What could cause this:

GFCI on a supposedly grounded circuit. Wires appear to be installed
correctly at the GFCI.

3 light tester shows Open Ground.
115V between hot and neutral
54V between hot and ground lug
46V between neutral and ground lug

Thoughts?

"DerbyDad03" wrote in message
...
What could cause this:

GFCI on a supposedly grounded circuit. Wires appear to be installed
correctly at the GFCI.

3 light tester shows Open Ground.
115V between hot and neutral
54V between hot and ground lug
46V between neutral and ground lug

Thoughts?

Sounds like a typical digital meter with the ground wire not attached at the
breaker box. Or at least some where between the GFCI and breaker box.

Check the ground at the breaker and also at any place it may connect in
between.

On 2/26/2012 1:54 PM, DerbyDad03 wrote:
What could cause this:

GFCI on a supposedly grounded circuit. Wires appear to be installed
correctly at the GFCI.

3 light tester shows Open Ground.
115V between hot and neutral
54V between hot and ground lug
46V between neutral and ground lug

Thoughts?

first thought is to **** can the digital and use an analog meter. Then
retest.


--
Steve Barker
remove the "not" from my address to email

On Feb 26, 4:04*pm, Steve Barker wrote:
On 2/26/2012 1:54 PM, DerbyDad03 wrote:

What could cause this:

GFCI on a supposedly grounded circuit. Wires appear to be installed
correctly at the GFCI.

3 light tester shows Open Ground.
115V between hot and neutral
54V between hot and ground lug
46V between neutral and ground lug

Thoughts?

first thought is to **** can the digital and use an analog meter. *Then
retest.

--
Steve Barker
remove the "not" from my address to email

....or I could test 3 other grounded circuits in the same house and see
what that same meter reads.

In all three cases the meter reads:

115V between hot and neutral
115V between hot and ground lug
0V between neutral and ground lug

I'm guessing the meter's fine and the problem is with the circuit.

"DerbyDad03" wrote in message
...
On Feb 26, 4:04 pm, Steve Barker wrote:

first thought is to **** can the digital and use an analog meter. Then
retest.

--
Steve Barker
remove the "not" from my address to email

...or I could test 3 other grounded circuits in the same house and see
what that same meter reads.

In all three cases the meter reads:

115V between hot and neutral
115V between hot and ground lug
0V between neutral and ground lug

I'm guessing the meter's fine and the problem is with the circuit.

The meter is fine, it is the meter operator.

Many digital meters will not load a circuit so that even the smallest
ammount of induced voltage will show up. Most do not understand this.
That is why you are seeing such oddball voltages on the circuit.

DerbyDad03 wrote:
What could cause this:

GFCI on a supposedly grounded circuit. Wires appear to be installed
correctly at the GFCI.

3 light tester shows Open Ground.
115V between hot and neutral
54V between hot and ground lug
46V between neutral and ground lug

Thoughts?
Hmmm,
Are you using DVM? No analog meter?
GND is missing.

Ralph Mowery wrote:
wrote in message
...
On Feb 26, 4:04 pm, Steve wrote:

first thought is to **** can the digital and use an analog meter. Then
retest.

--
Steve Barker
remove the "not" from my address to email

...or I could test 3 other grounded circuits in the same house and see
what that same meter reads.

In all three cases the meter reads:

115V between hot and neutral
115V between hot and ground lug
0V between neutral and ground lug

I'm guessing the meter's fine and the problem is with the circuit.

The meter is fine, it is the meter operator.

Many digital meters will not load a circuit so that even the smallest
ammount of induced voltage will show up. Most do not understand this.
That is why you are seeing such oddball voltages on the circuit.


Hi,
I always have Simpson 260 handy working on problem like this. Fluke is
for some other use.

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"
wrote:


"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

When wires are next to each other, and carrying AC (such as in Romex, and
under load). The power wire tends to put a tiny bit of voltage into the
nearby wire, acting as though it were a transformer. What you are reading
is probably that stray, induced voltage. it shows on your meter, because the
meter consumes nearly zero of the power in what's being read. If you had any
kind of load at all between H and G, the ghost voltage would disappear
instantly.

The voltage is also called "ghost reading" or some such name.

Christopher A. Young
Learn more about Jesus
www.lds.org
..

"DerbyDad03" wrote in message
...
What could cause this:

GFCI on a supposedly grounded circuit. Wires appear to be installed
correctly at the GFCI.

3 light tester shows Open Ground.
115V between hot and neutral
54V between hot and ground lug
46V between neutral and ground lug

Thoughts?

On Feb 26, 4:41*pm, "Ralph Mowery" wrote:
"DerbyDad03" wrote in message

...
On Feb 26, 4:04 pm, Steve Barker wrote:



first thought is to **** can the digital and use an analog meter. Then
retest.

--
Steve Barker
remove the "not" from my address to email
...or I could test 3 other grounded circuits in the same house and see
what that same meter reads.
In all three cases the meter reads:
115V between hot and neutral
115V between hot and ground lug
0V between neutral and ground lug
I'm guessing the meter's fine and the problem is with the circuit.

The meter is fine, it is the meter operator.

Many digital meters will not load a circuit so that even the smallest
ammount of induced voltage will show up. *Most do not understand this.
That is why you are seeing such oddball voltages on the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Still think it's operator error?

On Feb 26, 5:32*pm, "
wrote:
On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"









wrote:

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

On Sun, 26 Feb 2012 14:58:44 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 4:41*pm, "Ralph Mowery" wrote:
"DerbyDad03" wrote in message

...
On Feb 26, 4:04 pm, Steve Barker wrote:



first thought is to **** can the digital and use an analog meter. Then
retest.

--
Steve Barker
remove the "not" from my address to email
...or I could test 3 other grounded circuits in the same house and see
what that same meter reads.
In all three cases the meter reads:
115V between hot and neutral
115V between hot and ground lug
0V between neutral and ground lug
I'm guessing the meter's fine and the problem is with the circuit.

The meter is fine, it is the meter operator.

Many digital meters will not load a circuit so that even the smallest
ammount of induced voltage will show up. *Most do not understand this.
That is why you are seeing such oddball voltages on the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Still think it's operator error?

No, it sounds like the open ground your three-light tester told you it was.
Are there any other outlets on that circuit? Test them, too.

On Sun, 26 Feb 2012 14:59:42 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 5:32*pm, "
wrote:
On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"









wrote:

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Read my later reply, after you'd posted this information. Essentially, you do
have an open ground, like the TLT was telling you. Where it is in the next
question. Use the TLT on another outlet in this circuit. Testing both an
outlet before and one after will nail it down pretty well. If the one after
is fine, then there is a good chance that the fault is within the GFCI itself.

"DerbyDad03" wrote in message
...

--
Steve Barker
remove the "not" from my address to email
...or I could test 3 other grounded circuits in the same house and see
what that same meter reads.
In all three cases the meter reads:
115V between hot and neutral
115V between hot and ground lug
0V between neutral and ground lug
I'm guessing the meter's fine and the problem is with the circuit.

The meter is fine, it is the meter operator.

Many digital meters will not load a circuit so that even the smallest
ammount of induced voltage will show up. Most do not understand this.
That is why you are seeing such oddball voltages on the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Still think it's operator error?

Not so much as operator error, but a case where the operator does not
understand how it is possiable.

We (the others and myself) are just trying to educate you about how it is
possiable for a meter to show voltages like what you have. WE go through
this almost once a month here where someone has a digital meter and will see
the same thing.

