FEATURED LIGHTING SOLUTION SUPPLIERS
excelsys
TZL111.CA
LIGHTING SOLUTIONS
Digi-Key Corporation brings you TechZonesSM featuring suppliers,
products and resources for Lighting, Microcontroller, Power, Sensor,
and Wireless technologies, with more sites to come. Resources
include application notes, reference designs, white papers, links to
product training modules, and more!
Quick Way to Find Digi-Key’s TechZonesSM:
•
•
•
•
Links located on our homepage
Links located in our header under Resources
Links located in our toolbar
www.digikey.ca/techzones
LED Luminaire
Design Guide
contributed by Cree, Inc.
There is a huge market for high-power LEDs in
luminaires, but replacing incandescent in traditional
fixtures is not without design challenges. This article
provides guidelines for the process of designing highpower LEDs into luminaires. While it uses an indoor
luminaire as an example, the process can be applied to
the design of any LED luminaire.
46
2011 DIGI-KEY CATALOGUE
AVAILABLE NOW!
View our environmentally-friendly interactive
and PDF catalogues online or request your
annual Digi-Key print catalogue today!
www.digikey.ca/catalogue
www.digikey.ca/lighting
3
Table of Contents
Reliability and Lifetime of LEDs ................................................8
contributed by OSRAM Opto Semiconductors
Reliability and lifetime are different though related concepts. This article
details the numerous factors involved with each and the tradeoffs you’ll
need to consider to achieve optimal results.
Spectral Design Considerations
for White LED Color Rendering................................................16
by Dr. Yoshi Ohno, National Institute of Standards and Technology
The Color Rendering Index is unreliable for describing the color-rendering
performance of white LEDs as well as for conventional sources. An NIST
expert explains where the CRI falls short and why the industry needs a
new, improved metric.
Color Rendering of Light Sources ...........................................24
by Dr. Wendy Davis, National Institute of Standards and Technology
NIST is proposing a Color Quality Scale (CQS) that is designed to address
many of the shortcomings of the widely used Color Rendering Index with
regard to solid-state lighting.
Radiant and Luminous Flux.....................................................26
by William F. Long, Ph.D., O.D., Associate Professor Emeritus,
University of Missouri-St. Louis
Because the sensitivity of the eye varies considerably with the actual
wavelength of visible light, the units quantifying light energy in vision
science are different from those used in physics. A physicist explains
how the eye responds to the power spectra of different light sources.
A Halogen Desk Lamp Conversion to LEDs.............................28
by Jim Young and Bernie Weir, ON Semiconductor
Phosphor converted “white” LEDs are highly efficient and mercury
free. However, using them in standard fixtures can be a challenge. This
article demonstrates the use of an off-the-shelf halogen desk lamp to
demonstrate the real world performance of today’s latest production LED
light sources.
Compensating and Measuring the
Control Loop of a High-Power LED Driver ..............................43
by Jeff Falin, Texas Instruments
Measuring the control loop of a WLED current regulating boost converter
with traditional methods is cumbersome at best. This article presents
a simplified, small-signal control-loop model for a boost converter with
current-mode control and explains how to measure the converter’s
control loop.
Phosphor Film Conversion for White LEDs .............................53
by Bit Tie Chan, Avago Technologies
The color consistency of white LEDS may be enhanced by employing
a film where the phosphor material is uniformly incorporated in
GaInN- based white LEDs. The color consistency improvement of
phosphor film conversion is 50% compared with conventional phosphorbased white LEDs.
Drivers for HB LEDs Allow Designers to Build
Halogen and Incandescent Lamp Replacements ...................56
by Piero Bianco, Maxim Integrated Products Inc.
Designing HB LED replacements for popular halogen and incandescent
lamps such as the MR16, PAR20 and A19 presents physical, electronic
and thermal design challenges. An active PFC solution can provide
excellent dimmability with no flicker of the lamp, while providing more
than good enough power factor.
LED Drivers for Automotive Applications ...............................60
contributed by ROHM Semiconductor
From headlights to taillights and plenty of places in between, LEDs
have come to dominate automotive lighting applications. Automotive
applications have tough requirements that are reflected in industry
standards and purchasing specifications. Highly integrated driver
solutions provide a way to meet them.
High-Power LED Drivers Find Their Niche ..............................32
Color Management of a Red, Green,
and Blue LED Combination Light Source ................................63
by Tony Armstrong, Linear Technology Corporation
In automotive headlights and industrial applications, high-brightness
LED arrays are finding wide acceptance. Efficiently driving these arrays
isn’t a straightforward task, but the driver manufacturers aren’t short
of solutions.
RGB LED lighting is an appealing lighting solution. However, the
variability of LED characteristics causes the RGB light to deviate from
the intended color. Avago Technologies’ feedback controller simplifies
the implementation of a tricolor optical feedback solution.
contributed by Avago Technologies
Effective Thermal
Management of LED Arrays ....................................................35
contributed by Bridgelux, Inc.
Although a critical design parameter, thermal management is
not as difficult as many would believe. Understanding the basics
allows every lighting designer to optimize their products and meet
specification requirements.
Copyrights: The masthead, logo, design, articles, content and format of TechZone is Copyright 2011, Digi-Key Corporation. All
rights are reserved. No portion of this publication may be reproduced in part or in whole without express permission, in writing,
from Digi-Key.
Trademarks: DIGI-KEY, the Digi-Key logo, TECHZONE, and the TechZone logo are trademarks of Digi-Key Corporation. All other
trademarks, service marks or product names are the property of their respective holders.
All product names, descriptions, specifications, prices and other information are subject to change without notice. While
the information contained in this magazine is believed to be accurate, Digi-Key takes no responsibility for incorrect, false or
misleading information, errors or omissions. Your use of the information in this magazine is at your own risk. Some portions of
the magazine may offer information regarding a particular design or application of a product from a variety of sources; such
information is intended only as a starting point for further investigation by you as to its suitability and availability for your particular
circumstances and should not be relied upon in the absence of your own independent investigation and review. Everything in
this magazine is provided to you “AS IS.” Digi-Key expressly disclaim any express or implied warranty, including any warranty of
fitness for a particular purpose or non-infringement. Digi-Key cannot guarantee and does not promise any specific results from
use of any information contained in this magazine. Any comments may be addressed to techzone@digikey.com.
4
Editorial Comment
Recently, I had the opportunity to sit down with Mark Despotes,
Cree’s vice president of global channel sales, to discuss the trends
that are shaping the future of the Lighting industry. It is a very
dynamic time and all signals point towards rapid expansion. Mark
has a unique insight into what lies ahead for the industry; the
following are excerpts from our discussion:
“An exciting dynamic is that the LED lighting market is
still in its early stages. As an example, there are over 200
million streetlights installed worldwide and only a very small
percentage now use LED technology.
About Digi-Key Corporation
Mark Zack
Lighting is not a singular offering—its hundreds, even
Director, Semiconductor
thousands, of specific applications—each with unique needs
and considerations. That’s why you’ve seen Cree develop application-targeted products to
best enable our lighting customers to be successful.”
In this issue of Digi-Key’s TechZone™ Magazine, you will find a number of articles that appeal
to lighting design engineers including:
• Cree contributes an article on guidelines for the process of designing high-power LEDs
into Luminaires (page 46).
• OSRAM Opto Semiconductors contributes an article on Reliability and Lifetime of LEDs
(page 8).
• The National Institute of Standards and Technology contributes two articles on Color
Rendering (pages 16 and 24).
It’s easy to be enthusiastic about the growth opportunities in the Lighting industry after listening
Despotes’ closing comment, “no single application will dominate as the market expands.”
As one of the world’s fastest growing
distributors of electronic components,
Digi-Key Corporation has earned
its reputation as an industry leader
through its total commitment to service
and performance. As a full-service
provider of both prototype/design and
production quantities of electronic
components, Digi-Key has been ranked
#1 for Overall Performance for 18
consecutive years from among the
nation’s more than 200 distributors (EE
Times Distribution Study/August 2009).
Offering more than 1.8 million products
from more than 440 quality name-brand
manufacturers, Digi-Key’s commitment
to inventory is unparalleled. Access to
the company’s broad product offering
is available 24/7 at Digi-Key’s
top-rated website. www.digikey.ca
As one of the world’s leading electronic component distributors, Digi-Key is here for all of your
design support and production needs. We believe that the information and insights in this
issue of TechZone™ will be a linchpin in your very successful 2011.
Sincerely,
Mark Zack
Director, Semiconductor
Digi-Key Corporation
www.digikey.ca/lighting
5
Lighting TechZoneSM Q & A
In the rapidly evolving lighting market, it seems as if there is something new every day – technologies, products,
consumer trends, or regulations. Time-to-market demands have never been greater with the design of many of today’s
new lighting products requiring expertise in more than one discipline.
You have questions. We have answers. On target to field more than 260,000 calls this year,
our technical support specialists are available 24/7/365 to answer your questions and assist
you with your lighting needs.
If you have a question, we invite you to contact our technical staff via telephone, live web
chat, or by emailing your question to techzone@digikey.com.
Does a high CRI value guarantee that all colors will be
rendered well?
No. CRI only measures the shift in chromaticity (color) of eight
standard color test samples produced by a light source as compared
to the chromaticity values under a standard reference illuminant of
the same correlated color temperature (CCT). The reference illuminant
for CCT values below 5000K is the blackbody radiator and the CIE
Daylight source for CCT values above 5000K. However, since the CCT
values of the reference illuminant and the test source are matched
(i.e. a 3000 K reference to a 3000K test light source), the CRI score
can be very high even though the actual color rendering is degraded
at the far ends of the CCT range.
For example at lower color temperatures of 2700K to 3000K which is
typical of incandescent lamps, the majority of the emitted light is in
the red and yellow portions of the spectrum with only a small amount
in the blue spectrum. This can produce significant color shifts in
objects as compared to light sources with higher color temperatures
where the emitted light spectrum is more balanced. Therefore,
incandescent lamps have CRI values of 100 not because they render
all colors well, but rather, because they render colors in a similar
(degraded) manner as the standard reference (blackbody radiator) at
the lower color temperatures.
What is UL 8750?
UL 8750 is an Underwriters Laboratories (UL) standard that specifies the
minimum safety requirements for the components that are an integral
part of a LED luminaire. This includes discrete LEDs, LED modules,
constant current drivers, control circuitry, and power supplies.
What is an integrating sphere?
An integrating sphere is a hollow sphere that collects and spatially
integrates radiant flux through multiple internal diffuse reflections to
produce a uniform light density distribution. The spatially integrated
radiant flux can then be measured through an opening in the sphere
with a photometer, colorimeter, or spectroradiomter.
6
What is a photometer?
A photometer is an instrument that measures the luminous flux
incident on a surface per unit area (lumens/m2). It consists of a
silicon photodiode and a glass filter that adjusts the detector’s
spectral response to match the CIE photopic luminosity function
(human eye response). With some additional equipment, it is also
used to measure total luminous flux (photometer + integrating
sphere), luminous intensity distribution (photometer + goniometry),
and luminance (photometer + special optics).
What is a tristimulus colorimeter?
A tristimulus colorimeter is an instrument that uses optical filters to
match the spectral response of silicon photodetectors as close as
possible to the tristimulus color response of the human eye as defined
by the CIE color matching functions. The recorded tristimulus values
can then be used to calculate chromaticity coordinates and represent
the color in a CIE color space.
What is a spectroradiometer?
A spectroradiometer is an instrument that measures the
radiant power of a light source at each wavelength (Watt/nm).
The recorded spectral power distribution can then be used to
accurately compute photometric, colorimetric, and color rendering
properties of a light source.
Are spectroradiometers more accurate than filter-based
photometers and colorimeters?
Yes, spectroradiometers are more accurate. Currently, the IES
LM-79 standard recommends the use of spectroradiometers as
opposed to photometers and colorimeters because the spectral
filters used to approximate the response of the human eye in
these devices can cause significant measurement errors for solid
state light sources.
How is the spectral power distribution of a light source
related to the perceived color of illuminated objects?
Light reaching the eye is usually radiation that has been reflected by
a surface or transmitted through a layer of material. Each substance
modifies the emitted spectral power distribution of a light source
by absorbing, reflecting, and transmitting light differently at each
wavelength. As a result, the perceived color of an object depends
on the emission spectrum of the light source and the object’s
spectral characteristics (reflected, absorbed, and transmitted radiant
power at each wavelength). Although the spectral characteristics of
substances are relatively stable, the spectral power distributions of
different light sources are highly variable. This can cause significant
perceived color shifts in illuminated objects as compared to their
perceived color under a reference light source.
What is a goniophotometer?
A goniophotometer is an instrument that is typically used to measure
the luminous intensity distribution of a light source as a function
of angular position. It consists of a computer operated mechanical
positioning system and a photometer which records the spatial
intensity distribution of a light source.
What are Type C goniophotometers?
Type C goniophotometers (also known as moving mirror
goniophotometers) use a large mirror mounted on a rotating arm
to reflect light on to a photodetector from a stationary test source.
This configuration allows the light source to be mounted in a fixed
position similar to its intended application. In contrast, Type A
and Type B goniophotometers tilt the light source relative to the
detector to measure the angular intensity distribution.
Technical and Design Support Services
Digi-Key offers live technical support 24/7 via telephone, e-mail
and live web chat. Digi-Key’s 130 technicians on staff are
trained by manufacturers to answer product-specific questions.
Additionally, these technicians cross-reference part numbers,
assist customers in choosing products, research and aid in
selecting new product, and provide access to in-depth productspecific information as well as specifications and performance
data on new products.
Digi-Key’s Design Support Services (DSS) team of application
engineers and technicians provides general information and
complimentary project-specific assistance. DSS provides service
to engineers ranging from one-time contacts regarding product
recommendations to ongoing prototype-to-production design
support. DSS strives to guide the customer through the design
process while achieving the best solutions and, ultimately,
streamlining the design cycle. The DSS team provides support
and advice on system design, aids with product selection and
development tools, and provides assistance with other applicable
design issues. Additionally, members of the DSS team produce
application notes, webinars and instructional videos. The DSS
team is available from 8:30 a.m.-5:00 p.m. CST via telephone,
e-mail and web-conferencing software.
Are Type C goniophotometers required for LM-79 testing?
Yes, Type C goniophotometers are required for LM-79 testing
because the light source is installed and maintained in the same
orientation as it would be in the field. This is important for LED
luminaries because their preference characteristics are sensitive
to changes in heat dissipation.
Do you have a question about lighting solutions?
Digi-Key has more than 130 technical support specialists, product managers, and applications engineers who are eager to
answer your questions and assist you with your lighting projects.
Send your questions to techzone@digikey.com.
www.digikey.ca/lighting
7
Reliability and
Lifetime of LEDs
contributed by OSRAM Opto Semiconductors
With the increasing complexity of technical
equipment, modules or even individual
components, reliability and lifetime and
the costs involved with exchange and
revision become increasingly important for
the customer. Here, one must consider an
optimization between requirements, functions
and costs over the lifetime of the product.
Introduction
The single requirement that an LED will not fail is no longer sufficient
for modern, powerful components or devices. More often, it is
additionally expected that they perform their required functions
without failure. However, it is only possible to make a prognosis
(probability) supported by statistics and trials as to what extent such
requirements can be fulfilled. A direct answer or statement as to
whether an individual device or component will operate without failure
for a certain period of time cannot be given.
Now days, modern methods of quality management and
reliability modeling are used in order to investigate and verify these
types of questions.
This article provides a fundamental insight into the matters of
“reliability” and “lifetime.” The terms lifetime and reliability are
explained in further detail with respect to light emitting diodes (LEDs)
and how these terms are understood by OSRAM Opto Semiconductors.
In addition, important factors that influence the lifetime and reliability
of LEDs are described.
The Appendix provides descriptions of the mathematical foundations
that are needed in practice.
The concept of reliability at OSRAM Opto Semiconductors
Zero Tolerance to Defects (ZTTD) is a rigid part of the corporate
culture at OSRAM Opto Semiconductors. Only in this way is it
possible for our customers to also aim for zero defects in their
production and applications.
OSRAM Opto Semiconductors associates the term reliability with the
fulfillment of customer expectations over the expected lifetime. In
8
material
function
process
Figure 1: Basis for reliability of LEDs
other words, the LED does not fail during its lifetime under the given
environmental and functional conditions.
The reliability of the products is thus based on the chain of the
materials, the manufacturing process, and the function of the
component (Figure 1). In addition, the final application must
also be taken into consideration. High reliability can only be
achieved if the changing effects and interdependencies of the
individual components are already taken into account during the
development phase.
Neglecting this entirely or only focusing on one or two elements
leads to a reduction in the quality of the product and thus to a
decrease in reliability.
Reliability of LEDs
The reliability of a semiconductor element is the property that states
how reliably a function assigned to the product is fulfilled within a
period of time. It is subject to a stochastic process and is described by
the probability of survival R(t).
A fault or failure is indicated if the component can no longer fulfill the
functionality assigned to it.
Failures and failure rates are subdivided into three phases:
1. Early failures
2. Random or spontaneous failures
3. Wearout period
In reliability mathematics, this failure period is described by an
exponential distribution. An exponential distribution is based on a
constant failure rate over time. The average failure rate is given in FIT
(Failure unITs).
1: Early failure period
2: Spontanous failure period
3: Wearout period
As a rule, an experimental determination of the middle failure rate is
extremely difficult. For this reason, OSRAM OS uses the SN 29500
standard from Siemens AG, which incorporates the experience of
failures in the field into the typical failure rates for LEDs (Figure 3).
In the process, no distinction is made in regard to the cause of the
individual failures.
Figure 2: Failure rate over time (“bathtub” curve)
Since the failure rate is especially high at the beginning and end
of the product cycle, the failure rate over time takes the form of a
“bathtub” curve (Figure 2).
Thereby, each single failure mechanism exhibits its own chronological
progression and shows therefore an individual “bathtub” curve.
For each of these phases, many different types of definitions, analysis
methods, and mathematical formulas for reliability can be found in
the literature. The most important definitions and methods that apply
to LEDs are described in this section and in the Appendix.
For the sake of simplicity, the first two phases are combined
into a so-called “extrinsic reliability period.” The third phase, the
wearout period, is correspondingly designated as the “intrinsic
reliability period.”
12 –
10 –
8–
6–
4–
2–
0–
For LEDs, the most significant degradation parameters are the
changes in brightness or color coordinates. Other parameters such as
forward voltage generally play a subordinate role.
During operation, LEDs experience a gradual decrease in luminous
flux, measured in lumens. As a rule, this is accelerated by the
operating current and temperature of the LED and also appears when
the LED is driven within specifications.
The term “lumen maintenance” (L) is used in connection with the
degradation of light in LEDs. This describes the remaining luminous
flux over time with respect to the original luminous flux of the LED.
Due to continuous degradation, a failure criterion must be established
in order to obtain a concrete evaluation of the LED failure. The point
in time at which the luminous flux of the LED reaches the failure
criterion is then described as the failure time or lifetime of the LED.
As a rule, the failure criterion is determined by the application.
Typical values are 50 percent (L50) or 70 percent (L70), depending
on the general illumination.
110
Relative light output [%]
SN 29500-12 (2008)
for large power packages
100
90
80
70
60
L70
50
40
30
20
10
Time
–
8000
-
–
6000
-
–
4000
-
–
2000
-
–
0
-
0
–
Random failure rate λ [FIT]
Extrinsic reliability period
Extrinsic failures (early and spontaneous failures) are generated by
defective materials, deviations in the manufacturing process or by
incorrect handling and operation by the customer.
Intrinsic reliability period
The intrinsic reliability period describes the so-called wearout period
of the component at the end of the product cycle. It is based on
increased wear and aging of the material. This continuous change
over time is generally measurable and is referred to as degradation.
10000
Time [h]
Figure 3: LED failure rate in the extrinsic period according to Siemens Standard SN 29500
More than 99 percent of these extrinsic failures can be observed during
installation of the parts in the application (e.g. by soldering) or in the first
hours of operation. In contrast, between the early failure period and the
wearout period, the spontaneous failure rate for LEDs is extremely low.
Figure 4: Degradation curve
Since aging is based on a change in the material properties and is
therefore subject to statistical processes, the lifetime values also are
based on a statistical distribution.
The percentage of components that have failed is described by the
term “mortality” (B). A value of B50 thus describes the point in time at
which 50 percent of the components have failed. This value is generally
www.digikey.ca/lighting
9
–
40k —
—
—
–
—
–
–
Tj
Ts
—
–
–
20k —
—
If = 1.5A
80
100
120
—
–
—
–
–
–
60
0,1%
–
—
–
—
—
0—
1%
—
140
Temperature [°C]
100ppm
10ppm
Figure 7: Dependence of lifetime on the junction temperature and solder point temperature
1ppm
t10 tml
Time
Figure 5: Distribution of the probability of failure over the lifetime
For thermo-mechanical stress on a component (e.g. temperature
cycles), the continuous aging process generally cannot be
measured. This means that the constant aging process that
leads to failure cannot be described by means of a characteristic
measurement parameter such as light degradation during electrical
operation. An extrapolation of the degradation curve to a defined
failure criterion as is shown in Figure 4 is not possible here. In this
case, in order to be able to make statements about the time of
failure or the failure distribution, tests must be carried out until the
most abrupt failures occur. An example of this is fatigue in adhesive
or bonded connections.
Influencing factors with respect to reliability and lifetime
Similar to conventional lights, the reliability and lifetime of LED light
sources is also dependent on various factors or can be influenced by
these factors.
The most important physical influencing factors include humidity,
temperature, current and voltage, mechanical forces, chemicals, and
light radiation (Figure 6).
These can directly lead to total failure or influence the aging
characteristics in the long term (e.g. brightness) and thus produce a
change in the reliability and lifetime.
humidity
temperature
chemicals
LED
While such direct influencing factors are the temperature and
resulting junction temperature Tj of the LED, for example, but the
amount of current used to drive the LED is also an influencing factor.
Under otherwise equal operating conditions, an increase in the
ambient temperature as well as an increase in current produces an
increase in the junction temperature. In general, however, an increase
in junction temperature brings about a decrease in lifetime.
Another direct influencing factor is mechanical force. If large
mechanical forces are applied to the LED, for example, this generally
results in damage that can additionally lead to total failure of the LED.
The sources of the individual factors can be found in different areas
such as LED design, LED processing, the customer application,
and the environment and from there, can be traced back to various
aspects and parameters (Figure 8).
If these four areas are examined in more detail, it can be
determined that three of the four areas can be directly influenced by
the LED manufacturer or the user. The last area, the environment,
ultimately cannot be changed and must be considered as a given in
the application.
For example, the source of the influencing factor, temperature, can be
assigned to two areas: LED design and the customer application.
In the area of LED design, the source of the temperature influence lies
both with the electrical parameters and the transfer of heat.
Depending on the current applied (IF) and the associated voltage (UF),
a power dissipation is created, which to a large extent, is converted
LED design
Customer application
• electrical parameters
• heat transport
• ESD layout
• package
• heat transport
• PCB layout and material
• ESD protection
• electrical circuit
Reliability
light
mechanical forces
current and voltage
Figure 6: Influencing factors on reliability and lifetime
10
60k —
—
B10
–
–
10%
B50
—
–
—
90%
Diamond Dragon
–
Lifetime B50/L70 [h]
Cumulated failure distribution
99%
50%
80k —
—
specified as typical median lifetime, t50 or tml, for LEDs. In addition to
the median value (B50), a value can also be specified when 10 percent
of the components have failed (B10 value). This allows one to draw a
conclusion about the width of the lifetime distribution (Figure 5).
Reliability
Assembly
• storage condition
• soldering process
• handling (pick & place;
ESD)
Figure 8: Sources of influencing factors
Environment
• ambient temperature
• temperature cycles
• humidity
• pollution
• light radiation
—
–
∆Tj-s
60k —
Low Rth
—
Tj
–
—
Tj
–
—
–
—
–
Ts
Ts
High Rth
—
–
–
—
40k —
∆Tj-s
–
–
—
60
80
100
120
—
–
—
–
—
—
–
–
0—
If = 1,5A
—
–
–
20k —
—
Lifetime B50/L70 [h]
–
—
80k —
140
Temperature [°C]
In the area of customer applications, the influencing factor of
temperature can be traced back to heat dissipation. Here, the layout
and material of the circuit board play an important role. Summarized
under the term “thermal management,” which among other things,
includes the selection of an appropriate circuit board material (e.g.
FR4, IMS), the layout of LEDs, the component density, additional
cooling, etc., the user also has the opportunity to specifically target
his application to accommodate for the influencing factors.
The following measures can be taken:
• Optimal thermal board management
• Optimal design for efficient use of the LED
• Handling the LED according to specifications
Figure 9: The dependency of lifetime on temperature due to the influence of various Rth
values (example)
into heat. This leads to an increase in temperature in the junction of
the LED. The amount of power dissipation is proportional to changes
in the junction temperature.
The proportionality factor is the thermal resistance of the housing
(Rth, Junction-Solderpoint) of the LED. This reflects the heat transfer
characteristics of the LED.
The lower the thermal resistance of the LED, the better the thermal
properties of the LED become. If heat is dissipated quickly and
efficiently, the junction temperature does not increase as rapidly.
As an example, two components with differing Rth values (2.5 and 8
K/W) are examined at the same solder point temperature TS = 90°C
and the same operating conditions (current) (Figure 9).
The junction temperature of the component with low thermal
resistance only increases to ~104°C. In contrast, however, the
component with the higher thermal resistance exhibits a junction
temperature of >135°C.
As mentioned previously, the lifetime of an LED is reduced with an
increase in the junction temperature.
At the same solder point temperature, the component with the lower
Rth achieves a longer lifetime than the component with the higher Rth.
In addition to an increased lifetime, lower thermal resistance offers an
additional advantage: At the same solder temperature, a component
with a low Rth achieves a higher light output. The reason is the decrease
in efficiency of an LED with an increase in junction temperature.
For the LED manufacturer, the influencing factors that have a
significant influence on lifetime and reliability can already be taken
into consideration in the development phase.
The effects of these factors can be reduced through the
following measures:
• Robust design
• Optimal thermal management
• Stable and optimized production processes in order to minimize
the risk of spontaneous failure
• Customer support for including LED designs in the customer
application
• Considering the strengths and weaknesses of LEDs
An insufficient thermal management directly leads to a reduction of
the reliability and lifetime of the LED (see application note “Thermal
Resistance of LEDs” by OSRAM OS). This application note provides an
exact description of how the thermal resistance is determined for the
individual packaging types at OSRAM OS.
In general, it can be ascertained, however, that in spite of the high
reliability of OSRAM OS LEDs, only through the consideration of all
areas and all changing effects and dependencies can a high overall or
system reliability be achieved.
Validation and confirmation of reliability and lifetime
All LED packages and chip families from OSRAM OS undergo a
number of tests for validation and confirmation of reliability and
lifetime. The selection of tests, test conditions, and duration occurs by
means of an internal OSRAM OS qualification specification based on
JEDED, MIL, and IEC standards. In addition, the requirements profile of
the component is also included.
Table 1 shows the list of typically performed tests. In addition, the
various test conditions, the test duration and the stress factors
involved are listed.
Based on the internal OSRAM OS qualification specification and the
requirements profile, the selection of the test, the test conditions, and
test duration can be set.
The mechanical stability of an LED is checked by means of a solder
heat resistance test as well as powered and unpowered temperature
cycle tests. Here, the cycle count and the temperature difference
serve as measures of stability. These types of tests are also drawn
upon to evaluate the failure rate.
For proof of reliability, the LEDs undergo individual tests of up to 1000
hours in duration. If the properties and interactions of the integral
parts of the LED are known, results can be taken from already tested
products and applied to other types of LEDs with the same material
characteristics. As a result, the general test scope is reduced, since
fewer products must be tested. This allows the test duration of
individual products to be increased to a longer period. At OSRAM OS,
tests sequences are carried out for up to 10,000 hours, for example,
in order to investigate general effects. Individual technology platforms
are even evaluated for more than 35,000 hours.
www.digikey.ca/lighting
11
TEST
CONDITIONS
Resistance to soldering heat (RTSH)
DURATION
STRESS FACTORS
Convection soldering
260°C/10 sec
3 runs
Temperature, chemicals,
mechanical forces
Wave soldering
260°C/10 sec
3 runs
Temperature, chemicals,
mechanical forces
T = 85°C
R.H. = 85%
IF = 5 mA/10 mA
1000 h
Temperature, humidity
-40°C/+100°C
15 min at extreme temps
300/500/1000 cycles
Mechanical forces
-40/+85°C
IF = [max derating]
ton/off = 5 min
1000 h
Temperature, current,
mechanical forces
1000 h
Temperature, current
JESD22-A108
T = 25°C
IF = [max derating]
1000 h
Temperature, current
JESD22-A108
T = 85°C
IF = [max derating]
1000 h
Temperature, current
JESD22-A108
T = 25°C
IF = [max derating]
Human body model
2000 V
1 pulse per polarity direction
Voltage
JESD22-A114
JESD22-A113
Resistance to soldering heat (RTSH)
JESD22-A106
Temperature & humidity bias (T&HB)
JESD22-A101
Temperature cycle (TC)
JESD22-A104
Power temp. cycling (PTC)
JESD22-A105
Steady state life test (SSLT)
Steady state life test (SSLT)
Pulsed life test (PLT)
ESD
Table 1: Example reliability test matrix for OSRAM OS LED
These types of selective, extremely longterm investigations provide a solid
basis for calculation of lifetime. According to OSRAM OS, however, the
resulting test data should not be “blindly” extrapolated to determine the
average lifetime. Rather, this data should make it possible to understand
why the different materials used behave the way they do.
Reliability and failure probability
Reliability R(t) states the probability P, that a system or individual
component remains functional during a timeframe t under normal
operating and environmental conditions. In complementary terms, one
speaks of the probability of failure F(t) or unreliability.
This allows a highly reliable extrapolation or prediction of the product
performance characteristics to be made, which is confirmed by a
small deviation from target values.
Thus, if n components are driven under the same conditions and the
number of failures is r(t) at time t, then the following applies:
Statements about lifetime that are based on mathematical results
without test data and background knowledge should generally be
viewed with caution.
On-site support from OSRAM OS regarding reliability and lifetime
OSRAM Opto Semiconductors supports its customers worldwide. This
support begins in the presales phase. OSRAM Opto Semiconductors
offers its customers assistance with the selection of appropriate
light sources and advises them in implementing optimally executed
applications. In addition, OSRAM Opto Semiconductors provides
technical expertise regarding quality and reliability.
APPENDIX
Fundamentals – definition of terms
In the following, the most important and relevant terms and definitions
from the areas of quality management and statistics are presented, as
well as an example for a reliability distribution.
12
( )
( )
( )
( )
At the beginning,
( )all components
) ( (time )t=0) and at
(
) function(properly
some point, they all are defective. That is,
(
( ))
( ) ( (
))
( )
( ) of failure
( (F(t)))of a component
( ) starts
This means that the probability
( )
( )
( )
at 0 (0%) and increases
( ) to 1 (100%) over time – an inverse relation
( )
( )
to reliability.
( )
( )
Probability of(failure
) density
(Failure Density) ( )( )
( )
( )( )
The failure density f(t) states the probability of a failure at a time t,
with respect to a time interval dt.
( it represents
)
( ( )of the) probability of failure.
Mathematically,
the derivation
( )
( )
( )
( )
( )
( )
( )( )
(( ))
( )
(
( )
( )
( )
( )( )
)
( )
( )
( )
Probability
1
Phase II with constant failure rate
This phase corresponds to the actual period of economic usefulness.
In this phase, the failure rate is constant and can be described with
an exponential or Poisson distribution. Here, failures mostly appear
suddenly and purely at random.
F(t)
0.5
( )
0
R(t) ( )
t
( )
Phase III - Wearout failures
In this phase, the failure rate increases at a faster rate due to aging,
wearout, fatigue, etc. with continuous operation. This phase also can
be described by a Weibull distribution.
Figure 10: Relation between reliability R(t) and probability of failure F(t)
(
( ))
(
)
Failure rate ( )
( )
The failure rate λ(t) is an important
( )indicator
( for the )reliability or lifetime
( )
( ) the probability of a failure within a time interval
of an object. It describes
)
( functional
) at time t.
dt, with respect(to the components
that are
( )
( )
( )
( )
( )
( )
)
( )
(
( )
The failure rate states how many units fail on average within a period
( ) are given(in)units of [1/time unit] such
of time. Usually, failure rates
(
)
as 1 failure per hour (-1/h).
rate of electronic
( )Due to the low
( failure
)
components, this is often stated as a FIT:
( )
( )
Component hours = Number of(components
)
( )
( )
* hours of operation
( )
In general, the failure rate is not constant. In many cases, the failure
( ) over the entire
rate usually follows(the
curve”
) so-called
( “bathtub
)
( )
) (Figure 11).
component life(cycle
Failure rate
∆(t)
( ) ( ) ( )( )
( )
Early failure due to
weaknesses in the material,
quality fluctuations, application
failures
( )
( )
Early failure period
( )
( )
( )
( )
( )
Economic life
Increase due to
aging, wearout, fatigue,
etc.
Spontanous failure period
The mathematical description is somewhat imprecise due to deviations
and test-related scattering. A reliable representation is therefore only
possible if a statistically large quantity of data has been obtained.
In practice, it can also happen that the time periods of the individual
phases are significantly different. Depending on the complexity of
the object and the maturity of the manufacturing process, the initial
failure period may not be present at all or may be characterized by a
period of up to a few thousand hours of operation.
In order to minimize the failure rate, specific preventative measures
are already carried out by OSRAM OS during the development phase as
well as in the subsequent manufacturing phase. In addition, the failure
rate is strongly influenced by the predominant operating conditions. For
example, for classic semiconductor elements, the failure rate doubles
when the junction temperature increases by 10 to 20°C.
(( ))
Distributions for reliability analysis
(
)
(( ))used for describing
(( ))
The Weibull and exponential
functions
(( )) distribution
( )
(
)
the bathtub curve are described in more detail (in )the following.
))
((
)
Weibull Distribution((
(( Weibull
)) distribution is well
(( suited)))to statistical
Due to its flexibility, the
analysis of all types (areas I to III (of )the bathtub
(( ))
( curve).
)) The primary
))
(( )) is that(((the))) curve (((can be adjusted
advantage of this function
with ((the))
(( this
)) way, a large number of well established
shape parameter b. In
fixed-form distributions (such as normal, log, exponential distributions,
((( )))
((( )))
etc.) can be realized.( )
(( ))
( )( )
(( ))
(
)
(
)
)
With a shape parameter b<1, a decreasing
failure((rate
(( ))
( )) (area I) is described,
with b=1, a constant failure rate (area II - exponential distribution) is
described and with b>1, an increasing failure rate (area III) is described.
Random Failure
( )
With the representation and interpretation of bathtub curves, it should
generally be kept in mind that in most cases, the curve is only based
on a few test points.
Wearout period
In the biparametric form of the Weibull distribution, the probability of
failure F(t) and its complement, reliability R(t), become:
((
((
Figure 11: Chronological
( ) progression of failure rate
( )
The chronological sequence
is comprised of three phases:
( )
Phase I - the early failure period
At the beginning of the product lifetime, a higher
failure rate
(
) can
(quickly falls
) off over time.
be observed, which
This
phase
can
be
( )
described with a Weibull distribution. This is generally caused by
design defects, weaknesses in the material, quality fluctuations in
(
)
production, or through application failures
( (dimensioning,
) handling,
( etc,) or) unreal, unconfirmed
testing, operation,
failures.
(
(
(
)
)
)
(
(((
))
))
)
)))
(( ))
(( ))
((
((
))
))
Their density function f(t) and failure rate λ(t) result in:
www.digikey.ca/lighting
((
((
((
((
(((
(
((
((
))
))
))
))
)))
)
))
))
((
((
(((
(
))
))
)))
)
((
((
))
))
13
Failure density ƒ(t)
Failure rate λ (t)
lifetime t
Reliability R(t)
Failure density F(t)
lifetime t
lifetime t
lifetime t
Figure 12: Weibull distribution for various shape parameters b and with a characteristic
lifetime of T=1
Where: b = shape factor
T = characteristic lifetime
Exponential distribution
The exponential distribution particularly represents the lifetime
distribution in Phase II of the bathtub curve, the area of random
failures. The failure rate is assumed to be constant over time.
High Performance Delivers
Flicker-Free Illumination
( )
(( ))
(
(
((
(
(
(
)
)
))
)
)
)
( )
( )
( )
( )
( )
( )
( )
( )
( ) ( )( )
( )
( )
( ) ( )( )
( )
(
)
(
)
(
)
With the exponential distribution,
( () ) (( )) ( ) the following applies:
( )
( ) of failure:
Probability
( () )
( )
( )
( () )
( )
( )
(( ()))
( () )
( )
() )
(
( )
Reliability:
( )
(( ) )
( )
( )
(( ) )
( )
(
)
(( ))
(( ))
( )
( )function:
Density
(( ))
(( ))
( )
(( ))
(( ))
( )
( )
Failure( rate:
)
( )
(( ))
( )
(( ))
( )
(( ))
( )
) and(in)connection with irreparable systems, the term
In this( area
( ))
(
)
(( )to failure)
) is used to describe
MTTF((mean
time
the average lifetime.
( )
(
)
(
)with a constant failure rate, this
)
For a (lifetime
distribution
(
) means
( the probability
)
)) MTTF,
that at(( the
of failure
( is around 63%
) or that on
(
)
average,
( )approximately 2/3 of all components have failed.
(
)
( )
)
(
)) (
( )
)((
) (
)
(
)(
) (
)
(
)(
) (
)
((
))(
)
LM3450 LED driver integrates power factor
correction and phase dimming decoding for
flicker-free, uniform dimming.
National’s LM3450 phase dimmable LED driver integrates active
power factor correction and a phase dimming decoder, making
it ideal for 10W-100W phase dimmable LED fixtures. It accepts
universal input voltages, features unique dynamic hold circuitry
for excellent dimming performance, and an analog adjust pin
for differentiated features such as thermal foldback, interface to
sensors, or dimmer range adjust.
Target Applications
• Dimmable downlights, troffers, and lowbays
• Indoor and outdoor area SSL
• Power supply PFC
LM3450 Key Features
• Critical conduction mode PFC
• Over-voltage protection
• Feedback short circuit protection
• 70:1 PWM decoded from phase
dimmer
• Analog dimming
• Programmable dimming range
• Digital angle and dimmer
detection
•
•
•
•
•
•
Dynamic holding current
Smooth dimming transitions
Low power operation
Start-up pre-regulator bias
Precision voltage reference
Forward phase (TRIAC) and
reverse phase dimmer compatible
www.digikey.ca/nationalsemiconductor-lighting
14
1.
What is scotopic vision?
A: Daytime vision
B: Star gazing
C: Night vision
D: Machine vision
Are you a Lighting Solution Expert? Submit your
answers to find out and automatically be entered for a
chance to win a Cree LED Module LMR4 Evaluation Kit.
Cree LED
Modules
provide a
simple solution
for lighting
designers and
manufacturers
|to adopt best in
class LED lighting
from Cree.
2.
What is radiometric power?
A: Current
B: Voltage
C: Radiant electromagnetic energy per unit time
D: Light
3.
What is LM80 report?
A: LM80 report is a component test with a minimum of 6,000
hours of data at 3 test temperatures.
B: LM80 report is a luminaire report with a minimum of 6,000
hours of data at 3 different temperatures.
C: LM80 is an extrapolation of component performance.
D: LM80 predict LED lifetime performance.
4.
In addition to providing a comprehensive LED solution, the
LED Module LMR4 offers lighting manufacturers access to
Cree TrueWhite® Technology. This technology mixes the light
from red and unsaturated yellow LEDs to create beautiful,
warm, white light. This patented approach enables color
management to preserve high color consistency over life of
the product. Cree TrueWhite Technology also enables a CRI of
at least 90 while maintaining high luminous efficacy – a no
comparison solution.
A: 137 years
B: 34.2 years
5.
C: 17.1 years
D: 5.7 years
What does CCT mean?
A: Closed Circuit TV
B: Correlated Color Temperature
C: Cold Cathode Tube
D: Capacitive Circuit Topology
6.
Submit your answers to
www.digikey.ca/trivia-lighting
for your chance to win!
How long is 50,000 hours?
What is the definition of a Footcandle?
A: Quantity of light falling on a surface
B: Amount of light emitted from a light source
C: Intensity of a light source
D: Quantity of candles in a linear foot
Sponsored By:
7.
Official Rules: No purchase necessary to enter or claim prize. A purchase will not improve an individual’s chances of
winning with such entry. Employees of Digi-Key (the “sponsor”), business partners, members of immediate families are not
eligible. Void where prohibited by law or internal company policy. Valid entries contingent upon participation in the trivia
contest and providing name and e-mail address. Prizes will be randomly drawn among eligible entries on or about March
7, 2011. Two Cree LMR040-0700-27F9-1KIT Evaluation Kits will be awarded. Total value of all prizes will not exceed
$500.00. Contest rules declare only one Cree LMR040-0700-27F9-1KIT Evaluation Kit will be awarded per winner. Winners
will be contacted via e-mail. If a winner cannot be reached within 7 days of initial attempt, winner will be disqualified and
an alternate winner selected. Odds of winning will be determined by the number of eligible entries received. Taxes on prizes
are the sole responsibility of winners. Sponsor reserves the right in its sole discretion to cancel or suspend the promotion
or any portion thereof should viruses, bugs, or other causes beyond control of Sponsor corrupt the administration, security
or proper play of the promotion, in which case the prizes will be awarded via a random drawing from among all eligible
entries received prior to cancellation. Sponsor is not responsible or liable for late, lost, damaged or misdirected entries. Any
attempts to deliberately damage any web site or undermine the legitimate operation of the promotion may be subject to
prosecution of criminal and civil law. If due to a printing, production or other error, more prizes are claimed than are
intended to be awarded, the intended prizes will be awarded in a random drawing from among all verified and validated
prize claims received. In no event will more than the stated number of prizes be awarded. Entries become the property of
Sponsor and will be used in accordance with Sponsor’s privacy policy, available at: www.digikey.ca/privacy. For a list of
winners, mail a self-addressed, stamped envelope to be received by 4/01/2011 to: Digi-Key Corporation, TechZone Trivia,
701 Brooks Ave South, Thief River Falls, MN 56701. If you do not want to receive any future mailings for contests of this
nature, you can remove your name by calling 800-344-4539.
Which wavelength produces the highest perceived
brightness?
A: 1 W of radiant power at a wavelength of 555nm
B: 1 W of radiant power at a wavelength of 650nm
C: 100 W of radiant power at a wavelength of 940nm
D: 50 W of radiant power at a wavelength of 320nm
8.
Which produces the highest perceived brightness?
A: 1 lm of light at a wavelength of 555nm
B: 1 lm of light at a wavelength of 650nm
C: Both produce the same perceived brightness
www.digikey.ca/lighting
15
Spectral Design Considerations
for White LED Color Rendering
by Dr. Yoshi Ohno, National Institute of Standards and Technology
In this article a National Institute of Standards
and Technology (NIST) expert analyzes various
white LED models by simulating their colorrendering performance and luminous efficacy.
The results provide some guidance for designing
multichip and phosphor-type white LEDs.
Introduction
One of the most important characteristics of light sources for general
lighting is color rendering. Color rendering is a property of a light
source that tells how natural the colors of objects look under the given
illumination. If color rendering is poor, the light source will not be
useful for general lighting. The U.S. Energy Policy Act of 19921 specifies
minimum requirements for both the luminous efficacy (lumens per
watt) and the Color Rendering Index (CRI)2 for several common types
of lamp products sold in the USA. This is an important aspect to be
considered for white LEDs being developed for general lighting.
White light from LEDs is realized by a mixture of multicolor LEDs or by
combinations of phosphors excited by blue or UV LED emission, and
thus they have greater freedom in spectral design than conventional
sources. Questions arise on how the spectra of white LEDs should
be designed for good color-rendering performance, e.g., whether
RGB white LEDs can satisfy the need or a four-color mixture is
needed or whether much broader, continuous spectra are required. To
evaluate the color-rendering performance of light sources, the CRI,2
recommended by the Commission Internationale de l’Éclairage (CIE),
is available and widely used, but it is known to have deficiencies,3,4
especially when used for sources having narrowband spectra. A poor
correlation between visual evaluation of RGB white LEDs and the CRI
is reported.5 The color-rendering problems of white LEDs are being
investigated by the CIE Technical Committee 1-62 with a future plan
to develop a new metric.
The main driving force for solid-state lighting is the potential of
huge energy savings on the national or global scale.6 Thus, when
considering spectra of light sources for general illumination,
another important aspect to consider is luminous efficacy (lumens
per watt). The term luminous efficacy is normally used for the
conversion efficiency from the input electrical power (watts) to the
output luminous flux (lumens). The luminous efficacy of a source is
determined by two factors: the conversion efficiency from electrical
16
power to optical power (called radiant efficiency or external quantum
efficiency7) and the conversion factor from optical power (watts) to
luminous flux (lumens). The latter is called the luminous efficacy of
radiation (LER). Since the LER and color rendering are determined
solely by the spectrum of the source, white LED spectra should be
optimized for both of these aspects.
The difficulty is that color rendering and the LER are generally in
a trade-off. Based on the CRI, color rendering is best achieved by
broadband spectra distributed throughout the visible region, while
luminous efficacy is highest with monochromatic radiation at 555 nm.
This trade-off is evident in many existing lamps. By studying the CRI,
some people are led to believe that white LED spectra should mimic the
spectrum of the sun or a blackbody. While such spectra would give high
CRI values, they would suffer significantly from low LER. The challenge
in creating LEDs for use as illumination sources is to provide the highest
possible energy efficiency while achieving the best color rendering
possible. For this purpose, an accurate metric of color rendering is of
importance. If the metric is incorrect, energy will be wasted.
To analyze the possible performance of white LEDs and also the
problems of the CRI, a simulation program has been developed.
Various white LED spectra of multichip type and phosphor type were
modeled and analyzed in comparison with conventional lamps.
The results of the simulation are presented, and the problems and
necessary improvements of the CRI are discussed.
Color-Rendering Index
The CRI is currently the only internationally agreed upon metric for
color rendering evaluation. The procedure for its calculation is, first,
to calculate the color differences ∆Ei (in the 1964 W*U*V* uniform
color space—now obsolete) of 14 selected Munsell samples when
illuminated by a reference illuminant and when illuminated by the
given illuminant. The first eight samples are medium-saturated
colors, and the last six are highly saturated colors (red, yellow, green,
and blue), complexion, and leaf green. The reference illuminant is
the Planckian radiation for test sources having a correlated color
temperature (CCT) <5000K or a phase of daylight† for test sources
having CCT ≥5000K. The process incorporates the von Kries
chromatic adaptation transformation. The Special Color Rendering
Indices Ri for each color sample are obtained by:
(1)
One of daylight spectra at varied correlated color temperatures. The formula is available
in Reference 8.
†
This gives the evaluation of color rendering for each particular
color. The maximum value of Ri (zero color difference) is 100, and
the values can be negative if color differences are very large. The
General Color Rendering Index Ra is given as the average of the first
eight color samples:
(2)
The score for perfect color rendering (zero color differences) is
100. Note that “CRI” is often used to refer to Ra, but the CRI actually
consists of 15 numbers: Ra and Ri (i =1 to 14).
Luminous efficacy of radiation
The energy efficiency of a light source is evaluated as its luminous
efficacy v, which is the ratio of the luminous flux (lumens) emitted
by the source to the input electrical power (watts). It is determined by
two factors:
(3)
where e is the radiant efficiency of the source (ratio of output
radiant flux to input electrical power; “external quantum efficiency”
is often used with the same meaning), and K is the luminous
efficacy of radiation (ratio of luminous flux to radiant flux,
abbreviated as LER in this paper) and is determined by the spectral
distribution S(λ) of the source.
(5)
where g(λ, λ0,∆λ0.5)=exp{−[(λ−λ0)/∆λ0.5]2}. The unit of wavelength
is the nanometer. Figure 1 shows an example of this LED model
compared with the SPD of a typical real blue LED spectrum
(measured at NIST with a relative expanded uncertainty (k=2) less
than 5 percent, depending on the wavelength).
Using the LED model described, spectra of a three-chip (RGB) white
LED and four-chip white LEDs with various combinations of peak
wavelengths and spectral widths can be created. For these white LED
spectra, the simulation program calculates the general CRI, Ra, and
special CRIs, R1 to R14, as well as color differences ∆E*ab in the CIELAB
color space8 and the LER K. In addition, a broadband phosphor-type
white LED model has been developed based on Planckian radiation
in a limited spectral range with some modification. The details of
the phosphor LED model are described later in this article under
“Phosphor-Type White LEDs.”
For three-chip and four-chip LED models, the program performs
automatic color mixing of each LED to bring its chromaticity
coordinate exactly on the Planckian locus for a given CCT. This allows
the use of an iterative method to optimize LED spectra for maximizing
)
)
)
(a)
)
Figure 1: LED model SLED(λ) at 464 nm compared with the SPD of a typical real blue LED.
∫
∫
)
)
)
, where
(4)
Here Km is the maximum LER, and its value, 683lm/W (for
monochromatic radiation at 555nm), is defined in the international
definition of the candela. While various other terms are used in
the LED industry, the terms introduced here are the ones officially
recommended internationally.7
White LED simulation program
Mathematical models have been developed for multichip LEDs and
phosphor-type LEDs in order to analyze numerous spectral designs of
white LEDs. To simulate multichip LEDs, the following mathematical
model for LED spectra has been developed. The spectral power
distribution (SPD) of a model LED, SLED(λ), for a peak wavelength λ0
and half spectral width ∆λ0.5, is given by:
(b)
Figure 2: Examples of optimization of RGB white LED spectra. The peak wavelengths of
LEDs range from 452 to 472nm for blue, from 543 to 553nm for green, and from 598
to 620nm for red. ∆λ0.5 =20nm except for green (30nm). (a) Maximum Ra obtained at
varied CCT. (b) Maximum LER, K(lm/W), obtained at varied Ra.
www.digikey.ca/lighting
17
Symbol
Description
CCT (K)
Duv
Ra
R9
R(9-12)
LER(lm/W)
CW FL
Cool white fluorescent lamp
4290
0.001
63
-89
13
341
DL FL
Daylight fluorescent lamp
6480
0.005
77
-39
13
290
TRI-P
Triphosphor fluorescent lamp
3380
0.001
82
17
47
347
MH
Metal halide lamp
4280
0.007
64
-120
19
296
MER
High-pressure mercury lamp
3750
0.000
43
-101
-29
341
HPS
High-pressure sodium lamp
2070
0.001
20
-214
-43
380
3-LED-1
3-chip LED model (457/540/605)
3300
0.000
80
-90
27
409
3-LED-2
3-chip LED model (474/545/616)
3300
0.000
80
89
88
359
3-LED-3
3-chip LED model (465/546/614)
4000
0.000
89
65
64
370
4-LED-1
4-chip LED model (461/527/586/637)
3300
0.000
97
96
87
361
4-LED-2
4-chip LED model (447/512/573/627)
3300
0.000
91
99
99
347
PHOS-1
Phosphor model, warm white (400 to 700nm)
3013
0.000
99
97
99
253
PHOS-2
Phosphor model, warm white (450 to 650nm)
3007
0.011
86
26
67
370
PHOS-3
PHOS-2 with narrow dip at 560nm
3000
0.000
81
47
61
341
PHOS-4
PHOS-2 with broad dip in green
3000
0.000
88
46
75
345
P-LED YAG
Phosphor LED (YAG phosphor)
6810
0.004
81
24
61
294
P-LED WW
Phosphor LED (warm white)
2880
0.008
92
72
80
294
NEOD
Incand. lamp with neodymium glass
2757
-0.005
77
15
60
-
Illum. A vis
Illum. A (only in 400 to 700nm)
2856
0.000
99
98
100
248
D65 vis
D65 (only in 400 to 700nm)
6500
0.003
100
98
100
248
Table 1: Summarized results for the light sources and LED models analyzed.
the Ra or the average of Ri for specific colors or maximizing K under
given conditions. Figure 2 shows examples of such optimization for an
RGB white LED model. The index Ra or K was maximized by varying
the peak wavelengths of the three LEDs under given conditions. The
spectral widths, ∆λ0.5=20, 30, and 20nm for blue, green, and red
LEDs were used, which are typical of LEDs currently available. Figure
2(a) shows the maximum Ra obtained at varied CCT (the values of
the LER are also plotted), which demonstrates that an RGB white LED
can achieve as high as Ra ≈90 and also indicates that Ra is not very
dependent on the CCT. It is also observed that the LER decreases for
higher CCT. This is because larger power for the blue LED is necessary
for higher CCT, while the blue component (≈450nm) has a very low
lumen contribution compared to green or red. Figure 2(b) shows the
maximum LER obtained at varied Ra, which demonstrates that RGB
white LEDs can produce K≈400lm/ W with decent Ra values (>80).
The data also demonstrates the trade-off between Ra and K, though
the slope is not very large. Note that the maximum Ra and K values
presented here may not be the highest values under each condition,
because the iterative method yields only local maxima. Also, these
results are only examples of what the program can do and are not
intended to recommend optimization of source spectra for maximum Ra.
There are some serious problems in judging the color rendering of white
LEDs with Ra alone, as discussed in later sections. The optimization can
be done for various other parameters such as the average of Ri for other
sets of samples, or the lowest average ∆E*ab for a given set of color
samples. When optimizing for the LER in real developments, the radiant
efficiencies of available LEDs should also be considered. For example,
the white LED models shown in Figure 2 are currently not realistic,
because the radiant efficiency (and thus the luminous efficacy) of LEDs
with 540 to 555nm peaks is very low.
18
The simulation program also presents the actual colors of the 14 color
samples of CIE 13.3 under the reference illuminant and test illuminant
on the computer display, which provides a visual impression of the
color differences of each sample. The color presentation is achieved
by conversion from XYZ to the display RGB space and applying the
gamma correction.9 By calibrating each primary color of the computer
display used, accurate colors (within the screen gamut) can be
presented on the display, and it might be possible to use this for
visual experiments in the future.
