRENT SINK
CURRENT SOURCE
Figure 4
...
As previously described, the LM317 regulates when there is +1
...
In Figure 4
...
Current flowing through R1 will produce a voltage drop, with the
www
...
com
36
Chapter 4
OUT pin becoming more positive than the ADJ pin
...
25 V, the LM317 will regulate the current
...
1:25
R1
4
...
3 Constant Current Circuits
There are many constant current circuits; some using integrated circuits, some using
discrete components, and others using a combination of both ICs and discrete
devices
...
A simple constant current sink uses an op-amp with an input voltage range that
extends to the negative rail, as shown in Figure 4
...
In order to set the current level, a
voltage reference is required
...
The voltage reference can be a temperature compensated
precision reference, or a Zener diode
...
2 V
...
4: Constant Current Sink Using Op-amp
...
newnespress
...
2 Advantages and Disadvantages
The advantage of linear power supplies is that they produce no EMI radiation
...
A switching power supply may appear to have few components, but this does not
take into account the EMI filtering and screening
...
If the LEDs are distributed, such as in channel
lighting where there is no opportunity to shield any EMI, both common mode and
differential filtering are required
...
The efficiency is low only if the supply
voltage is somewhat higher than the LED voltage
...
A heatsink may be required, which is
bulky and moderately expensive
...
Linear mains powered LED drivers have the disadvantage of large size, because a stepdown transformer is almost always required (unless the LED string voltage is very near
to the peak AC supply voltage)
...
Smoothing capacitors after the bridge rectifier are also very bulky
...
4
...
Linear voltage and current sources cannot boost the
output voltage so that the output is higher than the input
...
These will be discussed in the next few chapters
...
4 Common Errors in Designing Linear LED Drivers
The most common error is to ignore the power dissipation
...
www
...
com
38
Chapter 4
If the voltage drop is high, the current must be limited to stay within the device
package power dissipation limits
...
Heatsinks are now available for surface mount packages, which eases the problem
...
The voltage rating of the regulator
must be high enough to allow for the output being connected to 0 V (ground)
...
Only after
operating for a short period does the output capacitor charge, which reduces the
voltage drop across the regulator
...
www
...
com
CHAPTER 5
Buck-Based LED Drivers
The first switching LED driver that we will study is the buck converter
...
The limit of about 85% is due to switching delays in the control system
...
The inductor is used to store
energy when the MOSFET is turned on; this energy is then used to provide current
for the LED when the MOSFET is turned off
...
A simple
schematic is shown in Figure 5
...
CIRCULATING CURRENT
(MOSFET OFF)
V+ SUPPLY
350 mA LED
D1
L1
CURRENT
(MOSFET ON)
CONTROLLER
Q1
Figure 5
...
www
...
com
40
Chapter 5
Buck converters are an attractive choice for LED drivers in offline and in low voltage
applications as they can produce a constant LED current at very high efficiencies
...
Coupled with the fact that these converters can be designed to operate
at above 90% efficiencies, the buck-based driver becomes an attractive solution to
drive high brightness LEDs
...
1 An Example Buck Converter Control IC
The Supertex HV9910B integrated circuit was designed especially for LED driving
...
Linear or PWM dimming can also be easily implemented using the IC
...
2
...
5 V
OSC
Rosc
VDD
250 mV
–
CM
+
S
R
LD
Q
–
CM
+
GATE
CS
PWM_D
100 k
HV9910
GND
Figure 5
...
www
...
com
Buck-Based LED Drivers
41
The HV9910B has two current sense threshold voltages – an internally set 250 mV
and an external voltage at the LD pin
...
The low value of sense voltage allows the use of
low resistor values for the current sense, which means high efficiency
...
The IC has an internal regulator that supplies 7
...
The IC is capable of driving the external MOSFET
directly, without the need for additional driver circuitry
...
2 Buck Circuits for DC Applications
For DC applications, the schematic shown in Figure 5
...
10–30 V DC
C1
4
...
3: Buck Converter for DC Applications
...
newnespress
...
2
...
2
...
A larger switching
frequency will result in a smaller inductor, but will increase the switching losses in
the circuit
...
From the HV9910B datasheet, the timing
resistor between the RT pin and ground that is needed to achieve this frequency is 150 k
...
In a buck converter, the duty cycle of the MOSFET switch (proportion of the
time that the switch is turned on), is given by D= VOUT and will also be 80%
...
To prevent instability, it is necessary to operate in constant off-time mode
...
The timing circuit only charges an internal capacitor when the timing
resistor is connected to 0 V; the gate pin is at 0 V when the MOSFET is turned off
...
If we choose a timing resistor that gives a constant off-time of say 5 ms, with an 80%
duty cycle the on-time will be 20 ms
...
At the
other extreme, with a 30 V supply and a 4 V load, the duty cycle will be just 13
...
Now the switching frequency is 173
...
The average
switching frequency will be about 100 kHz, so we can base the selection of other
components on this
...
5
...
3 Choosing the Input Capacitor (C1)
An electrolytic capacitor is good to hold the voltage, but the large ESR of these
capacitors makes it unsuitable to absorb the high frequency ripple current generated
by the buck converter
...
newnespress
...
The
required high frequency capacitance can be computed as
C1 ¼
Io ÂTOFF
ð0:05ÂVmin Þ
In this design example, the high frequency capacitance required is about 4
...
This capacitor should be located close to the inductor L1 and MOSFET switch
Q1, to keep the high frequency loop current within a small area on the PCB
...
5
...
4
Choosing the Inductor (L1)
The inductor value we use depends on the allowed level of ripple current in the LEDs
...
di
The familiar equation for an inductor is E ¼ LÂ dt
...
Another way of writing this is L ¼ Vo,max  dt
...
3 Â Io,max and dt is the off-time
...
Since this value is
a little higher than the calculated value, the ripple current will be less than 30%
...
e
...
www
...
com
44
Chapter 5
5
...
5 Choosing the MOSFET (Q1) and Diode (D2)
The peak voltage seen by the MOSFET is equal to the maximum input voltage
...
Hence, the current rating of the MOSFET is
IFET % Io,max  0:8 ¼ 0:28 A:
Typically a MOSFET with about three times the current is chosen to minimize the
resistive losses in the switch
...
5 A
...
Hence,
Vdiode ¼ VFET ¼ 45 V
The average current through the diode under worst case conditions (minimum duty
cycle) is
Idiode ¼ 0:87 Â Io,max ¼ 0:305 A
Choose a 60 V, 1 A Schottky diode
...
5
...
6 Choosing the Sense Resistor (R2)
The sense resistor value is given by
R2 ¼
0:25
1:15 Â Io,max
This is true if the internal voltage threshold of 0
...
Otherwise,
substitute the voltage at the LD pin instead of the 0
...
Note
that the current limit is set to 15% above the maximum required current, due to
the total 30% ripple specified
...
newnespress
...
625
...
62
...
Say
we want to use a 0
...
25 V at the LD pin by
0
...
625 = 0
...
Note that capacitor C3 is a bypass capacitor for holding up the HV9910B internal
supply VDD during MOSFET switching, when high frequency current pulses are
required for charging the gate
...
2 mF, 16 V is recommended,
although in this design the MOSFET gate charge is very low, so a 1 mF, 16 V can be
used instead
...
2
...
Using an inductor that has too high inductance
...
The
voltage seen across the current sense resistor at switch-on will be almost at the
current sense comparator reference voltage
...
The smallest current surge will
create a voltage spike across the current sense resistor and hence the current
sense comparator will trip
...
A typical switching pattern is one proper switching cycle, where energy is
stored in the inductor, followed by one short switching pulse
...
The result is a less efficient circuit that could suffer from overheating
and EMI problems
...
Using the wrong type of flywheel diode
...
However, in low duty cycle applications the LED current is
flowing in the flywheel diode most of the time
...
45 V
www
...
com
46
Chapter 5
at 350 mA results in 157
...
Note that the forward voltage drop of Schottky diodes
increases with their current rating, so a 30 V Schottky has much lower Vf than
a 100 V Schottky
...
3 Buck Circuits for AC Input
I will now discuss the design of a buck-based LED driver using the HV9910B with
the help of an AC mains input application example
...
The schematic is shown in
Figure 5
...
Live
V+
230 VAC
+
C2
330 nF
C1
33uF
350 mA LED
NTC
D1
UF4005
1
6
C3
2,2uF 10 V
VIN
L1
4,7 mH
VDD
HV9910
GATE
7
LD
5
3
CS
Q1
STD2NM60T4
4
2
PWM_D
Rosc 8
GND
R1
R2
0,62R
470K
Figure 5
...
Designs for an AC input have two problem areas to address
...
Because we are applying a low frequency sinusoidal
high voltage supply, high value input capacitors are needed to hold up the supply
www
...
com
Buck-Based LED Drivers
47
voltage during the cusps between each half-cycle of the input
...
5
...
1
Target Specification
Input voltage = 90 V to 265 V AC (nominal 230 V AC)
LED string voltage = 20–40 V
LED current = 350 mA
Expected efficiency = 90%
5
...
2
Choosing the Switching Frequency and Resistor (R1)
The switching frequency determines the size of the inductor L1
...
A typical switching frequency for high input voltage applications is
fs = 80 kHz, which is a good compromise
...
5
...
3
Choosing the Input Diode Bridge (D1) and the
Thermistor (NTC)
The voltage rating of the diode bridge will depend on the maximum value of the input
voltage
...
5 multiplication factor gives a 50% safety margin
...
The minimum input voltage must be more than half the maximum
LED string voltage, to make sure that the duty cycle stays below 50% and thus
remains stable
...
newnespress
...
The thermistor should limit the inrush current to not more than five times the
steady state current, assuming maximum voltage is applied
...
The calculations suggest that we choose a
thermistor whose resistance is around 380
and RMS current greater than 0
...
5
...
4 Choosing the Input Capacitors (C1 and C2)
The first design criterion to meet is that the maximum LED string voltage must be
less than half the minimum input voltage
...
As we have already seen, the
minimum rectified voltage should be
Vmin,dc ¼ 2 Â Vo,max ¼ 80 V
The hold-up capacitor required at the output of the diode bridge will have to be
calculated at the minimum AC input voltage
...
newnespress
...
pffiffiffi
Vmax,cap ! 2 Â Vmax,ac
) Vmax,cap ! 375 V
Choose a 450 V, 33 mF electrolytic capacitor
...
Thus, a metallized polypropylene capacitor is needed in
parallel with the electrolytic capacitor to absorb the high frequency ripple current
...
33 mF, 400 V
...
5
...
5
Choosing the Inductor (L1)
The inductor value we use depends on the allowed level of ripple current in the LEDs
...
di
The familiar equation for an inductor is E ¼ L Â dt : Considering the time when
the MOSFET switch is off, so that the inductor is supplying energy to the
di
LEDs, E ¼ VLED ¼ Vo,max ¼ L  dt : Another way of writing this is L ¼ Vo,max  dt
...
3 Â Io,max and dt is the off-time dt ¼
...
Vin
fs
Then, the inductor L1 can be computed at the rectified value of the nominal input
voltage as
Vo,max  1 À pffiffiVo,max
2ÂVac,nom
L1 ¼
0:3  Io,max  fs
www
...
com
50
Chapter 5
In this example, L1 = 4
...
The nearest standard value is 4
...
Since this value
is a little higher than the calculated value, the ripple current will be less than 30%
...
e
...
Note that with a large inductance value, the parasitic capacitance across the coil
could be significant and will affect switching losses
...
3
...
Using
a 50% safety rating,
pffiffiffi
VFET ¼ 1:5 Â
2 Â 265 ¼ 562 V
The maximum RMS current through the MOSFET depends on the maximum duty
cycle, which is 50% by design
...
For this application, choose a 600 V, >1 A MOSFET; a
suitable device is an ST part, STD2NM60, rated at 600 V 2 A
...
8
on-resistance
...
Although a MOSFET with lower on-resistance could be used to reduce the
conduction losses, the switching losses, which are caused by parasitic capacitance
and diode reverse recovery current, will then be higher
...
The analogy can be applied to diodes, because they have free
electrons in their conduction band that have to be swept out by the reverse potential
before current flow stops
...
newnespress
...
The peak voltage rating of the diode is the same as the MOSFET
...
The UF4005 is a low cost ultra-fast type, but for
greatest efficiency a faster diode like STTH1R06 should be used
...
The switching loss could be higher that this value, but is less of a
problem in faster diodes because the reverse conduction is for a shorter time period
...
3
...
25 V is being used
...
25 V into the equation
...
For this design, R2 = 0
...
The nearest standard value is R2 = 0
...
Note that capacitor C3 is a bypass capacitor for holding up the HV9910B internal
supply VDD during MOSFET switching, when high frequency current pulses are
required for charging the gate
...
2 mF, 16 V is recommended,
although for AC applications smaller capacitors as low as 0
...
The switching frequency tends to be lower and so the MOSFET gate
current requirements are low
...
www
...
com
52
Chapter 5
5
...
These additional circuits are required because of the phase dimmer circuit
...
Because of
switching transients, which would otherwise cause serious EMI problems, the triac is
bypassed by a capacitor (typically 10 nF) and has an inductor in series with its output
...
5
...
1 mH
TRIAC
47 nF
Figure 5
...
The input of an inactive LED driver is high impedance, with a large capacitor on
the DC side of the bridge rectifier
...
When the voltage builds up, the LED driver will try to operate
...
What is required is a discharge circuit, to keep the smoothing capacitor voltage
below that required to start the LED driver
...
To prevent high power loss when
the circuit is active, a simple voltage detector can be used to disconnect the
390
resistor when a voltage above about 8 V is detected
...
6
...
newnespress
...
6: Smoothing Capacitor Discharge Circuit
...
Once a triac is triggered, it is the load current that
keeps it switched on; the triac is a self-sustaining switch
...
For this reason, an additional load must be switched across the LED
driver input at low voltages
...
A latching circuit to provide
this function is shown in Figure 5
...
These circuits can be combined
...
newnespress
...
7: Additional Load Switch
...
Thus when the triac is off, the LED driver is also off
...
8
...
5 Common Errors in AC Input Buck Circuits
The most common error is trying to drive a single LED from the AC mains supply
...
The worst case is the higher voltage; consider driving
a white LED with 3
...
The duty cycle will be 3
...
9333%
duty cycle
...
02 ms second period, the
MOSFET on-time will be just 186 ns
...
newnespress
...
8: Complete Phase Dimmable LED Driver
...
Operating at 20 kHz will give an on-time
of 466 ns, which is close to the limit for accurate control
...
Another error is not taking into account the parasitic capacitance of the inductor
windings and the reverse current in the flywheel diode
...
The current peak through the MOSFET can be high enough to trip
the current sense circuit, resulting in erratic switching
...
A 2
...
5
...
9
...
Diodes steer the current
in L2, which must operate in discontinuous conduction mode (DCM) for correct
operation
...
newnespress
...
1 mH
V+
L2
2 mH
C2
47 nF
C1
47 nF
D1
36 V
D3
BYD57J
1
C3
2,2uF 10 V
6
+
D2
BYD57J
C3
470 nF
L3
0
...
9: Double Buck
...
An example is driving a single power LED from an AC supply line
...
Assume the maximum duty cycle, Dmax, is less than 0
...
Boundary conduction
mode means that the current through the inductor only just falls to zero and the next
switching cycle begins
...
www
...
com
Buck-Based LED Drivers
57
The storage capacitor voltage at Vin min and Dmax is given by the equation:
Vc min ¼ Vin min à Dmin
The peak current through the input stage inductor, at Vin min equals:
IL2
pk
¼ 2 Ã IL2 avg
Vo à Io
¼2Ã
Vc min
Thus the primary stage inductor L2 has a value given by:
L2 ¼
ðVin min À Vc min Þ Ã Dmax à Ts
IL2 pk
The transfer ratio for a DCM buck converter (where R is load resistor seen by the
converter) is given by:
Vc
2
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
¼
Vin 1 þ 1 þ 8 Â L2
R  Ts  D2
The resistor R seen by the first stage (and assuming second stage is in CCM) is
given by:
R¼
Vc 2
Po
) R Â D2 ¼
ðVc  DÞ2 Vo 2
¼
Po
Po
Combining the previous two equations (which turn out to be a constant):
Vc
2
ffi
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
¼K¼
Vin
1 þ 1 þ 8 Â L2 Â Po
Ts  Vo 2
We find that D is inversely proportional to Vin:
D¼
Vo
Vo
¼
Vc KÂVin
www
...
com
58
Chapter 5
And we can now show that the peak inductor current through L2 is a constant
over the operating input voltage:
Setting D = K0 /Vin, K0 = Vo/K
K0 is a constant, since Vo is constant
...
newnespress
...
7 Hysteretic Buck
As an alternative to the peak current control buck, hysteretic control can be used
...
The input to the
comparator is a high side current sense circuit, where the voltage across a resistor
in the positive power feed to the LED load is monitored
...
10
...
10: Hysteretic Current Control Circuit
...
The MOSFET is turned off when the current is at or above a maximum
reference voltage
...
11
...
Vcs (high)
Vcs (Average)
Vcs
Vcs (Low)
Figure 5
...
www
...
com
60
Chapter 5
The current level is set by a suitable resistor value, given by:
RSENSE
À
Á
1 VCSðhighÞ þ VCSðlowÞ
¼ Á
2
ILED
In words, the average current sense voltage (midway between the high and low levels)
divided by the average LED current required
...
www
...
com
CHAPTER 6
Boost Converters
Boost converters (see Figure 6
...
Normally, a boost converter would
only be used when the output voltage minimum is about 1
...
•
The converter can easily be designed to operate at efficiencies greater than 90%
...
This
simplifies sensing of the LED current, unlike the buck converter where we
have to choose either a high side MOSFET driver or a high side current
sensor
...
L1
V+
D1
Cin
Cout
CONTROLLER
Q1
Figure 6
...
www
...
com
62
Chapter 6
Boost converters have some disadvantages, especially when used as LED drivers,
due to the low dynamic impedance of the LED string
...
Thus,
a large output capacitor is required to reduce the ripple in the LED current
...
Turning
the boost converter on and off to achieve PWM dimming means the capacitor
will have to be charged and discharged every PWM dimming cycle
...
• Open loop control of the boost converter to control the LED current (as in the
case of an HV9910-based buck control) is not possible
...
This also complicates PWM dimming, since the controller
will have to have a large bandwidth to achieve the required response times
...
There is a path from the input to the output via a diode and
inductor, so turning off the switching MOSFET will have no effect on the
short circuit current
...
If the surge current is
high enough, the LEDs will be damaged
...
1
Boost Converter Operating Modes
A boost converter can be operated in two modes – either continuous conduction mode
(CCM) or discontinuous conduction mode (DCM)
...
Figure 6
...
2(b) is the
inductor current waveform for a DCM boost converter
...
If larger boost ratios are required,
the DCM boost converter is used
...
Thus DCM boost converters are typically less efficient than CCM boost converters,
can create more EMI problems and are usually limited to lower power levels
...
newnespress
...
2: Inductor Current CCM and DCM
...
2
HV9912 Boost Controller
Supertex’s HV9912 integrated circuit is a closed-loop, peak current controlled, switch-mode
converter LED driver
...
In particular, it features a disconnect MOSFET driver output
...
This disconnect MOSFET is also used by the
HV9912 to dramatically improve the PWM dimming response of the converter (see PWM
Dimming section)
...
The most significant functions within the HV9912 are shown in Figure 6
...
The internal high voltage regulator in the HV9912 provides a regulated 7
...
This voltage range is good for
most boost applications, but the IC can also be used in buck and SEPIC circuits when
accurate current control is required
...
The VDD pin of the IC can be overdriven (if necessary) with an external voltage
source fed through a low voltage ( >10 V), low current diode
...
newnespress
...
75 V
Reg
1
...
3: HV9912 Internal Structure (Simplified)
...
The maximum steady state voltage that can be applied to the
HV9912 VDD pin is 12 V (with a transient voltage rating of 13
...
Allowing for the
diode forward voltage drop a 12 V ± 5% power supply would be ideal
...
25 V, 2% accurate reference voltage
...
This reference is also used internally to set the over-voltage set point
...
If the resistor is
connected between the RT and GND pins, the converter operates in a constant frequency
mode, whereas if it is connected between the RT and GATE pins, the converter operates
in a constant off-time mode (slope compensation is not necessary to stabilize the
www
...
com
Boost Converters
65
converter operating in a constant off-time)
...
8 ms to 40 ms using the equation given in Section 6
...
12
...
This is sometimes necessary in RGB
lighting systems, or when EMI filters are designed to remove a certain frequency
...
The HV9912 tries to keep
the feedback signal equal to the voltage on the IREF pin
...
When the feedback falls below the voltage at the IREF pin, switching is started again
...
What is not shown in Figure 6
...
When the PWM dimming
signal is low, this switch disconnects the output of the amplifier
...
When the PWM
dimming signal goes high again, the compensation network is reconnected to the
amplifier
...
The FAULT pin is used to drive an external disconnect MOSFET (see Figure 6
...
During the start-up of the HV9912, the FAULT pin is held low and once the IC starts
up the pin is pulled high
...
4: Disconnect MOSFET
...
newnespress
...
In case of an output over-voltage condition or an
output short circuit condition, the FAULT pin is pulled low and an external
MOSFET switched off to disconnect the LEDs
...
This disconnects the
LEDs and makes sure that the output capacitor does not have to be charged/
discharged every PWM dimming cycle
...
Output short circuit protection is provided by a comparator that triggers when the
output current sense voltage (at the FDBK pin) is twice that of the reference voltage
(at the IREF pin)
...
Both these fault signals are fed into the hiccup control
...
Once the IC goes into the fault mode, either by an output over-voltage
condition or a short circuit, the hiccup control is activated
...
At the same time, a timer is started to keep the output
turned off for a short period (determined by the capacitance on the COMP pin)
...
If the fault condition persists, the
output is turned off again and the timer is reset
...
Linear dimming is achieved by varying the voltage level at the IREF pin
...
This allows the current to be linearly dimmed
...
This output voltage limit restricts the linear dimming range to about 10:1
...
The PWM dimming signal controls
three nodes in the IC
...
newnespress
...
At the same time, the output of the
transconductance op-amp is connected to the compensation network
...
When PWMD goes low, the GATE of the switching MOSFET is disabled to stop
energy transfer from the input to the output
...
This discharge of the capacitor also means that when the circuit
restarts, the output capacitor has to charge again, causing an increase in the rise time
of the LED current
...
Thus, it is important to prevent the discharge of the output capacitor
...
This causes the LED current to
fall to zero almost instantaneously
...
This enables a
very fast rise time as well
...
This
means that the feedback amplifier sees a very large error signal across its input
terminals, which would cause the voltage across the compensation capacitor to
increase to the positive rail
...
The current will come back into
regulation depending on the speed of the controller
...
Thus, when PWMD goes high again, the circuit will
already be at the steady state condition, eliminating the large turn-on spike in the
LED current
...
3 Design of a Continuous Conduction Mode Boost
LED Driver
As a reminder, continuous conduction mode is valid when the output voltage is
between 1
...
www
...
com
68
Chapter 6
6
...
1 Design Specification
Input voltage range = 22–26 V
LED string voltage range = 40–70 V
LED current = 350 mA
LED current ripple = 10% (35 mA)
LED string dynamic impedance = 18 ohms
Desired efficiency >90%
6
...
2
Typical Circuit
A typical boost converter circuit is shown in Figure 6
...
L1
Vin +
C1
D1
Vout+
C2
C3
LED
R8
RT
VIN
OVP
RT
R9
HV9912
R-slope
SC
REF
GATE
PWM_D
R3
R5
SYNC
Q1
R7
CS
IREF
CLIM
COMP
R4
R6
VDD
GND
LED
R1
Q2
FAULT
FDBK
R2
Figure 6
...
www
...
com
Boost Converters
6
...
3
69
Selecting the Switching Frequency (fs)
For low voltage applications (output voltage < 100 V), and moderate power levels
(<30W), a switching frequency of fs = 200 kHz is a good compromise between
switching power loss and size of the components
...
6
...
4
Computing the Maximum Duty Cycle (Dmax)
The maximum duty cycle of operation can be computed as
min Á Vin min
Vo max
¼ 0:717
Dmax ¼ 1 À
Note: If Dmax = 0
...
The converter cannot operate in
continuous conduction mode and has to be operated in discontinuous conduction
mode to achieve the required step-up ratio
...
3
...
3
...
Vin min Á Dmax
0:25 Á Iin max Á fs
¼ 254 mH
L1 ¼
www
...
com
70
Chapter 6
Choose a standard 330 mH inductor
...
Using a 3% loss in the inductor
Pind ¼ 0:03 Á Vo max Á Io max
¼ 0:735W
Assuming an 80–20% split in the inductor losses between resistive and core losses, the
DC resistance of the chosen inductor has to be less than
0:8 Á Pind
Iin max 2
) DCR < 0:38
DCR <
The saturation current of the inductor has to be at least 20% higher than its
peak current; otherwise the core losses will be too great
...
38
and a saturation
current greater than 1
...
Note: Choosing an inductor with an RMS current rating equal to Iin max would also yield
acceptable results, although meeting the minimum efficiency requirement might not be
possible
...
3
...
Using a 20% overhead to account to switching spikes, the minimum
voltage rating of the MOSFET has to be
VFET ¼ 1:2 Á Vo max
¼ 84V
www
...
com
Boost Converters
71
The RMS current through the MOSFET is
IFET % Iin max Á
¼ 1:05A
pffiffiffiffiffiffiffiffiffiffiffi
Dmax
To get the best performance from the converter, the MOSFET chosen has to have a
current rating about three times the MOSFET RMS current with minimum gate
charge Qg
...
It is recommended
that for designs with the HV9912, the gate charge of the chosen MOSFET be less
than 25 nC
...
5 A MOSFET with a
Qg of 11 nC
...
3
...
The average current through the diode is equal to the maximum output
current (350 mA)
...
Thus, it is a
better design approach to choose the current rating of the diode somewhere in
between the maximum input current and the average output current (preferably
closer to the maximum input current)
...
6
...
9
Choosing the Output Capacitor (Co)
The value of the output capacitor Co (labeled C3 in Figure 5
...
In designs using the HV9912, a larger output capacitor (lower output
current ripple) will yield better PWM dimming results
...
www
...
com
72
Chapter 6
+
Icap
VJ
Co
I
Idiode
RLED
Figure 6
...
The output stage of the boost converter is modeled in Figure 6
...
The output impedance (parallel combination of RLED and Co) is driven by the diode
current
...
7;
the capacitor is charged during the off-time, as the energy stored in the inductor is
transferred to the capacitor
...
Icap
Time
Figure 6
...
Using the 10% peak-to-peak current ripple given in the design parameters table, the
maximum voltage ripple across the LED string has to be
DvpÀp ¼ DIo Á RLED
¼ 0:63V
www
...
com
Boost Converters
73
Assuming a constant discharging current of 350 mA when the switch is ON, the
equation for the voltage across the capacitor can be written as
Io max ¼ Co Á
DvpÀp
Dmax Á Ts
Substituting values into the above equation, we can calculate the value for Co
...
Note: The proper types of capacitors to use are either metal film capacitors or ceramic
capacitors, since they are capable of carrying this high ripple current
...
Also, high
value ceramic capacitors are normally only rated up to 50 V
...
6
...
10
Choosing the Disconnect MOSFET (Q2)
The disconnect MOSFET should have the same voltage rating as the switching
MOSFET Q1
...
Thus,
0:01 Á Vo max
Io max Á 1:4
¼ 1:43
Ron,25C ¼
www
...
com
74
Chapter 6
The 1
...
In this case, a MOSFET with
high gate charge, Qg, can be chosen if desired (as it is not switching regularly)
...
In this case, the
MOSFET chosen is a 100 V, 0
...
6
...
11 Choosing the Input Capacitors (C1 and C2)
The values of input capacitors C1 and C2 have to be calculated to meet closed loop
stability requirements
...
These feed
across the input capacitors (C1 and C2) and so form an LC resonant circuit
...
How do we determine the inductance Lsource? A pair of 22AWG connecting wires
1 foot (30 cm) long will have an inductance of about 1 mH
...
If necessary, the wires can be twisted together to reduce the inductance
...
CIN !
1
ð2 Á p Á fLC Þ2 Á LSOURCE
¼ 3:95 mF
C1 ¼ C2 ¼ 2:2 mF, 50 V ceramic:
The magnitude of the reflected converter impedance at the LC resonant frequency is
given by:
REQ ¼ ð1 À DMAX Þ2 Á RLED
REQ ¼ ð1 À 0:717Þ2 Á 18
REQ ¼ 1:4416
RSOURCE,MAX ¼ 1:44
6
...
12 Choosing the Timing Resistor (RT)
The HV9912 oscillator has an 18 pF capacitor charged by a current mirror circuit
...
When RT is connected to 0 V, current flows and the timing process begins
...
newnespress
...
The timing resistor value can be calculated by using
the equation:
1
% RT Á 18 pF
fs
In this case, for a constant 200 kHz switching frequency, the timing resistor value
works out to 274 k
...
6
...
13
Choosing the Two Current Sense Resistors (R1 and R2)
The value of output current sense resistor R2 is calculated to limit its power
dissipation to about 0
...
Using this
criterion,
0:15W
Io max 2
¼ 1:22
R2 ¼
In this case, the resistor chosen is a 1
...
The MOSFET current sense resistor R1 is calculated by limiting the voltage across
the resistor to about 250 mV at maximum input current
...
18
, 1/2 W, 1% resistor
...
newnespress
...
3
...
In the present design, it is assumed that the voltage at
the IREF pin is set using a voltage divider from the REF pin
...
3
...
The slope
added to the current sense signal has to be one-half the maximum down slope
of the inductor current to ensure stability of the peak current mode control
scheme for all operating conditions
...
For the present design, the down slope of the inductor current is
Vo max À Vin min
L
¼ 0:145 A=ms
DS ¼
www
...
com
Boost Converters
77
The programming resistors can then be calculated as
Rslope ¼
10 Á R7 Á fs
DSðA=sÞ Á 106 Á R1
Assuming R7 = 1 k
,
10 Á 1k Á 200k
0:2682 Á 106 Á 0:15
¼ 76:62 k
Rslope ¼
Note: The maximum current that can be sourced out of the SC pin is limited to
100 mA
...
If the equation
for slope compensation produces a value Rslope less than this value, then R7 would have
to be increased accordingly
...
Based on this recommendation, the calculated values can be scaled by 0
...
The selected
resistor values are
R7 ¼ 510, 1=8 W, 1%
Rslope ¼ 39k, 1=8W, 1%
6
...
16
Setting the Inductor Current Limit (R5 and R6)
The inductor current limit value depends on two factors – the maximum inductor
current and the slope compensation signal added to the sensed current
...
The voltage at the
CLIM pin can be computed as
VCLIM ! 1:35 Á Iin max Á R1 þ
4:5 Á R7
Rslope
This equation assumes that the current limit level is set at about 120% of the
maximum inductor current Iin max and that the operating duty cycle is at 90%
(maximum for the HV9912)
...
newnespress
...
36 V from a 1
...
Using a maximum
current sourced out of the REF pin of 50 mA the two resistors in series should
be >25 k
, and can be calculated as:
R5 ¼ 20k, 1=8 W, 1%
R6 ¼ 8:06k, 1=8 W, 1%
Note: It is recommended that no capacitor be connected at the CLIM pin
...
3
...
For the VDD pin, the capacitor used is a 1 mF ceramic chip capacitor
...
2 mF
...
1 mF ceramic chip capacitor
...
3
...
Using a 20% margin, the maximum output voltage during open
LED condition will be
Vopen ¼ 1:2 Á Vo max
¼ 84V
www
...
com
Boost Converters
79
Then, the resistors that set the over-voltage set point can be computed as
À
Vopen À 5
R8 ¼
0:1
¼ 64 k
Á2
The above equation will allow us to select a 1/8 W resistor by limiting the power
dissipation in the resistor
...
For this design, it varies from 80 V
to 88
...
6
...
19
Designing the Compensation Network
The compensation needed to stabilize the converter could be either a Type-I circuit
(a simple integrator) or a Type-II circuit (an integrator with an additional pole-zero
pair)
...
The loop gain of the closed loop system is given by
Loop Gain ¼ Rs Á Gm Á Zc ðsÞ Á
1 1
Á Gps ðsÞ
Á
15 Rcs
Where Gm is the transconductance of the op-amp (435 mA/V), Zc(s) is the impedance
of the compensation network, and Gps(s) is the transfer function of the power stage
...
www
...
com
80
Chapter 6
For the continuous conduction mode boost converter in peak current control mode
and for frequencies less than one tenth of the switching frequency, the power stage
transfer function is given by
Gps ðsÞ ¼
ð1 À Dmax Þ
Á
2
1ÀsÁ
L1
ð1 À Dmax Þ2 Á RLED
RLED Á Co
1þsÁ
2
For the present design, choose a crossover frequency 0
...
The
low crossover frequency will result in large values for Cc and Cz, which will indirectly
provide a soft-start for the circuit
...
0:283
Á
Gps ðsÞ ¼
2
1ÀsÁ
330 Á 10À6
ð0:283Þ2 Á18
18 Á 2 Á 10À6
1þsÁ
2
Gps ðsÞ ¼ 0:1415 Á
1 À s Á 2:28912 Á 10À4
1 þ s Á 1:8 Á 10À5
Substituting s = i Á (2p Á fc), where fc = 2 kHz, s = i Á 12566
...
newnespress
...
0
90
boost 0 ) Type À I controller
boost 90 ) Type À II controller
boost
180 ) Type À III controller
Type-III controllers are usually not required to compensate an HV9912-based boost
LED driver and thus will not be discussed here
...
1
...
1: Compensation Networks
...
For the present design, however, we
www
...
com
82
Chapter 6
need to use a Type-II controller
...
0
1
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 þ K2 C 1 1
1
B
Á qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiAÁ Á
Á Aps ¼ 1
Rs Á Gm Á @
2 Á p Á fc Á ðCz þ Cc Þ
15 Rcs
2
1 þ ð1=KÞ
C z þ Cc ¼ 10 nF
!z
Cc ¼ ðCz þ Cc Þ Á
!p
¼ 2:32 nF
Cz ¼ 7:68 nF
1
! z Á Cz
¼ 21:522 k
Rz ¼
Choose
Cc ¼ 2:2 nF, 50 V, C0G capacitor
Cz = 6
...
0k, 1/8 V, 1% resistor
www
...
com
Boost Converters
6
...
20
83
Output Clamping Circuit
One problem encountered with a continuous mode boost converter, when operating
with Vout < 2 Â Vin, is L-C resonance between the inductor and Cout
...
This diode
is shown as D2 in Figure 6
...
D2
L1
Vin +
C1
D1
C2
C4
1
R2
R11
Vout+
7
6
VIN
LED
R1
OVP
12
RT
HV9912
C11
R3
SC
10
3
REF
GATE
PWM_D
5
8
CS
SYNC
15
IREF
9
CLIM
14
COMP
11
C5 2
C7
VDD FAULT
16
4
FDBK
GND
Q2
13
C6
R12
R13
R8
R9
R5
LED
R6
Q1
R14
C10
R10
Figure 6
...
This completes the design of the HV9912-based boost converter operating in
continuous conduction mode
...
4 Design of a Discontinuous Conduction Mode
Boost LED Driver
As a reminder, discontinuous mode is used when the output voltage is more than
six times the input voltage
...
newnespress
...
4
...
8)
LED current = 100 mA
LED current ripple = 10% (10 mA)
LED dynamic impedance = 55 ohms
Efficiency > 85%
6
...
2
Typical Circuit
A typical circuit for a discontinuous mode boost converter, using the HV9912 IC
identical to the continuous mode circuit shown in Figure 6
...
9
...
4
...
At higher voltage or power levels,
the switching frequency might have to be reduced to lower the switching losses in the
external MOSFET
...
4
...
newnespress
...
9: Discontinuous Mode Boost Converter
...
4
...
10)
...
The rate of fall is controlled by (Vo À Vin)/L1 and the falling
period is the time that the diode is conducting
...
newnespress
...
Ts
Ton_diode
Ts
Figure 6
...
The average input current at the minimum input voltage is equal to the average
inductor current and can be computed from
1
4:75 ms
Á iLpk Á
2
5 ms
¼ 0:475 Á iLpk
Iin max ¼
Transposing the equation, the peak input current is
Iin max
0:475
% 1:93A
iLpk ¼
Substituting for iLpk in the equation for L1
0:95 9V Á ð70V À 9V Þ
Á
200k
70V Á 1:93A
¼ 19:3 mH
L1 ¼
Note that the value of L1 computed is the absolute maximum value for the
inductor
...
www
...
com
Boost Converters
87
The RMS current through the inductor is
ILrms
rffiffiffiffiffiffiffi
0:9
¼ iLpk Á
3
¼ 1:057A
Choose a 15 mH inductor (±20% tolerance)
...
However, if a standard
value inductor is preferred, the saturation current rating of the inductor should be
at least 1
...
The inductor chosen in this case is a 15 mH inductor with an RMS current rating of
1
...
6
...
6
Computing the On and Off Times of the Converter
The on-time of the switch can be computed as
ton
sw
L1nom Á iLpk
Vin min
¼ 3:22 ms
¼
The on-time of the diode is
ton
diode
L1nom Á iLpk
Vo max À Vin min
¼ 467 ns
¼
The maximum duty cycle can then be computed as
Dmax ¼ ton
sw
Á fs
¼ 0:644
The diode conduction time ratio can be expressed as
D1 ¼ ton
diode
Á fs
¼ 0:0934
www
...
com
88
6
...
7
Chapter 6
Choosing the Switching MOSFET (Q1)
The maximum voltage across the MOSFET in a boost converter is equal to the
output voltage
...
It is recommended that for designs with the HV9912, the gate charge of
the chosen MOSFET be less than 25 nC
...
5 A MOSFET with a Qg of
11 nC
...
4
...
The average current through the diode is equal to the maximum output current
(350 mA)
...
Thus, it is a better design approach to choose
the current rating of the diode somewhere in between the peak input current and the
average output current (preferably closer to the peak input current)
...
6
...
9
Choosing the Output Capacitor (Co)
The value of the output capacitor depends on the dynamic resistance of the LED
string as well as the ripple current desired in the LED string
...
newnespress
...
