Incandescent lamp longevity at half rated voltage

[Ken1]

Hi, Can someone tell me what the life expectancy of a 25 watt 120 volt AC 1000 hour bulb is when operated at 60 volts AC? I will explain why I would like to know this information in my next post.

[reloadron]

HI Ken

I am not sure how long the bulb would last. However, I can share his. Years ago my mother-in-law was total invalid under home care. I got tired of replacing the nite light bulbs so I tried something. I placed 1N4006 (simply because I had a pile of them) in the AC lines. So the bulbs saw half wave rectified DC. Running on the lower voltage they had at least double their expected life. Obviously less light output but saved me constantly replacing nite lights.

Ron

[MrAl]

Hi there,

I have tried that diode thing the past too and found it to help, and they
even used to make "bulb savers" that were just a diode that fit inside
the base of any ordinary bulb screw base. I havent seen these in a long
time though.

But anyway, reducing the voltage to a bulb reduces the amount of
degradation of the filament and so should make it last longer, and it
has been my experience also that it does make it last longer.
I have read articles in engineering books that state otherwise, that
the vibration of the bulb and the thermal shock contribute mostly
to the shortness of life of a typical filament bulb, and so reducing the
voltage doesnt help much, but i have found this to not really be true
because it depends on where the bulb is located which varies the
amount of every day vibration, and the lower voltage has to produce
lower thermal shock.
It's not hard to do a test...run one bulb with a diode and one without.
After maybe 5 bulb changes see which one had to be changed the
most often.

[k7elp60]

To add to what MrAl has said.
Incandescent lamp life ratings should be taken with a grain of salt. Life ratings are based on shock-free, vibration-free, ripple free testing. Lamps are not turned on and off during life testing, so there is no thermal cycling.
Increased life can be obtained by derating the lamp—using it below rated voltage. This will diminish the light output, but greatly increase lamp life. Just decreasing the operating voltage by 10% can increase the life 5 times.

I developed a circuit to bring 12V lamps up slow, to eliminate the initial surge current, as the cold resistance of a lamp is about 1/10 that of a hot lamp. As a result the initial surge current is about 10 time normal current. In my case I think I extended the not only the life of the lamp but the life of the switch to turn on the lamp. I use 12V RV lamp fixtures on my bed for reading. The existing lamps (#1141) have been in use over 7 years.

I noticed a long time ago that most lamps burn out with the power is first turned on. I reasoned that the surge current was one reason for the shortened life of bulbs, and was my motivation to develop the initial surge current.

Ned

[Externet]

I would venture to guess 10 times life.

I installed 230V 200 W bulbs instead of 120V 60 W bulbs in exteriors of my mother's house to avoid the difficult access periodic replacement.
It has been 12 years now, burning all nights.

Did the same to a radio antenna tower; never replaced again until lost connection with that activity.

BUT I have a doubt about the practice... Somewhere I read the filament should burn at a certain high temperature to promote re-deposition of vaporized metal back onto the filament itself; otherwise it deposits in the glass surface... if I got that clear... Undervoltage may promote the blackening due to that... or to 'sucking too much darkness'

Miguel

[dacflyer]

years ago i think in the carol wright catalog, they had a lightlight that had 2 night lights in series, and they claimed that it could burn for many many years..
one of my friends got one, and its been about every 8 yrs before he has to change the bulbs..
not sure how long it would last with a 25 watt lamp.. i have some night lights here in a staind glass clock, 4 lamps, on a dimmer. i have it burning at about 30% full power, and i haven't changed them in 10 yrs,, and it still gives a decent light for nightlighting.

[jwax]

Miguel- You definitely do not want to suck too much darkness!
Years ago I built a prototype for a small company that was a "Zero-Crossing" incandescent lamp switch. Primary advantage was you only needed tiny wire to the switch, and the big 14GA that handled current was a shorter run. Less copper use, and only microamps available at the switch, so safer.
Side benefit was, when the switch went to "ON", the device would wait until the line voltage was at zero volts before connecting to the load, so the lamp always saw a zero to max initial application of current. A "Soft-Start". Consequently, his testing indicated double to triple lamp life, but without a "Certified" lab to test it, couldn't make that claim himself.
Seemed like a great product- wonder if it ever went into production?

[Dean Huster]

Miguel, the redeposition thing is with halogen lamps. They have to operate at temperature for the halogen cycle to work. I don't think it matters with standard incandescents. In fact, it would seem that the farther away from the melting point of tungsten you get, the less filament evaporation there would be and the less "dark sucking" that would occur.

