Lamparas Con Led
Lamparas Con Led
O f the three key technologies that all artificial light sources in the
market are based on, two of them (incandescence and plasma-based
sources, otherwise known as discharges) have reached a plateau in efficacies,
but nevertheless there is still ongoing research, and fascinating developments
continue to take place. The third technology, although it started developing
and was marketed much later than the other two, and perhaps for that very
reason, is now showing such fast progress that not only can it compete with
the other technologies, but in some cases it is preferred to them.
This chapter is an introduction to this technology of solid-state light
sources, known as LEDs, which has attracted the interest of many profes-
sionals and covers a range of topics answering the most frequently asked
questions. Some of the topics include the semiconductor and diode tech-
nology on which LEDs are based, ways of creating different colors and
white light, modes of operation, their thermal management, applications,
and comparisons with the other two technologies of light sources. The
chapter addresses a wide range of professionals for whom light and its
sources and, in particular, solid-state lamps, are part of their work.
Solid-state lamps, that is, light-emitting diodes, are considered by many
to be the future of lighting. Indeed, they have evolved to such an extent
that they demonstrate several advantages over the other technologies, and
they have already dominated some applications. Of course, there are still
some issues that are limiting factors for their further development, such as
the issue of thermal management. A large number of scientists are work-
ing intensively on the technology of solid-state lamps and, more particu-
larly, they are focused on understanding how to create light through the
125
Si Si
Si Sb Si Si B Si
Si Si
p-type n-type
– – + +
– +
– – + +
Light
emission
+
I
R –
Figures 5.4 and 5.5 show the anatomy of a modern LED unit, while
Figure 5.6 shows a variety of shapes and forms that LEDs can take.
The following list is representative of compounds that with appropri-
ate additions of materials (doping) and the connection of the p-type and
n-type semiconductors created, result in emission of radiation after flow
of electricity. A general rule is that the energy difference increases (wave-
length of emission decreases) with increasing aluminum (Al) concentra-
tion and decreasing with increasing indium (In) concentration.
AlGaAs—red and infrared:
• AlGaP—green
• AlGaInP—high brightness, orange—red, orange, yellow, and green
• GaAsP—red, orange—red, orange, and yellow
• GaP—red, yellow, and green
• GaN—green and blue
• InGaN—near-ultraviolet, blue—green and blue
• SiC as substrate—blue
• Sapphire (Al2O3) as substrate—blue
• ZnSe—blue
• Diamond (C)—ultraviolet
• AlN, AlGaN—ultraviolet
+
-
Figure 5.4 Light-emitting diode diagram (LED).
Figure 5.7 shows the emission spectra of three different LEDs in three dif-
ferent regions of the visible range. LEDs emit radiation of relatively nar-
row bandwidth, and this is depicted in Figure 5.8.
The efficiency of LEDs is defined by several factors such as
Epoxy lens/case
Wire bond
Reflective cavity
Semiconductor die
Anvil
Leadframe
Post
Flat spot
+ –
Anode Cathode
30°
30°
40° 40°
50° 50°
LED
5.2 Organic LEDs
If the material used is an organic compound, then it is known as an OLEDs
(organic light-emitting diode). For the organic compound to function as
a semiconductor, it must have a large number of conjugated double bonds
between the carbon atoms. The organic compound may be a molecule
with a relatively small number of atoms, in crystalline form, or a polymer
(PLEDs), which offers the advantage of flexibility. For the time being, the
organic LEDs offer lower luminous efficiencies and average lifetimes than
their inorganic cousins.
OLEDs are steadily making their way into commercial devices such
as cell phones and flat-screen displays. They are fabricated with layers of
organic polymers, which make them flexible, and they use less power and
less expensive materials than liquid crystal displays.
The downside is that because the polymers react easily with oxygen
and water, OLEDs are expensive to produce—they have to be created in
high-vacuum chambers—and they need extra protective packaging layers
to ensure that once they are integrated into display devices, they do not
degrade when exposed to air or moisture.
OLEDs can be made from a wide range of materials (see Figures 5.9 and
5.10 for examples), so achieving good-quality white light is less challenging.
It has not been the quality of light that has let OLEDs down but rather their
efficiencies. Fluorescent lighting typically operates at around 60 to 70 lm/W,
while incandescent bulbs operate at about 10 to 17 lm/W. In contrast, the
best reported power efficiency of an OLED until now has been 44 lm/W.