I think several others mentioned a digital meter before you told us what you
were using.

On Feb 26, 6:45*pm, "
wrote:
On Sun, 26 Feb 2012 14:59:42 -0800 (PST), DerbyDad03
wrote:









On Feb 26, 5:32*pm, "
wrote:
On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"

wrote:

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Read my later reply, after you'd posted this information. *Essentially, you do
have an open ground, like the TLT was telling you. *Where it is in the next
question. *Use the TLT on another outlet in this circuit. *Testing both an
outlet before and one after will nail it down pretty well. *If the one after
is fine, then there is a good chance that the fault is within the GFCI itself.

Well, at least we've stopped talking about the meter and/or the
operator, which so many wanted to blame as the cause.

"Testing both an outlet before and one after will nail it down
pretty well."

Can't do that. There is no receptacle before the GFCI and everything
after it is ungrounded. The GFCI was installed so that 3 pronged
receptacles could be installed downstream. Obviously, they'll all read
"open ground".

However, here's the latest:

I opened the junction box just before the GFCI and disconnected the
line in Romex feed, which had a ground wire. I then ran a suicide cord
from a known good receptacle in my shop to the Romex going to the
GFCI. The GFCI now reads correctly when using the tester and there are
no strange readings between the current carriers and ground. Same
meter, same operator. ;-)

I traced the original Romex back to another junction box which houses
a light fixture over the utility tub. That box has 2 other wires
associated with it, one obviously being the feed but the other one
disappears up into a wall. I'll need to open that box to see if
there's a ground in there, otherwise I'll be pulling a new wire from
the panel to the GFCI's junction box.

Too late to play now, SWMBO is cooking and killing that circuit kills
the kitchen lights too.

On Feb 26, 7:19*pm, "Ralph Mowery" wrote:
"DerbyDad03" wrote in message

...









--
Steve Barker
remove the "not" from my address to email
...or I could test 3 other grounded circuits in the same house and see
what that same meter reads.
In all three cases the meter reads:
115V between hot and neutral
115V between hot and ground lug
0V between neutral and ground lug
I'm guessing the meter's fine and the problem is with the circuit.

The meter is fine, it is the meter operator.

Many digital meters will not load a circuit so that even the smallest
ammount of induced voltage will show up. Most do not understand this.
That is why you are seeing such oddball voltages on the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Still think it's operator error?

Not so much as operator error, but a case where the operator does not
understand how it is possiable.

We (the others and myself) are just trying to educate you about how it is
possiable for a meter to show voltages like what you have. *WE go through
this almost once a month here where someone has a digital meter and will see
the same thing.

I think several others mentioned a digital meter before you told us what you
were using.

So are you saying that with an open ground an analog meter would not
show any reading between the hot or neutral and the ground prong at
the receptacle?

On Sun, 26 Feb 2012 11:54:58 -0800 (PST), DerbyDad03
wrote:

What could cause this:

GFCI on a supposedly grounded circuit. Wires appear to be installed
correctly at the GFCI.

3 light tester shows Open Ground.
115V between hot and neutral
54V between hot and ground lug
46V between neutral and ground lug

Thoughts?

Get an analog meter, with a needle, and use that.

It has other advantages too. You can watch a capaciitor charge from
the battery in the ohmmeter and see it accept current more slowly as
it charges.
You can with all but the
smallest caps tell if it's good or not, IMO, excpet for whether it is
no good at higher voltages than the meter battery.

I think I read that the cheap harbor freight digiital meters only have
1Meg resistance, or 1 meg per volt conceivabley, when meaasureing
voltage, but even that is a lot higher than most analog meters at
50K/v iirc.

On Feb 26, 7:19*pm, "Ralph Mowery" wrote:
"DerbyDad03" wrote in message

...





--
Steve Barker
remove the "not" from my address to email
...or I could test 3 other grounded circuits in the same house and see
what that same meter reads.
In all three cases the meter reads:
115V between hot and neutral
115V between hot and ground lug
0V between neutral and ground lug
I'm guessing the meter's fine and the problem is with the circuit.

The meter is fine, it is the meter operator.

Many digital meters will not load a circuit so that even the smallest
ammount of induced voltage will show up. Most do not understand this.
That is why you are seeing such oddball voltages on the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Still think it's operator error?

Not so much as operator error, but a case where the operator does not
understand how it is possiable.

We (the others and myself) are just trying to educate you about how it is
possiable for a meter to show voltages like what you have. *WE go through
this almost once a month here where someone has a digital meter and will see
the same thing.

I think several others mentioned a digital meter before you told us what you
were using.- Hide quoted text -

- Show quoted text -

Plugging a light bulb into the circuit doesn't change
anything because it doesn't introduce any ground current.
It's just flowing current between hot and neutral
and you already know that voltage is normal.
In a properly grounded circuit, you could hook a light bulb
between hot and ground and it should light. If
you try that in the problem circuit, my bet is that
it will not light, because you have an open ground
somewhere along the circuit.

The circuit tester showed there was no ground.
With the high impedance meter whenyou connect
it between hot and an unconnected ground wire,
you're going to get a reading of some in
between voltage because whatever ground
wire is there has capacitance, inductance, etc.

On Sun, 26 Feb 2012 13:31:48 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 4:04*pm, Steve Barker wrote:
On 2/26/2012 1:54 PM, DerbyDad03 wrote:

What could cause this:

GFCI on a supposedly grounded circuit. Wires appear to be installed
correctly at the GFCI.

3 light tester shows Open Ground.
115V between hot and neutral
54V between hot and ground lug
46V between neutral and ground lug

Thoughts?

first thought is to **** can the digital and use an analog meter. *Then
retest.

--
Steve Barker
remove the "not" from my address to email

...or I could test 3 other grounded circuits in the same house and see
what that same meter reads.

In all three cases the meter reads:

115V between hot and neutral
115V between hot and ground lug
0V between neutral and ground lug

I'm guessing the meter's fine and the problem is with the circuit.


Steve did seem to say the meter was bad, but my first reply to you
doesn't say that. It's just not suitable for this applicatoin,
perhaps. I don't think the other receptacles prove much. Induced
voltages depend on what is nearby to induce them. .

On Sun, 26 Feb 2012 11:54:58 -0800 (PST), DerbyDad03
wrote:

What could cause this:

GFCI on a supposedly grounded circuit. Wires appear to be installed
correctly at the GFCI.

3 light tester shows Open Ground.

It doesn't change my other posts much, but I must admit I didnt'
notice this. Duh. Why not fix the open ground and then measure
again.

115V between hot and neutral
54V between hot and ground lug
46V between neutral and ground lug

Thoughts?

On Sun, 26 Feb 2012 14:58:44 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 4:41*pm, "Ralph Mowery" wrote:
"DerbyDad03" wrote in message

...
On Feb 26, 4:04 pm, Steve Barker wrote:



first thought is to **** can the digital and use an analog meter. Then
retest.

--
Steve Barker
remove the "not" from my address to email
...or I could test 3 other grounded circuits in the same house and see
what that same meter reads.
In all three cases the meter reads:
115V between hot and neutral
115V between hot and ground lug
0V between neutral and ground lug
I'm guessing the meter's fine and the problem is with the circuit.

The meter is fine, it is the meter operator.

Many digital meters will not load a circuit so that even the smallest
ammount of induced voltage will show up. *Most do not understand this.
That is why you are seeing such oddball voltages on the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

I don't think the stray voltage we talked about is affected by turning
on a lamp. The lamp draws only from the hot wire and normally returns
only though the neutral . You have a problem with the ground.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Still think it's operator error?

DerbyDad03 wrote:
On Feb 26, 7:19 pm, "Ralph wrote:
wrote in message

...









--
Steve Barker
remove the "not" from my address to email
...or I could test 3 other grounded circuits in the same house and see
what that same meter reads.
In all three cases the meter reads:
115V between hot and neutral
115V between hot and ground lug
0V between neutral and ground lug
I'm guessing the meter's fine and the problem is with the circuit.

The meter is fine, it is the meter operator.

Many digital meters will not load a circuit so that even the smallest
ammount of induced voltage will show up. Most do not understand this.
That is why you are seeing such oddball voltages on the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Still think it's operator error?

Not so much as operator error, but a case where the operator does not
understand how it is possiable.

We (the others and myself) are just trying to educate you about how it is
possiable for a meter to show voltages like what you have. WE go through
this almost once a month here where someone has a digital meter and will see
the same thing.

I think several others mentioned a digital meter before you told us what you
were using.

So are you saying that with an open ground an analog meter would not
show any reading between the hot or neutral and the ground prong at
the receptacle?

Hi,
B4 you ask that question, do you understand the difference between high
impedance and low impedance measuring instrument and inductive coupling?
Low impedance ones will load the circuit down so ghost voltages won't
show vs. high impedance meter. You have to know what to expect with your
meter at certain test condition lest you should get carried away
chasing the ghost. Do an experiment. With your meter set in voltage
reading, hold test leads with your fingers, most likely you will see
some reading. Another experiment, put a light bulb across hot and ground
lead which showed voltage reading, does the bulb come on? at 57V, it
should burn at least half brightness. That voltage was bogus(ghost) value.

"DerbyDad03" wrote in message
...
So are you saying that with an open ground an analog meter would not
show any reading between the hot or neutral and the ground prong at
the receptacle?


Yes.

It may show a voltage, but it will be a lot lower if it does.

YOu may think of what is going on similar to a car battery. It will start
the car. If you have 8 D cell batteries in series for 12 volts and try to
start the car, it will not. The D cells can not supply the current for the
load.

The induced voltage or whatever people want to call it is at a very low
current. Most digital meters will not have much of a load and will show a
higher voltage than the analog meters.
If you put a digital meter and analog meter on the same wires at the same
time, they will both show the same voltage. Take the digital meter off and
the analog meter will hardly show any more voltage. Put both back on and
then take the analog meter off and the digital meter will show a much higher
voltage.

On Sun, 26 Feb 2012 16:30:14 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 6:45*pm, "
wrote:
On Sun, 26 Feb 2012 14:59:42 -0800 (PST), DerbyDad03
wrote:









On Feb 26, 5:32*pm, "
wrote:
On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"

wrote:

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Read my later reply, after you'd posted this information. *Essentially, you do
have an open ground, like the TLT was telling you. *Where it is in the next
question. *Use the TLT on another outlet in this circuit. *Testing both an
outlet before and one after will nail it down pretty well. *If the one after
is fine, then there is a good chance that the fault is within the GFCI itself.

Well, at least we've stopped talking about the meter and/or the
operator, which so many wanted to blame as the cause.

We? Well, the reason the meter is reading what it is, is because of its high
impedance.

"Testing both an outlet before and one after will nail it down
pretty well."

Can't do that. There is no receptacle before the GFCI and everything
after it is ungrounded. The GFCI was installed so that 3 pronged
receptacles could be installed downstream. Obviously, they'll all read
"open ground".

I guess your only option is to shut off the circuit and pull the GFCI. The
easiest thing to do is to measure the continuity of the outlet itself from the
ground hole contacts to the screw. You can probably do that without removing
the outlet from the box. You should also be able to inspect the connection
without pulling it. Else, you're going to have to pull it.

BTW, since there are no grounds in this house, how did the GFCI get one? You
do know that it really doesn't need one, in this case.


However, here's the latest:

I opened the junction box just before the GFCI and disconnected the
line in Romex feed, which had a ground wire. I then ran a suicide cord
from a known good receptacle in my shop to the Romex going to the
GFCI. The GFCI now reads correctly when using the tester and there are
no strange readings between the current carriers and ground. Same
meter, same operator. ;-)

Oops. That means it's a broken wire - much work.

I traced the original Romex back to another junction box which houses
a light fixture over the utility tub. That box has 2 other wires
associated with it, one obviously being the feed but the other one
disappears up into a wall. I'll need to open that box to see if
there's a ground in there, otherwise I'll be pulling a new wire from
the panel to the GFCI's junction box.

Again, why bother?

Too late to play now, SWMBO is cooking and killing that circuit kills
the kitchen lights too.

Dinner here was ruined (priority interrupt from real estate agent).

On Feb 26, 4:30*pm, DerbyDad03 wrote:
On Feb 26, 6:45*pm, "









wrote:
On Sun, 26 Feb 2012 14:59:42 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 5:32*pm, "
wrote:
On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"

wrote:

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Read my later reply, after you'd posted this information. *Essentially, you do
have an open ground, like the TLT was telling you. *Where it is in the next
question. *Use the TLT on another outlet in this circuit. *Testing both an
outlet before and one after will nail it down pretty well. *If the one after
is fine, then there is a good chance that the fault is within the GFCI itself.

Well, at least we've stopped talking about the meter and/or the
operator, which so many wanted to blame as the cause.

"Testing both an outlet before and one after will nail it down
pretty well."

Can't do that. There is no receptacle before the GFCI and everything
after it is ungrounded. The GFCI was installed so that 3 pronged
receptacles could be installed downstream. Obviously, they'll all read
"open ground".

However, here's the latest:

I opened the junction box just before the GFCI and disconnected the
line in Romex feed, which had a ground wire. I then ran a suicide cord
from a known good receptacle in my shop to the Romex going to the
GFCI. The GFCI now reads correctly when using the tester and there are
no strange readings between the current carriers and ground. Same
meter, same operator. ;-)

I traced the original Romex back to another junction box which houses
a light fixture over the utility tub. That box has 2 other wires
associated with it, one obviously being the feed but the other one
disappears up into a wall. I'll need to open that box to see if
there's a ground in there, otherwise I'll be pulling a new wire from
the panel to the GFCI's junction box.

Too late to play now, SWMBO is cooking and killing that circuit kills
the kitchen lights too.

You're homing in on the solution.

My best guess..... loose connection between panel & GFI box.

btw those three light testers can be fooled by some unusual conditions

here's a link to a website

http://www.thecircuitdetective.com/test.htm

Scroll down about half way & there is a detailed discussion of the 3-
prong plug in tester.

These plug-in testers are handy but not perfect, they can give
misleading results in some situations.

cheers
Bob

On Feb 27, 2:57*am, DD_BobK wrote:
On Feb 26, 4:30*pm, DerbyDad03 wrote:





On Feb 26, 6:45*pm, "

wrote:
On Sun, 26 Feb 2012 14:59:42 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 5:32*pm, "
wrote:
On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"

wrote:

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Read my later reply, after you'd posted this information. *Essentially, you do
have an open ground, like the TLT was telling you. *Where it is in the next
question. *Use the TLT on another outlet in this circuit. *Testing both an
outlet before and one after will nail it down pretty well. *If the one after
is fine, then there is a good chance that the fault is within the GFCI itself.

