Solid-state lighting is beginning to deliver on long-promised energy savings, explains ALI FAWAZ, but to broaden adoption developers need to continue to drive cost out of the control circuitry while features such as dimming must still be supported.
LED-based lighting has begun to impact the huge amount of energy used for lighting in developed regions of the globe, but more is expected from solid-state lighting (SSL) technology. To fulfill forecasts, for instance, of a greater than 50% penetration of the commercial lighting sector by the end of this decade, SSL manufacturers need to further reduce costs for the associated lamps and fixtures. The electronics used for LED control are a prime target for cost-reduction efforts, and integration at the silicon chip or IC level is focused on helping to meet that goal. Still, the electronics must be full featured with dimming support as dimming can further impact energy savings while also delivering a better environment for workers and enhancing ambience in many applications.
Indeed, consider the aforementioned commercial lighting sector. According to a paper published in 2014 by Worcester Polytechnic Institute, lighting accounts for nearly half (349 TWh/yr) of the electricity budget of commercial buildings in the US. The huge energy cost associated with such usage levels means that the cost-of-ownership advantage of LED-based lamps is often sufficient to win market share in new projects. But developers must still strive for improvements in functionality that lower up-front acquisition costs.
The dimming electronics are a particularly challenging area for SSL developers. Despite the LED itself being inherently dimmable, many early generations of LED lamps were not compatible with conventional dimmers. Moreover, traditional switch-mode power supply IC controllers proved to be inadequate for LED ballast applications. As a result, controller ICs for LED ballasts increasingly use digital technology, especially in the dimming portion of the IC. With this concentration on the controller IC, the interface between the LED controller IC and the dimming control (dimmer) has been largely ignored. A well-designed and stable dimming interface is critical for consistent light quality and to achieve the reliability needed for commercial and industrial applications.
Methods for incorporating 0-10V dimming control
While phase-cut dimming is commonly used for mass-market residential applications, there are issues around inherent flicker that limit its use in commercial markets. In commercial indoor and outdoor scenarios and even in high-end residential lighting where color changing is not required, 0-10V dimming is preferred by many lighting designers and specifiers. There are two methods of 0-10V dimming control. In one method, the controller (dimmer) sources current to the LED driver; this is defined and supported by the ESTA E1.3 standard and is a preferred method in theatrical or entertainment technology applications.
In the second method, the controller (dimmer) sinks current from the LED driver. Relative ease-of-use makes this second method popular for the widest range of commercial applications. Key technical specifications for the second method, which are defined in the IEC60929 Annex E Technical Standard, are:
• Minimum sinking current to the dimming controller (dimmer) is 10 μA and maximum sinking current is 2 mA.
• Under no circumstances should the interface circuit terminals to the dimming controller (dimmer) produce a voltage exceeding +20V nor can it be less than -20V. The driver/ballast should not be damaged when dimming voltage is between +20V and -20V.
• The control terminals of the interface circuit shall be reverse polarity protected. In the case of reverse polarity of the interface control terminals, output light should be at minimum or turned off.
• The dimming circuit interface should produce stable output light for a dimming control voltage between 0-11V.
• When the signal of the dimming controller (dimmer) is 10V or higher, output light should be at maximum. When the signal of the dimming controller (dimmer) is 1V or lower, output light should be at minimum or off.
• If no dimming controller (dimmer) is used, the dimming terminals are usually kept open and output light should be at maximum. If the dimming terminals are shorted together, output light should be at minimum.
• The supply wire of the dimming terminal is purple and the return is grey.
Additionally, double or reinforced insulation/isolation from all hazardous voltages including the input voltage is required for safety in all cases where the dimming controller (dimmer) circuitry is user-accessible. Isolation further improves dimming performance by keeping high switching noise away from the dimming signals.
Developing designs for a dimming interface
Two earlier figures show typical design solutions for a dimming interface circuit. Fig. 1 is a transformer-based dimming interface circuit and Fig. 2 shows an opto-coupler-based dimming interface circuit. In both circuits, the dimmer signal is converted into a pulse to facilitate having a signal proportional to the dimming signal on the other side of the boundary where the main LED IC controller resides. The pulse is usually averaged and fed to the dimming pin of the LED controller.
The transformer-based dimming interface does not need a bias voltage but suffers from inaccuracies over temperature change. Moreover, the transformer is costly and demands a relatively large PCB (printed-circuit board) footprint. An external square wave pulse is also needed; this is usually implemented by using the gate drive of a low-side power MOSFET. The gate drive signal has fast rise and fall edges that could exacerbate EMI at the dimmer. High voltage transients, primarily surge voltage, at the gate of the power MOSFET further imposes clamping requirements on the dimming signal around the isolation transformer. Many of these issues are addressed and the performance of the transformer-based dimming interface is significantly improved by the use of a main LED controller that can generate a controlled square pulse and provides temperature compensation for the dimming signal.
The opto-coupler-based dimming interface shown in Fig. 2 is more complex than a transformer-based design but also can be more accurate. It requires a bias voltage and at least two op-amps in addition to the opto-coupler. It generates the square wave on its own, eliminating some of the issues that come with using the gate drive signal of the power MOSFET.
In both interface circuits, the design parameters are fixed to a given specification and not readily changeable. Changes in parameters such as sink current into the dimmer, minimum duty cycle (which determines the minimum programming voltage at the LED IC controller), and the operational mode of the dimming interface all require redesign.
Benefits of simplified circuit designs for dimming interfaces
As is often the case in the semiconductor industry, the best way to solve a circuit design problem can be through an IC that is purpose built for the task at hand. Assuming that the application at hand is sufficiently large to support the design and manufacturing of such an IC, the benefits can include higher performance and lower cost. And the SSL sector is certainly growing to the point to make purpose-built ICs feasible.
Fig. 3 illustrates an approach in which a large portion of the discrete elements in the opto-coupler-based design are integrated into such a compact IC, the Infineon CDM10V. Essentially all of the circuitry shown on the left side of the opto-coupler in Fig. 2 is reduced to this IC.
In addition to simplified design, the IC approach reduces the cost and size of the overall system, lowers assembly costs, and improves reliability. Programmable parameters also provide flexibility to reuse the circuit design in multiple luminaire developments. One-time programmable settings include resistor current, minimum duty cycle, pulsewidth modulation signal frequency, and dim-to-off functionality. The nearby table summarizes the programmable features. The IC can even be configured in a transparent mode for direct output of a source pulsewidth-modulation signal.
This simpler approach to a dimming interface circuit provides flexibility across a wide range of dimming applications in industrial and commercial lighting such as troffers, downlights, sconces, under-cabinet lighting, office lighting, and more. The circuit could even be used for applications outside of general illumination such as LED-based signage. Indeed, lighting manufacturer could deploy one hardware design for an entire platform of commercial LED ballasts, thereby enabling volume application of the dimming technology.
ALI FAWAZ is senior staff application engineer at Infineon Technologies Americas (infineon.com).