OLEDs have the potential to grow into a very energy-efficient light
source. In production, levels of between 15 and 20 lumens per watt have
O O
N Al
O
N
H5C2 C2H5
H5C2 C2H5
Pd
H5C2 C2H5
H5C2 C2H5
been achieved, but the ultimate potential is for the technology to reach
efficiencies as high as 150 lumens per watt.
A combination of these technologies can also lead to the future light sources.
One idea proposed is a hybrid light-emitting diode, or HLED. The device
would incorporate both organic and inorganic layers, combining the flexibil-
ity of an OLED with the stability of an inorganic light-emitting material.
Figure 5.11 (See color insert following page 20.) Combination of dif-
ferent color LEDs for the creation of white or dynamic lighting.
Figure 5.12 (See color insert following page 20.) Combination of dif-
ferent color LEDs for the creation of white or dynamic lighting.
4000
Blue LED
3500
3000
2500
Intensity
1500
1000
500
0
300 350 400 450 500 550 600 650 700 750 800
Wavelength (nanometers)
Figure 5.13 Blue LED emission spectrum with phosphor that converts
part of the blue light into other colors/wavelengths (mainly yellow).
of using a powder is the issue of distribution of light. The light emission
angle from the crystal is different from that from the powder; hence, mix-
ing is not very good.
Apart from the UV LEDs that can be used with phosphors, blue LEDs
can also be coated with a powder that converts part of the blue light into
yellow (Figure 5.13). The ratio of blue to yellow can be controlled through
the quantity of the powder used, allowing us to control the color tempera-
ture of the source. However, this method gives us a source with a low color
rendering index, as there is a deficiency in red emissions, something that
can be an issue in some applications such as general lighting but not in
other applications such as signage. This method allows us to create white
light with a color temperature of up to about 5500 K, but with the addition
of another powder that emits in the red part of the spectrum we can also
create a warm white light temperature of 3200 K and better color render-
ing, at the cost of reducing the source efficiency.
The use of powder on a blue LED is the most economical way to cre-
ate white light, and there are even proposals to use a blue LED (InGaN)
with a green powder to replace the low-efficiency green LED or a
blue LED with a red powder to replace the temperature-sensitive red
(AlInGaP) LED.
The use of phosphors on an UV LED can give white light of significantly
higher color rendering index, but at the expense of efficiency (mainly due
to Stokes losses), similar to how a fluorescent mercury lamp operates. The
light from the ZnSe substrate). The absence of fluorescent powder means
higher efficacy.
Finally, a method that is being developed rapidly is the use of quantum
dots: nanocrystalline semiconductor materials with dimensions equal to
a few dozen atoms that emit light (fluoresce) with high efficiency under
electrical or optical stimulation. The wavelength of radiation can be con-
trolled by controlling the size of the nanocrystals, and this method is in
the experimental stage (see Figure 5.14 for related spectra).
This color-tailoring ability solves one of the major problems with using
LEDs for general lighting applications. LEDs are appealing because they last
for years, use perhaps 20% of the electricity of a standard incandescent bulb,
and are highly efficient at converting electricity into visible light instead of
into heat. But to make white light, you either have to mix together LEDs of dif-
ferent colors or use a blue LED coated with a phosphor that emits yellow light
1.4
1.2
Relative Intensity
1.0
Relative Intensity
0.2
0.8
0.6
0.1
0.4
0.2
0.0 0.0
400 500 600 700 400 500 600 700
Wavelength/nm Wavelength/nm
(a) (b)
1.4 1.4
1.2 1.2
Relative Intensity
Relative Intensity
1.0 1.0
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
400 500 600 700 400 500 600 700
Wavelength/nm Wavelength/nm
(c) (d)
Figure 5.14 Various emission spectra of white LEDs: (a) blue LED with
phosphor while the other three spectra, (b), (c), and (d), belong to dichro-
matic, trichromatic, and tetrachromatic quantum dot LEDs, respectively.
to produce a whitish mix. The problem with the phosphors is that they do not
emit evenly across the visible spectrum. They tend to have gaps in the green
section and even more so in the red, leading to the harsher, bluish light.
Due to their low levels of emitted light, a lamp must make use of mul-
tiple LEDs in order to be functional as a light source for general lighting.
Many LEDs are needed for most applications (Figure 5.15).
