More accurately, if the Hot and Neutral currents are not equal within a
tolerance range, a gfi will trip. It doesn't matter where the current
leaks to; it can be to other than ground. It's just that usually it will
be ground. So, it's "Only if it leaks current from one conductor more
than another, regardless of whether it's to ground or not.". Ground is
irrelevant to the operation of a GFCI.
You are quite right that the current can also leak to a conductor of
another circuit. And GFCIs will work on an ungrounded circuit and
increase safety. But I wouldn't go quite so far as to say the ground
is irrelevant; grounding still improves safety. For example, if an
ungrounded appliance on a GFCI circuit has a high resistance short
from hot to the case, the GFCI won't trip until you complete the
circuit from the case to ground or another circuit's conductor. While
if the appliance is grounded, the GFCI will trip as soon as current
through the short exceeds the 5ma threshhold.
I only mention this because too many people thing GFIs work on the
It's strictly an imbalance between Hot and Neutral: whether the current
goes to gound or anywhere else. GFCIs do not require the ground lead to
operate and don't care about it. It's an assumption that a voltge
difference will be to ground but it could be to any other place such as
another line or short between adjoining cables by a nail, etc. etc..
Yup; that's the inductive/capacitive arguement for their tripping. But
if you think about it, even surges/spikes from a motor may be
latitudinal and not longitudinal. I know it's true that they don't
false-trip nearly as much these days so I think I might have to see if
there are any articles that explain how they do that. Maybe it's the
dv/dt they've started to look at.
I have a fridge outside that has always been on a GFCI and never
I guarantee you, if you put a 2-3 prong adapter on the plug of a
fridge that trips a GFCI (floating the ground) and then measure the
case to ground, you will see 120v. It may be spikes when it starts and
stops or it may be solid. Be careful not to get killed.
They develop shorts inside the compressor and that is why you have
burnt smell when you open up the freon line of an old fridge.
By measuring from the case (intended to be but not grounded and metal)
to earth gnd is a measurement of two different references, one of which
is floating. In a non-fault system, the green gnd wire only has a
reference back to the service box and carries zero current.
You are extremely unlikely to see 120Vac on that wire. If you do see
a solid 120Vac, then there is another ground connection somewhere on the
unit meaning a FAULT exists that needs attention.
Depending on the length of that floating ground wire, one end open,
the other connected at the service box, you will see, usually between
15Vac and 90Vac.
You will specifically NOT see the identical voltage you see between
Hot and Neutral.
Depending on what other current carrying wires may be parallel to the
floating ground, crossing it (little effect), or any other fields that
may impact it, including the earth's magnetic field and some other
rudimentary sources, along with the measuring instrument's input
impedance, battery or line operated, and whether one lead is earth
grounded (capacitively, usually), and the distance from the box, you
will see differing "phantom" voltages appear on the measuring
It's a very simple task to prove whether what you're seeing is a
phantom voltage or an actual improper connection of some sort. Just
lower the measureing instrument's input impedance with a bulb, resistor,
even a wet finger and you'll see the voltage drop to 0.0x V quickly.
You would have the same condition if you disconnected the wire at the
service box and connected it to the unit, but you'd have to go measure
at the Service Box to see it.
It may be spikes when it starts and
Phantom voltages are completely safe and have no potential behind them.
With one end of a wire open, current can NOT flow! Removing earth
ground creates an open ckt!
NO current can flow in an open ckt. With no source behind the
phantom voltage, as soon as it gets a path to anywhere, the voltage will
disappear. As soon as you remove your measuring equipment, the voltage
You will not measure any current unless you have something capable of
measureing micro-amps across a known resistor, or just using the meter's
input Z for that figure.
There is no need to fear a phantom voltage.
There IS a reason to fear a voltage though, in the event it's NOT a
phantom and is instead a fault situation where it is actually connected
somehow to something with a connection outside the intended ckt. That
has to be fixed and quickly! Else make sure the fire insurance is paid
What? If there ever WERE a burnt smell to it, you'd have a pretty
special nose to pick it out from the intentionally added "perfume" they
put into it. Freon is odorless and deadly; that bad smell is in all
freon perfume; they don't make freon without it.
Capacitive /inductive vs resistive coupling to ground. You can have
insulation good for 50,000 volts on a 110 device and still get an
inbalance on startup if the inductive OR capacitive reactance is too
I have never had a problem and I have an extensive network of lights
and receptacles in my 2600 square foot screen cage and out to a remote
shed and boat lift. It is all wire in EMT or RNC and all on GFCIs. All
told there is close to a half mile of THHN/THWN out there.
I'm curious--you said earlier that several hundred feet of NM cable
might be enough to allow 5 ma of leakage current. But then you
mention several thousand feet of THHN/THWN in conduit without any
trouble. Is the upshot that the leakage current in NM cable is much
greater (per unit length) because the outer sheath holds the hot and
ground in close proximity?
On Mon, 14 Sep 2009 15:29:55 +0000 (UTC), Wayne Whitney
There are generally several wires in each pipe (as in 333' of RX is
1000' of wire).
I imagine the longest single run an a GFCI is about 250' of wire in
EMT. The only time it has ever tripped was when I had water in a box.
One tip. Point all wirenuts up and keep them towards the top of the
box. They all collect water now and then. All outside wiring is a "wet
location". That is particularly true underground
While this may be true it has frequently been mentioned here on these
pages that any 115 volt motor equipped domestic appliance, fridge,
freezer, washer etc. should NOT be plugged into a GFCI equipped
circuit. Too much chance of a momentary unbalance! And they can't all
have defective winding insualtion? Especially those all-enclosed
fridge compressor units?
GFCI (So called Ground Fault ...... ) operate when there is a 'slight
imbalance' of a few milliamps (thousandths of amps) between the live
and neutral current flow.
During motor starting of any AC induction or other types of motors,
due to capacitance of motor windings to the grounded appliance
framework etc. there 'might' be a momentary slight current unbalance
which is quite normal and OK.
GFCI are designed to protect humans against a fault such as a wire
inside touching the metal frame of an appliance especially in damp/wet
conditions; such as an operating but faulty electric lawn mower, or
electric drill. (But they both have electric motors! So what gives?)
The human touching the defective appliance can provide a path to
ground and get a potentially lethal shock. The faulty path to ground
(through the human) unbalances the current and 'trips' the GFCI for
Can somebody make a reference to an electrical code that confirms the
Section 210.8 of the new code spells it out. As Wayne Whitney points out,
many of the responses are out dated, the new code have very few exceptions
for the GFCI outlets in required areas. If you stick your fridge, or washer
in a garage or unfinished basement, for example, they must be GFCI protected
Just for preciseness because too many people think gnd is neccessary for
a GFI to work:
It's not the current to ground that is detected; it is the difference in
current between the Hot and Neutral wires that is detected. The ckt to
ground is where the current went, but what's detected is the Hot/Neutral
current difference > 5 mA.
That's good to know. I use GFCI's for those, even on my shop tools but
I didn't know it was a requirement. It seems that as long as I keep one
tool to one breaker, I don't have any problems nowadays. I did think
once I had one that was nuisance tripping, until I realized it was one
particular tool when I plugged it into my bench. Rewired the tool; all
OK. Never did figure out the "problem" but it's gone now. And I've had
one GFCI that just plain quit working; no test, no set, nothing; it's
straight thru like it's not there anymore. Guess NOW, I'm planning to
switch that one out this weekend<g>!
HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.