To compare the color rendering of white LEDs with that of common
existing lamps, the simulation program is also provided with the SPD
data on several different types of fluorescent lamps, high-intensity
discharge (HID) lamps, and some real white LEDs. The spectral
reflectance data on the samples in the program can be shifted in 10
nm steps in either direction in order to examine the sensitivity of the
results to small changes of the color of the samples.
Results
Table 1 summarizes the results of the calculation for the light sources
and LED models analyzed in this study, showing the CCT (unit: K); the
general CRI, Ra; a special CRI for strong red, R9; the LER; etc. The LER
and Ra of these sources are also plotted in Figure 3. The index R9 is
included in the table, because the red-green contrast is very important
for color rendering,10,11 and red tends to be problematic. Lack of the
red component shrinks the reproducible color gamut and makes the
illuminated scene look dull. This is the problem with many existing
discharge lamps. The index R(9–12) is the average of the special color
rendering indices R9 to R12 of the four saturated colors (red, yellow,
green, and blue). Duv, introduced in this paper, is the distance from
the chromaticity coordinate of the source to the Planckian locus on
the CIE 1960 UV chromaticity diagram with polarity plus (above the
Planckian locus) or minus (below the Planckian locus).‡ It is important
that the chromaticity coordinate of illumination is very close to the
Planckian locus, since greenish or pinkish white light is not accepted
for general illumination, and Duv of fluorescent lamps is typically
controlled to less than ±0.005. For multichip LED models, the spectral
width ∆λ0.5=20nm is used for all LEDs except for green ones (30nm).
Figure 5: Special CRI of the two three-chip white LED models shown in Figure 4.
Figure 3: LER and the general CRI Ra of the conventional sources and LED models analyzed
Conventional light sources
The first six sources in Table 1 are conventional discharge lamps
commonly used, including fluorescent lamps and HID lamps. The data
on these lamps are only samples and not representative of the type of
lamp. Among these lamps, the triphosphor lamp has the highest CRI,
Ra=82. It should be noted that the values of R9 of most of these lamps
are very poor, though R9 values are exaggerated (by factor of 2 or
more) by the nonuniformity of the W*U*V* color space used in the CRI
formula. For example, R9=17 (TRI-P) would correspond to ≈60 based
on the CIELAB color space. The values of R(9–12) for these lamps,
thus, are not good either. Even though R9 is important, it has not been
paid much attention, because R9 is not included in the calculation of
Ra and also probably because increasing the deeper red component
reduces the LER and thus the lumen output of the lamp. This has
been one of the problems with the CRI. The metric for color rendering
is important in that it drives manufacturers to design light spectra to
maximize the index Ra.
Three-chip white LEDs
The second group in Table 1 and Figure 3 (3-LED-1 to 4-LED-2) is
a group of multichip white LED models. 3-LED-1 is a three-chip
LED model optimized for the highest LER at Ra=80 and 3300K and
has a very high LER (K=409lm/W). 3-LED-2 is optimized for the
highest R(9–12) (=88) at the same Ra (=80) and the same CCT, with
Figure 6: The changes of Ra and R(9–12) for three-chip white LED models when the
wavelengths of the sample spectral reflectance data are shifted.
K=359lm/W. The spectra and the special CRI, R1 to R14, of these
three-chip LED models are shown in Figs. 4 and 5. Both models have
the same Ra value of 80, but 3-LED-1 exhibits very poor rendering of
red (R9=−90, appearing brown) and an R(9–12) of only 27, whereas
3-LED-2 exhibits good rendering of all the four saturated colors as
well as the medium-saturated colors. This is a case where sources
having the same Ra can exhibit very different color-rendering
performance (possibly having serious problems with saturated colors).
‡
Figure 4: The SPDs of the two three-chip LED models, both having Ra=80 at 3300K.
The symbol ∆uv is commonly used for this distance, but with no signs (no information
on the direction of the deviation).
www.digikey.ca/lighting
19
This demonstrates that Ra is unreliable for judging the color rendering
of three-chip white LEDs and possibly also for conventional light
sources having only a few narrow peaks.
Then, is R(9–12) a good indicator? Since saturated colors have sharp
changes in spectral reflectance curves, R(9–12) may cause some
irregular results with SPDs having large valleys between peaks in the
spectral distribution curve. As a simple test, all the sample spectral
reflectance data were shifted by amounts from −20 to +20nm to
examine the sensitivity of the results to small changes of the colors of
the samples. Figure 6 shows the changes in Ra and R(9–12) caused
by the shifts. As expected, R(9–12) is found to be very sensitive to
the wavelength shift of the samples, while Ra is fairly stable. This
means that, even if R(9–12) is good, color rendering of some other
saturated colors (orange, purple, etc.) may not be accurately rendered
(the hue will shift). 3-LED-3 is optimized for highest CRI (Ra=89),
K=370lm/W, at 4000 K. This model also exhibits strong sensitivity of
R(9–12) to sample color shifts. While R(9–12) is an important number
to look at, one should be aware that the results do not apply to all
the saturated colors. 3-LED-2 and 3-LED-3 seem to have fairly good
color-rendering performance except for this problem, which should be
studied further.
Four-chip white LEDs
Figures 7 and 8 show the SPDs and the special CRI values, R1 to R14,
of two four-chip LED models. 4-LED-1 is optimized for the highest Ra
(=97) at 3300K, with R(9–12) =87 and K=361lm/W. The ∆E*ab of all
the samples is less than 3.1 except for R12 (blue), which is 11.9. The
model 4-LED-2 is optimized for the highest R(9–12) (=99) at 3300K,
with Ra=91 and K=347lm/W. The ∆E*ab of all the samples is less than
2.4. With both models, all the sample colors are presented excellently.
Figure 7: The SPDs of the two four-chip white LED models 4-LED-1 and 4-LED-2.
Figure 8: Special CRI of the four-chip white LED models shown in Figure 7.
Figure 9 shows the results of the wavelength-shifting test. The
sensitivity of R(9–12) is much less than in the case of three-chip LED
models (Figure 6) and considered to be not significant.
Phosphor-type white LEDs
Figure 10(a) shows the SPD of one of the commercially available
warm-white LEDs using phosphors, denoted P-LED-WW in Table 1
and Figure 3. The spectrum is designed to mimic Planckian radiation.
Following this example, a simple model for phosphor-type white
LEDs is made using Planckian radiation that is cut off smoothly at
both ends of the spectrum, using a half of a Gaussian function. The
temperature of the Planckian radiation, both cutoff wavelengths (the
half point of the rise or drop), and the width of the half Gaussian
function can be varied. Then, another Gaussian function of a given
width and height is subtracted from the quasi-Planckian function
to produce a valley in the curve. The center wavelength, depth, and
width of the valley can be varied.
Figure 10(b) shows the result of simply trying to mimic Planckian
radiation as closely as possible for good color rendering, in which
case the cutoff wavelengths are set at 400 and 700nm (denoted
PHOS-1 in Table 1). As found in Table 1, the color rendering of this
source is excellent, with Ra =99. However, the LER is 253lm/W,
only 68 percent of that of the good three-chip white LED (370lm/W,
3-LED-3). If such white LEDs are used, a great amount of energy will
be wasted. To improve this, one may think of cutting off both ends
of the spectrum, which contribute very little to the luminous output.
Figure 10(c) is such an example, where the cutoff wavelengths are
20
Figure 9: The changes of Ra and R(9–12) of the four-chip LED models when the
wavelengths of the sample spectral reflectance data are shifted.
set at 450 and 650nm (PHOS-2 in Table 1). This spectrum produces
Ra=86 and K=370lm/W, which are comparable to those of the good
three-chip LEDs. However, one should pay attention to Duv. It is
+0.011, which indicates that the light is fairly yellowish and may not
be acceptable for indoor lighting. To reduce the Duv value, the green
(or yellow-green) part of the spectrum should be reduced. The SPD
shown in Figure 10(d) is one solution to this, where a narrow valley
is made at 560nm (PHOS-3 in Table 1). The Duv value is reduced to
zero, with Ra=81 and K=341lm/W. From this condition, the spectrum
is optimized for the highest Ra value by varying the valley parameters.
The result is shown in Figure 10(e). This yields Ra=88, R(9–12)=75,
and K =345lm/W, while keeping Duv=0.000. The color rendering of
Figure 11: The SPD of an incandescent lamp with neodymium glass.
Figure 10: 10 SPDs of (a) commercially available warm white LED model and (b to e)
phosphor-type LED models.
this source is probably good enough for office and home lighting. The
example of a commercially available warm white LED shown in Figure
10(a) has a high value of Ra(=92), but Duv= +0.008, rather yellowish,
and also K =294lm/W, which can be further improved.
The same considerations should apply when white LED spectra are
designed to mimic daylight spectra. For example, the D65 spectrum
cut out in the 400 to 700nm region (D65-vis in Table 1) yields an LER
of only 248lm/W, much lower than those of the good three-chip and
four-chip LED models (350 to 400 lm/W). There are proposals by a
few groups to judge color rendering performance by the closeness of
the SPD curve to the Planckian radiation or daylight spectrum (of the
same CCT) in the 400 to 700nm region. This is not recommended,
because it would drive manufacturers to design white LEDs having
low luminous efficacy. In addition, as already mentioned, four-chip
LEDs, for example, can have as good color rendering as full-spectrum
broadband light sources, and need to be studied further.
As indicated, the deviation of chromaticity coordinates of the source
from the Planckian locus is not treated well by the CRI. For example,
the RGB ratio of the three-chip LED model, 3-LED-2 (3300K, Ra=80,
Duv=0.000), is modified so that the chromaticity coordinate deviates
in the yellow direction (Duv= +0.015), keeping the same CCT. This
light would be very yellowish and will not be acceptable for indoor
lighting. However, the Ra value increased to 85 rather than decreased.
This is a problem related to color constancy and how to handle
chromatic adaptation.
CCT and color preference
Some manufacturers are considering a goal of realizing sunlight
spectra or daylight spectra with white LEDs, because these are the
most natural light that the human eyes have been adapted to and
because LED technology makes it possible. However, two points should
be considered. First is the energy aspect. If such full-spectrum white
LEDs mimicking Illuminant D65 or D50 in the 400 to 700nm region
were made, their LER would be only about 250lm/W, as discussed in
the previous section. Secondly, “natural daylight” implies that the CCT
of the source would be 6500 K (D65) or 5000K (D50) at least. The CCT
of fluorescent lamps, for example, has been designed for people’s
preference in the targeted markets (different countries). For homes in
the USA, warm white (2800K to 3000K) is dominant; 6500K white light
would not be accepted for homes in the USA. However, in Japan, for
example, 5000K is dominant. Some other countries prefer even higher
CCT, up to 7500K. Preferences for offices are different. For example,
4200K is common in the USA. Therefore, “natural daylight” does not
describe all markets and applications.
Another aspect to be considered for acceptance in the market is color
preference. As an example, incandescent lamps with neodymium
glass have been in the market for many years, and they have been
gaining popularity recently. The spectrum of this type of lamp is
shown in Figure 11. There is strong absorption in the yellow region.
The color rendering characteristics are shown in Table 1 (see NEOD).
It shows Ra=77 and R9=15, fairly poor, but the lamps are advertised
for more brilliant colors than normal incandescent lamps and are
actually preferred by many people. The reason for the popularity
of this type of lamp is explained in Figure 12, which shows the
plots of colors of the 14 samples in the CIELAB color space under
illumination by the neodymium-glass lamp and the reference source
(Planckian). It is observed that the chroma of red and green samples
is increased by the lamp compared to the reference source. These
deviations discount the values of CRI; however, the red-green contrast
is enhanced and the color gamut area is increased. This provides
more colorfulness to the illuminated scene. It is known that people
prefer slightly enhanced chroma of illuminated objects.12,13 Another
study11 shows that visual clarity is well correlated with the gamut
area produced by the four saturated colors (red, green, yellow, blue).
If visual clarity is increased, this is not just a matter of preference.
The present CRI simply evaluates the color shifts from the reference
source to test source. Color shifts in all directions, whether decreasing
or increasing chroma, are counted equally; therefore the results are
www.digikey.ca/lighting
21
Figure 12: Colors of the 14 samples in CIELAB space under illumination by the
neodymium glass lamp and the reference source (Planckian).
Figure 14: Colors of the 14 samples in CIELAB space under illumination by the three-chip
LED model shown in Figure 13 and by the reference source (Planckian).
Discussion of CRI
In the analyses reported here, it is demonstrated that such an
index as Ra, if it is accurate, would be a useful tool to design
spectra of white LEDs. However, as already shown, Ra alone is not
a reliable metric for color rendering, especially for white LEDs. The
additional indices for saturated colors such as R9 and R(9–12) also
need to be examined. Several problems with the CRI (particularly,
Ra) that have been identified or demonstrated in this study are
summarized below.
1. Since Ra is determined only with medium-saturated colors, the
color rendering of saturated colors (R9 to R12), particularly R9,
can be very poor even though Ra is fairly good. Saturated colors
should somehow be considered.
Figure 13: The SPD of a three-chip white LED model with peak wavelengths 455, 547,
and 623nm.
only for color fidelity. For overall color rendering, decreased chroma is
worse than increased chroma or hue shift, so the directions of color
differences should somehow be considered.
Such light source spectra that produce enhanced chroma can be
realized by a three-chip white LED. An example is shown in Figure
13. This is a 3-LED model with the peak wavelengths 455, 547,
and 623nm and with spectral half-widths 20, 30, and 20nm, for
blue, green, and red, respectively, yielding CCT =3300K, Ra= 73,
R(9–12)=50, K=363lm/W. The CIELAB a*, b* coordinates of the 14
samples are plotted in Figure 14. The color fidelity of this source will
not be good, but the color gamut is notably enlarged. This may be
an interesting white light spectrum to be studied from a preference
point of view.
22
2. The results for three-chip LEDs tend to be sensitive to small
variation of color samples, especially for saturated colors. Even
though the values of R9 to R12 are good for the given set of
samples, rendering of other saturated colors can be poor.
3. The CRI does not take good account for the shift in chromaticity
coordinates across the Planckian locus. The index Ra hardly
changes with a change of light source chromaticity from Duv=0
to Duv= +0.015, for example. This is a problem related to
handling chromatic adaptation and color constancy.
4. The CRI does not consider the direction of color shift. A
decrease of chroma has negative effects, and an increase has
rather positive effects (increased visual clarity). The directions
of color shift should somehow be considered.
5. The plots of color differences in the W*U*V* space (outdated)
indicates significant nonuniformity compared to the CIELAB
space. The distortion is notable particularly in the red region.
6. The 2000 K (very reddish) blackbody spectrum or a daylight
spectrum at 20,000K (twilight) gives Ra =100, though colors
do not render well. This indicates a problem in the reference
source (the CCT of the reference source moves with that of the
test source). Color constancy is assumed to be too perfect. Very
low or very high CCT should be penalized.
Conclusions
Various white LED models have been analyzed by simulation
of their color-rendering performance together with energy
efficiency aspects. The results provided some guidance for design
of multichip and phosphor-type white LEDs. It is shown that
well-designed three-chip white LEDs may have acceptable color
rendering (for indoor lighting) as well as good luminous efficacy,
but further study is needed. Four-chip white LEDs with appropriate
design are shown to have excellent color rendering as well as
good luminous efficacy. Phosphor-type LEDs can have excellent
color rendering but tend to have lower luminous efficacy. Attention
should be paid to the value of Duv when designing spectra of
phosphor-type white LEDs.
Finally, several problems with the CRI have been identified or
confirmed in this study. The index Ra is unreliable for the colorrendering performance of white LEDs (as well as for conventional
sources). Some of the problems can be addressed by examining
R9 to R12 (especially R9) additionally, but this will not solve the
fundamental problems. Also, the need for describing color-rendering
performance in one number for general users is strong. A new,
improved metric for color rendering, solving these problems, is an
urgent need.
References
1. Energy Policy Act of 1992, U.S. Public Law 486, 102nd Congress, 24 October 1992.
2. CIE 13.3:1995, “Method of measuring and specifying colour rendering properties
of light sources” (1995).
3. CIE 135/2:1999, “Colour rendering,” TC 1-33 closing remarks (1999).
4. J. Schanda and N. Sandor, “Colour rendering, past—present— future,” in Proc.
Int. Lighting and Colour Conf., pp. 76–85, SANCI (2003).
5. N. Narendran and L. Deng, “Color rendering properties of LED light sources,”
Proc. SPIE 4776, 61–67 (2002).
6. U.S. Department of Energy, “Illuminating the challenges—solid state lighting
program planning workshop report,” (2003).
7. CIE 17.4: 1989/ IEC 50(845), “International Lighting Vocabulary” (1989).
8. CIE 15:2004, “Colorimetry,” 3rd ed. (2004).
9. CIE 122-1996, “The relationship between digital and colorimetric data for
computer-controlled CRT displays” (1996).
10. J. Worthey, “Color rendering: asking the questions,” Color Res. Appl. 28(6),
403–412 (2003).
11. K. Hashimoto and Y. Nayatani, “Visual clarity and feeling of contrast,” Color Res.
Appl. 19(3), 171–185 (1994).
12. D. B. Judd, “A flattery index for artificial illuminants,” Illum. Eng. (N.Y.) 62,
593–598 (1967).
13. W. A. Thornton, “A validation of the color-preference index,” J. Illum. Eng. Soc. 4,
48–52 (Oct. 1974).
The DOE CALiPER (Commercially Available LED Product
Evaluation and Reporting) program independently tests
and provides unbiased information on the performance
of commercially available SSL products. The DOE
conducts at least two rounds of CALiPER testing each
year through independent testing laboratories. The results
are available on the DOE Solid State Lighting website and
included detailed test reports, analysis of test results, and
summary reports.
The DOE and EPA have implemented a new set of stringent
product testing and verification procedures for the ENERGY
STAR® program. Effective January 1, 2011, all products
participating in the ENERGY STAR® program are required
to follow a new set of third-party certification procedures.
This includes independent performance testing by an EPArecognized laboratory, product review and certification by
an EPA-recognized Certification Body, and on going offthe-shelf verification testing to ensure products continue to
meet ENERGY STAR® requirements.
Introduction to
High Brightness LEDs
Cypress Semiconductor
This comprehensive High Brightness LED introduction
runs 25 minutes. Cypress explains the basics behind
color science, LED
operation, industry
terminology, and
uses of HB LEDs.
*This paper is a revision of a paper presented at the SPIE conference on Solid State
Lighting, August 2004, Denver, Colorado.
This article originally appeared in Optical Engineering 44(11), 111302 November 2005.
Reprinted with permission.
www.digikey.ca/ptm
www.digikey.ca/lighting
23
Color Rendering of
Light Sources
by Dr. Wendy Davis, National Institute of Standards and Technology
The National Institute of Standards and
Technology (NIST) is proposing a Color Quality
Scale (CQS) that is designed to address many
shortcomings of the widely used Color Rendering
Index with regard to solid-state lighting.
This Color Quality Scale (CQS) is being developed by (NIST) with
input from the lighting industry and the International Commission
on Illumination (CIE) to address the problems of the CIE Color
Rendering Index (CRI) for solid-state light sources and to meet
the new needs of the lighting industry and consumers for
communicating color quality of lighting products. The method for
calculating the CQS is based on modifications to the method used
for the CRI. Although simulations support the appropriateness
and usefulness of the proposed CQS, vision experiments will be
conducted to improve and validate the CQS. It will then be proposed
as a new international standard.
Color Rendering Index vs. Color Quality Scale
The color appearance of objects under artificial lighting is an
important characteristic of light sources, particularly to consumers.
Currently, the only internationally-accepted assessment procedure
for evaluating the color appearance of objects under light source
illumination is the CIE color rendering index (CRI).
In the calculation of the CRI, the color appearance of 14 reflective
samples is simulated when illuminated by a reference illuminant
and the test source. After accounting for chromatic adaptation
with a Von Kries correction, the difference in color appearance ∆Ei,
for each sample, between the test and reference illumination, is
computed in CIE 1964 W*U*V* uniform color space. The special
color rendering index (Ri) is calculated for each reflective sample by:
= 100
4.6
The general color rendering
simply the average of Ri for
= index
100(Ra) is4.6
the first eight samples (shown in Figure 1), all of which have low to
moderate chromatic saturation:
8
=
=
24
1
8
1
8
8
=1
=1
Figure 1: The eight color samples used in the calculation of Ra.
A perfect score of 100 represents no color differences in any of the
eight samples under the test source and reference illuminant.
Unfortunately, the CRI only evaluates color rendering, or color
fidelity, and ignores other aspects of color quality, such as chromatic
discrimination and observer preferences. The Color Quality Scale
(CQS) is designed to incorporate these other aspects of color
appearance and address many of the shortcomings of the CRI,
particularly with regard to solid-state lighting.
Below is a description of several of the CRI shortcomings and the
CQS approach to addressing the shortcomings.
Problem:
The set of reflective samples used to calculate the Ra in the CRI
contains too few samples and has low chromatic saturation.
Solution:
The CQS uses a set of 15 Munsell samples, as shown in
Figure 2. These samples have the much higher chroma, span
the entire hue circle in approximately even spacing, and are
commercially available.
Figure 2: The 15 samples used by the Color Quality Scale (CQS).
Problem:
The uniform object color space (1964 W*U*V) used by the
CRI to calculate sample color differences is outdated and no
longer recommended.
Solution:
The CQS is calculated using the CIE 1976 L*a*b (CIELAB)
uniform object color space, which is more uniform and currently
recommended by CIE.
Problem:
The CRI penalizes lamps for shifts in hue, chroma (chromatic
saturation), and lightness, in all directions, of the reflective samples
under the test source (compared to under the reference illuminant).
Solution:
The CQS will only penalize a lamp’s score for hue shifts, lightness
shifts, and reductions in chroma. Lamps that increase object chroma
relative to the reference illuminant are not penalized because these
positive effects are generally preferred. This is shown in Figure 3.
Problem:
The color differences for all reflective samples in the CRI are
averaged, enabling a high score despite poor rendering of one or
two colors.
Solution:
To ensure that large hue shifts of any sample have notable influence
on the CQS, the root-mean-square of color shifts of all of the
individual samples is used.
Problem:
The CRI can result in negative values for some lamps.
Figure 3: Effects of the saturation factor illustrated in CIELAB color space. According to
the CQS, when the chroma increases under the test illuminant (with no change in hue)
in Figure 3A, there is no change in score. In Figure 3B, when the chroma decreases
under the test illuminant, the CQS score is decreased. In Figure 3C, when the chroma
increases and the hue shifts, the CQS score is decreased for the hue shift but not for
the increase in chroma.
Problem:
Because the CCT of the reference illuminant is matched to that of the
test source, the CRI score can be perfect (Ra = 100) for reference
illuminants of any CCT, even though actual color rendering is
degraded at extreme CCTs.
Solution:
The CCT of the reference source is matched to that of the test, but a
multiplication factor, the CCT factor, is included to penalize sources
with extreme CCTs.
Problem:
The CRI uses the Von Kries chromatic adaptation correction in its
calculation. It is outdated.
Solution:
The CQS uses the Colour Measurement Committee of the Society
of Dyers and Colourists Chromatic Adaptation Transform of 2000
(CMCCAT2000).
The Lighting Facts Label, jointly developed by the DOE
and the Next Generation Lighting Industry Alliance
(NGLIA), provides a quick and unbiased summary of
performance data for solid state lighting products. The
performance data includes luminous flux, system efficacy,
electrical power, correlated color temperature (CCT)
and color rendering index (CRI). To qualify for the label,
the product must be tested according to the IES LM-79
industry standard by a NVLAP accredited laboratory or an
independent testing lab recognized by the DOE CALiPER
program. In addition, the DOE monitors the accuracy
of reported product performance through off-the-shelf
verification testing.
Solution:
The CQS imposes a 0-100 scale.
Testing the CQS
Many computational simulations have been performed and, at the
level of subjective visual impression, appear to confirm the ideas used
in the CQS. However, a series of thorough and well-controlled vision
experiments are necessary to test, improve upon, and validate the
computational analyses. Experiments testing observers’ chromatic
discrimination and hue perception of illuminated objects will be
complemented by subjective rankings of naturalistic scenes. Current
experiments are also testing the relationship between illuminance
and object chroma. Since the CQS is intended to be a metric of overall
color quality, the data from several types of experiments will be used
to assess and improve its performance. Experiments are currently
being conducted on color quality and the CQS using the Spectrally
tunable lighting facility.
The CQS is being proposed to CIE TC 1-69 “Colour Rendition by White
Light Sources.” This committee is tasked with recommending a new
assessment procedure for evaluating the color rendering/quality
properties of light sources and is chaired by Wendy Davis.
Brand X
Light Output/Lumens
Measures light output. The higher the
number, the more light is emitted.
Reported as “Total Integrated Flux (Lumens)” on
LM-79 test report.
Watts
Measures the energy required to light
the product. The lower the wattage,
the less energy used.
Reported as “Input Power (Watts)” on LM-79 report.
Lumens per
Watt/Efficacy
Measures efficiency. The higher the
number, the more effecient the product.
Reported as “Efficacy” on LM-79 test report.
A Program of the U.S. DOE
Industry standardized test procedure that
measures performance qualities of LED luminaires
and integral lamps. It allows for a true comparison
of luminaires regardless of the light source.
E
840
9
93
9
3
Light Output (Lumens)
Watts
Lumens per Watt (Efficacy)
L
P
M
Color Accuracy
Color Rendering Index (CRI)
XA
Light Color
E
Correlated Color Temperature (CCT)
Warm White
IESNA LM-79-2008
Brand
2700K
Bright White
3000K
87
Measures color accuracy.
Color rendition is the effect of the lamp’s light
spectrum on the color appearance of objects.
Correlated Color
Temperature (CCT)
Measures light color.
Daylight
4500K
Color Rendering
Index (CRI)
6500K
All results are according to ESNA LM-79-2008: Approved Method for the Electrical and
Photometric Testing of Solid-State Lighting. The U.S. Department of Energy (DOE) verifies
“Cool” colors have higher Kelvin temperatures (3600-5500 K).
“Warm” colors have lower color temperatures (2700-3500 K).
Color temperatures higher than 6500 are outside of the
defined region for white light, but may be appropriate for
outdoor applications.
product test data and results.
Visit www.lightingfacts.com for the Label Reference Guide.
Registration Number
Model Number
Type
www.digikey.ca/lighting
Registration Number: ABC435TH4792023
Model Number: 18756CHT56428954RGHT1234HG
Type: 18750CHT56428954RGHT1234HG
25
Radiant and Luminous Flux
by William F. Long, Ph.D., O.D.,
Associate Professor Emeritus, University of Missouri-St. Louis
The radiant flux from a light source varies with
the source and the wavelength. Understanding
the variations is key to lighting design.
A physicist studying a system radiating electromagnetic energy would
use instruments that measure that energy in the usual MKS units. The
energy emitted by the system, U, would be given in joules. The time
rate of production of radiant energy, radiant flux P, would be defined as:
=
where t is time. The MKS unit of radiant flux would be, therefore,
joules per second or watts.
