Saturday, 24 June 2017

Eco Friendly LED & CFL - A Comparative Statement

The problem of harmonics cannot be neglected in cases of installations with high lighting load. This paper presents an analysis of harmonics in a network where lighting is one of the main loads. CFLs and LED lamps with electronic gear are characterized by extremely distorted current, with high total current harmonic distortions. Hence they cause a significant voltage distortion in electrical installations. A comparative analysis is performed on the power quality, maximum loading and economics of CFL lamps and LED lamps.
Greenhouse gases & LED
  The heat generated by conventional electric light bulbs may have been significantly reducing the release of greenhouse gases from natural gas. If all homes switch from (incandescent) bulbs to CFLs, there would be an increase of almost 220,000 tonnes in CO2 emissions in the province, equivalent to the annual emissions from more than 40,000 automobiles. As CFL Contains Mercury, Net mercury emissions for CFL and incandescent lamps is 0.012 mg of mercury per kilowatt-hour and 14% of CFL mercury contents escapes to environment after land fill disposal. CFLs, like all fluorescent lamps, contain mercury as vapor inside the glass tubing. Most CFLs contain 3–5 mg per bulb. As mercury is poisonous, even these small amounts contribute to air and water pollution.
  According to the European Commission Scientific Committee on Emerging and Newly Identified Health Risks (SCENIHR) in 2008, CFLs may pose an added health risk due to the ultraviolet and blue light emitted. This radiation could aggravate symptoms in people who already suffer skin conditions that make them exceptionally sensitive to light. The light produced by some single-envelope CFLs at distances of less than 20 cm (7.9 in) could lead to ultraviolet exposures approaching the current workplace limit set to protect workers from skin and retinal damage. However, industry sources claim the UV radiation received from CFLs is too small to contribute to skin cancer and the use of double-envelope CFLs "largely or entirely" mitigates any other risks.
.
  An LED lamp is a light-emitting diode (LED) product that is assembled into a lamp (or light bulb) for use in lighting fixtures. LED lamps have a lifespan and electrical efficiency that is several times better than incandescent lamps, and significantly better than most fluorescent lamps, with some chips able to emit more than 100 lumens per watt. 
  Like incandescent lamps and unlike most fluorescent lamps (e.g. tubes and compact fluorescent lamps or CFLs), LEDs come to full brightness without need for a warm-up time; the life of fluorescent lighting is also reduced by frequent switching on and off.
  Some governments around the world have passed measures to phase out incandescent light bulbs for general lighting. The aim is to encourage the use and technological development of more energy-efficient lighting alternatives, such as Compact Fluorescent Lamp & LED lamps. Consumers are being encouraged to switch outdated incandescent bulbs to these more energy efficient alternatives. LEDs are more efficient than CFLs but the initial cost is higher so it takes longer to recoup the cost of the bulb. However LEDs last much longer-over 20 years-so they will pay for themselves many times over their lifespan. While an 11w CFL bulb costs $1.25 in bulk, an 8w LED (which produces the same number of lumens as a 40w incandescent bulb) can run anywhere from $10 (available at local hardware stores) to $20.
Environmental impact of LED lamps compared to halogen lamps
  • The environmental benefits of using LED lamps to replace Halogen lamps are unquestionable.
  • At least 4 times less impact on all environmental impact categories throughout its product life cycle.
  • Still significantly lower than that of low voltage halogen lamps even when, extremely high halogen specs are considered.
  The environmental impact calculations are corrected for flux or central beam intensity differences.
  The LED lamp power is doubled and life-time of the LED lamp is reduced by half (sensitivity analysis).
Why Only LEDs
  • LEDs are ideal for use in applications that are subject to frequent on-off cycling, unlike fluorescent lamps that burn out more quickly when cycled frequently, or HID lamps that require a long time before restarting.
  • LEDs can very easily be dimmed or strobed.
  • LEDs light up very quickly. A typical red indicator LED will achieve full brightness in microseconds. 
  • LEDs mostly fail by dimming over time, rather than the abrupt burn-out of incandescent bulbs.
  • LEDs, being solid state components, are difficult to damage with external shock, unlike fluorescent and incandescent bulbs which are fragile. 
  • LEDs can be very small and are easily populated onto printed circuit boards. 
  • LEDs do not contain mercury, unlike CFL.
Basic advantages of LED Light
  • Energy efficient - LED’s are now capable of outputting 135 lumens/watt
  • Long Lifetime - 50,000 hours or more if properly engineered
  • Rugged - LED’s are also called Solid State Lighting (SSL) as they are made of solid material with no filament or tube or bulb to break
  • No warm-up period - LED’s light instantly – in nanoseconds 
  • Not affected by cold temperatures - LED’s “like” low temperatures and will startup even in subzero weather
  • Directional - With LED’s you can direct the light where you want it, thus no light is wasted
  • Excellent Color Rendering - LED’s do not wash out colors like other light sources such as fluorescents, making them perfect for displays and retail applications
  • Environmentally friendly - LED’s contain no mercury or other hazardous substances
  • Controllable - LED’s can be controlled for brightness and color.
Comparison Between Different Light Sources
Energy efficiency
  Energy usage for different types of light bulbs operating at different light outputs. Points lower on the graph correspond to lower energy use Because the eye's sensitivity changes with the wavelength, the output of lamps is commonly measured in lumens, a measure of the power of light as perceived by the human eye. The luminous efficacy of lamps is the number of lumens produced for each watt of electrical power used. The luminous efficacy of a typical CFL is 50–70 lumens per watt (lm/W) and that of a typical incandescent lamp is 10–17 lm/W. Compared to a theoretical 100%-efficient lamp (680 lm/W), CFL lamps have lighting efficiency ranges of 7–10%, versus 1.5–2.5% for incandescent Because of their higher efficacy, CFLs use between one-seventh and one-third of the power of equivalent incandescent lamps. Fifty to seventy percent of the world's total lighting market sales were incandescent in 2010. Replacing all inefficient lighting with CFLs would save 409 terawatt hours (TWh) per year, 2.5% of the world's electricity consumption. In the US, it is estimated that replacing all the incandescent would save 80 TWh yearly. Since CFLs use much less energy than incandescent lamps (ILs), a phase-out of ILs would result in less carbon dioxide (CO2) being emitted into the atmosphere. Exchanging ILs for efficient CFLs on a global scale would achieve annual CO2 reductions of 230 Mt (million tons), more than the combined yearly CO2 emissions of the Netherlands and Portugal.
Conclusion
  LEDs are Competitive, Eco Friendly & Likely to Get Better.Conclusion is that based on eco-friendly, life-cycle assessments and competitiveness, LEDs are about as energy efficient as CFLs as far as their whole life-cycle is concerned. But that seems likely to change, since LED lighting technology is still growing and improving its own performance day-by-day.

Khule Shridhar Shantaram is Associate Professor and Head of Electrical Engineering Department. Kakad Haridas is undergoing ME Electrical Power System at Matoshri COE and Research, Eklahare, Near Odha, Nashik. Birar Dhanshri P is undergoing Diploma Electrical Engineering at K K Wagh Polytechnic, Chandori, Tal- Niphad, Nashik.

A tailored approach to LED Thermal management enhances design while reducing costs

A critical aspect of effective LED –based product designs is ensuring that individual lamps are thermally managed. Although device manufacturers are continuing to improve the durability of LEDs at raised operating temperatures, excess heat in LEDs still has both an immediate impact on performance as well as longer term consequences as can be seen in Fig.1. Short-term, mostly reversible effects include colour shift and reduced light output, while the long-term impact of poor thermal management is accelerated lumen (light output) depreciation meaning a curtailed useful life.
Fig. 1: Impact of temperature on LED life expectancy (Source: Cree)
  As design approaches change and LED technology advances, there is a definite market shift towards the use of mid-power (sub one watt) as opposed to high-power devices (one to three watt) for many LED lighting applications.
  By using a higher number of mid-power devices more closely positioned to one another in an array as opposed to a lesser number of high power devices, a more even light can be achieved. This is especially important in larger designs such as where LEDs arrays are used instead of CFL tubes. This kind of arrangement also simplifies the design, and therefore cost, of the lens/cover fitted over the array as it is not required to mix and spread the light output so much as with ‘point’ light source of a small number of high power devices.
  The approach to thermal design for high-power LEDs is inappropriate and ‘overkill’ for mid-power devices and with careful and early consideration to the design there are opportunities to reduce overall design cost, weight and complexity without impacting performance, reliability and expected life of the lighting assembly in any way.
Thermally managing high-power LED designs
  Despite the growing popularity of mid-power LEDs, higher power devices are still likely to remain the best choice for a significant number of applications until further advances in technology allow even greater lumen per watt output. The power dissipated by high power LEDs is still of course much, much lower than that of other older lighting technologies, however, heat generated in normal operation is concentrated in a much smaller area, base of the individual LED device. By effectively managing the heat in the base, the junction temperature of the diode can be maintained at a level where short-term & long-term performance is not impacted.
  The use of insulated metal circuit boards offers one of the most efficient way of dissipating heat at the required rate and magnitude from designs that utilise high power LEDs. The construction of insulated metal circuit boards comprises a sandwich construction of an aluminium sheet, coated with a very thin electrically insulating dielectric material that possesses a high thermal conductivity. This is then topped with a bonded copper layer that is etched using conventional PCB processes. The LEDs & any other devices required to complete the circuit are then soldered onto the etched copper traces.
  The thin, high thermal efficiency dielectric layer ensures that heat is dissipated to the aluminium baseplate efficiently and effectively. However, in most high power LED designs the thermal performance of the baseplate alone is not enough to provide effective thermal management during the prolonged operating cycles typical in lighting applications. Therefore, the insulated metal circuit assembly must be attached to a larger heatsink or the metal chassis of the equipment or overall product. Using mechanical fixings and a material such as thermal grease tend to provide inconsistent results due to the difficulty in maintaining repeatability of thermal grease application. Also, performance over time may degrade due to the grease flowing from the interface or a relaxing of the closure force of the assembly. A more effective approach that also yields cost savings due to the need for mechanical fixings being alleviated is to use a thermally efficient structural adhesive tape. Materials such as Bondline 200 from Universal Science feature a strong, pressure sensitive adhesive on both sides of a thin foil of aluminium having a thermal conductivity of around 140 W/mK. These types of material exhibit cold flow characteristics when installed between two metals surfaces which means that micro air voids are effectively removed and heat transfer between insulated metal circuit & chassis or heatsink is maximised; this performance is enhanced due to a thin bondline of around 0.16mm which shortens the thermal path between the metallic surfaces. Typically available in rolls in a range of standard widths, or for high volume production, as custom die-cut shapes thermally adhesive tapes allow fast repeatable assembly & deliver long-term, high levels of stable thermal performance.
Taking a different approach for mid-power device-based applications
  Mid-Power LEDs – those rated at around 0.5W to 1.0W dissipate less power and so do not require the high thermal performance of an insulated metal circuit to dissipate heat from their junctions. This presents the designer with an opportunity to reduce the cost and weight of their design by using a standard thin (0.6 – 0.8mm) FR4 printed circuit board material. Mid-power LED lighting product designs can use a format of FR4 that is familiar to designers and PCB fabricators/ processors throughout the electronics industry: that is a thin sheet of glass-reinforced epoxy laminate clad top and bottom with a thin copper foil. During the chemical etching process the copper is selectively removed from the board to leave a finished circuit of copper lands where the individual devices – in this case LEDs and supporting passive components – are mounted, and joined by interconnecting tracks to complete the electrical circuit. On the underside of the board it may be necessary to have some tracks connected to the top side using plated via holes in order to achieve the desired circuit layout. However, to maximise the heat spreading effect, as much copper as possible should be left in place on the underside of the board. In order to promote the optimum heat flow from the individual LEDs mounted on the PCB, the copper pads on which the individual devices are mounted are drilled with multiple vias which are then plated in order to create a small matrix of ‘heat pipes’ to draw the heat from the die of the LED through the PCB to the larger copper area on the underside of the board. If required, the thermal efficiency of this top-to-bottom connection can be enhanced further by ‘flooding’ the via holes with solder.
Photo Credit: Bondline 200 thermal adhesive tape star
Photo Credit: Bondline 700 thermal adhesive tape 
  While mid-power LEDs generate a lot less heat than high-power devices, it is still necessary to attach the FR4 circuit to a more substantial metal surface in order to continually remove the heat and maintain acceptable device operating temperatures. Using a thermal adhesive tape can provide the best approach for achieving this and also removes the need for screws and other mechanical fixings. For the reasons previously described, thermal grease does not provide a consistent, reliable long-term solution and also necessitates the need for screws and other fixings. Modern thermal adhesive tapes such as Universal Science’s Bondline 700 can provide the required combination of adhesion and thermal performance and are specifically designed for bonding FR4 circuits to chassis or other heatspreaders/heatsinks.
  In addition to providing a permanent bond that increases to a maximum level over the first 24 hours after installation, Bondline 700 also flows to fill the micro air voids that can form between the adhesive layer and the underside of the FR4 circuit due to the small ‘steps’ that exist between the copper areas and the FR4 where the copper has been selectively etched away. Other undulations and surface irregularities that need to be filled may also exist on the chassis / heat spreader side of the sandwich assembly. With FR4 the right selection of the interface material and mechanical attachment is much more critical than with MCPCB as can be seen in Fig. 2. Where even higher levels of thermal performance are required Bondline 1800 provides conductivity of 1.8W/mK.
Fig. 2: Image of FR4 using thermal camera
Summary
  With both high power and mid-power LEDs providing options for designers a wide range of commercial and industrial lighting solutions, it is important that the correct decisions are made with regard to thermal management. As well as giving the appropriate degree of heat spreading and dissipation, the right choices can also save money, reduce weight and minimize design complexity.

James Stratford, CEO of Universal Science Group, got trained as an electro-mechanical design engineer with Her Majesty’s Government Communications. 

Thursday, 15 June 2017

EOT Characteristics of LEDs

This makes measuring LED output an important activity but their small size, directionality and narrow spectral emissions that traditional photometric and colorimetric measurement methods are not always the most appropriate. Measurements of major characteristics of a LED that manufacturers and end-users find most important are like total spectral flux, luminous flux, chromaticity, CCT (Correlated Colour Temperature), CRI (Colour Rendering Index), peak wavelength, dominant wavelength, current (I), voltage (V) and luminous efficacy in the field of lighting.
Techniques for measuring EOT
  The lamp measurement system will depend on three things that is measurement needed, physical size of the lamp, and magnitude of the lamp output. Lamp physical dimensions must also be considered. The sphere diameter must accommodate the maximum length of the lamp under test. The CIE recommends that for tubular fluorescent lamps, the sphere diameter should be at least twice the longest dimension of the lamp, and that for compact lamps the sphere diameter should be at least ten times the largest dimension of the lamp. There are different methods and considerations for lamp measurement radiometry and photometry. In general, the measurement of a lamp's luminous flux and its spectral radiant flux are identical except in how the radiation is detected. There are two methods for measuring the total flux of a lamp. The first employs a goniometer that scans the whole sphere of radiation from the source. Integration over this scan yields a very accurate measurement of the lamp output, but the process is not always easy or quick. The second method employs an integrating sphere to capture all the radiation from the source at the same time associated colour parameters are measured. A detector is used to measure the output of the sphere. By comparing the output with a test lamp in the sphere to the output of a known standard in the sphere, the total flux emitted from the lamp can be calculated. This procedure can be done easily and quickly.
Integrating Sphere
  The integrating sphere integrates the total light output of a tested source to produce a single measurement. It is a hollow sphere coated with Barium Sulphate of fundamental optical technology, commonly used in photometry, radiometry, for applications including the test and measurement of light sources (lamps, LEDs), sensors (spectrometers, camera systems), and material properties (reflectance and transmittance). It has become standard instrument in photometry and radiometry. Luminous flux is the total photometric power emitted in all directions from a light source, measured in lumens. With the proliferation of high brightness, LEDs are used for illumination, and measuring the total flux of LEDs has become widespread. The variation in the spatial and angular response of many detector leads to incorrect results when the flux to be measured is non-uniform or the beam shifts slightly. Non-uniformity can originate from the source or optical path. Beam movement can come from moving components or refractive index variation in the air path. Measurements with fiber optics can be influenced by launch or fiber bend conditions. The recorded output pattern changes as the fiber output pattern changes, and light moves across or off a detector.
  The integrating sphere light measurement systems are in sphere sizes ranging from 0.5 meters. The system consist of a spectrometers, source meter, thermal element, DC power supply, Auxiliary lamp (AUX), and calibration lamp (CAL). The integrating sphere must include a lamp socket fixture, baffles, and a viewing port for the photodetector. In high accuracy measurements, an auxiliary lamp may be used to correct for lamp self absorption.
