research
 Patents
 Books
 Literature
 Articles
 Timeline
 Auction Archive

about
 About this site
 Wanted to buy

bulb gallery

Incandescent:
C
carbon
WD
drawn tungsten
WC
coiled tungsten
WM
mini tungsten
WS
pressed tung.
FG
figural bulbs
XL
christmas
XS
christmas sets
T
tantalum

Discharge:
NE
neon lamps
AR
argon lamps
XE
xenon lamps
MA
mercury
MC
fluorescent
MS
special mercury

Hardware:
F
fuses
FX
fixtures
PF
plugs & fittings
SA
sockets
SW
switches

tube gallery

 X-ray
 Geissler
 Crookes
 Radio
 Box art

museum pics

 Dr. Hugh Hicks
 
Fort Myers, FL.
 S.Slabyhoudek

links

 Related links
 Submit a link

 

Author Topic: A bulb with a non-continuous filament that still works!?  (Read 25172 times)

Offline Yoshi

  • Hero Member
  • ****
  • Posts: 103
A bulb with a non-continuous filament that still works!?
« on: September 19, 2004, 01:41:00 am »
Hi everyone,

A few days ago I made an interesting electrical phenomena discovery that just baffled me. The best part is that I have pictures, audio, AND video to prove it!

I was sitting at my computer, when I suddenly started to hear a ballast-like noise coming from my room's fan. After taking a good look at it, I was totally amazed to see that one of the fan's bulbs, which was an incandescent 100w Duro-Mex bulb that was nearing its useful life, had a broken, non-continuous filament that was still giving off light!! I grabbed my camera and took a shot of it by covering my camera's lens with film negatives to protect its CCD, and to be able to get more than just a blank picture. Here is what came out:


 


Here we can clearly see that the filament is split in two, and that the points where the filament is separated are brighter. I had heard incandescent bulbs make ballast-like noise before, and when I did it was always when they were in their last moments. But I would have never guessed that this was because the filament was broken and still operating!

Here's an mp3 of the sound emitted by the bulb:

http://bulbs.2yr.net/images/forum/bulb-noise.mp3


Here's an additional shot of the bulb, with the "Duro-Mex 100w" stamping visible:


 


And here's a video of the filament:

http://bulbs.2yr.net/images/forum/bulb-video.avi


I theorized that there must be some sort of electrical bridge in the filament's gap, which kept a constant flow of electricity through the filament. So I assumed that this bulb would never light again if I turned it off. I tried it, and yes, it never turned on again (here's a video):

http://bulbs.2yr.net/images/forum/bulb-turnoff.avi


Surely there must be a logical explanation for this. Have any of you come across similar phenomena?


Regards,


-Yoshi

« Last Edit: November 14, 2004, 01:14:09 am by tim »
How do I add custom avatars? The feature doesn't seem to be working

Offline Alan Franzman

  • Nixie Tube Addict
  • Hero Member
  • ****
  • Posts: 104
  • Email a.j.franzman [A T] verizon [D O T] net
    • Nixie Tube Cross-Reference
A bulb with a non-continuous filament that still works!?
« Reply #1 on: September 19, 2004, 02:23:00 am »
While I've never seen nor heard of this phenomenon before, I do have a theory:

Since the filament was operating when it opened, and apparently the now-unconnected ends remained very close together (it looks like a "rough service" type bulb with multiple filament supports), the filament must have enough electron emission to operate as a pair of thermionic (vacuum tube type, directly heated) diodes in reverse-parallel.  As you guessed and verified, this arrangement can not cold-start unless you apply a high enough voltage across it to break down the gap without the aid of an electron cloud.

------------------
 
Alan "A.J." Franzman

Email: a.j.franzman at verizon dot net

------------------
A.J.

Offline Max

  • Sr. Member
  • ***
  • Posts: 72
  • 100W cleartop mercury lamp during run-up
    • the Lamp Review
A bulb with a non-continuous filament that still works!?
« Reply #2 on: September 20, 2004, 04:46:00 am »
Hi Yoshi,

You have here some fantastic pictures!
Your and Alan's hypotheses are correct, you have here a sort of self-ballasted arc lamp akin to both mercury blended-light lamps and pointolite sources!
The two bright tungsten beads terminating your filaments are definitely thermionic cathods, which as they operate in a gaseous atmospheres ensure that a high current runs in the plasma created between them - pretty much like in tungars!

The remarkable thing in your lamp is that neither its internal fuses blew up, nor its filament destroyed itself. In normal GLS lamp burnout, the created arc tends to spread a bit along the filament, which of course has the effect of short-circuiting a certain section lenght of filament. This naturally turns out to decreases the total resistance of the lamp, and hence increase the current drawn from the mains. This has the effect of increasing the filament temperure, and hence its thermionic emission which helps the plasma to spread and to decrease its resistance ... so in a loop-process, the current normally increases until something breaks...
In your lamp, the arc is stable certainly because the fill pressure is higher, and the gas mixture is different than in GLS lamps... That's a very interesting case!

You have also noticed a sort of ballast-like noise emanating from the lamp. This noise originates from the filament itself that mechanically constricts suddenly at each half cycle, during the brief period when the plasma is letting the current go through the circuit. The inionized gas does not bridge permanently the filament, but only when the applied voltage is higher than a certain level permitting re-ignition.

This reignition voltage depending on the temperature of the electrodes, it is then normal that when swithed off, the lamp will not restart. A simple tesla coil might not be enough to restart the lamp though, you may also need an autoleak transformer that will provide a high open voltage and high current that will help re-heat the filament.

Anyway, thanks for sharing this interesting sight.

Max

Offline Mónico González

  • Hero Member
  • ****
  • Posts: 146
  • Philips HP-80w Mercury Lamp
    • Mis Bombillas, peque?o museo virtual de la l?mpara el?ctrica.
A bulb with a non-continuous filament that still works!?
« Reply #3 on: September 20, 2004, 05:44:00 am »
Hi,
I've observed these same phenomenon a couple of times.
I'ts due, as have been answered yet, to the thermionic emission from both filament's ends in the gap.
The whole lamp are gasfilled, so the building up of an arc is an unavoidable thing when filament breaks during service.
The noisy "hum" from the lamp are due to the contractions an dilatations experimented by the plasma cloud around the arc, and this itself. Vibrations on filament at AC frequency (magnetostrictive, dilatation-contraction, etc.) does transmit mecanical waves to the wole bulb structure and its supports (socket, fan...) so, you can hear the 60Hz "buzz" coming from the lamp.
Due to the lower voltage in your Country, the ionisation into the lamp are not enough to extend the arc to the whole filament, or worse: between the lead-in wires, giving a strong short circuit and blowing-up the built-in fuses.
Here in Europe, the things are a little bit violent, because the higher line voltage of 230 volts. makes that the ionisation could be stronger; so, when the bulb fails during service, the arc strikes quickly and suddenly, resulting in a brigt and very white flash.
But in some scarce cases, when the filament brokes in two almost equally long halves, the arc strikes between both ends of the gap, being ballasted by these filament halves enough to avoid any direct arcing and fuse blowing.
During this time, it can be heard a glassy 50 Hz buzz.
When you turn off the bulb, and re-connect it, the phenomenon dissapears because the voltage accross the gap is under the ionisation value of gas when both halves of filament are cold.
Best regards,
M. Gonz?lez.

[This message has been edited by M?nico Gonz?lez (edited September 20, 2004).]

Offline Dave

  • Jr. Member
  • **
  • Posts: 30
A bulb with a non-continuous filament that still works!?
« Reply #4 on: October 14, 2004, 08:20:00 pm »
I have seen this weird thing happening with a number of incandescent bulbs that are failing! Recently, at the auditorium where I work, in one of the dressing rooms in the back (for the peformers to dress for the show) I turned on the lights which are 60w clear vanity globes. I noticed a slight flicker and I checked the lights, sure enough one of the bulbs' filament (it's a CC-9 w/ two supports, and was a Satco made in Korea) had a bright, hot spot! I turned it off and it never came back on. I have since replaced the bulb with an Ace brand which was made by GE for them.

Years ago, I had a Feit 75w R-30 reflector floodlight (long before the switch to the BR shape) in the living room and it operated for a year and a half. The bulb was made in Korea and had a C-9 filament. One day in early 1992 my mom reported hearing a funny noise from the lamp and she turned it off and it never came back on! I peered into the bulb's insides through the clear portion at the base, and I could see a gap in the filament and there were beads of melted tungsten on the ends of the filament at the gap. So the same exact thing happened to this bulb (and the one above) as your Duromex. BTW how long has your Duromex bulb been operating? Probably at least a couple of years, Duro-Test/Duro-Mex bulbs have always been the best!

Also, speaking of this funny noise that Max pointed out, note that large 750-1000 watt PS-52 lamps with CC-8 filaments make a quite audible hum when lit due to the AC current passing through the huge, heavy filament with lots of amps!
DMD

Offline Ed Covington

  • Sr. Member
  • ***
  • Posts: 91
    • http://www.frognet.net/~ejcov/
A bulb with a non-continuous filament that still works!?
« Reply #5 on: October 15, 2004, 07:36:00 pm »
I studied this effect some years ago in the testing rack area at GE. Other people before me and I called them "singing arcs."

Offline Hemingray

  • Hero Member
  • ****
  • Posts: 136
  • Lightbulb Junkie
Re: A bulb with a non-continuous filament that still works!?
« Reply #6 on: May 31, 2006, 07:33:05 pm »
Bringing this one back, just had this same thing happen to a house-brand 40W/120V bulb in an old brass lamp on my desk, made the same buzzing noise, no more light when turned off then back on.

Also noticed that light output was decreased dramatically when this happened.
« Last Edit: May 31, 2006, 09:29:11 pm by Hemingray »

Offline Mónico González

  • Hero Member
  • ****
  • Posts: 146
  • Philips HP-80w Mercury Lamp
    • Mis Bombillas, peque?o museo virtual de la l?mpara el?ctrica.
Re: A bulb with a non-continuous filament that still works!?
« Reply #7 on: June 13, 2006, 08:39:58 pm »
Three weeks ago, I've observed again the same phenomenon in a 40 watt 230 volt R63 reflector lamp.
When I went to my bedroom, I turned on the lights, a fixture fitted with three of these lamps. During the first seconds, nothing did happen, but half minute or so after the switch was closed, a typical glassy whistle modulated by a faint 50 Hz buzz, can be heard coming from one of the lamps, while the emited light by it was noticeably faded, but without a significant fall in colour temperature.
I watched directly to the lamp, within which, could be seen (through the frosted front of the glass envelope) a 5 to 7 mm gap at the filament, undoubtely "closed" by the arc. I turned off and on quickly the circuit, but that bulb did not back to give light any more when the plasma in the arc was de-ionized.
It's a curious thing to consider that most of times that I've seen this phenomenon have been precisely in these kind of lamps (R63). Why? I don't know, but perhaps it could have to do with the brand of these bulbs more than with the type itself.
The fact that the Kelvins did not went down while the flux did, could be relationed with the shortening in filament lenght, because the arc itself isn't almost luminous, and the remaining filament pieces were loaded almost at nominal current/voltage ratings as when the lamp was intact.
A new question are exposed: how many time could a lamp working at these conditions to last burning before complete fail?
Perhaps a "Novel Price's" question,? :wink:

Regards,
M. Gonz?lez.

Offline Lightbulb Collector

  • Jr. Member
  • **
  • Posts: 10
Re: A bulb with a non-continuous filament that still works!?
« Reply #8 on: September 18, 2008, 12:32:55 am »
I've got a better theory than the vacuum tube electron emission theory. First of all, only bulbs 40W and lower have a vacuum needed to act like a vacuum tube. Now here is my theory. Your 100W bulb is filled with pure nitrogen gas (air that has the oxygen removed).

When a bulb burns out, it usually makes a bright flash. That is an arc. But then it stops because the intense arc vaporizes more filament and then the gap is too wide to keep the arc going. But with YOUR bulb, you describe a humming or buzzing like a ballast. That is the sound a continuous arc would make. So I believe that you had an arc that did NOT vaporize the filament, but carried enough current to keep the filament heated. That is what kept your bulb glowing. From personal experience, I know that a nitrogen based arc (nitrogen is the gas in bulbs over 40W), is MUCH dimmer than the filament when carrying the same current. I have a neon sign trans former and when I use it to make an arc in a burned out light, the points on the filament that touches the arc get SO bright that you can't see the arc. However in my setup, only the points that touched the arc were heated. But that was high voltage + low current. With a 120V arc, you have low voltage at high current so the ENTIRE filament glowed (not just the ends that touched the arc). However I would still expect the ends of the filament touching the arc to be a bit brighter than the rest of the filament. And that is EXACTLY what I see when I look at your picture. And one last thing that shows it is an arc is the fact that after you turned off the light, you couldn't turn it back on. The fact is that with an arc, at 120V, that voltage is enough to keep the gas ionized, but it is NOT enough voltage to strike an arc. That's the same concept that makes a Jacob's ladder work. The arc heats the air above it, which makes it easier to ionize, so then the arc move up to the next position, and that effect happens all the way up.

And the fact that this forum seems to be in Europe, which means your voltage is 240 (not 120), means that you have a better chance of holding an arc that gets struck.
« Last Edit: September 18, 2008, 12:43:00 am by Lightbulb Collector »

Offline Max

  • Sr. Member
  • ***
  • Posts: 72
  • 100W cleartop mercury lamp during run-up
    • the Lamp Review
Re: A bulb with a non-continuous filament that still works!?
« Reply #9 on: September 18, 2008, 05:00:35 pm »
Hi,

If by "the vacuum tube electron emission theory" you mean thermionic electron emission, then it is useful to point out that this is this precise mechanism that is responsible for the emission of electrons from the filament and feeds the arc. The temperature at each filament's ends is high enough (i.e. see the bright ends) for the electrons to be "evaporated" from the metal. Thermionic emission does not only occur under vacuum, but under gaseous atmospheres as well. When the electric field is strong enough to create a plasma (i.e. a ionized gas), then the ions accumulate in front of the thermionic cathode and actually increase the electron current drawn out of the electrode (due to a local increase of the E field, and the cancellation of the electron space charge than tends to form and impede on the thermionic emission) - this is what precisely occurs in a gas-filled GLS lamp presenting the effect shown here.

Max

Offline Lightbulb Collector

  • Jr. Member
  • **
  • Posts: 10
Re: A bulb with a non-continuous filament that still works!?
« Reply #10 on: September 19, 2008, 11:28:07 pm »
I thought it was emitting electrons, simply because the arc itself acted like a wire between the parts of the filiment, so simply the electrons were flowing through the plasma of the arc like they would normally flow through a wire WITHOUT thermoionic emissions.

Offline Max

  • Sr. Member
  • ***
  • Posts: 72
  • 100W cleartop mercury lamp during run-up
    • the Lamp Review
Re: A bulb with a non-continuous filament that still works!?
« Reply #11 on: September 20, 2008, 06:17:27 pm »
You are right, the electrons flow in the plasma, so in a sense the plasma act as a "wire". However, that's only a part of the story because a crucial element is also how the electrons leave the filament to enter the plasma, this is where the thermionic emission comes into play. This transfer of electrons from one medium (i.e. the metal) to another (i.e. the plasma) costs some energy and requires a certain electric field (a voltage drop across a so-called sheath present in front of the cathode). Out of all electron-emission mechanisms only the thermionic emission can ensure a low-enough voltage drop needed to feed the plasma with the right electron current. If this voltage drop is too high, then the 220V of the mains would not be enough to sustain the whole process.

Max

Offline Lightbulb Collector

  • Jr. Member
  • **
  • Posts: 10
Re: A bulb with a non-continuous filament that still works!?
« Reply #12 on: September 21, 2008, 09:25:46 pm »
Moving to different substances requires the electron have more energy? Wouldn't that mean that the electron would need more energy to move between copper and aluminum?

Offline Max

  • Sr. Member
  • ***
  • Posts: 72
  • 100W cleartop mercury lamp during run-up
    • the Lamp Review
Re: A bulb with a non-continuous filament that still works!?
« Reply #13 on: September 22, 2008, 03:57:24 pm »
Actually it depends on which direction the electrons flow - but indeed, in a contact between two different metals the electron energy changes when it moves from one material to the other. One caracteristic energy used to describe the "median" energy of the free electron population in metals is the Fermi energy Ef (this has to do with stastistical physics .. I will not go into that). Only electrons having an energy higher than Ef will drift in the cristal lattice of the metal and participate to electrical conduction upon the application of an E-field.
To take your example of copper and aluminum, these metals have a Fermi energy of 7 and 11.8 eV, respectively (1 eV is 1.6E-19J). This means that when electrons are forced to flow from copper to aluminum, these mobile electrons will need to gain some energy in order to rise up to the step in fermi energy when passing from Cu to Al ... this gain in energy will partly be acquired via the lattice vibrations of the material (i.e. thermal energy) leading to a cooling of the metal junction. If we take the inverse case, forcing an electron flow from aluminum to copper, the electrons will lose their excess energy when the fermi energy drops from one metal to the other. This energy loss results in the heating of the metal junction. Both effects are in fact the basis of the Peltier effect.
In case the metal junction is left at rest, the electrons will naturally flow from the material of high Ef to the other until an equilibrium is reached. This leads to an excess of electrons on one side of the junction, and a lack of these on the other side (i.e. that with higher Ef). This is the Seebeck effect, used in thermocouples.