Generally, the radiant flux from a source differs for different
wavelengths. The flux at a given
= wavelength, λ and in a spectral
interval dλ is given by Pλdλ. A plot of Pλ versus λ gives the power
spectrum of the radiator. Power spectra may be either continuous
or discrete. The usual tungsten filament lamp has a continous
power spectrum like that of Figure 1. The mercury lamp spectrum
shown in Figure 2 is a typical line
40 spectrum with the spectral lines
= Fluorescent lamps have
corresponding to energy level=transitions.
00
discrete spectra superimposed on a continuous spectra. The total
radiant flux P emitted by a light source is the integral of Pλdλ over
all wavelengths,
= 68 =
This is the area under the Pλ curve.
=
40
00
=
= 68
=
0
26
=
00
=
Figure 1: A typical continuous power spectrum.
0
=410
0
410
Because visible light occupies only a small part of the
electromagnetic spectrum and because the sensitivity of the
eye varies considerably with the actual wavelength of the visible
light, the units quantifying light energy in vision science are
different from those used in =
physics. These units are based on the
experimentally measured spectral sensitivity of the eye quantified
by the spectral sensitivity function Vλ, or the relative spectral
sensitivity function vλ. One way to determine vλ, in principle at
least, would be to have a typical observer compare the amount
of energy required to make=the light sensation or “brightness” of
light at one wavelength match that of light at another wavelength.
If, for example, three times as much radiant energy were required
at 500nm as at 540nm to produce the
same magnitude of visual
=
sensation, then:
40
410
=
Figure 2: A typical line spectrum, that of a mercury vapor lamp.
=
= in the two stimuli confound this
(In practice, the color differences
type of experiment, and more elaborate techniques are necessary.
The exact results vary from observer to observer and with the state
= 68
of light adaptation of the observer, but we’ll ignore those effects here.
The vλ curve accepted as standard is actually an average over a
great many normal observers.) 40 =
00
The relative spectral sensitivity function
= is given by the graph of
Figure 3 or the numerical tabulation of Table 1. The spectral sensitivity
function vλ is proportional to itself according to the equation:
= 68
=
0
410
=
The value of the constant has been re-established from time to time
so some texts may use 680 while others use 685. The value 683 is
the currently accepted value.
==
TechZone™
On Demand
Get your next
TechZone
Magazine how
you want it; hot
off the printing
press or delivered
digitally. Choose
from our Lighting,
Microcontroller,
Sensors, and
Wireless TechZone Magazines to
receive current information on each
technology. Use our request page,
www.digikey.ca/request, to pick
the Digi-Key TechZone Magazine
publications you want to receive, and
choose how you want to receive them!
Figure 3: The relative spectral luminosity curve, vλ.
λ(nm)
vλ
λ(nm)
400
0.000
500
410
0.001
510
420
0.004
520
430
0.012
530
440
0.023
540
450
0.038
550
vλ
λ(nm)
0.323
=
600
0.631
700
0.004
610
0.503
710
0.002
0.503
0.710
=
=
0.862
0.954
=
0.995
460
0.060
560
0.995
470
0.091
570
480
0.139
580
490
0.208
590
0.95240
40
Table 1: A tabulation of the values of vλ.
00
λ(nm)
vλ
620
0.381
720
0.001
630
0.265
730
0.000
640
0.175
-
-
650
0.107
-
-
660
0.061
-
-
670
=
0.032
-
-
=
0.8700000 680
0.75740
vλ
690
=
0.017
-
-
0.008
-
-
Microcontroller Solutions
Page 39
TZM111.US
TL101.US
Look Inside Today’s
Microcontroller Technology
See page 22
www.digikey.com/microcontroller
TZM111.US
TZW101.US
Page 34
Designs
Page 39
Considerations for
Low-Power
Your Wireless
Processing
Designs
See page
Page 26
34
TZM111.US
Usually ∆λ is taken to be 10nm.
=410
Optimize
Low-Power
Embedded
Simplify
the
Designsof an RF
Design
Remote Control
Optimize
Thermal Management:
Low-Power
Processing
TurnLow-Power
down the
HEAT!
Embedded
using Simpson’s rule. This=
means
= that (1) is approximated by the sum:
0
=410
=410
=
Look Today’s
Deeper
Look Inside
Inside
Today’s
Microcontroller
Technology
Wireless
Technology
MICROCONTROLLER
LIGHTING SOLUTIONS
SOLUTIONS
TechZoneTM
Page 39
www.digikey.com/microcontroller
www.digikey.com/wireless
The integral is carried out in the range from 410nm to 720nm since
that is the non-vanishing range410
of vλ .
0 0 performed numerically
In practice the integral in equation (1) is always
Save $120 Billion
Microcontroller
Solutions
Wireless Solutions
0
410
410
TZW101.US
TL101.US
TZM111.US
TL101.US
LIGHTING SOLUTIONS
WIRELESS
SOLUTIONS
MICROCONTROLLER
SOLUTIONS
Optimize
Low-Power
Embedded
Adopt LEDs,
Designs
TZM111.US
TZL101.US
=
(1)
TM
TechZoneTM
TechZone
==
Look Inside Today’s
Microcontroller Technology
www.digikey.com/microcontroller
www.digikey.com/lighting
The total luminous flux F is obtained by integrating the above equation
=
to obtain:
0 0
TZL101.US
TL101.US
TZM111.US
TL101.US
LIGHTING
MICROCONTROLLER
LIGHTING SOLUTIONS
SOLUTIONS
SOLUTIONS
Microcontroller
Solutions
Lighting Solutions
==
TM
TechZoneTM
TechZone
The lumen (abbreviated
lm) in the previous equation is a unit
==6868
defined to characterize luminous flux in the same way the watt
characterizes radiant energy. The luminous flux Fλ at wavelength
λ in a range dλ is=
related
68 to the radiant flux in that interval by:
Low-Power Processing
Page 34
Values of Pλ and Vλ can be taken from a table like table 1 or
interpolated from graphs like those in Figures 1-3.
www.digikey.ca/request
www.digikey.ca/lighting
27
A Halogen Desk Lamp
Conversion to LEDs
by Jim Young and Bernie Weir, ON Semiconductor
It is estimated by the International Energy
Agency that more than 19% of electrical
energy demand globally is used for lighting.
As a result, there has been significant effort
around the globe in replacing inefficient
incandescent light sources with higher
efficiency solutions.
production LED light sources. A Cree XLAMP® MC-E LED was used
as the light source. This product incorporates four LEDs in a single
compact package, which is ideal for directed light applications
such as MR-16 and portable task lamps. The original lamp was
characterized before and after modification to highlight the real
world performance differences.
Much of the focus in curbing the electrical energy demand at the
consumer level has been in programs to drive adoption of compact
fluorescent lamps (CFL). The choice of using CFLs is not without
consequence; each bulb contains mercury which, when not disposed
of properly, ends up in landfills. Moreover, the form factor is not ideal
for all lighting applications. An alternative, which has been gaining
significant attention, is phosphor converted “white” LEDs. LED lumen
output and efficacy – the measure of lumen output versus input
power – continues to make significant progress year over year while,
at the same time, the cost in terms of lumen/$ is falling. Beyond
energy savings, high brightness LEDs can have lifetimes of greater
than 50,000 hours when properly designed and operated, which
eliminates the cost of replacing bulbs.
An off-the-shelf halogen desk lamp was used as the basis for
a demonstration of the real world performance of today’s latest
Figure 2: 50W bulb in lamp housing with protective glass cover and handle
Figure 1: Off-the-shelf 50W halogen portable desk lamp
Figure 3: 60Hz iron-core transformer, 2.4 pounds
28
The DoE ENERGY STAR Standard for Solid State Lighting Luminaires
(Version 1.1 - 12/19/08) includes requirements for a minimum
power factor (PF) of 0.7 for portable residential desk lamps. Typical
LED drivers at this power level normally have a PF of 0.5-0.6 that
falls below the requirements. The solution presented here has been
specifically designed to exceed the residential ENERGY STAR® PF
requirements without the addition of extra circuitry.
Halogen incandescent bulbs typically operate at temperatures in excess
of 250-350ºC, and consequently, luminaries must provide safeguards
for users. In this case, a glass cover plate restricts access to the bulb
to protect the user, and a plastic shell provides a thermal shield above
the bulb. Even with these safeguards, this halogen luminaire includes a
handle on the end of a rod, allowing the user to safely aim the light. The
glass is special, as it incorporates a filter to reduce unwanted UV light
generated by the halogen bulb. A typical 12-volt halogen bulb used in
these desk lamps can have efficacies ranging from 14 to 18lm/W. Note
as illustrated in Figure 3, this is only the rated efficacy of the bulb and
does not include fixture losses or the losses from the transformer that is
located at the base of the lamp.
High brightness LEDs display significantly higher light conversion
efficiency and therefore produce less waste heat for a given light
output. Less waste heat means the lamp assembly operates much
Figure 4: Custom heatsink for lamp shell
Figure 5: Cree MC-E mounted on sub-mount and connected to heatsink
Figure 5a: Cree MC-E including reflector optic
cooler and does not expose the user to dangerous high temperatures.
The bulb assembly was redesigned with a heat sink (Figure 4) that
fits within the existing plastic housing of the bulb fixture. The cast
aluminum insert shown below was installed in the lamp shell. A
neutral white Cree MCE (code - 000-KE5 – 4000K) was mounted on
a metal core PCB sub-mount to aid in assembly. LED performance
was evaluated with and without secondary optics. This LED was
selected to meet the ENERGY STAR correlated color temperature
(CCT) requirement and had ample flux to far exceed the minimum
requirement of 200 lumens out of the luminaire for the SSL
Residential Desk Lamp requirement. The LED is rated for 370 to 430
lumens at a test current of 350mA and at 25°C. Taking into account
the actual drive current (630mA) and the steady state operating
temperature, the estimated lumen output from the LED is in the
range of 450 to 600 lumens dependent on the heat sink design. The
secondary optic is an off-the-shelf 32 degree reflector (FRC-M1-MCE0R) from FRAEN that is specifically designed for the MC-E.
The LED assembly is powered with an ON Semiconductor
NCP1014LEDGTGEVB LED driver evaluation board. This high efficiency
driver provides the required galvanic isolation and regulated drive
current to power the LED array. It features a universal input power
range from 90 to 265VAC, which allows one design to be used in
multiple regions with only a cabling change for the wall plug. This is
not the case with the existing halogen iron-core transformer, as it is
designed specifically for one range of line voltage.
Figure 6: NCP1014LEDGTGEVB power factor performance at normal load
www.digikey.ca/lighting
29
The evaluation board supports the use of a potentiometer for
variable dimming with a current adjustment range of 100 to 630mA.
The 630mA maximum current was set to stay safely below the
700mA maximum operating current of the Cree MC-E. Figure 6
illustrates a typical curve of the NCP1014LEDGTGEVB LED driver
power factor correction performance as a function in input line
voltage. As illustrated, for the US line voltage range, the power
factor is well above 0.8, which far exceeds the ENERGY STAR
Residential SSL Luminaire requirement of 0.7. The driver also meets
the harmonic content requirements of IEC61000-3-2 Class C.
Shown in Figure 7 is the LED driver board mounted in the base of the
desk lamp. Note that a ballast weight was required to stabilize the
lamp since the original transformer, which provided a counterbalance,
was removed from the base. In a portable desk lamp design optimized
specifically for LEDs, the base would be wider and flatter to stabilize
the lamp assembly without the need for added weight. Figure 8
illustrates the LED driver demo board mounted in a display case with
no counterweight needed to support the head.
Data shown in Table 1 was collected on the desk lamp with the
original 50-watt halogen bulb and a replacement 35-watt halogen
Configuration
Setting
50W Halogen
compared to Cree
MC-E with FRAEN
Reflector Optics
High
35W Halogen
compared to Cree
MC-E without
secondary optics
Low
High
Low
Source
Halogen
LED (IDrive = 630mA)
Halogen
LED (IDrive = 100mA)
Halogen
LED (IDrive = 630mA)
Halogen
LED (IDrive = 100mA)
Illuminance
(lux)
Pin (W)
1462
56.6
2596
10.9
744
40.9
646
1.67
847
40.4
962
10.9
435
29.4
236
1.67
Table 1: Before and after comparison
bulb. Additional measurements made after conversion to the LED
source are shown with and without an additional optical reflector that
redirects stray light to a more central location. Illuminance expressed
in lux is the measurement of luminous intensity on a surface at a
distance of 0.5 meters from the light source.
While a comparison of the illuminance directly below the light
source is important, even distribution of light on the surface is also
significant, and thus, other test points were selected at a 0.25m offset
from the center to characterize the distribution of the light under
normal operation conditions. A summary is shown in Table 2.
Lamp
Illuminance @ 0.25m Offset (lux)
Analysis
50 W Halogen
Left
853
Right
800
Front
727
Mean
793
Sigma
51.6
LED w/optic
580
577
529
562
23.4
35 W Halogen
496
485
443
475
22.8
LED
518
527
490
512
15.7
Table 2: Light distribution comparison
Figure 7: NCP1014LEDGTGEVB constant current driver board mounted in base of desk lamp
In this example, the converted LED desk lamp produced 13 percent
greater illuminance than the 35-watt halogen bulb yet consumed 73
percent less electricity. Interestingly, after reviewing the performance
of the magnetic transformer alone in the original 50-watt desk lamp,
it had approximately 8.3 watts of loss, which is more than the 8.0
watts used by the LEDs in the converted desk lamp. In other words,
the transformer in the base of the original halogen lamp consumed
more power than the LEDs in the converted desk lamp.
Figure 9 displays a comparison of light patterns between the halogen
lamp on the left and the converted LED lamp on the right. The
comparison is between a 35-watt halogen bulb and the LED without
any secondary optics. Note the non-uniform pattern created by
the tubular shape of the GY6.35 halogen bulb. Conversely, the LED
assembly is mounted in a circular recessed area of the heat sink as
shown in Figure 5. This provides a more circular light pattern with
uniform illumination in all directions. High brightness LEDs excel at
providing uniform directional light distribution due to their lambertian
light structure.
Figure 8: Demo lamp with alternate base
30
Recessing the LED inside the heat sink facilitates the possibility of
using secondary optics and avoids possible glare if the lamp were
at eye level. As seen in Table 1, center illuminance increased by 170
percent by incorporating additional reflector optics. Combining LED
performance with a high efficiency driver and an off-the-shelf optic
provides a lighting solution far superior to a halogen luminaire.
constrained by the limitations of halogen bulbs. Effective solutions
are possible when coupled with an appropriate LED driver circuit to
simplify the product design so that, with only minor changes, the
same product could be sold in all regions of the world. Moreover,
the adjustable control allows the user to further optimize the light
output for their specific environmental needs instead of the one or
two light levels of a traditional halogen-based desk lamp. Energy
savings increase substantially at lower brightness settings.
Figure 9: Light pattern of 35W halogen and LED lamps
The surface of the exposed glass plate covering the halogen bulb
measured over 250°C. By comparison, the maximum temperature of
the LED mounting substrate was only 77°C. Since this was a retrofit,
the existing unvented plastic housing for the halogen bulb was
reused. With the plastic cover removed, the LED mount temperature
dropped to less than 63°C; if the product was optimized for LED
operation, the thermal environment would improve further, as the
non-vented cover could be redesigned or removed depending on the
end product design considerations. Note as well, under normal drive
conditions (350mA), the LED mount temperature with the cover on
was 49°C.
The benefits in energy consumption are clear. The higher efficiency,
smaller size and weight, and lower LED power dissipation opens
the door for innovative luminaire designs that have historically been
A driver optimized for universal AC input of 90 to 265VAC allows a
manufacturer to utilize one basic luminaire design for all markets,
and only the power cable would need to be changed by region.
Luminaires based around state-of-the-art LEDs such as the Cree
XLAMP MC-E and driven by high efficiency constant current source
drivers will allow the introduction of new long lasting, energy
saving general lighting products. While this example shows the
performance of a retrofit application, further optimization can be
achieved if the luminaire is designed specifically with LEDs from
the start.
Reference Material
1) ENERGY STAR® SSL Luminaire Specification, Version 1.1
http://www.ENERGY STAR.gov/index.cfm?c=new_specs.ssl_luminaires
2) Cree XLAMP MC-E Specification
http://www.cree.com/products/xlamp_mce.asp
3) Fraen Reflector Optics for Cree MC-E
http://www.fraensrl.com/prodinfo.html
4) ON Semiconductor Design Note DN06051 “Improving the Power Factor of Isolated
Flyback Converters for Residential ENERGY STAR® LED Luminaire Power Supplies”
http://www.onsemi.com/pub_link/Collateral/DN06051-D.PDF
High Performance, Mid-Power LED Provides a Cost-Effective
Lighting Solution without Sacrificing Reliability
The STW8Q2PA from Seoul Semiconductor is a high-performance
Surface Mount LED suitable for use in a wide variety of costconscious lighting applications. This 0.3 watt device provides
high-quality white light, excellent luminus efficacy and great
thermal performance.
Key Features
• 33 lumens at 100mA
• 0.32 W Typical Power
• 95 lm/W at 3500K
• ANSI Color Binning
•
•
•
•
CRI > 80
Thermal Slug
LM-80 Test Report
120° viewing angle
The STW8Q2PA’s low power, high efficacy and low thermal
resistance make it a perfect solution for high-density arrays, strip
lighting and backlighting applications where a more uniform light
distribution is desired. It has tightly controlled flux, color and
voltage binning allowing repeatability and easy blending of the
individual sources within a single fixture.
www.digikey.ca/seoulsemiconductor-lighting
www.digikey.ca/lighting
31
High-Power LED
Drivers Find their Niche
by Tony Armstrong, Linear Technology Corporation
Driving a string of highbrightness LEDs at full
capacity requires careful
design, as does achieving a
wide dimming ratio while at
the same time maintaining the
color spectrum. Fortunately,
there are some integrated
driver packages that are up to
the challenge.
Background
A high-power LED’s light output has already
achieved the critical milestone of 100
lumens/W, with some manufacturers claiming
120 lumens/W in the laboratory. This means
that the LED has now surpassed the CFL
(80 lumens/W) in terms of energy efficiency.
Nevertheless, it is further projected that by
2012, the LED will attain 150 lumens/W
output. Furthermore, with all the current focus
on being “green,” the LED does not contain
any hazardous materials like the CFL, which
has toxic mercury vapor inside the tube.
The cost of LED lighting has come down very
quickly. The cost of individual white-light
diodes, several of which go into an LED bulb
and make up much of the cost, have come
down in price from about $5 a few years ago
to less than $1 in the last twelve months.
Many LED industry analysts predict that over
the course of the next 12 months, LED bulb
replacements for the incandescent light bulb
will be priced at a level that will be acceptable
for the consumer. Some LED manufacturers
have already claimed that they have designed
light-emitting chips that could power an LED
bulb producing light comparable to the 75Watt incandescent bulbs so commonly used in
32
American homes. This type of LED chip usually
requires ~12W of power in order to be able to
output this amount of light.
One key performance feature that an LED
driver IC must have today is the ability to
adequately dim LEDs. Since LEDs are driven
with a constant current, where the DC current
level is proportional to LED brightness, to vary
the LED brightness, there are two methods
of dimming the light by controlling the LED
current. The first method is analog dimming,
where the LED DC current level is reduced
proportionally by reducing the constant LED
current level. Reducing the LED current can
result in a change in LED color or inaccurate
control of the LED current. The second method
is digital or pulse-width-modulation (PWM)
dimming. PWM dimming switches the LED
on and off at a frequency at, or above 100Hz,
which is not perceivable to the human eye.
The PWM dimming duty cycle is proportional
to LED brightness while the on-time LED
current remains at the same level (as set by an
LED driver IC), maintaining constant LED color
during high dimming ratios. This method of
PWM dimming can be used with ratios as high
as 10,000:1 in certain applications.
Specifically, in the case of driving high
brightness (HB) LEDs, LED driver ICs must
be capable of delivering sufficient current
and voltage for many different types of LED
configurations - in a conversion topology
that satisfies both the input voltage range
and required output voltage and current
requirements. Thus, HB LED driver ICs should
ideally have the following features:
• Wide input voltage range – up to 100V
• Wide output voltage range – up to 100V
• High efficiency conversion – up to 95%
• Tightly regulated LED current matching
– less than 2% over temperature
• Low noise, constant frequency operation
– as high as 2.5MHz
• Independent current and dimming
control
• Wide dimming range ratios – up to
10,000:1
• Multiple conversion topologies including
buck, boost, buck-boost and SEPIC
• Many protection features, such as
protection for open LED strings, LED
pin to VOUT shorts and accurate
undervoltage lockout thresholds
• Small compact solution footprints with
minimal external components
HB LED examples
There is no question that the majority
of automotive headlights still use the
incandescent light bulb. However, this
dominance is under pressure from both high
intensity discharge (HID) lamps and HB LED
headlights going forward. HID lamps include
all high intensity discharge lamps used
in general lighting such as High Pressure
Mercury Vapor, High Pressure Sodium, Low
Pressure Sodium and Metal Halide. General
lighting sources are sufficiently bright to
enable a working or living environment in a
room, a building or an external space. This
includes residential lighting, commercial
and industrial lighting, street lighting
and automotive headlights. HID Xenon
lamps were first introduced for use as an
automotive headlight in the late 1990s.
However, they are very expensive to produce
and make, so their use has been limited to
high-end vehicles only. Going forward, due to
the recent introduction of HB LEDs, the use
of such HID Xenon lamps will quickly decline.
Thus, the HB LED headlight will have the
largest growth rate in next decade.
Figure 1: 94% Efficient 25W
White LED Headlamp Driver.
A 25W white LED headlamp can be
configured using an array of 18 LEDs in
series and driving them with 370mA to
to produce the necessary light output.
However, a major obstacle is how to
efficiently and simply drive such a
configuration. One possible solution is to use
the recently introduced LT3956 monolithic
LED driver from Linear Technology. The
LT3956 is a DC-DC converter designed
to operate as a constant-current and
constant-voltage regulator. It is ideally suited
for driving high current, high brightness LEDs
(Figure 1).
The LT3956 features an internal low side
N-channel power MOSFET rated for 84V at 3.3A
and is driven from an internal regulated 7.15V
supply. The fixed frequency, current-mode
architecture results in stable operation over
a wide range of supply and output voltages.
A ground-based referenced voltage feedback
(FB) pin serves as the input for several LED
protection features and also makes it possible
for the converter to operate as a constantvoltage source.
The LT3956 senses the output current at the
high side of the LED string. High side sensing
is the most flexible scheme for driving LEDs,
allowing boost, buck mode or buck-boost
mode configurations. The PWM input
provides LED dimming ratios of up to 3,000:1
Figure 2: LT3743 Typical
Application Schematic
Delivering 20A of LED
Current.
www.digikey.ca/lighting
33
while the CTRL input provides additional
analog dimming capability.
color mixing resolution and allows very fast
current pulsing in laser driver applications.
Nevertheless, there are other end market
applications that have their own unique
needs for high-power HB LED drivers
where LED currents of greater than
10A are necessary. These include DLP
projectors, laser drivers and architectural
lighting. However, delivering currents
greater than 10A can bring a host of
design problems, not least of which is
the thermal management aspects within
the end product itself. To this end, Linear
Technology has developed a unique
LED driver to specifically address the
design considerations in these types
of applications.
The LT3743’s 5.5V to 36V input voltage
range makes it ideal for a wide variety
of applications, including industrial, DLP
projection and architectural lighting. The
LT3743 provides up to 20A of continuous LED
current from a nominal 12V input, delivering
in excess of 80 Watts of power. In pulsed
LED applications, it can deliver up to 40A
of LED current or 160 Watts of peak power
from a 12V input. Efficiencies as high as 95%
eliminate any need for external heat sinking
and significantly simplify the thermal design.
A frequency adjust pin enables the user to
program the frequency between 100kHz and
1MHz so designers can optimize efficiency
while minimizing external component size.
Combined with a 4mm x 5mm QFN or
thermally enhanced TSSOP-28 package, the
LT3743 offers a very compact high-power
LED driver solution. See Figure 2 for its
typical application schematic.
The LT3743 is a synchronous step-down
DC-DC converter designed to deliver
constant current to drive high current LEDs.
It can deliver up to 20A of continuous LED
current or up to 40A of pulsed LED current.
The LT3743’s unique design enables threestate current control, which makes it ideal
for color mixing applications required in DLP
projectors. It also offers very fast (<2µs)
transition times current level. This optimizes
The LT3743 offers both PWM and
CTRL_SELECT dimming, enabling 3,000:1
dimming capability at three LED current
levels. This makes it ideal for color mixing
applications such as those required in DLP
projectors. Similarly, the LT3743’s unique
topology enables it to transition between
two regulated LED current levels in less
than 2µsec, enabling more accurate color
mixing in RGB applications. An LED current
accuracy of ±6% is maintained in order to
ensure the most accurate luminosity of light
output from the LED. Additional features
include output voltage regulation, open-LED
protection, over current protection and a
thermal derating circuit.
Conclusion
Despite the long list of performance
characteristics necessary in HB LED drivers,
the LED being driven by the LED driver has
to be capable of delivering the necessary
lumens of light output from the lowest
possible level of power without causing
significant thermal design constraint.
Fortunately for the designers of high-power
HB LED products, there exists both high
efficacy LEDs and high performance LED
drivers to drive them.
Connect to Digi-Key
Like Never Before.
www.digikey.ca/connect
Digi-Key’s Mobile Applications
With instant access to www.digikey.ca, you
can view more than 500,000 board level
components in stock and our online offering of
more than 1.7 million products. Our application
is designed to help you source the electronic
components you need whether you’re at work,
at home, or anywhere in between!
Become a Fan on Facebook
Become a Digi-Key fan and stay in
touch with us on Facebook. Read our
latest press releases, new and featured
product highlights, as well as contests
and promotions. Check out our new
and featured products highlights. Be
among the first to find out when a new
PTM Online… On Demand® product
training module is available.
See us on YouTube
Catch us on YouTube, where you
will find excerpts from many of our
exclusive PTM Online… On Demand®
product training modules as well as
our aptly named Another Geek Moment
videos. These original videos by
Digi-Key engineers might be just what
you need to inspire your next project.
TM
Follow us on Twitter
Follow @digikey on Twitter to receive daily tweets
about new products and supplies, news and website
enhancements. Join us weekly for your chance to
win during “Monday Mash” and “Thursday Twrivia”.
34
Effective Thermal
Management of LED Arrays
contributed by Bridgelux
The performance, reliability, and design
life of LED arrays depend heavily on proper
thermal management. This article examines
a variety of conduction, convection, and
radiation techniques to help ensure optimal
performance and long life for your next
lighting design.
Introduction
The Bridgelux family of LED Array products delivers high performance,
compact, and cost-effective solid-state lighting solutions to serve the
general lighting market. These products combine the higher efficiency,
lifetime, and reliability benefits of LEDs with the light output levels of
many conventional lighting sources.
As noted in Bridgelux LED Array product datasheets, several
performance characteristics of the LED Array products, including flux,
forward voltage, color, and reliability are dependant upon temperature.
This is a common characteristic for all LEDs, a characteristic of the
semiconductor-based technology. As temperature increases, several
performance parameters experience a temporary and recoverable
shift. With increasing temperature, light output (or flux) decreases,
forward voltage (or Vf) decreases, and the color temperature shifts
towards blue.
Furthermore, absolute maximum ratings, such as maximum case
temperature and maximum junction temperature, must not be
exceeded. Exceeding the absolute maximum ratings, as listed in the
product datasheets, may irreversibly damage the product and cause
permanent shifts in performance.
Optimization of performance and reliability in a lighting system
using Bridgelux LED Arrays requires proper thermal management.
Although a critical design parameter, thermal management is
not as difficult as many would believe. Understanding the basics
allows every lighting designer to optimize his product and meet
specification requirements.
This article describes basic thermal management concepts and
guidelines for proper use of Bridgelux LED Arrays in a lighting
system. Included is an overview of basic heat transfer concepts,
a description of a thermal model, a sample calculation using this
thermal model, a description of various thermal components,
and recommendations for measuring the case temperature of the
Bridgelux LED Array to validate the performance of the thermal
management solution.
Heat generation
When voltage is applied across the junction of an LED, current flows
through the junction generating light. It is a common misconception
that LEDs do not generate heat. While essentially no heat is generated
in the light beam (unlike conventional light sources), heat is generated
at the case of the LED Array and must be effectively managed. As
LEDs are not 100 percent efficient at converting input power to light,
some of the energy is converted into heat and must be transferred to
the ambient. The amount of heat generated from the LED Array that
must be transferred to the ambient may be conservatively estimated
to be the power that is applied to the LED Array, or Pd, and may be
calculated by multiplying the forward voltage (Vf) times the current (If).
This is described in Equation 1.
Equation 1: Power Dissipated
Where:
Pd is the thermal power dissipated
Vf is the forward voltage of the device
If is the current flowing through the device
The maximum dissipated power for each of the Bridgelux LED Array
products is listed in Table 3 of this article. It should be noted that
any input power (Vf * If) that is converted to light is usually ignored
when calculating the total thermal load. Ignoring the devices
thermal efficiency provides a safety margin in the design of the
thermal solution.
www.digikey.ca/lighting
35
Heat generated by additional sources, such as a power supply located
near the LED Array, must also be managed. In order to reduce the
size and cost of the thermal management solution and to minimize
the amount of additional heat added to the system, power supplies
and other heat generating components should not be located in close
proximity to the LED Array.
Thermal path
Heat generated at the LED junction must be transferred to the
ambient via all elements that make up the thermal management
solution. These elements include the LED Array, the thermal interface
material used between the LED Array and heat sink, the heat sink, the
luminaire enclosure (if applicable), and other components that come
in contact with the lighting assembly. These elements transfer heat to
the ambient through conduction, convection, or radiation. These heat
transfer modes will be discussed in greater detail in the next section
of this application note. For a simple thermal management solution
that consists of an LED Array mounted to a heat sink, we consider
that all the heat from the LED junction is typically transferred to the
ambient through the following thermal path:
1. Heat is conducted from the semiconductor chip within the LED
Array to the elements that make up the LED Array, including the
metal board and the silicone resin.
2. Heat is then conducted from the LED Array through a thermal
interface material to the heat sink of the lighting system. This is
a critical component in transferring the heat. We will discuss this
point in detail later.
3. Heat is then conducted through the heat sink.
4. Heat is then convected to the air around the heat sink and is
radiated to the ambient.
Modes of heat transfer
There are three basic modes of heat transfer; conduction, convection,
and radiation. All heat flow is driven by temperature gradients; heat
moves from hot to cold. Each heat transfer mode plays an important
role in transferring heat away from the LED junction to the ambient.
Table 1 provides a summary describing the various heat transfer
modes and the equations that govern them.
The first of these, conduction, is the transfer of heat between
adjacent molecules of a material, usually a solid. Equation 2 provides
the basis for our first set of guidelines for designing meaningful
thermal solutions.
Table 1: Heat transfer modes and governing equations
* In the real world, the ideal models of Physics do not absolutely apply. Typically, engineers assume
that heat radiation does not depend on the sample’s surface conditions. In these cases, the heat
source is known as a “Gray Body."
• Use heat sinks with large surface areas (A).
• Eliminate air gaps and voids between the LED Array and the
heat sink. Air is a very poor conductor of heat. During assembly,
the flat bottom surface of the LED Array should be in full
contact with the flat surface of a heat sink. If air gaps or voids
exist between the two, a thermal interface material should be
used to fill the gaps.
Nature has two great thermal insulators. The first great insulator is
a vacuum, and the second is air. Therefore, it is critical to have a
sufficient volume of the thermal interface material to displace the
air. However, if there is any excess thermal interface material, for
example more than 0.3mm, then the system thermal resistance
will begin to increase. On the surface, this may not appear to
make sense. What one needs to consider is that while the thermal
interface material is better at removing heat than air, it is much
worse than the metal in the heat sink that conducts the heat away
from the LED Array.
Considerations for heat sink design or selection:
• Minimize the distance heat must travel (dx). In practical
terms, this means that if a heat sink is too large, it may loose
effectiveness. However, this distance should not be so small that
a bottleneck is created, constricting the flow of heat.
• Select heat sinks made of materials that have a high thermal
conductivity (k). As a reference, Table 2 compares thermal
conductivity of various metals typically used for heat sinks
and the thermal conductivity of air. Although aluminum is not
as an effective heat conductor as copper, it is frequently the
material of choice as it minimizes the cost and the weight of
the thermal solution.
Table 2: Thermal conductivity of common heat sink materials and air
36
Convection describes heat transfer due to random molecular motion
and bulk motion of a fluid. In other words, convection is the heat
transfer from the heat sink to air (or water) and is directly dependant
upon the amount of flow of the air or water. In the case of the natural
convection in air, where, for example, the movement of air molecules
in not aided by a fan, the convection heat transfer coefficient ranges
from 2 to 15W/m2K.
Thermal model
A simple thermal model or thermal circuit can illustrate the heat
flowing through an LED Array. This model is analogous to an electrical
circuit where heat flow is represented by current, voltages represent
temperatures, heat sources are represented by constant current
sources, and resistors represent thermal resistances. Figure 1 shows
a thermal circuit for a single LED Array mounted to a heat sink.
Forced convection is a result of movement in the fluid (water or air),
which results from other forces, such as the use of a fan or pump. With
forced convection, the convection heat transfer coefficients range from
25 to 250 for gases and from 100 to 20,000 for liquids. Both natural
convection and forced convection may be used to effectively convect
heat away from the LED Array. Equation 3 can be used to develop
guidelines for enabling heat transfer through convection.
Considerations for heat sink design or selection:
• Use heat sinks with the largest surface area (A) that is physically
or economically feasible. As a general rule of thumb, for a wellventilated heat sink, there should be 10 square inches of heat
sink for every 1 watt of power dissipated. The use of pinned heat
sinks, however, is not recommended.
• Orient heat sink fins in a manner that allows hot air to flow
upward and away from the heat sink and cold air to flow onto the
surfaces of the heat sink.
Figure 1: Simple thermal circuit
Where:
Q is heat flowing from hot to cold through the LED
• Avoid constricting the airflow.
Tj is the temperature at the junction of the device
• If possible, use natural convection to transfer heat from the
heat sink to the ambient. Doing so avoids potential reliability
issues of fans.
Tc is the temperature at the case of the LED Array
• If natural convection is insufficient, then consider using fans,
heat pumps, or liquid cooling elements that can dramatically
increase the convection heat transfer coefficient and hence
dramatically increase heat transfer.
• Use heat sinks with surfaces that have high emissivity values.
• Radiation is energy transfer by electromagnetic waves. By
analyzing Equation 4, guidelines for enabling heat transfer
through radiation can be developed.
• Radiation heat transfer has a very strong dependency on
temperature. The hotter the heats sink, the more significant
the heat transfer through radiation. However, as the maximum
case temperature of the LED Array is 105°C, heat transfer
due to radiation is very small when compared to other heat
transfer modes.
Keeping in mind how heat is transferred and consciously optimizing
specific heat transfer modes should make the task of designing a
cost effective and optimally sized thermal management solution
easier. In a typical single LED Array lighting assembly using simple
heat sinks, the amount of heat that is conducted, convected, and
radiated is a function of the heat sink geometry, surface area,
material properties, surface properties, fin geometry (including
thickness and spacing), and heat sink temperature. While this may
sound complicated, the application of effective heat sink design
is straightforward.
Th is the temperature at the point where the heat sink is
attached to the LED Array
Tamb is the ambient air temperature
R jc is the thermal resistance from junction to case of the
LED Array
R ch is the thermal resistance between the case of the LED
Array and the heat sink
R
ha
is the thermal resistance of the heat sink
With the exception of Tj, all temperatures listed above can be easily
measured. For Bridgelux LED Arrays, case temperature measurements
are made in the area noted in the mechanical drawings section of
the product datasheets. Note that this area is on the same surface
of the LED Array as the resin area, providing an easy to measure
location after assembly to the heat sink. Although traditionally case
temperature measurements are conducted on the back side of device,
the difference in temperature between the defined case temperature
measurement point location on the top of the Bridgelux LED Array
and bottom are very small. Bridgelux has characterized the difference
between these two points to be typically 1°C or less and considers
the difference negligible.
The thermal resistances that make up the thermal model may be
calculated and solved for using the following equation:
Equation 5: Thermal Resistance
www.digikey.ca/lighting
37
Where:
R xy is the thermal resistance from x to y, where x and y are
points along the thermal circuit
Tx is the temperature at x
Ty is the temperature at y
Q is heat flow and may be approximated to be Pd (see
Equation 1)
Substituting “x” with “j “and “y” with “c”, we get R jc, or the thermal
resistance from junction to case of the LED Array. R jc values are
included in the Product Data Sheets for all Bridgelux LED Arrays and
therefore do not need to be calculated. Instead, by knowing R jc
values and by using Equation 5, we may solve for Tj, the temperature
at the junction.
Figure 2: Simple thermal circuit of multiple LED Arrays on a single heat sink
By substituting “x” and “y” with appropriate values, both R ch and
R ha may be solved for using Equation 5. When using thermal
interface materials, such as thermal pastes and adhesives, R ch
describes the thermal resistance of the thermal interface material.
Thermal interface materials and their impact on thermal resistance
will be discussed in greater detail later in this article. R ha describes
the thermal resistance of a heat sink or heat sink assembly. Once
maximum operating conditions are known, including drive current,
forward voltage, and maximum ambient temperature, this value is
solved for, and a heat sink may be sized and selected to achieve
this value.
Rules governing a thermal circuit are also analogous to those of
an electric circuit. Equation 6 depicts the method for adding series
thermal resistance values.
Equation 6: Summation of Series Thermal Resistances
Where R
ja
is typically referred to as the system thermal resistance.
Furthermore, when assembling multiple LED Arrays on a single heat
sink, the rule of parallels applies. This is depicted in Equation 7.
Equation 7: Summation of Parallel Thermal Resistances
In Equation 7, “n” refers to the number of LED Arrays mounted
onto a single heat sink and R n jh is the thermal resistance from the
LED junction to the heat sink of each of the individual LED Arrays.
It should be noted that when using this model the total power
(sum of all LED Arrays mounted to the single heat sink) must be
multiplied by the system thermal resistance to calculate the case or
junction temperature.
Thermal Model Example
Design Challenge:
A luminaire is to be designed using the BXRA-C1202-00000 LED
Array driven at 1050 mA. The maximum ambient temperature will
be 40°C and the case temperature cannot exceed 70°C in the
application. A thermal paste, with a thermal conductivity of 4W/mK
and a resulting thermal resistance of 0.07°C per watt has been
selected. Shifts in forward voltage due to temperature are assumed
negligible. Given this information, the thermal resistance of the heat
38
sink required for this design must be calculated and the minimum
surface area for an extruded aluminum heat sink must be determined.
Solution:
To solve for the thermal resistance of the heat sink, we use
equation 5, where “x” is “h” and “y” is “a”:
Here, Ta is 40°C and Th is unknown at this time, but may be solved
for. Q is the dissipated power, which is the product of the voltage and
current. In order to design a thermal management system to account
for the entire range of product performance, the maximum forward
voltage of the LED Array must be used (see Table 3 or the relevant
product datasheet).
As the thermal resistance of the thermal interface material is known,
to solve for Th we use equation 5, substituting “x” with “c” and “y”
with “h”:
Table 3 illustrates that when relying on natural convection to transfer
heat to the ambient, the minimum required surface area for an
extruded aluminum heat sink with a black anodized finish is 10 in2 per
watt of dissipated power. This area may be provided as a flat plate or
via a finned extruded heat sink to minimize the volume of the thermal
management solution. The required heat sink surface area must be
determined and validated by the luminaire designer and will depend
on many luminaire design variables, including fin orientation and cold
and hot air flow paths.
Solving for Th, we determine:
Now solving for R
ha
:
The surface area of a black anodized aluminum extruded heat sink
with at thermal resistance of 2°C/W is estimated to be at least 101.5
square inches for a heat sink with fins that are oriented vertically or
145 square inches if the fins are oriented non-vertically. While these
values may appear large, this is the total surface area required and
may consist of surfaces that make up the luminaire itself such as the
housing and reflector, in addition to the surfaces that make-up the
heat sink. The final required surface area of the heat sink depends on
many variables, including, but not limited to, fin orientation, the ability
of hot air to move away from the heat sink, the ability of cold air to
enter and flow through the heat sink, and the existence of other heat
sources near the LED Array.
Thermal solution components – heat sinks
There are many commercially available components, including heat
sinks and heat pipes, which may be used with Bridgelux LED Arrays.
The most commonly used components are heat sinks, typically made
of aluminum or copper. Heat sinks conduct heat from a heat source
and then convect the heat to the ambient. The size of a required
heat sink depends on many variables, including the temperature
requirements for the application (such as maximum ambient and case
temperature), the material of the heat sink, the surface characteristics
of the heat sink, and the physical constraints for the application.
Table 3 lists minimum surface areas and sample dimensions for
reference heat sinks in simple cases where a single LED Array is used
in a light fixture, and natural convection and radiation are used to
transfer heat to the ambient. Calculated thermal resistance values of
the reference heat sinks (Rha) in Table 3 are based on the following:
In some applications, space requirements may not allow for such
a large heat sink. In these cases, designers should consider using
forced convection elements, such as fans. Consult heat sink suppliers
to explore forced convection heat sink options. Heat sink suppliers
can provide detailed technical data on their heat sinks to customers.
Data supplied includes heat sink temperature rise above ambient
as a function of air flow speed and heat sink thermal resistance as
a function of air flow speed. This information aids in the design and
heat sink selection process.
Table 3a: Critical parameters and sample heat sink dimensions for Warm White LED
Arrays by part number
• The maximum ambient temperature of the application is 40°C.
• The maximum case temperature of the array is 70°C.
• The thermal interface material used has a thermal conductivity
of 4Wm/K (like that of Dow Corning TC-5022). The thermal
resistance contributions for various thermal interface materials
are shown in Table 5. These values vary by LED Array product
and range from 0.19°C/W for the LS Array Series products to
0.035°C/W for the RS Array Series products.
• Maximum dissipated power values have been calculated
using forward voltage values reached at elevated junction
temperatures, those reached when the LED Arrays reach a case
temperature of 70°C.
The surface area and dimension values listed in Table 3 are based on
the reference heat sink (Rha) values.
Table 3b: Critical parameters and sample heat sink dimensions for Neutral White LED
Arrays by part number
Impact of fin orientation on heat sink performance
Thermal performance presented by heat sink manufacturers
usually pertains to the most favorable heat sink orientation, which
is with the fins of the heat sink oriented vertically (see Table 4). In
this orientation, hot air rises readily, allowing cool air to circulate
through the fins. Other orientations give varying results. Rules of
thumb on the impact of varying fin orientation are shown in Table 4.
However, please note that the actual performance of a heat sink is
www.digikey.ca/lighting
39
Table 3c: Critical parameters and sample heat sink dimensions for Cool White LED
Arrays by part number
dependant upon many variables such as heat sink location within an
assembly, the location of other heat generating elements such as a
power supply, effective airflow, fin spacing, fin height, fin thickness,
base thickness, base surface area, shape, fin geometry, and overall
length. Consequently, the effectiveness of a heat sink mounted
in varying orientations is not a fixed number, and depends on the
inter-relationship between many variables. A much larger poorly
positioned heat sink will not be as effective as a properly oriented
thermal system.
To illustrate the effect of fin orientation and limiting airflow,
measurements of case temperature and heat sink temperature
of a BXRA-C1200 LED Array were conducted. The LED Array was
mounted to a 121 x 121 x 32 mm heat sink with a surface area of
774 cm2. A thin layer of thermal paste was applied between the LED
Array and the heat sink. The dissipated power for this LED Array
was 18.4 watts. The heat sink was then placed in the orientations
listed in Table 4. When sitting on a flat surface, the heat sink sat
on a 1-inch thick foam board to minimize conduction to the bottom
surface. Measurements were taken after one hour of operation in
each configuration.
Some results vary from the guidelines listed in Table 4, illustrating
the need to measure a proposed or calculated design to gauge true
effectiveness of a thermal management solution. There may be many
variables that cause differences in results, such as uncontrolled
forced convection from central heating units and differing amounts of
heat conduction to surroundings. All must be taken into consideration
when designing a thermal solution.
Note that the worst performance was achieved by placing the LED
Array in a box and limiting both conduction and natural convection
airflow paths. The thermal resistance of the system could be
significantly improved in this case by providing a convective path to
ambient. Typically LED-based luminaires are constructed in such a
manner as to conduct the heat to the exterior case of the luminaire,
significantly improving the thermal performance of the system.
As mentioned previously, fans may be used to dramatically increase
the convection heat transfer coefficient and hence dramatically
increase heat transfer from the heat sink to the ambient. If fans are
used, make sure that they pull in the cold ambient air to the heat sink
surfaces and that they push hot air away from the heat sink. Also,
40
Table 4: Impact of heat sink fin orientation on heat sink effectiveness, general rules
and experimental results
make sure the lifetime of any active thermal management system
is evaluated to ensure it is suitable, matching, or exceeding the
expected life of the lighting assembly.
Thermal interface management
To ensure heat flow from the LED Array to a heat sink, pay close
attention to air gaps or voids located between the bottom of the LED
Array and the heat sink. Such voids will significantly impede the
flow of heat and therefore must be eliminated. The use of thermal
interface materials, such as thermal greases, pastes, or adhesives, is
recommended to ensure that air gaps and voids are eliminated.
When selecting a thermal interface material, many factors must
be considered. These include thermal conductivity, operating
temperature range, cost, workability (dispensability for pastes),
electrical conduction, and the ability to control the thickness of the
bond line. If using a paste, the amount of paste that is dispensed
should be enough to cover the entire base of the LED Array, but not so
much as to result in a thick bond line, which will increase the thermal
resistance. Application of excess thermal interface material can also
create side fillets. Fillet size should be kept at a minimum and must
not touch the top of the LED Array.
The following equation shows the relationship between thermal
interface material thickness and thermal resistance:
Equation 8: Thermal resistance of an interface material
Where:
L is the thermal interface material thickness (mm)
K is the thermal conductivity (W/m-K)
Ac is the contact area (mm2)
For design purposes, the thermal interface material thickness typically
ranges from 0.15 to 0.30mm, depending on the LED Array product
and the thermal interface material selected. This range of thicknesses
assumes that the planarity of the bottom surface of the LED Array is
maintained at or below 0.1mm for all products except RS Array Series
products. The planarity for the RS Array Series products is maintained
at or below 0.25mm. Hence, when selecting a thermal interface
material ensure that the thickness of the material is sufficient to fill
gaps between the base of the LED Array and the heat sink while at
the same time minimizing the thickness. Thermal conductivities of
interface materials vary from product to product. Table 5 lists a few
examples and the impact on the thermal resistance of a lighting
system for various Bridgelux LED Array products.
All thermal interface materials must be applied in a way that ensures
the entire bottom of the LED Array is covered and that air gaps or
voids between the LED Array and the heat sink are filled. One method
of achieving this, which may be used in volume production, is to
silkscreen precise quantities of a dispensable thermal interface
material on a heat sink surface prior to attaching the LED Array.
Silkscreen templates are made in sizes that mirror the base of the
LED Array or are scaled to a slightly larger size. The final thickness
of the interface material is controlled by the force applied on the LED
Array from the top when pressing it down on the thermal interface
material. These forces can vary from a few hundred grams, which
pick and place equipment is capable of handling, to several kilograms,
which would require special tooling. If pick and place equipment is
used for this processes, consider using “Scrub” mode, while pressing
the LED Array onto the heat sink. In all cases, care must be taken to
avoid contact with the resin area of the LED Array during assembly.
Please consult application note AN11 – Handling and Assembly of
Bridgelux LED Arrays, for further information.
The customer must evaluate the performance of the thermal interface
material to ensure adequacy in terms of thermal performance,
manufacturability, and durability.
Use of current derating curves
Current derating curves are included in the Bridgelux LED Array
Product Data Sheets. These curves provide guidance to customers in
developing effective thermal management solutions that meet system
design requirements.
A derating curve is included for each Bridgelux LED Array product.
When using these graphs, the required system thermal resistance
can be estimated when the LED Array is used at the rated test current
under various ambient conditions. An example of one of these derating
curves is included in Figure 3. Please consult the relevant product
datasheets for the most recent versions of these derating curves.
Figure 3: Derating Curve for BXRA-W0400, 900 mA Drive Current
Table 5: Thermal conductivity and estimated thermal resistance of some thermal
interface materials
Notes for Table 5:
1. For all products except RS Array series products, flatness of the back surface
of the LED Array is maintained at < 0.1 mm across the LED Array. For RS Array
Series products, flatness is maintained at < 0.25 mm across the LED Array.
Examples shown assume this worst-case scenario.
2. Thermal interface material thickness is assumed to be 0.05mm for all materials
except the Berquist A1500 Sil-Pad. The Berquist A1500 Sil-Pad is 0.254mm thick.
The thermal resistance values indicated on the derating curves are
total system thermal resistance values (junction to ambient). Although
limited options are included, it is possible to interpolate between
these curves for approximation purposes. The safest approach,
however, is to calculate the required system thermal resistance using
the equations contained in this application note.
The thermal resistance for the BXRA-W0400 product from the
Bridgelux LED Array product datasheet is listed as 1.0°C/W. If in a
given lighting system the thermal resistance from case to ambient is
www.digikey.ca/lighting
41
designed to deliver 3°C/W, the curve in Figure 3 labeled 4°C/W would
be applicable for this lighting system (sum of junction to case and
case to ambient thermal resistance values).
• Note that it is critical that the entire thermocouple bead be
secured tightly against the case. There should be no air gaps
between the thermocouple tip and the case of the LED Array.
In this example, as long as the ambient temperature is maintained
below 80°C, the maximum temperature ratings of the product will not
be violated at the rated forward current of 900mA. If, however, the
ambient temperature was to rise to 85°C, the forward current would
need to be reduced to 700mA based on the 4°C/W system thermal
resistance. Alternatively, a heat sink could be designed to deliver a
case-to-ambient thermal resistance of 2°C/W. This would result in a
system thermal resistance of 3°C/W, allowing for 900mA operation at
the 85°C ambient condition.
After turning on the LED Array, Tcase will increase with time as the
assembly heats up. Eventually, the lighting assembly should reach
a steady state temperature. The time required to reach a steady
state temperature depends on the time-constant of the assembly,
but is likely to be in the range of 45 minutes to an hour. Maintain
the Bridgelux LED Array Tcase at or below the maximum case
temperatures listed in the product datasheet to ensure functionality
and reliability.
Measuring effectiveness of a thermal solution
After a thermal management solution has been designed, it is
critical to experimentally validate the effectiveness of the solution.
This is typically done by building a prototype, simulating the worst
case use conditions, and measuring Tcase. When simulating worst
case use conditions ensure the following:
• Convection conditions are realistic.
Video:
Digi-Key Lighting Technology
Symposium - Laird Technologies
Dr. Rich Hill (Laird, Director of Technology ) overviews the
different methods of LED thermal management.
• Material properties and dimensions, including wall thicknesses,
surface areas, and component sizes are representative of
the design.
• Surface properties, including color and roughness properties,
are representative of the design.
• Additional heat sources that may impact the thermal
performance of the device are included (such as a power
supply that is placed inside a luminaire enclosure or imbedded
into a heat sink).
Once the representative prototype is buil, and realistic use conditions
are simulated, Tcase may be measured to validate the design.
Special care is required when measuring the case temperature
to ensure an accurate measurement. The following approach is
recommended to minimize measurement errors for attaching the
thermocouple to the case temperature measurement point of the
LED Array:
• Use 36 gauge or smaller diameter K-type thermocouples.
• Ensure that the thermocouple is properly calibrated.
• Attach the thermocouple bead, or junction, to the area on top
of the LED Array in the prescribed area (refer to the mechanical
drawings section in the relevant product datasheet).
• Attach the thermocouple to the LED Array using an adhesive
that has high thermal conductivity. To do this, first place the
thermocouple bead on to the prescribed area. Temporarily
secure the thermocouple using Kapton tape. Next, using the
back of a thin diameter wood stick, such as a tooth pick or the
wooden end of a cotton swab, press on the thermocouple bead,
ensuring contact with the LED Array board. Lastly, apply a small
amount of a fast curing, low viscosity, and thermally conductive
adhesive around the base of the thermocouple bead. Allow
the adhesive to cure. Remove as much as the wooden stick
as possible.
42
www.digikey.ca/video-lighting
LED Thermal Management
Bergquist
This product training module explains how insulated
metal substrate boards work and provides an overview
of the T-Clad
PA-Bond-Ply 450
thermal interface
material.
www.digikey.ca/ptm
Compensating and
Measuring the Control Loop
of a High-Power LED Driver
by Jeff Falin, Texas Instruments
A mathematical model is always helpful
in determining the optimal compensation
components for a particular design. However,
compensating the loop of a WLED currentregulating boost converter is a bit different
than compensating the same converter
configured to regulate voltage.
Measuring the control loop of a WLED current regulating boost
converter with traditional methods is cumbersome because of
low impedance at the feedback (FB) pin and the lack of a top-side
FB resistor. In “Designer’s Series, Part V: Current-Mode Control
Modeling”1, Ray Ridley has presented a simplified, small-signal
control-loop model for a boost converter with current-mode control.
The following explains how to modify Ridley’s model so that it fits
a WLED current-regulating boost converter; it also explains how to
measure the boost converter’s control loop.
Loop components
As shown in Figure 1, any adjustable DC/DC converter can be
modified to provide a higher or lower regulated output voltage from
an input voltage. In this configuration, if we assume ROUT is a purely
resistive load, then VOUT = IOUT × ROUT. When used to power LEDs, a
DC/DC converter actually controls the current through the LEDs by
regulating the voltage across the low-side FB resistor as shown in
Figure 2. Because the load itself (the LEDs) replaces the upper FB
resistor, the traditional small-signal control-loop equations no longer
apply. The DC load resistance is:
OUT
ILED
RR
EQEQ VV
OUT
/ I/LED
(1)
VOUT n nVFWD
VFWD VFB
VFB
VOUT
(2)
with
VFWD, taken either from the diodes’ datasheet or from measurements, is
the forward voltage at ILED; and n is the number of LEDs in the string.
