4

What is or are, the most likely and common ways - the weak links in the chain (including I assume poor electricity wiring / configured housing) - that cause a traditional incandescent light bulb to blow?

And how much does manufacturing / materials quality of the bulb make a difference in any of those factors? (In other words, should an expensively, state-of-the-art constructed bulb in terms of materials purity and construction robustness, generally last a lot longer than the average bulb on the market?)

I understand it is a fairly simple lighting technology, so what complications cause them to not last longer than they generally do? Is 'blowing' the only way that an incandescent can ever expire?

3 Answers3

7

I'd say that an incandescent bulb is quite forgiving about power quality. They generally blow up because tungsten, that is, the material the filament is made of, slowly evaporates until the filament breaks. So what really comes into play is the filament manufacturing, how it's held in place, if its thickness is regular or not, if the tungsten is pure enough, and so on.

A filament breaks because it eventually becomes a little thinner at a point. That point heats more, causing more tungsten to evaporate, making it even thinner... That's good old nasty positive reaction. Thermal shock can do the job too, that's why you tell your children to stop disco-switching the lights.

Interestingly enough since the filament usually just breaks in one point, it's possible to try to give a broken lamp some afterlife time. If you have the broken bulb plugged in\$^1\$ and shake it, it's possible for the two filament ends to touch: the point heats a lot and the filament might weld itself and give you ten or so more hours of life.

\$^{(1)}\$ Please, just don't try that if you are not sure it's safe.

Vladimir Cravero
  • 16,007
  • 2
  • 38
  • 71
5

Most light bulbs have a number of service hours designed in. This is achieved fairly accurately and on purpose during manufacturing. Consumer grade light bulbs burn out faster and the awful truth is that in that way the manufacturer ensures it can keep producing the light bulbs and make money. The main cause is the tungsten filament slowly evaporating until it gets too thin to carry the current. The trick during manufacturing is to etch the filament to a carefully designed thickness so the lifetime is reduced programmed.

The origins of programmed life light bulbs began with the Phoebus Cartel in 1924.

There are also light bulbs with a special stronger filament (basically they skip the 'etch-cycle'), these are for use in high reliable applications and last longer. They are commonly used in places that are hard to reach and are more expensive. Not sure about the proper name of these light bulbs. Because of the extended life time these bulbs are more expensive as a manufacturer has only limited room for producing spares. (What need is there for spares if the bulb doesn't fail?)

Where Australia and EU banned the consumer grade bulbs, the strengthened filament is not banned because of its specialized application. The latter type is just not available in your regular shop around the corner.

NealeU
  • 103
  • 2
jippie
  • 33,033
  • 16
  • 93
  • 160
  • This is more the sort of answer that speaks to the reason why I brought up this question in the first place - being such a simple technology, I wondered why they don't last so long if it is that simple a product. Thank you for your information, it helps to enlighten very much. –  Jul 03 '14 at 05:42
  • 1
    I can also add, that a relative just told me that his mother had an incandescent that was used as the outside house light (so even exposed to the elements quite a bit!), that lasted *for 15 years*, and I assume it was used quite a lot if it was an outside light too! And furthermore, I'll have to Google now, but I am reminded now of reading once of a famous incandescent bulb that apparently lasted 100 years (of, IIRC, CONTINUOUS use), and now it sounds like it's actually a credible story based on what we know. 'They don't make em like they used to', has never applied more. Disgusting –  Jul 03 '14 at 06:06
  • 1
    @jippie: While I'm sure you answer is credible, it would be nice to know where you got this information from. – nijoakim Jul 03 '14 at 08:50
  • 1
    http://en.wikipedia.org/wiki/Centennial_Light – Vladimir Cravero Jul 03 '14 at 09:09
  • 2
    Isn't the original Edison bulb still burning? – Scott Seidman Jul 03 '14 at 11:06
  • Any idea why the market effect doesn't cause manufacturers to etch the fillament less and less in order to beat their competitors until the fillament is not etched at all? – DBedrenko Jul 03 '14 at 11:27
  • 1
    It's easy to believe there is some conspiracy by the light bulb industry to sell us all crappy bulbs, but do you have any references? – Phil Frost Jul 03 '14 at 11:50
  • 11
    I'm downvoting even though the basic explanation (filament evaporation) is correct. The "programmed lifetime" is just silly. For a given voltage the thinner the filament the hotter and more efficient it gets, and the shorter its lifetime (faster evaporation). The extended-lifetime bulbs you describe (and they do have thicker filaments) do exist. The reason they last is that they don't get as hot as regular bulbs. This also means they produce less light for a given power. Some of the old bulbs are the same - they last but (because) they don't put out as much light. – WhatRoughBeast Jul 03 '14 at 12:29
  • 1
    To expand on my previous comment. Emergency exit signs used to be illuminated by incandescent bulbs (they may still be, but I'd expect LEDs). Such illumination obviously requires very high lifetimes for the source. The standard solution was not to provide an expensive long-lifetime bulb. It was simply to put a diode in series with the bulb. This cuts its power by nearly half, and the cooler-running filament lasts a LOT longer. It's dim, of course, and the color has changed, but who cares? – WhatRoughBeast Jul 03 '14 at 13:07
  • @WhatRoughBeast: exactly, this is how most long-life bulbs essentially operate: reduce temperature, and thereby efficacy, which increases the lifetime extremely nonlinearly (i.e. 2x less power, 10x more lifetime or something like that). – user36129 Jul 03 '14 at 15:15
  • @WhatRoughBeast thank you for your very illuminating information and insight. But I think the basic science, and facts, behind the ability and history of planned obsolescence in light bulbs is there (now that I've discovered it). There is no loss being had, (only financial gain), for light bulb companies by making their products deliberately expire sooner because energy costs can rise and fall completely independent of bulb costs and technology. Some people may have their own, (cheaper, or even completely renewable) power supply and prefer a bulb that costs them less due to it lasting longer. –  Jul 03 '14 at 22:45
  • 1
    Sure there's a loss to be had. If a competitor makes a longer-lasting bulb with the same brightness at the same price, one would expect the company with the shorter lifetime bulbs to lose sales. And referring to a cartel which was exposed 70 years ago does not seem particularly relevant, unless you are going to knowingly assert that such a cartel exists today - in which case a little proof is in order. As you say, the basic science behind bulb life is there. So why are you ignoring it in favor of a conspiracy theory? – WhatRoughBeast Jul 03 '14 at 23:09
  • 1
    The fact that my grandmother's bulb from the 60's I assume didn't cost much more (in fact probably less, and this is inflation-allowed) than the MUCH shorter lasting incandescents of today (and I'm pretty sure it was 60W - and I don't think incandescent technology has changed much between 1960 and now in terms of lumens per watt) (and this is also despite, unless tungsten reserves have depleted, CHEAPER manufacturing costs now, compared to THEN), speaks volumes to planned obsolescence applying to incandescent bulbs today. This topic is not just about electrical science, but business realities. –  Jul 03 '14 at 23:37
5

Most of the above data is incorrect. Tungsten was tested as a lamp filament about 1900 and found to be useless due to sag, a coiled filament was impossible. In about 1915, GE found by adding some other materials, they could produce NON SAG Tungsten which is still in use.

However, non sag tungsten is subject to DC etching such that the development of surface roughness changes the emissivity such that after 1000 hours operation, the lumens have decreased 30 percent. The filament is cooler and lamp life increases. The DC etching continues but much slower. There is no cure, the only solution in the case of automobile headlamps is to replace the bulbs. Most people do not realize their headlamps are below legal limits.

Another incandescent lamp problem that is not what people think, is halogen lamps. Halogen does nothing for an incandescent lamp except allow the use of a smaller bulb by keeping the bulb free of condensed tungsten, which in turn allows a higher fill gas pressure. Double the fill pressure, lamp live doubles. Result, higher lumens per watt with the same lamp life.

David R. Dayton
  • 51
  • 1
  • 1
  • 2
    Not quite. Halogen lamps re-deposit the evaporated tungsten back on the filament, thus making it last a lot longer. The smaller bulb is the result of a higher temperature requirement of the reaction - they don't want to make it smaller, they have to make it smaller. – paul Nov 02 '14 at 23:49
  • 1
    True, the Halogen cycle does deposit the tungsten back on to the filament. However, the tungsten is deposited at the coolest part of the filament and does nothing to extent the life of the filament. Dave Dayton – David R. Dayton Nov 09 '14 at 23:46
  • An interesting question is, why does a straight wire filament increase in temperature when it is coiled at a fixed current? True, the tighter the coil pitch, the higher the temperature but that is only part of the answer. – David R. Dayton Nov 10 '14 at 00:43