What is your experience with LED light bulbs?

Could be either, both need to work for it to light.

But yes usually it's the ballast that fails. It is possible that one LED chip fails taking out the whole unit too.

They're not user serviceable anyway so it's a unit failure.

Here is a new issue with LED lamps. But first a little background info. At some point, three-way incandescent lamps were introduced. They had an additional filament typically rated at 1/2 the wattage of the main filament, e.g. 50W and 100W. Using a lampholder (socket) designed for them, the switch could select OFF, 0N, 100W, and 150W (both filaments ON, hence the name three-way). A ring contact was added between the screw shell and the center contact to accommodate the additional filament. Likewise, the lampholder was modified by adding a matching tab-shaped contact between the center contact and shell. This all worked fine, even when using legacy single-filament lamps, because the latter were designed to have insulation between the shell and center contact such that the third contact of a three-way lampholder would just run into that insulation with no consequences. And now for the LED situation.

My next door neighbor bought many LED lamps at the 99¢-only Store. The ones purchased of course had a metal screw shell. Unfortunately, the shell extended downwards about 1/8" further than the shell of typical incandescent lamps, such that it touches the third contact of a three-way holder. The following consequences ensue:



Notice that two of the possible combinations of switch setting and lamp type result in dead shorts if the lamp is screwed in tight. This trips the branch circuit breaker or blows the fuse. If the holder center contact is bent upwards a bit, it is possible to screw in the lamp loosely and get away with it, but that is a tenuous situation.

My neighbor (who should know better) tossed the LED lamp after the short pitted the screw shell. It works fine in my 1-way holders. Silly goose.

I have contacted Underwriters Laboratories and they have an online form for submitting problems to their surveillance people. I'll fill it out after I get the E number off the lamp. Either UL missed this problem (either as a design or manufacturing matter), or unapproved changes were made to the product. Will this result in product recalls?

As an additional matter, I notice that this lamp is marked "NOT FOR USE WITH A DIMMER," yet the package was labelled "Dimmable." I wonder if the UL has any rules about mismarking the packaging? Or does the mfr. even submit the packaging to UL? Those sneaky Chinese!

I don't understand all the technical stuff, but I have some three-way lamps with one-way incandescent bulbs in them. There are two clicks off and two clicks on. I haven't noticed any difference in brightness between the on clicks. Is it dangerous to do this? I don't use those lamps too much now, as I mostly use the ones with compact fluorescents in them. Yeah, all old technology, but I usually use things until they are used up, and these aren't.

I have replaced most of the bulbs with energy efficient bulbs.  Began with cfl, and when prices came down, with led's.


Vanities in the bathroom makes a big difference, since wife didn't like the warm up phase for the cfl's there.  I replaced with LED's in her bathroom and she is now a fan.. 


 

With some embarrassment, I may have to suck back some of the stuff I said about using the LED in a 3-way lampholder. Looking at an incandescent lamp, I don't see much difference from the LED that gave the short with regards as to how long the screw shell is. It could be that the lampholder's low-watt contact—the one built in between the shell and the center contact—is bent out of position and that caused the short. That only causes a problem when the lamp is screwed in tight enough to hit the bent contact. So now we have to throw into the mix how tightly the lamp is screwed in. Imagine that, everybody and his brother maybe didn't screw up, and it's all due to a bent piece of metal. Nahhhh. It couldn't be that simple, could it? I'll let you know if my neighbor hasn't thrown out the fixture and I can still examine it.

With regards to my everyday use of LEDs, they are all holding up well at my place. With the small draw of those darlings, it may be that no future technology need be pursued to replace them, that is to say, they are good enough from now on. The only exception I can see is that something newer may be cheaper to build, although that is hard to imagine.

Once you say that something lasts for 50 years, are you really saying anything at all? There might be some accountant type with OCD that daily logs all his lamp usage in order to find out if the claim is true, but a soon as you get beyond the time where you can actually remember installing a particular lamp, it is all a blur and they might has well have claimed a hundred years for all the difference it makes.

MARTHA, MARTHA! I DISTINCTLY REMEMBER INSTALLING THIS LAMP ON CHRISTMAS MORNING OF '04, AND NOW IT JUST DIED AFTER ONLY 86 YEARS! WE'VE BEEN CHEATED!

I found some "F32T8" LED tubes cheap and tried one in a fixture with a real F32T8 tube in the other slot... didn't work! Flashed and then went dark. Uhoh! Put the real F32T8 back in...works fine. Hmm!

Then I had this other known electronic ballast, unfortunately it's "277VAC" for stealing a phase off of 3-phase industrial. Luckily I had a 240V step up, which seems to power it just fine. In any case this electronic ballast lights up this LED F32T8 just fine...

Unfortunately I have a surplus of regular F40T12's to use up...