• Newsroom Home
  • Astronomy
  • Biomedical Optics & Medical Imaging
  • Defense & Security
  • Electronic Imaging & Signal Processing
  • Illumination & Displays
  • Lasers & Sources
  • Micro/Nano Lithography
  • Nanotechnology
  • Optical Design & Engineering
  • Optoelectronics & Communications
  • Remote Sensing
  • Sensing & Measurement
  • Solar & Alternative Energy
  • Sign up for Newsroom E-Alerts
  • Information for:
SPIE Photonics West 2018 | Register Today

SPIE Defense + Commercial Sensing 2018 | Call for Papers

SPIE Photonics Europe 2018 | Call for Papers




Print PageEmail PageView PDF

Illumination & Displays

Non-imaging optics combine LEDs into one bright source

A single optical element combines light from several LEDs, forming an ultrabright light source with flexible shape and emission characteristics.
26 June 2006, SPIE Newsroom. DOI: 10.1117/2.1200605.0197

In the coming decades, bright white LEDs will ultimately replace most conventional light sources. These devices will bring us superior light quality, unsurpassed lifetime, compactness, and drastically reduced power consumption. However, today's LEDs produce neither as much light-per-area as discharge or incandescent lamps, nor sufficient light-per-device: the brightest-white LEDs currently produce only about 50lm each, while many applications require several thousand lumens. Also, although prices are falling steadily, LED light flux is expensive compared to conventional light sources. Furthermore, they show visible color and luminance variations across the chip surface as well as color and flux variations between different chips. We can eliminate some of these problems by combining and mixing light from several devices before using refractive and reflective surfaces to shape the desired irradiance or intensity pattern.

Standard optics for high-brightness LEDs use a modular approach, in which every LED has its own collimation optics. In most automotive headlamp prototypes, the number of lenses visible from the outside is the same as the number of LEDs used. In illumination designs, homogeneity and color problems are commonly solved by using diffusers. This may be acceptable for wide beam patterns (i.e general illumination) but it is counterproductive for creating collimated light. Some applications require sharp gradients in their intensity pattern: for example, car low beams must illuminate the road but avoid blinding oncoming traffic. Most LED designs—such as conventional headlamps—use shutters or baffles, whose shadow is projected into the far field to create the desired pattern features, resulting in efficiency losses. The light-source's etendue—which is basically the emitting area multiplied by the solid angle into which the light is emitted—determines the minimum aperture size of the optics required to collimate the light into a given narrow angle range. Therefore highly efficient and compact illumination systems—such as condensers for video projector illumination, car headlamps, and other applications—must be carefully designed to avoid increasing the etendue.

We combined two potent non-imaging design methods to create a new optical architecture that is viable for many applications. Here we consider the example of an automotive headlamp. The two basic building blocks are a monolithic lightguide with a freeform exit surface, called an LED combiner,1 and a freeform mirror (see Figure 1). The LED combiner was designed using the flow-line method,2 while the exit surface and the mirror were designed by the simultaneous multiple surface (SMS) method.3–6 The flow-line method ensures that the light from several LEDs mixes together, thus creating a secondary source with the desired dimensions and angular spread. Then, a lens and a mirrorr transform the light from this secondary source into the needed headlight far-field pattern. (By law, headlights must illuminate a number of points: they are typically tested at 20.) The two design methods also ensure that the lit aperture of the headlamp, which must have an intense output and a very tight vertical spread, is as small as possible.

Figure 1. Light from three LEDs mounted on a printed circuit board is combined, mixed, and shaped in this compact module, designed using the simultaneous multiple surface method. The full depth of the module in the projection direction is 66mm.

The straightforward approach to making devices that can emit kilolumens would be to place the LED chips side by side. The heat generated by these chips, however, and the difficulty of removing the heat, would reduce the total luminance. Also, commercial LEDs have pronounced variations in luminance across the emitter. Such patterning can cause undesirable artifacts in the beams of collimating or condensing lenses. The LED combiner (Figures 2 and 3) eliminates these artifacts because it does not preserve the topography of the light (i.e., it doesn't transmit an image of the LED).

Figure 2. An LED combiner with a freeform lens. The assembly's total height is only 24mm.

Figure 3. A combiner mounted on a printed circuit board gathers and mixes light from three emitters. Each LED is 0.9mm wide.

