- Jardins de les Dones de Negre
- Nº1, Floor 2
- Sant Adrià de Besòs
- 08930 - Barcelona
- Dr. Josep Carreras
- Director of the Lighting Group
- (+34) 933 562 625
Different spectrally tuneable lighting modules can be found in the background art for research purposes. In particular, the NIST (National Institute of Standards and Technology) was the first institution to implement a spectrally tuneable lighting facility (www.nist.gov), in which they conduct important research in colour science, vision experiments and user preferences.
Other implementations of similar concepts can be found also in Europe, as that developed under the framework of the SSL4EU Project (http://www.ssl4.eu/), by the University of Pannonia (PANNON, “Spectrally Tuneable LED Lighting Simulator Laboratory Room at University of Pannonia”. Csuti, P. et al, SPIE Press Room).
However, all this implementations of multi-channel spectrally-tuneable environments have to be regarded as research infrastructures where thousands of LEDs are used, with huge amounts of electrical power needs, and with no integrated intelligence (they are driven by an external computer).
Different wavelength (inorganic) LEDs and associated technologies used in the lighting module (wavelength and LED requirements may change depending on the application, so this should be considered only a typical approach, and does not include additions from OLED research)
In order to achieve the goals of this project, intelligent and autonomous tuneable modules will be designed and implemented with the following characteristics (see Figure 5):
- The light module will consist of three main parts: an optical cavity containing the printed circuit board (PCB) with the LEDs, driver circuits for each channel, and a microcontroller managing communication between all parts.
- The optical cavity will be a hollow cylinder measuring with a highly reflective interior. A plastic housing surrounds the cavity. The LED PCB will include LEDs and OLEDs of different energies arranged onto a PCB).
- Microchip temperature sensors will be used to compensate for colour shifts.
- Good thermal conductivity from the LEDs to the PCB and to the heatsink is essential.
- The LED drivers provide current to the individual channels, each having its own IC constant-current buck regulator and surrounding components to source a constant current. The regulation of each channel’s radiative intensity is accomplished using current pulse width modulation (PWM).
- A microcontroller manages the I2C communication between the different parts and the temperature monitoring of the LED PCB.
CAD renders (not real) of typical light modules specifically designed for this project. The normalized spectra for all the LED channels (in this case 12) are also shown.
OLED lighting panels have been commercially available since 2009. First commercial panels on market have been supplied by European companies OSRAM and PHILIPS. A number of additional companies are now offering OLED lighting panels (e.g., European company Black Body and Asian companies LG Chem, Lumiotec, Panasonic). Luminous efficacies of the panels are in the range of 10 to 45 lm/W, typical brightness is 1000 to 3000 cd/m2 and operating lifetime is in the range of 10.000 h. Module size is up to 145 x 145 mm2. Luminous efficacies of more than 100 lm/W and lifetimes up to 100.000 h were already reached on laboratory level for small sized white OLED devices.
The weak point within a white OLED device is the blue component within the three emitters (red/orange, green and blue). The current efficiencies and lifetimes of the red/orange and green emitters are very high, as very efficient phosphorescent emitter materials can be used. Furthermore, one can obtain high performance OLED devices when those emitters are integrated in OLED stack architectures using the so-called p-i-n structure, which was pioneered by IAPP (Institute for Applied Photo Physics) of the Technical University of Dresden . Their strategy consists of a multi-layer structure that can, currently, only be prepared using unique vacuum evaporation techniques. COMEDD has a Sunicel200 research and development tool in operation which will be used to develop the OLED components for the HI-LED SSL modules.
Top emission OLED device design
Generally, OLEDs are bottom-emitting devices, as the substrate is glass and a transparent bottom electrode like ITO is used. For some applications, the top emitting OLED design approach is used. This is the case when the substrate cannot be transparent, e.g. in display devices or if the OLED is built on other substrate like a thin metal sheet.
Top-emitting OLED cross section
In the figure below, the typical emission of a bottom-emitting, white OLED (incorporating orange, green and blue emitters) is shown. It can be seen that the emission distribution is very wide, which is typical for OLED devices. However, it is also known that the lack of the blue emitter stability leads to a colour shift of the white OLED during its lifetime, resulting in a warm-white light with a CCT that is not very stable.
Emission spectra of a bottom-emitting white OLED
By using the top emitting OLED design and single emitters, it is possible to avoid colour shifts. The ageing of the OLED will only be reflected as a luminance drop over the operational lifetime. Furthermore, the top-emitting approach allows narrower emission characteristics. In the next figure, simulation results of a green top emitting OLED device are shown. In the simulations, the OLED stack architecture was slightly varied to obtain different wavelength maxima and FWHM values (full width at half maximum):
Simulation results of organic stack in the green visible región
Within the HI-LED project, top-emitting green OLEDs (and potentially other colours) will be developed to meet the requirements of different specific applications and also cover the wavelengths not available with LED technology.