The capacitor required to filter the current
appropriately will be designed by considering the fundamental component of the
diode current only
...
11, where the LEDs
are modeled as a constant voltage load with series dynamic impedance
...
11: Model of Boost Converter Output
...
12
...
12: Output Capacitor Current
...
newnespress
...
Note: The proper type of capacitor to use is either metal film capacitors or ceramic
capacitors, since they are capable of carrying this high ripple current
...
High value ceramic capacitors are usually limited to
50 V rating
...
6
...
10 Choose the Disconnect MOSFET (Q2)
The disconnect MOSFET should have the same voltage rating as the switching
MOSFET Q1
...
Thus,
0:01 Á Vo max
Io max Á 1:4
¼ 5
Ron,25C ¼
The 1
...
In this case, a high Qg MOSFET
www
...
com
Boost Converters
91
can be chosen if desired (as it is not switching regularly), but a high Qg MOSFET will
slow down the turn-on and turn-off times (which might be allowable based on PWM
dimming frequency)
...
7
, SOT-23
MOSFET with a Qg of 2
...
6
...
11
Choosing the Input Capacitors (C1 and C2)
The values of input capacitors C1 and C2 have to be calculated to meet closed loop
stability requirements
...
These feed
across the input capacitors (C1 and C2) and so form an LC resonant circuit
...
A pair of 22AWG connecting wires 1 foot (30 cm) long will have an inductance of
about 1 mH
...
If necessary, the wires can be twisted
together to reduce the inductance
...
CIN !
1
ð2 Á p Á fLC Þ2 ÁLSOURCE
¼ 3:95 mF
C1 ¼ C2 ¼ 2:2 mF, 50 V ceramic:
The maximum source impedance is found using:
M¼
VO,MAX 70
¼
¼ 7:778
9
VIN,MIN
RSOURCE,MAX ¼
6
...
12
M2
MÀ1
Á RLED ¼ 1:404
Á ðM À 2Þ
Choosing the Timing Resistor (RT)
The HV9912 oscillator has an 18 pF capacitor charged by a current mirror circuit
...
www
...
com
92
Chapter 6
When RT is connected to 0 V, current flows and the timing process begins
...
The timing resistor can be calculated by using the
following equation:
1
% RT Á 18pF
fs
In this case, for a constant 200 kHz switching frequency, the timing resistor value
works out to 274 k
...
6
...
13 Choosing the Two Current Sense Resistors (R1 and R2)
The value of the output current sense resistor R2 can be calculated by limiting its
voltage drop to below 0
...
Using this criterion,
R2 ¼
0:4V
Io max
¼ 4
The power dissipation will be 0
...
04 W
...
9
, 1/8 W, 1% resistor
...
R1 ¼
0:25
iLpk
¼ 0:12
The power dissipated in this resistor is
PR1 ¼ IFET 2 Á R1
¼ 0:096 W
Thus, the chosen current sense resistor is a 0
...
www
...
com
Boost Converters
6
...
14
93
Selecting the Current Reference Resistors (R3 and R4)
The voltage at the current reference pin IREF can be set either by using the reference
voltage provided at the REF pin (through a voltage divider) or with an external
voltage source
...
The current reference resistors R3 and
R4 can be computed using the following two equations:
R3 þ R4 ¼
1:25V
50 mA
25 k
1:25V
Á R4 ¼ Io max Á R2 ¼ 0:1 Á 3:9 ¼ 0:39V
R3 þ R4
For this design, the values of the two resistors can be computed to be
R3 ¼ 19:1 k
, 1=8W, 1%
R4 ¼ 8:66 k
, 1=8W, 1%
6
...
15
Setting the Inductor Current Limit (R5 and R6)
The inductor current limit value depends on two factors – the maximum inductor
current and the slope compensation signal added to the sensed current
...
The voltage at the CLIM pin can be
computed as
VCLIM ! 1:2 Á iIpk Á R1
This equation assumes that the current limit level is set at about 120% of the
maximum inductor current Iin max
...
newnespress
...
6
...
16 Capacitors at VDD and REF Pins
It is recommended that bypass capacitors be connected to both VDD and REF pins
...
For low
power designs, a 1 mF is adequate
...
2 mF
...
1 mF ceramic chip capacitor
...
4
...
Using a 15% margin, the maximum output voltage during open
LED condition will be
Vopen ¼ 1:15 Á Vo max
¼ 80:5V
Then, the resistors that set the over-voltage set point can be computed as
À
Á2
Vopen À 5
R8 ¼
0:1
¼ 57 k
The above equation will allow us to select a 1/8 W resistor by limiting the power
dissipation in the resistor
...
newnespress
...
For this design, it varies from 76
...
52 V
...
4
...
The type of the compensation circuit required will be dependent on the phase of
the power stage at the crossover frequency
...
Please note that although the resistors give a 1:14 ratio, the overall effect when
including the diode drop is effectively 1:15
...
M¼
Vo max Á Io max
Vo max Á Io max À 0:5 Á L1nom Á iLpk2 Á fs
M¼
70 Á 0:1
70 Á 0:1 À 0:5 Á 15 Á 10À6 Á 1:932 Á 200 Á 103
M¼
GR ¼
7
¼ 4:9552
1:41265
MÀ1
3:95522
¼
¼ 0:4439
2ÁMÀ1
8:9104
www
...
com
96
Chapter 6
For frequencies less than one tenth of the switching frequency, the power stage
transfer function is given by
Gps ðsÞ ¼ 2 Á
Gps ðsÞ ¼ 2 Á
Io max
GR
Á
iLpk 1 þ s Á RLED Á Co Á GR
0:1
0:4439
0:4439
Á
¼
À6 Á 0:4439
1:93 1 þ s Á 55 Á 2 Á 10
1 þ s Á 48:829 Á 10À6
For the present design, choose a crossover frequency $ 0
...
The low
crossover frequency will result in large values for CC and CZ, which will indirectly
provide a soft start for the circuit
...
By
substituting s = i Á (2p Á fc) = i Á 12566 into the transfer function, we get:
Gps ðsÞ ¼
0:046
1 þ s Á 0:6136
The magnitude and frequency of the power stage transfer function are:
Gps ðsÞ
¼ Aps ¼ 0:039
fc¼2kHz
ffGps ðsÞfc¼2kHz ¼ ps ¼ À31:5
To get a phase margin of about m = 45° (the recommended phase margin range is
45°–60°), the phase boost required will be
boost ¼ m À ps À 90
¼ 45 þ 31:5 À 90
¼ À13:5
Based on the value of the phase boost required, the type of compensation can be
determined
...
newnespress
...
The implementations for the
Type-I and Type-II systems for use with the HV9912 are given in Table 6
...
Table 6
...
Type
Circuit diagram
Transfer function
COMP
Zc ðsÞ ¼
Cc
1
sCc
Zc ðsÞ ¼
I
1
Á
sðCc þ Cz Þ
COMP
Cz
II
Cc
Rz
1 þ s Á R z Á Cz
Cz Á Cc
1þsÁ
Á Rz
Cz þ Cc
For the present design, a simple Type-I controller will suffice
...
One more equation can be obtained by equating the magnitude of the loop gain to
1 at the crossover frequency
...
3 nF, 50 V, C0G capacitor
This completes our DCM boost converter design
...
newnespress
...
5
Common Mistakes
1
...
If the LEDs are disconnected while the circuit is operating,
the output voltage will rise until components start to break down
...
2
...
The forward voltage drop
may be lower than the supply voltage, and in this case there is little to prevent
the LEDs being destroyed by the high current that will flow
...
6
Conclusions
Boost converters are used when the minimum output voltage is at least 1
...
Continuous conduction mode should be used when the output voltage
is a maximum of six times the input voltage
...
The EMI
produced by a discontinuous mode boost converter is higher than for a continuous
conduction mode boost converter of similar power output
...
newnespress
...
The power train of typical boost-buck
circuit topology (used as an LED driver) is shown in Figure 7
...
C1
L1
IL1
VIN
+
–
L2
IL2
+ Vc –
Q1
VO–
D1
VO+
Figure 7
...
The converter has many advantages:
•
The converter can both boost and buck the input voltage
...
This condition is most common in automotive
applications, or when a customer wants a single driver design to cover a wide
range of voltage supply and load conditions
...
newnespress
...
Operating both
stages in continuous conduction mode (CCM) will enable continuous currents
in both inductors with low current ripple, which would greatly reduce the filter
capacitor requirements at both input and output
...
• All the switching nodes in the circuit are isolated between the two inductors
...
This will minimize the
radiated EMI from the converter
...
• One of the advantages of the boost-buck converter is the capacitive isolation
...
Thus, the LEDs are protected from failure of the MOSFET
...
When
coupled on a single core, the ripple in the inductor current from one side can be
transferred completely to another side (ripple cancellation technique)
...
7
...
• The converter is difficult to stabilize
...
This compensation also tends
to slow down the response of the converter, which inhibits the PWM dimming
capability of the converter (essential for LEDs)
...
The
resonance of L1 and C1 leads to excessive voltages across the capacitor, which
can damage the circuit
...
However, the problem of compensating the circuit so that it is stable is more complex
...
newnespress
...
This uses fast
comparators to control a MOSFET gate by setting upper and lower limits, which
ensures fast response and accurate current levels
...
To overcome this problem, the HV9930 has two hysteretic current mode controllers –
one for the input current and another for the output current
...
The MOSFET turns on and the input current rises until
the input current limit is reached, it then turns off so that the input current drops
until a lower current limit is reached
...
The output current is then maintained between the set upper and lower current limits
...
The hysteretic approach will also help in limiting the input current during start-up
(thus providing soft-start); also current is limited in the case of an output overload or
input under-voltage condition
...
Thus six resistors determine the input and output
performance
...
The design example is specifically designed for
automotive applications, but it can also be applied for any DC/DC applications
...
The AT9933 has an automotive temperature
specification (up to 125°C operation), whereas the HV9930 has an industrial
temperature range
...
1
...
1
...
newnespress
...
Thus, the
voltage, Vc, across the middle capacitor C1 is equal to the sum of the input and
output voltages
...
2)
...
2: Cuk Circuit, MOSFET On
...
3)
...
3: Cuk Circuit, MOSFET Off
...
newnespress
...
5 and boost operation for D > 0
...
The steady state
waveforms for the converter are shown in Figure 7
...
The maximum voltage seen by Q1 and D1 is equal to the voltage across the
capacitor C1
...
5)
...
These additional
components stabilize the circuit
...
This diode is necessary for PWM dimming
operation (in case of automobile applications, this could be the reverse polarity
protection diode)
...
Thus, when the HV9930 is enabled, the
steady state output current level will be reached quickly
...
1
...
As previously shown in Figure 7
...
newnespress
...
4: Cuk Converter Steady State Waveforms
...
Thus, the hysteretic control scheme turns the switch OFF when the
inductor current reaches the upper limit and turns the switch ON when it reaches
the lower limit
...
The ON and OFF times (and thus the switching frequency) vary as the
input and output voltages change to maintain the inductor current levels
...
newnespress
...
5: Modified Boost-Buck Circuit
...
The switch will not turn ON and OFF at the instant the inductor current
hits the limits, but after a small delay time, as illustrated in Figure 7
...
Turn-on Delay
Desired Io
Actual Io
Turn-off Delay
Figure 7
...
7
...
3
The Effects of Delay in Hysteretic Control
This delay time introduces two unwanted effects:
•
It alters the average output current value
...
•
It decreases the switching frequency, which may make it more difficult for the
circuit to meet EMI regulations
...
newnespress
...
Assume a peak-to-peak current ripple setting of Dio (using the programming
resistors) and a desired average current lo
...
Thus, the converter
is designed assuming a constant off-time Toff (the method to determine the off-time
will be discussed later)
...
In these cases, the output
inductor can be determined by
L2 ¼
Vo Á Toff
Dio
If the inductor chosen is significantly different from the computed value, the actual
off-time Toff,ac can be recomputed using the same equation
...
This will place the fundamental frequency of the conducted and radiated
EMI outside of the restricted bands making it easier for the converter to pass
automotive EMI regulations
...
Figure 7
...
From this figure,
Toff ¼ Tf 1 þ Tf 2 þ Tf
Vin
¼
Á Tr þ Tf 2 þ Tf
Vo
The desired output current ripple Dio and the down-slope of the inductor current
m2 determine Tf 2
...
For the
HV9930, the delay time of the comparators is related to the overdrive voltage
www
...
com
Boost-Buck Converter
107
Tr
Tf1
ΔIo, ac
ΔIo
m1 = Vin
L2
–m2 = –Vo
L2
Tf
Tf 2
Toff
Figure 7
...
(voltage difference between the two input terminals of the current sense comparator)
applied as
K
Tdelay % qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
3
m Á 0:1=Dio
Where ‘m’ is the rising or falling slope of the inductor current
...
This cubic has one real root and two complex roots
...
www
...
com
108
Chapter 7
a ¼ K2
Vin
b¼
Á K1 þ K3
Vo
c ¼ Toff
pffiffiffi
D ¼ 12 Á 3 Á
L2 ¼
8
>
>
>
<
ffi
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
4 Á b3 þ 27 Á a Á c2
a
"
1
ð108 Á c þ DÞ Á a2
>6 Á a
>
>
:
#1=
3
À
93
>
>
>
=
2Áb
h
ð108 Á c þ DÞ Á
a2
i1= >
3>
>
;
The actual off-time Toff,ac can be computed by substituting the chosen inductor value
back into the equations for Tr, Tf and Tf 2, to get Tr,ac, Tf,ac and Tf 2,ac
...
1
...
8)
...
newnespress
...
8: Boost-Buck Converter
...
7 rather than as the single-switch Cuk
converter
...
www
...
com
110
Chapter 7
In hysteretic control of the boost-buck converter using the HV9930, the
output buck stage is controlled and the input boost stage is uncontrolled
...
9
...
9: Boost-Buck Controller
...
So, for the purposes of average
modeling, the load seen by the capacitor C1 can be modeled as a current source
equal to d Á lo, where d is the instantaneous duty cycle and lo is the constant output
current
...
For the system to be stable, it is necessary that the control
system will act to reduce any disturbance in capacitor voltage
...
e
...
This is due to the undamped LC pole-pair and causes the system to be unstable
...
newnespress
...
When testing the circuit, if it is close to becoming unstable,
the switching frequency rises and falls with a low frequency beat and a low frequency
ripple in the average output current can be seen
...
Also, the presence of Cd
ensures that Rd will not see the DC component of the voltage Vc across it, reducing
the power dissipated in the damping resistor (Cd blocks the DC component of the
voltage)
...
Damping (and ESR) zero at !Z ¼ Rd 1 Cd :
Á
2
Vo
2
...
D
1ÀD
3
...
L1
4
...
The latter condition is easily met by selecting Cd >> C1
...
This
condition is usually met for the worst-case calculations at minimum input voltage,
since the DC gain is the highest at this condition
...
Then !o can be approximately calculated from
rffiffiffiffiffiffiffiffiffiffiffiffi
rffiffiffiffiffiffiffiffiffiffiffiffi
1 À D !RHP
1ÀD
!O ¼ !C Á
¼
Á
D
D
N
www
...
com
112
Chapter 7
Substituting for !o and !RHP in (21) gives the equation for computing Cd:
Cd ¼
N2 Á D3 L1 Á IO 2
ð1 À DÞ3 VO 2
Selecting Rd such that !z = !c results in a good phase margin with minimum power
dissipation
...
Rd ¼
NÁD
L1 Á IO
ð1 À DÞ Cd Á VO
2
Using the equations above, the approximate values for the damping network can be
computed using the following equations:
D
Cd ¼ 9 Á
1ÀD
Rd ¼
3ÁD
2
ð 1 À DÞ
Á
3
Io
Á L1 Á
Vo
2
L1 Á Io
Cd Á Vo
Note that the damping resistor value includes the ESR of the damping capacitor
...
In such cases, the damping
resistor can be reduced accordingly
...
1
...
For a converter designed to operate at a minimum switching frequency of 300 kHz,
one switching time period equals 3
...
This is the minimum on-time of the PWM
dimming cycle
...
33 ms
equals a minimum duty cycle of 0
...
This corresponds to a 1:1500 dimming
range
...
33% or a PWM dimming range
of 1:300
...
newnespress
...
Hence the average current will
be less
...
7
...
6
Design of the Boost-Buck Converter with HV9930
Specification
Input voltage: 9–16 V (13
...
6 ohms
Estimated efficiencies: 72% minimum, 82% maximum (80% typical)
These efficiency values do not take into account the power loss in the reverse blocking
diode
...
5 V across it and thus will dissipate
power in the range 0
...
6 W
...
The efficiency values used in this design are typical values for the given input
voltages and output power level
...
e
...
The efficiency values will depend on the operating conditions and,
except in very high power designs, these values can be used as a good
approximation
...
However, because
of automotive EMI requirements, the higher efficiencies are traded off for higher
switching frequencies (which increase switching losses in the system)
...
newnespress
...
10
...
10: Boost-Buck Converter Using HV9930
...
In the case of automotive converters, designing
with a switching frequency in the range between 300 kHz and 530 kHz would avoid
the restricted radio broadcast bands and make it easier to meet the conducted and
radiated EMI specifications
...
Calculating the Duty Cycle
The switch duty cycle will have to be computed at the minimum input voltage
...
newnespress
...
The value obtained will be used to work
out the current ratings of the various components
...
The off-time of the converter can be
calculated as
Toff ¼
1 À Dmax
fs,min
¼ 598 ns
Assuming a 25% peak-to-peak ripple in the output current ðDio ¼ 87:5 mAÞ, and
accounting for the diode drop in the input voltage by substituting Vin,min À Vd in
place of Vin, yields
598 ns ¼ 0:887 m Á
pffiffiffiffiffiffi
pffiffiffiffiffiffi
3
L2 þ 3:125 m Á L2 þ 1:89 m Á 3 L2
Solving for L2 gives
L2 ¼ ð0:052Þ3 ¼ 145 mH
The closest standard value is a 150 mH, 0
...
4 A saturation inductor
...
newnespress
...
5 mA), the actual ripple is almost double that value
...
A capacitor will be required at the output of the
converter (across the LEDs) to reduce the ripple to the desired level
...
Large output capacitors are to be avoided in applications
that use PWM dimming, because the stored charge will reduce the dimming ratio that
can be obtained
...
This will help determine how the average current changes
due to the delays
...
7 mA
...
Thus, it is better to use the actual value of the
off-time calculated in order to work out the rest of the component values
...
www
...
com
Boost-Buck Converter
117
Calculating the Input Inductor
We can assume a 15% peak-to-peak ripple in the input current at minimum input
voltage (this low input ripple will minimize the input filtering capacitance needed)
...
L1 ¼
Vo Á Toff,ac
0:15 Á Iin,max
¼ 72 mH
The closest standard value inductor is an 82 mH inductor
...
The peak-to-peak ripple in the input current is
DIin ¼
Vo Á Toff,ac
L1,ac
¼ 0:21A
Calculating the Value of the Middle Capacitor (C1)
Assuming a 10% ripple across the capacitor at minimum input voltage
(Dvc ¼ 0:1 Á ðVin,min À Vd þ Vo Þ ¼ 3:65 V ), capacitor C1 can be calculated as
Iin,max Á Toff,ac
Dvc
¼ 0:257 mF
C1 ¼
Irms,C1 ¼
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
I2
in,max Á ð1 À Dmax Þ þ I o Á Dmax
¼ 0:72 A
The voltage rating and type of this capacitor have to be chosen carefully
...
