refrigerator tripping GFI

Status
Not open for further replies.

DaveBowden

Senior Member
Location
St Petersburg FL
Just finished a job where HO wanted 2 prong outlets replaced with 3 prong ones. House was built in 1954 and there was no ground wire in the romex.
We installed GFI recept's on the first outlet in each circuit and put the GFI protected No Equipment ground stickers on all the new receptacles we put in. The refrigerator is the last outlet on one of the living room recept circuits.
Plugging in the refrigerator causes the GFI to trip immediately. Plugging it in to another circuit trips that circuit's GFI immediately. We connected drills, lamps, and a heat gun to both of these circuits and they all worked fine.
My problem is convincing the HO that his refrigerator has a problem. I told him we could put GFI recepts in each outlet on the circuit with the fridge so it would not be GFI protected, but that it would not be safe as far as him (or his renter) not getting shocked. He thinks I should do this for free since the fridge worked fine before the GFI's went in.
I'm thinking I'm not going to do it at all until we plug the fridge into a non-GFI protected outlet, wet his concrete floor, and he grabs the fridge while he's barefoot. Just kidding about that part, but not about the not doing it part.
Any reasons I'm unaware of that the fridge would trip a GFI.
There is no continuity between the ground prong on the fridge cord and either hot or neutral prong.









w
 

ceb58

Senior Member
Location
Raeford, NC
Just finished a job where HO wanted 2 prong outlets replaced with 3 prong ones. House was built in 1954 and there was no ground wire in the romex.
We installed GFI recept's on the first outlet in each circuit and put the GFI protected No Equipment ground stickers on all the new receptacles we put in. The refrigerator is the last outlet on one of the living room recept circuits.
Plugging in the refrigerator causes the GFI to trip immediately. Plugging it in to another circuit trips that circuit's GFI immediately. We connected drills, lamps, and a heat gun to both of these circuits and they all worked fine.
My problem is convincing the HO that his refrigerator has a problem. I told him we could put GFI recepts in each outlet on the circuit with the fridge so it would not be GFI protected, but that it would not be safe as far as him (or his renter) not getting shocked. He thinks I should do this for free since the fridge worked fine before the GFI's went in.
I'm thinking I'm not going to do it at all until we plug the fridge into a non-GFI protected outlet, wet his concrete floor, and he grabs the fridge while he's barefoot. Just kidding about that part, but not about the not doing it part.
Any reasons I'm unaware of that the fridge would trip a GFI.
There is no continuity between the ground prong on the fridge cord and either hot or neutral prong.









w
What you are seeing is common for older refg. to do this. They can leak a small amount of current, enough to trip a GFCI, in the compressor. You can try going back to the panel and installing a GFCI breaker, If it has one and not a screw in fuse panel, they have a higher trip limit. Or run a dedicated circuit with ground just for the refg. Or tell him to get the refrigerator fixed or buy a new one.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
There is no continuity between the ground prong on the fridge cord and either hot or neutral prong.

The difference in current between the hot line and the neutral/grounded line has to be going somewhere.
The two possibilities are to an incidental ground like a copper water supply line for water dispenser or ice maker, or like conduction through the legs to the floor, or else
Capacitance of the entire appliance shell. Assuming that the ground wire in the cord is bonded to the metal of the fridge, you should be able to measure the voltage (mostly irrelevant but impressive to the owner) and the current (hopefully a small number of milliamps) with a meter connected to any nearby ground like a water pipe. Or you can just meter from fridge chassis bare metal or bare screw in sheet metal to ground. That may convince the homeowner that the fridge has a problem.
 

GerryB

Senior Member
My problem is convincing the HO that his refrigerator has a problem. I told him we could put GFI recepts in each outlet on the circuit with the fridge so it would not be GFI protected, but that it would not be safe as far as him (or his renter) not getting shocked. He thinks I should do this for free since the fridge worked fine before the GFI's went in.

I don't think you can do what you say above. If the outlet is ungrounded you have to put a gfi or a 2-prong receptacle.









w[/QUOTE]
 

growler

Senior Member
Location
Atlanta,GA
What you are seeing is common for older refg. to do this.



I have even run into a couple of new models where the manufacturer recommends not useing GFCI protection.


The way I see it there are two options.

1. The owner buys a new refrigerator (hopes for the best ).

2. The owner pays to have a new refrigerator circuit run. At least it will be grounded and the breaker won't trip and spoil their food.

I normally just explain to the owner that refrigerators may not work on GFCI and let them know their options and that I don't do anything for free.
 

PetrosA

Senior Member
Why is it that people have such low expectations of their refrigerators until the refrigerator gets really, really old. Then, suddenly, it should work perfectly! He'd have thrown out a computer half the age of the fridge without a second thought...
 

hurk27

Senior Member
Just finished a job where HO wanted 2 prong outlets replaced with 3 prong ones. House was built in 1954 and there was no ground wire in the romex.
We installed GFI recept's on the first outlet in each circuit and put the GFI protected No Equipment ground stickers on all the new receptacles we put in. The refrigerator is the last outlet on one of the living room recept circuits.
Plugging in the refrigerator causes the GFI to trip immediately. Plugging it in to another circuit trips that circuit's GFI immediately. We connected drills, lamps, and a heat gun to both of these circuits and they all worked fine.
My problem is convincing the HO that his refrigerator has a problem. I told him we could put GFI recepts in each outlet on the circuit with the fridge so it would not be GFI protected, but that it would not be safe as far as him (or his renter) not getting shocked. He thinks I should do this for free since the fridge worked fine before the GFI's went in.
I'm thinking I'm not going to do it at all until we plug the fridge into a non-GFI protected outlet, wet his concrete floor, and he grabs the fridge while he's barefoot. Just kidding about that part, but not about the not doing it part.
Any reasons I'm unaware of that the fridge would trip a GFI.
There is no continuity between the ground prong on the fridge cord and either hot or neutral prong.w

Make sure that you do not have a reversed polarity in the circuit somewhere, many times I have run into where old non-grounded circuits that had the cloth covered rubber insulation it was very hard to tell which wire was the hot or the neutral, especially at the lights where the heat from over lamping the fixture would bake the color out of the conductors, someone comes along as installs a new fixture and pulls apart the conductors but doesn't remark them and the hot and neutral gets switched, while this shouldn't cause this problem but if you bootlegged the ground terminal off the neutral or what you thought was the neutral then now the ground prong on the receptacle is hot in reference to earth so the case of the appliance is hot and since it is making connection to a concrete floor the GFCI does what it is supposed to do trip the GFCI.
Because you stated that you measured no continuity between the ground prong and the hot or neutral on the cord of the appliance then I suspect that the ground prong on the receptacle would have to be giving reference to another conductor, even using a GFCI you are not allowed to bootleg the ground to the neutral and is a very dangerous mistake to make, if the neutral and hot are swapped ahead of this receptacle you would not know it as AC doesn't care if only two conductors are referenced it will still show correct polarity, now if you run an extension cord to a truly grounded receptacle that maybe is close to the panel that can be verified at the panel or even temporally install a receptacle off the panel to use as a test reference, then plug in a long enough extension cord to reach the receptacles in question so you can use it as a reference to check which conductor is the true hot and which is the true neutral, verify if the neutral and hot has not been reversed, this and only this is the best way to make sure you are not going to have a dangerous problem after you do an update like this.

Also make sure you or your workers did not bootleg the ground prong to the neutral at this or any receptacles, even using a GFCI will not make this a safe installation and is a very big code violation that some thing is allowed if the GFCI is used, I have electricians even think this was allowed even when a GFCI is not used.


I had a service call to a house that an elderly lady was getting shocked if she touched the freezer in her garage, about a month before I was called she had a friend run a circuit to the garage to feed this freezer, the house had old cloth UN-grounded type circuits and he ran the new NM to a receptacle in the crawl space that was for the sump pump, he had mistakingly reversed the polarity of the old cloth covered wire on this receptacle then bootlegged the ground to what he thought was the neutral at the sump pump receptacle, after checking with an extension cord to a receptacle I pig tailed into the panel to make sure I had a good reference point, I found that the freezer had 120 volts on its case, upon checking the receptacle both the neutral and the ground pins both were hot, I removed the receptacle to find it wired correctly, but the ground and neutral wire was still hot and the hot showed as a neutral, I traced the new NM into the crawl space to this pump receptacle after pulling it apart I saw what had happened, he had unknowingly reversed the hot and neutral when he tied this new NM into the old conductors, he had connected the EGC, and the new neutral to what he thought was the neutral in the old NM, I even had a hard time telling which one was which but the extension cord made it very clear.

The problem with this vary dangerous mistake is that the GFCI or a 3-light plug tester can not tell you that you have a reversed polarity if the EGC or ground prong on the receptacle is also bonded to the neutral at the receptacle, it will light up the lights that show the receptacle is correctly wired, but the danger just sits there in wait for a person to make contact with this hot EGC and a true path to another grounded appliance or Earth like the concrete floor.

So I would go back and run the test using the extension cord to a known good reference receptacle or if you don't have one pig tail one into the panel temporally, then check this circuit with a voltage tester from the ground pin or the neutral pin to the suspect neutral to see if it shows hot, if so find where the neutral and ground has been reversed, other wise if someone was to get hurt you could be found liable even if it was reversed before you touched anything, also never bootleg the ground pin off the neutral even when using a GFCI, because if the neutral was ever lost it can cause any appliance that is plugged into any receptacle wired like this that has a grounding type plug that grounds the case of that appliance to be hot to earth even if the circuit has the correct polarity, this is why we never are allowed to do this.

Also I should add tht the reason I suspected the bootleg ground to neutral was that you made the statment that you checked the continuity to th eground pin on the fridgorators plug, if these is no wires connecting to the ground terminal at the receptacle then a referance of the ground pin on the fridges plug is meaning less.

Also keep in mind that it takes an impedance of 20k to 25k ohms or less to trip a GFCI most common continuity testers will only show continuity if the impedance is lower then 1k ohms, some even less, so a leakage could get over looked, using a megger would be the best way to test this.
 
Last edited:

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
never bootleg the ground pin off the neutral even when using a GFCI, because if the neutral was ever lost it can cause any appliance that is plugged into any receptacle wired like this that has a grounding type plug that grounds the case of that appliance to be hot to earth even if the circuit has the correct polarity, this is why we never are allowed to do this.
That is why the correct procedure, if there is no suitable ground present, is to leave the ground terminal open and attach a "No Ground Present" label to the face of the receptacle.
Better to have no ground and know it than to have an undependable or dangerous ground, as your examples clearly illustrate.
 

DaveBowden

Senior Member
Location
St Petersburg FL
There is nothing connected to the ground terminal of the receptacle. The code allows a 3 wire receptacle on a two wire circuit if it is GFI protected, but specifically states a ground wire is NOT to be run from the GFI recept to any downstream recepts. If he wants the GFI protection removed, I will be re-installing a 2 wire recept on everything that doesn't have a ground or GFI protection.
As for the reversed polarity, this is cloth jacketed romex but the conductors are not cloth insulated. The plastic (I think) insulation is easily identified. We also tested the wires with a proximity- type tester (one that beeps when its near a hot without touching the wire) and it shows the hot as a hot and the neutral s a neutral. I think tomorrow when I plug his fridge into an extension cord plugged into a non-GFI protected outlet, that tester will beep when I hold it near the refrigerator.
Hopefully this problem won't prevent the owner from giving me the $1900 he owes me for what we've done so far.
 

hurk27

Senior Member
There is nothing connected to the ground terminal of the receptacle. The code allows a 3 wire receptacle on a two wire circuit if it is GFI protected, but specifically states a ground wire is NOT to be run from the GFI recept to any downstream recepts. If he wants the GFI protection removed, I will be re-installing a 2 wire recept on everything that doesn't have a ground or GFI protection.
As for the reversed polarity, this is cloth jacketed romex but the conductors are not cloth insulated. The plastic (I think) insulation is easily identified. We also tested the wires with a proximity- type tester (one that beeps when its near a hot without touching the wire) and it shows the hot as a hot and the neutral s a neutral. I think tomorrow when I plug his fridge into an extension cord plugged into a non-GFI protected outlet, that tester will beep when I hold it near the refrigerator.
Hopefully this problem won't prevent the owner from giving me the $1900 he owes me for what we've done so far.

Well from this info then I agree that the fridge must have some forum of leakage that the continuity tester doesn't show up, my last post was to just point out some common things I have found over the years that are easily missed, another method since there are no other paths other then the concrete floor that would provide the alternate path to trip the GFCI since there is no ground path at the receptacle, would be to keep the GFCI and have him insulate the feet of the fridge from the floor by some kind of insulating means unless there is a metal water line to the fridge for an ice maker if there is then it can be changed to plastic, at least this way if some one was to become part of the path of this leakage current then at least the GFCI is still in line to protect them and you have some what covered your liability.

FYI a GFCI does not have any connection internally from the ground terminal to any of the electronics that are used to detect a ground fault, it does it by the balance of the current on the neutral versus the current on the hot, this is why when one trips and there is no ground path to the EGC terminal back to the service neutral point, then the only other path is a water line or Earth which would have to be the concrete, this is also why a plug in GFCI tester will not work on a GFCI that has no ground wire connected to the ground terminal as they use the ground terminal for the test current path because unlike the test button on the GFCI which uses the line side neutral as the test current path.

I too hope he doesn't hold this against you for you have to follow what the code requires, homeowners don't understand that not only do you have to follow the code but to cover liability and safety issues as as you have to protect yourself and your company from law suits and even possibly criminal charges if someone was to be hurt of killed from actions you have done, lets say a child was to touch this fridge and was killed, they don't think we have liabilities but I would bet they would in a second go after one of us if something like that was to happen, I have had home owners want me to remove GFCI's where required but I walked from those jobs in a heart beat, because I know they can tell me all they want that they wouldn't sue or would say they would sign off, but you cant sign off on liability, so a job like that is just not worth taking the risk.

Wish you luck
 

jumper

Senior Member
There is a copper ice maker line from the refrigerator to the water heater, so it might not be leaking through the feet of the fridge.

Regardless of where the leakage is, unless there is an improper bond-which I believe you said you have discounted, if the fridge is bad - it is not your problem. Motors and Appliances work perfectly well on circuits with GFCI protection.
 

kwired

Electron manager
Location
NE Nebraska
Regardless of where the leakage is, unless there is an improper bond-which I believe you said you have discounted, if the fridge is bad - it is not your problem. Motors and Appliances work perfectly well on circuits with GFCI protection.
But it is his problem - the owner thinks nothing was wrong with the fridge, now he put this device that causes fridge to not work.

Test fridge with a megger instead of a typical multimeter.

Run a new circuit to fridge that includes EGC. If the fault is hot to ground it will be blowing fuses or breakers.

Try plugging the fridge into a circuit for a short time to a circuit with an equipment grounding conductor, maybe whatever fault is there will clear itself - sometimes this does work.
 

kwired

Electron manager
Location
NE Nebraska
What you are seeing is common for older refg. to do this. They can leak a small amount of current, enough to trip a GFCI, in the compressor. You can try going back to the panel and installing a GFCI breaker, If it has one and not a screw in fuse panel, they have a higher trip limit. Or run a dedicated circuit with ground just for the refg. Or tell him to get the refrigerator fixed or buy a new one.
4-6 mA from a receptacle type GFCI is different than 4-6 mA from a GFCI breaker?? The devices ability to filter out inductive kick back and other false trip conditions is more likely what is different about them.
 

ELA

Senior Member
Occupation
Electrical Test Engineer
I use a "current leakage tester" for this type of issue :slaphead:

This would allow you to directly demonstrate to the customer that his fridge leaks vs another that does not.
 

DaveBowden

Senior Member
Location
St Petersburg FL
Just got back from meeting HO. Plugged fridge into washing machine outlet (has EGC) and of course it worked fine. HO's 3 prong adaptor was still on the counter so I plugged fridge into that and then into extension cord to washer outlet - still worked fine and I got nothing from the frame of the fridge to the frame of the stove next to it. I did however get 120 volts from the hot lead of my cord (it was 1 with 3 outlets) to the frame of the fridge - with the ground prong on the cord isolated.
I think the leakage is downstream of some internal control on the fridge which has to be energized to work - either a solenoid or contactor type control - and that's why I got no continuity thru the cord cap.
The problem is going to be resolved by us going back next week to run a new circuit to the fridge (not for free obviously) and I got paid for work already done, so everything should be okay in the end.
 
Status
Not open for further replies.
Top