Extended life of solid state light emitters...

[Externet]

Greetings.

A LED emitter can have a long life.  Even longer if underdriven from specifications.

Asking for an educated guess...  A LED designed to operate on 1 Watt of power, if energized to 1/10W; would it last 10 years ? 100 years ?

Same question if it is a laser diode.

[John Cuthber]

Depending on circumstances, the limiting factor might be something like a capacitor in the power supply.
Good temperature control is also a big factor so dropping the power is almost certain to help.
 

[studiot]

11 hours ago, Externet said:

A LED emitter can have a long life.

Can have a long life yes, if properly manufactured and installed.

They said this about early transistors but then came the 'tin whisker' effect due to migration of metal from the leads into the semiconductor material and disabling the transistor.

https://nepp.nasa.gov/whisker/anecdote/af114-transistor/index.html

I said maunfactured and installed because this effect can also effect inappropriately chosen circuit board materials, solder, leads and other components as John has noted.

We do not yet know if another unforseen effect will rear up and bite us in the ass in years to come.

I have never seen a long life compact flourescent that achieved more than half the promoted service life bulb for instance.

[Externet]

Hi studiot.  That link took me to several others and saw/learned about impressive whiskers formation; for sure unexpected when equipment was built decades ago. Thank you !

[Enthalpy]

With properly built components, my educated guess (my first profession) is 10 years for a LED at normal operating point, more if you derate. At I/5 you still get light, possibly you get 10* more life. And yes, temperature matters a lot, more than the current.

For a laser diode it's much less: 1 years continuous operation is already difficult. Alas, you can't derate a laser diode, because it won't lase at all if you reduce the current, even /3. Try the current form factor if possible. Cooling better would bring a lot but it's already optimized.

[swansont]

33 minutes ago, Enthalpy said:

For a laser diode it's much less: 1 years continuous operation is already difficult.

The ones I use last 1-2 years, typically, and the failure is that it mode-hops, which is likely because if degradation of the AR coating. If you aren’t worried about the exact frequency of the light, the diode itself would last much longer.

33 minutes ago, Enthalpy said:

Alas, you can't derate a laser diode, because it won't lase at all if you reduce the current, even /3. Try the current form factor if possible. Cooling better would bring a lot but it's already optimized.

Well, no. It won’t lase below threshold, but there’s no reason to assume you’d normally operate it there. If threshold is e.g. 25 mA and you normally operate it at 100 mA, there is plenty of room to lower the current. 

[Enthalpy]

9 hours ago, swansont said:

The ones I use last 1-2 years, typically, and the failure is that it mode-hops, which is likely because if degradation of the AR coating. If you aren’t worried about the exact frequency of the light, the diode itself would last much longer.

Well, no. It won’t lase below threshold, but there’s no reason to assume you’d normally operate it there. If threshold is e.g. 25 mA and you normally operate it at 100 mA, there is plenty of room to lower the current. 

Everything degrades in a laser diode, including the metal conductors and their contacts with the semiconductor, the drift of varied species (including the metal) into the semiconductor, and so on. Expect the laser diode to fail completely soon after it begins to degrade.

Not every laser diode has a factor-of-four between threshold and normal operating point, but even then, if you half the current, you half the efficiency too, already through the non-lasing recombinations. Plus some other effects, like the natural emission not matching well the  resonator.

[swansont]

4 hours ago, Enthalpy said:

Everything degrades in a laser diode, including the metal conductors and their contacts with the semiconductor, the drift of varied species (including the metal) into the semiconductor, and so on. Expect the laser diode to fail completely soon after it begins to degrade.

http://www.worldstartech.com/what-determines-the-lifetime-of-a-laser-module/

Laser distributor disagrees

Manufacturers, too, that I’ve talked to in DARPA program reviews. They were working toward 100k hour lifetimes