Figure 1: Adjustable DC/DC converter used to regulate voltage.
Figure 2: Adjustable DC/DC converter used to regulate current through LEDs.
LED. Being a dynamic (or small-signal) quantity, rD is defined as the
change in voltage divided by the change in current, or rD = ∆VFWD/∆ILED.
To extract rD from Figure 3, we simply drive a straight tangent line
from the VFWD and ILED for the application and compute the slope. For
example, using the dotted tangent line in Figure 3, we get rD = (3.5 –
2.0 V)/(1.000 – 0.010 A) = 1.51Ω at ILED = 350mA.
s s Small-signal
s smodel
However, from a small-signal standpoint, the load resistance consists
11
11example of a small-signal model, the TPS61165 peakAs
an
of REQ as well as the dynamic resistances of1the
1 LEDs,
RHP
DD rD, at the ILED. z z current-mode
RHP
converter driving three series OSRAM LW W5SM parts
Gp(s)provide
R Rtypical values of r at various
Gp(s)
KK
While some LED manufacturers
2
D
RiRi
s
s Figure
will
be
used.
s
s
s 2s 4a shows an equivalent small-signal model of
current levels, the best way to determine rD is to extract it1from
1 the
1
1a current-regulating2 boost
2
converter, while Figure 4b shows a more
QQ
p pn n
n n
typical LED I-V curve, which all manufacturers provide. Figure 3 p p
simplified
model.
shows an example I-V curve of an OSRAM LW W5SM high-power
z z
11
ESR CC
OUT
ESR
OUT
www.digikey.ca/lighting
43
Evaluation of CurrentRegulating Boost Converter
Term
Evaluation of VoltageRegulating Boost Convertor
Table 1: Differences in Equation 3 terms for two converter models.
REQ
REQ
VOUT / ILED
VOUT / ILED
VOUT n VFWD VFB
EQ
VnOUT
/ ILEDVFB
VROUT
VFWD
s
n VFWD VFB
1
1
D
sz
R
EQ
V
OUT / ILED1
Gp(s) KR
1
1
REQ VOUT / ILED
1 RiD
z
s
s
R
Gp(s)
1
1
Figure 3: I-V curve
of OSRAMKLW
W5SM.
Ri
Qps
sp s
VOUT n VFWD VFB
1 1
1 1
VOUT n 3 Vshows
FWD V
Equation
aFBfrequency-based
(s-domain)
model
for
computing
1 D
Qp
p
z
KR
DC gain in Gp(s)
both the current-regulating
and the voltage-regulating
Ri
s
s
boost converters:
1
1
1
Qp
p
s
s s
z
1
1 1
1
ESR
1COUT 1 D z
1 D
RHP z
Gp(s) Gp(s)
KzR
R
K
(3)
Ri COUT Ri
ESR
s
ss
s2
s
1
1
2
1 Qp n
p
n1
1
Qp
p
z
ESR COUT
VOUT
where the common variables are:
z
Qp
and
n
RHP
1
ESR Qp
COUT
1
Se
z
Qp ESR 1 COUT
Sn
Se
1
Sn
Qp
Se
1 1
Sn
n Se
f
SW
1
1 D 0.5
n Sn
fSW
Qp
Se
1
n
fSW Sn
REQ
fSWRHP
1 D
REQ2
RHP
1 D 2
n
fSWREQ
EQ
RRHP
2
1 D 2 L1 D
RHP
1
1
1 D
1 D
1
1 D
1
1 D
L
L
0.5
0.5
0.5
0.5
RHP
s
Since the
value of RSENSE is typically much lower than that of ROUT in
s2
a converter
configured to regulate voltage, the gain for a current2
ns
sn2 converter, where R = REQ, will almost always be lower
regulating
OUT
2
than the
nRHP
n gain for a voltage-regulating converter.
RHP
2
s
Measuring
the loop
2
To
measure
the control loop gain and phase of a voltage-regulating
n
n
s
converter, a network or dedicated loop-gain/phase analyzer typically
RHP
uses a 1:1 transformer to inject a small signal into the loop via a small
resistance
s 2 (RINJ). The analyzer then measures and compares, over
frequency,
2 the injected signal at point A to the returned signal at point
n
n
R and reports the ratio in terms of amplitude difference (gain) and time
delay (phase). This resistance can be inserted anywhere in the loop as
long as point A has much lower impedance than point R; otherwise, the
injected signal will be too large and disturb the converter’s operating
point. As shown in Figure 5, the high-impedance node where the
FB resistors sense the output voltage at the output capacitor (lowimpedance node) is the typical place for such a resistor.
In a current-regulating configuration, with the load itself being the
upper FB resistor, the injection resistor cannot be inserted in series
with the LEDs. The converter’s operating point must first be changed
so the resistor can be inserted between the FB pin and the sense
resistor as shown in Figure 6. In some cases, a non-inverting, unitygain buffer amplifier may be necessary to lower the impedance at the
injection point and reduce measurement noise.
L
REQ
1 D 2 L
Figure 4: Small-signal model of current-regulating boost converter.
44
The duty cycle, D, and the modified values for VOUT and REQ are
computed the same way for both circuits. Sn and Se are the natural
inductor and compensation slopes, respectively, for the boost
converter; and fSW is the switching frequency. The only real differences
between the small-signal model for the voltage-regulating boost
converter and the model for a current-regulating boost converter is
the resistance KR, which multiplies by the transconductance term,
s(1 – D)/Ri, and the dominant pole, p. These differences are
summarized in Table 1. See Reference 1 for more information.
Figure 5: Control-loop measurement for voltage-regulating converter.
With the measurement setup in Figure 6 but without the amplifier, and
with RINJ = 51.1Ω, a Venable loop analyzer was used to measure the
loop. The model of a current-regulating converter was constructed in
Mathcad® using the datasheet design parameters of the TPS61170,
which has the same core as the TPS61165. With VIN = 5V and ILED
set to 350mA, the model gives the predicted loop response for
the TPS61165EVM as shown in Figure 7, which provides an easy
comparison with measured data.
We can easily explain the differences between the measured and
simulated gain by observing variations in the WLED dynamic resistance
Figure 7: Measured and simulated loop gain and phase at VIN = 5V and ILED = 350mA.
and using the typical LED I-V curve as well as chip-to-chip variations
in the IC’s amplifier gain.
Conclusion
While not exact, the mathematical model gives the designer a good
starting point for designing the compensation of a WLED currentregulating boost converter. In addition, the designer can measure the
control loop with one of the alternate methods.
Reference
1. Ray Ridley. (2006). “Designer’s Series, Part V: Current-Mode Control Modeling.”
Switching Power Magazine [Online]. Available: http://www.switchingpowermagazine.
com/downloads/5%20Current%20Mode%20Control%20Modeling.pdf
Figure 6: Control-loop measurement for current-regulating converter.
Pyrolytic Graphite Sheet
State-of-the-art material that answers the demands
of today’s applications.
High Thermal-Conductivity Material Opens Up Future
Opportunities.
Panasonic Electronic Components’ PGS or Pyrolytic Graphite
Sheet is a heat spreading material with high thermal conductivity.
This material is lightweight and highly flexible. It can be cut into
custom shapes, making it extremely useful for applications with
limited space. In addition, it has excellent thermal conductivity
properties, 600 to 800W/(m-K), which is twice as high as copper
and three times as high as aluminum. PGS can also be used as a
clean and efficient alternative to silicone grease.
PGS is extremely effective in the development of equipment
where temperature control must be attained in a package that
is restricted due to downsizing, and where the packaging is
lightweight. PGS is an ideal solution where there are I.C. heat
sources such as CPUs, processors, power amplifiers, etc. that
create hot spots that need to be diffused.
Key Features
• Excellent thermal
conductivity in its plane (600
to 1700W/m.K) or 2 to 4
times as high as copper, 3 to
6 times as high as aluminum
• Lightweight and ultra-thin:
Specific gravity: 0.85 to
2.1g/cm3 (1/4 to 1/10 of
copper, 1/1.3 to 1/3 of
aluminum in density)
Applications
• Telecommunications
• Lighting
• Computer and peripherals
• Power conversion
• Flexible and easy to be cut
or trimmed (withstands
repeated bending)
• Low thermal resistance
• Shielding effect
• Maintenance-free
• Long life
• RoHS compliant
• Between heat generating
semiconductors or
magnetic components and
a heat sink
www.digikey.ca/panasonic-lighting
www.digikey.ca/lighting
45
LED Luminaire
Design Guide
contributed by Cree, Inc.
This article provides guidelines for the process
of designing high-power LEDs into luminaires.
While it uses an indoor luminaire design as an
example, the process described can be applied
to the design of any LED luminaire.
Lighting-class LEDs are now available that deliver the brightness,
efficacy, lifetime, color temperatures, and white-point stability
required for general illumination. As a result, these LEDs are being
designed into most general lighting applications, including roadway,
parking area, and indoor directional lighting. LED-based luminaires
reduce total-cost-of-ownership (TCO) in these applications through
maintenance avoidance (since LEDs last much longer than traditional
lamps) and reduced energy costs.
There are over 20 billion light fixtures using incandescent, halogen, or
fluorescent lamps worldwide. Many of these fixtures are being used
for directional light applications but are based on lamps that put out
light in all directions. The United States Department of Energy (DOE)
states that recessed downlights are the most commonly installed
luminaire type in new residential construction.1 In addition, the DOE
reports that downlights using non-reflector lamps are typically only
50 percent efficient, meaning half the light produced by the lamp is
wasted inside the fixture.
In contrast, lighting-class LEDs offer efficient, directional light that
lasts at least 50,000 hours. Indoor luminaires designed to take
advantage of all the benefits of lighting-class LEDs can:
luminaire design will have better optical, thermal, and electrical
performance than the retrofit lamp, since the existing fixture does not
constrain the design. It is up to the designer to decide whether the total
system performance of a new luminaire or the convenience of a retrofit
lamp is more important in the target application.
Target Existing Luminaires
If the target application is better served by creating a new LED
luminaire, then designing the luminaire’s light output to match or
exceed an existing luminaire has several advantages. First, an existing
design is already optimized to target a known application and can
provide guidance for setting the design goals around light output,
cost, and operating environment. Secondly, an existing design is
already in an accepted form factor. Switching to the LED luminaire is
easier for the end user if the form factors are the same.
Unfortunately, some LED luminaire manufacturers are misreporting or
inflating claims of LED luminaire efficacy and lifetime characteristics.
The lighting industry saw similar problems during the early years
of CFL replacement bulbs. The lack of industry standards and wide
variations in early product quality delayed the adoption of CFL
technology for many years. The United States Department of Energy is
aware that the same standards and quality problems may exist with
early LED luminaires and that these problems may delay the adoption
of LED lighting in a similar fashion. In response, it launched the DOE
SSL Commercial Product Testing Program (CPTP) to test the claims of
LED-luminaire makers. This program anonymously tests LED-based
luminaires for the following four characteristics:
• Luminaire light output (lumens)
• Exceed the efficacy of any incandescent and halogen luminaire
• Luminaire efficacy (lumens per watt)
• Match the performance of even the best CFL (compact
fluorescent) recessed downlights
• Correlated color temperature (degrees Kelvin)
• Provide a lifetime five to fifty times longer than these lamps
before requiring maintenance
• Reduce the environmental impact of light: no mercury, less
power-plant pollution, and less landfill waste
Design approach
Luminaires or Lamps?
Designing LEDs into general illumination requires a choice between
designing either a complete luminaire based on LEDs or an LED-based
lamp meant to install into an existing fixture. Generally, a complete
46
• Color-rendering index
DOE’s CPTP sets a good precedent for LED luminaire design by
focusing on the usable light output of a luminaire — not just the light
output of the light source.
The Idea of a Lamp May Be Outdated
The long lifetime of LED light may make the idea of a lamp outdated.
Lighting-class LEDs do not fail catastrophically like light bulbs. Instead,
they can provide at least 50,000 hours of useful lifetime before they
gracefully degrade below 70 percent of their initial light output (also called
lumen maintenance). That is equal to 5.7 years if left on continuously!
Importance
Critical
Potentially
Important
Characteristic
Units
Luminous flux
lumens (lm)
Illuminance distribution
footcandle (fc)
Electrical power consumption
watts (W)
Luminaire aesthetics
-
Price
-
Lifetime
hours
Operating temperatures
°C
Operating humidity
% RH
Color temperature
K
CRI
-
Manufacturability
-
Ease of installation
-
Form factor
-
Chart 1: What 50,000 hours means in practical terms.
Table 2: Important characteristics of target luminaire.
However, in most lighting situations, the lights are switched off
regularly. This off period can extend the lifetime of the LED well
past three decades, as shown in Chart 1. After the years it will take
for an LED luminaire to “burn out,” LED lighting technology will be
brighter, more efficient, and probably offer TCO savings over the
older LED luminaire.
CFL downlight. This design process is repeatable for all kinds of
luminaires and not just the example included.
Keep in mind how much environmental impact was avoided over that
50,000 hours of LED luminaire lifetime:
• At least 25 fewer incandescent bulbs were sent to landfills and five
times less energy used. (About 50 percent of power in the United
States comes from burning coal, which releases mercury into the air.)
• Or, at least five fewer CFL bulbs containing mercury sent for disposal.
As mentioned previously, maintenance avoidance is an important benefit
for LED luminaires. Therefore, designing the LED luminaire to deliver
maximum lifetime and provide TCO savings is an excellent strategy to
overcome the hurdle of the higher initial cost of LED-based luminaires.
Design process
Table 1 lists a general process for designing high-power LEDs into a
luminaire. The rest of the article walks through these design steps in
order. To better illustrate these design concepts, this article includes
example calculations for an LED luminaire meant to replace a 23-W
Step
Explanation
1. Define lighting requirements
•
The design goals should be based either on an existing fixture’s
performance or on the application’s lighting requirements.
2. Define design goals
•
Specify design goals, which will be based on the application’s lighting
requirements.
The designer should specify any other goals that will influence the design,
such as special optical requirements or being
able to withstand high temps.
•
3. Estimate efficiencies of
the optical, thermal and
electrical systems
•
•
•
1. Define lighting requirements
The LED luminaire must meet or exceed the lighting requirements for
the target application. Therefore, the lighting requirements must be
defined before establishing the design goals. For some applications,
there are existing lighting standards that will define the requirements
directly. For other applications, a good approach is to characterize an
existing luminaire.
Table 2 lists important characteristics to consider when characterizing
an existing luminaire. The light output and power characteristics
are always critical, while other characteristics may or may not be
important, depending on the application.
All lighting companies can provide data files or documentation that
detail the “critical” characteristics for each of their fixtures. The
“potentially important” characteristics are more subjective or may not
be listed in the manufacturer’s documentation. In this case, it is up to
the designer to characterize the existing luminaire.
Figure 1 illustrates the critical characteristics for the example
CFL downlight. Table 3 shows the full characterization of the
existing luminaire.
2. Define design goals
With the lighting requirements defined, the design goals for the LED
luminaire can be set. Just as when defining the lighting requirements,
the critical design goals will be related to light output and power
Design goals will place constraints on the optical,
thermal, and electrical systems
Good estimations of efficiencies of each system can be made based on
these constraints.
The combination of lighting goals and system efficiencies will drive the
number of LEDs needed in the luminaire.
4. Calculate the number
of LEDs needed
•
Based on the design goals and estimated losses, the designer can
calculate the number of LEDs needed to meet the design goals.
5. Consider all design
possibilities and
choose the best
•
•
As with any design, there are many different ways to best achieve the goals.
LED lighting is still a new field, so assumptions that work for conventional
lighting sources may not apply to LED lighting design.
6. Complete the final steps
•
•
•
•
•
Complete circuit board layout.
Test design choices by building a prototype luminaire.
Make sure the design achieves all the design goals.
Use the prototype design to further refine the luminaire design.
Record observations and ideas for improvement.
Table 1: Process for designing high-power LEDs into luminaires.
Figure 1: Critical characteristics for example CFL downlight.
www.digikey.ca/lighting
47
Part
Lamp
Fixture
Characteristic
Technology
Light output
Power
Efficacy
Lifetime
Average selling price
Coefficient of utilization (CU)
Light output
Power (excluding ballast)
Efficacy
ASP
Unit
lm
W
lm/W
hrs
%
lm
W
lm/W
-
Value
CFL
1,500
23
65
10,000
US $6
54
810
23
35
US $20
Table 3: Example CFL Downlight Characterization.
Characteristic
Light output
Illuminance distribution
Power (excluding driver)
Efficacy
Lifetime
CCT
CRI
Circular opening
Maximum ambient temp.
High volume BOM cost
Unit
lm
fc
W
lm/W
hours
K
in
°C
-
Goal
810
23
35
50,000
4,000
75
6
55
US $50
Figure 2: Comparison of CFL & LED Coefficient of Utilization.
Notes
Same shape as target; matching or better illuminance
Commercial building with vented ceilings
-
Table 4: Design Goals for CFL Downlight Replacement Luminaire.
consumption. Make sure to include other design goals that may
also be important for the target application, including operating
environment, bill-of-materials (BOM) cost, and lifetime.
Table 4 lists the design goals for the example LED luminaire.
3. Estimate efficiencies of the optical,
thermal & electrical Systems
One of the most important parameters in the design process is how
many LEDs are required to meet the design goals. The rest of the
design decisions revolve around the number of LEDs, since it directly
impacts the light output, power consumption, and cost of the luminaire.
It is tempting to calculate the number of LEDs by looking at the typical
luminous flux listed on an LED’s data sheet and divide the target
lumens from the design goals by that number. However, this approach
is too simplified and will lead to a design that will not meet the
application’s lighting requirements.
An LED’s luminous flux depends on a variety of factors, including
drive current and junction temperature. To accurately calculate the
necessary number of LEDs, the inefficiencies of the optical, thermal,
and electrical systems must be estimated first. Personal experience
with previous prototypes or the example numbers provided in this
article can serve as a guide to estimate these losses. This section
walks through the process of estimating these system losses.
Optical System Efficiency
Optical system efficacy is estimated by examining light loss. There are
two main sources of light loss to estimate:
Chart 2: Light Ouput vs. Angle for CFL Fixture & XLamp XR-E LED.
by the fixture housing, while some is reflected back into the fixture.
The efficiency of the fixture is dictated by placement of the light
source, the shape of the fixture housing, and materials used in the
fixture housing. As Figure 2 shows, the directional nature of LED
light enables much higher fixture efficiencies than is possible with
omni-directional light sources.
For the example luminaire, there will only be secondary optics loss
if the luminaire requires secondary optics. The main purpose of
secondary optics is to change the light output pattern of the LED.
Chart 2 (above) compares the beam angle of the Cree® XLamp® XR-E
LED to the light output pattern of the target fixture. The beam angle of
the bare LED is similar enough to the target fixture that no secondary
optic is required. Therefore, there is no optical loss due to secondary
optics for the example luminaire.
To calculate fixture loss for the CFL example, we assumed 85 percent
reflectivity for the fixture reflector cup and that 60 percent of the light
will hit the reflector cup. Therefore, the optical efficiency will be:
Optical Efficiency = (100% x 40%) + (85% x 60%)
Optical Efficiency = 91%
1. Secondary Optics
Secondary optics are any optical system that is not part of the LED
itself, such as a lens or diffuser placed over the LED. The losses
associated with secondary optics vary depending on the particular
element used. Typical optical efficiency through each secondary
optical element is between 85 and 90 percent.
Thermal Loss
LEDs will decrease in relative flux output as junction temperature (Tj)
rises. Most LED data sheets list typical luminous flux at Tj = 25°C,
while most LED applications use higher junction temperatures. When
using Tj > 25°C, the luminous flux must be derated from the value
listed on the LED’s data sheet.
2. Light Loss Within the Fixture
Fixture light loss occurs when light rays from the light source strike
the fixture housing before hitting the target. Some light is absorbed
LED data sheets include a chart showing the relative light output
versus junction temperature such as the one shown in Chart 3 for
XLamp XR-E white LEDs. By choosing either a specific relative light
48
output or a specific junction temperature, this graph shows the value
for the other characteristic.
For the CFL example, this luminaire is only designed for commercial
buildings with vented ceilings. Based on the listed design goals, this
design will prioritize light output, efficacy, and lifetime.
XLamp XR-E LEDs are rated to provide an average of 70 percent
lumen maintenance after 50,000 hours provided the junction
temperature is kept at 80°C or lower. Therefore, the appropriate
maximum junction temperature for the CFL example is 80°C. This
corresponds to a minimum relative luminous flux of 85 percent as
shown in Chart 3. This 85 percent relative luminous flux is the thermal
efficacy estimate for the example luminaire.
Table 5: Summary of Example CFL Replacement Efficiencies.
Be aware that driver efficiency can vary with output load as shown in
Chart 4. Drivers should be specified to run at greater than 50 percent
output load to maximize efficiency and minimize cost.
For indoor applications, 87 percent is a good estimate for driver
efficiency. Drivers meant for outdoor use or very long lifetimes will
probably have lower efficiency.
Table 5 summarizes the efficiencies of the optical, thermal, and
electrical systems for the example luminaire.
4. Calculate number of LEDs needed
Actual Lumens Needed
With all the system efficiencies estimated, the actual number of LED
lumens required to achieve the design goals can be calculated. For
this calculation, only the light efficiencies (optical and thermal) are
used. The electrical efficiency affects only the total power consumed
and fixture efficacy, not the amount of light coming out of the
luminaire. The calculation of “actual lumens needed” for the example
luminaire is shown below:
Chart 3: Example Relative Intensity versus Junction Temperature Graph for XLamp
XR-E White LED.
Electrical Loss
The LED driving electronics convert the available power source (e.g.,
wall-plug AC or battery) to a stable current source. Just as with any
power supply, this process is not 100% efficient. The electrical losses
in the driver decrease the overall luminaire efficacy by wasting input
power on heat instead of light. The electrical loss should be taken into
account when beginning the LED system design.
Typical LED drivers have efficiencies between 80 and 90 percent.
Drivers with efficiency over 90 percent will have much higher costs.
Actual Lumens Needed = Target Lumens / (Optical Efficiency x
Thermal Efficiency)
Actual Lumens Needed = 810 / (91% x 85%)
Actual Lumens Needed = 1,050 lm
Operating Current
Another decision to be made is what operating current to
use for the LEDs. Operating current plays an important role
in determining the efficacy and lifetime of the LED luminaire.
Increasing the operating current will result in more light output
from each LED, thus reducing the number of LEDs needed.
However, increasing operating current also has several drawbacks,
shown in Table 6. Depending on the application, these drawbacks
may be acceptable trade-offs for the higher per-LED lumen output.
For the example luminaire, lifetime and efficacy are top priority
design goals. The luminaire will run at the minimum operating
current listed on the XLamp XR-E datasheet (350 mA) to maximize
LED efficacy and lifetime.
Chart 4: Example Efficiency vs Load Graph for LED Driver.
Drawback
Explanation
Reduced efficacy
Higher operating current reduces the efficacy of current generation power
LEDs. In general, the size of the power supply will increase as operating current
increases, since it takes more power to generate the same number of lumens
Reduced
maximum
ambient
temperature
OR
decreased lifetime
Higher current will increase the temperature difference between the LED junction and
the LED’s thermal path. In practical terms, since the maximum junction temperature is
already decided, this reduces the maximum ambient temperature for the luminaire.
If instead of lowered maximum ambient temperature, the maximum junction temperature
is raised, the LED will degrade in light output faster over its operational life.
Table 6: Drawbacks of High Operating Current in LED Luminaires.
www.digikey.ca/lighting
49
Number of LEDs
After deciding on operating current, the lumen output of each LED
can be calculated. Since the thermal loss of the LED has already
been taken into account through the actual-lumens-needed
calculation, the numbers specified in LED supplier documentation
can be used directly without further interpretation.
For this calculation, it is important to use the minimum flux
listed for your LED order code and not the typical number on the
data sheet. Most LED companies sell to minimum flux ranges.
By designing against this minimum number, you are ensuring
that all luminaires made with that LED order code meet the
target requirements.
The example luminaire will use XLamp XR-E LEDs at 4000K CCT
with minimum luminous flux of 67.2 (P2 flux bin) @ 350 mA. The
number of LEDs is calculated below:
Number of LEDs = Actual Lumens Needed / Lumens per LED
Number of LEDs = 1,050 lm / 67.2 lm
Number of LEDs = 16 LEDs
5. Consider all design possibilities and choose the best
With the number of LEDs calculated, consider all the
possibilities to accomplish the design goals. Since each
LED is a small light source and has a much longer lifetime
than traditional light sources, LEDs can be integrated into
luminaires with new and unusual design elements. Designers
can take full advantage of LED light’s directionality and wide
variety of available secondary optics to create original designs.
At the same time, keep in mind that there are many different
regulations that constrain the design choices. While providing
a comprehensive list of worldwide standards applicable to LED
luminaires is beyond the scope of this article, Table 7 gives
examples of regulations that will apply in some portions of
the world.
The rest of this section explains three design options for each
system of our example LED luminaire: optical, thermal, and
electrical. For each system, guidelines for choosing the best
option are provided.
Type of Standard
EMI (Electro-Magnetic
Interference)
Safety
Efficiency
Example Standards for Lighting
• FCC CFR Title 47, Part 15
• EN61000
• EN55015
• UL 1310, Class 2
• UL 48
California Title 24, Part 6, of the California Code of
Regulations: California’s Energy Efficiency Standards
for Residential and Non-residential Buildings
Table 7: Example Standards that Pertain to LED Lighting
50
Optical system options
1. Bare LEDs & existing lamp reflector
As discussed earlier, the beam angle of the existing CFL fixture
and the LEDs are very similar. Thus, one available option is to use
no secondary optics. This option provides the lowest cost and
lowest optical loss for the system. Using fewer components and
less labor makes the luminaire easier and cheaper to assemble.
The drawback is the multiple-source shadow effect. Additionally,
if the light distribution of the LED is significantly different than the
target luminaire’s distribution, this option is not available.
2. LEDs with secondary optics & existing lamp reflector
Secondary optics are optical elements used in addition to the
LED’s primary optic to shape the LED’s light output. The general
types of secondary optics are reflecting (where light is reflected
off a surface) or refracting (where light is bent through a refractive
material, usually glass or plastic). Secondary optics are available
either by buying a standard, off-the-shelf part or by designing
a custom optic through ray-trace simulation with an optical
source model.
By using a secondary optic per LED, the beam angle of each
LED can be customized to provide the exact light output pattern
necessary. For instance, the beam angle of each LED can be
narrowed to make the luminaire optimized for spot lighting instead
of general lighting.
There are several drawbacks to this approach. First, the
luminaire will have a higher cost because of additional
components and more complicated assembly. Secondly, since
the optics are attached to each LED, there may still be multiplesource shadowing. Finally, the secondary optics will reduce the
optical system efficacy.