The combiner is a special case of an optical manifold that provides an etendue-limited combination of multiple phosphor-coated white or multicolored LEDs. In the embodiment shown, the combiner gathers the light from three devices via total internal reflection and creates a single prescribed distribution at its exit aperture. The combiner has several advantages. First, it eliminates color and flux variations among the different LEDs and the illuminance and color variations across the emitting face of each chip. Second, sharp edges at the exit aperture of the combiner are used by the secondary optical elements to create the vertical intensity gradient at the horizon. Further, the luminance distribution at the combiner exit aperture is not sensitive to variations of position of the LEDs with respect to the combiner. The resulting light source has better uniformity and definition than other high-luminance sources such as incandescent filaments or arcs. One additional advantage of the device is that the LED combiner can be injection-molded using tools with optically smooth surfaces, which avoid problems with scattering losses. We use ultraclean optical polymers (i.e., PC or PMMA) to suppress bulk scattering and absorption.

The thin-film Osram O-Star LED is an almost perfect match for the our combiner because it produces almost all of its emitted flux from its top surface. The combiner typically increases the LED output by 30–40% compared to flat-cover LEDs by eliminating light confinement within the package. The coupling of the LEDs to the guide is achieved by using either an optical gel, index-matching fluid, or a UV-curable optical adhesive. The thickness of this coupling layer and the encapsulant layer that covers the LED chip must be carefully controlled to ensure high coupling efficiency to the combiner.

The SMS design of the two freeform surfaces is based on coupling two pairs of wavefronts perfectly, and another pair of wavefronts partially. In this case, the two wavefronts are emitted from two points chosen on the edge of the LED combiner's exit aperture. The two outgoing wavefronts emitted by the mirror are derived from the target intensity pattern. The SMS calculation ensures that all rays emitted from the edges of the light guide will leave the optical device exactly as dictated by the two outgoing wavefronts.

To form a full LED headlamp, several modules—each consisting of three LEDs, the combiner, and a mirror—are placed side by side. More or more-powerful LEDs can be combined so that a single module can work as a headlamp. We have designed both low and high beams that meet ECE (Economic Commission for Europe) headlamp standards. Detailed raytrace simulations predict extremely-high overall optical efficiencies (76% and 82% respectively for low- and high-beam designs) and intensities (44000cd for low and 82000cd for high beam). The LED-to-light-guide placement error, which was as large as ±0.1mm, has virtually no effect on the beam pattern.

We have demonstrated that the architecture described here—which consists of a group of flat top-emitting LEDs, a combining optic, and two freeform surfaces—is versatile and very efficient. It preserves the etendue, tolerates misalignments, and handles luminance variations across the chip surfaces. It also handles color and flux mixing of several LEDs. This architecture is particularly suited for high-flux high-collimation applications.

The presented design ‘LED combiner/SMS projector’ was developed for the EU project TST3-CT-2003-506316: ‘Integrated communicating solid-stage light engine for use in automotive forward lighting and information exchange between vehicles and infrastructure’.

Oliver Dross
LPI Europe
Oliver Dross heads LPI's European research and development activities and designs and develops LED and incandescent exterior automotive lighting. He previously worked for Hewlett Packard and LumiLeds Lighting in the field of high-brightness LEDs. His MS in Physics is from the University of Konstanz, Germany. He has written and presented several papers on non-imaging optics at SPIE conferences.
J.C. Miñano and Pablo Benitez 
CEDINT, Technical University of Madrid (UPM)
A. Cvetkovic
E.T.S.I. Telecomunicación, Universidad Politéica de Madrid
Julio Chavez
Light Prescriptions Innovators LLC
Altadena, CA

1. J. C. Chavez, W. Falicoff, J. C. Minano, P. Benítez, O. Dross, W. A. Parkyn, and R. Alvarez,
Optical manifold for light-emitting diodes,
Patents pending, US patent publication 20050243570.
2. R. Winston, J. C. Miñano, P. Benítez,
Nonimaging Optics,
Elsevier Academic Press, Oxford, 2005.
3. P. Benítez, J. C. Miñano, Simultaneous multiple surface optical design method in three dimensions,
Opt. Eng.,
Vol: 43, no. 7, pp. 1489-1502, 2004.
4. O. Dross, P. Benítez, J. C. Miñano, Review of SMS Design Methods and Real World Applications,
Proc. SPIE
5529, 2004.
5. J. C. Miñano, P. Benítez, J. C. Gonzalez, W. Falicoff, H. J. Caulfield, SMS 2D Design method,
High Efficiency Non-Imaging Optics,
Patent No 6,639,733, 2003.
6. P. Benítez, J. C. Miñano,
Three-Dimensional Simultaneous Multiple-Surface Method and Free-Form Illumination-Optics Designed Therefrom,
Patents pending, US patent publication No 20050086032.