Powering (driving) these LED's is not always simple. Linear driving is inefficient and generates far too much heat. With a switching supply, the main issues are EMI and efficiency, and of course cost. The problem is to get a design that meets legal requirements and is efficient, while costing the least. This book covers the design trade-offs involved in LED driving applications, from low-power to UB-LEDs and beyond.
* Practical, "hands-on" approach to power supply design for LED drivers
* Detailed examples of what works and why, throughout the design process
* Commentary on how the calculated component value compares with the actual value used, including a description of why the choice was made
|Product dimensions:||7.50(w) x 9.25(h) x (d)|
About the Author
Prior to joining Intersil Inc., Steve worked for US based Supertex Inc. in 2002, where he was instrumental in encouraging Supertex’s management to start developing LED drivers. One of Steve’s German customers had started using a relay driver for LEDs and once Steve had explained the technical detail of this application to Supertex’s management, they decided to start an applications team to develop LED specific products. Supertex then invested heavily to became a leader in this field. Microchip acquired Supertex in 2014.
Until 2002, Steve was for many years a team leader at British Telecom Research Laboratories, based at Martlesham Heath, Ipswich in the UK. Here he designed analog circuits for wideband transmission systems, mostly high frequency, and designed many active and passive filters.
Steve has studied electronics and related topics since 1973, receiving an Ordinary National Certificate (ONC) in 1975 and Higher National Certificate (HNC) in 1977 with Endorsements in 1978. He studied Mathematics and Physics part time with the Open University for 10 years, receiving a Bachelor of Arts Degree with 1st Class Honours in 1989. He received a Master’s Degree in 1991, in Telecommunications and Information Systems after studying at Essex University. Since 1991, he has continued with self-study of electronics, to keep up-to-date with new innovations and developments.
Read an Excerpt
Power Supplies for LED Driving
By Steve Winder
NewnesCopyright © 2008 Elsevier Inc.
All right reserved.
As a field applications engineer for one of the pioneering developers of integrated circuits for driving power LEDs, I meet many potential customers who have little or no idea of how to drive an LED properly. The older type of LED requiring a 20mA supply can be abused to some extent. However, power requirements have been increasing; current ratings of 30mA, 50mA, 100mA, 350 mA and higher are becoming common. There are several manufacturers that produce power levels up to 20W, and more; these higher powers use LED chip arrays. If a power LED is abused, it tends to die very quickly.
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. Names such as HB-LEDs (high bright) and UB-LEDs (ultra-bright) are becoming meaningless as the power levels continue to rise. This book will cover all types of LED drivers, from low power to UB-LEDs and beyond.
Is power LED driving simple? No, not usually. 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. Linear driving is inefficient and generates far too much heat. With a switching supply, the main issues are EMI and efficiency, and of course cost. The problem is to produce a design that meets legal requirements and is efficient, with minimal cost.
1.1 Objectives and General Approach
The approach of this book will be very practical, although some theory is introduced when necessary for understanding of later chapters. It is important to understand the characteristics of components before they can be used effectively.
In most chapters, I will include a section called 'Common Errors'. 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. It is said that people learn from their mistakes, but it is also true that we can learn from the mistakes of others. Our own mistakes are more memorable, but also more costly!
Usually the first problem for a designer is to choose between different topologies. When is a buck preferred to a buck-boost or a boost? Why is a Cuk boost-buck better than a fly-back type? This book will cover these topics at the beginning of the switching supplies section.
Power supply design equations will be given and example designs of practical supplies will be worked through. With switching power supplies, equations are needed to make the correct component choice; a wrong component can make a poor power supply and require a lot of corrective action. 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.
The implications of changing the calculated component values into standard values, which is more practical, will be discussed. In many cases, customers want to use standard off-the-shelf parts, because of ease of purchase and cost. Calculations rarely produce a standard value, so a compromise has to be made. In some cases the difference is negligible. In others it may be better to choose a higher (or lower) value. All component value changes will introduce some 'error' in the final result.
Having proven worked examples in the book will help the reader to understand the design process: the order in which the design progresses. 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.
1.2 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. It is also important to understand the characteristics of LEDs in order to understand how to drive them properly. One of the characteristics is colour; an LED emits a very narrow band of wavelengths so the colour is fairly pure. The LED color determines the different voltage drop across the LED while it is conducting, and I will show how that varies with the current level. But the current level determines the light output level: higher currents give higher luminosity from a given LED. The light output has the characteristic of intensity and the amount of beam spreading.
Chapter 3 will show that there are several ways to drive LEDs. Because most electronic circuits have traditionally been driven by a voltage source, it is natural for designers to continue this custom when driving an LED. The trouble is that this is not a good match for the LED power requirement. A constant current load needs a constant voltage source, but a constant voltage load (which is what an LED is) needs a constant current supply.
So, if we have a constant voltage supply, we need to have some form of current control in series with the LED. With a series resistor or active regulator circuit we are trying to create a constant current supply. 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. Detecting an LED failure is possible using a current monitoring circuit. This could also be used to detect an open circuit. Instead of having a constant voltage supply, followed by a current limiter, it seems sensible to just use a constant current supply! There are some merits of using both constant voltage supply and a current regulator, which will be described in Chapter 4.