3. Bare LEDs, existing lamp reflector & diffuser
Instead of using one optic per LED, a diffuser can be used over the
entire LED array to spread the light. The benefits of this approach
are a wider beam angle than is possible with the bare LEDs and
eliminating the multiple-source shadow effect.
As with Option 2, the drawbacks are higher cost and reduced optical
system efficacy. This is also not an option if the light distribution
must be narrower than the bare LED, since diffusers can only
spread light, not collect it.
Illuminance distribution, the multiple-source shadow effect, and
aesthetics will usually drive the decisions on the optical system.
Option 2 is the only option if the light output must be narrower than
the bare LED. If not, Option 1 is better in terms of cost, efficacy, and
brightness. However, both Options 1 and 2 will exhibit the multiplesource shadow effect.
Also, users looking up at Options 1 and 2 will notice each
individual LED. Users of Option 3 will see only a diffuse, uniform
light source.
Multiple-Source Shadow Effect
Multiple-source shadow effect is a phenomenon where an
object placed between multiple light sources and a surface will
create multiple shadows. Most people have seen multiple light
bulbs mounted above a sink in a bathroom. If you have noticed
multiple shadows of yourself on the wall behind you, then you
have seen the multiple-source shadow effect.
LEDs placed close together create multiple shadows that are
close together. The appearance of these close shadows may be
undesirable in the target application. It is the designer’s job to
determine how important the multiple shadow effect is for the
target application and whether it is worth additional optical loss
to add a diffuser to minimize this effect.
Tooling and manufacturing fees may drive the per-unit cost of the
custom heat sink higher than an off-the-shelf design.
Target luminaire cost, available heat sink development time,
and target maximum ambient temperature will usually drive the
decisions for the thermal system. In general, Option 2 is better for
situations where low cost is more important than maximum ambient
temperature. Option 3 is better when maximum ambient temperature
is more important (e.g., outdoor lighting or indoor lighting in
unconditioned spaces).
The example LED luminaire will use an off-the-shelf heat sink with a
thermal resistance of 0.47°C/W. With the heat sink thermal resistance
value, the maximum ambient temperature can be calculated with the
following formula:
Tj = Ta + ( Rth b-a x Ptotal ) + ( Rth j-sp x PLED )
Tj
Ta
Rth b-a
PLED
Ptotal
Rth j-sp
= LED junction temperature
= Ambient temperature
= Heat sink thermal resistance
= Single LED power consumption
= (Operating current) x (Typical Vf @ Operating current)
= Total power consumption = (# LEDs) x PLED
= LED package thermal resistance
Example luminaire values:
Tj MAX
= 80°C
Rth b-a
= 0.47°C/W
PLED
= 0.35 A x 3.3 V = 1.155 W
Ptotal
= 16 x 1.155 W = 18.48 W
Rth j-sp = 8°C/W
b = ( a x L2 ) / L1
a = LED spacing
b = Shadow spacing
L1 = Distance between LED & object
L2 = Distance between object & surface
Ta MAX = Tj MAX – ( Rth b-a x Ptotal ) – ( Rth j-sp x PLED )
Ta MAX = 80°C – ( 0.47°C/W x 18.48 W ) – ( 8°C/W x 1.155 W )
Ta MAX = 80°C – 8.6856°C – 9.24°C
Ta MAX = 62°C
Thermal system options
1. Existing fixture housing
The lowest-cost option is to reuse the fixture housing of an existing
design as the housing and heat sink for the LED luminaire.
Obviously, this is not an option for new luminaire designs. Additionally,
most existing housings are made of steel, which is a poor thermal
conductor. Generally, a steel housing will be a bad choice for a
heat sink.
2. Off-the-shelf heat sink
Another option is to buy an off-the-shelf heat sink. This heat sink
will be a proven design and come with full specifications from
the manufacturer.
However, it may not be optimized in performance, size, or shape for
the target application.
3. Custom heat sink
While a custom solution provides the best opportunity to optimize the
heat sink for the application, there are several drawbacks.
This option requires the designer to have access to thermal simulation
software or access to a third party with thermal design expertise.
A maximum ambient temperature of 62°C for the example
luminaire is acceptable for this indoor application. For an
operating environment needing higher maximum ambient
temperature, either the maximum junction temperature should
be raised (which may impact lifetime) or the thermal system
(Rth b-a) improved (e.g., better heat sink).
Electrical system options
1. Off-the-shelf LED driver
An existing LED driver will provide the quickest design time, since it
is already available and will come with a reference circuit design. All
parts will be tested for EMI and safety regulations and will typically
have the lowest per-unit cost in volume.
The drawbacks are that existing LED driver efficiencies are typically in
the mid-80 percent range. Lifetime and operating temperatures may
also be an issue, depending on the vendor and the application.
2. Next generation LED driver
As LED lighting is gaining in popularity, more semiconductor
companies are turning their attention to optimizing LED driver
www.digikey.ca/lighting
51
designs. Another option is to partner with one of these companies
on the next generation of LED drivers, which will have higher
efficiencies and full regulatory approval.
However, waiting for the product development may delay the
development of the LED luminaire. Additionally, smaller companies
may not be able to work together with a driver company on an
unreleased product.
3. Custom design
As with thermal design, a fully customized electrical system
is an option. While it may be possible to get a higher efficacy
than by using an off-the-shelf part, there are many potential
drawbacks.
The burden of development and regulatory approval is now on the
designer. Even after development, the per-unit cost may be higher
than an existing solution. Also keep in mind that driver companies
will continue to develop more efficient and cheaper drivers during
the LED luminaire development period.
Available development resources and target efficiency will usually
drive the decisions for the electrical system. In today’s high-power
LED environment, improvements in the overall luminaire efficacy are
driven more by the LEDs themselves and not the drivers. It may be
advantageous to get a product out sooner rather than trying to wait
until the electrical design is perfect.
6. Final steps
Once the design decisions have been made, Table 8 details the final
steps to build and evaluate a prototype luminaire.
Step
Explanation
Board layout
•
•
•
Build a prototype
•
•
•
Complete the circuit board layout.
Choose board material (FR4 vs. MCPCB) based
on thermal and cost contraints.
Keep in mind how the layout and positioning of parts will affect the light
output and thermal flow of the luminaire.
Building on prototype (or several) is a valuable way to validate the design.
Verify that the optical, thermal and electrical
systems perform as they should.
Test how easy the unit is to assemble.
Test prototype against
design goals
•
•
Test the prototype to make sure it achieves all the design goals.
Testing can be done either internally or externally by a contracted luminairemeasuring company.
Finalize design & BOM
•
Make final changes to the design (if any) based on the new information
learned from analyzing the prototype.
Draw conclusions
•
How could the existing design be improved if a
different design choice was made?
Are all of the original design goals still applicable, or are
some less important than they seemed initially?
Are there other applications that would benefit from LED light?
•
•
Table 8: Final Steps in LED Luminaire Design
Where to get help
The entire LED luminaire design process can be overwhelming. Table
9 provides URLs to the current list of Cree partners that can assist
getting to a final design.
Source for Help
Can Assist With...
Cree XLamp LED Optics Solutions
http://www.cree.com/products/xlamp_part.asp
•
Optical system design
•
Electrical system design
•
•
•
•
•
Thermal system design
Off-the-shelf optics & LED drivers
Board layout
Assembly
Small quantity orders
Cree XLamp LED Driver Solutions
http://www.cree.com/products/xlamp_drivers.asp
Cree XLamp LED Distributors
http://www.cree.com/products/xlamp_dist.htm
Table 9: Cree XLamp LED Solutions Providers
Cree® XLamp® XP-G LEDs
New Cree® XLamp® LED Performance
Breakthrough Delivers Both High Efficacy &
High Color Rendering
The XLamp XP-G LED delivers unprecedented levels of light output
and efficacy for a single die LED. The XLamp XP-G LED continues
Cree’s history of innovation in LEDs for lighting applications with
wide viewing angle, symmetrical package, unlimited floor life and
electrically neutral thermal path.
With up to 90-CRI, the XLamp XP-G LEDs can deliver light quality
comparable to halogen with better efficacy than fluorescents.
Lamp designers can now create systems that are 70 percent more
efficient than traditional halogen PAR38 lamps and deliver similar
high color rendering. The high-CRI XP-G LEDs are also similar to
standard XP-G LEDs, allowing designers the potential to upgrade
existing systems without any redesign. The new high-CRI XLamp
LEDs bring lighting-class performance to the most demanding color
sensitive lighting applications.
Key Features
• Available in white, outdoor
white and 80-CRI, 85-CRI
and 90-CRI white
• ANSI-compatible
chromaticity bins
• Maximum drive current:
1500 mA
• Low thermal resistance:
6°C/W
• Wide viewing angle: 125°
• Unlimited floor life at ≤
30°C/85% RH
• Reflow solderable - JEDEC
J-STD-020C
• Electrically neutral thermal
path
• RoHS and REACH-compliant
• UL-recognized component
(E326295)
Applications & Key Markets
• Wherever high CRI is
important
• Bulb retrofits (Par 38, MR16)
• Specialty retail
• Interior design
• Medical lighting
• Jewelry and department
stores
• Fabric retailers
• Paint retailers
www.digikey.ca/cree-lighting
52
Phosphor Film
Conversion for White LEDs
by Bit Tie Chan, Avago Technologies
Existing white LED technology is
manufactured using a combination of a blue
LED and a yellow phosphor. White light is
perceived when blue light from the LED is
mixed with the yellow light that is emitted
from the phosphor.
In such a device, it is important that the color of the light output
is uniform, and that the colors are consistent among different
devices. A conventional phosphor-dispensing method is typically
used to fabricate this device, where phosphor in epoxy slurry is
directly dispensed on top of the LED die.
The difficulty of accurately dispensing a consistent amount
of phosphor during the manufacturing process is widely
acknowledged. Furthermore, two other phenomena take place after
the dispensing process. First, the phosphor tends to settle down
over time prior to curing (See Figure 1). Second, the encapsulate
materials will also shrink down, leading to different casting height
before it is fully cured. These two changes cause significant
interaction effects between the LED and phosphor. The higher the
casting height, the higher the amount of phosphor, which causes
the white light to be more yellowish since more blue light is being
converted by the phosphor. In the case in which more phosphor
settles to the bottom, a bluish white light is seen (See Figure 2).
The end result is a non-uniform color output from one device
to another, leading to a wide color spread along the axes of the
chromaticity chart in the white LED.
Figure 2: Illustration on white light output as a result of different casting (different
amount of phosphor).
The film-based method helps minimize the color spread of a white
LED, thus improving manufacturing yield and reducing product cost.
In the prior method, the color spread spans as much as 0.04 along the
axes of the chromaticity chart, while the film-based method is able to
reduce the spread to less than 0.02. (Please refer to the experimental
results in Figure 9).
Efficient light collection, mixing, and transmission is ensured by
placing the LED and phosphor film within a cavity. The phosphor
layer is less affected by the heat generated by the LED by avoiding
direct contact, thus preserving the emission efficiency of the
phosphor material. Phosphor tends to lose conversion efficiency
with increased temperature, of course, and therefore the more
distant the phosphor film from the LED, the less an undesirable
heating effect is experienced.
Furthermore, by embedding the phosphor film inside a cavity, it is
protected from other elements in the environment that are known to
adversely affect phosphor, such as moisture.
Yet another advantage of embedding the phosphor film between two
encapsulate layers is to ensure there is no mismatch in the coefficient
of thermal expansion, which can give rise to de-lamination between
the dissimilar layers.
Figure 1: LED
cross-section
showing
Phosphor
sendimentation.
Experimental approach and discussion
Phosphor in polymer binder (silicone) is drawn across a substrate
by using a Doctor Blade to produce phosphor film. Carrier tape
was used as a substrate in this experiment. Selecting the correct
silicone to produce a continuous film with good surface properties is
one of the important steps to produce phosphor film with consistent
film thickness.
www.digikey.ca/lighting
53
Three types of silicone were evaluated to identify the best silicone
to create continuous film with good surface properties: no voids or
bubbles. Silicone A was used in this evaluation as it created excellent
films with the desired surface properties. Figure 3 shows the quality
of phosphor film using different types of silicone.
The phosphor film with carrier tape was first fixed on an adhesive
board and then diced with the laser (Figure 5).
Figure 5: The converter layer is processed on a glass carrier and cut with the laser.
Figure 3: Quality of phosphor film after curing with different types of silicone.
Next, the highest possible fill-phosphor ratio is mixed with silicone
to avoid phosphor settling. Full films with the thickness described
in Table 1 were fabricated. The films were scanned in a 1cm raster
with Avago’s Blue Moonstone device (Figure 4), and Correlated Color
Temperature (CCT) was measured. The results revealed that phosphor
film with CCT of either 3000K or 4100K was non-uniform due to
small film thickness. Phosphor film with CCT of 6500K or 9300K
was more uniform due to larger film thickness. The optimum film
thickness of 180µm was then used to produce the phosphor film for
the subsequent evaluations.
Correlated Color
Temperature
Phosphor Ratio
Film Thickness
Figure 6 illustrates a successful model of the experiment. An LED is
placed inside a cavity and a wire bond is made from one terminal
of the LED to a terminal (not shown) at the base of the cavity. A first
layer of encapsulant is placed above the LED. The phosphor film is
placed such that one side is in contact with the first encapsulant.
Then a second encapsulant is placed such that the other side of
the phosphor film is in contact with it. As a result, the phosphor film
received nearly all the blue light and converted at least a portion
of the blue light to yellow light. As illustrated in Figure 7, the walls
of the cavity acted as a reflector and channeled the combination of
blue and yellow light in the direction desired to further improve color
mixing, thus enhancing the uniformity of the emitted white light.
Film Uniformity
6500K
60%
180µm
Uniform
9300K
60%
180µm
Uniform
3000K
40%
100µm
Not uniform
4100K
40%
75µm
Not uniform
Figure 6: Illustration of phosphor film assembly in Avago's Moostone LED package.
Table 1: Optimum film thickness and phosphor ratio for the required correlated color
temperature (CCT).
Figure 7: Illustration on the cavity of wall acts as reflector to shape the ray light to
desired direction.
Figure 4: Isometric view of Avago’s Moonstone
Power LED Star package (ASMT-Mx09).
54
With the proposed assembly method, the phosphor film is placed in
close proximity to the die, allowing the blue light emitted by the blue
LED to see a consistent phosphor thickness. Thus, the conversion of
blue to yellow is consistent and uniform. The uniform ratio of blue
to yellow light yields an end result of more consistent perceived
white light. As shown in Figure 8, the current practice of ‘binning’
white LEDs is simply a work-around to manage the variation in
white color and tint that are the result of today’s manufacturing
processes. The inefficient binning process creates poor yields to the
manufacturer, as LEDS that don’t fit well in the desired color bins are
discarded. To date, there have been no solutions to address the supply
chain risk waste produced by this method. Results shown in Figure 9
reveal that contemporary color spread spans as much as 0.04 along
the axes of the chromaticity chart, while the phosphor-film method is
able to reduce the spread to less than 0.02.
White LEDs should be color binned to within a 2-step
MacAdam ellipse to avoid having noticeable color
temperature differences when they are positioned sideby-side or when they are used to illuminate a white
scene such as a wall. However, a wider 4-step MacAdam
ellipse can be used in applications where the LEDs are
not directly visible or when they are used to illuminate a
multicolored scene.
Figure 8: White LEDs Color Binning
Cree MP-L EasyWhite 2- and 4-step MacAdam ellipse chromaticity binning.
Video:
Digi-Key Lighting Technology
Symposium - LEDIL
Figure 9: Comparison of Color Spread with Dispensing vs Phosphor Film
Conclusion
The color consistency of white LEDs may be enhanced by employing a
film where the phosphor material is uniformly incorporated in GaInNbased white LEDs. Incorporating the phosphor into a film provides an
accurate and consistent amount. A consistent color converting effect
is achieved as the blue light ‘sees’ a consistent layer of phosphor;
consequently, the ratio of blue light and yellow light is maintained,
leading to the perception of a consistent white light. This technology
enables the manufacturer to control the color temperature and
minimize production variance. The color consistency improvement
of phosphor film conversion is 50% compared with conventional
phosphor-based white LEDs.
Panelist Tomi Kuntize (Director, Design and Engineering,
LEDIL) specifies how optics can affect the reliability and
lifetime of SSL products for the topic, “How real is a
50,000-hour lifetime for LEDs?”
Phosphor film is proven to be feasible to achieve narrow color binning.
However, further enhancements of the homogeneity are required as it
is limited by the current process.
Acknowledgment
The author would like to thank Margaret Tan Kheng Leng for providing
the support and information needed to write this article.
www.digikey.ca/video-lighting
References
[1]. E. F. Schubert, Light-emitting Diodes, Cambridge, 2003
www.digikey.ca/lighting
55
Drivers for HB LEDs Allow
Designers to Build Halogen and
Incandescent Lamp Replacements
by Piero Bianco, Maxim Integrated Products Inc.
Halogen and incandescent
lamps, although popular,
present some key
concerns in today’s
power conscious, green
world. The lamps
consume a lot of
power and typically
burn out after a
few thousand
Figure 1: Typical HBLED
replacement for popular halogen
hours of use.
and incandescent lamps
LED retrofit lamp solutions
Halogen and incandescent lamps, although popular, present some
key concerns in today’s power conscious, green world. The lamps
consume a lot of power and typically burn out after a few thousand
hours of use. The latest high-brightness LEDs (HBLEDs) offer a great
alternative – they use a lot less power and can last about ten times
as long. However, designing HBLED replacements for popular halogen
and incandescent lamps such as the MR16, PAR20, A19 (Figure 1),
and others present three key design challenges:
1.
The retrofit lamp must fit in the same socket as the lamp it
replaces, which means it should have the same form factor
2.
The retrofit must not only manage the high amount of power
dissipated by the LEDs by having proper heat sinking but
operate at high temperature while maintaining high reliability
and a long lifetime
3.
Finally, retrofits must be electrically compatible with the existing
lighting infrastructure (wiring, dimmers, etc.)
While previous generation LED drivers could implement retrofit LED
lamps that meet the first challenge, most of the drivers do not have
the circuitry to meet the third challenge when cut-angle (triac or
trailing edge) dimmers are present and can have problems keeping
long lifetimes at high operating temperatures due to the lifetime
limitations of electrolytic capacitors. However, the latest generation
drivers, such as Maxim’s MAX16834 solution, incorporate additional
56
circuitry to handle dimming functions and provide work alike
functionality to halogen and incandescent lamps while delivering
what is probably the most important advantage of LED technology
today – lamps with very long lifetimes and as a consequence, low
lamp replacement costs.
Fitting in the existing form factor
The existing form factor imposes both a physical limitation (i.e.,
the driver board has to be small enough) and a thermal limitation
on a retrofit lamp. While both of these limitations are particularly
challenging for MR16 and GU10 form factors, they pose a challenge
for the design of other replacement lamps (including PAR, R, and A19
form factors) as well.
While size is important for a retrofit, thermal limitation is often more
critical. LEDs emit only visible light; they do not radiate energy at
infrared wavelengths like other technologies. Thus, while LEDs are
more energy efficient than incandescent and halogen lamps, they
dissipate much more heat through thermal conduction in the lamps.
Thermal dissipation is also the main limiting factor for the amount
of light that a lamp can produce. Today’s LED technology in retrofit
lamps can just achieve a level of brightness that is acceptable
for the mainstream market. Pushing the limits of brightness
and, consequently, thermal design is essential for designing a
commercially successful product.
A corollary issue to the thermal dissipation is the lifetime of the
driver board. To emit more light, the lamp must work at a fairly
high temperature (often +80°C to +100°C). At this temperature,
the lifetime of the driver board can limit the lifetime of the whole
lamp. Electrolytic capacitors are, in particular, the biggest challenge.
Since they dry quickly at those temperatures, the lifetime of those
capacitors is limited to little more than 10,000 hours, and this
becomes the limiting factor for the lifetime of the whole lamp.
The graph in Figure 2 shows an example of LED lamp lifetime
degradation (B50/L70 lifetime, i.e. when 50 percent of the LEDs
have lost at least 30 percent of their brightness) as a function of
lamp’s internal operating temperature. As you can see, at about
80°C, lifetimes shorten for lamps that use electrolytic capacitors
versus lamps that do not use the electrolytics, because the lifetimes
of the driver board start becoming the limiting factor. At 100°C,
lifetimes with electrolytic capacitors are much shorter. As previously
mentioned, internal operating temperatures of 80°C to 100°C are
common for LED retrofit lamps.
requirements for their load currents that are fulfilled by resistive
loads:
• During the off part of the voltage half cycle, the triac dimmer
cannot have an open circuit as its load but instead needs a
resistive load. In general, dimmers have RC networks that time
off time, and their loads (lamps) are the only return paths for the
current flowing through these RC networks.
• After the end of the off part, a dimmer latches on; in order to
remain latched on during the remaining part of the voltage half
cycle, a dimmer needs to have at least a certain amount of
load current. If the current falls below this amount, the dimmer
unlatches and turns off inappropriately, and as a consequence,
the light of the lamp flickers. High transient spikes in the load
current can also be a problem, because they can cause the load
current to fall below this minimum level.
Figure 2: As the internal temperature of a lamp increases, its operating lifetime
decreases. Lamps that employ electrolytic capacitors in their driver boards (red line)
have shorter lifetimes than lamps that do not have electrolytics (blue line).
A long lifetime is a major selling point for an LED lamp and probably
the main reason why businesses are switching to LED lighting
today—a long lifetime means much lower lamp replacement cost,
which can more than offset the considerably higher sticker price of
an LED lamp. For this reason, lamp makers need to provide more than
10,000 hours lifetime if they want to make a successful product.
Matching the electrical infrastructure
Retrofit LED lamps must work correctly in infrastructures
than include cut-angle (triac or trailing-edge) dimmers and
electronic transformers.
A triac (i.e. cut-angle) dimmer reduces the amount of light produced
by a lamp connected as its load by keeping the lamp for the initial
part of each AC supply voltage cycle. The dimmer is off for an amount
of time, which is adjustable, and then turns on and latches on for the
rest of the half cycle. The effect is that the voltage applied to the lamp
looks like the waveform in Figure 3.
Triac dimmers are designed to work with incandescent and halogen
lamps, which are purely resistive loads. In fact, they have some
LED lamps that are not designed to be dimmable do not work well
with triac dimmers. Their internal driver circuits typically include
rectifiers and buck or flyback converters. The input current of such
a driver consists of short, high spikes at each half cycle of the input
voltage. Such an input current is not compatible with the conditions
listed above, and, in fact, such lamps do not dim, flicker, or turn on at
all when used with triac dimmers.
The electrical infrastructure is even more complicated for a 12VAC
input lamp, because an electronic transformer and dimmer can be
connected at the lamp’s input.
An electronic transformer typically includes an oscillating circuit
that modulates the input 50/60Hz AC voltage with a frequency
of approximately 40kHz. The resulting higher frequency passes
through a transformer that provides isolation and converts the input
120/230VAC to the output 12VAC. By modulating the input voltage
with a higher frequency, it is possible to have a much smaller
transformer, thus reducing size, weight, and cost.
Similar to a triac dimmer, an electronic transformer needs a certain
amount of load current to remain on for the full cycle of the input
voltage. If the load current is not sufficient or has high spikes, the
transformer can turn off, causing the light to flicker. For the same
reason stated above, a traditional AC/DC converter driver can be
incompatible with the transformer and dimmer and cause the light to
flicker.
Active power factor correction as the
solution for dimmable LED lamps
While, for the remainder of this article, we will focus in particular on
a design for offline 12VAC lamps, the same considerations apply to
230VAC input lamps.
As described above, the dimmability of an LED lamp and compatibility
with and an electronic transformer has a lot to do with shaping the
input current of the lamp appropriately.
Another typical requirement for LED lamps that also has to do with
the shape of the input current is power factor correction. For LED
lamps, a power factor of at least 0.7 is needed for most residential
applications, and a power factor of at least 0.9 is needed for most
commercial applications.
Figure 3: Output voltage of a triac dimmer
www.digikey.ca/lighting
57
As the problems of dimmability and power factor correction are
similar, it is likely that a single solution can solve both challenges. In
this article in particular, we propose active power factor correction
as the best solution for these challenges. There are several reasons
why active power factor correction is superior to passive power factor
correction in this case:
• With active power factor correction, a power factor of 0.9 is easily
achievable; with passive power factor correction, a power factor of
0.7 is fairly easy to achieve with 0.9 being a much bigger challenge.
• Active power factor correction allows a very fine control of the
input current, and therefore, it can keep the input current above
the level required for the dimmer to work properly for the whole
cycle of the input voltage. With passive (or valley fill) PFC, the
input current remains zero or close to zero for a certain part
of the input cycle and/or is phase shifted with respect to the
input voltage.
• Passive PFC, in particular if done with a valley fill circuit, causes
spikes in the input current, which can cause flicker of the lamp
as mentioned above. With active PFC, it is possible to reduce the
amplitude of those spikes.
Another choice that the designer faces is between a fixed frequency
switching regulator topology and a variable frequency type (e.g.
transition mode) and also between continuous conduction mode and
discontinuous or transition mode.
Regarding the first matter, fixed frequency generally has an advantage
for the management of EMI issues. With a fixed frequency solution,
the designer must only filter EMI noise at that particular frequency,
while with a variable frequency design (e.g. a transition mode design),
the switching frequency varies along the cycle of the input voltage,
so it causes noise over a wide frequency range, which can be more
difficult to filter.
With regard to the conduction mode, the continuous mode has the
obvious advantage of keeping the peak current lower, thus reducing
conduction losses (which increase as the square of the currents);
with discontinuous or transition mode, the switching losses are lower,
because the MOSFET turns on at zero inductor/transformer current.
However, the gains in conduction losses in continuous conduction
mode are often greater than the difference in switching losses.
The solution in Figure 4 uses a single stage conversion, which
minimizes size and cost, to drive LEDs in replacement lamp. It uses
active PFC and works at a fixed frequency in continuous conduction
mode using a patent-pending technique.
Figure 4: Block diagram of the electrolytic-free LED driver
58
In this solution, the input current is shaped as a square wave at
the same frequency as the input voltage to maximize its value over
the whole cycle of the AC voltage in order to fulfill the requirements
of triac dimmers. It uses a fully released product from Maxim, the
MAX16834. The square input current is obtained by controlling its
average value and keeping it constant throughout the cycle of the
rectified input voltage. Resistor R1 senses the MOSFET current, which
is basically the same as the input current, and components R2 and
C2 extract the average of this value and feed this information to the
MAX16834, which keeps it constant with its control loop.
As mentioned before, an LED driver compatible with triac dimmers
needs to behave like a resistive load for the dimmer during the off
part of the input voltage cycle. In this design, components R3, Q1, and
“Startup Current Control” block performs this function by providing an
input resistance whenever the input current of the driver falls below a
certain level.
Inductor L1 and capacitor C1, which is a small value ceramic
capacitor, make a filter for EMC purposes.
The “IC Bias Circuit” provides a 15V supply for the MAX16834 IC. At
start-up, a linear regulator circuit generates this voltage from the AC
supply. Once the IC starts switching, a second circuit generates this
voltage with a level translator supplied by the switching node and
overrides the linear regulator. This second supply circuit allows for
an increase in efficiency of the solution, because it avoids the power
dissipation that takes place in a linear regulator.
This design uses a non-isolated buck topology composed of inductor
L2, diode D1, and MOSFET Q2. It is possible to design a similar
solution that uses a flyback isolated topology. Thus, this solution
works regardless of if the safety isolation of the LED lamp from the
input voltage is done in the driver or in the enclosure of the lamp.
Option to avoid electrolytic capacitors
The electrolytic output capacitor C3 is optional. If it is included, the LED
current has a small amount of ripple at twice the input voltage frequency.
If a smaller value ceramic capacitor is present, the LED current is rectified
sinusoid at twice the input frequency as mentioned above. However, the
lifetime of the lamp can extend to 50,000 hours or more, since there are
no electrolytic capacitors in the circuit, and electrolytic capacitors are
typically the limiting factor for the lamp’s lifetime.
Performance of this solution
The circuit in the schematic (Figure 4) has been tested on a demo
board with a 120VAC/60Hz input and eight LEDs at the output with
a total output power of 10.1 watts. It has been tested to work with a
wide range of triac dimmers, including those listed below:
Input Voltage
Maker
Model
120VAC
Lutron
Skylark S-600P
120VAC
Lutron
Rotary DNG-600PH
120VAC
Lutron
Maestro MA-600
120VAC
Panasonic
WN576159
120VAC
Leviton
Sureslide 6633-P
120VAC
Leviton
Illumatech IPI06
120VAC
Cooper
Aspire
120VAC
GE
18022
With output electrolytic capacitors, this driver dims to zero light
intensity with no flicker. Without electrolytic capacitors, it can dim to
about five percent of the maximum light intensity without flicker.
OSLON SSL
OSRAM Opto Semiconductors
The observed efficiency is 85 percent, and the input power factor is 0.95.
Conclusions
Designing LED retrofit lamps is a big challenge. They have to
fit into the physical and electrical infrastructure that was made
for incandescent and halogen lamps that have very different
requirements and limitations. This creates heat dissipation, lifetime
limitation, and dimmability issues for the LED lamp maker.
This product training module introduces the OSLON
SSL (solid state lighting) LEDs and discusses their
features, specifications, and target applications.
At present, numerous solutions are being proposed to try to
overcome dimmability limitations using different topologies and
power factor correction techniques. In this article, we discussed
the advantages of using an active PFC solution, which provides
excellent dimmability with no flicker of the lamp while providing a
more than good enough power factor. We have provided an example
of such a solution using Maxim’s MAX16834 IC. This is a solution
that is available for use in mass production today, and that, in
addition to the dimmability, provides the advantage of giving the
designer the option to avoid electrolytic capacitors, thus solving the
driver lifetime issue mentioned previously in the article.
www.digikey.ca/ptm
SORT BY PRICE
SEARCH FEATURE
Digi-Key’s Sort by Price feature makes it easier
for engineers and purchasers around the world to
purchase top-quality product from Digi-Key.
Sort by Price is based on tier 1 or unit pricing and
applies to all parts, all of Digi-Key’s 82 international
websites, and all Digi-Key supported currencies.
Customers are able to sort by price in ascending
or descending order in addition to applying filters
for product in stock, lead-free product, and RoHS
Compliant product. An Advanced Sort option is also
available and allows customers to request parts in
specific quantities.
Customers can view pricing of products from
Digi-Key’s vast line card of suppliers at a glance and
choose the parts that best fit their particular needs.
WWW.DIGIKEY.CA
www.digikey.ca/lighting
59
LED Drivers for Automotive
Applications
contributed by ROHM Semiconductor
Design options include both external and
integrated output versions, all of which meet
exacting automotive qualification standards.
Introduction
The usage of lighting-emitting diodes (LEDs) in automotive
applications is increasing for many of the same reasons that LED
lighting is penetrating non-automotive sectors. LEDs are more
efficient and smaller in size, have a substantially longer life, allow
considerably greater design freedom for improved aesthetics, and
more. In brake lamp applications, the fast turn on of LEDs provides
an added margin of safety to warn the driver of the car behind.
LEDs can respond as much as ten times faster than traditional
incandescent bulbs. In addition to this safety aspect, the color of LED
lighting is more natural, making it safer for forward illumination. The
ease of controlling LEDs also makes them a natural for intelligent
lighting systems that adjust based on vehicle sensor inputs. An
essential aspect of the control is the power management provided by
integrated circuit (IC) drivers.
According to Strategies Unlimited, a market intelligence firm that
focuses on LED lighting, LED lighting penetration of exterior and
interior lighting functions continues. Strategies Unlimited analyst
Dr. Robert Steele says that as auto production recovers from the
recession, so too will sales of LEDs for these applications. For 2010,
Steele predicted that the automobile segment would grow by more
than ten percent from 2009.
Automotive applications for LEDs include interior lighting (such as
dome, dash and footwell lighting), indicator and telltale lights and
infotainment backlighting as well as exterior (signaling) functions
such as tail lights, turn signals, brake lights (including center
high-mount stop lamps (CHMSL)), parking lights, side marker lights,
fog lamps and daytime running lights (DRLs). More recently, a few
vehicle manufacturers have introduced LED headlamps on production
models based on high-brightness (HB) LEDs. In some cases, the
capabilities of an LED driver can enable more than one application to
be addressed with the same LEDs.
With leading automotive headlamp manufacturers AL-Automotive
Lighting (Magneti Marelli), Hella, Ichikoh, Koito, Valeo, Visteon and
others providing prototypes with HB-LEDs, almost all carmakers
have displayed concept vehicles with LED headlights. In fact, at least
one headlight manufacturer predicts that several standard vehicles
60
will have LED headlights in 2012. As LEDs continue to improve in
efficiency and decrease in cost based on projections from Haitz’s Law
(the light output levels from packaged LED devices roughly doubles
every 18 months), an increasing amount of LEDs and LED drivers will
be used in vehicles.
With the low-power consumption of LEDs compared to conventional
lighting, an estimated 0.2 liters of fuel per 100km and about 4 grams
lower CO2 emissions per kilometer are being cited as the ultimate
advantage of replacing incandescent lighting with LEDs in the DRL
application alone. In electric and hybrid vehicles, an 85% reduction
in energy consumption from LED usage instead of incandescent
bulbs translates into increased range. As a result, there are several
compelling reasons to implement LEDs in automotive applications.
LED driver capabilities
LEDs require a constant current to produce consistent lighting.
Consequently, this forms the basic operating requirements for an LED
driver. The accuracy of the current source determines its customer
appeal. Current fluctuations than can occur with voltage supply
variations in vehicles must be avoided. Linear regulators provide a
simple control and do not require electromagnetic interference (EMI)
filters. However, their power dissipation can become excessive for
higher power applications. Buck DC-DC converters are commonly
used in vehicles as the next step up from a linear regulator. When the
driver must control several LEDs in series, a boost converter topology
is used. In some cases, a buck-boost topology provides the capability
to address a variety of application requirements including the ability
to handle voltage extremes.
LED drivers can be designed to offer a combination of series and
parallel LED control. Devices with this capability built-in provide
circuit designers flexibility to control LEDs in different applications
with a single driver rather than requiring different devices that
increase qualification testing.
Dimming the light level is a common requirement for interior lighting.
However, exterior lighting has applications for brighter and normal
requirements from the same LEDs. For example, brake lights/tail lights,
low beam/daytime running lights and high beam/low beam headlights
are bi-level lighting. In some cases, lighting design may be able to
address both situations with the same LEDs with the right LED driver.
For the harsh automotive environments, several protection circuits are
required to prevent device failure under fault conditions.
Automotive design considerations
Unlike other market segments, automotive applications have several
tougher requirements that are reflected in industry standards and
purchasing specifications. These unique criteria include temperature
and humidity range, voltage range, ability to withstand harsh
chemicals, electromagnetic interference and electromagnetic
compatibility (EMC) as well as reliability requirements dictated
by qualification testing. For example, the automotive passenger
compartment temperature range is -40 to 85°C.
The automotive voltage range extends from normal operation of 9
to 16V (nominally 14V) to charge the 12.6V battery under ambient
temperatures from Arizona to Alaska and includes extreme conditions
such as reverse battery (-12V), jump start conditions of continuous
double battery voltage (+24V) to faults conditions such as load dump,
which occurs when the battery is disconnected from the alternator,
and other voltage transients. An unclamped load dump can be as
long as several hundred milliseconds and can easily exceed 80V but
today many manufacturers have centralized load dump clamping
circuits and subsequently require that components must withstand
the transient for levels of 40 or 60V. In addition to higher voltage
requirements, cranking conditions cause lower voltage that require
protection for worst case situations.
High reliability in automotive applications is indicated by the need
for protection circuitry such as overvoltage, undervoltage, reverse
polarity, overcurrent, short circuit and overtemperature protection in
many ICs. Also, the component’s life must be verified by testing to
meet the vehicle manufacturer’s target life and warranty requirements
that could be 10 years or 100,000 miles. In general, automotive
IC qualification requirements are reflected in tests such as the
Automotive Electronic Council’s AEC Q100 series for integrated
circuits but individual automotive suppliers and carmakers can
require tighter criteria and even further testing.
Discrete versus integrated LED outputs
A power IC process with analog and digital as well as power
circuitry allows device designers to integrate the LED power
switches with the control circuit - as long as the package can
dissipate the power. Integrated LED switches reduce the number
of components, saving board space and simplifying inventory and
manufacturing. As power levels increase, such as the LED drivers
for HB LEDs or when the option for driving LED arrays is desired, a
driver, or pre-driver, that controls external discrete output devices
provides flexibility with the output switches selected based on the
circuit requirements. As a result, a single LED driver can cover the
requirements of several applications.
ROHM Semiconductor LED driver solutions
for automotive applications
ROHM Semiconductor has highly-integrated LED driver solutions for
passenger compartment and forward illumination applications. Drivers
have integrated switches or in some cases, pre-drivers designed to
switch external power MOSFETs. Three products demonstrate the
different approaches that are provided for automotive applications.
ROHM Semiconductor’s BD8119FM-M is a 4-channel constant
current backlight LED driver for medium to large automotive
displays such as navigation or dashboard panels. The driver uses
an original current mode buck-boost DC-DC converter and requires
Figure 1: The highly-integrated BD8112EFV-M can drive 150mA through two LED lines
and requires a minimal number of external components.
only a single coil for simplified design. It is packaged in an 18.5
x 9.9 x 2.31mm HSOP-M28. The BD8112EFV-M is a 2-channel
version, suitable for backlighting small-to-medium size TFT
displays, for example, the TFT installed in an instrument cluster. It
is offered in a HTSSOP-B24 package. The flexible buck-boost LED
drive is shown in Figure 1. It has external PWM control and voltage
control of the VDAC terminal for brightness control.
Operating from a supply voltage of 5 to 30V, the switching frequency
of the BD8112EFV-M and BD8119FM-M can range from 250 to
550kHz with external synchronization option to avoid EMI problems
with vehicle radios or other sensitive circuits. A total of 28 LEDs from
a 7 x 4-channel matrix can easily be driven by the BD8119FM-M
driver with a maximum current of 150mA per line and even more with
a properly configured external circuitry. The driver has built-in LED
abnormal state detection for Open and Shorted conditions as well as
built-in protection functions including undervoltage lockout (UVLO),
overvoltage protection (OVP), thermal shutdown (TSD), overcurrent
protection (OCP) and short circuit protection (SCP).
ROHM Semiconductor’s BD8105FV and BD8115F fully-integrated
drivers provide serial and parallel control to address telltale indicators
for instrument clusters, especially clusters and center stack controls
such as HVAC, radio and more. The units require only a few external
components minimizing board space. The BD8105FV consists of 12
open-drain outputs (see Figure 2) and the BD8115F has eight opendrain outputs. The maximum DC current for each output is 50 mA with
a pulsed maximum value of 150mA.
The drivers can have at least two devices cascaded in series as
shown in Figure 3, so that more LEDs can be controlled without
increasing the number of I/O pins of the microprocessor. Packaged in
an SSOP-B20W, the devices have thermal shut-down (TSD) circuitry
(nominal detection point at 175°C) built in to prevent overheating.
www.digikey.ca/lighting
61
SDWN
D11
TSD
D10
D9
D8
D7
<11:0>
SERN
Driver
LATCH
CLK
Serial
I/F
RST_B
D6
D5
D4
<11:0> <11:0>
D3
<11:0>
D2
D1
ROHM Semiconductor LED drivers
With the introduction of the BD8381EFV-M white LED driver IC for
automotive lamps, ROHM Semiconductor has significantly extended
its range of highly integrated LED driver ICs. These LED drivers
provide designers a variety of design options with integrated or
externally switched outputs, parallel/series control and extensive
protection and fault detection functions in small surface mount
packages. For more information about the expanding portfolio
of automotive LED drivers in ROHM Semiconductor’s power
management portfolio visit our website or contact your local ROHM
Semiconductor Automotive sales office.
D0
VCC
SEROUT
SEROUT
GND
Figure 2: The block diagram of the BD8105FV shows the 12 drain connections to the
internal power MOSFETs.
Since the power devices are integrated in the IC, the TSD can protect
the 12 outputs from excessive temperature.
ROHM Semiconductor’s newest LED driver is the BD8381EFV-M for
driving multiple HB-LEDs in high- and low-beam headlamps as well
as daytime running light applications. The BD8381EFV-M is a white
LED driver with the capability of withstanding high input voltage
(50V MAX). As shown in Figure 4, a current-mode buck-boost DC-DC
controller is integrated to achieve stable operation over varying
voltage input and also to remove the constraint of the number of LEDs
connected in series.
Figure 4: The BD8381EFV-M uses external power MOSFETs so designers can select
the switch rating they need for the high-beam, low-beam and/or DRL applications.
Another Geek Moment video:
ROHM LED Driver
Jamie Pederson (Design Support Services) demonstrates
the use of the 11 pin SIP LED driver module from ROHM
Semiconductor.
Figure 3: Application circuit for two cascaded BD8105FVs being controlled by a
microcomputer with a 4-wire (3-line plus enable) I/F serial input.
The BD8381EFV-M provides dimming by either a built-in PWM or
linear control. Operation with or without a microcomputer is possible.
The operating frequency can be set internally between 100 to 600KHz
or externally synchronized from the internal oscillator frequency up to
600kHz. Built-in protection functions include UVLO, OVP, TSD, OCP and
SCP with LED error status detection function for OPEN/ SHORT circuit.
The circuitry is housed in an HTSSOP-B28 package.
62
www.digikey.ca/geekmoment
Color Management of a
Red, Green, and Blue LED
Combination Light Source
contributed by Avago Technologies
Tricolor optical feedback is a proven color
management solution, but its implementation
can be complex. It doesn’t need to be.
Abstract
Combining red, green, and blue LEDs (light emitting diodes) to form
a light source is an attractive proposition. Such a multi-color light
source can generate a wide range of colors. The LEDs themselves
are robust and highly efficient. LED light sources will improve many
existing applications, such as LCD backlighting, and enable new ones,
such as adaptive automotive interior lighting. There are, however,
challenges to overcome before a high-quality multi-color LED light
source can be produced. This paper describes an Application Specific
Standard Product (ASSP) integrated circuit that addresses many of
those challenges.
on the ratio of light intensity between the red, green, and blue sets
of LEDs. The light intensity of an LED can be controlled by varying its
drive current, or by varying the duty cycle of the LED using pulsewidth modulation (PWM). The PWM method is popular because the
relationship between duty factor and light intensity is more linear than
that of current vs. intensity.
This simple, open loop method of constructing an LED light source
has a critical flaw. Since an LED’s optical characteristic varies with
operating conditions, the brightness and chromaticity of the combined
RGB light output will vary. In addition, LEDs vary part-to-part,
contributing to more variation in the RGB light output. Figures 2 and 3
below illustrate examples of LED variation.
One solution is to use optical feedback to produce a closed-loop
system. A basic setup consists of a photo sensor to record the
Green LED
0.8
y - Color Coordinate
Blue LED
0.6
Red LED
0.4
0.2
Figure 2: Example of RGB LED spectral shift over temperature.
0
0
0.2
0.4
x - Color Coordinate
0.6
0.8
Figure 1: CIE 1931 Color Space. The triangle represents the color gamut formed by red,
green, and blue LEDs. W—white, R—red, G—green and B—blue. The x, y coordinate
defines the chromaticity of a light source. A third axis (not shown), Y, is needed to define
the brightness of the light source.
Using RGB LEDs
The simplest multi-color LED light source consists of three sets of
LEDs—red, green, and blue—with each set driven by a separate
driver module. The resultant color of the light source is dependent
Figure 3: Example of LED relative intensity shift over temperature. Normalized at 25°C.
www.digikey.ca/lighting
63
brightness of the LED light source and a method of modulating
the light output with respect to the photo sensor readings. This
permits the brightness of the LED light source to be consistent even
if the individual LEDs vary. (i.e., the sum is constant, though the
components vary).
In Figure 4, an integrator (labeled 22) outputs a voltage that is
dependent on the amount of light falling on a photodiode (11a). That
voltage is compared against VSET. A counter (34) increments or
decrements depending on the comparator decision. The counter value
is used to drive a DAC (37), which determines the LED driving current.
The tricolor optical feedback system
Essentially, a tricolor photo sensor creates a three-dimensional color
specification system (henceforth known as the RGB sensor color
space). This system allows specific colors to be specified in terms
of the sensor’s output voltage. For example, a D65 white with a
particular brightness can be specified as:
(Vred, Vgreen, Vblue) = (2.0, 2.2, 1.9) volts (see Figure 6)
Refer to Figure 5 and suppose that the D65 example above is used as
the “TARGET COLOR.” The feedback system will periodically measure
the red, green, and blue sensors (collectively known as the tricolor
photo sensor) and compare the “MEASURED COLOR” against the
“TARGET COLOR.” The goal of the feedback system is to adjust the
difference (error) between the “MEASURED COLOR” and “TARGET
COLOR” to zero.
Figure 6 illustrates the idea in a different way. All possible target
color setpoints are specified as coordinates within the RGB sensor
color space formed by the red, green, and blue sensors. As the
LEDs change in characteristic, the measured color moves away
from the target. The ASSP will detect this and adjust the LED PWM
signals accordingly.
Figure 4: Schematic showing a method of implementing optical feedback. This drawing
was extracted from an Avago Technologies’ patent US 6,344,641. The concept of using
color filters placed over the photo sensors was described in Avago Technologies’ patent
US 6,448,550.
A more advanced optical feedback method uses a tricolor photo
sensor, which typically consists of three independent photo sensors
with a tricolor filter placed over them. Thus, the photo sensors are
able to record color information, rather than just brightness. This
allows the ratio of light intensity between red, green, and blue sets
of LEDs to be controlled. This is a critical feature as it allows the
brightness and chromaticity of the RGB light source to be controlled. It
is in tricolor optical feedback where the ASSP plays an important role.
Figure 6: Trilinear coordinates representation of the RGB sensor color space. For
clarity this trilinear representation only considers chromaticity and not brightness. In
practice, the feedback system corrects for any deviation of the target color setpoint
including brightness.
It is important to understand that as LEDs age, they degrade in light
output intensity. Hence, over time, an RGB LED system’s maximum
achievable brightness will decrease. In most applications, a gradual and
graceful reduction in brightness is acceptable; what is not acceptable
is a change in the chromaticity of the RGB light system. The ASSP has
a feature that gracefully controls the diminishing brightness of the RGB
light system such that the chromaticity is kept steady (within tolerance)
even as the maximum achievable brightness decreases.
In applications where the system brightness must be maintained
throughout the lifetime of the application, the user must ensure
that the maximum selectable brightness is less than the maximum
achievable brightness over the required lifetime, as shown in Figure 7.
Figure 5: The tricolor optical feedback system. Each photo sensor has a color filter
placed over it. The low-pass filters perform a “time-average” on the sensor output such
that a DC voltage level is passed to the controller.
64
As attractive as RGB light systems are, there are challenges that
hinder widespread usage of this technology. There exists a need for a
device that “hides” the complexity of tricolor optical feedback behind
a simple user interface. The following sub-sections describe how the
ASSP accomplishes this.
Green LED
0.8
y - Color Coordinate
Blue LED
0.6
Red LED
0.4
0.2
0
Figure 7: The ASSP keeps the chromaticity (defined here as 1931 x, y coordinates)
of an RGB light source steady (within tolerance) even as the maximum achievable
brightness decreases over time.
Zero external processing
The ASSP incorporates a set of algorithms that analyze color
information from the tricolor photo sensor and compute the PWM
drive signals to achieve the target color setpoint. The ASSP samples
the photo sensor at about 100 times per second to ensure that
the periodic adjustment to the PWM signals is not perceptible to
the human eye. As mentioned previously, the ASSP also contains
an algorithm that prevents LED degradation from changing the
chromaticity of the RGB light output.
No other calculations are necessary to achieve and maintain a target color.
Color space standardization
This relates to device independence in selecting a target color setpoint.
Figure 8: The ASSP samples the tricolor photo sensor directly and translates the
readings into the required LED PWM signals.
The RGB sensor color space varies with differences in photo sensor
output, photo sensor location, LEDs, LED drivers, etc. Figure 9
illustrates the problem. Each system will have a slightly different RGB
sensor color space. Hence, a D65 specification for system A will be
different than that for system B.
0
0.2
0.4
x - Color Coordinate
0.6
0.8
Figure 9: Variations in the RGB sensor color space are calibrated out using the ASSP.
A transformation is performed to map the RGB sensor color space to a standard color
system such as 1931 CIE xyY, to allow every system to pick a target color using a
standard color system.
Design-in simplicity
Typically, the ASSP only requires only passive components for support
and an external PROM to store calibration data. In most cases,
memory space can be shared with other peripherals at a system level,
since calibration data is only 31 bytes.
The ASSP has a standard two-wire 100kHz I2C interface, and all major
functions are mapped to an 8-bit address space. For example, to
run the calibration calculator, simply write 0x01 to register CTRL2.
Implementation details can be found in the device datasheet.
In the manufacturing stage, the system is calibrated using a standard
CIE camera. The calibration data needs to be stored in an external
non-volatile memory. The calibration process is not required after
the system is deployed in an application. In application, the user first
configures the device and then writes the previously stored calibration
data into the calibration registers. This is a simple read-then-write
process. After that, the system is ready to accept a target color.
Color selection is simple. In the examples, a D65 target color was
specified in terms of sensor voltages. In actual application, the target
color can be specified as coordinates in the CIE 1931 xyY system.
Other color systems such as CIE uvY and CIE RGB can also be used.
For example, to select illuminant E as the target color, a value of (x, y,
Y) = (330, 330, 250) is sent to the appropriate registers in the ASSP.
• Illuminant E CIE x, y coordinates are 0.33, 0.33
• Scale that by 1000 to get 330, 330
Example:
System A (Vred, Vgreen, Vblue)D65 = (2.0, 2.2, 1.9) volts
System B (Vred, Vgreen, Vblue)D65 = (2.1, 2.4, 2.3) volts
• Choose a relative brightness level of Y = 250
The tricolor photo sensor in system A will produce the voltage levels
above when D65 light output is achieved.
• Write 330 into register address 235 and 234 to set the x
chromaticity coordinate
System B’s photo sensor will produce a different set of voltage levels
even though it generates the same D65 light output as system A. In
other words, the color specification system as defined by the RGB
sensor color space is unique system-to-system.
• Write 330 into register address 233 and 232 to set the y
chromaticity coordinate
• Write 250 into register address 237 and 236 to set brightness (Y
value)
The ASSP incorporates a calibration procedure to allow every system
to use a standard color specification system. CIE 1931 xyY and CIE
RGB are two such systems built into the ASSP. By having a standard
color space input, the user can send the same target color to every
system and expect that each system will produce the same color
(within tolerance).
• Write 0x12 to register address 1 (CTRL1) to update the new
target color
The ASSP will change the RGB light output immediately after the
“update” register bit is set.
www.digikey.ca/lighting
65
Figure 10: Typical configuration of a
tricolor color management system. The
internal reference and oscillator options
are enabled. Only passive components
support the device.
Figure 13: Measurement of chromaticity change over temperature.
Figure 11: A sample of the ASSP register space. Each bit is mapped to a function.
Figure 12: Typical design-in flow.
Figure 14: Spectral shift over temperature. Chromaticity shift duv <0.005.
Experimental results
Figure 13 shows the performance difference between an open-loop
and a closed-loop RGB light system. The experiment was conducted
using a 9000K white target color and uses duv as the figure-of-merit.
Figure 14 illustrates the severity of LED spectral shift as
temperature increases. This data was collected from a closed-loop
system with a 9000K white target color. Although the spectral
curves shifted significantly, duv was kept below 0.005 by the
feedback system.
where,
(u25, v25) = 1976 CIE u, v chromaticity coordinates at +25°C
(uT, vT) = 1976 CIE u, v chromaticity coordinates at temperature T
A good rule-of-thumb to judge performance is to use duv = 0.005
as the minimum change in chromaticity before the human eye can
detect a change.
66
Conclusion
RGB LED lighting is an appealing lighting solution. However,
the variability of LED characteristics causes the RGB light to
deviate from the intended color. Tricolor optical feedback is a
proven solution, but its implementation can be complex. Avago
Technologies’ feedback controller simplifies the implementation of
such a system.
Digi-Key, at your service.
Programming
Programmable Oscillators
Digi-Reel
Battery Packs
Cable Assembly Services
Departments
Phone ....................................................................800-344-4539
Sales/General Inquiries ..........................................E-mail: sales@digikey.com
Phone: 218-681-6674
Fax: 218-681-3380
Customer Service ..................................................E-mail: customer.service@digikey.com
Technical Support ..................................................E-mail: techs@digikey.com
Design Support Services (DSS) ...............................E-mail: design@digikey.com
Phone: 877-DK-DESIGN (877-353-3744)
Quote Request ........................................................E-mail: quotes@digikey.com
Comments/Suggestions .........................................E-mail: webmaster@digikey.com
Web Links:
Feedback ...............................................................www.digikey.ca/feedback
My Digi-Key ...........................................................www.digikey.ca/registration
TechZonesSM ..........................................................www.digikey.ca/techzones
PTM Online...On Demand® .....................................www.digikey.ca/ptm
New@Digi-Key ......................................................www.digikey.ca/new
DK Toolbar .............................................................www.digikey.ca/toolbar
BOM Manager ........................................................www.digikey.ca/bommanager
Digi-Key Catalogue ................................................www.digikey.ca/catalogue
Technical Support Web Chat ..................................www.digikey.ca/techchat
2
Digi-Key’s TechZones
SM
• Featured products from
leading suppliers
• Application notes,
references designs,
PTMs, exclusive editorial
• Parametric search
tools and more!
Text Paper Contains 10% Post-Consumer Fiber
Please Recycle
©2011 Digi-Key Corporation, 701 Brooks Ave. South, Thief River Falls, MN 56701, USA
www.digikey.ca/techzones