GFCI--Red LED stays on

[W. eWatson]

Subject.

Neither pressing the test or reset button turns the LED off. Neither
really depresses. It's a single line with nothing else on it. The switch
box shows no switch thrown. The line is buried 18-24" in black
electrical plastic pipe, and goes about 40' to the switch box. Comments?

 

[P E Schoen]

"W. eWatson" wrote in message

Neither pressing the test or reset button turns the LED off.
Neither really depresses. It's a single line with nothing else
on it. The switch box shows no switch thrown. The line is
buried 18-24" in black electrical plastic pipe, and goes about
40' to the switch box. Comments?

You may need to hold the switch down for up to ten seconds or so for it to
reset. Sometimes the button is hard to reach, and it doesn't have a snap
action "click", so you can use a plastic pen or similar tool to push it.

Something else you can do is disconnect the circuit at the breaker box, and
then read resistance from both the black and white wires to earth ground. Be
very careful. If you are not "dead sure" what you are doing, you will soon
BE dead for SURE!

Paul

 

[Winston]

Subject.

Neither pressing the test or reset button turns the LED off. Neither
really depresses. It's a single line with nothing else on it. The switch
box shows no switch thrown. The line is buried 18-24" in black
electrical plastic pipe, and goes about 40' to the switch box. Comments?

http://en.wikipedia.org/wiki/Residual-current_device

Is the GFCI a 'latching' or 'non-latching' type?
Make and model, please?

Active/passive latching/non-latching

RCDs may be obtained that have different behaviours if the circuit they
are protecting is de-energised.

* One type will trip on power failure and not re-make the circuit
when the circuit is re-energised. This type is known as
non-latching.[citation needed] [8]

* Another type will maintain the circuit after power is restored.
This type is known as latching or active.

--Winston

 

[W. eWatson]

Subject.

Neither pressing the test or reset button turns the LED off. Neither
really depresses. It's a single line with nothing else on it. The switch
box shows no switch thrown. The line is buried 18-24" in black
electrical plastic pipe, and goes about 40' to the switch box. Comments?

http://en.wikipedia.org/wiki/Residual-current_device

Is the GFCI a 'latching' or 'non-latching' type?
Make and model, please?

Active/passive latching/non-latching

RCDs may be obtained that have different behaviours if the circuit they
are protecting is de-energised.

* One type will trip on power failure and not re-make the circuit when
the circuit is re-energised. This type is known as
non-latching.[citation needed] [8]

* Another type will maintain the circuit after power is restored. This
type is known as latching or active.

--Winston

Pushing on Test or Reset for 10 to 15 seconds didn't change anything.

I have no idea which switch in the breaker box is for this outlet. The
breaker box was installed 10 years ago, and I had an electrician install
two more circuits about five years ago. I may have a diagram I drew
several years ago. The box is in a small 300' square foot building.

The device looks nowhere like the link Winston provided. I cannot
determine the mfer, but it's probably a Levitron. I found a instruction
sheet for a GFCI in my electric parts box, and it mentions a red LED.
The sheet was for a Levitron. The device is commonly found in households.

When our local h/w store is open tomorrow, I might be able to determine
the mfer.

I put the outlet outdoors about five years ago. I either bought it here
then or used one I had in my electric box that was about 15 years old.

What powers the LED while it's declaring a fault?

I have had the LED go on a number of times over our long winter. Pushing
the proper button got it back on. It doesn't get a lot of notice, since
a measuring device associated with it is mostly non-operational in the
winter. It's a sky camera, and there's not much to look at in the
winter. Typically the LED turns on after a rain storm. We get 60" a year
here, all in the winter.

 

[Winston]

Subject.

Neither pressing the test or reset button turns the LED off. Neither
really depresses. It's a single line with nothing else on it. The switch
box shows no switch thrown. The line is buried 18-24" in black
electrical plastic pipe, and goes about 40' to the switch box. Comments?

http://en.wikipedia.org/wiki/Residual-current_device

Is the GFCI a 'latching' or 'non-latching' type?
Make and model, please?

