GFCI breaker tripping


  #1  
Old 01-23-01, 05:06 AM
Guest
Visiting Guest
Posts: n/a
I have a 20 Amp GFCI breaker, supposedly wired to only one device, a whirlpool bath. The circuit worked properly for some time, and then failed. The breaker will not reset. It trips immediately. The outlet by the whirlpool is non-GFCI.

To diagnose, I have:
- Unplugged the whirlpool from the outlet. Still will not reset.
- Removed the duplex outlet and capped the wires. Still will not reset.
- Replaced the breaker. Still will not reset.
- Checked the voltage and resistance of the line. The circuit shows no voltage at the breaker (as expected, since it is tripped). The resistance is infinite at the breaker (across the black and the white and the black and the bus).
If I deliberately short the circuit at the outlet, the resistance across the breaker goes to 0.

These diagnostics seems to rule out a bad breaker, a bad device (whirlpool), a bad duplex outlet, or a break in the wiring. So what's left? The only thing that I can think of is that there is another (bed) device somewhere on the supposedly dedicated circuit. If so, how do I find it?
Are there other possible causes?


 
  #2  
Old 01-23-01, 09:42 AM
J
Member
Join Date: Sep 2000
Location: Fayetteville, NY, USA
Posts: 1,052
Received 1 Vote on 1 Post
You seem to have done some very good diagnostic work. One key thing is in the term "Ground Fault Circuit Interrupter". The problem could very well be a ground fault. That is hot juice leaking directly to ground. With power to the circuit off you should check resistance between hot & ground. Just a nick in the black wire's insulation anywhere along it could be touching either the bare ground wire in the cable going to the whirlpool's receptacle, or any grounded metalic surface anywhere that circuit goes. Finding it requires visually inspecting the entire length of that cable, and any junction box it goes thru.

If it were my house I would take ground fault protection out of the picture for a moment just to isolate another variable. I'd borrow a non-GFCI breaker from another slot in my breaker panel and snap it in in place of the GFCI breaker just to see what happens. (Being a whirlpool circuit, though, I would not use that tub without GFCI protection.) With everything hooked back up and the non-GFCI breaker switched on what happens?

One other thing to check - is there anywhere along that circuit where water could get into the cable or box associated with that circuit?

Also, is all or part of the wiring concealed in walls. If so, (a) visual inspection would be tough; and (b) is it possible someone was hanging pictures or otherwise driving nails or screws into a wall where that cable may travel just prior to this tripping problem? Stranger things have happened.

Is there event or unusual occurance that you can remember which happened at your house around the time the tripping started. No matter how remote it may seem to the tripping problem it's worth thinking about.

It's also probably worth your time to go through the whole house, I mean every wall of every room on every floor of the house, and see if you can find something else that's also without power. "Supposed to be" a dedicated circuit isn't necessarily the same thing as "IS" dedicated.

Without further information that's about all I can come up with. Hope that gives you some ideas that turn up something. Let us know what happens, or doesn't happen, and we'll do what we can for ya. Again, very good diagnostic work so far.

Juice
 
  #3  
Old 01-23-01, 10:05 AM
Guest
Visiting Guest
Posts: n/a
GFCI Tripping

Juice wrote: "With power to the circuit off you should check resistance between hot & ground..."

I tried that (hot (black at the breaker) to ground (bus in the breaker box)), and had infinite resistance. Is that what you meant, and does that definitively rule out a ground fault in the circuit?

The cable is hidden in walls for 99% of it's length, although I have a very good idea where it runs. This means that I cannot visually check it, although I can check the walls for nails, etc, and have done so, with negative results.

Juice wrote: "If it were my house I would take ground fault protection out of the picture for a moment just to isolate another variable. I'd borrow a non-GFCI breaker from another slot in my breaker panel and snap it in in place of the GFCI breaker just to see what happens. ... With everything hooked back up and the non-GFCI breaker switched on what happens?"

I haven't tried that yet, but it's a good idea. What are the implications if a non-GFCI breaker trips, or does not trip, on the circuit? What does that rule in or out as possible causes?

Water in the box? Not at the whirlpool end, or at the breaker box. Unlikely in the walls, but possible I suppose. I see no evidence of water anywhere.

The search and destroy mission for another powerless device will be the next step. It's the most likely scenario at the moment.

Is there any simple way to prove or disprove the presence of another device on the circuit?

Bill J.
 
  #4  
Old 01-23-01, 10:37 AM
J
Member
Join Date: Sep 2000
Location: Fayetteville, NY, USA
Posts: 1,052
Received 1 Vote on 1 Post
Checking hot to ground, I would have the black wire disconnected and isolated from everything at both ends so I was sure it was completely floating. Then I would put a probe on one end and the other probe on the ground wire of that cable. I would do it at the receptacle end of the cable. Then I'd check black to white. It's just the piece of cable and the conductors in it that you're concerned about testing, which is why I'd be isolated from the panel and it's ground and neutral buses. Too many variiables for me there.

Implications if a non-GFCI trips: Most probably a dead short. A GFCI breaker of receptacle will trip upon miliamps of leakage, which a standard non-GFCI will often not. We're talking about a much more sensitive device. If the less sensitive non-GFCI does not trip there could be all kinds of things going on, but if it does trip a dead short is indicated, and usually easier to find the problem, in my experience (Such as a big black scorch mark, or flying sparks!).

Water in the system somewhere: not likely, but was worth mentioning.

Prove or disprove presence of another device on the circuit: As far as I know, just checking every single receptacle, light fixture, behind appliances, in closets, the attic, garage. Anywhere something electrical is or could be used, check and prove power there. You're looking for something dead, or even a GFCI receptacle that's tripped.

As the day grows long some of the other guys will start popping in and they will undoubtedly have some good ideas. I'm sort of out at the moment, but will continue to stop in and see how it's going. Later.

Juice
 
  #5  
Old 01-24-01, 04:17 PM
Wgoodrich
Visiting Guest
Posts: n/a
Do not use that whirlpool tub if powered without GFI protection. However, if you do temporarily energize that whirlpool tub with a normal breaker on a test basis, and all works fine, then I would suspect the white and the bare to be touching in the circuit somewhere.

If the circuit works fine without GFI protection, then de-energize the circuit again and test resistance between the white and bare. If you find the white and bare to be touching then this is your reason the GFI won't hold. May be a nail driven into the wall to hang a picture or any other type cause. GFI will not hold if black, white, or bare touch any conductor in that circuit.

Also if someone added a load to this circuit, and you have a situation where a neutral is shared with another hot conductor then the GFI won't hold on this situation either.

To confirm that that white wire is being shared try de-energizing the hot conductor of that circuit and carefully disconnect that white wire from the neutral bar. Strike that white wire to the neutral bar. If there is an arc or something quits running then the white is being shared with another hot. Sharing a neutral on a 120 volt GFI device will not work.

Let us know what you find.

Wg
 
  #6  
Old 01-24-01, 04:35 PM
J
Member
Join Date: Sep 2000
Location: Fayetteville, NY, USA
Posts: 1,052
Received 1 Vote on 1 Post
Drove a picture nail through it? Now why didn't I think of that? ...Oh yeah, I did!

Hey Wg, this guy has an electrical tester, so before you have him strike a disconnected white onto the bus I'd first advise him to test it to the bus with a meter. Reminds me of one of the first editions of the NEC. It recommended that one touch the wires to test for the presence of voltage. Heard that in a seminar given by the International Association of Electrical Inspectors, so it must be true!

"Still having fun."

Juice
 
  #7  
Old 01-24-01, 05:26 PM
Guest
Visiting Guest
Posts: n/a
Juice and Wg,

Just tried the white to ground, and found that they have continuity! This is with all wires hanging in the air at the known outlet, and the wires still connected at the breaker end. I can see that the white and ground wires do not touch in the box, but I won't reach in there and disconnect them until the weekend (when I can shut the breaker box down).

So perhaps it is a real ground fault. I will do the other tests on the weekend.

When the place was being built, I took pictures of the wires in the walls, and there are definitely no new nails (or other construction)anywhere along the circuit in the past few months (since the last time the circuit was known to work).

Perhaps there was wire damage during construction that only just became apparent when something settled and the wires moved.

The search and destroy for another dead device came up empty. So if it really is a damaged wire, I suppose I will have to abandon it in place. Short of running a new wire the entire length, is the following approach reasonable? Can I tie into another curcuit in that part of the house, as long as I add a GFCI outlet at the whirlpool, and upgrade the breaker on that line to 20A?
 
  #8  
Old 01-24-01, 06:50 PM
Wgoodrich
Visiting Guest
Posts: n/a
As long as this circuit is a dedictated circuit for that tub then disconnect the white and bare from both ends, both in the receptacle and from the neutral bar in the panel. Then do you continuity test. Otherwise you will show a short because both the bare and the white is connceted to the neutral bar in the panel.

Juicehead, I asked him to arc the white to the neutral bar not the buss bar. Just to keep him from being confused and confuse him to do what you were having fun with. I suggested that he disconnect the white wire from the neutral bar then strike the bar that the white wire was disconnected from [neutral bar]if a spark shows then it would prove that that white is being shared with another hot wire.

Striking the buss would probably cause him to have heart palpatations. I pictured that in my mind, not a good picture.

Wg
 
  #9  
Old 01-25-01, 06:05 AM
J
Member
Join Date: Sep 2000
Location: Fayetteville, NY, USA
Posts: 1,052
Received 1 Vote on 1 Post
Wg, I meant the neutral bus, which is what I thought you meant. "Bus" to me does not necessarily imply ground bus. A bus is a bus. I work with hot buses in MCCs and other equipment. I should have clarified.

(Also, FYI "Buss" is a brand name of a fuse manufacturer. The correct spelling for a common connection point for multiple conductors is "bus".)

I also felt that taking a voltage reading from a disconnected white wire, which may share a hot in that circuit, to the neutral bus would show what you suspect.

BillJ, I would disconnect all three wires at BOTH ends of the circuit so that you have essentially a piece of cable completely floating, and you are testing only the conductors of a single cable and nothing else. Just as if it's still in the package on the shelf at home depot. There are no variables whatsoever in this scenario, only 6 bare ends, period.

To find out if there's anything else concealed in the walls which may share this circuit it is possible to find it as follows: Test each of the three completely isolated wires to the neutral bus. If Wg's right I'm thinking that the white to neutral bus test will reveal at least some voltage. But if any of the three show voltage to the neutral or ground buses there is something else connected in that circuit somewhere in the house. If this test reveals nothing go to the next:

To find a short between any two of the wires in this cable betwen the box & the whirlpool receptacle, you would of course test continuity between black & white, black & bare, and white & bare with all wires completely separated from each other or anything else at both ends. (I know I'm repeating myself but I am stressing this because any variation will poison the test results.) If you don't find continuity you don't have a short.

But you could have an open. When ends are distant from one another there is an easy method of testing for an open. After testing for the short and finding none, take a bare & a white and twist them together at one end only. (Everything is definitely disconnected from power and everything else at this point of course.) Go to the other end and test continuity between the separated white & bare at the other end. If there's no continuity you have one or the other open. To find out which one go back to the other end and untwist white & bare and twist black & bare together. Repeat continuity test at other end, but on black & bare this time. If it's still open you have an open ground. If the first two tests revealed continuity, all three have no open. With your initial report of the testing you've done I'm fairly certain you follow what I'm doing clearly. If there are readings other than what I suggested you might find, there must be something else on that circuit somewhere.

Let us know what these tests reveal. This is a fascinating problem to say the least. Good luck.

Juice
 
  #10  
Old 01-25-01, 06:12 AM
Guest
Visiting Guest
Posts: n/a
Juice,

That series of tests is perfectly clear and definitive.
Only reason I didn't disconnect all wires at the breaker end last night is because I could not shut down all power to the breaker box, and I do not want to stick my hands in there while the box is hot. I need to find an opportunity in the next few days to do that.

I will get back with the results.
 
  #11  
Old 01-30-01, 05:40 AM
Guest
Visiting Guest
Posts: n/a
Juice and WG,

Ran some tests over the weekend.

With all wires disconnected at both ends (3 wires hanging free at the box and at the outlet), the tests and results were:

Test for voltage on wires. At Breaker Box, check voltage between each wire and ground bus.
Copper to bus - 0 volts
Black to bus - 0 volts
White to bus - 0 volts

Test for Short. At breaker box, test resistance between each pair of wires.
Black to white - infinite ohms
Black to copper - infinite ohms
White to copper - infinite ohms

Test for Open.
At outlet, connect white and copper with wire nut.
Test resistance of white and copper at box. Result - 0 ohms
At outlet, connect black and copper with wire nut.
Test resistance of black and copper at box. Result - 0 ohms

So we have no stray voltage (measured with my meter), no short in the wires and no break in the wires.

Didn't try hooking to a non-GFCI temporarily, because the test for short seemed to rule that out.

What do you make of all this?

 
  #12  
Old 01-30-01, 08:58 AM
s1nuber
Visiting Guest
Posts: n/a
I would recommend that you try the standard circuit breaker connected to a standard receptacle for testing purposes only. If this works, then install a GFCI receptacle at the tub location, leaving the standard breaker in place. This will bypass your minor ground fault situation in your wiring, while still providing GFCI protection to the tub. The only other solution I can think of is to replace your wiring.
 
  #13  
Old 01-30-01, 09:21 AM
Guest
Visiting Guest
Posts: n/a
OK, but...

What is a "minor" ground fault? I had assumed that such a condition was binary; either you had one or you didn't.

Can you describe how this condition coule exist? Why would I not detect the condition with my ohmmeter?
 
  #14  
Old 01-30-01, 12:03 PM
abNORMal's Avatar
Member
Join Date: Nov 2000
Location: New Jersey USA
Posts: 535
Received 0 Votes on 0 Posts
Lightbulb

When you tested with the new breaker (assuming it's CGFI), was the outlet reinstalled, and grounded at the tub end? Otherwise it would trip for an open ground.

If so, then this can only be a "bad" breaker. Your tests prove the wire to be good.

Is it possible you have two bad breakers? Try them in another circuit that works now.
 
  #15  
Old 01-31-01, 06:54 AM
s1nuber
Visiting Guest
Posts: n/a
BillJ-

Good questions. Let me clarify.

A GFCI device (breaker or receptacle) typically measures a difference in hot current versus nuetral current. If this value exceeds 5 milliamps (.005 amps), then the downstream circuit is shut down.

This makes GFCI devices extremely sensitive to any type of fault. A 5 milliamp fault will not be detected by a typical circuit breaker, or your multimeter. The reason for this is because a .005 amp falut is below the level that a regular breaker or a typical multimeter can detect (this level is likely smaller than your meter's accuracy). This does approach the level where a human being can be killed, however.

The engineers say .01 amps across a human heart for one second can kill. .005 amps is half of that amount, and so GFCI devices are thought to be very safe. This is why they are used near water, where the dangerous aspects of electricity are magnified. If GFCI protection were installed on every residential circuit, many circuits would not function.

Some reasons for this are material manufacturing defects, poor installation, and poor job site procedures. For example, a cable is only allowed to be bent at a radius of 5 times the wire's diameter, but I have seen many cases where cable is caught on a drilled hole during installation, kinking the cable in half. This would cause a breakdown in the wire's insulation, but not necessarily a direct ground fault. This minor defect would be seen by a GFCI, but not a typical circuit breaker.

This is the type of problem that you have. It could have been caused at the manufacturing facilty that made your wire, I cannot tell you for sure. If you install a regular circuit breaker to provide overcurrent and ground fault protection for the circuit, then install GFCI protection at the outlet for the load (tub), you will have met all NEC requirements, as well as providing a safe installation.

One other thing to point out, you should install a 20 amp GFCI receptacle if your tub is on a 20 amp circuit.

If you don't feel comfortable with this solution (keep in mind it meets all code and safety requirements), your only other choice as I see it is to replace the circuit wiring.

P.S. - one other field test that you can try. Use the GFCI breaker and run new wiring from the breaker to your tub location through your house. Install the load end into a spare plastic box with a receptacle. Keep your kids and animals away from the wire when you do this. Plug in your tub and turn it on. This will tell you if your breaker is adequate, and if the fault is truly in your wire. Be sure to use new wire, right off the spool/out of the box!

Enjoy your day!

 
  #16  
Old 01-31-01, 07:04 AM
Guest
Visiting Guest
Posts: n/a
Great answer. I can understand that, and you answered another question that had been in my head for a while, which was: "If GFCI is so great, why isn't it the standard for all circuits". Now I understand why it is not.

I will go with the GFCI (20A) at the whirlpool, and replace the GFCI breaker with a standard 20A. Thanks for the help.

Replacing the wire would have been a very expensive option.
 
  #17  
Old 01-31-01, 10:40 AM
J
Member
Join Date: Sep 2000
Location: Fayetteville, NY, USA
Posts: 1,052
Received 1 Vote on 1 Post
s1nuber strikes again with a remarkable and remarkably clear explanation. Nice job!

Juice
 
  #18  
Old 01-31-01, 11:20 AM
Guest
Visiting Guest
Posts: n/a
Talking

Solution is in place and works like a champ.

Thanks to all you folks who contributed to the solution, Juice, Wg, and s1nuber.
 
  #19  
Old 01-31-01, 11:24 AM
J
Member
Join Date: Sep 2000
Location: Fayetteville, NY, USA
Posts: 1,052
Received 1 Vote on 1 Post
Very, very glad to hear it. And thank you for letting us know the results. You deserve a long, hot bath!

Juice
 
 

Thread Tools
Search this Thread
 
Ask a Question
Question Title:
Description:
Your question will be posted in: