Creating Tunable White Lightwith High Colour-Rendering Index
by Muhamad Moussa, Technical Marketing Engineer, Future Lighting Solutions
Insufficient colour rendering has been a major challenge
standing in the way of LED technology’s adoption in mass
markets such as medical, general, retail or display lighting.
Muhamad Moussa, Technical Marketing Engineer, Future
Lighting Solutions, explains a systematic method for
overcoming that challenge with solid-state lighting
designed to produce tunable white light.
Colour-rendering index
Colour-Rendering Index (CRI) represents how well a light source renders the
true colours of different objects with a value ranging from 0 to 100. It is
possible to calculate a negative CRI value, but this would contain no useful
information.
In order to calculate the CRI of a light source, referred to as the test
illuminant, a set of predefined light sources are used. These are known as
reference illuminants and they all have a CRI of 100. A set of test colours is used
to compare the test illuminant with the reference illuminant, which is the 15
Munsell test-colour samples shown in Table 1. Standard CRI calculations only
take the first eight colour samples into consideration. The ninth colour sample
(strong red) is of great significance in applications such as medical illumination
and retail displays.
Table 1: Munsell test-colour samples.
The choice of reference illuminant varies depending on the target
Correlated Colour Temperature (CCT) at the test illuminant. For CCTs less than
5,000K, the reference illuminant should be a black-body radiator, such as an
incandescent bulb, otherwise it should be daylight.
Any variation between the Spectral Power Distribution (SPD) of the test
illuminant compared with the reference illuminant results in a reduced CRI
value. Beingmonochromatic light sources, LEDs inherently have a spectral
power distribution that peaks at certain wavelengths and falls at other
wavelengths. For example, white LEDs are created using a blue die covered
with a layer of phosphor and, as a result, their SPD includes a peak in the blue
region with the phosphor filling the rest of the spectrumto a lesser extent (see
Figure 1).
Fig. 1: Typical relative spectral power distribution of a white LED.
Such peaks and valleysmay not affect the xy chromaticity coordinates of
the white light, nor the Correlated Colour Temperature (CCT). However, they
do significantly affect the CRI of the white light. In other words, two white
lights can have the same CCT and xy chromaticity coordinates, but have
significantly different CRI values.
An example is shown in Figure 2 where two normalised SPDs are plotted.
One is for a test illuminant created using Red, Green and Blue (RGB) LEDs, and
the other SPD represents the reference illuminant. Both light sources have the
same CCT and xy chromaticity coordinates. The only difference is that the
reference illuminant has a CRI of 100 and the test illuminant has a CRI less than
20 resulting fromthe difference between the two SPDs.
To create a white light (test illuminant) with a high CRI value, the principle
of colourmixing can be applied in order to fill the gaps in its SPD. Although
thismight sound like an easy solution, it involves some complexmathematical
calculations.
Future Lighting Solutions has developed tools, concepts and solutions to
support the development of high-CRI tunable white-light applications. These
tools can also optimise the design in terms of light output, number of LEDs
and efficacy.
Fig. 2: SPDs for RGB test illuminant and its reference illuminant.
Commercial recessed downlight
To illustrate the concept of colour mixing to create a high-CRI tunable white
light, an example 6”commercial recessed down-light will be analysed. To set
the performance criteria for this application the following Energy Star®
requirements make a good starting point:
- Minimum light output: 575 lumens
- Acceptable CCT range: 2,700K to 6,500K
- Minimum efficacy: 35 lm/W
- Minimum CRI: 75
Table 2 summarises the commercial down-light’s target performance. It is
worth noting that the actual CCT range can be a subset of the Energy-Star
CCT range. A range from 3,000K to 4,500K will be used in this example. Also,
although Energy Star lists a minimum CRI of 75, this example will improve on
that value and target a CRI of 90 throughout the CCT range.
Table 2: Summary of down-light requirements and target values.
LED selection: colour and quantity
The first step is to determine the combination of LEDs that will produce
575 lumens at the given CCT and CRI values. The CRI optimisation tool,
developed by Future Lighting Solutions, provides the minimum number
of LEDs needed to achieve the target CCTs and CRI, while maximising the
luminous flux performance in order to enhance the system’s efficacy.
