MSW inverter use GFCI to protect short to ground?

Discussion in 'General Electronics Chat' started by sdowney717, Mar 11, 2015.

  1. sdowney717

    Thread Starter Member

    Jul 18, 2012
    385
    13
    These MSW inverters have 60 vac power on both black and white wires. The 'neutral' is not bonded to ground.
    I know the inverter is internally protected against shorts from black to white wires.
    But the manuals say short from black or white to ground will damage them.

    so
    What gets damaged? the internal fuses or is it more serious damage?
    Can I use a GFCI breaker (receptacle style) to protect the inverter from a potential ground fault?
    Will the GFCI react trip fast enough to prevent damage to the inverter?

    Would a back feed from another AC mains source into the inverter using a gfci inline protect it or will it just burn out?
    Would a backfeed into an inverter only damage it if it is running?

    It seems if a ground fault short can easily damage these, then why dont they design all of them to prevent it seeing it might easily happen?

    How much current flow from white or black to ground can they take without burning up?
     
    Last edited: Mar 11, 2015
  2. Hypatia's Protege

    Distinguished Member

    Mar 1, 2015
    2,780
    1,220
    I believe the concern is output to input 'coupling' via a ground path --- A GFI may help where the anticipated failure etiology is over-current, Over EMF induced failure, however, occurs FAIAP instantly...

    Best Regards
    HP
     
  3. sdowney717

    Thread Starter Member

    Jul 18, 2012
    385
    13
    Won't the short of a hot to ground on the AC output be an overcurrent event?

    What would be an EMF failure with an inverter?
     
  4. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,509
    2,369
    We are talking about an invertor isolated from an AC supply? If so I don't see what harm grounding one side would do?
    Max.
     
  5. sdowney717

    Thread Starter Member

    Jul 18, 2012
    385
    13
    Yes, inverter isolated from AC supply.
    The problem is when one of the white or black wires is grounded in a MSW inverter.

    Also I just read about GFCI outlet breakers seems they only break the black wire leaving the white wire connected, so then if a short between white and ground occurs with a MSW inverter, the GFCI would trip leaving the short still shorting the inverter. So maybe not as useful to me. Somewhere I have a double pole GFCI from an old electric yellow pressure washer, and they do disconnect both poles.

    http://en.wikipedia.org/wiki/Residual-current_device#Break_time_.28response_speed.29
    Anyway does this make sense?
     
  6. Hypatia's Protege

    Distinguished Member

    Mar 1, 2015
    2,780
    1,220
    Not (necessarily) in the absence of a ground reference

    There would be no trouble so long as one was certain the entire system was ground-isolated...

    Best regards
    HP
     
    Last edited: Mar 11, 2015
  7. Hypatia's Protege

    Distinguished Member

    Mar 1, 2015
    2,780
    1,220
    As, for instance, output EMF applied to the input via 'ground' leakage...

    Best regards
    HP
     
  8. sdowney717

    Thread Starter Member

    Jul 18, 2012
    385
    13
    I just tested the Leviton GFCI outlet and when tripped it DOES disconnect both neutral and hot.

    Question is does it detect a neutral to ground short and trip? My thinking is it should do that.

    Adding some more, maybe the GFCI outlet is bad.
    Touch ground to neutral -- GFCI trips off
    Touch ground to black -- sparks fly! blows the house breaker.

    I verified the outlet polarity and I have it wired up with a cord from a Microwave oven, black wire to black side, white to white side in inverter.
     
    Last edited: Mar 11, 2015
  9. Hypatia's Protege

    Distinguished Member

    Mar 1, 2015
    2,780
    1,220
    A functional GFI will trip upon detection of an N-H imbalance in excess of Spec.

    In all cases a functional, properly installed, GFI should trip long before the load center interrupter.

    BTW, in the interest of fire safety, instead of a conductor, you may wish to investigate the GFIs performance via use of a 220K resistor

    Best regards
    HP
     
    Last edited: Mar 11, 2015
  10. sdowney717

    Thread Starter Member

    Jul 18, 2012
    385
    13
    I just thought it through some more. It is a working GFCI outlet.
    Since a house has the ground and neutral bonded in the panel they are as one wire. So the GFCI wont trip from black hot to ground. It is like running black hot to neutral to send current to a device.
    GFCI outlet will only detect h-n imbalance or neutral to ground imbalance.
    Even disconnecting the line ground on the GFCI wont help me as the boat has a DC to AC ground connection bond, so a short would just flow around by way of the DC negative wire back to inverter. I would have to rig a relay to break the AC-DC bond and that I dont know if I want to do.

    So I would need a true dual pole GFCI breaker of 15 amps I think to protect both hot and neutral. I suppose using a GFCI designed for a house with neutral to ground panel bond is better than no GFCI, unless it nuisance trips on the boat.

    My goal was to aid in keeping the inverter from self destructing in case a hot wire gets grounded. I dont think normal fuses or breakers can act fast enough to prevent it from frying on a short.

    The inverter is on a boat and is never connected to another source of power unless something unfortunate happens in the relay transfer box. Boat has onboard gen with neutral-ground bond and shore power and inverter power.

    Maybe something like this, wonder if it would work with an MSW inverter?
    http://www.homedepot.com/p/Siemens-20-Amp-2-Pole-GFCI-Breaker-QF220P/100184710

    The white wire would I think have to be the ground reference. But I wont spend money if it will not function, anyone know?
     
    Last edited: Mar 11, 2015
  11. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,509
    2,369
    This is exactly when a GFI trips?
    If an AC supply has no conductor connected to earth ground then there is no neutral.
    Both conductors are considered live and should be individually fused.
    There are very fast acting fuses know as Rectifier Fuses, but quite expensive.
    mAX.
     
  12. Hypatia's Protege

    Distinguished Member

    Mar 1, 2015
    2,780
    1,220
    Correct, at line frequency, wired as described, the G and N receptacles are electrically identical

    Actually only H-N --- Ground is irrelevant (a [true] ground fault will upset H-N balance)

    Re: the remainder of your post --- As per the forum's TOS I cannot help you with vehicular topics... :(

    Best regards
    HP
     
    Last edited: Mar 11, 2015
  13. sdowney717

    Thread Starter Member

    Jul 18, 2012
    385
    13
    Do you know if in a standard wall GFCI with a hot to ground short, is it supposed to trip a GFCI?
    I need to do some more testing myself. We have one in the bathroom which when I short hot to ground blows the panel breaker. That one though I cannot test or reset, the test button is stuck and when I short neutral to ground is does nothing. So for all I know, I have two defective GFCI outlets here at the house.
     
  14. Hypatia's Protege

    Distinguished Member

    Mar 1, 2015
    2,780
    1,220
    The operation you describe is not anomalous --- it would behoove you, however, to 'embrace' the distinction between 'earth ground' and 'house ground' :)

    IOW Hot to ground absolutely! -- hot to neutral-tied ground receptacle? Of course not...

    Best regards
    HP
     
    Last edited: Mar 11, 2015
  15. sdowney717

    Thread Starter Member

    Jul 18, 2012
    385
    13
    I just tested three more GFCI's
    two Levitons out on the deck, a third part of an extension cord.
    Shorting Hot to ground produce sparks and a tripped panel circuit breaker.
    Shorting neutral to ground trips just the GFCI. The one part of a cord is a GFCI from an electric pressure washer, the kind which everytime you plug in, you have to reset.

    So, unless I have 4 defective GFCI devices, a hot to ground short is not GFCI protected.
    What I did was simply energize circuit GFCI, so the line is on.
    Then short the neutral to ground terminal using an old wire with a grounded plug.

    This is for a standard single hot wire 120vac USA system with bonded neutral to ground in the distribution panel.

    It also appears I discovered our downstairs bathroom GFCI is defective, a short from neutral to ground leaves the power on, the GFCI never trips. So that is a plus to all my testing.

    I will go over to my inlaw's house tonight and test their GFCI outlets. They have three more.
     
    Last edited: Mar 11, 2015
  16. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,509
    2,369
    Sounds like you have some serious issues somewhere?
    Max.
     
  17. Hypatia's Protege

    Distinguished Member

    Mar 1, 2015
    2,780
    1,220
    A hot to *neutral* short is not protected!

    To see the unit react to a hot-side fault try this: correctly wire a GFI to hot and neutral only. Then tie the GFIs hot receptacle to earth ground (preferably through a current-limiting load --but you're gonna do what you're gonna do)... ;)

    FWIW: Often, an H-G fault to the G position of an outlet on another branch occasions an imbalance sufficient to trip the GFI --- Despite the load center N-G connection
    Best regards
    HP
     
    Last edited: Mar 11, 2015
  18. sdowney717

    Thread Starter Member

    Jul 18, 2012
    385
    13
    Hi, thanks for that test.
    A hot to ground trips the GFCI. I used a solder pencil to limit the current.
    A white to ground does not trip the GFCI. Of course in the house system, white and neutral are bonded.

    Interesting to me that neutral to ground will not trip it with either the line side ground wire connected to the GFCI line side terminal or not connected.
    While a cord plugged into the GFCI outlet will trip a neutral to ground short.

    I wonder if I can safely test the GFCI the same way with the MSW inverter using a current limiter like the solder pencil, and not have the MSW inverter short out? What do you think?
     
  19. Hypatia's Protege

    Distinguished Member

    Mar 1, 2015
    2,780
    1,220
    'Tis merely the wiles of The Imp of The Small Signal!;) --- Pleased I was of assistance! :D

    As a current limiter I strongly suggest a 220k resistor...

    Best regards
    HP
     
    Last edited: Mar 11, 2015
  20. sdowney717

    Thread Starter Member

    Jul 18, 2012
    385
    13
    Still though, you know what is bad about it?
    A direct short from hot to ground uses a lot of current with a GFCI, even while tripping the GFCI, sparks flew and the 15 amp breaker popped (with just a wire)
    So that makes me think this, if you had a direct short from a loose wire in an appliance internally ground itself, it might trip the GFCI and burn out the MSW inverter.
    Or say you plug in a power cord and it has worn insulation and it is internally shorted, it might also destroy the inverter.

    I find it really mind blowing these inverter producers protect the MSW inverter from overcurrent from hot to neutral (hot) but not hot to ground.
    http://gpelectric.com/files/gpelectric/Docs/Manuals/Go_Power_MAN_Inverter_GP-1750HD-2500_vA.pdf
    From the manual
    Am I wrong, what do you think about this?

    If I am right, How do you protect the MSW inverter from self combusting?

    Seems like you plug in a defective appliance and the MSW is toasted.

    Can the MSW ground be fused? What size? will it be fast enough to prevent destruction of the MSW inverter? I think a fuse wont help, because the MSW inverter is connected to the DC ground and internally bonded AC to DC ground. And that DC wire is huge can carry 200 AMPS or more. I doubt the MSW DC fuse would blow if you shorted AC hot to ground.

    Adding, I emailed the company to see if they know how to prevent damage. If they have a good idea, I will repost it here.
     
    Last edited: Mar 11, 2015
Loading...