Next-Generation Solid-State Lighting Fixtures Need Optimized LED Drivers

Sep 1, 2009 12:00 PM
IRENE SIGNORINO Director of Marketing Microsemi Corp., Irvine, Calif.

For optimum performance, high-power LED light sources require intelligent drivers that are able to take into account specific colorimetric, electrical, and temperature behavior.

HIGH-POWER LEDS are a completely different light source compared to previous technologies used in general illumination applications. Therefore, to leverage the optimal fixture life-expectancy and performance enabled by progress in LED manufacturing and packaging processes, we must also consider a specialized approach for driving LEDs.

Today's LED drivers must manage complicated behavior that is specific to the light source, including tightly interdependent parameters with highly nonlinear behavior such as:

• Photometric traits related to luminous flux and efficacy
• Electrical traits related to current, voltage, and power
• Thermal characteristics related to junction temperature

For example, the LED spectrum shifts to longer wavelength as temperature rises, and the absolute value of these shifts varies for red, green, and blue LEDs. The opposite happens when LED current rises. To deal with this behavior, LED driver systems must be optimized for realistic operating ranges and specific application requirements.

Using a non-optimized power supply developed for a different light source might lead to very unpleasant and unsatisfying experiences — flickering, cold and un-tunable light, low energy efficiency, and premature fixture failures — that defy the value proposition underlying the users' choice of LEDs.

A key step is picking the right power supply. High-quality white or RGB illumination, and LED drivers simply cannot be achieved with low-cost, off-the-shelf power supplies. A total system approach is required spanning colorimetry, optics, power electronics, thermal management, and control theory.

KEY LED FIXTURE PARAMETERS

Many fixture variables, which are tightly connected to regulatory requirements, are made possible only by using intelligent LED driver systems. Among them:

• Light efficacy as measured by l m/W
• Life expectancy as measured when light output is decreased to 70% of the initial value
• Light quality measured by the Color Rending Index (CRI)
• Power-factor correction (PFC)
• Total harmonic distortion (THD)

Other key requirements affected by the LED driver are color temperature (CCT), color spatial uniformity, and color maintenance. Finally, an intelligent LED driver will support smart dimming and a communications interface to get the most out of the fixture.

Most LEDs exhibit significant parametric differences in luminous flux, spectral power distribution, efficacy, and efficacy degradation due to variations in materials and manufacturing processes. Luminous efficacy alone can decrease from 0.2% to 1% per degree (°C) rise in temperature, and the percentage is higher with aged LEDs.

Some variations can be mitigated using typically costly LED binning practices (i.e. sorting LEDs by forward voltage (Vf) or other key parameters). However, LED fixtures can't meet stringent luminary requirements unless LED drivers can address the following issues:

• Current average and ripple control to extend the life expectancy of the LED and improve end-user experience
• Protection against events such as overvoltage and overtemperature, etc.
• Effects of LED aging and temperature on its spectrum and light intensity to ensure end-user consistency
• Accurate color representation and maintenance
• More complex dimming and other remote-control commands to ensure proper behavior of the LED fixture in response to an external event

LED life expectancy varies whether it is used indoors or outdoors, and across operating conditions. Although LED manufacturers claim life expectancies in excess of 50,000 hours, failures can occur after as little as 10,000 hours without an optimized power supply. Life expectancy claims should include an estimate of Mean Time Between Failures (MTBF) for the entire fixture, not only for the LEDs.

Finally, PFC and THD parameters are also important and related. Power factor is the ratio of real power to complex power, with two components: (1) the phase shift induced between the sinusoidal input voltage and current due to a load's inductive or capacitive nature (called displacement power factor — DPF — and defined as the cosine of the phase-shift angle, ϕ), (2) the distortion factor, which is related to the non-linear load characteristics and expressed as a function of the load's THD.

The load's power factor is, therefore, the product of its distortion factor and its displacement power factor. When power factor is low, electric utilities must supply more current per given amount of real power, which impacts energy savings. Controlling the driving system's THD is critical to meeting power-factor requirements.

Intelligent LED drivers can optimize each of the aforementioned parameters and enable many advanced features including non-flickering, dimming, remote LED fixture monitoring/management, color mixing, and custom color setting.

DIMMING

Dimming adapts light output to specific applications, and can substantially increase power efficiency and life expectancy of the fixture while improving the end-user experience. On one hand, increasing energy savings and time between failures directly affects the value proposition of the LED fixture, often substantially reducing the payback period of such fixtures. On the other hand, proper dimming is critical as LED light sources pose different challenges than traditional sources, and legacy approaches such as TRIACs do not work well with LEDs. Poor dimming methods cause many problems, such as RFI, power harmonics, audible noise, and flickering.

An intelligent LED driver can handle more traditional analog dimming (Constant Current - CC) methods, where the current amplitude is varied to obtain a varied light-output result, as well as pulse-width modulation (PWM) dimming methods used in high-volume LED display applications to avoid color shifting. With PWM, the LED is driven with a pulsed current in either a fully on or off state.

By varying the pulse duty cycle from 0% (off) to 100% (full-scale), you can program a full brightness range. The light source appears dimmed because the pulse frequency is set above 100 Hz, where the human eye averages the pulsed display's brightness. Because the pulses occur with only full-scale amplitude, there is no change in forward current and, theoretically, no color shift to degrade the end-user experience.

REMOTE CONTROL AND COLOR MANAGEMENT

Besides dimming, next-generation solutions must remotely control LED fixtures. This requires support for standard communications protocols, signals, and interfaces including linear voltage control (0 to 10 V), digital multiplex (DMX512), digital addressable lighting interface (DALI), power-line communication (PLC), and relevant domestic standards including Universal Power-Line Bus (UPB) and ZigBee.

Finally, fixture products can be differentiated across extended applications by using advanced color management. By enabling various colored LEDs to be mixed, users are given full control of color-selection sequencing.

To create LED-based white light, you can use either phosphor-based white LEDs or RGB LEDs. The variable color point of RGB-based white light is superior for high-end luminaries and architectural lighting. Color-management systems ensure high color accuracy and support color-maintenance requirements. One example is the very fast and accurate system for RGB luminaries and displays (Fig. 1), which includes RGB luminaries, color sensor, color manager, and drivers.

Color-management accuracy depends on the performance of the RGB color sensor which, ideally, should have CIE color-matching functions as illustrated in Fig. 2. However, most commercially available RGB sensors don't support these functions, so compensation methods are necessary. Microsemi has developed one such compensation method. During the calibration process, the output of the color sensor [R, G, B] is connected to the input of the color manager through low-pass filters. The color manager's calibration block then converts the color-sensor value to a tri-stimulus value, and a digital controller compares reference and measured data.

Using a digital PI algorithm, the color manager develops RGB PWM for the LED drivers. Meanwhile, LED drivers control luminary RGB LEDs so that mixed light will be represented as a targeted white-point temperature. In an advanced LED fixture, color temperature can be changed by the user through a communication protocol such as power-line communication (PLC), DXM412, or DALI. To improve system accuracy, the temperature sensor can be connected to the color manager to adjust the influence of the temperature on LED spectral characteristics. The Table below presents major parameters of this color-management system.

As LED performance improves, so too must the capabilities of LED driver systems. Advances in dimming, remote fixture control, color mixing, and customer color setting contribute to improved fixture performance, energy efficiency, life expectancy, and an enhanced user experience. Specially designed LED drivers enabling these capabilities represent a dramatic departure from earlier low-cost power supplies that were optimized for legacy lighting technologies.

PARAMETER	VALUE
Color accuracy	Δu`v`<=0.002	Color loop settling time	<180 msec	PWM resolution	12 bits	PWM frequency	120 Hz to 2000 Hz	White point set up	CT and Y, or X,Y,Z

Acceptable Use Policy blog comments powered by Disqus