GFCI puzzling me

Status
Not open for further replies.
S

ShockedMonkey

Member
Location
North SF Bay Area, CA
I have a GFCI receptacle and outlets fed on the load side (as well as a garage on the same 20 amp circuit) that a tester won't trip, don't know why.

Details:
20 amp 120 vac circuit off of older service, with a cold water ground to the first GFCI outlet (GFCI 'A' let's say.)

Another 3 receptacles are on load side, as well as a 20amp 120vac single-breaker box in garage. The breaker box has an isolated neutral bar and can grounded ground bus bar (unsure if it has a ufer, or anything and the "installer" has all kinds of hokey-non-sparky creations :roll:

Additional GFCI receptacles on the load side of the garage breaker trip via tester.

GFCI-A is good and will trip via it's 'test' button. It will also trip if I jump the isolated neutral bar with the ground bar in garage.

I thought it was something wonky at the garage, but I lifted the wires off all the bars to isolate it out of the circuit and GFCI still won't trip via tester.

GFCI tester works and trips other GFCI's in house that have a ground (old house not all have bare-wire.)

Any ideas why I can't get the tester to trip any outlets on GFCI-A's load side? I've only had a tester not trip a GFCI, if it didn't have an ECG.

Thanks
 
Gfci testers will not operate without an equipment grounding conductor. It works by testing to the equipment grounding conductor. So if you lift the equipment grounding conductor from the circuit the tester won't work however the gfci will still have some protection and will operate without the equipment grounding conductor.

The tester is the problem because of how it functions.
 
ShockedMonkey said:
GFCI tester works and trips other GFCI's in house that have a ground (old house not all have bare-wire.)

Any ideas why I can't get the tester to trip any outlets on GFCI-A's load side? I've only had a tester not trip a GFCI, if it didn't have an ECG.
Click to expand...

Dennis Alwon said:
Gfci testers will not operate without an equipment grounding conductor. It works by testing to the equipment grounding conductor. So if you lift the equipment grounding conductor from the circuit the tester won't work however the gfci will still have some protection and will operate without the equipment grounding conductor.

The tester is the problem because of how it functions.
Click to expand...
I agree with Dennis.

You can still use your tester, but, for receptacles that show they do not have an EGC, you'll have to provide an EGC. One way to do this would be to have a simple three-prong grounding to two-prong ungrounded cheater prewired with fifty feet or so of #14 that you can stretch out and connect to a "known ground". Plug your GFI tester into the cheater with the known ground and then plug the two-prong end into the receptacle without an EGC. Operate your tester, and it should trip GFCI "A" if wired correctly to the LOAD side of the GFCI.
 
ShockedMonkey said:
I have a GFCI receptacle and outlets fed on the load side (as well as a garage on the same 20 amp circuit) that a tester won't trip, don't know why.

Details:
20 amp 120 vac circuit off of older service, with a cold water ground to the first GFCI outlet (GFCI 'A' let's say.)

Another 3 receptacles are on load side, as well as a 20amp 120vac single-breaker box in garage. The breaker box has an isolated neutral bar and can grounded ground bus bar (unsure if it has a ufer, or anything and the "installer" has all kinds of hokey-non-sparky creations :roll:

Additional GFCI receptacles on the load side of the garage breaker trip via tester.

GFCI-A is good and will trip via it's 'test' button. It will also trip if I jump the isolated neutral bar with the ground bar in garage.

I thought it was something wonky at the garage, but I lifted the wires off all the bars to isolate it out of the circuit and GFCI still won't trip via tester.

GFCI tester works and trips other GFCI's in house that have a ground (old house not all have bare-wire.)

Any ideas why I can't get the tester to trip any outlets on GFCI-A's load side? I've only had a tester not trip a GFCI, if it didn't have an ECG.

Thanks
Click to expand...

Check the ground at one of these receptacles where you are trying to use the tester. The easiest way is probably to check continuity to the neutral wire in the same junction box. I think you will find a poor connection somewhere along the line.

Even if the plug in tester shows that an EGC is present there can still be a poor connection.

The moral of the story is don't trust plug in testers. They are good for some things but when it comes to troubleshooting it's better to use more reliable test equipment.
 
These all have an ecg, and the tester (ideal plug-in type) shows 2 green lights, which = correct polarity with ground. I also get 120 to ground same as to the neut. 0 volts between neut and ground.