Well, at least we've stopped talking about the meter and/or the
operator, which so many wanted to blame as the cause.

"Testing both an outlet before and one after will nail it down
pretty well."

Can't do that. There is no receptacle before the GFCI and everything
after it is ungrounded. The GFCI was installed so that 3 pronged
receptacles could be installed downstream. Obviously, they'll all read
"open ground".

However, here's the latest:

I opened the junction box just before the GFCI and disconnected the
line in Romex feed, which had a ground wire. I then ran a suicide cord
from a known good receptacle in my shop to the Romex going to the
GFCI. The GFCI now reads correctly when using the tester and there are
no strange readings between the current carriers and ground. Same
meter, same operator. ;-)

I traced the original Romex back to another junction box which houses
a light fixture over the utility tub. That box has 2 other wires
associated with it, one obviously being the feed but the other one
disappears up into a wall. I'll need to open that box to see if
there's a ground in there, otherwise I'll be pulling a new wire from
the panel to the GFCI's junction box.

Too late to play now, SWMBO is cooking and killing that circuit kills
the kitchen lights too.

You're homing in on the solution.

My best guess..... *loose connection between panel & GFI box.

btw those three light testers can be fooled by some unusual conditions

here's a link to a website

http://www.thecircuitdetective.com/test.htm

Scroll down about half way & there is a detailed discussion of the 3-
prong plug in tester.

These plug-in testers are handy but not perfect, they can give
misleading results in some situations.

cheers
Bob- Hide quoted text -

- Show quoted text -

There's also a minor, but interesting issue with those testers.

With the latest method of installing 3 prong receptacles with the
ground prong up, the testers (at least mine) are manufactured upside
down.

When you plug one into a "ground prong up" receptacle, you can't read
the chart and have to remember to reverse the light indications since
they are "backwards" when the tester is plugged in.

On Feb 26, 10:53*pm, "
wrote:
On Sun, 26 Feb 2012 16:30:14 -0800 (PST), DerbyDad03
wrote:





On Feb 26, 6:45*pm, "
wrote:
On Sun, 26 Feb 2012 14:59:42 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 5:32*pm, "
wrote:
On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"

wrote:

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground..

Fluke 75 multimeter.

Read my later reply, after you'd posted this information. *Essentially, you do
have an open ground, like the TLT was telling you. *Where it is in the next
question. *Use the TLT on another outlet in this circuit. *Testing both an
outlet before and one after will nail it down pretty well. *If the one after
is fine, then there is a good chance that the fault is within the GFCI itself.

Well, at least we've stopped talking about the meter and/or the
operator, which so many wanted to blame as the cause.

We? *Well, the reason the meter is reading what it is, is because of its high
impedance.

"Testing both an outlet before and one after will nail it down
pretty well."

Can't do that. There is no receptacle before the GFCI and everything
after it is ungrounded. The GFCI was installed so that 3 pronged
receptacles could be installed downstream. Obviously, they'll all read
"open ground".

I guess your only option is to shut off the circuit and pull the GFCI. *The
easiest thing to do is to measure the continuity of the outlet itself from the
ground hole contacts to the screw. *You can probably do that without removing
the outlet from the box. *You should also be able to inspect the connection
without pulling it. *Else, you're going to have to pull it.

BTW, since there are no grounds in this house, how did the GFCI get one? *You
do know that it really doesn't need one, in this case.

However, here's the latest:

I opened the junction box just before the GFCI and disconnected the
line in Romex feed, which had a ground wire. I then ran a suicide cord
from a known good receptacle in my shop to the Romex going to the
GFCI. The GFCI now reads correctly when using the tester and there are
no strange readings between the current carriers and ground. Same
meter, same operator. ;-)

Oops. *That means it's a broken wire - much work.

I traced the original Romex back to another junction box which houses
a light fixture over the utility tub. That box has 2 other wires
associated with it, one obviously being the feed but the other one
disappears up into a wall. I'll need to open that box to see if
there's a ground in there, otherwise I'll be pulling a new wire from
the panel to the GFCI's junction box.

Again, why bother?

Too late to play now, SWMBO is cooking and killing that circuit kills
the kitchen lights too.

Dinner here was ruined (priority interrupt from real estate agent).- Hide quoted text -

- Show quoted text -

In all honesty, I had already pulled the GFCI just to eliminate it as
a factor before I ran my suicide cord. I replaced it with a 3 prong
receptacle and disconnected the downstream devices so that there was
only the single 3 prong receptacle on the run.

When it still showed the open ground - even though there was a ground
wire attached - that's when I started tracing back and found the
prvious junction box. I'll be opening that tonight to look at the
ground situation.

As far as pulling a new grounded run from the panel to the junction
box, you asked "Why bother?"

Well, this is the last room in the house without 3 prong grounded
receptacles. That's why I started by adding the GFCI - so I could at
least install 3 prong receptacles downstream, mainly for the
convenience of having 3 prong receptacles available. Eventually, I
want to run grounded romex to all of the downstream receptacles, but
that is going to require moving some storage in the basement so it's
going to take some time.

Since the feed to the junction box has a ground wire, it appeared that
things would flow fairly smoothly in the forward direction from the
junction box on. Now that I'm showing an open ground, I have to go
backward and at a minimum find out why. I'm just not comfortable
knowing that a non-grounded ground wire is attached in the junction
box and at the GFCI. It'll gnaw at me until I fix it.

On 2/26/2012 4:27 PM, Ralph Mowery wrote:
"Tony wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

This can be rather bizarre the first time you see it.


I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Fluke has a small gizmo (SV225) that plugs into the meter jacks, then
the meter leads plug into it. It adds a relatively low resistance across
the leads to eliminate phantom voltage.

http://www.fluke.com/Fluke/usen/Acce....htm?PID=56696

--
bud--

On Feb 27, 10:29*am, bud-- wrote:
On 2/26/2012 4:27 PM, Ralph Mowery wrote:

"Tony *wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

This can be rather bizarre the first time you see it.



I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Fluke has a small gizmo (SV225) that plugs into the meter jacks, then
the meter leads plug into it. It adds a relatively low resistance across
the leads to eliminate phantom voltage.

http://www.fluke.com/Fluke/usen/Acce...Chargers-and-A...

--
bud--

The site says:

"It can be used with all modern meters with standard input spacing."

I wonder if their own Model 75 series is considered "modern" and/or
"standard". It would have been nice if they followed that statement
with an actual measurement so that readers could easily determine if
the device was compatible with the meter they own.

BTW...does that statement mean that it can't be used with "old" meters
with standard input spacing? ;-)

On 2/27/2012 10:38 AM, DerbyDad03 wrote:
On Feb 27, 10:29 am, wrote:
On 2/26/2012 4:27 PM, Ralph Mowery wrote:

"Tony wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

This can be rather bizarre the first time you see it.



I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Fluke has a small gizmo (SV225) that plugs into the meter jacks, then
the meter leads plug into it. It adds a relatively low resistance across
the leads to eliminate phantom voltage.

http://www.fluke.com/Fluke/usen/Acce...Chargers-and-A...

--
bud--

The site says:

"It can be used with all modern meters with standard input spacing."

I wonder if their own Model 75 series is considered "modern" and/or
"standard". It would have been nice if they followed that statement
with an actual measurement so that readers could easily determine if
the device was compatible with the meter they own.

BTW...does that statement mean that it can't be used with "old" meters
with standard input spacing? ;-)

My Fluke has the jacks on 3/4" centers. I would call the probe plugs
shielded banana plugs (plastic cylinder around the plug).

I assume the Fluke adampter can be used with any meter with 3/4" spaced
shielded banana plugs. Or I would guess you could cut off the insulating
shield and use it with 3/4" spaced banana jacks. My Simpson 260 has
jacks on 5/8" centers - wouldn't work.

All measuring tools have limitations. Part of the challenge of making
measurements is knowing what the limitations are.


For something like you are doing I would commonly use a pig-tail socket
and a 40W bulb as a 'voltmeter'. I use a neon test light a lot more than
I use a meter.

--
bud--

On Mon, 27 Feb 2012 06:27:01 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 10:53*pm, "
wrote:
On Sun, 26 Feb 2012 16:30:14 -0800 (PST), DerbyDad03
wrote:





On Feb 26, 6:45*pm, "
wrote:
On Sun, 26 Feb 2012 14:59:42 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 5:32*pm, "
wrote:
On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"

wrote:

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Read my later reply, after you'd posted this information. *Essentially, you do
have an open ground, like the TLT was telling you. *Where it is in the next
question. *Use the TLT on another outlet in this circuit. *Testing both an
outlet before and one after will nail it down pretty well. *If the one after
is fine, then there is a good chance that the fault is within the GFCI itself.

Well, at least we've stopped talking about the meter and/or the
operator, which so many wanted to blame as the cause.

We? *Well, the reason the meter is reading what it is, is because of its high
impedance.

"Testing both an outlet before and one after will nail it down
pretty well."

Can't do that. There is no receptacle before the GFCI and everything
after it is ungrounded. The GFCI was installed so that 3 pronged
receptacles could be installed downstream. Obviously, they'll all read
"open ground".

I guess your only option is to shut off the circuit and pull the GFCI. *The
easiest thing to do is to measure the continuity of the outlet itself from the
ground hole contacts to the screw. *You can probably do that without removing
the outlet from the box. *You should also be able to inspect the connection
without pulling it. *Else, you're going to have to pull it.

BTW, since there are no grounds in this house, how did the GFCI get one? *You
do know that it really doesn't need one, in this case.

However, here's the latest:

I opened the junction box just before the GFCI and disconnected the
line in Romex feed, which had a ground wire. I then ran a suicide cord
from a known good receptacle in my shop to the Romex going to the
GFCI. The GFCI now reads correctly when using the tester and there are
no strange readings between the current carriers and ground. Same
meter, same operator. ;-)

Oops. *That means it's a broken wire - much work.

I traced the original Romex back to another junction box which houses
a light fixture over the utility tub. That box has 2 other wires
associated with it, one obviously being the feed but the other one
disappears up into a wall. I'll need to open that box to see if
there's a ground in there, otherwise I'll be pulling a new wire from
the panel to the GFCI's junction box.

Again, why bother?

Too late to play now, SWMBO is cooking and killing that circuit kills
the kitchen lights too.

Dinner here was ruined (priority interrupt from real estate agent).- Hide quoted text -

- Show quoted text -

In all honesty, I had already pulled the GFCI just to eliminate it as
a factor before I ran my suicide cord. I replaced it with a 3 prong
receptacle and disconnected the downstream devices so that there was
only the single 3 prong receptacle on the run.

When it still showed the open ground - even though there was a ground
wire attached - that's when I started tracing back and found the
prvious junction box. I'll be opening that tonight to look at the
ground situation.

As far as pulling a new grounded run from the panel to the junction
box, you asked "Why bother?"

Well, this is the last room in the house without 3 prong grounded
receptacles. That's why I started by adding the GFCI - so I could at
least install 3 prong receptacles downstream, mainly for the
convenience of having 3 prong receptacles available. Eventually, I
want to run grounded romex to all of the downstream receptacles, but
that is going to require moving some storage in the basement so it's
going to take some time.

My point is that, in your case, you don't need a ground to install 3-prong
receptacles. The GFCI does the job. The GFCI doesn't need a ground, either.
Everything should be good as it is. Well, except the ungrounded 3-prong
outlets have to be marked "ungrounded".

Since the feed to the junction box has a ground wire, it appeared that
things would flow fairly smoothly in the forward direction from the
junction box on. Now that I'm showing an open ground, I have to go
backward and at a minimum find out why. I'm just not comfortable
knowing that a non-grounded ground wire is attached in the junction
box and at the GFCI. It'll gnaw at me until I fix it.

OK, there is that. There's more than one reason I wouldn't buy a house with
no grounds. ;-)

On Feb 27, 9:27*am, DerbyDad03 wrote:
On Feb 26, 10:53*pm, "





wrote:
On Sun, 26 Feb 2012 16:30:14 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 6:45*pm, "
wrote:
On Sun, 26 Feb 2012 14:59:42 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 5:32*pm, "
wrote:
On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"

wrote:

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Read my later reply, after you'd posted this information. *Essentially, you do
have an open ground, like the TLT was telling you. *Where it is in the next
question. *Use the TLT on another outlet in this circuit. *Testing both an
outlet before and one after will nail it down pretty well. *If the one after
is fine, then there is a good chance that the fault is within the GFCI itself.

Well, at least we've stopped talking about the meter and/or the
operator, which so many wanted to blame as the cause.

We? *Well, the reason the meter is reading what it is, is because of its high
impedance.

"Testing both an outlet before and one after will nail it down
pretty well."

Can't do that. There is no receptacle before the GFCI and everything
after it is ungrounded. The GFCI was installed so that 3 pronged
receptacles could be installed downstream. Obviously, they'll all read
"open ground".

I guess your only option is to shut off the circuit and pull the GFCI. *The
easiest thing to do is to measure the continuity of the outlet itself from the
ground hole contacts to the screw. *You can probably do that without removing
the outlet from the box. *You should also be able to inspect the connection
without pulling it. *Else, you're going to have to pull it.

BTW, since there are no grounds in this house, how did the GFCI get one? *You
do know that it really doesn't need one, in this case.

However, here's the latest:

I opened the junction box just before the GFCI and disconnected the
line in Romex feed, which had a ground wire. I then ran a suicide cord
from a known good receptacle in my shop to the Romex going to the
GFCI. The GFCI now reads correctly when using the tester and there are
no strange readings between the current carriers and ground. Same
meter, same operator. ;-)

Oops. *That means it's a broken wire - much work.

I traced the original Romex back to another junction box which houses
a light fixture over the utility tub. That box has 2 other wires
associated with it, one obviously being the feed but the other one
disappears up into a wall. I'll need to open that box to see if
there's a ground in there, otherwise I'll be pulling a new wire from
the panel to the GFCI's junction box.

Again, why bother?

Too late to play now, SWMBO is cooking and killing that circuit kills
the kitchen lights too.

Dinner here was ruined (priority interrupt from real estate agent).- Hide quoted text -

- Show quoted text -

In all honesty, I had already pulled the GFCI just to eliminate it as
a factor before I ran my suicide cord. I replaced it with a 3 prong
receptacle and disconnected the downstream devices so that there was
only the single 3 prong receptacle on the run.

When it still showed the open ground - even though there was a ground
wire attached - that's when I started tracing back and found the
prvious junction box.

You could just take a light bulb and temporarily
connect it between hot and ground as a simple
test. Open ground, it's not going to light. No need
to remove or change anything.

"DD_BobK" wrote in message
news:8e96f96c-2f03-468c-b2fd-

btw those three light testers can be fooled by some unusual conditions

here's a link to a website

http://www.thecircuitdetective.com/test.htm

Scroll down about half way & there is a detailed discussion of the 3-
prong plug in tester.

That's a great resource! Thanks for posting it. Especially useful were the
possible causes of "dim" readings which I've seen once or twice.

I saved a couple of thousand on this house just by going around to every
outlet with a 3 pronged tester and discovering that the seller's son had
replaced most of the two prong outlets with grounded outlets without any
connection to ground!!!! (Except for his room in the basement, where he had
run a ground to act as a neutral to install a programmble light switch.)
Didn't discover that until after the sale when I pulled down the ceiling.
The older the house, the more likely someone's done something stupid. I
wouldn't buy a house without using one to test as many outlets as I could
possibly test. I discovered back then, too, that in a really old house with
very few outlets, one that has nothing plugged into or that has been painted
over is very likely to be bad.

--
Bobby G.

On Feb 27, 1:40*pm, "
wrote:
On Mon, 27 Feb 2012 06:27:01 -0800 (PST), DerbyDad03
wrote:





On Feb 26, 10:53*pm, "
wrote:
On Sun, 26 Feb 2012 16:30:14 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 6:45*pm, "
wrote:
On Sun, 26 Feb 2012 14:59:42 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 5:32*pm, "
wrote:
On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"

wrote:

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Read my later reply, after you'd posted this information. *Essentially, you do
have an open ground, like the TLT was telling you. *Where it is in the next
question. *Use the TLT on another outlet in this circuit. *Testing both an
outlet before and one after will nail it down pretty well. *If the one after
is fine, then there is a good chance that the fault is within the GFCI itself.

Well, at least we've stopped talking about the meter and/or the
operator, which so many wanted to blame as the cause.

We? *Well, the reason the meter is reading what it is, is because of its high
impedance.

"Testing both an outlet before and one after will nail it down
pretty well."

Can't do that. There is no receptacle before the GFCI and everything
after it is ungrounded. The GFCI was installed so that 3 pronged
receptacles could be installed downstream. Obviously, they'll all read
"open ground".

I guess your only option is to shut off the circuit and pull the GFCI. *The
easiest thing to do is to measure the continuity of the outlet itself from the
ground hole contacts to the screw. *You can probably do that without removing
the outlet from the box. *You should also be able to inspect the connection
without pulling it. *Else, you're going to have to pull it.

BTW, since there are no grounds in this house, how did the GFCI get one? *You
do know that it really doesn't need one, in this case.

However, here's the latest:

I opened the junction box just before the GFCI and disconnected the
line in Romex feed, which had a ground wire. I then ran a suicide cord
from a known good receptacle in my shop to the Romex going to the
GFCI. The GFCI now reads correctly when using the tester and there are
no strange readings between the current carriers and ground. Same
meter, same operator. ;-)

Oops. *That means it's a broken wire - much work.

I traced the original Romex back to another junction box which houses
a light fixture over the utility tub. That box has 2 other wires
associated with it, one obviously being the feed but the other one
disappears up into a wall. I'll need to open that box to see if
there's a ground in there, otherwise I'll be pulling a new wire from
the panel to the GFCI's junction box.

Again, why bother?

Too late to play now, SWMBO is cooking and killing that circuit kills
the kitchen lights too.

Dinner here was ruined (priority interrupt from real estate agent).- Hide quoted text -

- Show quoted text -

In all honesty, I had already pulled the GFCI just to eliminate it as
a factor before I ran my suicide cord. I replaced it with a 3 prong
receptacle and disconnected the downstream devices so that there was
only the single 3 prong receptacle on the run.

When it still showed the open ground - even though there was a ground
wire attached - that's when I started tracing back and found the
prvious junction box. I'll be opening that tonight to look at the
ground situation.

As far as pulling a new grounded run from the panel to the junction
box, you asked "Why bother?"

Well, this is the last room in the house without 3 prong grounded
receptacles. That's why I started by adding the GFCI - so I could at
least install 3 prong receptacles downstream, mainly for the
convenience of having 3 prong receptacles available. Eventually, I
want to run grounded romex to all of the downstream receptacles, but
that is going to require moving some storage in the basement so it's
going to take some time.

My point is that, in your case, you don't need a ground to install 3-prong
receptacles. *The GFCI does the job. *The GFCI doesn't need a ground, either.
Everything should be good as it is. *Well, except the ungrounded 3-prong
outlets have to be marked "ungrounded".

Since the feed to the junction box has a ground wire, it appeared that
things would flow fairly smoothly in the forward direction from the
junction box on. Now that I'm showing an open ground, I have to go
backward and at a minimum find out why. I'm just not comfortable
knowing that a non-grounded ground wire is attached in the junction
box and at the GFCI. It'll gnaw at me until I fix it.

OK, there is that. *There's more than one reason I wouldn't buy a house with
no grounds. *;-)- Hide quoted text -

- Show quoted text -

I know how a GFCI works with regards to the grounds. I've posted on
that very subject in this forum many times.

At this point it is really just the presence of ground wires that
don't do anything that irks me.

In the long run (no pun intended) I do plan to rewire the circuit to
provide the equipment ground to the downstream receptacles but the
GFCI will suffice for now.

What's that old saying?

It's only temporary...unless it works.

On Feb 27, 1:40*pm, "
wrote:
On Mon, 27 Feb 2012 06:27:01 -0800 (PST), DerbyDad03
wrote:





On Feb 26, 10:53*pm, "
wrote:
On Sun, 26 Feb 2012 16:30:14 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 6:45*pm, "
wrote:
On Sun, 26 Feb 2012 14:59:42 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 5:32*pm, "
wrote:
On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"

wrote:

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Read my later reply, after you'd posted this information. *Essentially, you do
have an open ground, like the TLT was telling you. *Where it is in the next
question. *Use the TLT on another outlet in this circuit. *Testing both an
outlet before and one after will nail it down pretty well. *If the one after
is fine, then there is a good chance that the fault is within the GFCI itself.

Well, at least we've stopped talking about the meter and/or the
operator, which so many wanted to blame as the cause.

We? *Well, the reason the meter is reading what it is, is because of its high
impedance.

"Testing both an outlet before and one after will nail it down
pretty well."

Can't do that. There is no receptacle before the GFCI and everything
after it is ungrounded. The GFCI was installed so that 3 pronged
receptacles could be installed downstream. Obviously, they'll all read
"open ground".

I guess your only option is to shut off the circuit and pull the GFCI. *The
easiest thing to do is to measure the continuity of the outlet itself from the
ground hole contacts to the screw. *You can probably do that without removing
the outlet from the box. *You should also be able to inspect the connection
without pulling it. *Else, you're going to have to pull it.

BTW, since there are no grounds in this house, how did the GFCI get one? *You
do know that it really doesn't need one, in this case.

However, here's the latest:

I opened the junction box just before the GFCI and disconnected the
line in Romex feed, which had a ground wire. I then ran a suicide cord
from a known good receptacle in my shop to the Romex going to the
GFCI. The GFCI now reads correctly when using the tester and there are
no strange readings between the current carriers and ground. Same
meter, same operator. ;-)

Oops. *That means it's a broken wire - much work.

I traced the original Romex back to another junction box which houses
a light fixture over the utility tub. That box has 2 other wires
associated with it, one obviously being the feed but the other one
disappears up into a wall. I'll need to open that box to see if
there's a ground in there, otherwise I'll be pulling a new wire from
the panel to the GFCI's junction box.

Again, why bother?

Too late to play now, SWMBO is cooking and killing that circuit kills
the kitchen lights too.

Dinner here was ruined (priority interrupt from real estate agent).- Hide quoted text -

- Show quoted text -

In all honesty, I had already pulled the GFCI just to eliminate it as
a factor before I ran my suicide cord. I replaced it with a 3 prong
receptacle and disconnected the downstream devices so that there was
only the single 3 prong receptacle on the run.

When it still showed the open ground - even though there was a ground
wire attached - that's when I started tracing back and found the
prvious junction box. I'll be opening that tonight to look at the
ground situation.

As far as pulling a new grounded run from the panel to the junction
box, you asked "Why bother?"

Well, this is the last room in the house without 3 prong grounded
receptacles. That's why I started by adding the GFCI - so I could at
least install 3 prong receptacles downstream, mainly for the
convenience of having 3 prong receptacles available. Eventually, I
want to run grounded romex to all of the downstream receptacles, but
that is going to require moving some storage in the basement so it's
going to take some time.

My point is that, in your case, you don't need a ground to install 3-prong
receptacles. *The GFCI does the job. *The GFCI doesn't need a ground, either.
Everything should be good as it is. *Well, except the ungrounded 3-prong
outlets have to be marked "ungrounded".

Since the feed to the junction box has a ground wire, it appeared that
things would flow fairly smoothly in the forward direction from the
junction box on. Now that I'm showing an open ground, I have to go
backward and at a minimum find out why. I'm just not comfortable
knowing that a non-grounded ground wire is attached in the junction
box and at the GFCI. It'll gnaw at me until I fix it.

OK, there is that. *There's more than one reason I wouldn't buy a house with
no grounds. *;-)- Hide quoted text -

- Show quoted text -

Forgot to mention this:

It's not a house with no grounds, it's a house with *some* grounds.

Even though the older wires seem to be of the same cloth covered
vintage, some of them have grounds, some of them don't and in (at
least) this particular case, some have ground wires that aren't really
grounded.

Any new circuits that I've run in the past 20 or so years have all
been grounded. Except for some of the lighting circuits (two of which
include a shared neutral) this group is the last of the ungrounded
receptacles.

"Robert Green" wrote in message
...
"DD_BobK" wrote in message
news:8e96f96c-2f03-468c-b2fd-

btw those three light testers can be fooled by some unusual conditions

here's a link to a website

http://www.thecircuitdetective.com/test.htm

Scroll down about half way & there is a detailed discussion of the 3-
prong plug in tester.

That's a great resource! Thanks for posting it. Especially useful were
the
possible causes of "dim" readings which I've seen once or twice.

I saved a couple of thousand on this house just by going around to every
outlet with a 3 pronged tester and discovering that the seller's son had
replaced most of the two prong outlets with grounded outlets without any
connection to ground!!!! (Except for his room in the basement, where he
had
run a ground to act as a neutral to install a programmble light switch.)

I do not recall seeing any mention where someone connects the ground and
neutral together at the outlet. This has been done in the older homes
where there is no ground wire and fakes out the tester.

On Mon, 27 Feb 2012 13:29:42 -0800 (PST), DerbyDad03
wrote:

On Feb 27, 1:40*pm, "
wrote:
On Mon, 27 Feb 2012 06:27:01 -0800 (PST), DerbyDad03
wrote:





On Feb 26, 10:53*pm, "
wrote:
On Sun, 26 Feb 2012 16:30:14 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 6:45*pm, "
wrote:
On Sun, 26 Feb 2012 14:59:42 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 5:32*pm, "
wrote:
On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"

wrote:

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Read my later reply, after you'd posted this information. *Essentially, you do
have an open ground, like the TLT was telling you. *Where it is in the next
question. *Use the TLT on another outlet in this circuit. *Testing both an
outlet before and one after will nail it down pretty well. *If the one after
is fine, then there is a good chance that the fault is within the GFCI itself.

Well, at least we've stopped talking about the meter and/or the
operator, which so many wanted to blame as the cause.

We? *Well, the reason the meter is reading what it is, is because of its high
impedance.

"Testing both an outlet before and one after will nail it down
pretty well."

Can't do that. There is no receptacle before the GFCI and everything
after it is ungrounded. The GFCI was installed so that 3 pronged
receptacles could be installed downstream. Obviously, they'll all read
"open ground".

I guess your only option is to shut off the circuit and pull the GFCI. *The
easiest thing to do is to measure the continuity of the outlet itself from the
ground hole contacts to the screw. *You can probably do that without removing
the outlet from the box. *You should also be able to inspect the connection
without pulling it. *Else, you're going to have to pull it.

BTW, since there are no grounds in this house, how did the GFCI get one? *You
do know that it really doesn't need one, in this case.

However, here's the latest:

I opened the junction box just before the GFCI and disconnected the
line in Romex feed, which had a ground wire. I then ran a suicide cord
from a known good receptacle in my shop to the Romex going to the
GFCI. The GFCI now reads correctly when using the tester and there are
no strange readings between the current carriers and ground. Same
meter, same operator. ;-)

Oops. *That means it's a broken wire - much work.

I traced the original Romex back to another junction box which houses
a light fixture over the utility tub. That box has 2 other wires
associated with it, one obviously being the feed but the other one
disappears up into a wall. I'll need to open that box to see if
there's a ground in there, otherwise I'll be pulling a new wire from
the panel to the GFCI's junction box.

Again, why bother?

Too late to play now, SWMBO is cooking and killing that circuit kills
the kitchen lights too.

Dinner here was ruined (priority interrupt from real estate agent).- Hide quoted text -

- Show quoted text -

In all honesty, I had already pulled the GFCI just to eliminate it as
a factor before I ran my suicide cord. I replaced it with a 3 prong
receptacle and disconnected the downstream devices so that there was
only the single 3 prong receptacle on the run.

When it still showed the open ground - even though there was a ground
wire attached - that's when I started tracing back and found the
prvious junction box. I'll be opening that tonight to look at the
ground situation.

As far as pulling a new grounded run from the panel to the junction
box, you asked "Why bother?"

Well, this is the last room in the house without 3 prong grounded
receptacles. That's why I started by adding the GFCI - so I could at
least install 3 prong receptacles downstream, mainly for the
convenience of having 3 prong receptacles available. Eventually, I
want to run grounded romex to all of the downstream receptacles, but
that is going to require moving some storage in the basement so it's
going to take some time.

My point is that, in your case, you don't need a ground to install 3-prong
receptacles. *The GFCI does the job. *The GFCI doesn't need a ground, either.
Everything should be good as it is. *Well, except the ungrounded 3-prong
outlets have to be marked "ungrounded".

Since the feed to the junction box has a ground wire, it appeared that
things would flow fairly smoothly in the forward direction from the
junction box on. Now that I'm showing an open ground, I have to go
backward and at a minimum find out why. I'm just not comfortable
knowing that a non-grounded ground wire is attached in the junction
box and at the GFCI. It'll gnaw at me until I fix it.

OK, there is that. *There's more than one reason I wouldn't buy a house with
no grounds. *;-)- Hide quoted text -

- Show quoted text -

Forgot to mention this:

It's not a house with no grounds, it's a house with *some* grounds.

Yeah, I got that.

Even though the older wires seem to be of the same cloth covered
vintage, some of them have grounds, some of them don't and in (at
least) this particular case, some have ground wires that aren't really
grounded.

