Connect with us

fluorescent lights - electric ballasts

Discussion in 'Lighting' started by Tim923, Sep 15, 2004.

Scroll to continue with content
  1. Tim923

    Tim923 Guest

    Do all energy efficient electric ballasts upconvert the source
    frequency of 60Hz to 20-50kHz?

    I have a history of symptoms with normal fluorescent lighting
    (60Hz/120 blinks per second). I usually put up with it, but feel
    better in sun or incandescent lighting.
  2. TKM

    TKM Guest

    Yes. Typical electronic ballasts for fluorescent lamps operate the lamp
    somewhere in the high frequency range that you indicate. The reason is that
    lamp efficacy increases abut 10% compared to 50/60 Hz operation; but
    elimination of flicker is a side benefit.

    Incandescent lamps flicker on 50/60 Hz. too, of course, and some people are
    bothered by that if they are particularly sensitive to flicker. Low
    wattage lamps (thin filaments with low thermal mass) flicker more than
    high-wattage lamps. Google "flicker index" to find out more.

    Terry McGowan
  3. I agree with Terry but need to add that some electronic ballasts have
    power line ripple superimposed on the high frequency output that can
    reintroduce much of the 120 Hz ripple that you don't like. Usually
    these are lower cost, lower quality designs.
  4. Tim923

    Tim923 Guest

    I agree with Terry but need to add that some electronic ballasts have
    The SunLight Jr.:

    Any idea if this one has 120Hz ripple?
  5. There is no way to tell from the information provided. If you have the
    data sheet for a particular ballast, the lamp current crest factor
    (CCF) is a good indication. A high frequency electronic ballast with
    no power line frequency ripple will have a lamp CCF not much above
    1.414, the value for a pure high frequency sine wave. Since not all
    electronic ballasts produce "pure" sine waves, the lamp CCF lamp be
    higher. values up to 1.7 are allowed by the lamp manufacturers.
    However, if the value of not 1.414, then there is no way to know how
    much of the increase above 1.414 is due to a non-sinusoidal high
    frequency signal and how much is due to power line ripple, without
    actually measuring the output current waveform or the light produced
    by the lamp with an oscilloscope.

    As a practical matter, any ballast that meets has a lamp CCF of 1.7 or
    less (the industry standard) is unlikely to have enough power line
    ripple to cause you problems.

    Perhaps Terry has some additional thoughts on this subject as he is
    more involved with health effect of lamps than I am.
  6. The conversion is slow. It starts about 1 kHz and is complete before
    20 kHz.
    I disagree. A pure current source will determine the current waveform
    of any load: resistor, capacitor, inductor or any combination of
    those. You can also say the same thing for non-linear loads, as ling
    as the current source is good enough.
    The strongest dependence is on gas fill and pressure and then the tube
    diameter. Tube length is probably the least important consideration,
    unless the tune is so short that the electrode regions are much more
    important than the positive column.
  7. TKM

    TKM Guest

    while CCF data are typically published for ballasts, as Vic says, CCF is not
    a good measure of modulation and so may not indicate if there is visible
    flicker or not. It's a gap in electronic lamp/ballast specifications only
    since flicker index and percent flicker are both defined for electromagnetic
    lamp/ballast systems. From a practical standpoint, I haven't seen any
    electronic fluorescent ballasts with CCFs of 1.7 or less that also cause the
    lamps (either linear or CFLs) to visibly flicker; but sensitivity to flicker
    is subjective. Phosphor persistence is another variable. But if Tim923
    isn't bothered by incandescent flicker, chances are that fluorescent lamps
    on electronic ballasts will be O.K. too.

    Terry McGowan
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