In recent years, Light Emitting Diodes (LEDs) have become a viable alternative to conventional light sources. The overriding advantages long life, high efficiency, small size and short reaction time have lead to the displacement, in ever increasing numbers, of incandescent bulbs. One of the markets where this change has become most evident is Automotive, where LEDs are used now not only for backlighting dashboards and switches, but also for exterior illumination in Center High Mounted Stop Lights (CHMSL), Rear Combination Lamps (RCL), turn signals and puddle lighting.

Despite the long life and low failure rates of LEDs, cars can be found, on occasion, with failed LEDs in their CHMSL. Most often this is due to a flawed circuit design wherein the LEDs were allowed to be overdriven. It is with that supposition in mind that this application note is written: to identify, characterize and comment on LED behavior and failure modes in serial and matrix circuits.

The purpose of this application brief is to show a method to determine the maximum permissible power dissipation of the LEDs in a display application, and allow the junction temperature (T_J) to remain below its rated value. Junction refers to the p-n junction within the LED.

Nowadays, applications increasingly make use of LEDs as a replacement for traditional light bulbs. For example, LEDs are frequently used in the design of automobile tail lights, signal lights, traffic signals, variable message signs, ...

LEDs provide several advantages over traditional light bulbs, such as smaller size and longer life. In many applications, the LEDs must be driven with intelligent control circuitry. According to the task at hand, this control circuitry must be able to fulfill various functions and tasks. In the following pages, solutions are provided for various application areas. These solutions are principal suggestions, not a concept ready for series production.

One possible task for control circuitry is regulation of intensity, in case the LED brightness must be set to various levels. A solution is described in the section “Dimming“. In addition, the specified brightness should be maintained at a constant level. Fluctuations in the supply voltage, for example, could lead to significant variations in current. In this case, one must insure that the current through theLEDs and thus the brightness is maintained at a constant level. This problem is covered in more detail in the section “CurrentRegulation“.

Another task for control circuitry is failure recognition. Modules consist of individual LEDs which can be tested for total failure. Additional information can be found in thesection “Failure Recognition“. A particular characteristic of LEDs is their strong temperature dependency. Since LED brightness is strongly dependent on temperature, the driver circuitry can carry out temperature compensation. Two possible approaches are described in thesection “Temperature Compensation“.

Furthermore, it may be necessary to adapt the driver for LEDs in different brightness groups by means of hardware selection. This is described in the section “Adjusting for Different Brightness Groups“. In the following applications, a PICmicrocontroller is used as a controlling unit.

The first true ancestors to the Indium Gallium Nitride (InGaN) LED evolved last decade. These took the form of blue LEDs utilizing Silicon Carbide (SiC) as the active, light-emitting material. These early LEDs were characterized by very low light output, less than 2cd/m². The next generation of blue LEDs relied upon SiC as a base layer only and employed Gallium Nitride (GaN), grown directly on the SiC substrate, as the active, light-emitting epitaxial layer. This process initially increased light output by a factor of eight. The final iteration saw the introduction of Indium (In) to the epitaxial layer to form InGaN. This development further boosted light output by a factor of five - a full 1300% increase in intensity over the first SiC LEDs. Today, through advances in process, packaging and thermal transfer technologies, light output continues to evolve.

Besides increasing the intensity of blue, and by extension, white LEDs (since all white LEDs use a blue chip in conjunction with a light converter, or phosphor), the InGaN process has replicated two new colors: verde and true green. These unique colors, alongside InGaN’s high intensity and inherent reliability, has greatly increased their proliferation into applications once reserved solely for incandescent lighting: traffic signals, realcolor displays, message boards, moving signs, dashboard backlighting, battery flashlights and toys.

While the InGaN process produces the brightest light output across blue, verde, true green and white, it is important to understand that the wavelength of the light emitted is strongly dependent upon the forward current driven through the device, and that in order to avoid shifts in color, careful consideration must be paid to dimming strategies. This application note, then, will examine methods for dimming InGaN LEDs with little or no effect on wavelength.

What is Surface Mounting? In conventional board assembly technology the component leads are inserted into holes through the PC board and connected to the solder pads by wave soldering on the reverse side (through-hole assembly). In hybrid circuits (thick and thin film circuits) ìchipsî, i.e. Ieadless components, are reflow soldered onto the ceramic or glass substrate in addition to the components already integrated on the substrate. Surface mounting evolved from these two techniques.