Fullscreen HTML5Med Res (??MB)HTML4

LED replacements for fluorescent lamps: are they any good? Half the power – lots of light – more than twice 36W fluorescent tubes are the main source of light in supermarkets & retail stores, offices and many factories. There are hundreds of millions of them in use around Australia and many billions in use around the world. But while they are the most efficient light source in offices and the home, they are now being challenged by LED tubes which can directly replace them. By LEO SIMPSON B ACK IN THE MAY 2010 issue we had a feature article on how to slash office lighting bills by using quad phosphor fluorescent tubes made by NEC. In our own office, we substituted one quad phosphor tube for the two existing tubes in each twin-lamp luminaire. The result was a much brighter office and a cut in electricity consumption due to lighting of about 50%. Since then we have heard from readers who are delighted with the results of using quad phosphor fluorescent tubes. In one industrial warehouse complex, they replaced all the tubes in what was a dingy and dangerous un- derground section and the increase in lighting was a revelation. In that case they did not save power but the improvement in illumination, especially at night, made the complex much safer and less prone to graffiti and vandalism. As a bonus, the improved lighting also made their CCTV security system much more effective. At the end of the above article, we made a brief reference to the existence of LED replacements for fluorescent tubes but discounted them at the time because they were very expensive and the units that we knew of apparently did not comply with Australian This close-up shot of the end of a LED replacement tube shows the different LEDs used in the Cool White model. It is apparently powered as four sets of 63 LEDs. 16  Silicon Chip standards. We concluded with the remark that “In the next few years that is bound to change.” Well, the future has a habit of arriving quickly these days. In the very next issue (June 2010), local company Tenrod Australia introduced a range of ecoLED replacement tubes. Naturally, we had to obtain some samples and put them to some comparative tests. The first point to make about these LED replacements is that they are exactly that. They can be fitted in place of conventional fluorescent tubes since they have the same length and the same 2-pin connector at each end. There are a number of provisos though. First, while they can be used in standard 36W fluorescent fittings which have a conventional ballast, the starter MUST be removed. If the starter is left in place, it is immediate death to the LED replacement tube. Ideally, the fitting should be rewired so that the ballast, starter and power factor correction capacitor (if fitted) are all removed, to provide maximum efficiency. In effect, the full mains supply is applied across the LED replacement tube. siliconchip.com.au the life The tubular extruded aluminum housing provides rigidity and acts as a heatsink. It barely gets warm. If the existing iron ballast is left in place, there is a small loss in efficiency but since the extra power consumption amounts to less than one watt, we think that most users will leave the ballast in place. It is too much trouble to remove it. LED replacement tubes cannot be fitted in fluorescent fittings with elec- tronic ballasts, unless of course, the fitting is rewired to remove the ballast. How many LEDs? The ecoLED T8 replacement is housed in a tubular aluminium extrusion with a clear polycarbonate lens section. It has three rows of 84 LEDs each, making a total of 252. The LEDs are surface-mount types, on a long narrow PC board. The driving electronics is in the tube housing. Switch on a LED replacement tube and there is a short delay of about two seconds and then it lights up at full brilliance. There is no flick-flickflickering and no gradual build-up in light output. There is no strobing which can be evident with conventional fluorescent tubes and nor is there any buzz, hum or whistle. Radio interference is also very low. And they are really bright. Looked at directly from less than a metre away, they are almost painful to behold. The aluminium tubular housing barely gets warm and that is what you would expect since the power consumption for the tube is around 20W or less – less than half the consumption of a fluorescent tube running with a conventional ballast. Looked at from further away, their apparent surface brightness is also higher than any conventional fluorescent tube. However, that is not the whole story. While, the LED replacement type radiates