Active/passive latching/non-latching

RCDs may be obtained that have different behaviours if the circuit they
are protecting is de-energised.

* One type will trip on power failure and not re-make the circuit when
the circuit is re-energised. This type is known as
non-latching.[citation needed] [8]

* Another type will maintain the circuit after power is restored. This
type is known as latching or active.

--Winston

Pushing on Test or Reset for 10 to 15 seconds didn't change anything.

I have no idea which switch in the breaker box is for this outlet. The
breaker box was installed 10 years ago, and I had an electrician install
two more circuits about five years ago. I may have a diagram I drew
several years ago. The box is in a small 300' square foot building.

The device looks nowhere like the link Winston provided. I cannot
determine the mfer, but it's probably a Levitron. I found a instruction
sheet for a GFCI in my electric parts box, and it mentions a red LED.
The sheet was for a Levitron. The device is commonly found in households.

When our local h/w store is open tomorrow, I might be able to determine
the mfer.

I put the outlet outdoors about five years ago. I either bought it here
then or used one I had in my electric box that was about 15 years old.

What powers the LED while it's declaring a fault?

Apparently the LED is powered from the input side of the GFCI.

It is telling you that some amount of current that it sends out
the 'line' side is *not* returning via the 'neutral' side.
It's saying you have a dangerous condition involving
current not going exactly where you want it to go.

People have suffered nasty shocks under these circumstances.

I have had the LED go on a number of times over our long winter. Pushing
the proper button got it back on.

So, in the past after a ground fault event, you could press the 'reset'
button and bring power back to the cable, until the next rain, yes?

I suppose that your plastic pipe is no longer doing it's
job of keeping your cable nice and dry.

It doesn't get a lot of notice, since
a measuring device associated with it is mostly non-operational in the
winter. It's a sky camera, and there's not much to look at in the
winter. Typically the LED turns on after a rain storm. We get 60" a year
here, all in the winter.

Based on what you've said here, I would conjecture that either
the 'line' or 'neutral' side of the 40' cable came into contact
with that water intermittently during winter and has now found
a more permanent place to leak current to ground.

Diagnosis and repair are in order!

Here is something that could be educational and fun.

I would first turn off all the power to the outbuilding.
(Shouldn't take long to determine which breakers
power those circuits.)

Then I would attach my shop vacuum to the near side
of the plastic pipe and watch the vacuum fill
with water. (Use about 6' too much duct tape.)
I would go to the outbuilding listen at the switch
box for the sound of entering air.

If you do detect moisture in the plastic pipe, you will
want to uncover and repair the crack or installation
error that is allowing water in.

--Winston

 

[W. eWatson]

....

Apparently the LED is powered from the input side of the GFCI.

It is telling you that some amount of current that it sends out
the 'line' side is *not* returning via the 'neutral' side.
It's saying you have a dangerous condition involving
current not going exactly where you want it to go.

People have suffered nasty shocks under these circumstances.


So, in the past after a ground fault event, you could press the 'reset'
button and bring power back to the cable, until the next rain, yes?

I suppose that your plastic pipe is no longer doing it's
job of keeping your cable nice and dry.


Based on what you've said here, I would conjecture that either
the 'line' or 'neutral' side of the 40' cable came into contact
with that water intermittently during winter and has now found
a more permanent place to leak current to ground.

Diagnosis and repair are in order!

Here is something that could be educational and fun.

I would first turn off all the power to the outbuilding.
(Shouldn't take long to determine which breakers
power those circuits.)

Then I would attach my shop vacuum to the near side
of the plastic pipe and watch the vacuum fill
with water. (Use about 6' too much duct tape.)
I would go to the outbuilding listen at the switch
box for the sound of entering air.

If you do detect moisture in the plastic pipe, you will
want to uncover and repair the crack or installation
error that is allowing water in.

--Winston

Yes, up until the other night resetting was very easy. Interestingly
though, it has not rained really hard for more than a week, but it may
have been a month since I had to reset.

I have a shop vac. 6" duct tape? I have a sub-pump, which might be
easier to attach. I hope I do not have to dig up ground. It's clay, and
either loaded with water or hardened.

