When it comes to creating white light using LEDs, there are many different methods, all of which have their advantages and disadvantages. First, the most common and traditional is chip-level conversion, where the converting material is integrated directly onto a blue LED die or fills up the LED package volumetrically. A second method converts the blue light of a diode far away from the chip, which utilizes a mixing chamber and remote optic integrating a converting phosphor, fluorescent dye, quantum dots, or other converting material. Yet another method is color mixing using LEDs of different colors such as red, green, and blue, where each LED’s intensity is varied additively to create the desired color temperature. Each method presents different challenges and brings to the designer different advantages from a system performance perspective with regards to system efficacy, CRI, optical efficiency, energy efficiency, heat sinking and other design considerations. It is important for designers to understand the nuances of each method when deciding on which method to use, whether one is selecting the light engine to be used in design of a new fixture, or selecting the type of fixture to be used in a given application.


Figure 1. OSRAM’s Golden Dragon Plus exhibits true chip-level conversion, where the phosphor is deposited directly on the die.

The first method to generate white light using LEDs, which is also the oldest and most common, is chiplevel (or package level) conversion. There are essentially two methods of package level conversion. In what I call true chip-level conversion, the active converting material (such as a phosphor) is deposited directly on the blue LED die and the exiting blue light is converted almost immediately, as in the case of the OSRAM Golden Dragon Plus (Figure 1).The other most widely used method for phosphor application and conversion is known as volumetric conversion such as in the Nichia 183A (Figure 2). In volumetric conversion, the phosphor effectively floods the chip package and the blue die rests underneath a sort of pool or film of phosphor.

While either method can be argued to be equally efficient and effective at converting blue light to white light as phosphors are intended to do, each presents a different set of challenges for secondary optic design.

Figure 2. Nichia’s 183A Series is an example of an LED using volumetric phosphor conversion, where the blue die sits beneath a die film.

An LED utilizing true chip-level conversion better approximates a point source of light compared to an LED which uses the volumetric method. This has a major influence in optical design applications where a point source is essential to light quality and efficiency. For example, most parabolic optics will require a point light source to maintain efficient beam control. When using a volumetrically converted chip, shadowing and dark/light rings are often observed due to its diffuse nature, where the larger emitting area creates superfluous reflections within the optic. Another example of the importance of focal point location occurs in refractory optics, such as batwing style lenses, which perform very extreme light bending and have a very low tolerance on focal point location, such that much of the light escaping from a volumetrically converted chip enters the optic at incorrect incident angles, resulting in color separation and undesirable Fresnel scattering losses which can result in lower optical efficiency.

There are indeed times when a volumetrically converted LED is desirable, however. In light guides, the larger diffuse surface of the LED widens the area of injection, and effectively narrows the spacing between LEDs which will promote a shallower mixing area, thus enabling larger emitting areas for the light guide. This can lead to thinner and more attractive bezel design in, for example, LCD panels. Also, volumetric conversion can be a reason to choose a particular LED when the LEDs are placed very closely behind a diffuser lens that requires higher backlighting uniformity. As can be expected, the extra few millimeters of emitting surface area provided by a volumetric conversion can help with uniformity at the surface, especially in the case of very high clarity diffusers.

In the case of reflective optics where flat and angled reflectors are used to direct and control the light, the chip-level phosphor conversion method perhaps matters a little less. A key point to remember in a situation such as this is that highly specular surfaces can be glary and distracting to people in the space. However, when flat reflectors are used, the concept of a focal point usually does not come into play and for the most part light exiting from an LED, or array of LEDs, has already achieved a far-field characteristic by the time it first reflects off a flat optic.

Figure 3. Philips Fortimo LED Module uses an optic coated with a remote phosphor, separated by a mixing chamber from the blue LEDs beneath it.

Designing for high optical efficiency is one of the most cost-effective ways of maximizing the efficacy of an LED lighting system. By paying attention to the type of optics needed for the application, and selecting the right LEDs to be used with those optics based on knowledge of the phosphor conversion method, designers can realize easy and significant gains in system efficiency.

In a remote wavelength conversion optic, discrete blue LEDs are mounted in an array and directly illuminate the inside surface of a lens which contains the conversion material. As previously noted, there are many different types of conversion methods available on the market today including phosphors, fluorescent dyes, quantum dots, or other conversion technology. Typically, the lens is mounted as part of a mixing chamber several centimeters away from the LEDs to provide uniform light at the surface. This method results in a number of optical and thermal effects designers should be aware of.

Side Effects

Thermally, there will be heat generated on the lens from the wavelength conversion. Designers should be sure to select materials with high conversion efficacies to minimize this heat, which can cause rapid deterioration of the lens substrate or the conversion material itself. This results in the need for creative heat sinking of the lens, which is often a challenge since lenses are not often made from highly thermal conductive material, and they do not frequently have a very large or tight thermal interface with heat sinking materials.

Remote conversion optics, such as those used in the Philips’ Fortimo™ LED Module (Figure 3), often result in large, diffuse emitting areas that can pose challenges in down lighting and spot lighting for tight beam control. Just as there will be losses from the conversion, a designer should be aware of high losses from using a large diffuse source where a point source should be, e.g. in a parabolic lens.

Figure 4. In additive color mixing, red, green and blue light combine to make white light. Depending on the spectral output of the LEDs, not all three colors are always necessary.

Remote conversion methods can offer advantages in binning and color uniformity, which can reduce the overall cost of a system. Especially when used to control white light exiting from chip level LEDs, using wavelength conversion to more finely tune output during production can allow OEMs to purchase larger, and therefore cheaper, bins of white LEDs.

Aesthetically, many remote conversion technologies have a yellow, green, or red tint when the light engine is off (as opposed to the normal white diffuser or just fully visible CFL or incandescent light source), an unfamiliar effect that some lighting designers have found hard to swallow. This is a matter of taste, and not necessarily a good reason to move away from remote conversion technology.

Beyond wavelength conversion technologies such as chip level, volumetric, or remote conversion technologies, the third way to create white light using LEDs is by color mixing.Like an RGB pixel of any standard video monitor, color mixing uses the principles of additive color (Figure 4) to combine two or more colored LEDs to create white light. By far it is the most electronically complex method of creating white light, as it depends on some type of control of the LED RMS current through either Pulse Width Modulation (PWM) or Amplitude Modulation, which dims LEDs of different color combinations (such as red, green, blue, mint green, amber, even yellow or cool white) at appropriate intensity levels to achieve white light of a desired color temperature. As with any lighting system, there will always be advantages and challenges.

Color mixing, when done correctly, can offer improved fixture to fixture color uniformity. Capitalized upon by Cree through their TrueWhite™ technology (Figure 5), LEDs of different color are actively controlled via dimming through either some type of optical, thermal, or electrical feedback system, which helps improve consistency among systems.

Where chip level or remote conversion techniques require different chemistries to achieve different CCTs such as 2700K, 3000K or 4100K, LED color mixing often requires nothing more than a simple firmware change to set the CCT of the system. This introduces a higher level of simplicity on the manufacturing side, which helps lower the overall cost of the system.

The PWM controls used in an LED color mixing system enable the system to be inherently more controllable during dimming in application, which opens the door for better color performance in a dimmed down state. A common complaint observed with dimmed LED systems is the even further absence of a red component, which gives dimmed LED lighting a faint, ghostly appearance, far from the warmer dimmed tone of an incandescent bulb which is more generally preferred. By designing for increasing levels of red and/or amber LED light while dimming, the system can more accurately approximate the performance of an incandescent bulb.

Figure 5. Cool white and Red LEDs are used to create warm white light with high CRI in this Cree down lighting module.

Since issues such as color uniformity and dimming can be effectively solved using the control techniques described above, LED binning becomes less of a concern, which enables the usage of less expensive and larger bins. This lower LED cost, however, may be offset by higher costs in electronics and firmware.

Luminous efficacy and CRI can be significantly increased using LED color mixing with the use of red and green LEDs. Since the value of a “lumen” is based off the human eye sensitivity curve and CRI is dependent on the blackbody spectrum of Tungsten, adding strong green and red components can give a significant boost to lumen and CRI performance.

As can be seen, LED color mixing in general does enable improved color controllability and potentially superior CRI and efficacy performance compared to the other two methods discussed. Of course, the types of systems mentioned above do present some challenges. A color mixing system will invariably require an array of LEDs which can limit beam control options, and will most certainly require an efficient diffuser to reduce shadowing on the application surface. The inclusion of some mixing chamber or mixing distance into any system using this technology is often desirable as well. Additionally, care must be taken during the design to account for varying lumen maintenance of different color LEDs. Since Red LEDs often have a much longer lifetime and shallower lumen maintenance curve than their cooler counterparts, care must be taken to balance system color performance from a programming perspective as well as heat sinking to make sure this higher lumen maintenance is accounted for so the system does not experience a red shift over the course of its life.

Since all three conversion methods to create white light discussed in this series — chip level (either volumetric or true chip level), remote, or color mixing — all offer their advantages and disadvantages, each may lend itself more pertinently to one application over another. It is ultimately up to the designer to determine which method to choose whether it is choosing the appropriate light engine for a given fixture, or choosing the appropriate fixture for a given application.

For more information, Click Here