I started dating and tracking my CFLs because they seemed to fail quite often. I would typically get 30 - 90 days life. Some have lasted longer, but a short life is typical in my home.
Unfortunately, must of my fixtures are built in to the house and are upside down enclosed or both. Some would say that I'm not using the CFLs properly. I'd counter that they aren't designed properly.
What good is a device that works near perfect in the lab, but prematurely fails such a large percentage of the time in the field? If it's not designed to be used in its end application setting, it's simply not a product.
@ divide_by_zero: "As with lots of other consumer goods, price is the only thing most people consider when shopping. "
Very true. Which gives governments a good reason to legislate for quality of production in these items. Which will cost to enforce, pushing the price up more. Maybe enforcing guaranteed lifetimes might help, but that's difficult to prove.....
Unlike Bert, I've had plenty of CFLs fail in less than, or not much more than, the average life of an incandescent. Those I have taken apart (mainly cheap chinese ones) often exhibit signs of cheap electrolytics (swollen and/or popped case).
And Roger Watkins comments in the article mentioned the mercury issue with CFL disposal....at least LEDs are a bit better in that regard.
I forgot to mention one possible issue with CFLs, which may also apply to LEDs. It seems that you're better off with the full size Edison base, than the miniature base. My take is, the full size base dissipates heat better.
One lamp we have had a habit of eating incandescent bulbs on a rgular basis. I tried a CFL, and it too failed prematurely. So I changed the lamp socket to a full size Edison, and so far so good.
I've been testing a CFL in an enclosed, upside down configuration, in a bathroom fixture. Worst case scenario, I would think, although it's a full size Edison base. It's been okay so far, about maybe 6 months.
(continuation because I am too long winded)
IRT divide_by_zero: We are in agreement in all of what you have said. One additional point is that newer CFL and LED units that are available at consumer outlets are almost all part of your "race to the bottom".
If it makes you all feel a little better, I started using CFL devices pretty extensively in my non-enclosed applications (free-standing lamps, bathroom exposed bulb fixtures, ceiling-look-down lamps, etc.) back in the mid '90s, and have been using LED devices in similar applications for a few years now. Even without payback, it is my opinion that we who are technologists should help lead the charge to more efficient use of our resources. I keep trying refrigerator and enclosed fixture applications occasionally, and yard light applications fairly regularly, and have yet to be satisfied there is a good replacement for the incandescent, I use halogen if available, bulb in these types of applications.
In response to these well considered comments, it seems some clarifications and responses may be appropriate.
IRT sharps_eng "Good quality power supplies use 105C electrolytics running at maybe 65-70C": You are very correct. The problem is what happens when "cost reductions" are applied or when a CFL/LED device is placed in an enclosed incandescent fixture. The CFL/LED device is way under the typical 60W rating of the fixture, but fails in typically only a week or two. A big piece of this failure where I have post-mortemed seems to be aluminum electrolytics, and yes sometimes magnetics.
IRT Bert22306: I am sorry for sounding negative, but both as a consumer and a designer of electronics it can be very frustrating to be "sold a bill of goods" where a much higher priced device is supposed to pay for itself over life, if only its life was as advertised. My experience has been pretty good for exposed bulb lighting (though the dimming-with-age issue becomes very obvious when replacing devices that have failed), but direct replacement of incandescent lamps in incandescent fixtures is a BIG problem for those of us in older homes that could be easily addressed by proper choice of components and design for efficient drive circuitry especially in LED devices. We both agree on reducing electrical load. I have for years been looking at off-grid power, and lighting loads are a noticeable part of the budget...
Cheap electrolytics will fail sooner than good ones, but especially at elevated temperatures. Most of the big cap manufacturers have a few lines of high-ripple long-life 'lytics which are typically used in linear fluorescent ballasts. The office I'm sitting in now is lit by high-efficiency electronic ballasts made in 1995. Very few have failed since installation. Getting long life out of aluminum caps is simply a matter of selecting the correct cap and applying the necessary level of derating.
The same applies to magnetics used in lighting. Good transformer and inductor design entails using core material appropriate for the switching frequency used and careful attention to issues with skin effect in windings. Insulation must be selected based on worst-case temperatures in the rated range of the lamp. I have performed a few autopsies on CFLs in my home which died after only 3000 hours or so and found evidence of overheated windings in inductors.
Heat sinking for switching devices in most cheap CFLs I have taken apart is pretty minimal. This is a case where size matters. Tiny packages are nice if you can get the heat out, but this is not typically a priority in these products. I haven't bought any cheap LED lamps yet, but the more expensive ones I've bought have pretty generous heat sinking for both the ballast and the LEDs.
As with lots of other consumer goods, price is the only thing most people consider when shopping. This market in yet another race to the bottom, with junk designed by amateurs leading the way. Efficient lighting is in bad need of some standards.
Interesting comments, even though they initially sounded negative.
Of course, cheap components will fail. But I have only had one CFL fail prematurely, i.e. about as soon as an incandescent. Every single other one has lasted considerably longer. I mean, years and years. But true, I have not used CFLs in the oven or in the refrigerator.
I do not understand the comment about heat; "... the actual life of the device is generally much shorter due to higher temperature of operation of the device and due to the non-LED components in the assembly."
"Higher temperature of operation" compared to what? Surely, not to incandescents! I'll agree that if the electronic components overheat, especially when these are cheap, they will fail prematurely. But, you know, we now have computers operating in side the engine compartment of cars, even on hot summer days.
Lighting loads in average homes are a hefty percentage of the overall electric load, especially in homes with gas heat and gas stove. It makes a lot of sense to get rid of the waste caused by ancient design white-hot filaments. If there is one silver lining behind the ban of incandescents, it is that this will encourage serious development efforts in alternatives, LEDs in particular. I see CFLs as a transitional solution at best.
Necessity is the mother of invention.
Designers have always been able to make cheap and nasty products using electrolytics; toys, power supplies and other gadgets. LED lamps are just another example. Good quality power supplies use 105C electrolytics running at maybe 65-70C, giving many years lifetime. Only if there are mass returns, customer feedback or bad publicity will manufacturers put more quality into their designs.
Like with water quality, ordinary consumers cannot judge this kind of quality issue, so the only solution has to be a market standard protecting consumer rights, with statuatory enforcement. This could be checked at the point of import, just as with toys.
Replay available now: A handful of emerging network technologies are competing to be the preferred wide-area connection for the Internet of Things. All claim lower costs and power use than cellular but none have wide deployment yet. Listen in as proponents of leading contenders make their case to be the metro or national IoT network of the future. Rick Merritt, EE Times Silicon Valley Bureau Chief, moderators this discussion. Join in and ask his guests questions.