At present time, there are two major goals for the optics in Solid State Lighting. The first deals with keeping the luminance (brightness, amount of light flux per unit area and unit solid angle) as high as possible and the second with the ability to mix the light coming from non-uniform light sources, such as the multi-chip LED arrays that are needed when the application needs a high lumen output and/or a wide set of different output spectra, CRI and colour temperatures. Accomplishing both goals at the same time is not an easy task.
High luminance is preferred because the lamps doing a specific function can be more compact (and cost-effective), since the apparent source they are dealing with is also “smaller” (either dimensionally or angularly, i.e. its ètendue , is smaller), for a constant flux output. When high flux is needed, SSL light engines tend to locate several chips altogether into an array.
The maximum luminance of an array of LEDs is typically smaller than that of the LEDs alone, since these are typically placed with a sizable gap in the PCB and the total area is larger than the sum of the chips areas. The addition of primary optics lower the luminance as well, not only due to the flux decrease linked to optical losses, but to the increase of the apparent size of the source, when, for instance, diffusers to scatter the light are used, or when a dome lens is installed to increase the flux extraction from the dies.
In fact, diffusers and light scattering are the first option to achieve the desired light mixing mentioned as second goal before, but do not fulfil the luminance constraint: they produce large area sources with low efficiency (owing to back scattering and absorption) and in the end, the lumens per square meter and stereo radian are reduced.
A successful attempt to overcome these limitations is the SSL4EU project where a new type of light engine was designed, based on a tightly-packed chip array combined with a special type of high efficiency optics, the so-called shell integrator, developed by LPI that barely decreases the luminance of the light source. The high density of chips in this light source requires advanced thermal management strategies since the amount of power to be dissipated is not compatible with inexpensive passive cooling.
This trend of packing the chips of a light source as close as possible deals with the preference of most luminaire manufacturers to retrofit solutions: if the lamps should maintain the current type of secondary optics the light source should be as similar as possible to conventional light sources.
Progress beyond the State-of-the-art in SSL illumination optics
Retrofitting is a constraint that may help to get a rapid market penetration but compromises the overall performance of the SSL lamp. New type of luminaires should be developed adapted to the special features of the new type of light sources, the LEDs.
The shell integrator mentioned above is a technology with room for improvement. For instance, its size, relative to that of the LED array, could be reduced to achieve a more compact light engine (currently, the shell mixer base diameter is 2.5 times the diameter of the array). Reducing the shell mixer size is a challenging design task but the benefits affect both the lamp costs and Zhaga compliance.
As explained, one way to keep the luminance of the LED array as high as possible is packing them as tightly as possible, to achieve the minimum cumulative area for the equivalent source. Another way to keep the ètendue as low as possible, and the luminance high, is to spread the chips over a wide area but add an optical feature that collimates the LED beam into a narrower beam, so the flux per solid angle increases, to compensate for the reduction of flux per unit area. A novel type of array optics that are being explored in this project is the so-called Phase Space Combiner (PSC [patent pending]), which keeps the effective brightness of the LED array.
The advantage of this approach is that thermal management is much easier and low cost passive cooling heat sinks can keep the LEDS working at reasonable temperatures. The PSC is based on a thin lens array which covers the complete area and concentrates the LED light to some extent, thereby filling up the phase space “gaps”, as shown in the following figure:
Basic version of the PSC optics
The thermal benefits of architectures of this kind can be quantified in a simple way by assuming the Power density to be dissipated is reduced by the inverse of the concentration factor, i.e. ALED/Aoptics, where ALED is the chip area and Aoptics is the area of the lens facing each LED.
Considering 2x2 mm2 square lenses on top of 0.5x0.5 mm2 chips, the factor becomes 1/16 approximately. Although this is clearly an upper bound (we are assuming there are no conduction losses), it shows the potential of this approach to utilize simple dissipation strategies. We can expect the PSC engine would demand heat sinks having thermal resistances 4 times larger than those needed for a tightly packed chip array, for the same operating junction temperatures.
Passive heat sinks lower the investment on thermal components by a factor ranging 1/2 to 1/50 depending on the active cooling strategies they help to eliminate (usage of fans, heat pipes) but also have other practical benefits, like the elimination of noise (the case of fans), higher reliability and lower lamp power consumption.
This basic version of PSC light engine mixes the chips colours in the far field, but not in the near field, so it might not be compatible with applications where not only the illuminated area, but the light source itself, needs to look uniformly white. A novel advanced version of this light engine [patent pending, see figure below] has a light mixing cavity where a high reflectance lambertian white diffuser mixes the light coming from all chips, and at the same time lets the mixed light escape at a set of preferred locations: