Connect with us

Weird Bulb

Discussion in 'Electronic Design' started by N. Thornton, Apr 30, 2004.

Scroll to continue with content
  1. N. Thornton

    N. Thornton Guest


    A light bulb went today - but its way of going out is a complete
    puzzle. I know they can act up for a second or less, but this one was
    doing all sorts of weird things for a good 20 minutes before it
    finally gave up. Much of the time it was flashing at maybe 8 or 10Hz,
    most of the time its output was well below normal. At times it pulsed
    at praps 2 or 3 Hz, sometimes went very very dim for a few seconds
    then brightened up again.

    It cant be down to filament evaporation surely.
    It surely cant be down to arcing either, as arcs will destroy a
    filament in under a second.
    And I dont see how it would be a loose connection: the behaviour was
    far too regular, and the filament survived far too long.
    What on earth was going on?

    I got a bit of a look at the element, its arranged as a 3/4 circle,
    and all looked intact except the end 2 sections. Its a 100w 240v 2000
    hour bulb.

    Anyone know how it could do all this? Should I be asking News2020?

    Regards, NT
  2. On our 120VAC/60Hz, floodlights will often buzz at maybe 10-20Hz for
    some time before failing. If you turn the switch off, they invariably
    will not turn on again. I think the filament breaks and the magnetic
    forces excite a mechanical resonance in the element, causing it to
    flap back and forth, arcing as the end touches.

    Best regards,
    Spehro Pefhany
  3. The arcing from a break in the filament is really nothing at all. It's
    comparable in brightness to shorting a 9V battery.

    The spectacular arcs are when the filament shorts the leads, melts the
    leads, and then the leads fall on to each other. It draws enough
    current to start an electric plasma fire. The best I've seen yet lasted
    about 10 seconds then blew a 20A breaker. Better quality light bulbs
    have a fuse wire to prevent this. It's a fire hazard.
  4. Jeff

    Jeff Guest

    The best one I saw was on a old XL 80cc motorcycle I got running for some
    friends. The ex-car head lamp turned blue when I downshifted too hard
    (brakes didn't work too well) on a big hill!

    I had a similar experience with a halogen bulb on my snowmobile when the
    voltage regulator wire formed a bad connection, which allowed full magneto
    voltage (at considerable power) to burn the bulb out and turn the light beam
    to a unique purplish blue that lit all the snow up. To add insult, the bulbs
    cost $15, and it fried both beams within an hour or two, while in the dark
    in the middle of nowhere. It was also hard to find that bad connection, as
    it came and went.
  5. I can do Better - I once stuck a Mercury high-pressure lamp into a standard
    socket because - "if it goes in there it's supposes to work like that" ;-)

    This thing lit up like a Neutron Bomb before the fuses popped - I though I
    had gone blind and everyone was falling over furniture etc. Was great fun
    though - afterwards, sortof 15 minutes later when the eyesight came back so
    we could change the fuses.
  6. N. Thornton

    N. Thornton Guest

    That would explain a lot. I guess I was just surprised that it would
    arc for so long without either the filament end burning away or the
    arc growing and bypassing the filament, and ending things quickly.

    Thanks, NT
  7. Jim Thompson

    Jim Thompson Guest

    Every few months I have a light bulb go with a flash and blows the

    What brands have the fuse? I generally buy GE and Philips.

    ...Jim Thompson
  8. N. Thornton

    N. Thornton Guest

    In England they nearly all have a ballotini fuse, its a requirement
    here. If it arcs over, the bulb is liable to explode from the sudden
    heat applied to the glass. Perhaps arcing over is less of a problem at
    110v, I dont know.

    A ballotini fuse is easy enough to see: instead of just 2 straight
    leads going from filament to base, one lead has a glass capsule
    containing a very thin bit of fuse wire part way along - you can
    usually see it.

    Regards, NT
  9. You can see the fuse in the clear bulbs. One of the leads near the
    socket has a thin section.

    The GE clear bulbs have a fuse but their spot halogens don't.
  10. L. Fiar

    L. Fiar Guest

    You sat watching it for 20 minutes, rather than just putting it
    out it's misery?
    Wassup... getting every last penny's worth out of it, or did you
    turn your stereo up and enjoy that disco lights effect?

    Seriously, though, they can very often last more than the second you
    say. Some can last a while, but I really cannot say that I have stood
    watching one to see how long it can suffer for.
    Switching it off and back on will end it's torment.
    Not always.
    Also, it is known that a poor connection can act as a rectifier. This may
    lead to some strange effects that may take a while to kill the bulb
    Although they rate bulbs in hours, the life also depends upon how often it
    is switched on and off. The surge at switch on does have an effect on the
    life of the bulb.
    I remember seeing a design for a bulb life extender, where it delayed the
    power to the bulb until a low point in the mains cycle. Not sure if it
    worked... or if it actually justified the cost when you consider the life of
    the added components.

