GFCI INTERMITTENT TRIP

[mtnelect]

Maybe not...the Technical Committee for UL 943 was presented with evidence that power conversion equipment used in refrigeration and cooling systems can generate high frequency leakage current that can result in a listed GFCI tripping with the high frequency leakage current less than 5 mA. The TC voted not to even look at that issue. For the record, I sit on that TC and did not vote with the majority on this issue.

Then the issue was "Resolved".

 

[don_resqcapt19]

Then the issue was "Resolved".

UL Standards Panels do not use the same terms as the NFPA does. There is supposedly a new proposal in the works, but have not seen it yet.

 

[jim dungar]

But the tripping is still caused by the equipment. There is nothing the electrical contractor can do about it, they must follow the NEC.

It is a separate discussion whether or not the NEC requirements are valid, based on the industry product standards.

 

[synchro]

...the Technical Committee for UL 943 was presented with evidence that power conversion equipment used in refrigeration and cooling systems can generate high frequency leakage current that can result in a listed GFCI tripping with the high frequency leakage current less than 5 mA. The TC voted not to even look at that issue. For the record, I sit on that TC and did not vote with the majority on this issue.

Another factor is that the human body becomes less responsive to currents at frequencies above 60Hz, and so in my opinion a 5mA threshold independent of frequency is being too strict from a safety point of view. The plot below represents "average values for the threshold of sensation and the can't let go current as a function of frequency", and is from the webpage https://www.texasgateway.org/resource/36-electric-hazards-and-human-body .

 

[Jpflex]

EMC is Electromagnetic Compatibility. A generic term used to explain how different pieces of electric/electronic equipment should behave in order to prevent electrical interference between them ( to be electromagnetically compatible with each other).
GFCI devices will all false, or nuisance trip, if exposed to enough electrical interference. When they are subjected to enough radiated or conducted electromagnetic interference (EMI), currents flow through unintended paths within the internal circuitry. Those currents then trigger the SCR that trips the unit. This EMI essentially bypasses the differential current transformer that normally discriminates the 5ma differential current.

If you used a pico milliamp current probe and oscilloscope directly on the EGC, of small enough resolution, and measured nothing, then you may have an EMC issue. Did you monitor the current at the moment that the trip occurs?
If it is an EMC issue you could try inserting an EMI filter between the GFCI and the refrigerator as a diagnostic step.

Yes I watched the circuit several times before and during trip. I repeated test with same results, however only one time I think I may have spotted 0.18 amperes on ground instead of meter always showing 0.02 to 0.04 as default current on EGC

 

[Jpflex]

Another factor is that the human body becomes less responsive to currents at frequencies above 60Hz, and so in my opinion a 5mA threshold independent of frequency is being too strict from a safety point of view. The plot below represents "average values for the threshold of sensation and the can't let go current as a function of frequency", and is from the webpage https://www.texasgateway.org/resource/36-electric-hazards-and-human-body .

So what are electricians supposed to do? Follow code and have to return to customers without being able to alleviate this problem because AHJ won’t address it. Meanwhile the electrician gets called into BAR licensing with upset customers

 

[Jpflex]

Take the GFCI out and use a standard receptacle. No problems. Imagine that.

Imagine taking the GFCI out and violating NEC code?

 

[Jpflex]

Imagine taking the GFCI out and violating NEC code?

Work has to pass inspection

 

[Jpflex]

I don;'t think it needs rocket science technology ..you have 3 basic culprits,,,the microwave, the refrigerator, a circuitry problem.
Most feel the refrigerator is the source, Run the refrig (only) for a time and see if it trips... if not, try the same with the micro.

I didn’t say it was rocket science, as I stated before repeatedly, it is the refrigerator causing the problem.

The issue is trying to find the equipment, method or capability to pinpoint the leakage current to help some of us here who may come across a similar GFCI tripping service call

 

[arhythm]

What about pig-tailing the gfci and not going out of the load side to continue the circuit with the gfci? Unless you need the gfci to protect outlets down the line.

 

[ruxton.stanislaw]

I would suggest a cord and plug, UL approved isolation transformer to avoid tripping the GFCI while still remaining code compliant.

 

[kwired]

But the tripping is still caused by the equipment. There is nothing the electrical contractor can do about it, they must follow the NEC.

It is a separate discussion whether or not the NEC requirements are valid, based on the industry product standards.

But if it is because of a variable speed driven compressor the manufacturer can do something about it. I've run into this before with clothes washers with variable speed motors. The manufacturer electrically isolates the motor frame from the machine frame and equipment ground conductor - that solves the GFCI trip issues. But they may put a switch in an EGC to the motor frame that is "held open" as long as the back cover is in place. Had two different ones that started tripping GFCI's. Appliance repair guys in both cases said something wrong with the GFCI. I took back cover off and found broken mount on said "grounding switch" and unplugged lead to the motor frame, tried running it and was fine.

Another thing OP can try (temporarily) is plug the appliance in via one those two to three prong cheaters, I bet it holds that way, but one should still try to find out exactly what is causing the tripping and not leave that as a permanent solution. All this does is help show that something is finding a way back to the source via the EGC.

 

[Jpflex]

I would suggest a cord and plug, UL approved isolation transformer to avoid tripping the GFCI while still remaining code compliant.

How does this work?

transformers normally step up or step down voltage but if voltage available is only secondary 120v how would this work and how would this isolation system work since the GFCI would appear to still trip if there were an imbalance current on line and neutral?

 

[Jpflex]

But if it is because of a variable speed driven compressor the manufacturer can do something about it. I've run into this before with clothes washers with variable speed motors. The manufacturer electrically isolates the motor frame from the machine frame and equipment ground conductor - that solves the GFCI trip issues. But they may put a switch in an EGC to the motor frame that is "held open" as long as the back cover is in place. Had two different ones that started tripping GFCI's. Appliance repair guys in both cases said something wrong with the GFCI. I took back cover off and found broken mount on said "grounding switch" and unplugged lead to the motor frame, tried running it and was fine.

Another thing OP can try (temporarily) is plug the appliance in via one those two to three prong cheaters, I bet it holds that way, but one should still try to find out exactly what is causing the tripping and not leave that as a permanent solution. All this does is help show that something is finding a way back to the source via the EGC.

Very interesting info you have. I was not aware that cloth washers may have an NC isolation switch that opens when cover is in place.

Does this mean there is no EGC bonding between motor and appliance frame while motor is in operation?

There was a time when I used to take apart washers and driers and study their diagrams. I came across one that apparently transformed AC signal into DC again into what the manual called SIMULATED AC. On the scope, the voltage on the field windings looked like square wave version of a standard sinusoidal ac wave form with voltage having positive and negative polarity.

Has anyone come across this or know more about this?

 

[Jpflex]

What about pig-tailing the gfci and not going out of the load side to continue the circuit with the gfci? Unless you need the gfci to protect outlets down the line.

Perhaps but are you allowed to feed the line side of GFCI and put another set of wires on same line side terminals to power other outlets? It doesn’t seem like you shouldn’t in NEC but wouldn’t GFCI still trip since GFCI is monitoring line and neutral current on both receptacles top or bottom?

 

[ptonsparky]

Perhaps but are you allowed to feed the line side of GFCI and put another set of wires on same line side terminals to power other outlets? It doesn’t seem like you shouldn’t in NEC but wouldn’t GFCI still trip since GFCI is monitoring line and neutral current on both receptacles top or bottom?

The GFCI device only monitors the load out. A ground fault on the line side will not affect the devices operation or trip it.

 

[Jpflex]

The GFCI device only monitors the load out. A ground fault on the line side will not affect the devices operation or trip it.

If GFCI only measures load out, how would it compare this current out to current in from hot wire.

According to a book I have, a GFCI uses a current transformer with both line and neutral wires going through it. When current is balanced between neutral and line hot, the magnetic fields are canceled and no voltage is output on current transformer but when a fault is present line has more current than neutral such as little as 4 mili amperes and a voltage is output from internal current transformer which then trips to disrupt line power to GFCI

 

[retirede]

If GFCI only measures load out, how would it compare this current out to current in from hot wire.

According to a book I have, a GFCI uses a current transformer with both line and neutral wires going through it. When current is balanced between neutral and line hot, the magnetic fields are canceled and no voltage is output on current transformer but when a fault is present line has more current than neutral such as little as 4 mili amperes and a voltage is output from internal current transformer which then trips to disrupt line power to GFCI

Other than a tiny bit to run the electronics, the current in has to equal current out. I don’t understand your question.

Your second paragraph is essentially correct.

 

[Jpflex]

What I mean was that I did not understand what you meant by the GFCI only measures load out?

My point was that not just output but input current is measured by GFCI

 

[ptonsparky]

What I mean was that I did not understand what you meant by the GFCI only measures load out?

My point was that not just output but input current is measured by GFCI

Only the imbalance of the current flowing through the GFCI is measured.

The current of the branch circuit feeding multiple pigtailed devices could be imbalanced.