Thus, to prevent
excessive losses and overheating of the capacitor, it must have a very low ESR
...
If a ceramic capacitor cannot be used for reasons of
www
...
com
118
Chapter 7
cost or availability, a plastic film capacitor such as PET can be used instead, although
these are considerably bulkier
...
Ceramic
capacitors can easily withstand up to 2
...
Also, the actual capacitance value of these capacitors reduces
based on the bias voltage applied
...
Thus, a 0
...
Choosing the Switching Transistor (Q1)
The peak voltage across the MOSFET Q1 is 70 V
...
The maximum RMS current
through the MOSFET is
IFET,max ¼ ðIin,max þ Io Þ Á
pffiffiffiffiffiffiffiffiffiffiffi
Dmax
¼ 1:77A
A typical choice for the MOSFET is to pick one whose current rating is about three
times the maximum RMS current
...
5 A, 50 m
N-channel MOSFET)
...
The total gate charge Qg of the chosen MOSFET is a maximum of 15 nC
...
A higher gate charge
would be allowable if the switching frequency can be reduced appropriately
...
newnespress
...
Current transients are limited when the MOSFET turns on slowly, but this
reduces efficiency
...
Choosing the Switching Diode
The maximum voltage rating of the diode D2 is the same as the MOSFET voltage
rating
...
Idiode ¼ Io ¼ 350 mA
Although the average current of the diode is only 350 mA, the actual switching
current through the diode goes as high as 1
...
(Note: the calculations
were for 360 mA, to allow for 10 mA drop because of delays, but the actual average
current is 350 mA
...
79 A current safely, but
the voltage drop at such high current levels would be extremely large, increasing the
power dissipation
...
A 100 V, 2 A Schottky diode would be a good choice
...
Choosing the Input Diode
The input diode serves two purposes:
1
...
2
...
The current rating of the device should be at least equal to Iin,max
...
A higher current
rating often gives a lower forward voltage drop
...
www
...
com
120
Chapter 7
If neither reverse protection or PWM dimming is required, removing the input diode
from the LED driver circuit will increase the input supply voltage at the converter, which
will slightly increase the efficiency and slightly reduce the maximum input current
...
This
capacitance is mainly responsible for reducing the 2nd harmonic of the input current
ripple (which in this case falls in the AM radio band)
...
5 V
...
7 mF, 25 V, X7R ceramic capacitors
...
newnespress
...
A 70 mA peak-to-peak ripple in the LED
results in a 392 mV (Dvo = DlLED Á RLED) peak-to-peak ripple voltage
...
22 mF, 35 V ceramic capacitor
...
www
...
com
122
Chapter 7
Design of the Damping Circuit
The values for the damping network can be calculated as follows
Dmax
Cd ¼ 9 Á
1 À Dmax
3
Io
Á L1,ac Á
Vo
2
¼ 11 mF
Rd ¼
3 Á Dmax
2
ð1 À Dmax Þ
Á
L1,ac Á Io
Cd Á Vo
¼ 7:16
The power dissipated in Rd can be computed as
Dvc 2
12 Á Rd
3:652
¼
¼ 0:155 W
12 Á 7:16
PRd ¼
The RMS current through the damping capacitor will be
iCd ¼
Dvc
pffiffiffi
¼ 0:147 A
2 Á 3 Á Rd
Choose a 10 mF, 50 V electrolytic capacitor that can allow at least 150 mA RMS
current
...
This capacitor has about a 1
ESR, so Rd can be reduced to about 6
...
Internal Voltage Regulator of the HV9930
The HV9930 includes a built-in 8–200 V linear regulator
...
This regulator can be connected at either one of two nodes on the
circuit as shown in Figure 7
...
In the normal case, when the input voltage is always greater than 8 V, the VIN pin of
the IC can be connected to the cathode of the input protection diode (as shown in
www
...
com
Boost-Buck Converter
(A)
D2
VIN+
(B)
L1
D2
VIN+
L1
Q1
Q1
3
1
Vin
CS1
PWM
GATE
6
GND
VDD
HV9930
REF
123
CS2
C2
C2
1
Vin
CS1
PWM
3
GATE
6
GND
VDD
HV9930
REF
CS2
Figure 7
...
Figure 7
...
If reverse protection is not provided, the VIN pin can be connected
directly to the positive supply
...
11B
...
However, in this case, more hold-up capacitance
will be required at the VDD pin to supply the power to the IC when the MOSFET
is ON
...
2 mF or greater value ceramic capacitor is recommended at the
VDD pin
...
25 V (+3%) reference
...
It is
recommended that this pin be bypassed with at least a 0
...
www
...
com
124
Chapter 7
Programming the Hysteretic Controllers and Over Voltage Protection
The input and output current levels for the hysteretic controllers are set by means of
three resistors for each current – one current sense resistor and two divider resistors
...
25 V reference provided by the HV9930 is used to
set the current
...
25 V reference and
the output current thresholds are modified using the variable input voltage available
...
25 V reference is used for both the input and output programming
...
The minimum external voltage is given by
VLD ¼ 0:1 Á
www
...
com
Rref 2 þ Rs2
:
Rs2
Boost-Buck Converter
125
The programming of the output side is also linked to the over-voltage protection
...
This is achieved by adding Zener diode D3, and by
splitting the resistor Rs2 into two parts – Rs2a and Rs2b
...
When there is an open LED condition, the inductor current will flow through
diode D3
...
However,
since the diode cannot take the full design current, the current level has to be reduced
to more manageable levels
...
Thus, the effective current sense resistor seen by the IC is
Rcs2 þ Rs2a and the voltage drop across both of these will be sensed through Rs2b
...
Choosing the Output Side Resistors
For the output current, Io = 0
...
5 mA
...
Using these values in the above
equations,
Rs2a þ Rs2b
¼ 0:534
Rref 2
Rcs2 ¼ 1:64
PRcs2 ¼ 0:352 Á 1:64 ¼ 0:2 W
Before we complete the design of the output side, we also have to design the overvoltage protection
...
This is the voltage
at which the output will clamp in case of an open LED condition
...
Using a 5 mA current
level during open LED conditions, and assuming the same Rs/Rref ratio,
Rs2a þ Rcs2 ¼ 120
www
...
com
126
Chapter 7
Choose the following values for the resistors:
Rcs2 ¼ 1:65
, 1=4 W, 1%
Rref 2 ¼ 10 k
, 1=8 W, 1%
Rs2a ¼ 100
, 1=8 W, 1%
Rs2b ¼ 5:23 k
, 1=8 W, 1%
Design of the Input Side Resistors
For the input side, we first have to determine the input current level for limiting
...
The peak of the input current at minimum input voltage will be
Iin,pk ¼ Iin,max þ
DIin
2
¼ 1:706 A
Assuming a 30% peak-to-peak ripple when the converter is in input current limit
mode, the minimum value of the input current will be
Ilim,min ¼ 0:85 Á Iin,lim
We need to ensure that Ilim,min > Iin,pk for proper operation of the circuit
...
e
...
1 A
...
newnespress
...
At a nominal input voltage of 13
...
Iin,nom ¼
28 Á 0:35
0:8 Á ð13:5 À 0:5Þ
¼ 0:942 A
PRcs1 ¼ 0:9422 Á 0:228 ¼ 0:2 W
Thus, at nominal input voltage, the power dissipation reduces by about five times to a
reasonable 0
...
Choose the following values for the resistors:
Rcs1 ¼ parallel combination of three 0:68
, 1=2 W, 5% resistors
Rref 1 ¼ 10 k
, 1=8 W, 1%
Rs1 ¼ 4:42 k
, 1=8 W, 1%
Input Inductor Current Rating
The maximum current through the input inductor is Ilim,max ¼ 1:15 Á Iin,lim ¼ 2:4 A:
Thus, the saturation current rating of the inductor has to be at least 2
...
If the converter is going to be in input current limit for extended periods of time,
the RMS current rating needs to be 2 A, else a 1
...
Improving Efficiency
The input current sense resistor can be reduced in value, which gives reduced
power dissipation (loss)
...
This resistor allows a reduction in the hysteresis required by the input
comparator
...
12
...
12, RS1 = R4, RREF1 = R7, and RCS1 = the parallel combination R1//R3
...
newnespress
...
12: Modification of the Cuk Circuit
...
With the
MOSFET turned ON, the positive side of the capacitor C1 is grounded and the other
side of C1, which is connected to resistor RA, is at potential ÀVC1
...
The voltage reference for the comparator input
at CS1 is 0 V
...
Now the flywheel diode D3 is
conducting, so the negative side of capacitor C1 is grounded (the small forward
voltage of the diode can be ignored)
...
www
...
com
Boost-Buck Converter
129
DIIN
0:1V þ IIN,LIM À
Á R1==R3
VREF À 0:1V 0:1V
2
¼
þ
R7
RA
R4
Since R4 is a very large value and the voltage across it is small, we can ignore its effect
to simplify the calculations:
DIIN
0:1V þ IIN,LIM À
Á R1==R3
VREF À 0:1V
2
¼
R7
R4
We can thus ignore the addition of RA during the period that the MOSFET is turned
À
Á
OFF
...
The maximum current sense voltage occurs when the MOSFET is first turned ON
...
If we take another look at
the equation for current flow when the MOSFET is turned ON:
IIN,LIM þ
VREF VC1,NOM
¼
þ
R7
RA
DIIN
Á R1==R3
2
R4
In a Cuk topology, VC1 = VIN þ VOUT
...
The highest input current occurs at VIN,MIN
...
newnespress
...
In the modified circuit,
ISAT (of L1) must be much higher than these values in order to gain the loss reduction
benefit, which gives a higher input ripple at start-up
...
25 V in the standard configuration
...
A few precautions need to be taken during design and
PCB layout to be able to meet the EMI standards
...
In some cases, when the input current ripple is too large or the switching
frequency of the converter is above 150 kHz, it might not be possible to meet
the conducted EMI standards using only capacitors at the input
...
2
...
These inductors will minimize radiated magnetic fields
...
During layout, the IC and MOSFET ground connection should be connected
to a copper plane on one of the PCB layers with the copper plane extending
under the inductors
...
The loop consisting of Q1, C1 and D1 should be as small as possible
...
www
...
com
Boost-Buck Converter
131
5
...
A low
value resistor (10–47 ohms) in series with GATE connection will slow down
the switching edges and greatly reduce EMI, although this will cause
efficiency to decrease slightly
...
6
...
This concludes the Cuk converter design
...
7
...
A SEPIC is a boost-buck converter, like a Cuk, so its input
voltage range can overlap the output voltage
...
The SEPIC topology has been known for some time, but only recently has there
been a revival in its application because: (a) it needs low ESR capacitors and these are
now widely available and (b) it can be used to create AC input power supplies with
power-factor correction that are used to meet worldwide EMI standards
...
A 12 V supply used in automotive applications can have a wide
range of terminal voltage, typically 9 V to 16 V during normal operation using a leadacid battery, but can go as low as 6
...
The peak voltage is usually clamped to
about 40 V, using a voltage dependent resistor to absorb the energy
...
A single lithium cell provides an open voltage of 4
...
During discharge the cell still retains some energy down to 2
...
This input voltage
range can be both above and below the output of many DC/DC converters and so
discounts the possibility of using boost or buck converters
...
newnespress
...
Having a good power factor means that the current waveform from
the AC line is sinusoidal and in phase with the voltage
...
In Europe AC inputs of 190–265 V RMS are
found, which impose an output of at least 375 V, forcing the following converters to
work with elevated input voltages
...
By using a SEPIC topology, which has a boost-buck topology, the boost section
provides PFC and the buck section produces a lower output voltage
...
It provides the required output level even if the peak
input voltage is higher
...
2
...
13, is the basis for the
SEPIC converter
...
Second, the switch stops conducting
during the off-period, TOFF, but the current through L1 cannot change abruptly – it
continues to flow, but now through diode D1 and into Cout
...
Capacitor Cout filters the
current pulse that was generated by L1 when Q1 turned off
...
13: This Boost-Converter Topology is the Basis for SEPIC
Power-Supply Circuits
...
In cases where Vout is relatively low, the
www
...
com
Boost-Buck Converter
133
efficiency can be improved by using a Schottky diode with low forward voltage
(about 400 mV) for D1
...
If Vin is ever allowed to become greater than Vout, D1 will be forward biased
and nothing can prevent current flow from Vin to Vout
...
14 removes this limitation by inserting a capacitor
(Cp) between L1 and D1
...
The anode of D1, however, must connect to a known potential
...
L2 can be
separate from L1 or wound on the same core, depending on the needs of the application
...
14: SEPIC Topology
...
However, the
transformer leakage inductance, which is no problem in SEPIC schemes, often
requires a snubber network in fly-back schemes
...
Parasitic resistances that cause most of the conduction losses in a SEPIC are RL1,
RL2, RSW and RCP, and are associated with L1, L2, SW, and CP respectively
...
27
...
Thus the
limitation of the simple boost converter, that VIN had to always be less than VOUT,
has been overcome
...
newnespress
...
First, assume
that the values of current and voltage ripple are small with respect to the DC
components
...
By using these assumptions, Cp sees a DC potential of Vin at one
side (through L1) and ground on the other side (through L2)
...
The portion of T for which
switch Q1 is closed is the duty cycle, D, and the remaining part of the period
is thus 1 À D
...
e
...
e
...
Simplifying this we get:
D Á T Á VIN ¼ ð1 À DÞ Á T Á ðVOUT þ VD Þ
Transposing this, we get:
ðVOUT þ VD Þ
D
¼
¼ Ai
VIN
1ÀD
Ai is called the amplification factor, where ‘i’ represents the ideal case for which
parasitic resistances are null
...
5)
...
newnespress
...
The formula is recursive
(‘Aaxxx’ appears in both the result and the expression), but a few iterative
calculations lead to the solution
...
Those losses are usually negligible,
especially if Q1 is a fast MOSFET and its drain-voltage swing (Vin þ Vout þ Vd)
remains under 30 V
...
You can extrapolate the corresponding values of D:
D ¼ Aa=ð1 þ AaÞ
Or more generally:
Dxxx ¼ Aaxxx =ð1 þ Aaxxx Þ, where xxx is min, typ or max:
The DC current through Cp is zero, so the mean output current can only be
supplied by L2:
IOUT ¼ IL2
The power-dissipation requirement for L2 is eased, because the mean current into
L2 always equals IOUT and does not depend on variations of VIN
...
Thus, the coulomb charge flowing during D * T is perfectly
balanced by an opposite coulomb charge during (1 À D) * T
...
Since
the capacitor Cp was previously charged to voltage Vin, the anode of D1 will now
www
...
com
136
Chapter 7
have a potential of ÀVIN, which reverse-biases D1
...
When the switch is open during (1 À D) * T, current IL2 flows through D1
while IL1 flows through Cp:
D Á T Á IL2 ¼ ð1 À DÞ Á T Á IL1
Knowing that IL2 = IOUT,
IL1 ¼ Aa xxx Á IOUT
Input power equals output power divided by efficiency, so IL1 depends strongly on
VIN
...
Knowing that IL2
(and hence IOUT) flows into Cp during D*T, we choose Cp so that its ripple delta
Vcp is a very small fraction of Vcp (gamma = 1% to 5%)
...
Cp i
IOUT Á Dmin Á T
gamma Á VIN MIN
By using a high switching frequency, small multi-layer ceramic capacitors can be
used for Cp
...
This incurs the
following loss:
Psw ¼ Aa min à ð1 þ Aa minÞ Ã Rsw à IOUT 2
Losses PRL1 and PRL2, due to the internal resistances of L1 and L2, are easily
calculated:
PRL1 ¼ Aa min2 Ã RL1 Ã IOUT 2
PRL2 ¼ RL2 Ã IOUT 2
When calculating the loss due to D1, the average power loss is due to the output
current and the forward voltage drop of D1:
PD1 ¼ VD Ã IOUT
www
...
com
Boost-Buck Converter
137
L1 is chosen so its total current ripple (DIL1) is a fraction (b = 20% to 50%) of IL1
...
Assuming b = 0
...
Using a single core, the two windings should be bifilar
(twisted around each other before being wound on the core) and thus will have the
same number of turns and the same inductance values
...
If the
winding voltages are identical, they generate equal and additive current gradients
...
Thus, the inductance measured across each isolated winding (when there is nothing
connected to the other winding) should equal only half of the value calculated for
L1 and L2
...
If the windings’ cross-sections
are equivalent, the resistive losses will differ because their currents (IL1 and IL2) differ
...
This is particularly easy to do when the windings consist of insulated strands
www
...
com
138
Chapter 7
of wire (Litz) for counteracting the skin effect
...
The purpose of the output capacitor (COUT) is to average the current pulses
supplied by D1 during TOFF
...
Fortunately,
ceramic and plastic film capacitors meet these requirements
...
The input capacitor can be very small, thanks to the filtering properties
of the SEPIC topology
...
The result can be
misleading, because it doesn’t account for the switch-transition losses or core
losses and the real efficiency could be much lower:
¼ VOUT =AaVIN
Finally, the switch SW and diode D1 should be rated for breakdown voltages with a
15% margin:
VDS ðswitchÞ > 1:15ðVOUT þ VD þ VIN Þ
VR ðdiodeÞ > 1:15ðVOUT þ VIN Þ
Example
Let VIN = 50 À 150 V and VOUT = 15 V at 1 A maximum
...
Now VOUT ¼ 1ÀD , so Dmax = 0
...
091
...
newnespress
...
So Aamin occurs at
Dmin ¼ 0:091 and Aamin ¼ 0:1:
COUT ! Aa min Á IOUT Á Dmin Á T = DVOUT
COUT > 0:1 Ã 1 Ã 0:091 Ã 2 Ã 10À5 =0:1:
COUT > 1:82 mF: Let COUT ¼ 100 mF:
>
CIN > COUT =10: Let CIN ¼ 10 mF:
So, the fundamental component values have been calculated
...
7
...
It is a fly-back circuit and hence will be covered in
Chapter 9
...
4 Common Mistakes in Boost-Buck Circuits
Boost-buck circuits operate with both inductors in continuous conduction mode
...
Calculate the value, add 20%, and then pick the next highest standard value
...
So if temperature rise is an issue, pick a component with a higher
current rating
...
newnespress
...
5 Conclusions
The boost-buck is an ideal topology where the LED load voltage can be higher or
lower than the supply voltage
...
So if the LED
voltage (maximum) is 20 V and the supply voltage (minimum) is 23 V, the difference is
3 V, and 3/20 = 0
...
If the supply voltage
is more than 20% higher, use a buck topology
...
The boost-buck is less efficient compared to buck
or boost topologies
...
newnespress
...
1 Power Factor Correction
Power factor correction, or PFC, is a term used with AC mains powered circuits
...
A pure
resistive load has a power factor of 1, but active loads tend to have power factors
closer to 0
...
The most common power factor correction circuit is a boost converter
...
This is achieved by switching the
current on for short but constant periods: as the supply voltage rises and falls, so does
the amplitude of the current
...
1
...
It is common to switch the
primary current off when a certain current level is reached, but this leads to constant
average current
...
The secondary current will rise and fall at double the AC line
frequency and so a large secondary capacitor is required to absorb this ripple, to
prevent significant ripple in the output voltage
...
However, alternative solutions exist; these are the Bi-Bred and the BuckBoost-Buck (BBB)
...
newnespress
...
1: PFC Circuit
...
2
Bi-Bred
The Bi-Bred is very similar to the Cuk boost-buck that we described in the previous
chapter, see Figure 8
...
The main difference between the Cuk and the Bi-Bred is that, in a Bi-Bred, the input
inductor is in discontinuous conduction mode (DCM) and operation of the output
stage is in continuous conduction mode (CCM)
...
This means that in the design, the input
inductor L1 must have a small enough energy stored to ensure that conduction stops
before the end of each cycle
...
The output inductor L2 must have large enough energy stored (large
inductance value) so that the current only falls to about 85% of its nominal value at
the end of each switching cycle
...
At this time the storage capacitor C3 immediately
begins to charge from the supply voltage through D1 and L1, although the voltage
will not rise very high because, when the MOSFET M1 switches on, the charging
current is redirected to the 0 V rail
...
newnespress
...
2: Bi-Bred Circuit
...
Now the input
circuit acts like a boost converter because the current through L1 cannot change
immediately and it charges C3 to a high voltage
...
The current rises in inductor L2 and the load until the voltage drop across
resistor R7 is sufficient to trip a second internal comparator and turn M1 off again
...
Notice that the current sense resistor is not in this path, because the current level
measurement is not required until the MOSFET turns on again; this minimizes power loss
...
Energy is supplied from a bulk
storage capacitor, C3, with sufficiently large capacitance to provide a more or less
constant supply voltage over an AC line cycle period
...
The Bi-Bred draws a more or less sinusoidal AC line input current when
driven from a switch operating at constant duty cycle, hence a large capacitance value
for C3 helps to produce a good power factor
...
newnespress
...
Or put another way, D ¼ VIVO O
...
5 V (a typical white
þV
3:5
3:5
LED), D ¼ 350þ3:5 ¼ 353:5 ¼ 0:99%
...
This means that a
Bi-Bred is not really suitable for driving short LED strings
...
3
Buck-Boost-Buck (BBB)
The Buck-Boost-Buck (BBB) is a proprietary circuit, patented by Supertex, and is
illustrated in Figure 8
...
It resembles the Bi-Bred in some respects, except for two
current steering diodes D1 and D2
...
3: Buck-Boost-Buck Circuit
...
The
energy stored in each inductor is proportional to the inductance value
...
newnespress
...
This means that the input
inductor value must be relatively small
...
When power is first applied, MOSFET M1 is off and waiting for the first clock signal
to trigger the gate drive pulse
...
With M1 conducting, current begins to rise in amplitude through the inductor L1
until the voltage drop across R2 is sufficient for the internal comparator inside the
HV9931 to trigger, which turns M1 off
...
The voltage is typically midway between the input and
output voltage levels
...
The current rises in inductor L2 and the load until the
voltage drop across resistor R8 is sufficient to trip a second internal comparator and
turn M1 off again
...
Notice that, like in the Bi-Bred, the current sense resistor is
not in this path, because the current level measurement is not required until the
MOSFET turns on again; this minimizes power loss
...
Energy is supplied from
a bulk storage capacitor, C3, with sufficiently large capacitance to provide a more or
less constant supply voltage over an AC line cycle period
...
The BBB draws a more or less sinusoidal AC line input
current when driven from a switch operating at constant duty cycle, hence a large
capacitance value for C3 helps to produce a good power factor
...
This means that there will be some voltage ripple across C3, at a frequency
double that of the AC line (i
...
120 Hz when driven from a 60 Hz line)
...
By adding a simple circuit, the second harmonic can be
reduced; the MOSFET off-time is modulated by the ripple voltage and this acts like
negative feedback to reduce the second harmonic
...
4
...
newnespress
...
4: Buck-Boost-Buck with Harmonic Reduction
...
Between each switching cycle, resistor R7 discharges capacitor C5
...
Capacitor C7
acts as a DC block, to allow just the modulation across C5, rather than any DC level,
to vary the MOSFET off-time
...
2
D
The duty cycle of the switching in a BBB converter is given by VO ¼ 1ÀD
...
So if Vin = 350 V and Vo = 3
...
This is a considerably greater duty
700
700
cycle than the Bi-Bred or the buck converter with a similar low voltage load
...
www
...
com
LED Drivers with Power Factor Correction
8
...
Inductors are
sized for their magnetic saturation level and for resistive heating
...
This inductor can pass 500 mA
with a temperature rise of, say, 40°C
...
If this inductor were used in a PFC stage with a peak current
of 400 mA it would overheat
...
Inductor manufacturers do not normally specify magnetizing losses
...
So when designing a ferrite-based
inductor, a manufacturer will make his design based on this level
...
8
...
This chapter has
been intended to show readers some options and point out limitations
...
Application notes from ST Microelectronics and Supertex cover the PFC, Bi-Bred
and Buck-Boost-Buck circuits in detail
...
www
...
com
CHAPTER 9
Fly-Back Converters
A traditional fly-back converter uses an inductor with at least two windings (really,
this is a transformer)
...
The circuit is arranged so that magnetic energy is stored in
the inductor during the time that the switch is on, when current increases in the
primary winding
...
This is shown in Figure 9
...
The energy release is the ‘fly-back’, so called because in early television sets with a
cathode ray tube, a transformer winding was used to deflect the electron beam back
to the starting point on the screen
...
Fly-back power supplies are relatively easy to design, but are more suited to constant
voltage outputs
...
Driving an isolated LED load is then possible if the secondary winding is
isolated from the primary winding
...
An opto-coupler will be required, to maintain
isolation between primary and secondary, if accurate control of the output current
is required
...
newnespress
...
1: Fly-Back Principle
...
These are buckboost controllers and are an alternative to the boost-buck converters like Cuk and
SEPIC types that were discussed in Chapter 7
...
9
...
2
...
In
this case the start is connected to the MOSFET drain, which alternates between a
ground connection and open circuit
...
Conversely, the outer
layer (end of the winding) is at a fixed high voltage and tends to screen the inner
layers, which reduces radiated EMI
...
newnespress
...
2: Fly-Back Circuit for LEDs
...
The start point of the secondary is connected to the
output diode, but the end point is connected to ground and this tends to screen the secondary
winding for minimal EMI radiation
...
Calculation of the transformer characteristics, like inductance value and primary
to secondary turns ratio, are very important in the design
...
The equation is:
VPRI Á TON VSEC Á TOFF
¼
NPRI
NSEC
9
...
1
Fly-Back Example
Let us make an isolated 3 W lamp by connecting three white power LEDs in series
...
1
...
newnespress
...
4 ms)
...
4 ms in order to allow complete removal of the magnetic energy in the
transformer core
...
With 60 kHz switching, the period will be 16
...
If the average output current is
350 mA, the average in 2
...
43 A
...
86 A
...
L¼EÁ
dt
2:4 Á 10À6
¼ 10 Á
¼ 4:94 mH
di
4:86
Since the primary has ten times the turns of the secondary, the primary inductance
will be 100 times that of the secondary (the turns ratio, N, is squared)
...
Most current-mode power supplies control the switching so that the MOSFET turns
off when a certain peak current is reached in the primary winding
...
86 A and the turns ratio is 10:1, we need a peak primary
di
current of 486 mA
...
486/(5 * 10À6) = 48 V]
...
Actually this gives a constant power output, assuming a constant
voltage input, which is fine for non-critical designs
...
This means that the duty cycle will be reduced and hence the number of voltseconds on the primary will be unchanged
...
This
overshoot increases with increasing input voltage, this is because the delay is constant
but the rate of current rise is increasing with input voltage
...
This resistor injects a small DC bias that increases with
increasing supply voltage and thus triggers the comparator earlier as the supply
voltage rises
...
newnespress
...
1 turns ratio and 10 V output used in the above example cause a reflected
voltage of 100 V in the primary winding when the secondary conduction takes
place
...
The design example does not allow for efficiency
...
If we were designing a constant voltage circuit, we would allow the peak primary
current to be higher than that given in the example
...
We would then use feedback from the output to control the
switching, to reduce the power in the primary, as necessary
...
2 Three Winding Fly-Back
Some fly-back power supplies use a third winding, called a bootstrap or auxiliary
winding, as shown if Figure 9
...
This is used to power the control IC, once the circuit
is operating
...
In our example of a 10 V output from the secondary, the
bootstrap could have the same number of turns and thus give (approximately) 10 V
for the powering the control IC
...
Example start-up regulators are the LR645 and the LR8
from Supertex; these give a low voltage, low current output from an input with a
voltage as high as 450 V
...
The HV9120 shown in Figure 9
...
9
...
1
Design Rules for a Fly-Back Converter
This section gives design rules for a fly-back converter based on either turns ratio
selection determined by the maximum duty cycle allowed (case 1), or by the optimum
turns ratio based on the maximum working voltage of the MOSFET switch (case 2)
...
newnespress
...
3: Fly-Back Using a Three-Winding Transformer
...
In case 2, a design based on the maximum
voltage across the MOSFET allows a potentially lower cost solution
...
The transfer function of a fly-back converter is:
VO
D
¼
ÁN
VI
ð1 À DÞ
So the duty cycle can be found by transposing the equation:
V O Á ð 1 À DÞ ¼ V I Á D Á N
VO ¼ VI Á D Á N þ VO Á D ¼ D Á ðVI Á N þ VO Þ
D¼
www
...
com
VO
ðVI Á NÞ þ VO
Fly-Back Converters
155
Case 1: Turns Ratio Based on Maximum Duty Cycle
Given the minimum input voltage VI_MIN, output voltage VO and maximum duty
cycle DMAX, the turns ratio N can be calculated:
N¼
VO Á ð1 À DMAX Þ
VI MIN Á DMAX
DMAX is typically chosen as 45% (0
...
With DMAX < 50%, the system is inherently stable and there is no complex
compensation required
...
6 V for the output diode) and allow 45% duty cycle, we get:
N¼
10:6 Á ð0:55Þ
¼ 0:282
46 Á 0:45
This is the minimum value
...
33 (3:1)
could be used
...
This was illustrated at the beginning of this chapter,
when a 10 V output caused 100 V to be induced into the primary winding of a 1:0
...
Considering that the supply voltage was only 48 V, this
forced us to use a 200 V MOSFET as the primary switch
...
Because the voltage reflected into the primary often has some ringing, a snubber circuit is
used to limit the voltage across the primary winding
...
Parasitic inductance in the transformer
is often referred to as ‘leakage inductance’ because it is the proportion of the primary
inductance that is not coupled into the secondary, so the magnetic field ‘leaks out’
...
newnespress
...
The voltage across the Zener diode
will be greater than the voltage induced into the primary from the secondary (output)
voltage, otherwise power dissipation and losses will both be very high
...
N¼
VO þ VF
ðVSW À VZ À VIN
As a safety margin,ðVSW À VZ À VIN
MAX Þ
MAX Þ
! 10V
...
The output is 10 V, so allowing for VF this becomes 10
...
5 turns ratio (N = 0
...
The primary voltage
induced from the secondary winding will be 21
...
8 V, which is a reasonable margin to minimise power dissipation
...
With a turns ratio of 1:0
...
Let us use
case 1, with 41% as the maximum duty cycle
...
newnespress
...
35 A = 3
...
The input power is then 4
...
The input current at minimum input
voltage is then:
IAV ¼
PIN 4:12
¼
¼ 0:09A
46
VIN
IPK ¼
2 Á IAV
DMAX
At 46 Vin and 41% duty cycle:
IPK ¼
2 Á 0:09
¼ 0:439A
0:41
With 60 kHz switching, the period will be 16
...
With a 41% duty cycle, the switch
on-time will be 6
...
So we need the primary current to rise by 439 mA in 6
...
LPRI ¼
VIN Á dt 46 Á 6:835 Á 10À6
¼
¼ 716 mH
dI
0:439
The secondary has 1/3 the number of turns compared to the primary, so the
inductance of the secondary will be 1/9, or 79
...
The other design parameter for the transformer is the size and AL factor of the ferrite
core
...
The flux density (B) will depend on the cross-sectional area of the core (Ac),
given in square meters
...
For E20/10/6 cores, the core cross-sectional area is 32 mm2
...
The number of turns can be calculated, based on the design
parameters above and using B = 200 mT as the maximum flux density:
N¼
LPRI Á IPK
ðturnsÞ
AC Á BMAX
N1 ¼
716 Á 10À6 Á 0:439
¼ 49
32 Á 10À6 Á 0:2
AL ¼
LPRI 716 Á 10À6
¼
¼ 298 nH
N12
2401
www
...
com
158
Chapter 9
Refer to core manufacturer’s specifications and choose a core with a lower AL value
(larger gap) than calculated using the above equation
...
The number of turns can then be
calculated as:
rffiffiffiffiffiffi
L
, when L is expressed in nH: Thus 716 mH ¼ 716,000 nH:
N¼
AL
NPRI = 54 (rounding up to the next highest value)
...
9
...
This is shown in Figure 9
...
V+
C1
L1
C4
6
C2
5
150K
7
5K
C3
3
VDD
1
VIN
LED
D1
LED
PWM_D
4
GATE
CS 2
LD
HV9910
GND
Rosc 8
Q1
R2
R1
Figure 9
...
Current is forced through the inductor by a MOSFET connecting the inductor across
the power supply rail
...
At a
predetermined current level, the MOSFET is turned off and the current is forced to
flow through a diode to charge the output capacitor and drive the load
...
newnespress
...
Like the two-winding fly-back, the single winding fly-back can be
calculated from the number of volt-seconds on the charge cycle equalling the
number of volt-seconds on the discharge cycle
...
555
...
This means that the inductor current falls to zero at the
end of each cycle
...
The period is 10 ms, so the on-time is 5
...
45 ms
...
To average 350 mA output, the average current during the off-time must
be 350=0:445 mA ¼ 786:5 mA, so the peak current must be double this, or 1
...
This means that during the on-time, the current must rise from zero to 1
...
The voltage from the power supply is 24 V, so using the familiar equation:
E ¼ ÀL Á
L¼EÁ
di
dt
dt
5:55 Á 10À6
¼ 24 Á
¼ 84:67 mH
di
1:573
In practice there should be some dead time allowed, when the inductor carries no
current, to ensure discontinuous conduction mode
...
Too much dead time means that
the peak current is higher and this reduces the efficiency of the power supply
...
newnespress
...
44 ms; this will reduce the
inductance by 25%
...
E ¼ ÀL Á
À30 ¼ 68 Á 10À6 Á
di ¼
di
dt
di
4:45 Á 10À6
À30 Á 4:45 Á 10À6
¼ 1:963 A
68 Á 10À6
Increasing the peak current by 25% gives the desired result
...
www
...
com
CHAPTER 10
Essentials of Switching Power Supplies
This chapter will examine the advantages and disadvantages of the various driver
techniques, which have already been described
...
10
...
A linear LED driver is generally less efficient
than a switching driver
...
For
example, if you have a 12 V power source and three LEDs each having a 3
...
5 V
...
5 V will be 87
...
It would be difficult for a
switching LED driver to achieve this level of efficiency
...
On the other hand, driving one LED from a 12 V supply would give an
efficiency of 3
...
Here a buck switcher
would give closer to 90% efficiency
...
1
...
Otherwise cost usually takes
precedence and the cost of a switching regulator with EMI filters would be
somewhat higher
...
newnespress
...
5 V
6
1
VIN
VDD
HV9910 8
RT
5
7
3
...
1: Linear vs Switching Solutions
...
2 Switching Regulators
In Chapters 5 to 9 we looked at switching regulators, which have much higher
efficiency, but can generate electro-magnetic interference (EMI) which has to be
suppressed by careful circuit board design, screening and filtering
...
Although Supertex’s LED driver integrated circuits are used in examples, similar
drivers from other manufacturers can also be used
...
The National
Semiconductor LM5020 is a buck controller, like the HV9910B
...
Switching power supplies have the disadvantage of producing electromagnetic
interference (EMI)
...
This is a legal requirement and product cannot be sold unless the
equipment meets the standards laid down in law
...
Conversely, where EMI requirements are very demanding, such as medical and
automotive applications, linear LED driver techniques can be used instead
...
newnespress
...
10
...
1
Buck Regulator Considerations
In Chapter 5 we first looked at the simplest switching regulator, the buck converter
...
Buck circuits are used for mains
powered LED drivers, when driving a long string of LEDs
...
Buck regulators can be very efficient, maybe 90–95%, especially if the load is a long
string of LEDs with a moderately high forward voltage (i
...
high duty cycle)
...
The MOSFET dissipates
power during the on-time, when it is conducting, but the voltage drop across the
MOSFET switch is usually much lower than the forward drop of a fast rectifier
...
The
output current needs to reduce enough to allow the current sense comparators to be
reset
...
If the ripple current is below 10% of IO, the switching of
the MOSFET can be erratic
...
The ripple current can introduce a peak-to-average error in the output current
setting that needs to be accounted for
...
Therefore, the output current will remain unaffected by the varying input voltage
...
newnespress
...
This
capacitor reduces EMI at the output by providing a bypass path for any switching
current spikes, which may also improve the LED lifetime
...
Therefore, the initial output current accuracy might be sacrificed with
large ripple current levels in the inductor
...
These parasitic elements affect the efficiency of the switching converter
because they cause switching losses
...
2
...
2: Parasitic Elements
...
newnespress
...
Minimizing parasitic elements is
essential for efficient and reliable operation of the buck converter
...
pffiffiffiffiffiffiffiffiffiffiffiffiffi
SRF ¼ 1=ð2 L Á CL Þ;
Here L is the inductance value, and CL is the coil capacitance
...
Therefore, connecting a small capacitor CO ($10 nF) across the LED
string is recommended to bypass these spikes, as mentioned earlier
...
When the MOSFET turns on the diode changes from forward conduction to off
(reverse bias), but this cannot happen immediately because charges have to move
inside the semiconductor material, which takes time
...
Using
diodes with shorter reverse recovery time, TRR, and lower junction capacitance CJ
improves performance
...
The forward voltagedrop of diodes with very fast recovery times is sometimes relatively high and can
lead to high conduction losses, so also consider this when making a diode
selection
...
The discharge current is limited to the MOSFET
saturation current, so MOSFETs with a high on-resistance and a lower saturation
current can sometimes produce lower overall losses
...
newnespress
...
Note that the saturation current in
a MOSFET becomes lower at increased junction temperature
...
When the MOSFET gate drive is activated, the
control IC disables the current sense input for this time period, to avoid false
triggering from the switch-on current surge, previously described
...
Discharging the parasitic capacitance CP into the DRAIN output of the MOSFET is
responsible for the bulk of the switching power loss
...
The switching loss is the greatest at the maximum input voltage
...
This value for FS based on typical
values for VIN and VO can be used in the previous equation if a value of constant
switching frequency is not available
...
newnespress
...
Due to the large amount of parasitic capacitance connected to
this switching node, the turn-off transition occurs essentially at zero voltage
...