    How about a really wild case... a 100 Watt 240v standard mains bulb, used in
    an "inspection lamp", exploding with one hell of a bang and glass
    Not just one, but replacements doing the same.
    The cause was a 100 amp arc welder being used from the same double socket.

  11. I once saw a filament of this arrangement break and form a stable arc in
    the gap. I even have a photo of it - although it only came out well
    enough to show the bright spots at the broken filament ends being heated
    by the arc.
    If the electric field (voltage gradient) in the arc exceeds that along
    the filament, then the arc will not blow up into a big bright blue flash.
    Also, arcs in argon and the usual argon-nitrogen mixture at around
    atmospheric pressure to ~1/10 atmosphere with current in the several 10's
    of mA to a couple amps are impressively dim, with little radiation output
    other than argon's infrared lines and incandescence of whatever the arcs
    are connected to.

    I have also known linear tubular halogens to form stable burnout arcs.
    In those, the gas pressure is higher and the electric field within the arc
    is higher.

    If the filament breaks while it is hot rather than when it is cold, it
    is easier to have a stable burnout arc because the current is lower.
    Higher current makes the arc more conductive and easier to blow up into a
    big blue flash. Also, a more compact filament design has higher electric
    field in the gas and that also makes it easier for the arc to expand and
    "blow up".

    - Don Klipstein (,
  12. Sometimes it works that way, but more often what happens is that the arc
    expands until it goes from one filament end to the other, and then not
    much is limiting the current through the arc except wiring resistance.
    Arcs get more conductive when you push more current through them, to the
    point of acting almost like short circuits unless something is limiting
    current somewhere.

    When a light bulb filament breaks, if the electric field (voltage
    gradient) along the filament is greater than that within the arc, then the
    arc will expand - usually quite quickly. High current arcs can exist on
    cold tungsten - there is such a thing as a "cold cathode arc". That is
    usually what is in a xenon flashtube.

    When a burnout arc blows up into a big bright blue flash, it is common
    (but not guaranteed) to see the filament break in two, sometimes even
    three places: Where the original break was, and also at one or both ends
    from the arc melting it there once it "blew up" into something drawing a
    hundred or so, even more amps.

    - Don Klipstein (,
  13. 120V bulbs usually have the thin piece of fuse wire, but without a glass

    - Don Klipstein ()
  14. Surprisingly little for most bulbs, actually. What happens is that an
    aged filament has a thin spot that suffers a fatal temperature overshoot
    during a cold start, and that is why light bulbs so often burn out during
    cold starts.
    The thin spots that would have a temperature overshoot during a cold
    start will not last long if kept running - they tend to deteriorate at a
    rate that accelerates worse than exponentially.
    More details in a web page of mine:
    I have seen soft starters with negative temperature coefficient
    thermistors. I once did a test to determine the voltage drop of one in
    normal use fully warmed up - it dropped enough voltage to explain a 50%
    increase in bulb longevity.

    - Don Klipstein ()
  15. ~^Johnny^~

    ~^Johnny^~ Guest

    Putting incandescent lamps in a microwave oven is a lot of fun
    (even if foolhardy)...
    wide-open at throttle dot info

    "The first step in intelligent tinkering is to
    save all the parts." - Aldo Leopold
  16. L. Fiar

    L. Fiar Guest

    But we are talking pennies for replacement bulbs.
    The higher efficiency bulbs are now quite cheap here in the UK, just
    a couple of quid or cheaper. Not only do they save in electricity,
    but they are claimed to last much longer than a standard bulb.

  17. I replaced the 60W incandescent bulb next to the back door of the garage
    today with a helically-wound fluorescent. The glass globe over the bulb has
    been missing since we bought the house, but i found out that an empty dill
    pickle jar fits perfectly.
  18. L. Fiar

    L. Fiar Guest

    Very illuminating, it certainly threw some light on the subject.
    I am no longer in the dark about it.

    I do hate when sites use Flash, Active-X and scripting where it is not
    needed, so your site is a rather refreshing change.

    OK to link from my site?
  19. Rich Grise

    Rich Grise Guest

    Nice site. Do you write HOWTOS as well? :)

  20. Dave Cole

    Dave Cole Guest

    I have seen dozens of examples of 60-100 watt severe-service bulbs used in
    'drop light' portable work light fixtures fall to the floor and suddenly get
    much brighter for 5-30 minutes, then 'flash' out. I assume they are either
    arcing, oxygen getting in and burning the filiment, or the filiment shorts
    (literally) to itself. New bulbs won't do this; only ones used for several
    hours. I guess I ought to do a postmortem on one to be sure.
    Dave Cole
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day