>> leds are more luminous and they consume little, but they emanate less >> heat of the incandescent light bulbs. >> And they do not melt the snow... >> The article is in italian but the photo is enough! >> http://www.zeusnews.it/index.php3?ar=stampa&cod=11596&numero=999 > LEDs do of course emit plenty of heat (though obviously less than > incan), it's just that it tends to be extracted out of the back to > keep the LEDs themselves cool rather than emitted out the front. To a first order approximation ,yes. But it's interestingly more complex than that, and changing by the day as LED efficincies rise. At present about 1/3 of the energy input to a very top efficincy white LED exits as light. An incandescent bulb has an efficiency of about 15 - 20 l/W (lumen/Watt) A good quality coiled CFL about 70 l/W. A tube fluorescent about 100-120 l/W A typical good quality White LED in the 1 W - 5W range about 100-120 l/W and somewhat more or less epending on what luminance "bin" you buy or are given. So, to produce say 1500 - 2000 lumen you need. 100 W incandescent 25 W CFL 20 W tube fluro 20 W LED The 2000 lumen is equivalent to roughly 5 Watt of light energy So the heat output of the various emitters will be about. 95 W incandescent 20 W CFL 15 W tube fluro 15 W LED However, the incandescent radiates almost all the energy - both the 5W of light and the 95 W of heat energy (actually also light but not as we see it, Jim).So too the CFL and the tube fluorescent. In their cases the glass gets hot and acts as a radiator. I haven't worked out how much energy goes where and why, but a look in a modern CFL bulb shows that the switching transistors (often MOSFETS) have shunk from TO220 to something smaller, and that heat sinks aren't used. Electronic efficicny looks quite good. Only the LED relies mainly on conduction to remove its heat. In a decent calorimeter the radiated "light" and "heat" would all sum So heating wise the ratios are about 100 : .25 : 20 : 5 While the LED consumes about 20% of the energy of the incandescent it radiates only 5% as much (which is what you said) - and what heating you get is ALL from "light" ! :-). Interestingly, in the traffic light situation, a CFL would take about 20% more energy than a top LED, but produce about 5 times as much radiated heating output if all the output was absorbed (something you don't want in a traffic light). If youcan abosrb much of the radiated heat in the light structure you'd probably get 2 to 4 x as much radiated heating from a CFL than with an LED. Presumably traffic light designers could arrange for the LEDs conducted or convected heat to be directed into the from sructi\ure to assist with snow melting. It's 'funny' the requirements that can crop up along the way in apparently straight forward applications. Melting snow is unlikely to feature high on an LED designers checklist in many cases. Russell McMahon YMMV / E&OE - sure to have made a rubbish assertion somewhere in there, but the general gist thereof should be OK. -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist