GFCI Outlet Installation

[David]

I just installed a GFCI outlet in my bathroom. There are five wires that
connect to the outlet, 2 line, 2 load, and a ground. When I connect the
line wires to the line terminals and the ground wire to the ground terminal,
everything works fine. I get power. But when I then connect the load wires
to the load terminals, the reset switch keeps tripping and the power keeps
shutting off. I'm not sure why this is happening. Any ideas? I don't think
that there are any loads down the line from the GFCI outlet, so I'm
thinking that maybe I should just cap the load wires and call it a day. Is
this o.k.?

Thanks for the help.

Dave

 

[w_tom]

It is tripping because you have a wiring problem. Fastest
way to determine the problem is to disconnect all appliances
from the down stream (load) outlets, then use your multimeter
to measure for conductivity between neutral and safety ground
wires. You should not even see megohms of resistance between
any two (of three) 'load' wires. Meter makes locating this
problem quick and (more important) decisive. For human
safety, you want decisive answers; not speculation.

 

[John Ray]

I just installed a GFCI outlet in my bathroom. There are five wires that
connect to the outlet, 2 line, 2 load, and a ground. When I connect the
line wires to the line terminals and the ground wire to the ground
terminal, everything works fine. I get power. But when I then connect the
load wires to the load terminals, the reset switch keeps tripping and the
power keeps shutting off. I'm not sure why this is happening. Any ideas?
I don't think that there are any loads down the line from the GFCI outlet,
so I'm thinking that maybe I should just cap the load wires and call it a
day. Is this o.k.?

Thanks for the help.

Dave

Make sure you've got your line / load right on BOTH hot and neutral. Its
possible the old one was wired backwards, as older GFCIs would operate as a
regular receptacle if wired wrong. The newer models will not. Easiest way to
check is look for continuity between neutral and ground on the line side.

John

 

[Paul A]

Make sure you've got your line / load right on BOTH hot and neutral. Its
possible the old one was wired backwards, as older GFCIs would operate as
a regular receptacle if wired wrong. The newer models will not. Easiest
way to check is look for continuity between neutral and ground on the line
side.

John

Actually, it appears that the new ones will act as a regular receptacle when
wired backwards. I just did that with a brand new Leviton "Smart Lock" GFCI
(don't ask), and it was a live outlet. Neither the Test not Reset buttons
did anything. When I got line and load where they belonged, Test and Reset
worked as they should.

The instructions claim that they are shipped from the factory so they have
to be reset before use and will not reset when wired incorrectly, but that
was not my experience.

Paul

 

[David]

It looks like everything is wired right. According to my instructions, if
you connect the line wires to the load terminals, the receptacle will still
operate, but it will not trip (shut off) when you push the "test" button
like it is supposed to do I just tested the continuity between the neutral
(white) and ground wires. It measured 0.

Thanks,

Dave

 

[w_tom]

You tested 0 ohms between neutral (white) and ground.
Therefore GFCI complains about defective wiring. A wiring
problem that may have existed since the house was first
built. Find that short in one of the 'downstream' (load side)
wall receptacles. Open each cover plate and inspect for a
short between ground and neutral wires. And then confirm,
with meter, that the short has been eliminated. If necessary,
electrical tape those wires as well as separate them.

BTW, if any of those receptacles are powering computer
equipment, then every wire should be firmly attached via
receptacle screws on side; not connected by pushing into the
back. Each wire must be firmly wrapped around screw so that
it remains there even if screw is loose. Insulation on the
white wire should have never made contact to bare copper
safety ground. IOW the GFCI trip may be a symptom of a bad
electrical job inside one of those 'downstream' (load side)
wall receptacles.

Only white and neutral wire that should measure 0 ohms is
the wire that goes to breaker box. Wires that connect to the
GFCI's 'Line' screws. Those wires that connect to 'Load'
screws must measure infinite ohms (no conductivity). GFCI
will detect these wiring defects.

GFCI will not trip off when test is pressed? Did you
properly identify 'load' and 'line' wires? This can be a
symptom of reversed wires. But then this is also why you have
the meter. Use meter to determine with wire pair has 120 VAC
when circuit breaker is on. This would be the 'line' wires.

 

[David]

No.. I appreciate the suggestion. I like to install things, troubleshoot
problems on my own, whenever I can. I learn that way. Sometimes, though.
I've found it's better to let a professional do the install/repair or at
least watch a professional do it the first time around to see how it's done.
I definitely don't need a house fire, damaged electrical components, or a
possible electrical injury.

Dave

 

[David]

Just a couple things.. The 0 ohm reading was between the line neutral (hot
wire) and the ground wire . The reading between the load neutral and ground
reads infinite. Is this o.k.? I'm sorry if I didn't make this clear. You
write about the wire holes and the wire screws on the receptacle. When I
the installed the GFCI, I pushed all of the wires (2 line and 2 load)
through their respective holes. Was this a mistake? There are holes
besides each line and load screw.

Thanks,

Dave

 

[John Gilmer]

BTW, if any of those receptacles are powering computer
equipment, then every wire should be firmly attached via
receptacle screws on side; not connected by pushing into the
back.

Are you saying that the GFCIs that "clamp" the wire as you tighten the screw
aren't "good enough?"

 

[John Ray]

Are you saying that the GFCIs that "clamp" the wire as you tighten the
screw
aren't "good enough?"

I think that statement was in reference to inspecting the standard duplexes
that may exist load-side of the GFCI, In which case I agree with him
totally.

John

 

[[email protected]]

I just installed a GFCI outlet in my bathroom. There are five wires that
connect to the outlet, 2 line, 2 load, and a ground. When I connect the
line wires to the line terminals and the ground wire to the ground terminal,
everything works fine. I get power. But when I then connect the load wires
to the load terminals, the reset switch keeps tripping and the power keeps
shutting off. I'm not sure why this is happening. Any ideas? I don't think
that there are any loads down the line from the GFCI outlet, so I'm
thinking that maybe I should just cap the load wires and call it a day. Is
this o.k.?

It is ok. Cap it, and see what doesn't work. There has to be *something*
downstream (ie electricaly farther away from the voltage source) because
you have wires going out as well as coming in. You might find that a seldom
used receptacle - perhaps in the garage or outdoors is dead. When you find
something that doesn't work in the future, remember the capped wires. When
you call the electrician in to fix it, tell him about the bathroom GFCI
& capped
wires. That will reduce the repair time.

Ed

 

[[email protected]]

Just a couple things.. The 0 ohm reading was between the line neutral (hot
wire) and the ground wire . The reading between the load neutral and ground
reads infinite. Is this o.k.? I'm sorry if I didn't make this clear. You
write about the wire holes and the wire screws on the receptacle. When I
the installed the GFCI, I pushed all of the wires (2 line and 2 load)
through their respective holes. Was this a mistake? There are holes
besides each line and load screw.

If this was the newer type of GFCI (probable) then the right way is
to install the wires in the holes - NOT to wrap them around the
screws - and then tighten the screws. Tightening the screws on
this style GFCI receptacle clamps the wires. Simple to check - if
the screws are loose, the wires pull out of the holes easily.

Unlike it being ok to cap the load side wires, it is NOT ok to
leave those screws loose!

Ed

 

[w_tom]

John Ray has corrected a misconception. All those other
wall receptacles, if powering things that require high startup
currents (ie electric motors) or computers should have wires
wrapped completely around the screws; not use the push in
holes on the back of conventional wall receptacles. That was
not a reference to GFCIs that can only connect using screws.

Dave has apparently modified an earlier post. There is no
conductivity between any two 'load' side wires. He was
measuring neutral to ground on 'line' side. BTW the
conductivity meter must be set to read in the megohm region -
not just measure for short circuit. The 'beeper on a meter'
is not a valid reading. The meter must read more than a
megohm resistance which is more than testing for a short
circuit. Dave may have done this correctly. But if he has
not measured a leakage, then other problems may exist such as
a leaky capacitor installed in some receptacle as a noise
reducer or some other problem causing a non-linear leakage.
For example, a 1000 ohm resistance between two wires would not
'beep' the conductivity tester but would be resistance so low
as to always trip the GFCI. 100K ohm resistance might cause
marginal GFCI tripping. But the resistance between two wires
much be more than 1000K ohms. Anything at or less is
considered a short circuit - to the GFCI..

Non-linear in this case would mean a leakage between two
wires does not exist when at low voltage (ie using a meter)
but exists when a higher voltage or a higher frequency current
is applied. Inspector for something unusual in that electric
box (ie electronic bugging device).

Other possibilities could include a short inside walls from
the white neutral wire to, for example, some adjacent pipe.
One way to check for this fault is to measure for high megohm
resistance from each 'load' side wire to the safety ground
wire on the 'line' side cable.

This we know. Something is leaking excessive current. If I
read the post correctly, this leakage only exists when the
'load' wire is connected to 'load' screws on GFCI. But I am
rather confused about another post where the test button does
not work. That is a different inconsistency.

I am rather concerned about a GFCI not tripping when the
test button is pressed. One reason why this can happen is
that the 'load' and 'line' wires are reverse connected to
GFCI. Another reason observed is that the 'load' side neutral
wire was wrapped together (wire nutted) with a neutral wire
from some other circuit. I have seen amateur electricians do
this. Two switches were to be powered by different circuits.
But the electrician wired all neutrals from both circuits
together. When asked why he did this, well, he always did
this for years. It was only when AGFIs were required by code
that his circuits had strange problems.

IOW be suspicious of any receptacle or switch boxes that
control power from two different circuit breaker circuits.
Just another reason why the GFCI would see a failure and
complain.

Again, we do know one thing. The GFCI sees and is
complaining about a wiring problem. It could be a problem
existing for decades - before the GFCI was installed.

How do electricians learn of the wiring schematic? They use
testers that put radio waves on those wires, then follow the
wires inside wall with a 'receiver'. They cost as low as $20
even in Radio Shack. However, with experience, the meter, and
process of elimination, one can also figure out (discover) how
each circuit is wired.

 

[[email protected]]

John Ray has corrected a misconception. All those other
wall receptacles, if powering things that require high startup
currents (ie electric motors) or computers should have wires
wrapped completely around the screws; not use the push in
holes on the back of conventional wall receptacles. That was
not a reference to GFCIs that can only connect using screws.

Tom - look at a new GFCI. The new ones should NOT be wired
by wrapping the wire around the screws. The wires SHOULD be put
in the holes in the back, and the screws must be tightened. These are
not the "push-in" holes you are thinking of. Those make contact with the
wires via spring tension. The newer GFCI's have a clamp in the holes,
which is tightened onto the wires when you tighten the screws.

 

[[email protected]]

Great... Thanks. I capped the load wires and put the receptacle assembly
back together. It works fine. With my multimeter, I checked the voltage of
all the outlets and switches that are on the same circuit line with the GFCI
receptacle (the receptacles that power off with the GFCI receptacle when I
flip the breaker switch). Are these the receptacles that I need to check?
There were 7 outlets and 2 switches that I tested. I think I got them all.
They all tested at about 120 volts. I'm not sure what else I could/should
test. I'd like to get a wiring diagram of my town house. That way maybe I
could tell for sure if I missed anything. Does anyone know how I can get
one? Would a home wiring diagram be helpful in this case?

Yes! Home wiring diagrams don't exist until you make one - and
you can't make one that includes the wires that are capped off.
The closest you'll get - until you make your own - is looking at the
labeling that is supposed to be at your service panel.

But all is not lost - in making your own diagram, you are likely to
find out where those capped wires go. To make a diagram:
draw each room on its own piece of paper. Mark *EVERY*
outlet. An "outlet" is any place where a device can use
electricity from you house wiringt, not just receptacles. Mark
the diagram with every light, switch, receptacle and hard-wired
appliance. Don't forget to mark the thermostat and doorbell
transformers - you'll need to search for them. You'll also need
to make a diagram that shows *ALL* outdoor outlets. With a
helper at the panel, go from room to room with your diagrams,
and mark the breaker number that kills the power to each and
every outlet. Use a 3 lite tester to test for power at the receptacles.
It is much easier and better than a multimeter for this task.

In the process, you should find something that does not work.
If you do, it is highly probable that it is connected to the wires
that are capped. When you are done with the diagrams,
put them on the PC and print them. Then make one more -
a diagram of your service panel showing *EVERYTHING*
that is connected to each breaker. Tape a clear sheet of acetate
on the panel and put the diagram inside. You don't need
a CAD program for this - anything that allows drawing rectangles
and adding text will work - even a word processor.

The next time you have a problem, there will be no guesswork
needed. It takes a few hours, but it is well worth the effort.

Ed

 

[Roy Q.T.]

GFCI Outlet Installation

Group: alt.engineering.electrical Date: Sun, Apr 3, 2005, 3:09pm From:
[email protected] (David)

I just installed a GFCI outlet in my bathroom. There are five wires that
connect to the outlet, 2 line, 2 load, and a ground. When I connect the
line wires to the line terminals and the ground wire to the ground
terminal, everything works fine. I get power. But when I then connect
the load wires to the load terminals, the reset switch keeps tripping
and the power keeps shutting off. I'm not sure why this is happening.
Any ideas? I don't think that there are any loads down the line from the
GFCI outlet, so I'm thinking that maybe I should just cap the load wires
and call it a day. Is this o.k.?
Thanks for the help.
Dave
----------->
You know,the same darned thing happened to me Frist time, but with a gfi
20A circuit breaker., I opened a plateded old fuse box, now a splice
through, I found I had 2 different circuits on the same ground (neutral
white wire) the wires are covered with threads but the rubber inside is
still fine....... So I made a run of 12awg white stranded for the new
circuit & left the piggyback on t's won circuit and isolated it from my
gfi Kithchen/Bathroom Circuit and the GFI Breaker finally ReSet.

on gfci receptacles i find it hard to connect the line to the load side
(the loadside has a deeper groove), but with the mix of travelers and
adjacent circuit cables, it could happen to anyone I think it's best to
spice pigtails to the gfci taps/screws and attach from there , this coud
be a time saver if you have 1 or 2 fed-through wires and are attaching
more than 1 load.

caped load/end circuits shouldn't affect gfci performance at all,
reversed condunctors? yes, somethings they're better off on the line
side, like the lighting circuit in the units....

oh well back to the books.... i don't worry any more :>) i just do it
}

OOP; shheee't still thinking about that stranded no.6 bare neutral to
ground panel screw-in-lug., no barred earth bonding };-o

®oy cet

 

[John Gilmer]

Those make contact with the

wires via spring tension. The newer GFCI's have a clamp in the holes,
which is tightened onto the wires when you tighten the screws.

Some months ago "someone" on this NG was touting a replacement for the
wirenut that to all appearances is equivalent to the "push in" you see on
cheap outlets.

I purchased a recessed ceiling fixture (new work type) that has these
already "in the box" but I haven't used it yet. I haven't seen them at
the local BIG BOX store but I haven't looked lately.

Problem is, of course, is when "they" start selling GOOD "push ins" many
folks will remember the NOT so good "push ins." Maybe that's why I don't
see them at HD.

 

[Takoma Park Volunteer Fire Department Postmaster]

Tom - look at a new GFCI. The new ones should NOT be wired
by wrapping the wire around the screws. The wires SHOULD be put
in the holes in the back, and the screws must be tightened. These are
not the "push-in" holes you are thinking of. Those make contact with the
wires via spring tension. The newer GFCI's have a clamp in the holes,
which is tightened onto the wires when you tighten the screws.

Most of the GFCI receptacles you are talking about are listed to be
wired either way. The Leviton rear clamping type is listed for either
termination method for example. If the device is not listed for side
wire by using the screws the screws would be tightly shrouded by the
plastic housing to make mis wiring difficult.

 

[David]

Interesting.. I live in a small 2 level townhouse. There are 3 breakers
(switches) labeled "lighting" in the breaker panel. One of those breakers
controls the downstairs lighting, and the other 2 control the upstairs
lighting. There are other breakers in the breaker panel labeled range, A/C,
water, dryer, etc., and one labeled "main" that shuts everything off. I
tested all of the upstairs receptacle and switches that are controlled by
the same breaker as the GFCI receptacle. They all tested normal. Could a
problem with a receptacle/switch controlled by one breaker cause a problem
with a receptacle/switch controlled by another breaker? If yes, how might
that happen? If possible, please explain it to me at the high school level.
I know the basics of series/parallel circuit design, but not much more.

Thanks for all of your help (everybody),

Dave

 

[w_tom]

One answer was posted previously. This is simple point to
point wiring. Nothing that requires an advanced education.
But translating simple diagrams from and to text requires
careful reading. Now that answer which but one example:

... the 'load' side neutral wire was wrapped together (wire
nutted) with a neutral wire from some other circuit. I have
seen amateur electricians do this. Two switches were to be
powered by different circuits. But the electrician wired all
neutrals [white wires] from both circuits together. When
asked why he did this, well, he always did this for years.
It was only when AGFIs were required by code that his
circuits had strange problems.

IOW be suspicious of any receptacle or switch boxes that
controls power from two different circuit breaker circuits.
Just another reason why the GFCI would see a failure and
complain.

Wire wires from different circuits connected together will
not trip conventional circuit breakers but will be detected by
GFCI units.

BTW, if you have determined the circuit breaker number for
each outlet, then simply write that circuit breaker number on
the inside surface of each cover plate. If one need ever turn
off power to fix a receptacle, then he knows immediately which
breaker to cut off.