how are lightbulb lifetimes calculated?

[isaacmacdonald]

I'm wondering because I've had to replace a couple in my apt, and they seem to be dying a lot faster than the spec claims. Does usage frequency (off/on/off/on) have a signifigant impact? Is this too arcane for atot?

 

[Viper GTS]

Damn, maybe Jerboy does have a use.

I found this:

http://www.newsbulletin.msu.edu/oct25/lights.html

Myth #2: Turning fluorescent lamps on and off wears them out right away.

Reality: Electric lights have a published rating for expected life. This rating is in the hundreds of hours for many incandescent lights and in the thousands of hours for most fluorescents. Fluorescent lights have a life rating based on how many hours they are left on each time they are turned on. This is referred to as "burn time." Most full-size fluorescent lights are rated to last 20,000 hours when left on for three hours every time they are turned on. This means that the light has about 6,667 starts available to use up. If you "burn" your fluorescent lights less than three hours per start, you use up your potential starts faster. However, longer burns use more energy and add to your energy costs. The most efficient light is the one that is turned off when it is not needed.



Viper GTS

 

[gopunk]

dammit i should know this... we talked about it in stat class... i think it's something like they assume the failure rate is some distribution, run tests for some short (relatively) amount of time, and fit the data to the model. i dunno that's just my guess. look up hazards, weibull.... i remember that components usually have a surge in failure early on, but components remaining will gradually dwindle afterwards. that is, there are some products that are just defective.

 

[Orsorum]

Originally posted by: gopunk
dammit i should know this... we talked about it in stat class... i think it's something like they assume the failure rate is some distribution, run tests for some short (relatively) amount of time, and fit the data to the model. i dunno that's just my guess. look up hazards, weibull.... i remember that components usually have a surge in failure early on, but components remaining will gradually dwindle afterwards. that is, there are some products that are just defective.

I believe their lifetimes are estimated using a Poisson distribution.

 

[sandorski]

With the way light bulbs burn out in this house, I suspect they just slap a number on the buld to sell them.

 

[her209]

By the number of times you turn it on and off.

 

[MWink]

Are you talking about incandescent or fluorescent lights? If you're talking about incandescent lights burning out quick, it might be that your voltage is high. We have chronically high voltage (but not high enough to cause problems for most things) and we found out that that causes incandescent bulbs to burn out very fast. About a year ago we switched to compact fluorescent bulbs and only one has died (it died within 2 weeks so it was probably defective).

 

[Ryan]

This is why I buy those halogen/florescent light bulbs. I replaced all of the lights in my household with them a year and a half ago, and none have burned out yet

 

[Eli]

The lightbulbs in my room seem to burn out very fast, I haven't had an incandescent last for more than a month..

There's a CF bulb in there now.

Oh, and the number of on/off cycles definately has an impact. You'll notice that a bulb usually burns out right when you turn it on.

The sudden heating of the tungsten is quite stressful, you can imagine.