Layout is:

Service 20amp breaker -> burial box (ecg is pigtailed) -> GFCI outlet -->load side of GFCI -->back to same burial box --> duplex outlets --> 20amp breaker box garage that has isolated neutral bus and ground bar similar to a sub-panel/breaker disconnect --> garage GFCI's (they tester-trip, but won't trip 1st gfci in the circuit)

So the first GFCI and load side outlets won't tester trip, but those on the load side of the garage 20 amp breaker panel will.

It's wonkey. I unearthed 4 pvc conduit body 'T's (drilled in the bottom for "drainage" God love him) and replaced them with in-ground boxes. It was like digging for buried treasure. It would have been a nice bill -- except it's my house. :rant: I re-pulled everything from the service to the garage LB (don't know what happens between LB and 20 amp disconnect.)

Assuming I have a good ECG and polarity is all correct: is there anything else that that could cause it not to trip?
 
ShockedMonkey said:
These all have an ecg, and the tester (ideal plug-in type) shows 2 green lights, which = correct polarity with ground. I also get 120 to ground same as to the neut. 0 volts between neut and ground.

Layout is:

Service 20amp breaker -> burial box (ecg is pigtailed) -> GFCI outlet -->load side of GFCI -->back to same burial box --> duplex outlets --> 20amp breaker box garage that has isolated neutral bus and ground bar similar to a sub-panel/breaker disconnect --> garage GFCI's (they tester-trip, but won't trip 1st gfci in the circuit)

So the first GFCI and load side outlets won't tester trip, but those on the load side of the garage 20 amp breaker panel will.

It's wonkey. I unearthed 4 pvc conduit body 'T's (drilled in the bottom for "drainage" God love him) and replaced them with in-ground boxes. It was like digging for buried treasure. It would have been a nice bill -- except it's my house. :rant: I re-pulled everything from the service to the garage LB (don't know what happens between LB and 20 amp disconnect.)

Assuming I have a good ECG and polarity is all correct: is there anything else that that could cause it not to trip?
Click to expand...
There are two things that would cause the portable tester to fail to trip the GFCI:

1. No current is being drawn from the receptacle hot when you push the button and connect a resistor from receptacle hot to receptacle EGC. This is a fairly unlikely occurrence given that the correct lights show on the tester.
2. The current drawn by the tester is actually returning to the GFCI over the load neutral terminal.
One way this could happen is if the receptacle has a "bootleg" ground which just connects the EGC terminal of the receptacle to the neutral wire.
A more complicated path to the same result would be if there is an indirect connection of any sort from the receptacle EGC back to the GFCI neutral load terminal which has a lower resistance than the EGC connection back to the main panel (service) bond point. That will reduce the amount of current that bypasses the GFCI neutral to less than the trip threshold.

In any case, it is not compliant with current code to omit an EGC connection (conduit or wire) from the garage back to the main panel or to omit the EGC downstream in the circuit and instead connect to a local ground like the water pipe.
 
dibloafer said:
Sounds like your missing an equipment ground, or have a poor one, (which is entirely possible, given your scenario) somewhere in this part of your ckt.
Click to expand...

It sounds like the grounding conductors in the garage are not connected to the grounding system in the house where the garage is fed from. Instead, the earth is being used as a path back to the source.
 
It sounds like nothing is necessarily wrong to me. You described having two GFCI's connected in series and have described successfully tripping both of them. But when you put an actual fault condition on the load end of the second one, it apparently has a faster response time then the first and fault current is cleared by the second one before the first one trips.

Personally I would eliminate or bypass GFCI protection through the first device, or eliminate the second device and only rely on the first device for protection, creates less confusion when devices do trip to only have one potential place to go to reset things, especially if this were for a client instead of for yourself.
 
Thank you all for the replies.

A bootleg ground may be the cause. I'll eliminate the entire garage from the circuit by opening it at the LB, as I can't be sure what was done in the walls.

I have a continuous ECG, unless there was a factory break in the wire I pulled. I'll ring it out to be certain.

Agreed, the 2nd GFCI receptacle in detached garage is redundant, but existing. The first GFCI in the line needs to be there for the exterior (load side) receptacles. I was going to keep it for easy reset, if a power tool caused a nuisance trip (less walking.)

I'm questioning if I've ever understood how GFCI's work. I understood them to measure the current between the hot and neutral and open if the current differential passed a set amount of current. Is this correct?

Thanks
 
ShockedMonkey said:
...
I'm questioning if I've ever understood how GFCI's work. I understood them to measure the current between the hot and neutral and open if the current differential passed a set amount of current. Is this correct?

Thanks
Click to expand...
That is correct and the built in test button tests that circuit by connecting a resistor from the hot on the line side of the current sensor to the neutral on the load side. When you push the test button, you create the current differential required to trip the GFCI. If the test button works and the tester does not, there are only two possibilities, first the tester is not working, or second there is no equipment ground at the receptacle that has a path back to the main bonding jumper.
 
don_resqcapt19 said:
That is correct and the built in test button tests that circuit by connecting a resistor from the hot on the line side of the current sensor to the neutral on the load side. When you push the test button, you create the current differential required to trip the GFCI. If the test button works and the tester does not, there are only two possibilities, first the tester is not working, or second there is no equipment ground at the receptacle that has a [low resistance] path back to the main bonding jumper [other than through the line side neutral terminal].
Click to expand...
FIFY
 
ShockedMonkey said:
Thank you all for the replies.

A bootleg ground may be the cause. I'll eliminate the entire garage from the circuit by opening it at the LB, as I can't be sure what was done in the walls.

I have a continuous ECG, unless there was a factory break in the wire I pulled. I'll ring it out to be certain.

Agreed, the 2nd GFCI receptacle in detached garage is redundant, but existing. The first GFCI in the line needs to be there for the exterior (load side) receptacles. I was going to keep it for easy reset, if a power tool caused a nuisance trip (less walking.)

I'm questioning if I've ever understood how GFCI's work. I understood them to measure the current between the hot and neutral and open if the current differential passed a set amount of current. Is this correct?

Thanks
Click to expand...
If a power tool causes a trip there is no guarantee which GFCI will trip or even is possible that both will trip.

You are mostly correct about how they work. If you clamp an ammeter around both conductors of a two wire circuit you typically get no reading, because same current flowing opposite directions in each conductors has a cancelling effect on the magnetic field the ammeter will be reading. GFCI's do the same thing, as long as current through a CT is same on both conductors there is even canceling of the fields, the sensitivity of these is high enough they will detect 4-6 mA of differential and that initiates operation of the trip mechanism.
 
Appreciate the clarification everyone. Thank you.

kwired said:
If a power tool causes a trip there is no guarantee which GFCI will trip or even is possible that both will trip.

You're absolutely correct KW. I was thinking of it as a breaker, but I have experienced simultaneously tripped GFCI's in the past.

My mind is grasping at straws after being humbled/humiliated by the lowly GFCI.

Wife did her usual "Do we need to call an electrician?" joke, and it wasn't funny this time. :D

I'll see if I can isolate it sometime this week.

Thanks
Click to expand...
 
Follow-up, hopefully save someone else some grief.

Tester was only tripping the old (approx 10years) GFCI's but not the new ones. A $10 tester would have saved me a ton of time, but it was test/tripping the existing GFCI's so -- it works I thunked. :roll:

I'm curious if the newer GFCI's take a larger fault to trip them. The tester is 15+ years old. I'll compare the resistance between the two if I get time.

The good news is my circuit and ECG checked out perfectly -- several times :D Caused me to question everything I thought I knew about electricity, sheesh.
 
ShockedMonkey said:
Follow-up, hopefully save someone else some grief.

Tester was only tripping the old (approx 10years) GFCI's but not the new ones. A $10 tester would have saved me a ton of time, but it was test/tripping the existing GFCI's so -- it works I thunked. :roll:

I'm curious if the newer GFCI's take a larger fault to trip them. The tester is 15+ years old. I'll compare the resistance between the two if I get time.

The good news is my circuit and ECG checked out perfectly -- several times :D Caused me to question everything I thought I knew about electricity, sheesh.
Click to expand...
The older GFCIs may have been tripping at a lower fault due to age.
 
ShockedMonkey said:
...
I'm curious if the newer GFCI's take a larger fault to trip them. ...
Click to expand...
The standard for the trip value has always been the same as far as I know. Must not trip at less than 4 mA, is permitted to trip between 4mA and 6mA, and must trip above 6mA.

However there is also a maximum permitted time to trip. The maximum permitted time to trip in seconds is equal to the quantity (20/fault current in milliamps) raised to the 1.43 power. The application of this formula would permit a 7.25 second trip time for a 5 mA ground fault.

At 10 mA it is permitted to take almost 3 seconds before it trips. At 12 mA it could be a bit over 2 seconds. If you have a device that is operating at the high end of the permitted trip time, you would have to hold the test button in for that amount of time to make the GFCI trip.

Most GFCIs trip in a much shorter time than the maximum permitted by the standard.
 
Status
Not open for further replies.
Top