Fig. 1: Integrating Sphere (Courtesy: www.labsphere.com)
  A baffle is placed between the lamp and the detector port. Because the total luminous flux of the lamp is directly proportional to the illumination of the sphere wall, the detector must be baffled from direct illumination by the lamp. The CIE recommendation for the placement of this baffle is that it be 1/4 to 1/6 the sphere diameter from the photometer head. The baffle is coated with the same material as the integrating sphere wall. The most common mounting arrangement places the lamp at the centre of the integrating sphere and the baffle at approximately one-third of the radius from the viewing port.
Measuring procedure
  At first initialization of integration sphere is very important one because to get accurate measurement. For initialization purpose the CAL lamp can be calibrated for accurate total luminous flux and total spectral flux, proceeding with AUX lamp calibration for self-absorption correction. Each lamp screened, seasoned and calibrated then only integrating sphere ready for measurement.
  The LED is placed in the centre of the sphere and connected to the power port. After fixing the LED updating the auxiliary calibration compensates for errors due to size, shape and colour of device under test. Using temperature controller sets the temperature of the LED mounting block. LED power supply panel control the current or voltage to the LED. Then LED total flux is calculated from signals with a standard (known) flux source. But, anything placed in the sphere affects its throughput.
  Fig. 1 shows Integrating spheres measurement setup that collect the total flux emitted from a lamp. Because an integrating sphere reflects and integrates all the light entering the sphere, the light received by a small area of the sphere is directly proportional to the total flux from a light source mounted within the sphere. The total flux from a test lamp is determined by comparing it to a calibrated working standard. Then the instrument produced the output screen, it consists of total spectral flux, total luminous flux graph and parameters, CRI block and CIE chromaticity diagram and parameters.
  TOC analysis is the thermal transient response of the LEDs resulting from a step function excitation carried a large amount of information about the thermal behaviour of the device. The six types of tests that can be performed are:
  • ILV @ constant T: Step and control I, wait for T to stabilize, measure L and V 
  • VLI @ constant T: Step and control V, wait for T to stabilize, measure L and I 
  • TLV @ constant I: Step and control T, wait for T to stabilize, measure L and V 
  • TLI @ constant V: Step and control T, wait for T to stabilize, measure L and I 
  • ILV/T: Perform an ILV @ constant T, then step T and repeat at each T 
  • VLI/T: Perform a VLI @ constant T, then step T and repeat at each T
  Where, L=Lumens, V=Voltage, I=Current,T =Temperature.
Results and Discussion
  The EOT characteristics of LEDs are measured using laboratory Integrating sphere as shown in Fig. 2. The total spectral flux,total luminous flux, CIE-XYZ chromaticity, color rendering index and TOC analysis (case temperature control vs. electrical and optical parameters) of LEDs are measured.
Fig. 2: Laboratory Integrating Sphere
  Spectroradiometry is the techniques of how the energy distributed over visible range & the total flux is the measurement types of light emitted in all directions. Spectroradiometry + Total Flux = Total Spectral Flux (Watts/nm).
  Photometry involves the physical measurement of visible light energy and attempts to compensate for the psychophysical attributes of the human response and physical units of power. Photometry is just like radiometry except that everything is weighted by the spectral response (luminosity function) of the human eye as defined by the CIE. luminosity function as the transfer function of filters which approximate the sensitivity of the human eye. Photometry + Total Flux = Total Luminous Flux (lumens/ nm).
(a) Total spectral flux output
(b) Total luminous flux output 
Fig. 3: Single Green LED with Heatsink
  Spectroradiometry, photometry measurement result of single green LED with heatsink and single yellow LED with heatsink are shown in Fig. 3 and Fig. 4 respectively. Fig. 3(a) and 4(a), total spectral flux waveform x-axis represents visible wavelength range from 360 to 1000 (nm), y-axis represents spectral power (W), this graph indicates spectral parameters in terms of lumens, Kelvin’s, dominant wave, peak wavelength, purity. CIE-XYZ chromaticity diagram x-axis represents the normalized value of X-coordinates, y-axis represents the normalized value of Y-coordinates, this diagram indicates the value of CIEx, CIEy, DUV prime, u prime, v prime and also measure full width half max (FWHM) and center wavelength, centroid wavelength and temperature. Fig. 3(b) and 4(b), Total luminous flux waveform x-axis represents visible wavelength range from 360 to 1000(nm), y-axis represents total flux (lumen/nm). Luminous Flux is the flow of light from a source per unit of time and is measured in Lumens. Then other waveform & measurement parameters are same as that of total spectral flux waveform.
(a) Total spectral flux output
(b) Total luminous flux output
Fig. 4: Single Yellow LED with Heatsink
  From the Figures 3 & 4, it is proved that LED light sources are much more efficient at converting watts to lumens. Different materials can be used within the LED sources themselves, each of which has its own light extraction efficacy. For these and other reasons, Single green LED with heatsink and Single yellow LED with heatsink can consume the same power but differ widely in lumen output. Because power (Watts) can’t be used as an index of light output, evaluating the “brightness” of LED sources.
  In data analysis of redstrip, whitestrip, single LED without heatsink, single green LED with heatsink, single yellow LED with heatsink are display screen in Fig. 5. This LEDs spectral waveform of the total spectral flux (units Watts/nm) over the visible wavelength range from 360 to 1000 (nm).