That's a problem. It may be hard to get it all together if you can't trust
anything.

Any new circuits that I've run in the past 20 or so years have all
been grounded. Except for some of the lighting circuits (two of which
include a shared neutral) this group is the last of the ungrounded
receptacles.

Shared neutral? Ick. I'd never do that.

On Feb 27, 1:40*pm, "
wrote:
On Mon, 27 Feb 2012 06:27:01 -0800 (PST), DerbyDad03
wrote:









On Feb 26, 10:53*pm, "
wrote:
On Sun, 26 Feb 2012 16:30:14 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 6:45*pm, "
wrote:
On Sun, 26 Feb 2012 14:59:42 -0800 (PST), DerbyDad03
wrote:

On Feb 26, 5:32*pm, "
wrote:
On Sun, 26 Feb 2012 17:27:42 -0500, "Ralph Mowery"

wrote:

"Tony Hwang" wrote in message
...

Hi,
I always have Simpson 260 handy working on problem like this. Fluke is for
some other use.

It might not show up in the circuits in a house, but at work on some of the
480 volt 3 phase circuits I could show some of the same things with a
Simpson 260. *Start with the voltage on the highest scale and don't worry
about the voltage, just pay attention on how far up the meter goes, then
switch to the lower ranges in order. *On many of the 'phantom' voltages, the
pointer on the meter will not move very far up or down.

I have an assortment of Fluke and other meters I can use, but mostly grab
the 260 and the old analog Amprobe for current measurments.

I do not have one to check it out, but I understand that some of the newer
digital test meters have a low input resistance to help eliminate the
'phantom' voltage problem.

Just plug an incandescent light bulb into the circuit.

With a lamp plugged into the same GFCI, and turned on, I get the same
"oddball" readings at the GFCI.

With no load on 2 other circuits that I tested, I get the expected
readings of 0 between ground and neutral, 115V between hot and ground.

Fluke 75 multimeter.

Read my later reply, after you'd posted this information. *Essentially, you do
have an open ground, like the TLT was telling you. *Where it is in the next
question. *Use the TLT on another outlet in this circuit. *Testing both an
outlet before and one after will nail it down pretty well. *If the one after
is fine, then there is a good chance that the fault is within the GFCI itself.

Well, at least we've stopped talking about the meter and/or the
operator, which so many wanted to blame as the cause.

We? *Well, the reason the meter is reading what it is, is because of its high
impedance.

"Testing both an outlet before and one after will nail it down
pretty well."

Can't do that. There is no receptacle before the GFCI and everything
after it is ungrounded. The GFCI was installed so that 3 pronged
receptacles could be installed downstream. Obviously, they'll all read
"open ground".

I guess your only option is to shut off the circuit and pull the GFCI. *The
easiest thing to do is to measure the continuity of the outlet itself from the
ground hole contacts to the screw. *You can probably do that without removing
the outlet from the box. *You should also be able to inspect the connection
without pulling it. *Else, you're going to have to pull it.

BTW, since there are no grounds in this house, how did the GFCI get one? *You
do know that it really doesn't need one, in this case.

However, here's the latest:

I opened the junction box just before the GFCI and disconnected the
line in Romex feed, which had a ground wire. I then ran a suicide cord
from a known good receptacle in my shop to the Romex going to the
GFCI. The GFCI now reads correctly when using the tester and there are
no strange readings between the current carriers and ground. Same
meter, same operator. ;-)

Oops. *That means it's a broken wire - much work.

I traced the original Romex back to another junction box which houses
a light fixture over the utility tub. That box has 2 other wires
associated with it, one obviously being the feed but the other one
disappears up into a wall. I'll need to open that box to see if
there's a ground in there, otherwise I'll be pulling a new wire from
the panel to the GFCI's junction box.

Again, why bother?

Too late to play now, SWMBO is cooking and killing that circuit kills
the kitchen lights too.

Dinner here was ruined (priority interrupt from real estate agent).- Hide quoted text -

- Show quoted text -

In all honesty, I had already pulled the GFCI just to eliminate it as
a factor before I ran my suicide cord. I replaced it with a 3 prong
receptacle and disconnected the downstream devices so that there was
only the single 3 prong receptacle on the run.

When it still showed the open ground - even though there was a ground
wire attached - that's when I started tracing back and found the
prvious junction box. I'll be opening that tonight to look at the
ground situation.

As far as pulling a new grounded run from the panel to the junction
box, you asked "Why bother?"

Well, this is the last room in the house without 3 prong grounded
receptacles. That's why I started by adding the GFCI - so I could at
least install 3 prong receptacles downstream, mainly for the
convenience of having 3 prong receptacles available. Eventually, I
want to run grounded romex to all of the downstream receptacles, but
that is going to require moving some storage in the basement so it's
going to take some time.

My point is that, in your case, you don't need a ground to install 3-prong
receptacles. *The GFCI does the job. *The GFCI doesn't need a ground, either.
Everything should be good as it is. *Well, except the ungrounded 3-prong
outlets have to be marked "ungrounded".

Since the feed to the junction box has a ground wire, it appeared that
things would flow fairly smoothly in the forward direction from the
junction box on. Now that I'm showing an open ground, I have to go
backward and at a minimum find out why. I'm just not comfortable
knowing that a non-grounded ground wire is attached in the junction
box and at the GFCI. It'll gnaw at me until I fix it.

OK, there is that. *There's more than one reason I wouldn't buy a house with
no grounds. *;-)

I found the problem.

As a recap, the junction box just below the GFCI receptacle had a feed
wire with a ground, which I connected to the ground of the wire that
went up the GFCI. However, the GFCI read "open ground".

When I used a suicide cord with a known good ground as the feed to
that junction box, the tester showed that the GFCI was wired
correctly. So now I had to go looking for the open ground upstream of
that junction box.

I traced the original feed wire to another junction box that held a
light fixture above the utility sink. The box had a feed wire coming
in and 2 wires coming out.

When I dropped the light fixture, here's what I found:

The ground from the feed wire was wire nutted to the 3 neutrals. The
ground in the wire that ran to the GFCI's junction box was cut back to
about a 1/2 inch and folded neatly back along the insulation. The
other wire that left the box had no ground wire.

Luckily, there was enough slack in the wire I was working with that I
was able to pull enough into the junction box to expose the ground
wire and connect it to the ground wire of the feed.

The tester now reads correctly at the GFCI.

I won't even fathom a guess as to why they would have wired the ground
with the neutrals and then cut back the other ground. Why not just
connect the grounds as long as they were in the box? Who knows...

"DerbyDad03" wrote in message news:2f775e55-52cb-I
traced the original feed wire to another junction box that held a
light fixture above the utility sink. The box had a feed wire coming
in and 2 wires coming out.

When I dropped the light fixture, here's what I found:

The ground from the feed wire was wire nutted to the 3 neutrals. The
ground in the wire that ran to the GFCI's junction box was cut back to
about a 1/2 inch and folded neatly back along the insulation. The
other wire that left the box had no ground wire.

Luckily, there was enough slack in the wire I was working with that I
was able to pull enough into the junction box to expose the ground
wire and connect it to the ground wire of the feed.

The tester now reads correctly at the GFCI.

I won't even fathom a guess as to why they would have wired the ground
with the neutrals and then cut back the other ground. Why not just
connect the grounds as long as they were in the box? Who knows...

Just as most said, a disconnected ground somewhere between the GFCI and
breaker box.

Thanks for letting us know what the problem was.