Which lasts longer: a blinking light bulb or a steady one?

I was driving home from work today and noticed that one of the traffic lights I passed was in full yellow blinking mode. My first reaction was that doing that must wear the light bulbs out pretty quickly. But I can think of other reasons why they would last longer than ones that were on steady all the time.

Anyone have theories (or even gasp evidence) of which would last longer if you had 2 bulbs, one blinking, one steady?

all lights on 120 AC are blinking (just so fast you can’t see) and they do wear out faster (or so the maker of a product to convert the AC to DC to extend the bulb life would have you believe).

Other then that I would depend on how fast they were blinking (on time vs off time) and and starting surge.

if a bulb is on for a day, off for a day, etc. - that is technically a blinking light that would double the life compared to a constant on bulb

It depends on the type of lightbulb. Incandescent lights will definitely burn out faster if blinking, because there’s a bit of a power surge that causes the element to heat up more when you first turn it on. This is why a good 90+% of the time a light bulb burns out, it’s when you first turn the light on.

I’m not sure about florescents or whatnot, but i know LEDs last a lot longer and use less power when they’re blinking. Manufacturers frequently advertize this, and it’s one reason that battery powered bike lights flash.

I wouldn’t guess that the element heats up more when you first turn it on…it’s just that it’s going from cold to hot. When the light is sitting in it’s “on” state, it’s really flickering at 60Hz, but it’s probably so fast that the filament doesn’t cool down very much, so it’s effectively “on” full time.

I’d bet the fastest way to burn out a light would be to put it in a cycle of turning on just long enough to heat up all the way, then turning it off just long enough to cool off all the way (or most all the way). Or to shake it real hard. :slight_smile:

I’ve heard of light bulb life extender devices which sit inline with the bulb and act as a current buffer, so that when you first flick the switch, current flows slowly, and it gradually lets more current through until it’s up to full brightness. This way, the filament heats up slowly and (supposedly) lasts longer.

Just an interesting note to all you non-physics people(although i don’t know much about physics either). Have you ever been riding down the highway and noticed that the wheels of the car next to you seem to been spinning in the opposite directions they should be? Or at least the rims look like it. Well when i took physics I found out that it’s becauset hs street lights are strobing. Althought it looks like they are on constantly, they really flicker on and off with the lfow of current in electrical line- which switches every 1/60th of a second or something like that- although that statistic is probably wrong.

      • Yea, but in on the TV show, everybody climbed through the General Lee’s windows 'cause the doors didn’t open, but the poster showed that they did! And Hazzard County didn’t even have streetlights! What’s up with that?
  • Most of the damage to a filament is done in the rapid cooling cycle, the moment the power is turned off. -The reason it burns out when you turn it on is that it was damaged the last time you turned it off. - MC

Sure about the cooling cycle being more damaging than the heating one?

Why do I ask? Consider this: we all know the “fring!” kinda noise a dying lightbulb makes, yet I have often heard a fainter than normal version of this sound on the last SUCCESSFUL time I use the bulb. Sure enough, the bulb dies the following time I turn it on. Why then, do we never hear a noise when we turn the bulb off? And why, if cooling damages the bulb more than heating, does the filament only ever seem to break when the bulb is switched ON, not off?

I believe zyzzyva is mostly correct. Incandescent light filaments are most highly stressed when heating up or cooling down because of the thermal gradients; thus, a blinking bulb should fail more quickly than the same bulb that’s on all the time. I suspect that the AC/DC difference that k2dave alludes to is relatively minor, since the bulb doesn’t have much time to heat or cool in 1/120 of a second (filaments are pretty thin, though, so I might be wrong here).

I do recall reading about the oldest lightbulb in the world (in a publication that had more info than the linked snopes article). One supposition was that this particular lightbulb was so durable simply because it was in a room in a fire station and was never turned off and was protected from blackouts by the backup power supply of the fire station.

One thing I always thought was neat about incandescent lightbulbs was their expected failure rate, which does not depend on how long they’ve been in service. Example: You have a new bulb with an expected life of 1000 hours. You use it for 800 hours without failure. How much longer would you expect it to live? Answer: 1000 hours.

This is somewhat correct. At night, you’ll notice this effect at probably an earlier time, but you can also see this effect in the daytime, and I know the sun isn’t flickering at 60 Hz. It’s really a sampling problem, and the backwards motion you see is really just a result of aliasing. If the wheel is completing a half revolution in a time that’s less than the time your mind can process an image in, it will appear that the wheel is moving backwards., since it will see just a small movement to the rear, instead of the real movement, which is just shy of a full revolution. It’s most noticable in movies, where there’s a 24 fps sampling rate, but our eyes sample in a way, and at high wheel rpm, we get the same effect.


Was it named Byron?

  • Yes. Metal is damaged from rapid cooling, not heating. - MC

Incandescent lamps that operate at 60 Hz are actually modulating at 120 Hz. This is because the lamp does not care about polarity, i.e. the filament heats up when the polarity is positive or negative. Put another way, brightness is lowest (but certainly not zero!) at the zero crossing point, and maximum at “peak positive voltage” and “peak negative voltage”. (This assumes the filament’s impedance has no reactive component, otherwise there will be a phase shift.)