Dean

[Bob Scott]

I used to believe that a lamp dimmer would increase the life of my living room table lamps. As they burned out, I replaced both of the expensive and short lasting trilights with plain 200 watt bulbs, and a 650 watt dimmer-on-a cord I bought at Ace Hardware. This dimmer is the type that has to be turned down to the dimmest position before it switches off. Using this type of dimmer, there should be no surge through the cold filament because it turns on like an old radio with an on-off-volume control, at minimum power before it slides to whatever setting you choose.

I have found that the bulbs in these lamps still burn out as often as before, faster than any other bulbs in the house.

[Ken1]

The reason I asked this question is that I got tired of replacing the 15 watt bulbs in the exit signs in my church. They would burn out every 2 or 3 months. Since there are 2 bulbs in each sign, I decided to rewire the 2 bulbs in series and increase the wattage of the bulbs to get the same brilliance. The sockets have pigtail leads so it was a simple matter to rewire them and I found 25 watt chandelier bulbs have the same bases. I wish I would have marked down when I did this, but it's at least 6 years ago now and I have yet to replace one and they are just as bright as when they were new. It's also the cheapest solution to the problem of these bulbs burning out, certainly much cheaper than the LED arrays they sell nowadays. I thought I would submit this question to see if perhaps there is a formula to calculate longevity of incandescent lamps as voltage is increased or decreased fron the rated voltage.

[Externet]

Oh! you want the formula...
Longevity = Differential of supply voltage times phase correction factor divided by the RMS power dissipation differential of the original minus the replaced series bulbs; all that multiplied by the square root of the diameter of the venting holes in the enclosure in millimeters. Add the factor of ambient temperature in worst August condition plus one seventh the logarithm of the number of people attending mass as they contribute each with 37.5 Celsius.
As you see, it is not a constant due to attendance or season of the year, but will help you with a near figure.

If this does not give you a close enough calculation, is because the bulbs were made in China, with unreliable performance curves in their data sheet.

Miguel

[philba]

I have a lighting system in my house which allows me to control the power to the lights (via triac control). One of it's features is that you can slowly ramp the power to a given "zone". I do this on all my lights and do not find that it matters much. Even limiting the max power to 90% doesn't help much. However, I have a number of zones that never run about 60% (hallway, for example) where I have never changed a bulb in the 8 years since the system has been working. These lights are on every day.

The comment about vibration is totally true. I have several lights that are subject to my herd of elephants (light under the stairway, for example) which burn out at much faster pace. I estimate their lifespan is about 75% the rating.

By the way, I take a sharpie and write the installation date at the base of most bulbs so I can see what the reality is. this is how I know whether a bulb has failed prematurely.

[MrAl]

Hi there,


There might be something to be said about resistance vs Triac phase
control vs Diode. With the triac, the bulb still gets hit with a possible 170v
steep transient every half cycle, whereas the pure resistance method (two
bulbs in series) would not. This could make a big difference.
The diode method, BTW, does not hit the bulb with 170v transients,
but rather one half cycle per cycle of the line frequency.
The resistance method always provides a smooth turn on and turn off,
or at least 50 percent of that obtained by using a single bulb alone,
so i would think this would be the best overall. I havent extensively
tested these methods in a long time however.

These three techniques are all different and need to be evaluated
separately.

[haklesup]

I don't have the reference handy but I do recall that although bulbs tend to burn out right when switched on, switching a bulb on and off does not reduce its lifetime. In general, with normal use, its still pretty close to an elapsed on time, flipping the switch is just the "straw that broke the camels back" in most cases.

Running 130V bulbs at 120V is the typical way to obtain long life, so even a few percent below rated gives a significant boost in longevity. 1/2 rated voltage should result in a very long life but is probably uncharacterized


http://www.newton.dep.anl.gov/askasci/e ... g99244.htm

[MrAl]

Hi again,

I agree with the thermal shock being the "last straw" after the filament
has already worn down in perhaps one place, making it the weakest
place.

Also, here is a formula derived from
"Standard Handbook for Electrical Engineers", by Fink and Beaty...

Life_Actual=Life_Rated*(Volts_Rated/Volts_Actual)^d

where

d=13.1 to 13.5 for tungsten filament

For example, Life_Rated might say 1000 hours on the package
when run at Volts_Actual=120v.
Using the formula it can be seen that running a bulb at 90 percent
of nominal voltage will provide a lifetime that is about 4 times longer than
the nominal lifetime.
The formula is valid for voltages between 90 and 110 percent
of nominal, and this is the theoretical formula and does not take
into account things such as vibration, handling, cleaning, etc.

We might feel bold enough to extrapolate down to 80 percent of
nominal voltage too, accepting a somewhat larger error. This would
give us close to 20 times longer lifetime (assuming again other factors
dont take it out first) and after the error maybe 25 times longer.