Chapter 3 continues describing features of constant current circuit. If we have a constant current source, we may have to provide some voltage limiting arrangement, just in case the load is disconnected. Open circuit protection can take many forms. A failure (short) would make no difference to the current level, so voltage monitoring would be a preferred failure detection mechanism. If the circuit failed open the voltage would rise up to the level of the open circuit protection limit, which could also be detected.
Chapter 4 describes linear power supplies, which can be as simple as a voltage regulator configured for constant current. Advantages include no EMI generation, so no filtering is required. The main disadvantage is heat dissipation and the limitation of having to ensure that the load voltage is lower than the supply voltage; this leads to a further disadvantage of only allowing a limited supply voltage range.
Chapter 5 describes the most basic of switching LED drivers: the buck converter. The buck converter drives an output that has a lower voltage than the input; it is a step-down topology. The reader will be taken through the design process, followed by an example design.
Chapter 6 describes boost converters. These are used in many applications including LCD backlights for television, and computer and satellite navigation display screens. The boost converter drives an output that has a higher voltage than the input; it is a step-up topology. The reader will be taken through the design process, followed by an example design, for both continuous mode and discontinuous mode drivers.
Chapter 7 describes boost-buck converters. These have the ability to drive a load that is either higher or lower voltage compared to the input. However, this type of converter is less efficient than a simple buck or boost converter.
Chapter 8 describes specialist converters: buck-boost and buck (BBB), and Bi-Bred. These converters are intended for AC input applications, such as traffic lights, street lights and general lighting. 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. This extra functionality does come at a cost – the efficiency is much lower than a standard off-line buck converter.
Chapter 9 describes fly-back converters. This chapter describes simple switching circuits that can be used for constant voltage or constant current output. The use of two windings or more in an inductor permits isolation of the output. A single winding inductor is a non-isolated buck-boost circuit that is sometimes used for driving LEDs, although the Cuk and SEPIC generally produce less EMI (at the cost of an additional inductor).
Chapter 10 covers topics that are essential when considering a switch mode power supply. The most suitable topology for an application will be discussed. 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). Discussion will include snubber techniques for reducing EMI and improving efficiency, limiting switch-on surges using either in-rush current limiters or soft-start techniques.
Chapter 11 describes electronic components for power supplies. The best component is not always an obvious choice. There are so many different types of switching elements: MOSFETs, power bipolar transistors and diodes, each with characteristics that affect overall power supply performance. Current sensing can be achieved using resistors or transformers, but the type of resistor or transformer is important; similarly with the choice of capacitors and filter components.
Magnetic components are often a mystery for many electronic engineers, and these will be briefly described in Chapter 12. First, there are different materials: ferrite cores, iron dust cores and special material cores. Then there are different core shapes and sizes. One of the most important physical characteristics from a power supply design point of view is magnetisation and avoiding magnetic saturation.
EMI and EMC issues are the subjects of Chapter 13. It is a legally binding requirement that equipment should meet EMI standards. 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. Meeting EMC standards is also a legal requirement in many cases. It is no use having an otherwise excellent circuit that is destroyed by externally produced interference. In many areas, EMC practices are compatible with EMI practices.
Chapter 14 discusses thermal issues for both the LEDs and the LED driver. The LED driver has issues of efficiency and power loss. 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. Handling the heat by using cooling techniques is a largely mechanical process, using a metal heatsink and sometimes airflow to remove the heat energy. Calculating the temperature is important because there are operating temperature limits for all semiconductors.
Another legal requirement is safety, which is covered in Chapter 15. The product must not injure people when it is operating. This is related to the operating voltage and some designers try to keep below SELV (safety extra low voltage) limits for this reason. When the equipment is powered from the AC mains supply, the issues of isolation, circuit breakers and creepage distances must be considered.
Chapter TwoCharacteristics of LEDs
Most semiconductors are made by doping silicon with a material that creates free negative charge (N-type), or free positive charge (P-type). The fixed atoms have positive and negative charge, respectively. At the junction of these two materials, the free charges combine and this creates a narrow region devoid of free charge. This 'intrinsic region' now has the positive and negative charge of the fixed atoms, which opposes any further free charge combination. In effect, there is an energy barrier created; we have a diode junction.
In order for a P-N junction to conduct, we must make the P-type material more positive than the N-type. This forces more positive charge into the P-type material and more negative charge into the N-type material. Conduction takes place when (in silicon) there is about 0.7 V potential difference across the P-N junction. This potential difference gives electrons enough energy to conduct.
An LED is also made from a P-N junction, but silicon is unsuitable because the energy barrier is too low. The first LEDs were made from gallium arsenide (GaAs) and produced infrared light at about 905 nm. The reason for producing this color is the energy difference between the conduction band and the lowest energy level (valence band) in GaAs. When a voltage is applied across the LED, electrons are given enough energy to jump into the conduction band and current flows. When an electron loses energy and falls back into the low energy state (the valence band), a photon (light) is often emitted. See Figure 2.1.
Excerpted from Power Supplies for LED Driving by Steve Winder Copyright © 2008 by Elsevier Inc.. Excerpted by permission of Newnes. All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of ContentsIntroduction; Characteristics of LEDs; Driving LEDs; Linear Power supplies; Switching Power Supplies; Selecting electronic components for power supplies; Magnetic; EMI and EMC Issues; Thermal issues; Safety issues.