Lots of light bulbs going out 1

[BronYrAur]

I seem to have an inordinate number of light bulbs going out at my house. It seems like 5 or 6 have gone out in the last few weeks. I've only been in the house for a few months, so I don't have any history. The bulbs are all in recessed cans and are mostly 65W indoor flood bulbs.

Is this just a weird coincidence or do I have some strange power quality problem?

Any thoughts?

 

[electricpete]

A small correction (not important to the main point). I think the full-wave rectified current would contain components at 60, 120, 180 etc hz. Not 30hz.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

A small correction (again!)... I meant half-wave

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[powerwagon75]

Well, here's my pennies worth--I personally had a short bulb life issue years ago when I was living in Texas. I started noticing about a month after a direct lightning strike to my house. Decided to monitor before/after work. Morning voltages were anywhere from 128-134, afternoon was still running 127-131. Called utility company and told them what I had been monitoring. They were there within the hour, and had the pad transformer changed out in two hours. Their assumption was that "the lightning strike most likely caused the neutral connection to burn." [their quote, not mine]

 

[itsmoked]

P = V²/R
Then the temperature is some power of a power.

Then there is some tungsten in a vacuum evaporation curve that is probably a power owing to Area so you end up with a bunch of powers X powers of powers. <just my guess>

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[jraef]

e-pete,
What I meant was the bulb filament is only getting power to it 30 times per second instead of 60. With a sensitive enough light meter connected to a scope I have seen the flicker and it looks like 30Hz. My old analog light meter doesn't see it though, the movement is probably too slow to pick it up.

Eng-Tips: Help for your job, not for your homework Read faq731-376

 

[itsmoked]

Thanks for the info jraef.

I did try "Lights of America". True garbage of the lowest quality. None lasted more than two months.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[electricpete]

I guess I can see power^power. But who'dve thunk V^12?

I wonder if it is the steady state condition or the inrush that limits life?

60hz sinusoidal current has 60 positive peaks and 60 negative peaks per second. If you get rid of 60 negative peaks per second, you still have 60 positive peaks per second.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[MikeHalloran]

I'm getting really good bulb life on outdoor lights that are switched on dim at dusk by light sensors, and switched to bright by motion sensors. Bulb life was disappointing regardless of brand in a succession of new fixtures in the same location, switched on manually at dusk and off at dawn.

Some premium bulbs had longer lifetimes, but not enough longer to justify the higher price. Some premium bulbs are just high priced.



Mike Halloran
Pembroke Pines, FL, USA

 

[ko99]

Let's see, radiated power is related to Ttungsten^4 - Tair^4. Tungsten is about 2800K, so I can see filament temperature related to V^6. Don't know how life relates to tempearture but it's probably exponential...

ko (

http://www.ecooling.biz)

 

[itsmoked]

electricpete; It is purely the loss of tungsten through evaporation that kills the bulb. As a metal evaporates it has to do with the temperatures of the surfaces "it sees". This means the filament evaporates unevenly as some of the filament sees other filament some sees glass, etc. Ultimately some point becomes thin enough that it presents extra resistance, and drops a little more voltage, over heats, and opens. Often at turn on.

By the way the longest burning light bulb in the world is just down the street <sorta> actually closer to jraef than me. It has been burning for 105 years continuously. In a Livermore firehouse. It's so old it is a carbon filament bulb. It's 4 watts. And lights the fire truck bay. It was hand made and is likely built as a 150V bulb or something and it is making great use of the y¹².

Math test anybody?
Following the equation above:
And using the assumptions;
1) Actual Voltage = 120V
2) Life at design voltage = 1200 hrs
3) Actual life 110 years

Life of bulb = (Design voltage/Actual voltage)¹² X Life at design voltage

What was the design voltage?


Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[electricpete]

Ttungsten^4 - Tair^4 represents REMOVAL of heat energy from the lightbulb. That tends to LIMIT the operating temperature difference.

It brings to my mind an approximation.

Let's say resistance varies with temperautre
R = R0*(T/T0)^n where n>0

HeatIn=V^2 / [R0*(T/T0)^n]
HeatOut = T^4 (approximate that Tair is much smaller and that other energy loss mechanisms are insignificant compared to the radiation).

Equilibrium temperature occurs where HeatIn = HeatOut.
T4 = V^2 / [R0*(T/T0)^n]
multiply each side by T^n
T^(4+n) = V^2 / [R0*(1/T0)^n]
Raise each side to the power 1/(4+n)
T = V^2/(4+n) * [R0*(1/T0)^n]^1/(4+n)
Note the stuff on the right is all constant with respect to T and V. Remove it and change = to ~

T ~ V^2/(4+n)
Since n>1, this is less than V^0.5, i.e. T varies as less than the square root of V.

jraef - did you see my comment above?






=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[jraef]

I did electricpete, I see your point (all 60 of them I suppose ) and I should have known better considering I work with variable time base zero cross SCR controllers. D'oh!

The wave shape on luminosity from the scope threw me off course though because it looked like 30Hz. Can't think of why in light of your clarification, maybe measurement error or response delay in the filament. I was letting my observation cloud my reasoning. Dangerous thought process.

Eng-Tips: Help for your job, not for your homework Read faq731-376

 

[ko99]

Thanks for catching my mistake, electricpete.

ko (

http://www.ecooling.biz)

 

[wfowfo]

Now that we've re-invented the wheel, has anyone checked this guy's system for a bad neutral?

Get a voltmeter and check your voltage leg to ground. Turn on a heavy 120 volt load (like a microwave oven)on the other leg and see if your voltage goes up. If it does, go check your neutral connections.

 

[LionelHutz]

CFL's - I put 8 GE's in my garage a few years ago and I'm happy with them so far. They're the ones with 3 U loops coming out of the base, not the typical spiral. I think they are 18W - in any case they are very bright and completely silent in operation.

I was half expecting to have 8 extra CFL's to find a home for but they actually exceeded my expectations. At the time, they were selling for $12/4 with a $5 mail in rebate.

 

[SphincterBoy]

Many years ago an eye doctor I was seeing informed me about a problem caused by mercury vapor lights: subliminal flicker, which is hard on the eyes.

Regular filament bulbs do not exhibit this problem.

I also use the compact flourescent bulbs in my house, but for my reading lights, I use the filament bulbs (130 V).

I've noticed that when the compact flourescent bulbs run continuously, they burn out quickly. If you want 'em to last long, you have to make certain they do not operate 24/7.

 

[BronYrAur]

I'm glad this thread has interested so many of you. Thank you all for your ideas. I measured my voltaghe all weekend and found it to be an average of 125V, although it was not very hot here (Chicago). Should I wait and measure the voltage during a peak condition, or should I call the utility now. Had another bulb go out too.

 

[itsmoked]

LionelHutz; Interesting.. But names.. need names of quality bulb maker. (not shape)

SphincterBoy; I'm with you on that! I read at least an hour a night (SciFi) and dimmable tungsten is the only way to go for the color and the lack of flicker.

BronYrAur; We're getting a lot of mileage out of your question!! I think I would call them and complain that all your bulbs keep burning out quickly due to your voltage. See what happens and what they say. Can't hurt and could help!

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[ScottyUK]

'smoked,

In Europe the best CFL's I've found for longevity are made by Philips. I've used them in the exterior lighting at home with better life than I got from GE and some of the anonymous types out there. I'm not sure how much of the improvement is related to our climate - England is pretty damp, so perhaps the Philips design is better suited to it. I'm also suspicious of low amplitude vibration of the luminaire caused by wind. I'm certain this contributes to the premature demise of the tungsten filament lamps, but I wonder if some CFL cathodes are also susceptible to this failure mode.

Aside: I keep receipts for the mains voltage halogen lamps bceause they're expensive, and if they don't make it past 6 months I take 'em back. I've never been refused a refund!

----------------------------------

I don't suffer from insanity. I enjoy it...

 

[jraef]

Keith,
The one in my porch light fixture that has been operating for 7 years or so is a Phillips EL/O 18, 18W Outdoor rated. it's only rated officially for 10,000hrs, but I must have more than that by now.