mainly in one direction, set by the individual LED lenses, a conventional fluorescent tube radiates evenly in all directions around its main axis and so puts out more light. This may or may not be useful, depending on whether the luminaire has a reflector behind the tube and maybe a prismatic diffuser in front of it. For example, we did some comparisons between a Daylight White LED replacement and an NEC quad phosphor 36W fluorescent tube. The JOIN the teChNOLOgy age NOW WIth PICaXe Developed as a teaching tool, the PICAXE is a low-cost “brain” for almost any project. Easy to use and understand, professionals & hobbyists can be productive within minutes. Free software development system and low-cost in-circuit programming. Variety of hardware, project boards and kits to suit your application. Digital, analog, RS232, 1-Wire™, SPI and I2C.PC connectivity. Applications include: Distributed in Australia by 1[Datalogging 1[Robotics 1[Measurement & instruments 1[Motor & lighting control 1[Farming & agriculture 1[Internet server 1[Wireless links 1[Colour sensing 1[Fun games Microzed Computers Pty Ltd Phone 1300 735 420 Fax 1300 735 421 www.microzed.com.au www.siliconchip.com.au siliconchip.com.au NEW X2 HIPS now in sC tock! ovember 2010  17 2009  67 SNeptember These LED replacement tubes are exactly the same overall length and use the same 2-pin connector as on standard T8 fluorescent tubes. This is the Daylight model in which the surface-mount LEDs are all the same type and colour. light fitting was a single 36W batten with a conventional iron ballast and no reflector, mounted on the ceiling of a small study measuring about 2.5 x 3.5 metres. Comparative light outputs were measured by a Digitech multimeter with a Lux scale. It was sited on a desk 2.2 metres below the light fitting. Brightness When first turned on, the LED replacement produced a reading of about 138 Lux which was certainly adequate for general tasks such as reading or using a computer. However, since most of the light is projected straight down, it was noticeable that the ceiling itself was quite dark, as you might expect from the tube’s 120° viewing angle. The overall room illumination improved when the prismatic diffuser was fitted, as this bounced some light onto the ceiling. It did not make any difference to the Lux reading though. The second test was with the NEC quad phosphor tube and this produced a reading of 215 Lux. Not only was this 55% brighter but the illumination was more even, with considerable bounce of light from the sides of the tube off the ceiling. Fitting the prismatic diffuser made a slight difference, increasing the reading to 220 Lux. On the face of it, the NEC quad phosphor tube wins the light output race by a mile but there is a lot more to the story. On the other hand, if we 18  Silicon Chip had used a single phosphor tube the comparison would have been in favour of the LED replacement. However, we did not see much point in that. Since the LED replacement has a premium price it should be compared with a premium fluorescent tube – the NEC quad phosphor. As an aside, the LED replacement comes on with full brilliance and then dims very slightly as it warms up over a period of (say) 30 minutes. The dimming is so slight that you only know about it if you are measuring Lux – otherwise it is not apparent. This is in line with normal LED behaviour – the warmer they are, the less light they emit but it is not a large effect. By contrast, the quad phosphor tube started off with a relatively low output (it was a cold July evening with a room temperature of about 12°C) and came up to full brilliance in a period of about 30 minutes. As far as power consumption is concerned, the LED replacement wins hands down, with less than half the power use of a conventional fluorescent fitting. By the way, when installed in a fitting with a prismatic diffuser, these tubes appear little different to ordinary fluorescents, at least to the casual observer. How it works The Tenrod LED replacements are powered by an internal switchmode supply. It can accept a mains input supply from 80-265V AC. It has an EMI (electromagnetic interference) filter at the input followed by a bridge rectifier and capacitor filter. The supply is based on a PFM (pulse frequency modulation) chip and a Mosfet which regulate the current though the LEDs to within 5%. Hence the LED brightness is essentially constant for that full range of mains input voltage. In fact, we measured the brightness over the range 70-250V and there was virtually no variation at all. One thing’s for sure, with these tubes “brownouts” will be a thing of the past. The LEDs in the Tenrod tubes appear to be run in groups of 63 (4 x 63 = 252), possibly further split into three paralleled strings of 21 LEDs. This would give an operating voltage in the region of 70V and is in line with the minimum input supply of 80VAC. Mind you, we are hypothesising here because we were not able (unwilling, actually) to break into the package to find out the details. Power consumption was virtually constant over a fairly wide range of mains input voltages. For one of the sample tubes we tested, in a fitting without a ballast, it was 19.6W at 210250VAC. Power consumption drops very slightly after the unit has been running for 30 minutes or so. The Tenrod LED tubes come in three colours: Daylight White (5600 - 6300K), Cool White (3800 - 4200K) and Warm White (3200 - 3500K). Their brightness ratings were 400, 375 and 360 Lux, respectively (measured at a distance of one metre). Total light output, measured in Lumens, was 1500, 1400 and 1350, respectively. For the Daylight tube, that is equiv­ alent to 75 lumens/watt which compares very favourably with the NEC quad phosphor tubes at 100 lumens/ watt. Interestingly, on a Lux/watt basis, the ecoLED Daylight tube easily beats the NEC quad phosphor, especially if it is fitted in a ballasted luminaire. siliconchip.com.au ecoLED Tube For our money, we would go for the Daylight model for most applications. The different colours are obtained by mixing the LEDs in the rows but for the Daylight White the LEDs all appear to be the same – white. OK, so we have talked about light output and power consumption. A number of points remain to be discussed. The first is expected life. Typical fluorescent tubes have a stated life of 16,000 hours and can be expected to last four or five years in a typical office installation. This is far in excess of compact fluorescent lamps which often don’t last any longer than the much-maligned incandescent lamps. The above-mentioned NEC quad phosphor tubes have a 20,000 hour life but these ecoLED replacements have a stated life of 50,000 hours! Such a long life makes them ideal for installations where they are on permanently, such as in parking stations and railway stations. They could be expected to last almost six years. In practice, no-one really knows how long LEDs last so they could have an even longer life. As an aside, the fact that they work well at low temperatures means that they can be used where conventional fluorescent tubes don’t function efficiently and have difficulty starting, such as in freezer and cool-rooms and in glass-fronted refrigerators (as in supermarkets and convenience stores). Saving Energy & the Environment No mercury Friendlier alternative to fluorescent lamps Install in its place How much are they? The final point is the price. As you might expect, these ecoLED tubes are significantly more expensive than even the NEC quad phosphor tubes we cited as a benchmark. But with less than half the power consumption and very long life, many building owners will be looking very closely at installing these tubes. Pricing for the Daylight T8 tube is $80 + GST on a one-off basis, dropping to $60 + GST for 100-off quantities. With electricity prices rising inexorably, they are certainly attractive, especially in installations where they run 24 hours a day. At 24c/kilowatt hour (the Energy Australia commercial rate) it costs $100.92 to run one 36W ballasted fluorescent fitting (drawing around 48W total) for a year. By using an ecoLED tube drawing 20W instead, that drops to $42.00. You don’t have to be a chartered accountant to work out that an installation of these tubes into a parking station would have a very short payback period. Oh and we should mention that these LED replacement tubes have no mercury and no lead content. Our impression is that these LED replacement tubes will make very quick inroads into the Australian market and around the world. For further information, contact Tenrod Australia Pty Ltd, 1/24 Vore St, Silverwater, NSW 2128. Phone (02) 9748 0655. Their website is at www.tenrod.com.au SC siliconchip.com.au Half your energy bills 18W for 4ft (120cm) 9W for 2ft (60cm) No flickering, no noise - No irritation No glass to break (robustly built) No starters to change 50,000 hours of lifetime Daylight White, Cool White, Warm White CRI > 75; 76.7 Lm per W (off the wall) 500gm net IES Data available Website: www.tenrod.com.au E-mail: sales<at>tenrod.com.au Sydney: Melbourne: Brisbane: Auckland: Tel. 02 9748 0655 Tel. 03 9886 7800 Tel. 07 3879 2133 Tel 09 298 4346 Fax. 02 9748 0258 Fax. 03 9886 7799 Fax. 07 3879 2188 Fax. 09 353 1317 September 2010  19