Connect with us

Fluorescent fixture problem

Discussion in 'Electrical Engineering' started by Peter, Apr 23, 2010.

Scroll to continue with content
  1. Peter

    Peter Guest

    I've got a "no-name" clamp-on fluorescent fixture that no longer starts when the
    rocker switch is snapped from the off to the on position. The lamp has a
    polarized plug with an in-line black "brick" 7 1/2" x 2 1/2" x 1 3/8" that has
    the code "G0138" stamped above the code "GG10051F" on the bottom but no other
    markings. The 2 conductor cord from the brick goes to the base of the fixture
    where the clamp is located. The fixture uses a single 18W 4 pin double tube
    bulb with code G24q-2.

    I'm sure that the bulb is good because I have a second, identical fixture (that
    works normally) and when I swap the bulbs, the "good" fixture works perfectly
    with the bulb from the fixture that is not working. I also have a brand new
    spare bulb that I've tried in the non-functional fixture and it too fails to
    light in the bad fixture.

    The problem started spontaneously with no earlier indication of problems.
    Normally, when the fixture is turned on, there are a few quick white flashes in
    the bulb and the bulb lights and glows steadily. The behavior I observe is that
    when I snap the rocker switch to on, the bulb either has one quick white flash
    but then I only see the heaters glowing in each of the two tubes, or there is no
    white flash at all, and all I see is the glow of the heater filaments. I've
    tried plugging the fixture into another outlet in case the problem was related
    to grounding (I've read that these quick start bulbs need their circuits and
    fixtures to be grounded to work properly) but it did not help. The "brick" has
    always been entirely quiet and never got particularly warm, and that has not

    Any suggestions (besides ditching the fixture)?

    P.S. Brick contains a Fu-chi ballast PC-4526P-B for PL-C 18W, 300 mA, with 2
    wires at each end. No other components inside the "brick" case except for the 4
    plastic insulating twist nuts that connect the ballast wires to the line cord at
    one end and the cord to the base of the fixture at the other end. No foul
    smells or scorch marks.
  2. [SMF]

    [SMF] Guest

    Verify that the ballast has line voltage. If not, troubleshoot the

    Most likely the ballast tanked. I'm not sure what a replacement would
    cost versus purchasing a new fixture, but that is your call if you
    are attached to the fixture and they are no longer available.
  3. Peter

    Peter Guest

    OK, it seems unanimous here and on If the problem is the
    ballast, I'm ditching the fixture. Used every search engine I can think of,
    including metabrowsers and I couldn't locate a replacement (if this were a shop
    light I would have an embarrassment of riches deciding where to buy). Too bad,
    I probably didn't use the fixture for more than about 300 hours.

    I think I'll stick to the old fashioned fixture that requires you to hold down
    the on button while the filaments heat up, and then release the button. Those
    seem to last forever (or as long as the switch does!)
  4. Peter

    Peter Guest


    The fixture always did start with a "tink" "tink" "tink" each one corresponding
    to a flash of the bulb. I don't hear that noise at all now. Perhaps what I was
    hearing was a starter buried in the depths of the bulb socket?

    I was able to pull off the rotating shade, unscrew the bulb socket, pull it out
    about 2", and observe a 1" glass bulb that looks almost like a neon bulb with an
    opaque mercury-like metallic coating on the inside of the glass bulb. There are
    2 wires coming out of the base of this little bulb, 1 connected to the black,
    and the other to the white power wires that enter the base of the socket.
    Perhaps this is the hard-wired starter? There are no markings on it at all. I
    reattached the bulb, plugged in the fixture, and turn it on while observing the
    little glass bulb. Nothing at all; no glow, no sparks, no "tink" "tink" "tink".
    What do I replace it with?
  5. Peter

    Peter Guest

    OK. Can I use the glow tube I take from an FS-2 starter (rated for 14, 15, and
    20W bulbs)? I plan to cut the original wires off right at the glow tube bulb to
    give me nice long leads, and solder, not crimp them to the leads of the new one.
  6. James Sweet

    James Sweet Guest

    Yeah that should work fine, these things are really not very critical.
    The "tink tink" sound is the bimetal contact doing its thing in the
    glowbottle, eventually either the contacts wear out or the gas fill
    becomes trapped in metal sputtered from them.

    If you prefer the old switch start type, you can replace the starter
    with a simple momentary pushbutton, but the starters are cheap and
    readily available.

    Most of the new cheap fixtures are using electronic ballasts of
    essentially the same design as used in disposable compact fluorescent bulbs.
  7. Peter

    Peter Guest

    Follow up:

    I cannibalized an unused FS-2 starter I found in my "junk box" and wired in it's
    glow bulb in place of the defective glow bulb I clipped out. To my surprise and
    disappointment, when I replaced the CFL bulb, plugged in the fixture and turned
    it on, the fixture and the glow bulb both continuously flickered. I waited
    about 5-10 seconds to see if it would stabilze; it didn't. I turned off the
    fixture, waited about 10 seconds, tried again with the same result. I then
    added the capacitor from the FS-2 in parallel with the glow bulb (as it was
    wired within the FS-2). Same behavior.

    Should I assume that the glow bulb from the FS-2 is mismatched to this circuit
    (although the CLF is 18W and the FS-2 is rated for 14, 15, and 20W bulbs), or
    that something else is wrong in the circuit? Should I buy a starter with a
    higher rating and try again with that?
  8. James Sweet

    James Sweet Guest

    It's possible that the tube voltage is higher than the breakdown voltage
    of the glow starter. Try removing it entirely and touch the wires to the
    glow starter together for a second or two and pull them apart, if the
    tube lights normally then you just need a different glowbottle.
  9. Peter

    Peter Guest

    Which model starter would you suggest I obtain? I've been singularly
    unsuccessful on the web trying to find the specific electrical characteristics
    of the many different starter models to decide. There are quite a few that are
    supposed to be compatible with an 18W fluorescent lamp. Hint: The original
    glowbottle was larger, both in length and in diameter than the one I
    cannibalized from the FS-2.
  10. James Sweet

    James Sweet Guest

    You could measure the voltage. Just start it manually as I suggested,
    then measure across the tube with a multimeter. What you get will likely
    not be super accurate unless you have a true RMS meter, but it ought to
    be close enough. Otherwise just try the next size up and see what
    happens, starter for a 40W tube has a breakdown voltage of roughly 2x
    that of the FS-2 starter.
  11. Peter

    Peter Guest

    I can't see how measuring the voltage would be useful because I cannot find any
    starter electrical specifications (other than the wattage rating of the lamp
    they are intended for).

    However, your comment that a starter for a 40W tube shouldn't arc until the
    voltage across the internal switch is about twice as much as for an FS-2 gives
    me a great place to start. This weekend I'll pick up a starter rated for a 40W
    tube, cannibalize its glow bulb, and substitute it for the one from the FS-2.
  12. James Sweet

    James Sweet Guest

    Well I guess you'd have to measure the breakdown voltage of the starter
    as well in order to characterize them. In a nutshell though, assuming
    North America, 20W and smaller lamps are designed to run from a 120V
    line with a simple choke to limit current, while larger lamps require an
    autotransformer ballast to step up the voltage. In Europe, the larger
    lamps also use a choke as the 240V line is sufficient to light the tube.

    The starter has to be matched roughly to the sustaining voltage of the
    tube. If it is too low, the gas in the starter will break down even
    after the tube has struck, and if it's too high the gas will never break
    down at all and nothing will happen. To work, the breakdown voltage must
    be lower than the open circuit voltage available from the ballast, but
    higher than the worst case sustaining voltage of the tube.
  13. James Sweet

    James Sweet Guest

    My thoughts on true RMS came from past experimenting with HID ballast
    design where both the RMS and peak values are important, some lamp types
    are picky about the crest factor and if either the peak or RMS current
    values are out of line lamp life can be dramatically reduced and in some
    cases there is the risk of a violent failure. There is far less to worry
    about with fluorescent lamps though.

    Most cheap VOMs are averaging, their response to non-sinusoidal
    waveforms (which you'll have across a discharge lamp) and frequencies
    other than 60Hz (such as electronic ballasts) is unpredictable and are
    not the peak voltage either. Many inexpensive electronic ballasts
    produce an output that varies widely in frequency depending on lamp
    voltage and other factors so if you try to measure it with a cheap meter
    you get inconsistent results, and values measured with one cheap meter
    may be much different than the same measured with a different meter. A
    scope or power analyzer would be ideal, but in reality with this
    particular case you are better off just trying something, either it will
    work or it won't. Glowbottle starters and fluorescent lamps are not
    precision devices, their rated characteristics are approximate and vary
    throughout life. You have to be careful using ballasts and lamps in
    unsanctioned combinations but substituting a starter is unlikely to
    cause any harm.

    It's worth mentioning too that if you try to measure anything, don't
    connect the meter until after the lamp is lit. Choke ballasts can
    generate spikes big enough to damage digital meters while starting the lamp.
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