LED life... how did they come up with 100,000 ?

[sdl13]

There is a lot of confusion about the "B" mortality number.

Philips Lumileds defined mortality as performing below the designated lumen maintenance. So if L is defined as 70% (L70) then a failure would occur when performance lumen maintenance is below L70. There is no distinction between a catastrophic failure, say from infant mortality, or if the failure is simply a due to lumen degradation. Either way, performance would be below the L designation. In some cases an application may require L80 or L50, so there are many B/L combinations that could be calculated.

There is a white paper that explains all this and gives the statistical methodology and it can be found at http://www.lumileds.com/pdfs/WP12.pdf

 

[uk_caver]

so there is a simple solution, Socketing. Every other light in the world is socketed. mabey its about time , before they SOLDER them into home lighting fixtures , to have a copper heat sink socketing system.

patent anyone

By 'they' I guess you mean the people who hope to sell you a new lighting fixture in N years when the old one starts having enough failures to be worth bothering about?

At the low end of the market, that doesn't seem likely.

Even at the high end, you'd need a standard that could cope with all kinds of LED package changes over a timespan long enough to make it worth standardising. You'd also need a standard which could accommodate all kinds of drive currents/voltages, or possibly lots of different standards, or else have the drive electronics integrated, with some way of communicating control information, etc through the fixture. You don't really want a standard that allows someone to stick one of today's 1Amp LEDs into a fixture set up to drive 2009's 5Amp ones.

It's pretty easy with lightbulbs - the interface is two-wire constant-voltage, either 110-120 or 220-240, and that's highly unlikely to change every few years.