I mapped the switches this morning. There are five switches. #5 controls
the outlet 40' from the building, but there's another outlet 125'
further out from the building for a pond, and is on #5. It's elevated by
about 5' from the first outlet. I think it has its own length of wire,
possibly a different gauge wire. I disabled #5 until I can get this fixed.

I think I'll call my electrician this morning to see if he can be of any
help.

 

[Winston]

W. eWatson wrote:

(...)

Yes, up until the other night resetting was very easy. Interestingly
though, it has not rained really hard for more than a week, but it may
have been a month since I had to reset.

Could be soil movement cracked the plastic pipe, or something.

I'd be tempted to disconnect the GFCI and use a multimeter
to determine resistance from each downstream wire to ground.
If both 'line' and 'neutral' show infinite resistance to
ground and the GFCI will still not drive the lines, perhaps
the GFCI is broken. Any lightning in the area recently?

I have a shop vac. 6" duct tape?

An attempt at humor. I mentioned six feet of duct tape to
illustrate the lashup necessary to attach a shop vacuum to
a piece of plastic conduit.

I have a sub-pump, which might be
easier to attach.

I imagine that the shop vacuum is *much* more self-priming.

I hope I do not have to dig up ground. It's clay, and
either loaded with water or hardened.

I suffer from that stuff too.

Perhaps you can use the existing cable as a leader to
pull some 'submarine power cable' through the
plastic pipe.

I mapped the switches this morning. There are five switches. #5 controls
the outlet 40' from the building, but there's another outlet 125'
further out from the building for a pond, and is on #5. It's elevated by
about 5' from the first outlet. I think it has its own length of wire,
possibly a different gauge wire. I disabled #5 until I can get this fixed.

I think I'll call my electrician this morning to see if he can be of any
help.

Sounds like a good plan.

--Winston

 

[Rich Grise]

I have a shop vac. 6" duct tape? I have a sub-pump, which might be
easier to attach. I hope I do not have to dig up ground. It's clay, and
either loaded with water or hardened.

I thought he said six FEET of duct tape. ;-)

For the clay, make a 4" x 8" x 2" mold, get some straw, and cast the clay
into adobe bricks, then build a tunnel for your wires. ;-)

Of course, you'll need some kind of sump (or a runoff place) to drain the
wetness from the yard.

Have Fun!
Rich

 

[Rich Grise]

An attempt at humor. I mentioned six feet of duct tape to
illustrate the lashup necessary to attach a shop vacuum to
a piece of plastic conduit.

Truth be told, I've done this very thing. The shop has a brand-
named Shop-Vac(r?) shop vac, but it was lacking sufficient wands.
The kit with wands, hose, and nozzle was like forty-five bucks!
I looked for some PVC or ABS pipe, but an eight- or ten-foot
piece was almost ten bucks! So I thought I'd amble down to the
Home Depot to see if I could find anything, and there was gray
plastic (I think PVC) conduit, for like 2.99 for a ten foot piece!

I only needed about three feet of duct tape to wrap around the
PVC to make a snug fit to the hose cuff, and I still have another
five feet of conduit as a spare.

Cheers!
Rich

 

[W. eWatson]

W. eWatson wrote:

(...)


Could be soil movement cracked the plastic pipe, or something.

I'd be tempted to disconnect the GFCI and use a multimeter
to determine resistance from each downstream wire to ground.
If both 'line' and 'neutral' show infinite resistance to
ground and the GFCI will still not drive the lines, perhaps
the GFCI is broken. Any lightning in the area recently?


An attempt at humor. I mentioned six feet of duct tape to
illustrate the lashup necessary to attach a shop vacuum to
a piece of plastic conduit.


I imagine that the shop vacuum is *much* more self-priming.


I suffer from that stuff too.

Perhaps you can use the existing cable as a leader to
pull some 'submarine power cable' through the
plastic pipe.


Sounds like a good plan.

--Winston

Actually, my sub-pump will not do me any good, so I bought a "drill"
pump this morning for $10. My hose may not be flexible to make a bend,
although there's a bend where the vertical 4' conduit goes into the ground.

I'm just going to turn off the circuit and take out the GFCI, after
checking as you mention above. If the unit is defective, then I probably
do not need to pump water.

 

[W. eWatson]

I thought he said six FEET of duct tape. ;-)

For the clay, make a 4" x 8" x 2" mold, get some straw, and cast the clay
into adobe bricks, then build a tunnel for your wires. ;-)

Of course, you'll need some kind of sump (or a runoff place) to drain the
wetness from the yard.

Have Fun!
Rich

I could drop a little gasoline down there, and follow it with a match.

 

[Winston]

W. eWatson wrote:

(...)

I'm just going to turn off the circuit and take out the GFCI, after
checking as you mention above. If the unit is defective, then I probably
do not need to pump water.

If there *is* water in the plastic pipe and you are not
going to replace your wiring with 'underwater & direct bury'
cable, you really *do* need to remove the water and repair
the pipe so that no water can enter it. Think *dry*.

All bets are off if your wires are allowed to get damp.

--Winston

 

[Winston]

Rich Grise wrote:

(...)

I only needed about three feet of duct tape to wrap around the
PVC to make a snug fit to the hose cuff, and I still have another
five feet of conduit as a spare.

You *do* realize that comes awfully close to using duct tape
to do duct work? I'm not going to report you to the guys
at the Possum Lodge this time, but be careful in the future,
y'hear?

--Winston

 

[Jamie]

I could drop a little gasoline down there, and follow it with a match.

why don't you just use the existing wire as a pull wire and pull in some
direct burial type cable ?

Jamie

 

[W. eWatson]

A new branch to the thread. The other one has gotten frazzled.

I decided to push a 1/4" garden irrigation tube down the vertical tube.
It was about 20' long, and it went in smoothly. I pulled it out, and
most of it showed water.

I then decided to pump it out with the "drill" pump using the tube. It
wouldn't work. The pump requires a 3/4" or 1" diameter. I had used a cap
of the right size that had a hole for the 1/4" tube. After taking it
down to the h/w store, and experimenting with it, it became apparent I'd
have to attach a "full" sized tube.

It started raining, so I've given up for the evening. A pump for a 1/4"
tube would probably do the trick. However, I plan to check the GFCI
tomorrow. Maybe I won't need the pump.

 

[ehsjr]

A new branch to the thread. The other one has gotten frazzled.

I decided to push a 1/4" garden irrigation tube down the vertical tube.
It was about 20' long, and it went in smoothly. I pulled it out, and
most of it showed water.

I then decided to pump it out with the "drill" pump using the tube. It
wouldn't work. The pump requires a 3/4" or 1" diameter. I had used a cap
of the right size that had a hole for the 1/4" tube. After taking it
down to the h/w store, and experimenting with it, it became apparent I'd
have to attach a "full" sized tube.

It started raining, so I've given up for the evening. A pump for a 1/4"
tube would probably do the trick. However, I plan to check the GFCI
tomorrow. Maybe I won't need the pump.

The "drill pump" probably requires priming and a constant supply
of water at its input. Likely not a good choice for what you are
trying to do. The shop vac hooked up to blow air into the pipe
has a far batter chance of drying things than hooking it up to
pull the water out. Of course, that is not the solution to your
problem.

The weak points in your setup, aside from any defects, are at the
outlets at 125' and at 40'. Open them up, clean any dirt/gunk in
the j boxes and _replace_ whatever receptacles are there. Next,
replace the GFCI. Receptacles outdoors and in unheated outbuildings
collect condensation, spiderwebs, dirt etc. A GFCI that has tripped
multiple times may well have burned contacts. So, since you have
to invest time for diagnosis anyway, you might as well replace the
devices you need to remove for inspection/cleaning and the GFCI
(because of the multiple trips). NOW you can start troubleshooting
if you haven't fixed the problem. You DO NOT want to go through
all the effort of pulling new UF cables, digging trenches, replacing
conduit etc only to discover that all you needed to do was spend
$20 and replace two receptacles plus a GFCI.