Most LEDs on the market operate at low power, usually less than 1 W,
but some products operate at powers as high as 7 W with an efficacy of
20 lm/W. The highest efficacy ever reported is about 130 lm/W, but for
very low-power LEDs. Generally, LEDs emit from 1 lm to several tens of
lumens, while there is now on the market a 5 W LED 5W with 120 lm of
white light at 350 mA. Depending on their flux, LEDs are categorized as
follows:
If LEDs reach 250 lm/W in the next two decades, they could replace all
fluorescent lamps that are currently limited to 50–120 lm/W.
A general rule that has been stated and that so far seems to be hold-
ing is the doubling of light output from each LED every 24 months in
the last 40 years (similar to Moore’s law) and the halving of cost almost
every decade. This pattern was observed and described for the first time
by Roland Haitz. In most cases, LEDs operate with less than 100 mA, and
when the current exceeds that value, they are referred to as power LEDs.
For high-power LEDs, a silicon gel is used instead of a polymer epoxy coat-
ing, and an appropriate heat sink is used for proper heat management.
5.4 LED Operation
The LED is a diode; that is, electric charge flows only in one direction.
Once the LED starts to operate normally, the current is linearly propor-
tional to the voltage; that is, the relationship between voltage and current
for the LEDs is positive, as shown by Figure 5.16.
Vf = Vo + Rs If
Vo is the initial voltage that must be applied before charges start to flow.
The value depends on the material and the energy difference between
states (band gap). In general, the voltage is influenced by the temperature
of the diode. The resistance also depends on the material and is quite low.
During production, there are shifts in Vo and the resistance, so the cate-
gorization of the final products according to their electrical characteristics
(binning) is necessary as the light output is proportional to the intensity of
the current and, therefore, the voltage.
1000
900
Current Intensity (mA)
800
700
600
500
400
300
200
100
0
0.0 1.0 2.0 3.0 4.0 5.0
Voltage (V)
The two most important factors in limiting light emission are heat and
current density. At large currents, the materials wear out, mainly due to
high temperatures but also due to leakage of charges to the other layers
outside the active crystals (especially when the layer of the active crystal is
thin). The maximum values depend on the material but, in general, more
than 120 Α/cm2 is prohibitive for all cases. High temperature not only
damages the crystal but also causes shifts in the emitted spectrum. For
example, yellow can be converted to red at high temperatures just before
the destruction of the crystal and the diode junction. This change cor-
responds to about 0.1 nm per degree Celsius, something that is a great
disadvantage if color reliability is important.
LEDs generate heat during operation (not in emission), and the materi-
als are very sensitive to it. Thus, the thermal design is an important part of
the design of these light sources.
An LED is a low-power light source, so LED lamps cannot be connected
directly to the mains voltage as they need a current and voltage controller
to keep them at low values and also to be able to run several LEDs at once
(as shown in Figures 5.17 and 5.18). In other words, LED lamps need the
equivalent of ballast, which we call a driver.
The driver can be a constant voltage driver or a constant current driver.
The constant voltage driver is not as reliable because there is a dif-
ferentiation of the required voltage during production; and because the
resistance is small, variations in voltage result in significant variations
in intensity.
The constant current driver is preferable as the intensity is set to the
normal operating intensity of the LEDs, and the voltage is adjusted.
There is also a hybrid driver of constant voltage with a large resistor
parallel to the LED. However, the power consumption by the resistor ren-
ders such drivers inefficient.
A driver can be used for many LEDs, and it is important how they are
connected (series/parallel). If we have a constant voltage driver, then a
resistor should be added so that we have nominal current intensity, while
with constant current drivers, we need to take into account the number
of LEDs. In each case, the current intensity through each LED or series of
LEDs must be carefully estimated.
The position of each LED in a system plays an important role in the
functioning of the whole layout. If an LED fails, it can create an open cir-
cuit (e.g., in a series connection) or a closed circuit with different electrical
Figure 5.17 (See color insert following page 20.) LEDs in series mainly
for decorative lighting.
Figure 5.18 (See color insert following page 20.) LEDs in series mainly
for decorative lighting.
temperature) over the consumed power due to the current that flows
through the LED.
Rθ = (ΔTjunction − ambient/P)
Where
ΔT = Tjunction – Tambient
P = current intensity (I) * voltage (V)
The total thermal resistance must, of course, include the entire system
from the junction to the surfaces that are in contact with the surrounding
air. Each individual part of an LED is characterized by a different thermal
resistance (whose values depend on the geometry, material, and surface
area of each piece), and according to these we can define the thermal resis-
tance of the crystal, which is given by the manufacturers (the smaller the
value, the easier the transfer of heat). Figure 5.19 shows a thermal model
for an LED circuit that is analogous to an electrical circuit.
With the thermal resistance (obtained from the manufacturer) between
the junction and the material on which the crystal rests, and by measur-
ing the temperature differences between the other parts using infrared
TJunction
RJC
RCB RJA
RTIM
PLED
RH
TAmbient
1.20
0.80
Blue
0.60
Green
0.40
0.20
Red
0.00
–20 0 20 40 60 80 100 120 140
Junction Temperature in Degrees Celcius
5.5.2 Overheating Avoidance
There are two ways in which we can control the junction temperature to
avoid a malfunction or premature destruction of an LED. One is to decrease
the intensity of the current, so that the LED is operated at a lower power,
and the other is to carefully design so that the overall thermal resistance
of the LED is minimized. The second method, although preferable, is not
so simple and requires us to take into account several parameters such as
the total surface area of the materials in contact with the surrounding air.
The greater the contact area, the lower the thermal resistance.
A combination of both methods is common practice, but proper calcula-
tions must be done in each different case. Since the flux decreases with both
increasing temperature and decreasing current, and thus the power, we
must work out whether (taking into account the thermal resistance given
by the manufacturer) a decrease in power, which means operation with
lower brightness, ultimately pays. The voltage also depends on the junction
temperature, and it is reduced as the temperature increases. If we deal with
a single LED or several in a series, then one can control the current; other-
wise, there is a risk of having uneven distribution of the current intensity
due to temperature variations and, ultimately, destruction of the system.
The circuit connected to the LED may include some type of heat sink,
such as the MCPCB LEDs (metal-core printed circuit board), but higher-
power commercial LEDs are using extra heat sinks characterized by their
large size relative to the LED and their large surface area. A typical geom-
etry of such extra large heat sinks includes the use of fins that dissipate
heat more efficiently. Such a heat sink is incorporated in luminaires or can
be itself the luminaire. Finally, some LEDs have the circuit at a distance so
that the heat generated by the circuit does not contribute to the increase
of the junction temperature. See Figures 5.21 and 5.22, which depict LED
lamps with incorporated heat sinks.
Figure 5.21 LED lamps with heat sinks incorporated in the design.
Figure 5.22 LED lamps with heat sinks incorporated in the design.
5.8 Applications
Applications in which LEDs have the advantages are
Applications where LEDs have disadvantages and must not be widely used
yet are the following:
• In high-temperature environments
• Where high brightness is required
• Where color stability is necessary
• If accurate stability of color temperature and color rendering index
is essential
• If good knowledge of lifetime and lumen depreciation is needed
The goal of LED companies is for this light source technology to dominate
the following applications in the near future:
• Car headlights
• Interior lighting
5.8.1 Signage—Visual Contact
In applications in which high levels of brightness are not needed but the
creation of optical signals of specific color is the aim, LEDs have already
started dominating the market. Some of those applications are the follow-
ing (LED characteristics that offer the advantages are in brackets):
LEDs are used in a wide range of applications ranging from building light-
ing to large panel displays.
For LEDs to dominate in other markets, the following characteristics
must be improved or changed:
5.9 Closing Remarks
The future of LEDs and perhaps of lighting in general will depend on
the outcome of various research paths that scientists around the world
are following or must follow. It is essential that a better understanding
of the light-generation mechanism and improvements of internal quan-
tum efficiency be achieved. LED manufacturing techniques must also
improve so that consistency and better quality can lead to better market-
ing. Apart from the actual junction, other parts of an LED are also crucial
and require R&D, such as substrates, packaging, and lenses with proper
thermal management. The ultimate goal is not only to increase the effi-
cacy of each LED but also to achieve long lifetimes and tolerance to high
temperatures. In addition, it is important to develop phosphors capable
of absorbing and converting photons efficiently and, as is the case already
for fluorescent lamps, the development of “quantum-splitting phosphors”
could be a breakthrough for LEDs.
The technology of OLEDs also has the potential for rapid growth and
market penetration as the great variety of organic luminescent materials
could give a large number of emitting colors.
As with any other technology, high efficiencies, consistency, long life-
times, good color stability, uniformity over large surfaces, and relatively
low costs are the characteristics desired for a technology to succeed.
Standards
IESNA LM-79-08
ANSI C82.2 (efficacy)