Fig. 5: Analysis the characteristics of LEDs
  Fig. 5 reveals that lumens of the red strip is high compared than others, because array of led but the value of CCT is very low so it come under warm colour. White strip has lumens is low than LED arrays because is just a module of LED but the value of CCT is very high compared than others so it come under the cool colour. Single LED with heat sink is used to get the same lumens of module LEDs and it has its own power supply ratings. Single LED without heat sink its lumens level is very low. With different colour LED measurement results the lumen is same, the CCT and CRI value of each LED can be varied, so come under different category of quality.
  Table 1 gives EOT parameters of LED, i.e., CIE chromaticity parameters, CCT in Kelvin’s, lumens, wavelength, purity, integration time, temperature, CRI, voltage, current, power. Measures are characterized according to the measurement value, it is useful for customers to identifying the quality of LED and also this measurement is useful for production level industries to improve quality of LED.
  • Chromaticity is an objective specification of the quality of a colour regardless of its luminance. The CIE Chromaticity Diagram showing all visible colours. x and y are the normalized amounts of the X and Y primaries present, and hence z = 1 - x - y gives the amount of the Z primary required. These values used for colour matching purpose. 
  • The Correlated Colour Temperature (CCT) is a specification of the colour appearance of a light source, relating the colour to a reference source heated to a particular temperature, measured by the thermal unit Kelvin. The measurement can also be described as the warm, cool and neutral of a light source. Generally, sources below 3500K are considered "warm;" while those above 4100K are considered "cool" sources, in between that values are considered neutral. 
  • Lumen is a unit of light flow or luminous flux. The lumen rating of a lamp is a measure of the total light output of the lamp. As lamps and fixtures age and become dirty, their lumen output decreases. 
  • Dominant wavelength is the spectral colour which can be mixed with white light in order to reproduce the desired colour. 
  • Centre wavelength is the average of two wavelengths determined in the 3 dB width of measurements.
  • FWHM describes the spectral width of the half power points of the LED. Half Power points are where power spectral density is one half of the peak amplitude. FWHM in the LEDs are classed as high power level or low power level. 
  • Purity is defined as the ratio of the distance from the equal energy point E to the color coordinate. All intensity fairly close to the dominant wavelength (spectral colour is 100%). 
  • The integration time is the main parameter for the user to adapt the modulation of a spectrometric system to the signal level of the measurement. In many applications it is necessary to have a linear relation between the integration time and the signal, especially in radiometric measurements.
Conclusion
  The electrical, optical and thermal characteristics of various LEDs like18 Volts red LED strip(redstrip), 12 Volts white LED strip (whitestrip), 3 Volts single LED without heatsink, 3 Volts single green LED with heat sink, 3 Volts single yellow LED with heat sink. From the experimental results it is observed that two different LED sources can consume the same power but differ widely in lumen output. The illumination level of LED with heat sink holds good when compared to other LEDs. From the obtained results best lamp can be chosen for particular lighting purpose and also this measurement is useful for production level industries to improve the quality of LED.

Dr P S Manoharan is working as Associate professor in Thiagarajar College of Engineering, Madurai. N Uma, is pursuing Post-graduate (Control and Instrumentation) in Thiagarajar College of Engineering, Madurai.

Achievening High Quality Light Dimming with Texas Instruments LED Drivers

The growing popularity of LED light sources is rooted in energy savings, long life, and new fixture options that enable it to be used in almost any application. A 25W LED lamp can replace the light output of a 100W incandescent lamp, deliver a useful lifetime averaging 50,000 hours (compared to 10,000-20,000 hours for fluorescent lamps and 3000 hours for halogen lamps) these advantages ensure a strong future for LEDs but there are challenges associated with using LEDs to meet market expectations. Compatibility between LED lamps, drivers, and legacy controls can be confusing, if they are specified improperly performance will suffer. Issues with compatibility are probably the greatest source of frustration among lighting designers and their customers. Mock-up installations are expensive and time-consuming to smooth the pathway of a project, customers are look for lighting manufacturers who have already done the appropriate testing and research and can ensure successful LED lamp, driver, & control installation.
  Many consumers have turned to dimmers or automated dimming controls over standard light switches because dimmed lighting can reduce energy use and offer ambiance. When using an energy-efficient LED lights, the customer generally expects an experience similar to what incandescent lamps provide but correct combination of controls, drivers and LED sources are necessary. Although some LED lamps are marked as compatible with incandescent dimmers, there are various degrees of what can be defined as “compatible.” Dimmable LED lamps tend to interact quite differently when used with these legacy devices. A number of undesirable results may occur when you use a dimmable LED lamp with an incandescent dimmer, including reduced dimming range, flickering or fluttering of the lamp, inconsistent performance based on the number and assortment of lamps being controlled by one incandescent dimmer.
  Dimming LEDs, similar to the process with incandescent sources, saves energy at a roughly 1:1 ratio. This means that if you dim LEDs down to 50% of their light output, you save nearly 50% of the associated energy use. While it is true that LEDs are already very efficient compared to almost any other light source, you save even more energy by dimming them. Dimming LEDs also makes them run cooler, extending the life of the electronic components in the driver, as well as the phosphor in the LEDs. This will potentially double or triple the useful life of the LED lamp or module. The problem, however, is that nearly all dimmers in the market were designed for standard incandescent lamps. Unfortunately, the market expectation of dimming performance isn't being achieved by LED products over incandescent lamps with existing triac or phase-cut dimmers.
  Here we are analyzing the compatibility issues and the solutions currently industry is offering. The solutions must meet the current technological needs while being mindful of both past and future technological challenges.