Buck Converter AC Input Stage
An off-line LED driver requires a bridge rectifier and input filter; selecting an input
filter is critical to obtaining good EMI
...
As a ‘rule of thumb’, 2 $ 3 mF per each
watt of the input power is required
...
Large values of input capacitor will cause unacceptably high current surges when power is
first applied
...
Inrush
current limiters, usually a negative temperature coefficient (NTC) thermistor rated for
high current, are often connected in series with the AC line to prevent the current surge
...
3
...
The value of the inductor depends on the level of signal attenuation required,
when combined with the input capacitor shunt impedance, to meet the EMI standards in
force
...
318 mH
...
A capacitor connected between the switching side of the filter inductor and ground,
albeit of small value, is necessary in order to ensure low impedance to the high
www
...
com
168
Chapter 10
High
Impedance
XL
AC
Input
Low
Impedance
XC
Low Impedance
path for switching
current
Attenuation of Switching Signals
Figure 10
...
frequency switching currents of the converter
...
1–0
...
A 100 nF capacitor
can be used in a circuit that drives a single 1 W LED
...
2
...
Ignoring PFC applications, a
boost converter driving LEDs will always be powered from a low voltage DC supply
...
A boost regulator is used in this application to drive a
string of 20 mA LEDs from a 3–4 V battery
...
In this application a boost converter powered from a 12 V or
24 V DC supply is used to drive many 350 mA LEDs connected in series, with a
forward voltage in the range 40–80 V
...
Otherwise the output voltage will continue to rise and
eventually cause component breakdown
...
www
...
com
Essentials of Switching Power Supplies
10
...
3
169
Boost-Buck Regulator Considerations
To operate in an environment where the input voltage could be higher or lower than
the output voltage, a buck-boost (or boost-buck) circuit is necessary
...
The situation of having a load voltage range that
overlaps the supply voltage range is commonly found in automotive applications
...
The two types of converters often found in boost-buck applications are known as
SEPIC and Cuk
...
Like boost
converters, over-voltage protection should be provided to prevent excessively high
voltage in case of an open-load condition
...
Shunt capacitors across
the input and output strengthen this filtering, and provide a low impedance path for
the circulating currents
...
Sometimes common mode chokes are added at the input side, to
reduce the radiated signals from the whole circuit
...
5 m long
...
2
...
A power factor of 1 indicates that the voltage and current are
in-phase and have low harmonic content
...
In semiconductor circuits powered from the AC mains, a bridge rectifier converts the
AC power into DC
...
4, because charging of a large smoothing
capacitor takes place each half cycle
...
3–0
...
Power
factor correction is an active or passive circuit designed to correct phase errors and
www
...
com
170
Chapter 10
V
l
Figure 10
...
reduce harmonics, and make the power factor closer to 1
...
A circuit having a good power factor, approaching 1, has an input current that has
low harmonic content with a wave shape that closely follows the sinusoidal input
voltage
...
10
...
5 Fly-Back Converter Considerations
Transformer coupled switching regulators can be designed for a very wide range of
supply and output voltages
...
Fly-back
converters were described in Chapter 9
...
If a wide tolerance can be accepted
for the current regulation, a simpler and cheaper circuit can be built
...
Fly-back converters have the advantage of stepping up or down the output voltage
compared with the supply (buck-boost)
...
A single winding inductor is usually much lower
cost
...
newnespress
...
5
...
The output requires a large storage capacitor to
maintain current flow in the LEDs when the converter is on the first step
...
1A 800 V
T1
Luxeon
star X3
BYV26C
100 nF
22 μF
UF4006
12
VDD
220 μF 100 nF
200 V
1W
1
VIN
10 M
2
...
5: Discontinuous Fly-Back Current
...
newnespress
...
2
...
This current can be very high, causing
temporary heating in the capacitor and possible damage to switch contacts or
components connected in series
...
For AC mains applications, an NTC thermistor designed to carry high current is
often used
...
See Figure 10
...
NTC
33R
IF
AC
Input
VC
IC
+
IF
I
C1
22 μF
400 V
VC
IC
t
t
Figure 10
...
For DC applications, an active inrush limiter is more common because the losses can
be minimized during normal operation, when inrush limiting is not needed
...
7
...
newnespress
...
7: Active Inrush Circuit
...
2
...
This could be to reduce damage to switch contacts
by the risk of sparking
...
For example, a circuit for driving one or two power LEDs from the AC mains
could use a double-buck topology
...
But using a polyester film capacitor means that the
voltage dips between switching cycles; since the output power is normally constant,
this means that the input current will peak as the input voltage dips
...
If the output current was controlled, i
...
reduced as the supply voltage
dipped, the input current would remain constant when switching
...
Soft-start can also be implemented by connecting an RC filter to the analogue
dimming input (e
...
linear dimming pin of HV9910B)
...
newnespress
...
7 μF
350 mA LED
D1
10BQ060
1
6
C3
2,2 μF 10 V
5
R
VDD
HV9910
8
PWM_D RT
R1
100K
D
7
C
L1
470 μH
VIN
3
LD
GATE
CS
GND
4
Q1
VN3205N8
2
R2
0,62R
Figure 10
...
and grows as the capacitor charges
...
8
...
newnespress
...
It will describe how different
materials and component types can affect the performance of LED drivers
...
11
...
They have negatively charged electrons orbiting around this nucleus, like
planets around the Sun
...
Lighter atoms, like silicon, are most stable when there are
eight electrons in their outer orbit
...
The addition of a small amount of material (dopant) with either three or five electrons
in their atom’s outer orbit can create an imbalance because, when combined with the
four electrons of silicon, there are either seven or nine electrons in the outer orbit
...
This hole appears as a free positive
charge and is called P-type semiconductor
...
1, diagram A
...
This is shown in Figure 11
...
www
...
com
176
Chapter 11
DIAGRAM A
14+
13+
Si
Al
Free Hole
(P-Type)
DIAGRAM B
14+
15+
Si
P
Free Electron
(N-type)
Figure 11
...
When P-type and N-type semiconductor form a junction, the free electrons and
holes combine and are destroyed
...
Thus there is an energy barrier created; we have a diode junction
...
2
...
This forces more positive charge into the P-type material
and more negative charge into the N-type material
...
7 V potential difference across the P-N junction
...
www
...
com
Selecting Components for LED Drivers
P-type
177
N-type
–
–
–
–
–
+
+
+
+
+
–
–
–
–
–
+
+
+
+
+
–
–
–
–
–
+
+
+
+
+
–
–
–
–
–
+
+
+
+
+
Negative
Charge
Positive
Charge
Figure 11
...
11
...
1
MOSFETs
Metal oxide silicon field effect transistors (MOSFETs) are used as electronic switches
in switching and linear LED driver circuits
...
A MOSFET has three terminals – gate, drain and source; a fourth ‘body’
terminal is internally connected to the source
...
3
...
3: N-Channel MOSFET Construction
...
Also notice that there is a parasitic diode due to the P-type material of the
body and the N-type material of the drain
...
www
...
com
178
Chapter 11
To create a conducting channel in the body of the MOSFET requires a certain
amount of gate potential
...
Because the gate-body isolation is a dielectric, gate-source
and gate-drain capacitance values are usually found in the datasheet
...
A ‘logic-level’ device is defined as one that switches fully
on at Vgs equal to 5 V; this means that the gate threshold is typically about 2 V
...
A logic-level device can also be operated with Vgs equal to 10 V or higher,
in which case the on-resistance is lower
...
MOSFETs have two current ratings – peak current and continuous current
...
Peak current ratings are the maximum
current that is able to flow
...
However, note
that this current is normally quoted at 25°C; at 100°C the peak current is about
half this value
...
When the MOSFET is connected to a load, but turned off, the drain is at high
voltage
...
The gate-drain capacitance thus sees a large
voltage fall on the drain side and a slight rise on the gate side
...
Figure 11
...
Instead of considering the gate-drain capacitance and the gate-source capacitance, we
can consider the gate charge
...
In switching circuits, the gate charge is most significant and is usually quoted in
nano-coulombs (nC)
...
newnespress
...
4: MOSFET Circuit with Parasitic Capacitance
...
I ¼ IQ þ QG Ã FSW
This is important when calculating the power dissipation in a MOSFET driver
circuit
...
11
...
2
Bipolar Transistors
Bipolar transistors are used in switching and linear LED driver circuits
...
The base-emitter voltage is about 0
...
There is some base-emitter resistance, so the
forward voltage drop will increase slightly with base current
...
A current
mirror is one where two or more branches carry identical currents; the current in one
branch depends on the current in another, hence the ‘mirror’
...
newnespress
...
Transistors of the same type have very similar
characteristics, so by adding a low value resistor between the emitter and ground any
variation in the base-emitter voltage (Vbe) is negligible; see Figure 11
...
V+
Primary
String
–
Maximum
Voltage
Drop
Q1
Q2
Q3
Qn
V–
MATCHED NPN
V+
Primary
String
–
Maximum
Voltage
Drop
Q1
Q2
Q3
V–
NON-MATCHED NPN
Figure 11
...
www
...
com
Qn
Selecting Components for LED Drivers
11
...
3
181
Diodes
There are many different diodes (rectifiers)
...
Schottky diodes have the lowest forward voltage drop and the shortest reverse
recovery time, but they are more expensive than standard diodes and generally
have a limited reverse breakdown voltage range, although the company Cree has
recently introduced high voltage Schottky diodes
...
Reverse leakage is higher than in most P-N junction diodes
...
Note that the forward voltage
drop across a Schottky junction tends to increase with diode voltage rating, so
use the lowest voltage rating suitable to keep the conduction losses to a
minimum
...
When the voltage
across a diode is suddenly reversed, an initial current flow will occur in the reverse
direction
...
The labels fast, ultra-fast and hyper-fast are sometimes given
...
More recent devices are much faster, for example the STTH1R06
600V 1A rectifier with Trr $ 30 ns
...
This is because the
reverse current often flows through the MOSFET switch when the voltage across the
MOSFET is high, so the less time when this happens gives lower losses
...
In some
applications a ‘soft-recovery’ diode should be used, where the turn-off speed in the
reverse biased condition is fast but at a controlled rate of change
...
This
snubber often has a medium speed diode in series so that the diode is still conducting
for a period and allows any ringing current to flow through the RC network and
thus decay quickly
...
newnespress
...
1
...
These devices are typically semiconductors:
Zener diodes, Transorb suppressors or voltage dependent resistors (VDRs)
...
Low
voltage Zener diodes rated below 6 V have a soft knee in their current versus voltage
graph; the conduction increases gradually
...
Zener diodes can exhibit some noise when breaking down and are often used
with a small capacitor in parallel to reduce this effect
...
Transorbs can be uni-directional or bi-directional and rated from low
voltage $5 V up to several hundred volts
...
A voltage dependent resistor (VDR) has high resistance at low voltage and low
resistance at high voltage
...
A VDR can absorb high surge energy; the devices are often rated in
joules rather than watts, because the surge energy is short lived
...
11
...
2
...
There
are two types of storage, slow storage and fast storage
...
The purpose of this storage
is to supply energy to the LED driver between the peaks of the AC voltage, which
is twice every cycle
...
newnespress
...
For slow storage, an aluminium electrolytic capacitor is often used because it has
a high energy storage density (they take up less space for an equivalent amount of
storage, compared to other dielectric types)
...
Because of this construction, they
cannot be used in a high temperature environment for long periods; the dielectric
dries out and the capacitor eventually fails
...
The energy only has to be stored for a short time, as
short as a few microseconds, so the main characteristic of the capacitor for this
function is to have the ability to store and discharge energy quickly
...
Surface mount components generally
have lower self-inductance because they have no added lead inductance
...
Capacitors are constructed from two conducting surfaces (known as plates) separated
by an insulator (known as a dielectric)
...
The dielectric can be a
number of materials including ceramic, mica and plastic film
...
Ceramic and mica capacitors are made using flat dielectric sheets; the simplest
construction uses just one insulating layer with a conducting plate on either side
...
Higher valued
devices use several insulating layers with interleaving layers of metal film
...
Plastic film capacitors, such as polyester, polypropylene, polycarbonate, etc
...
One form of construction is identical to that
of ceramic capacitors, where flat sheets of metallized film are used
...
Another form of construction for plastic film capacitors uses rolled films
...
The films are
laterally offset from one another so that the conductor of ‘side A’ protrudes from one
www
...
com
184
Chapter 11
side, and the conductor of ‘side B’ protrudes from the other side (this technique is
sometimes known as extended foil)
...
The rolled form of construction provides a
metal film around the body of the capacitor; this can be connected to earth or the
‘earthy’ side of a circuit to reduce external electric field pickup
...
A capacitor’s behaviour is not ideal
...
Every capacitor will have some series inductance;
this is due to the plate conductors and the lead wires attached to them
...
Each capacitor will also have series resistance due to both the conductors
and the dielectric of the insulator, this is known as equivalent series resistance or ESR
...
An equivalent circuit for a capacitor is shown in
Figure 11
...
L
ESR
C
Figure 11
...
Generally, ESR and self-inductance is more of a problem with aluminium or tantalum
electrolytic capacitors
...
Digital circuit designers have long since become accustomed to connecting
10 nF ceramic capacitors across tantalum devices used for power supply decoupling
...
Dissipation factor (DF) and loss tangent are terms used to describe the effect of ESR
...
This is the tangent of
the angle between the reactance vector Xc, and the impedance vector (Xc + ESR),
where the ESR vector is at right angles to the reactance vector
...
newnespress
...
Self-resonance
occurs due to the device construction: leads are inductors (albeit low value) and
wound capacitors can have some inductance because the currents circulate through
the capacitor’s plates
...
5 mm (or 0
...
A rough idea of the self-resonant frequency can be found by calculating the
inductance of a component lead
...
5 mm diameter lead that is 5 mm
long (2
...
94 nH in free
space
...
Replacing the 1 nF capacitor in previous
calculations with a 10 nF capacitor, results in the self-resonant frequency falling to
29 MHz
...
5 mm
leads was about 20 MHz, not 29 MHz
...
Adding the inductance of the plates gives a lower self-resonant
frequency
...
For small value capacitors of less than 1 nF the self-resonant frequency can be
1
approximately calculated by the following equations: fR ¼ 2 pffiffiffiffiffi , where L is the lead
LC
À
inductance
...
All dimensions are in millimeters (mm)
and the inductance is in mH
...
25 mm (0
...
5 mm each
leg), the inductance is 2
...
This is 2
...
When substituted into the
frequency equation, with a 1 nF capacitor, the self-resonant frequency is calculated to
be 92
...
Surface mount capacitors are in common use now because of their small size
...
This reduction in inductance has benefits for switching
power supplies too; where fast pulse rise and fall times are needed
...
newnespress
...
Some conventional
leaded ceramic capacitors are surface mount devices with wire leads attached
...
Ceramic capacitors generally have a temperature coefficient that is zero or negative
...
Other ceramic dielectrics are
described by the temperature coefficient; N750 describes a dielectric that has a
negative temperature coefficient of À750 ppm/°C
...
The X7R and Y5U capacitors have a wide tolerance on
the component value
...
A high voltage AC signal can generate acoustic noise
...
The piezo-electric effect will also cause the
capacitance value to change with applied voltage
...
They are thus ideal for making LC filters
...
Polyester and polycarbonate capacitors are very common
...
Polyester capacitors are popular because they have a high
capacitance density (high capacitance value devices are small)
...
Another useful feature of polycarbonate capacitors is that they are ‘self-healing’: in
the event of an insulation breakdown due to over-voltage stress, the device will return
to its non-conducting state, rather than become short circuit
...
For universal AC
input, 275 V AC X2 rating is normally used
...
This capacitor reduces EMI emissions and absorbs fast transient surges
www
...
com
Selecting Components for LED Drivers
187
from the mains supply
...
Capacitors from each AC power line to ground (earth) are sometimes used and
must be 250 V AC Y2 rated
...
Capacitance values are readily available in the range
1 nF to 47 nF
...
2 nF is commonly found in power supply designs
...
2
...
Details of
custom-made components will be covered in Chapter 12
...
A length of wire creates inductance, but winding insulated wire into a coil can
magnify this; the wire is normally soft copper covered with a thin plastic film
...
Although a simple coil creates inductance, if a magnetic material is placed within the
coil the inductance increases considerably
...
The advantage of this
type of inductor is that the saturation current level is very high, considering the
inductor size
...
Many low value inductors look like wire ended resistors,
with colored bands marking their inductance value
...
The toroidal shape keeps the magnetic field contained
...
Toroidal inductor winding is not easy, and so these types of inductors can be more
expensive than their bobbin wound counterparts
...
The coil is wound on a
bobbin, within a closed ferrite material
...
The central ferrite core inside the coil
often has an air gap to increase the saturation current rating, although this reduces
the inductance value
...
newnespress
...
This is because the energy stored in an inductor is given by E ¼ 1LI 2
...
If we ignore physical
imperfections due to the construction of an inductor, when a voltage is applied across
it the current will increase linearly
...
If we alternately switch the voltage source and the load across
the inductor, the current will rise and fall, but remain fairly constant
...
Because of their energy storage characteristics, they tend to oppose any change in
current, so they present high impedance to unwanted interference
...
Inductors can be a source of many problems
...
This is
because they are usually made up from tens or hundreds of turns of enameled copper
wire that is wound on a ferrite core
...
This capacitance
causes switching losses in power supplies, or poor filtering in supply input filters
...
Inductors also possess some series resistance due to the intrinsic resistance of the
copper wire used
...
Heating effects due to this resistance can cause problems
...
Magnetizing (core) losses are also present and are due to the energy required to make
the magnetic fields in the core to align with each other
...
These losses increase rapidly if
the magnetisation is forced to operate outside its linear region
...
Transformer cores that have no air gap and are prone to saturate easily
...
If the current drops to or near zero each switching
cycle, the peak current should be kept well below the saturation level (I suggest
Imax = 0
...
25 * Isat)
...
newnespress
...
Some manufacturers quote a saturation current at the point where the
inductance has fallen to 60% of the zero current value
...
This is
the voltage or current magnification value in a tuned circuit
...
This is because of the ‘skin effect’
...
The effect is due
to an inductive force concentrated at the center of the wire, which forces the
electrons to travel down the outside surface (hence ‘skin’ effect)
...
Originally, the
wire strands were covered overall with a cotton braid and called Litz wire
...
It comprises several strands of enameled
copper wire inside a cotton braid
...
Off-the-shelf transformers are available with double or multiple windings, with or
without an air gap in the magnetic core
...
A forward converter is a
popular power supply topology, which uses ungapped cores because magnetic energy
is not stored in the core – it is immediately transferred to the secondary winding
...
Transformers with multiple windings are used to create a step-up or step-down
primary-to-secondary turns ratio
...
Very small duty cycles less than 5% should be avoided,
because of the difficulty in controlling the switching (due to delays in the system)
...
In some cases, such as where the input voltage range is very wide,
a wide range of duty cycle may be unavoidable
...
newnespress
...
A bootstrap
circuit creates a power supply for the switching circuit, typically in the range 8–15 V
...
Once switching starts, the
voltage developed on the bootstrap winding can be used to self-power the switching
circuit
...
11
...
3 Resistors
There are several types of resistor
...
They are used at the AC power input of some power supplies
to provide some impedance for fast transients and surges
...