To implement our down-light example using
LUXEON® Rebel LEDs, the CRI optimisation tool
specifies a colour combination of red, green,
amber and cool-white LEDs. In order to achieve
the target light output, three clusters of these
four selected LED colours are required as shown
in Figure 3. In order to set CCT values, the current
delivered to each LED is controlled. It is the ratio
of these currents that sets the CCT value of the
output light.
Fig 3: Three clusters of four LEDs with colour-mixing optics to create a commercial down-light.
Figure 4 shows the SPD for the reference
illuminant and the SPD of the white light generated using the four-LED
cluster. Compared to the RGB SPD shown in Figure 2, the gaps between
the test illuminant and the reference illuminant have been reduced. This
results in a higher CRI value. Adding more colours will further reduce the
gaps and enhance the CRI; however, there will be a point at which adding
more LEDs will significantly increase the system cost without a major
change in CRI. The goal is to minimise the number of LEDs used and still
achieve the target requirements. It is important to note that a CCT of
3,250K was specified as an example and the principle outlined can apply
to all target CCTs shown in Table 2.
Fig 4: SPDs of the reference illuminant and test illuminant at 3,250K
Thermal analysis and design
A comprehensive solution must be thermally managed. LED manufacturers
publish datasheets that contain information on the performance of LEDs at a
junction temperature of 25°C. In reality, LEDs rarely operate at such low
junction temperatures. Therefore, to ensure that the lighting system performs
as expected, analysing the system’s performance at elevated junction
temperatures is more practical.
As an LED’s junction temperature increases, it experiences three
phenomena:
- Degradation in light output
- Colour shift
- Degradation in lifetime
Since the four LEDs used in the example down-light consist of four
different colours, degradation in light output may not be consistent.
Uneven degradation in light output would result in a change in the light-output
ratios of the four LEDs, which affects the colour point, CCT and CRI
values. It is therefore critical to make sure that the ratios stay constant. This
can be achieved by optimising thermal design to make sure that the LEDs
do not reach extreme junction temperatures. This optimisation can
include heat-sink design, as well as a continuous thermal-feedback loop
that monitors the LED’s junction temperature and de-rates current
accordingly.
To ensure that the LED’s junction
temperature is within an acceptable
range a thermal analysis must be run
for each CCT. Thermal simulation
software, QLED, is used to perform
such analyses and optimise heatsink
design. QLED provides junction
temperature results along with light
output and power consumption for
all the system’s LEDs.
Fig 5: QLED simulation of thermal distribution.
These results can, for example, be used to ensure that the LED-based
commercial down-light delivers the required number of lumens (575lm) at
the actual operating temperature rather than 25°C. QLED’s results are also used
to calculate efficacy values.
Efficacy, measured in lumens per Watt, is calculated by taking three losses
into consideration:
- Optical losses due to secondary optics
- Power losses
- Thermal losses (calculated in QLED)
Table 3 shows QLED’s numerical results and efficacy calculations for the downlight
example, assuming 80% efficiency for the optical solution and 80%
efficiency for the power solution. This results in an equivalent efficiency of 64%
with losses of 36%.
Table 3: QLED light output and efficacy results for LED-based lighting system.
Table 4 shows a comparison between Energy Star requirements and the
proposed high-CRI tunable white commercial recessed down-light. Using
Future Lighting Solution’s software, these Energy Star requirements are met
while maintaining high-CRI colour mixing.
Table 4: Energy Star® requirements compared to the performance of the proposed commercial recessed down-light.
LED Lifetime
An overall lifetime analysis of any LED-based system must include all the
components used, including LEDs, power ICs, resistors, capacitors and
other components. Any failure in one or more component will result in a
system failure.
The LED Reliability Tool (LRT) is used to estimate the lifetime of LUXEON®
Rebel LEDs. Figure 6 shows the output of the LRT at a sample forward current
and junction temperature for InGaN and AlInGaP LEDs.
Fig 6: LED lifetime output from the LED reliability tool.
Conclusion
Demand for high-CRI white and tunable-white luminaires that meet Energy-Star
requirements is increasing. Overcoming the inherent optical, thermal and
power challenges is central to capitalising on this demand. It is now possible to
achieve any CCT at high CRI values using a colour-mixing methodology with
the tools and support provided by Future Lighting Solutions.
Previous Page
Previous Page