Facts and Challenges
  Dimming methodologies can impact flicker. In the output of an LED driver the percentage of ripple at twice the line frequency is the parameter that corresponds to the flicker in the light output. Many LED drivers produce dimming by switching the LED light on and off at a relatively high frequency with the process called pulse-width modulation (PWM) dimming or digital dimming. The human eye is completely oblivious to these high frequencies and simply perceives less light. Dimmable LED drivers exist that simply modulate the light on and off at twice the line frequency at low dim levels, the result can be a lot like the light output of old magnetic ballasts where the flicker may be easily perceived. In addition, if used with a triac dimmer, which doesn't dim positive and negative half-cycles equally, it may introduce a line frequency component to the PWM that will be perceptible to anyone. Other LED drivers produce a uniform DC current level, which is then adjusted downward to produce dimming. This methodology is sometimes referred to as analog dimming. For task and office lighting, this approach is the most trouble-free kind of dimming to use, though it's likely to be more expensive than digital dimming.
  There are applications where LED lamps will operate with an incandescent dimmer, in general, an incandescent dimmer will provide inconsistent performance with SSL. The incandescent lamp by nature represents a simple resistive load with a linear response to the dimmer set point. Standard incandescent dimmers work particularly well with this type of load by switching on at an adjustable phase angle after the start of each alternating current half-cycle, thereby altering the voltage waveform applied to lamps. By switching instead of absorbing part of the voltage supplied, minimal power is wasted, and dimming can occur almost instantaneously. In contrast, LED lamp loads can vary greatly across different manufacturers and designs. But most can be characterized by a diode-capacitor power supply feeding a constant current source. The diodes rectify the applied AC voltage, allowing it to charge the storage capacitor, while the LED elements draw a constant current from the power supply that is related to the desired dimming level and brightness.
  LED light loads are significantly different from incandescent lamps in which the applied voltage and the current flowing into the load are related. In incandescent lamps, the applied voltage across the load and the resulting current flowing through the load are related linearly by Ohms Law (V = IR). In this case, the resistance sets the scale, and the current waveform follows the voltage waveform, differing only by scale. In LED loads, the applied voltage and resulting current flow are not related by a simple linear relationship. In the diode-capacitor power supply model of the LED lamp, current flows from the applied voltage to the load only when the magnitude of the applied voltage exceeds the stored voltage on the power supply capacitor. The stored voltage on the power supply capacitor, in turn, depends on the current drawn by the LED elements themselves, which is a function of the LED brightness. Therefore, the current flowing from the supply to the lamp depends both on the instantaneous value of the input AC voltage waveform and the brightness of the LED lamp. Changing the intensity or dimming level of the LED lamp affects where in the AC line cycle the load begins to draw current. This inflection point also affects the amount of current that surges into the lamp.
Wiring, Dimmers and Regulations – Contributes to Light Flickering
  Wiring:-The inconsistency issues among dimmers and lamps are that most of the existing residential wiring infrastructure was built without a neutral wire at the switch box. The absence of the neutral wire is referred to as two-wire lighting control while the inclusion of a neutral at the switch box is referred to as three-wire lighting control. The need of supporting two different wiring scenarios poses certain challenges that lighting-control designers need to account for in planning to control a broader range of lamp types with a single dimmer.
  Some dimmers are designed to work with one type or the other only, while some are designed to work in both types of installations. But, for all dimmers, even those that are designed for both two-and three-wire installations, there are significant differences in performance between these two installations in terms of how the dimmer circuitry is powered and how the dimmer synchronizes with the line voltage. When used to drive incandescent lamp loads, these differences are mostly negligible. But, when used to drive LED loads, they present significant challenges to stable dimming and lighting control. Regardless of the circuit type, all phase-controlled dimmers need to synchronize with the AC line in order to work correctly. Without the ability to sense the AC line and its zero-crossings, a phase-controlled dimmer would not detect the correct timing for switching the AC voltage, and it would lose its ability to control and dim the lamp load. The end result is flickering and fluttering of the light output.
  In three-wire installations the line, load and neutral wires are connected to the wall control electrical box. The line wire comes from the AC power source and supplies power for both the dimmer and the load. The load wire is connected to the lamp load and provides a return path for the power delivered to the load. The third wire, the neutral connection, provides the essential return path for the dimmer even when the load is disconnected or is in a state that doesn’t draw any current. The neutral is an important feature of three-wire installations. It ensures that the dimmer device has a direct connection to the AC power source regardless of the state of the load. This third wire not only ensures that the dimmer has power to drive its own internal circuitry even when the load is disconnected or off, it also provides a clean signal of the incoming AC power source for detection of zero-crossings and synchronization with the line. Both of these are essential for stable phase-controlled dimming, and are easier to obtain in three-wire designs.
  In two-wire installations only two wires are present in the electrical box, the line wire and the load wire. In this case, the dimmer is simply placed in series between the line and the load. With only two wires, the dimmer must rely on the current passing through the load to both power its own internal circuitry and to detect zero-crossings for synchronization with the AC line. When LED lamps perform poorly with a dimmer, often times the blame is placed on the dimmer circuit. But the source of the problem really lies in how the LED load current differs from the incandescent lamp in two-wire applications. If the load current is regular, as in the case for incandescent lamps, then stable line synchronization and ample power for the dimmer’s internal circuitry are both easy to obtain. With LED lamps, however, the load current is much smaller and much less regular, and line synchronization becomes difficult. Similarly, the load current of LED lamps in their off state can be so small, that even obtaining a few milliamps to supply the internal dimmer circuitry can be challenging. Without adequate supply and stable line synchronization, lamp flickering may result.
Triac Dimmers
  In AC phase control widely used form of brightness control is the familiar triac-based dimmer that is present in many residential applications. Triac dimmers operate by cutting out a portion of the AC waveform.
Fig. 1: Typical TRIAC dimmer circuit
  During the start of AC cycle TRIAC will be off and during the operation cycle (refer Fig 1) C1 charges through R1 and light bulb, when voltage on C1 exceeds DIAC threshold voltage the TRIAC starts conducting. R1 is a variable resistor which controls when TRIAC turns ON, dimming function and defines the conduction angle.
  The most common type cuts out a portion of the leading edge of the AC waveform, as shown in Fig 2. The dimmer senses each zero-crossing of the AC input, and waits for a variable delay period before turning on the triac switch and delivering the AC to the load. The AC input to the light therefore has a bite out of the leading edge of each half sine wave. This forward phase dimming typically operates on two wires & avoids the labor associated with adding a third wire.
Fig. 2: Forward-phase dimming cuts the front edge of each half-cycle of the AC line input
  A second similar type of dimmer operates in the reverse manner, by cutting a portion of the trailing edge of each half sine wave, as shown in Fig 3. This type of dimming is sometimes called reverse phase control, and is designed for use in electronic low voltage (ELV) applications. Reverse phase dimming is considerably more expensive but minimizes electromagnetic interference (EMI) issues.
Fig. 3: Reverse-phase dimming cuts the trailing edge of each half-cycle
  Phase-control dimmers were originally developed for incandescent lighting, where the lamp brightness is directly dependent on the average power in the AC input. By cutting out a portion of the waveform, the power is reduced and the lamp becomes dimmer. However, this is not the case with LED lighting, because LED luminaires contain a power supply and driver whose primary function is to supply constant current to the LEDs regardless of the AC input voltage. If you connect a constant-current or constant-voltage power supply to the output of a phase-control dimmer, the power supply will attempt to compensate for the missing portions of the AC waveform. As the amount of phase-cut increases, the power supply will maintain its output voltage by drawing higher input current, and the LEDs will remain at normal brightness. Eventually, when the dimmer setting is very low, the power-supply feedback circuits will no longer be able to compensate and the power supply output will collapse.
  Performance of phase dimming circuits depends on certain TRIAC parameters, which are critical and should match the spec requirements.
  • To turn on the Triac, a gate signal is required and must exceed specified IGT and VGT requirements. 
  • Latching current (IL) is required to maintain the Triac in the on state immediately after the switching from off state to on state has occurred and the triggering signal has been removed. 
  • Then, Holding current (IH) is the required to maintain (hold) the Thyristor in the on state.
  For an LED luminaire to respond correctly to a phase-control dimmer, it is necessary to add several functional blocks into the driver electronics. A sensor block monitors the AC input waveform and generates an output signal proportional to the amount of phase cut and feeds to PWM controller and then drives the MOSFET. There is also the issue of how dimming information is conveyed to drivers. The driver selected must have the ability to work with the dimmers deployed in an application, especially in retrofit scenarios.
Regulations
  Compatibility issues between lamps and dimming devices are certainly due lack of dimming performance standards within the lighting industry and how each uniquely corresponds with LED drivers. This lack of standardization can be seen not only in varying characteristics between manufacturers, but also by product within some manufacturers’ product lines. Complications arise from the fact that any given lamp can require a set of electrical and electronic characteristics – current, voltage and control signals – that are vastly different from any other lamp. While one lamp may be able to be dimmed by a particular dimming device, others cannot.
  Under current UL standards, notably UL 14725, intended to regulate the safety of dimmers, an LED lamp driver is categorized as an “electronic ballast.” One notable issue addressed by UL 1472 is in-rush current which is generated at the startup of many LED lamp loads. High in-rush current can result in failure of switch contacts, which is a safety hazard in many field applications – such as dimmers – where the switch is serving as the disconnect means. To evaluate the safety of the combination of dimmer and electronic ballasts, UL has taken the systems approach by requiring dimmer manufacturers to provide information on the intended electronic load (i.e., CFL, LED or electronic ballast) for each dimmer. UL listing investigation will involve the use of the specified electronic ballasts or a synthetic load exhibiting the same in-rush and steady-state characteristics in the overload, endurance and temperature tests.
Texas Instruments LED Driver solutions to overcome current dimming challenges
  As energy efficient lamps continue to penetrate the lighting market, the availability of new lighting controls that meet the specific needs of these lamps is increasing. Consumers can take fuller advantage of all of the benefits of the newer, more energy efficient lamps by using TI’s industry-leading TRIAC dimmable offline LED driver solution which is perfect for any application where an LED driver must interface to a standard TRIAC wall dimmer. TI’s new TRIAC dimmable LED driver delivers a wide, uniform dimming range free of flicker, best-in-class dimming performance, & high efficiency all while maintaining ENERGY STAR® power factor requirements in a typical application.
Full Range Dimming Capability
  TRIAC dimmable LED driver offers 100:1 full range dimming capability, going from full light to nearly imperceptible light in a continuous range without being extinguished and maintains a constant current to large strings of LEDs driven in series off of a standard line voltage.
Easy to Use
  TRIAC dimmable LED driver enables a direct replacement of incandescent or halogen lamp systems that are currently interfaced to a TRIAC dimmer without having to change the original infrastructure or sacrifice performance. In addition the new TRIAC dimmable LED driver is available in WEBENCH® LED Designer to allow for easy and quick design in.
Uniform Dimming Without Flicker
  TRIAC dimmable LED driver allows master-slave operation control in multi-chipsolutions which enables a single TRIAC dimmer to control multiple strings of LEDs with-smooth consistent dimming, free of flicker.
  The LM3447 is TI’s newest product for offline, phase-dimmable LED lighting. It is a future genera-ti on LED lighting product as it offers significant features,
  • Designed for constant power operation with better line regulation,
  • Overall lower isolated solution cost.
  • Power Factor Correction.
  • Valley switching to improve efficiency and EMI performance
  • Improved dimmer hold circuitry
  • Addition of thermal fold-back to protect LED arrays.
  The LM3447 is a versatile power factor correction controller IC designed to meet the performance requirements of a residential and commercial (phase-cut) dimmer compatible LED lamps. The device incorporates a phase decoder circuit and adjustable hold current circuit to provide smooth and flicker free dimming operation. A proprietary primary side control technique based on input voltage feed forward is used to regulate the input power drawn by the LED driver and achieve line regulation over a wide range of input voltage. Valley switching operation is implemented to minimize switching loss and to reduce EMI. It contains an internal thermal foldback feature designed to protect the LEDs from damage based on the temperature sensed by a single external NTC resistor. Additional features include load open and short circuit protection, cycle-by-cycle FET over-current protection, burst mode fault operation and internal thermal shutdown. The LM3447 is ideal for implementing dimmable, isolated single-state LED lamp drivers where simplicity, low component count, and small solution size are of primary importance. The device is available in 14-pin TSSOP package.
  The LM3447 is TI's next generation controller for off-line, isolated, phase-dimmable retrofit LED lighting applications. Using primary side power regulation, the LM3447 can implement compact, isolated, constant power flyback designs which give lighting designers up to 10% better utilization of a given string of LEDs. Protection of LED arrays is enhanced using the integrated thermal fold-back capability which protects against high temperature conditions by reducing power, and therefore light output, to the LEDs until the high temperature condition is alleviated. Most competing solutions either lack this capability, or completely turn off under high temperature conditions. The LM3447 is well suited for retrofit bulb designs as A19, E26/27, PAR30/38 as well as TRIAC dimmable down lighting.
  Considering current market challenges and industries technological expectations LM3447 is robust with additional intelligent blocksfor flickerfree Triac dimming.
Angle Detection Block
  The LM3447 uses the input voltage, VREC, to detect the conduction phase angle. Fig 4 shows the LM3447 angle detect circuit, where the input voltage, VREC, is scaled by the current mirror circuits and re-generated across an internal 42kΩ resistor. This replica of the input vol-tage is compared with internal 280mV reference to obtain the conduction information. The resulting PWM signal, with its on-time proportional to the conduction period, is buffered and supplied through the FLT1 pin, as shown in Fig4. To match the external phase dimmer characteristics with the LM3447 decoding circuit and prevent EMI filter capacitors from interfering with dimming operation, it is necessary to select an angle detection threshold, VADET(TH). This threshold can then be programmed using the resistor, RAC, such that- 
Fig. 4: Phase Angle Detection and Hold Current Circuit
  For best results, set VADET(TH) as follows:
  • 25V to 40V for 120V systems
  • 50V to 80V for 230V systems
  Resistor RAC should also limit the VAC current under worst case operating conditions. The value of RAC should be optimized to meet both angle detect, VADET, & VAC current, IVAC constraints.
Hold Current Block
  The LM3447 incorporates an efficient hold current circuit to enhance compatibility with TRIAC based leading edge dimmers. Holding current from an external dimmer is drawn before the Flyback PFC circuit through the pass transistor, QPASS and limited by resistors RHLD1 and RHLD2, as shown in Fig 4. It should be noted that the additional current drawn has no effect on the rectified input voltage and therefore does not interfere with the input power regulation control scheme. To provide high efficiency, the hold circuit is enabled only when the presence of an external dimmer is detected based on the FLT2 input. The ENHOLD signal is asserted and hold operation is permitted when VFLT2 falls below 1V. The hold operation is halted when VFLT2 rises above 1.2V. During dimming, the hold current is drawn during the interval when rectified input voltage is below the VHOLD(TH), based on the external resistor RAC. The FET turn on is controlled by an internal comparator with a reference of 400mV (higher than angle detect reference), such that hold current is always asserted before angle detect threshold VADET(TH). The hold circuit operation is summarized in Fig 5. The hold trun-on threshold, VHOLD(TH) is given by
  The hold current is based on the BIAS voltage and set by the sum of resistors RHLD1 and RHLD2,
Fig. 5: Angle Detection circuit and Holding Current Circuit Operation
  In selecting the hold current level, it is critical to consider its impact on the average power dissipation and the operating junction temperature of pass transistor, QPASS under worst case operating conditions. The current should be limited to a safe value based on the pass transistor specifications or the ABS MAX rating of LM3447 (70mA). For best performance, it is recommended to set the hold current magnitude between 5mA and 20mA. A capacitor, CHLD of 2.2μF to 10μF, from RHLD2 to GND is connected to limit the rate of change of input current (diin/dt) caused by the step insertion of holding current. This prevents TRIAC based dimmers from misfiring at low dimming level.
Angle Decoding Block
  The LM3447 incorporates a linear decoding circuit that translates the sensed conduction angle into an internal dimming command, VDIM. The conduction angle information, represented by the PWM signal at FLT1 output, is processed by an external low pass filter consisting of resistor, RFLT and capacitor, CFLT, which attenuates the twice line frequency component from the signal. The resulting analog signal at FLT2 is converted into the dimming command by a linear analog processing circuit. The piecewise linear relationship between the FLT2 input and the dimming command is shown graphically in Fig 6.
Fig. 6: Relationship Between VFLT2 and VDIM
  The dimming command, VDIM is-
  • Held constant at 1V for VFLT2 ranging from 1.75V to 1.45V (conduction angle 180O to 150O)
  • Linearly varied with gain of 0.877 for VFLT2 ranging from 1.45V to 280mV (conduction angle 150O to 30O)
  • Saturated at 13mV for VFLT2 lower than 280mV (conduction angle less than 30O).

Shinu Mathew is Analog Application Engineer, Texas Instruments (India).