Wetness in the conduit is not causing the trip alone, and it could
be normal condensation. Of course, if rain water is getting into
the conduit, that's not normal, but it still isn't the sole cause
of the problem. The conductors still have to come into contact with
the water, so one or more of the conductors insulation has to be
compromised somewhere, too. If you have eliminated the other
possibilities, you could try pulling new UF cable in the existing
conduit. If you _know_ rainwater is getting into the conduit
you've got to fix whatever is causing that - which might mean
a lot of digging to run new conduit.

I'm working hard to get you to spend the ~20 bucks for the
receptacles and GFCI - can you tell?

Ed

 

[W. eWatson]

The "drill pump" probably requires priming and a constant supply
of water at its input. Likely not a good choice for what you are
trying to do. The shop vac hooked up to blow air into the pipe
has a far batter chance of drying things than hooking it up to
pull the water out. Of course, that is not the solution to your
problem.

The weak points in your setup, aside from any defects, are at the
outlets at 125' and at 40'. Open them up, clean any dirt/gunk in
the j boxes and _replace_ whatever receptacles are there. Next,
replace the GFCI. Receptacles outdoors and in unheated outbuildings
collect condensation, spiderwebs, dirt etc. A GFCI that has tripped
multiple times may well have burned contacts. So, since you have
to invest time for diagnosis anyway, you might as well replace the
devices you need to remove for inspection/cleaning and the GFCI
(because of the multiple trips). NOW you can start troubleshooting
if you haven't fixed the problem. You DO NOT want to go through
all the effort of pulling new UF cables, digging trenches, replacing
conduit etc only to discover that all you needed to do was spend
$20 and replace two receptacles plus a GFCI.

Wetness in the conduit is not causing the trip alone, and it could
be normal condensation. Of course, if rain water is getting into
the conduit, that's not normal, but it still isn't the sole cause
of the problem. The conductors still have to come into contact with
the water, so one or more of the conductors insulation has to be
compromised somewhere, too. If you have eliminated the other
possibilities, you could try pulling new UF cable in the existing
conduit. If you _know_ rainwater is getting into the conduit
you've got to fix whatever is causing that - which might mean
a lot of digging to run new conduit.

I'm working hard to get you to spend the ~20 bucks for the
receptacles and GFCI - can you tell?

Ed

Good info. I appreciate your effort, and very likely will just spend the
$20.

If I do not find a solution above ground, then it's highly unlikely I
will spend lots of dollars excavating a trench. For probably 4 or more
years I simply ran one of those orange power cords out of the building
along the ground out to the equipment. It worked very well. No animals
ever chewed on the cord, unlike damage they've done to 1/2" irrigation
tubing I've run around the property. The only noticeable "damage" was
the color of the cord dulled, but yet the flexibility remained.

I will likely take the drill pump back. However, I'm now curious about
any type of pump that might work in such a situation. BTW, the shop vac
approach is pretty much out of the question. There really is no good way
to attach it w/o removing the outlet box.

 

[Winston]

W. eWatson wrote:

(...)

Good info. I appreciate your effort, and very likely will just spend the
$20.

What were the results of your resistance measurements?
Infinite resistance to ground on line and neutral?

(...)

I will likely take the drill pump back.

They don't self-prime very well, under the best of circumstances.

However, I'm now curious about
any type of pump that might work in such a situation. BTW, the shop vac
approach is pretty much out of the question. There really is no good way
to attach it w/o removing the outlet box.

You could probably have the box gutted and removed in 10 minutes.

Doesn't matter. You answered the question about water incursion
and that is all the shop vac was for. Now that we know water is
getting into the plastic pipe, drying it is useless until it is
repaired.


--Winston

 

[W. eWatson]

I installed a new GFCI today and it is working just fine. It's the same
as the defective one. Pass and Seymour GFCI.

 

[Winston]

I installed a new GFCI today and it is working just fine. It's the same
as the defective one. Pass and Seymour GFCI.

Keep us posted.

--Winston