There have been cases where electronic capacitors in the small MR16/GU5.3 LED bulbs have failed after prolonged use where there was an increase in heat due to the use of electronic transformers and poor convection cooling.
My reading shows that the life time of various capacitors at 45°C ranges from 32,000 hours to greater than 15 years for long life ones, e.g., NAZV series. However, at 85°C the general purpose capacitor, e.g., NACE series is rated at 2,000 hours and the long life 40,000 hours, and others in the range of 6,000 to 32,000 hours, less than the life of the LED. We have measured heat sink surface temperatures of up to 72°C using compatible electronic transformers so capacitor life is probably a significant issue with heat being the real killer for caps.
I have seen no studies as yet comparing the performance of MR16 with constant voltage driver vs. GU10 with integrated driver, which would be interesting. My theory is the GU10 LED form factor will have a shorter life span than the MR16/GU5.3 due to greater componentry and additional heat, supporting the use of independent constant voltage drivers.
Does anyone know the temperature of the circuitry in a GU10 or A19/E27 LED bulb? don@ www.liteonled.com
"So far they seem to have one significant drawback; [LEDs] just don't have illuminating capability of the bulbs they purport to replace, whatever the figures say!"
As someone that has lived with almost total residential LED lighting every day for the last year I just haven't found that to be the case. (The only tungsten lamps in the entire house are in the stove vent hood.) You do have to pick your lamps and dimmers carefully and avoid the really cheap top-firing product. My under counter LED lamps run fully-dimmed almost continuously. I don't think I could stand running them at full brightness without shades. Most of the time my PAR and MR-16s are dimmed.
I've found that for a 12x12 room, a cluster of four of the GE62180 430 lumens in an overhead is about the equivalent of two conventional 75W incandescents. The former is 36W PF ~0.9 vs 150W PF 1.0. The LEDs look far cleaner and produce a fraction of the waste heat with attendant air conditioning load.
Ever tried to read laser-marked SMT ICs? Much easier with LEDs and I don't know why. Just better contrast.
For whatever reason LED lamps attract less bugs.
Go to Home Depot or Lowes and buy the cheapest LED you can find. You will be completely disappointed. There are (were?) quite a lot of losers and a few gems.
I've had good luck with GE A-series, Sylvania Osram PAR and MR-16, Philips A-series and Utilitech A-series. I prefer the 3000K but the 2700K Philips ambient light is gorgeous giving that Thomas Kinkade look.
Math has nothing to do with it. The electrical code requires a certain size wire. If you have a kitchen circuit with a refrigerator, toaster and coffee maker, you need a 20 amp circuit with #12 wire. If you have 24 watts of LEDs in your kitchen (that's pretty bright) and run them off a 12 V bus, you can use #18 wire with less insulation.
I treat claims of bulb lifetimes as highly suspect. I have used Quartz-Halide lamps that 'promise' 2000 hours that are failing after a few hundred, certainly no longer than the tungstan incandescents that they claim to replace!
LEDs seem to have one major benefit over CFLs - at least they switch-on at full brightness. So far they seem to have one significant drawback; they just don't have illuminating capability of the bulbs they purport to replace, whatever the figues say!
I believe the issue with running DC is in the math. Joule's Law and Ohm's Law state P = I^2 x R = V x I. For transmission, higher voltage means lower current means less loss in the wire while allows for a smaller gauge wire. I believe this is why there was a (albeit failed) push a while ago to increase car power rails from 12V to 42V. Running a low voltage DC bus in a house may lead to heavier gauge, more expensive wiring. Everyone, feel free to correct me if I am wrong.
LG's $10 dollar pricing is highly competitive. The LED used will be a 10 watts capacity which is made to work at 7.5 watts to get the optimum light output and life.The cost of 10 watt white LED of good make will be around $5 and the rest part electronics, enclosure, profit, manufacturing cost all put together is $5.That is why i felt it is priced rightly. Soon this market also will be captured by LG and may be Samsung also.Other companies watch out.
I'm not seeing anything particularly revolutionary in this "me-too" snow cone design. The GE 62180 A-19 "40W" 430 lumen (9W actually) is a far superior optical design and produces more even dispersion than the "top firing" model shown here. I have reviews of several LED retrofits here: http://www.proaudiodesignforum.com/forum/php/viewtopic.php?f=11&t=408
NASA's Orion Flight Software Production Systems Manager Darrel G. Raines joins Planet Analog Editor Steve Taranovich and Embedded.com Editor Max Maxfield to talk about embedded flight software used on the Mars on EE Times Radio. Live radio show and live chat. Get your questions ready.