They are
constructed using carbon particles set in a clay rod and the resistance depends on the
surface area of the touching particles
...
Carbon film resistors are low noise devices with a negative temperature coefficient
...
1% are available albeit
more expensive
...
The spiral conductor is actually a lossy inductor
...
The short length of carbon film has very little inductance
...
Component tolerances of 1% are standard, although
precision devices in an E96 range of values with 0
...
These resistors are constructed
by applying a number of metal film layers, of different metals, to a ceramic former to
achieve the correct resistance and a low temperature coefficient
...
www
...
com
Selecting Components for LED Drivers
191
All conductors have some series inductance, simply due to having a certain length
...
In fact some high frequency circuits just use a thin wire
bond to form an inductor
...
Some types have more inductance than others
...
Wire wound
resistors have a significant inductance because of their construction, when a wire is
wound into a coil its inductance increases in proportion to the number of turns
squared
...
All through-hole
components also have some inductance due to the wire leads at either end
...
The two ends have a certain cross-sectional area
and are spaced a certain distance apart, separated by a ceramic dielectric
...
2 pF, so has little effect in an LED driver circuit
operating up to 1 MHz
...
11
...
The notable exceptions are military and (ironically)
medical applications, although these will be forced to change due to the lack of RoHS
lead-free components
...
This means that soldering profiles have to
change – higher temperatures are needed for lead-free solder
...
At high frequencies, for example,
capacitance between tracks can cause a lower resonance frequency in a tuned circuit
...
There are several types of
board, with FR4 (fibreglass insulator) being the most common
...
For slow speed circuit prototypes they are ideal for
fault finding and fast construction
...
www
...
com
192
Chapter 11
11
...
1 Through-Hole PCBs
It is usual for an RF or high speed digital circuit to have an earth plane on the printed
circuit board (PCB) component side
...
The earth plane serves two purposes; it screens the components from tracks passing
underneath, and it provides part of a low loss transmission line
...
6 mm, 50 ohm transmission lines can be created by
making the printed circuit tracks 2
...
A transmission line is formed between
the earth plane and the track
...
This capacitance forms a
parallel tuned circuit with the inductance and may cause the filter to be detuned
...
An
alternative solution is to mount the inductor on spacers above the board, so
reducing the capacitance
...
3
...
Ceramic
capacitors are common but can be damaged by stress due to circuit board expansion
...
18 Â 0
...
Ceramic capacitors should be protected with a moisture resistant coating
...
Moisture
can also be absorbed into plastic packages, so a conformal coating over the whole
board is preferred
...
This will
prevent moisture being trapped into an assembled board and avert the risk of damage
during soldering (as the moisture boils off)
...
Surface mount boards do not need holes large enough for component
leads; hence they tend to be smaller in diameter
...
3 mm in
diameter are common (used to connect two tracks rather than for component leads)
...
newnespress
...
Glass and epoxy board,
e
...
FR4 type, has a high coefficient of expansion at temperatures above 125°C
...
Soldering causes a problem due to the heat applied to the board; in wave soldering
the board is heated to about 300°C, which is way above the glass transition
temperature
...
Temperature cycling of completed
boards also causes problems
...
Leadless chip carrier (LCC) devices have an expansion coefficient of
6 ppm/°C, but for the board it is 14 ppm/°C (below the glass transition temperature)
in the X-Y plane
...
Again, temperature cycling strains the solder joints and can
lead to failure
...
Printed circuit boards built on aluminium sheet are often used with power LEDs
...
Traditionally, copper
clad invar has been used within some PCBs to restrain expansion and to distribute
heat
...
Solder resist can be used to restrain solder, but this can create large blobs on the lead
or pad area
...
PCBs that have a fine track pitch sometimes have 0
...
If the gold is
thicker it causes embrittlement
...
11
...
The AC characteristics
are less affected by temperature
...
newnespress
...
The ideal op-amp has infinite
input impedance, zero output impedance and a flat frequency response with linear
phase
...
The output impedance is not zero, and can be up to about
100
...
There is, however, an assumption that the gain-bandwidth of the op-amp is far higher
than that required by the circuit
...
If the op-amp has insufficient gain-bandwidth product, excessive phase shifts occur
and the circuit can show peaking in the frequency response
...
A good frequency
response can be obtained by utilizing an op-amp that has a gain-bandwidth product
many times that of the circuit’s bandwidth
...
Comparators are used in many LED drivers to detect the current level in a sense
resistor
...
Often
a comparator has some inbuilt hysteresis to prevent jitter when the two inputs are
at or near the same potential
...
For example, the current sense comparator in the HV9910B
LED driver has an offset of about 10 mV, and the maximum threshold for
switching is 250 mV, so the threshold range is 10–250 mV, and could potentially
give a dimming range of over 20:1
...
It is possible to build a comparator by using an op-amp with positive feedback
...
www
...
com
CHAPTER 12
Magnetic Materials for Inductors
and Transformers
Standard off-the-shelf transformers and inductors were described in Chapter 11
...
The primary design requirement is to minimise losses,
but to do this we have to consider copper losses, core losses, magnetic saturation, size
and construction
...
For more detail, the reader should consult
specialist books on the subject
...
By soft, the meaning is that magnetisation is easy and
demagnetisation occurs when the magnetising force is removed
...
Most magnetic materials have some
remnance and the field strength required to return the magnetic flux to zero, to
overcome this remnance, is called the ‘coercivity’
...
But when the magnetic field
is reversed, the flux does not follow the same curve; it needs more field strength (more
energy) to return to the same point and thus forms a ‘fat S’ shape
...
The core can be rectangular or cylindrical in cross-section with two halves that
separate to allow the bobbin to be inserted
...
newnespress
...
This form of
inductor is suitable for values of a few micro-henries up to about one henry
...
Remembering that inductance is proportional to the number of turns squared,
qffiffiffiffiffiffiffiffiffiffi
the number of turns required is given by the simple formula: N ¼ LðnHÞ: Here L is the
AL
required inductance in nano-henries and AL is the core’s inductance factor (nanohenries per turn)
...
The AL factor is
the inductance, in nano-henries, that will be produced for a single turn of wire
...
Different magnetic materials are used, depending on the frequency at which the inductor
is operating
...
Note: an
air gap in the center of the core, rather than in the outer material, reduces the emission of
magnetic fields because the outer material behaves like a shield
...
A typical core gap is 0
...
5 mm, although it may be larger or smaller depending
on the magnetic material permeability and the required AL value
...
The presence of an air gap in inductor and transformer cores makes them suitable for
high magnetic saturation levels
...
In PFC
circuits, the current is switched on and off at high frequency with zero current flow
between each pulse
...
Thus
the power factor is close to unity (true sine wave)
...
Gapless inductor cores are often used in forward converters, in which
the secondary current flows at the same time as the primary current
...
If the coupling between windings must be very close, bifilar winding is often used
...
www
...
com
Magnetic Materials for Inductors and Transformers
197
Trifilar and higher order windings use multiple strands
...
g
...
Sometimes multiple winding strands are used to reduce the equivalent series
resistance, because at high switching frequencies the skin effect must be considered
...
A type of winding wire with multiple twisted strands is called Litz wire; each
strand has a thin polymer film surrounding the conductor, for insulation
...
1
Ferrite Cores
Ferrite cores are available in many shapes and material types
...
Ferrite is usually
a compound made from magnesium and zinc, or from nickel and zinc
...
Nickel-zinc ferrites are used in inductors intended for EMI filters, because they
have high losses at high frequency – the core absorbs most of the energy above
20 MHz, up to about 1 GHz
...
This characteristic makes them almost useless for EMI
filtering
...
Ferrite is less effective at very low or very high
frequencies
...
12
...
The iron dust is ferrous oxide and is mixed with clay-like
slurry, which sets when baked
...
www
...
com
198
Chapter 12
These cores are good for switching frequencies up to about 400 kHz
...
Above 20 MHz the core has little effect
and so cannot be used in EMI filtering applications
...
3
Special Cores
Proprietary compounds are used to make special cores
...
This has the ability to operate with high flux density of
typically 800 mT, rather than 200 mT of conventional ferrite cores
...
MPP cores possess many outstanding magnetic characteristics, such as high electrical
resistance (thus, low eddy current losses), low hysteresis (magnetizing) losses,
excellent inductance stability after high DC magnetization or under high DC bias
conditions and minimal inductance shift when subject to flux densities up to
2000 gauss (200 mT) under AC conditions
...
4
Core Shapes and Sizes
For custom inductors and transformers, E-cores are popular
...
The center segment is designed to pass through
the middle of a bobbin on which the windings are wound
...
1, to allow high magnetic
flux without saturation of the core
...
The EFD core is shaped so that the
center segment is thinner than the main body of the core, so that the bobbin has a
rectangular cross-section, rather than square
...
However, the area on the circuit board is essentially square
...
These cores are rarely used except in a tuned filter, when an
adjuster is provided in the central spigot
...
newnespress
...
1: E-Core
...
However, toroids are difficult to
wind, since the wire must loop many times through the central hole
...
Magnetic saturation can be a problem, so
MPP and iron powder tend to be used because they have the ability to carry a
high flux density
...
2
...
2: Toroidal Core
...
5
Magnetic Saturation
Magnetizing (core) losses are also present and are due to the energy required to make
the magnetic fields in the core to align with each other
...
These losses increase rapidly if
www
...
com
200
Chapter 12
the magnetization is forced to operate outside its linear region
...
If an inductor or transformer has a large discontinuous current flow, as in certain
fly-back transformers and input inductors, the magnetic flux density may need to
be lower than 200 mT
...
Inductors requiring the ability to handle high levels of flux variation sometimes
use special cores with low losses at high flux density, in which flux levels much
greater than 50 mT are used
...
LI
The flux density is given by the equation: B ¼ N Ae : Here, L is the inductance, I is
the peak current, N is the number of turns and Ae is the effective core area
...
We do not know the core area or the number of turns at this stage, but through
iteration we can find something suitable
...
The number of turns can be found by transposing
I
the previous equation: N ¼ BLAe
...
N2
Cores are usually available with standard AL sizes
...
However, if
a lower AL value is not available, a larger core size with a higher Ae value should be
selected and the above process repeated
...
12
...
In 99
...
73 Â 10À8 ohm meter
...
The wire resistance at any temperature can be estimated from Table 12
...
www
...
com
Magnetic Materials for Inductors and Transformers
201
Table 12
...
Temperature
Multiplying factor
20°C
40°C
1
1
...
157
1
...
314
Unfortunately, the resistance of wire also increases as the frequency of signals passing
through it increases
...
An alternating magnetic field produced by the current in the wire induces an
electric field, strongest at the center of the wire, which repels the electrons and forces
them to the outside surface of the wire
...
The skin depth is given in Table 12
...
Table 12
...
Frequency
Skin depth
50 Hz
1 kHz
100 kHz
9
...
09 mm
0
...
0662 mm
0
...
newnespress
...
63 mm
...
In fact, in an LED driver (or any PWM power
supply) there are harmonics at many times the switching frequency
...
In some cases, it is necessary to suffer higher copper losses that desirable, in order to
have a transformer of a reasonable size
...
www
...
com
CHAPTER 13
EMI and EMC Issues
The first two questions regarding EMI and EMC are: what is the difference between
EMI and EMC? And which standards apply? Subsequent questions relate to how
equipment can be made to meet the standards
...
EMI is electro-magnetic interference
...
EMI is caused by emissions in the radio spectrum, which
not only interfere with radio systems but also can cause other equipment to malfunction
...
An often seen warning notice at a gasoline station says ‘using a radio
transmitter can cause a fire’, but in reality the most likely effect is to cause an error in the
fuel measurement
...
Medical systems have a high
immunity requirement, because the consequences of a failure are death or injury
...
Power meters connected to lines
where they enter a building are subject to the highest potential surges, so they have
very high immunity requirements
...
www
...
com
204
Chapter 13
Before we look at EMI and EMC standards, and the design techniques used to meet
them, it is important that we understand signals
...
For example,
a square wave with a 50/50 duty cycle has a fundamental signal at the switching
frequency plus a 3rd harmonic of 1/3 amplitude, plus a 5th harmonic at 1/5 amplitude,
plus a 7th harmonic at 1/7 amplitude, etc
...
Typically, this is like the signal across a MOSFET switch in an LED
driver circuit
...
1 EMI Standards
13
...
1 AC Mains Connected LED Drivers
Any LED driver connected to AC mains supply has to meet the limited specified
in harmonic current emissions standard IEC/EN 61000-3-2
...
The harmonic
emission limits specified in IEC/EN 61000-3-2, Ed
...
1
...
1
Harmonic order ‘N’
Maximum current, Class C
(percentage of fundamental current)
2
3
2%
(30 Â Power factor) %
4–40 (even)
5
Not specified
10%
7
9
11–39 (odd)
7%
5%
3%
Conducted emission limits in the 150 kHz to 30 MHz frequency range are specified in
the standard IEC/EN 61000-6-3
...
newnespress
...
1
...
The standard is IEC/
EN 61000-6-3, which covers the frequency range 30 MHz to 1 GHz
...
The limits given in CISPR22 and EN55022 standards were intended for
computers and communications related equipment, but these have been adopted as
generic limits for all electronic products, including lighting
...
From 200 MHz to 1 GHz the emission level
increases to 37 dB mV/m
...
Since the signal power is proportional to 1/R2;
for example, at 1 meter from the EUT the emission limit will be 20 dB higher
(100 times the power), at 50 dB mV/m and 57 dB mV/m, respectively
...
2 Good EMI Design Techniques
It is important to look at the circuit diagram and determine where the possible
sources of EMI are located
...
The center point for EMI sources must be the MOSFET switch
...
When
looking at the circuit schematic, consider the effect of high frequencies (1–200 MHz)
...
Similarly, a capacitor
thought to have a low impedance characteristic behaves like an inductor at very high
frequency; a good example of this is an electrolytic capacitor
...
2
...
Figure 13
...
The integrated circuit is a PWM controller
...
The MOSFET
Q1 turns on and the current increases at a fairly constant rate, due to the inductance
of L1
...
newnespress
...
1: Buck Circuit
...
The MOSFET Q1 turns off but
current continues to flow in the LED and inductor due to the flywheel diode D1
...
When the gate pin of the HV9910B outputs a voltage of 7
...
The drain
voltage is very low, just a small voltage due to the current flowing in the drain-source
channel of Q1 and in the current sense resistor R2
...
When D1 conducts, the drain of Q1 is clamped to the positive supply rail
...
The fast rising and
falling edges create a broad spectrum of harmonics
...
2
...
Analysis
of the LED current gives a path from ground, through the decoupling capacitors
C1 and C2, through the LED and inductor, through Q1 and the current sense
resistor R2, and back to ground
...
Not shown is the current that flows through the flywheel diode D1
...
There is also a momentary reverse current that flows when Q1 first
www
...
com
EMI and EMC Issues
207
V+
C2
C1
C3
D1
LED
1
6
HV9910
C4
Rosc
7
5
3
L1
VIN
VDD
V drain
8
LD
PWM_D
4
GATE
2
CS
GND
Time
R1
Q1
R2
Figure 13
...
turns on
...
A small part of this current is due to the junction capacitance, but the main part
is reverse recovery current
...
The free electrons in the junction take
some time to clear and thus create a depletion region inside the silicon
...
The choice of capacitors in the circuit is important
...
The capacitor dielectric could be ceramic for low voltage supplies,
or metalized plastic film such as polyester
...
The inductor winding capacitance is simply
due to insulated wires being wound over each other in a coil
...
The capacitor
C3 must withstand the voltage across the LED, or the supply voltage if there is a
chance that the LED could be disconnected
...
A typical value is 100 nF
...
newnespress
...
2 mF
...
We briefly mentioned the inductor L1
...
But the
magnetic field must be considered too; a shielded inductor or a toroidal construction
should be used to minimize radiating magnetic fields
...
3
...
Basically adding L2 and C5 creates a low-pass filter to attenuate
(reduce) and high frequency signals from the switching element Q1
...
3: Buck Circuit with Filter
...
A value in the range 10 ohms to 100 ohms is likely to
be sufficient
...
The paths for the switching currents must be kept short and compact when laying
out a printed circuit board (PCB)
...
newnespress
...
Using the circuit
schematic of Figure 13
...
In Figure 13
...
C2
+V
I
L1
+
C5
D1
Drain
Q1
L2
C1
Figure 13
...
Notice how the ground connection goes from C5 to C1 and then on to C2 before
reaching the ground plane
...
High frequency
signals are taken from the grounded side of C2, which is low impedance at high
frequency
...
Figure 13
...
The positive supply from
the bridge rectifier BR1 flows to C5 and onto filter inductor L2
...
Notice that the C2 connection is a node where
BR1
AC in
C2
I
C5
+V
L1
+
D1
Drain
Q1
L2
C1
Figure 13
...
www
...
com
210
Chapter 13
current also returns from the cathode of D1
...
As with the ground connection, the high frequency current path is kept away from
the low frequency current flowing into capacitors C5 and C1
...
6 shows both sides of the circuit board overlaid
...
Both Q1 and L1 have high
frequency, high voltage switching, and a ground plane below helps to reduce the
radiation from this area by screening underneath and making the node low
impedance
...
BR1
AC in
C2
I
C5
+V
L1
+
D1
Drain
Q1
L2
C1
Figure 13
...
13
...
2 Cuk Circuit Example
A Cuk circuit is a boost-buck converter that performs well in a DC input application
...
7
...
An earth plane under the main switching elements will also help
reduce radiation
...
Dipole antennas radiate and receive
signals easily because their metallic elements are resonant at the transmit frequency and
www
...
com
EMI and EMC Issues
D1
VIN+
C1
L2
C5
+
R5
C2–C4
L3
R1//R3
VO–
R2
D5
C10
D3
Q3
211
R8//R12
Load
VO+
VIN–
C9
R4
2
PWM
4
1
5
Vin
CS1
GATE GND
VDD
HV9930
PWM
CS2
REF
C8
R9
3
6
7
R10
8
R7
R11
Figure 13
...
thus high impedance at the ends of the elements
...
An
earth plane under the circuit lowers the impedance and reduces radiation
...
High frequency emissions are caused by the fast rising and falling edges of the
MOSFET drain voltage
...
Not only does this reduce high frequency emissions, it
also reduces high frequency ringing that is caused by the drain-gate capacitance
resonating with stray circuit inductance
...
Slowing the
MOSFET switching speed reduces the efficiency of the LED driver circuit, but saves
the cost of additional filters
...
Figure 13
...
I will now describe the input filter, starting at the input of the switching circuit and
working outwards towards the power source
...
newnespress
...
5 uH
L4A
47 uH
L1
D1
L2
C5
+
C20
10 nF
C11
10 nF
R5
C3–C4
C2
VO–
R2
D5
D3
Q3
C10
R8//R12
R1//R3
VIN–
L3
C1
Load
VO+
L4B
C9
R4
1
2
PWM
5
4
C8
3
R9
Vin GATE GND
6
VDD
CS1
HV9930
PWM
CS2
REF
7
R10
8
R7
R11
Figure 13
...
Capacitors C3 and C4 provide the high frequency current source; these are
ceramic capacitors and have very little high frequency ripple across them
...
The circuit ground is not the same as the supply ground, because two parallel resistors
R1 and R3 break the ground connection
...
The path to circuit ground is provided by C20
...
The path
to supply ground is provided by C2, which is a high value ceramic capacitor
...
This means that a
differential capacitor, like C2, has no effect since the voltage is the same on both sides
of the capacitor and no current flows through it
...
A common mode inductor L4 has two windings on a common magnetic core
...
Common mode currents produce magnetic fields that add together and
thus have a high inductance
...
www
...
com
EMI and EMC Issues
213
Finally, a small value ceramic capacitor C11 is connected differentially across the
power supply input to provide a low impedance path at the higher frequencies
...
9
...
9: Output Filter
...
If the distance is very short, the only
filter needed is a differential capacitor (C10) across the load
...
Thus we may require a common mode inductor, L5,
and a second differential capacitor C23
...
As well as a ground plane on the circuit board to reduce the impedance of the
switching circuits at high frequency, a screen over the components may be needed
...
10
...
However, there will always be some
leakage due to signals being carried outside the enclosure by connections to the
remainder of the circuit
...
www
...
com
214
Chapter 13
SCREEN AROUND SWITCHING AREA
D1
VIN+
L2
C5
+
R5
C2–C4
L3
C1
R2
D5
R8//R12
R1//R3
1
2
PWM
Vin
CS1
4
PWM
C9
3
GATE GND
VDD
HV9930
5
C10
D3
Q3
VIN–
R4
VO–
C8
Load
VO+
R9
6
7
R10
CS2
REF
8
R7
R11
Figure 13
...
13
...
If radio frequency signals cannot get out of some equipment, they
cannot get in either
...
People generate high electrostatic voltages during normal activities, such as walking across
a carpet or opening a plastic envelope
...
Thus equipment must be protected against high voltage
discharge
...
The
standard voltage levels are 4 kV for a contact discharge and 8 kV for an air discharge
...
Each surge pulse has an open circuit rise time of 1
...
In domestic equipment, the peak surge voltage is 1 kV, which is
added to the AC mains supply
...
The test pulses are positive and negative, and are applied at 0, 90,
180 and 270 degree phases of the AC mains voltage
...
This comprises –2 kV pulses with a rise time of 5 ns and 50% decay at 50 ns
...
newnespress
...
There are 75 pulses in each burst, and the bursts are repeated every 300 ms, for
1 minute
...
13
...
The surges
are applied, which are added to the normal AC voltage, at times to coincide with
different phases of the AC line
...
The energy in surge pulses can be absorbed or
reflected to limit its damaging effects in the equipment under test
...
A varistor, which is a voltage dependent resistor made from a metal oxide, is commonly
used to absorb energy by clamping the voltage
...
The
amount of energy absorbed in a varistor depends on its physical size
...
For example, a 9 mm disc varistor from Epcos that is
rated for 275 V AC has a transient energy rating of 21 joules and a peak current rating
of 1200 amps
...
This device is a Zener diode made in silicon and has a stronger clamping
action
...
In
AC systems a bi-directional breakdown is required, but in automotive and other DC
applications, a uni-directional breakdown is sufficient
...
The oldest technology, and still sometimes used, is the gas discharge tube (GDT)
...
When the voltage
across the electrodes is high enough, the gas ionizes and conducts to clamp the voltage
...
This not only helps to reduce EMI emissions and susceptibility,
it also helps to absorb some of the energy in surge pulses
...
www
...
com
216
Chapter 13
Many systems have a large electrolytic capacitor across the power rails, after a bridge
rectifier
...
A plastic film capacitor connected in parallel with the electrolytic
capacitor will help to absorb high frequency energy
...
A fuse should be fitted to every piece of equipment powered from the AC mains
supply
...
When a high-energy surge causes the varistor to break down, the fuse will
blow
...
When laying out a PCB, the spacing between tracks should be carefully considered
...
An air gap of
3
...
On a PCB
the gap between conductors is known as the creepage distance
...
Integrated circuits that can be powered directly from the rectified AC supply
usually have ‘no connect’ or NC pins adjacent to the high voltage pin
...
Where no gap exists, a slot can be
cut in the PCB, or the contact pins can be coated with a conformal coating or a
resin to increase the insulation
...
newnespress
...
1 Efficiency and Power Loss
People sometime refer to LEDs as being a cold light source
...
However, LEDs do indeed generate heat and this has been the cause of failure
of several designs
...
A white LED with a 3
...
225 W of heat
...
Power LEDs should always be mounted on a heatsink
...
A heatsink could be mounted on the backside of the
PCB for removing heat from both the LEDs and the driver circuit
...
When designing analog or switching power sources, we discuss efficiency
...
What designers
sometimes overlook is that input power minus output power equals power loss in the
LED driver circuit; see Figure 14
...
Loss in the driver must be dissipated as heat
...
8 = 11
...
This means that 1
...
www
...
com
218
Chapter 14
POWER
IN
11
...
1: Power Loss in Driver Circuit
...
2 Calculating Temperature
The temperature of a device can be calculated using simple ‘Ohm’s law’ type
mathematics
...
Thermal
resistance can be equated to electrical resistance
...
2
...
2: Electrical Equivalent Calculations
...
3)
...
The thermal resistance between the silicon die and the package, added to the thermal
www
...
com
Thermal Considerations
219
HEAT
θ1
θ2
θ3
TEMPERATURE
θ = θ1 + θ2 + θ3
CURRENT
R1
R2
R3
VOLTAGE
R = R1 + R2 + R3
Figure 14
...
resistance of the package to heatsink interface and the thermal resistance of the
heatsink to air interface, can all be added to find the total thermal resistance from the
silicon junction to air
...
(Note, a 1 degree kelvin temperature rise = 1 degree Celsius temperature rise
...
For example, let uJC = 1
...
1 K/W, and uHA = 2
...
1 K/W
...
7 K/W, the silicon junction temperature will be 37 degree
hotter than the ambient temperature
...
Like electrical resistance, having parallel thermal resistance paths reduces the overall
resistance (see Figure 14
...
Two paths, each of 2 K/W, will create an effect single
path of 1 K/W
...
However, for a first
approximation calculating the temperature drop along obvious thermal paths will
give a sufficiently accurate result
...
Because parallel paths reduce the thermal resistance, in general a large surface area can
dissipate heat much better than a small area
...
newnespress
...
4: Thermal Resistances in Parallel
...
For this reason, a small driver is rarely able to drive a high power
load; remember this when the marketing department asks you to design a smaller driver!
Semiconductor component manufacturers usually specify the minimum and maximum
junction operating temperatures for their devices
...
Commercial device ambient
temperature ratings are 0°C to 70°C, industrial device ratings are À40°C to þ85°C
...
Component manufacturers also specify power dissipation (usually based on 25°C
ambient temperature)
...
14
...
If there is high thermal resistance from the source,
the source temperature will rise until sufficient heat is dissipated (or until components
are destroyed)
...
newnespress
...
One obvious cooling technique is to reduce
the thermal resistance, and thus dissipate heat easier, by using a heatsink
...
Surface mounted power MOSFETs are usually in a D-PAK or D2-PAK housing,
which have a tab for dissipating heat
...
5
...
Surface mount heatsinks are sometimes made
from tinned brass and are soldered to the PCB, either side of the MOSFET body
...
DRAIN TAB
SOLDERED
TO PCB
SURFACE MOUNT
HEATSINK
SOLDER
D-PAK
MOSFET
GATE
PAD
COPPER AREA
ON PCB
(DRAIN)
SOURCE
PAD
Figure 14
...
Through-hole MOSFETs in a TO-220 package can be fitted to various heatsinks with
a wide range of sizes
...
Larger heatsinks could increase
parasitic capacitance and cause an increase in switching losses, but this can be
prevented if the heatsink is connected to the ground plane
...
Switching losses will be due to the capacitance between
the MOSFET drain (tab) and the heatsink
...
This is because the surface of the
MOSFET tab and the heatsink surface are not smooth
...
newnespress
...
Micro-cavities between the two surfaces create air
pockets that have high thermal resistance, see Figure 14
...
The thermally conductive
pad or paste fills these cavities to create a uniform surface with low thermal resistance
...
6: Thermal Resistance Created by Air Pockets
...
Cooler air from outside the equipment can be
blown over warm components to reduce their temperature
...
Careful placing of cooling fans can make a big difference to the performance
...
If the air flows in the direction of heatsink fins, it will be more
effective
...
7
...
7: Fan Cooling of Heatsinks
...
newnespress
...
Fans mounted in the side of equipment are much more effective if two
fans are used, one on either side of the enclosure
...
Fans do have a reliability issue, so consider adding a fail-safe mechanism in case the
fan fails to operate
...
Driving the LEDs at a lower power or
turning them off when the temperature rises too high may be a solution
...
newnespress
...
The
information here is to show that many topics must be considered, rather than as a
reference for design work
...
15
...
This transformer can be
placed on the AC mains supply, or as part of the switching regulator circuit
...
Conversely, a transformer that isolates the output of a switching regulator can be
very small because it is operating at the switching regulator frequency of
typically 50 kHz or more
...
For products connected to AC mains supplies, 1500 V RMS (50 Hz or 60 Hz)
isolation is usually required
...
www
...
com
226
15
...
This is
basically a piece of wire that is heated by the current flow
...
Sometimes two wires are joined by solder, one wire being a
weak spring; as the heat softens the solder joint sufficiently the spring wire pulls the
joint apart
...
Electronic circuit breakers are also available
...
Tyco produce a fuse that becomes high impedance when an over-current is detected,
due to the current’s heating effect, but then remake the electrical connection once the
fuse has cooled down
...
3
Creepage Distance
In most electrical circuits connected to the AC mains supply, creepage distance is a
concern
...
In
either example, the current may not be high enough to blow the fuse, but could be
lethal to the user through electrocution or toxic smoke inhalation
...
Some devices
have ‘no connect’ (NC) pins between high and low voltage pins, so that a small piece
of solder cannot bridge any two points
...
The creepage distance in
air is much less than the creepage distance on a PCB
...
15
...
These tend to
be more expensive than standard capacitors, because they are rated to withstand
voltage surges
...
newnespress
...
Polyester or polypropylene (MKP) is the usual dielectric in X2 capacitors
...
The
typical DC operating voltage of Y2 capacitors is 1500 V
...
2 nF) and are usually made with a ceramic or
polypropylene dielectric
...
Since
these are not rated for operation above their nominal working voltage, they are often
smaller and lower cost compared to X2 capacitors
...
An EMI filter before the bridge rectifier
will tend to prevent voltage surges from reaching more sensitive components and is
thus preferred
...
15
...
The EN60950
standard was originally intended for information technology equipment (i
...
computers
and associated hardware) but, since it is one of the few ‘harmonized’ standards that
have been agreed by all of Europe and many other countries in the world, it has been
used as a reference for most safety regulations
...
The European Low Voltage Directive (LVD) is a safety regulation in Europe that
covers all products operating from voltages of 50–1000 V AC and 75–1500 V DC
...
These directives
require a CE mark to be placed on all goods offered for sale
...
Note that
sub-modules do not require CE marking, but the overall equipment does
...
The SELV (safety extra low voltage) requirements are that no
www
...
com
228
Chapter 15
touchable conducting parts have a voltage (relative to ground, or across any two
points) above 60 V DC, or 42
...
For example, a DC powered
(boost-buck) Cuk converter, with 24 V DC input must not have an output above
36 V
...
An AC mains powered LED lamp must be isolated to meet these regulations – in
addition to the output voltage being limited to 60 V if the electrical connections are
‘touchable’
...
However, the voltage limit can be
ignored if the cover has a micro-switch to disable the equipment in the event of the
cover being removed
...
www
...
com
Bibliography
1
...
3
...
5
...
7
...
9
...
11
...
13
...
Brown, Marty
...
Power Supply Cookbook
...
Pressman, Abraham I
...
Switching Power Supply Design
...
Billings, Keith
...
Switch-Mode Power Supply Handbook
...
Harrison, Linden T
...
Current Sources & Voltage References
...
Zukauskas, Arturas, Shur, Michael S
...
2002
...
New York: Wiley Interscience
...
1998
...
Oxford: Newnes
...
2006
...
Waldenburg Germany: Wurth Elektronik/Swiridoff Verlag
...
2001
...
Dallas TX: Texas Instruments
Incorporated
...
and Nakauchi, Edward M
...
Testing for EMC Compliance
...
Montrose, Mark I
...
Printed Circuit Board Design Techniques for EMC Compliance
...
Lenk, John D
...
Simplified Design of Switching Power Supplies
...
Williams, Tim
...
EMC for Product Designers
...
Oxford: Newnes
...
1998
...
Oxford: Newnes
...
2006
...
Sunnyvale CA: Supertex Inc
...
supertex
...
www
...
com
Index
AC Input, 46–60, 167–68
Active current control, 20–4,
172–73
Automotive lighting, 11, 19
Backlight, 12, 168
Bi-bred, 142–44
Bifilar, 137, 196–97
Bipolar transistor, 119, 131,
179–80
Blanking time, 166
Bobbin, 187
Boost, 61–98, 168
Boost-buck, 169
Buck, 39–60, 163–68
Buck-boost, 139, 169
Buck-Boost & Buck (BBB),
144–46
Capacitor, 42–3, 45, 48–9,
182–87, 226–27
Carbon film, 190
Ceramic, 183–86, 227
Channel lighting, 10
Circuit breaker, 226
Clamping circuit, 83
Color, 9, 14–5
Common-mode choke, 169,
212–13
Comparator, 193–94
Compensation network, 65,
79–82, 100–01
Constant current, 36, 164
Constant off-time, 42, 64–5
Electromagnetic interference
(EMI), 33, 37, 43, 61,
100, 116, 119, 130–31,
162–63, 167–68, 181,
188, 203–14
Energy gap, 8, 176
Energy storage, 188
Equivalent circuit (to a LED),
13–4
Equivalent series resistance
(ESR), 17–8, 184
Continuous conduction mode
(CCM), 42, 62–3, 67,
99–100
Copper losses, 189, 200–02
Core losses, 147, 188, 195
Creepage distance, 216, 226
Cuk converter, 100–30, 169
Current limiter, 20–3
Current mirror, 24–7, 179–80
Current sense, 20–1, 44–5,
75, 163
Current sink, 22–3, 35–6
Current source, 22–5, 35–6
Delay, 104–08
Depletion MOSFET, 22–4
Detecting failures, 23–4, 28–9
Dimming ratio, 66, 194
Diodes, 44–6, 50–1, 71, 165–66,
181–82
Discontinuous conduction
mode (DCM), 62–3,
84–6, 171
Dissipation factor, 184
Double buck, 55–8
Dummy load, 29–30
Duty cycle, 42, 69, 87, 153–56,
144, 146
E-core, 198–99
Efficiency, 33, 217–18
Electrolytic capacitor, 182–83
Electromagnetic compatibility
(EMC), 203–04, 214–16
Failure detection, 28–9, 65–6
Fans, 222–23
Feedback, 65
Ferrite, 147, 157–58, 197
Filter, 55, 167–68, 208
Flyback, 139, 149–60, 170–71,
189–90
Forward voltage drop,
14–5, 17–8
Gain bandwidth, 194
Gas discharge, 215
Gate charge, 71, 74, 78, 178–79
Gate drive, 51, 66–7, 71, 74,
177–78, 211
Harmonics, 120, 145–46,
204, 206
Heat, 217–23
Heatsink, 220–23
Hysteretic controller, 59–60,
101–08
www
...
com
232
Index
Inductor, 43, 49–50, 187–90
Input offset voltage, 194
Inrush current, 48, 167, 192
Iron dust core, 197–98
Isolation, 225, 228
Leading edge blanking, 166
LED equivalent circuit, 13–4
Light flux, 12
Linear dimming, 40–1, 66
Linear regulator, 22–4, 33–6
Litz wire, 137–38, 189, 197
Loss tangent, 184
Low voltage, 227–28
Magnetising losses, 188–89,
199–200
Metal film, 190
Molypermalloy powder (MPP)
core, 198
Mood lighting, 12
MOSFET, 20–2, 44, 50–1,
164–65, 177–79
NTC thermistor, 48, 167,
172–73
N-type, 7, 175–77
Open circuit protection, 28,
78–9, 125–26
Operational amplifier, 193–94
Opto-coupler, 149
Oscillator frequency, 47, 64–5,
69, 106
Over voltage protection, 27–8,
78–9, 125–26
Overshoot, 152
Parallel LEDs, 20
Parasitic elements, 164–67
Passive current control, 18–20
Peak current control, 40, 101
Phase dimmer, 52–4
Piezo electric effect, 186
Plastic film, 183–87
P-N Junction, 7–8, 164–65
Polycarbonate, 183–87
Polyester, 183–87, 227
Polypropylene, 73, 183–87, 227
Polystyrene, 186
Pot-core, 198
Power factor, 141, 169–70
Power factor correction (PFC),
132, 141–47, 169–70
Power loss, 217–23
Printed circuit board (PCB),
164, 191–93, 208–10
P-type, 7, 175–77
Pulse width modulation
(PWM), 9
PWM dimming, 40, 62, 66–7,
112–13, 199–20
Recovery time, 165–66, 181
Resistor, 190–91
Resonant frequency, 185
Ripple current, 71–3, 88–90,
116, 163–64
Safety, 27, 168, 216, 225–28
Saturation current, 147, 157,
188–89
Safety extra low voltage
(SELV), 168
Screen, 213–14
Schottky diode, 45–6, 119, 181
Single ended primary inductance converter (SEPIC),
131–39, 169
Self-resonance, 185
Semiconductor, 175–77
Series LEDs, 20
Short circuit protection, 23
Skin effect, 189, 197, 201–02
www
...
com
Slope compensation, 76–7
Snubber, 155–56
Soft-start, 173–74
Soldering, 193
Stability, 62, 103, 159
Step up and down, 139, 169
Step-down, 39–40, 163–68
Step-up, 61–98, 168
Streetlights, 11
Surface mount, 185–87,
192, 221
Switching frequency, 42,
64–5, 69, 106, 113–14,
197–99
Synchronization, 65
Temperature, 178, 191–93,
218–20
Temperature coefficient, 193,
200–01
Testing LED drivers, 29–30
Thermal resistance, 218–20
Thermistor, 48, 167, 172–73
Through-hole, 192
Toroidal, 187, 199
Traffic lights, 11, 217
Transformer, 189–90
Transorb suppressors, 182
Ultra-fast diodes, 181
Voltage dependent resistor
(VDR), 182, 187, 215
Voltage drop, 14–5, 17–8
Voltage limiting, 27–8
Voltage regulator, 22–4, 33–6
Voltage source, 17–24
Wire-wound, 190
Zener diodes, 13–4, 29–30,
36, 182
Author Biography
Steve Winder is now a European Field Applications Engineer for Supertex Inc
...
A large part of his time is spent helping customers with LED driving
applications
...
Here he designed analogue circuits for wideband
transmission systems, mostly high frequency, and designed many active and passive
filters
...
He is a Chartered Electrical Engineer
and a member of the Institute of Engineering and Technology (IET), based in
London, England
...
newnespress
Title: Power Supplies for LED Driving
Description: The essential guide to designing cost- and energy-efficient power systems for LED devices.
Description: The essential guide to designing cost- and energy-efficient power systems for LED devices.