Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Ground Fault Isolation

  1. Jun 18, 2007 #1
    We are using 120V commercial grade electronic ballasts, properly grounded, rated at 2Amps to drive two low pressure UV lamps submerged in water rated at 800mA. System requirements are to use GFI barkers with each ballast. This installation in a wastewater treatment plant, where there are motors, pumps and air compressors. GFI’s are randomly tripping at 5mA from ballast electronic current leakage up to 9.2mA. Can we resolve our nuisance GFI tripping by using 120V to 120V isolation transformers between the GFI and electronic ballast?
  2. jcsd
  3. Jun 18, 2007 #2


    User Avatar

    Well maybe the isolation transformers will help.
    It depends on what you're trying to achieve.

    a) you mention the facility has motors, pumps, compressors,
    etc. and there's some implication that maybe the power
    line fluctuations / noise from these may be relevant
    in causing the GFI's to trip. If the GFIs trip because of
    EMI / noise present on the power line, then they're
    erroneously tripping due to noise and not a 'real' ground
    fault. To solve problems relating to power line coupled
    noise, try installing power line EMI filters and/or
    filtering surge suppressors and/or filtered or "online" UPS
    units BEFORE the GFI so that the power supply INTO the
    GFI is clean and stable and the GFI won't be more likely
    to trip due to noise present on its INPUT.

    b) If EMI coupled through induction to the leads on the
    OUTPUT of the GFI (between the GFI and its load,
    the ballasts) is causing enough noise so as to contribute
    strongly to tripping the GFI, then you'll need to minimize
    the circuit loop inductance and EMI noise coupled from
    the ballast circuit loop to the GFI's output side. This
    can be done by using more closely spaced wires, twisted
    pairs of wires, series ferrite / iron chokes which each of
    the individual circuit wires is separately wound around,
    and by installing a series EMI filter (capacitor / inductor)
    unit just on the outside of the GFI before the leads run
    out toward the ballast load(s). Using more heavily
    shielded metallic conduit and boxes can also
    help somewhat in reducing the amount of induced
    EMI coupled to the GFI and its associated wiring.

    c) if you install an isolation transformer after the GFI
    and before the GFI's load, so that the transformer IS
    the GFI's only load then you've just created a somewhat
    distinct partition of your circuitry. Circuit A is between the
    GFI's output and the isolation transformer.
    Circuit B is between the isolation transformer and the
    ballast(s). The GFI will ensure there are no ground faults
    between it and its load, Circuit A, the isolation transformer.
    If you draw a significant imbalanced current from either
    the hot or neutral lines attached to the GFI to ground, the
    GFI will trip. If the current in the hot and neutral leads
    of the GFI is balanced (equal within a couple milliamps),
    the GFI should not trip. By using a transformer between
    the GFI and the ballast, and by (for the sake of
    example) locating the isolation transformer right next to
    the GFI, you're almost ensuring that there WON'T be any
    ground faults that the GFI will 'see' since unless you do
    cause a ground fault in the short circuit path between the
    GFI and the transformer primary circuit, there will be no
    ground fault for the GFI to detect IN THE CIRCUIT IT IS
    Now the isolated SECONDARY winding of the isolation
    transformer is CIRCUIT B and consists of the secondary
    of the iso xfmr and the ballast(s) which load it. One
    could create a legitimate ground fault on CIRCUIT B,
    relative ONLY to the SECONDARY of the ISO XFMR and
    the GFCI on CIRCUIT A would not detect it since the
    current in CIRCUIT A is still balanced (through the
    ISO transformer PRIMARY) and isn't ground-faulting.
    There could be unbalanced current from the secondary of
    the ISO XFMR but that's not what the GFCI is attached to
    and therefore it's not monitored.
    Certainly in such a case you'd be reducing the nuisance
    tripping of the GFCI due to a legitimate ground fault
    in the circuit leading to the ballast(s), but in such cases
    of a 'true' ground fault imbalance in CIRCUIT B, the GFCI
    on circuit A is providing no protection / detection.
    So could it help solve nuisance tripping? YES.
    Could it help to circumvent the very protection that
    the GFCI is supposed to provide in the major way it's
    supposed to work? YES, in that your ultimate circuit would
    include portions of it that wouldn't be GFCI protected
    against real ground faults.

    d) However since isolation transformers usually are
    somewhat effective line noise filters for some kinds of
    high frequency spikes, surges, EMI, the mere presence
    of an isolation transformer in the circuit COULD help reduce
    EMI related GFCI tripping due to the other motors /
    compressors / etc. in the vicinity. For maximum EMI /
    line noise protection, of course you should use an isolation
    transformer that also incorporates EMI filtration / surge
    suppression / choke type circuitry. That'd help against
    nuisance tripping that ISN'T due to a true ground fault.
  4. Jun 19, 2007 #3


    User Avatar
    Science Advisor
    Gold Member

    Are you sure that the UV lamps and wiring aren't leaking through the water? I would think this is the most likely reason for seeing a ground fault.

    Have you checked with other water treatment facilities to see if they have experienced such problems? And what they have done?

    If you use an insolation transformer then will your GFIs still be effective? I think not.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook