Samsung Refrigerator Tripping GFCI

[wwhitney]

The GFCI isn't measuring current on the EGC, but rather net current on the supply lines. Of course any capacitive leakage is ending up on the EGC or another path back to the transformer.

Agreed on how GFCIs function.

So I take your point to be that the idea of putting the ferrite core on the EGC will only work when the EGC is the only return path for leakage current. Which I was assuming, but I guess is sometimes not true. E.g. if the refrigerator happens to have internal metal water piping and has a metal water supply (pretty rare these days, I think), then that would be another path for leakage current, since the metal water supply system will be bonded, creating a ground loop.

Cheers, Wayne

 

[don_resqcapt19]

But in this case, if the issue is high frequency current noise on the EGC, and if a ferrite ring works by increasing the impedance at high frequency, while maintaining a low impedance at 60 Hz, then putting them on the EGC makes sense to me.

Cheers, Wayne

I don't see how the GFCI sees anything on the EGC. The internals only look at the hot and neutral. It sees the leakage current that has gone out to the load, but has not returned via the other circuit conductor.

 

[wwhitney]

I don't see how the GFCI sees anything on the EGC. The internals only look at the hot and neutral. It sees the leakage current that has gone out to the load, but has not returned via the other circuit conductor.

If the EGC is the only other return path, then a CT around all the supply conductors and the EGC must read zero. Therefore a CT around the supply conductors only will read a net current matching the current on the EGC. Therefore reducing the current on the EGC is the same as reducing the current imbalance on the supply conductors.

Cheers, Wayne

 

[ramsy]

Legrand has a "sensitive appliance" GFCI receptacle that they say helps eliminate the issue from power conversion equipment used in refrigerators.

Nice find. Thanks for posting

 

[kwired]

But in this case, if the issue is high frequency current noise on the EGC, and if a ferrite ring works by increasing the impedance at high frequency, while maintaining a low impedance at 60 Hz, then putting them on the EGC makes sense to me.

Cheers, Wayne

I guess that makes sense but you would need to have just one path or it won't work. Something with other grounded connections to it like a water pipe, gas pipe, etc. that is able to carry current probably won't work with just a ferrite ring on the EGC that is part of the supply conductors.

ETA: I see you kind of already addressed this in later post

 

[Doc13067]

What code cycle you on? 2023 you loose that allowance for if you are not within 6 ft or counter top. 2023 simply states all receptacles in a kitchen, 150V or less to ground, must have GFCI protection, in addition to the 6ft rule.

We’re on 2017

 

[JimsGT]

The receptacle is within 6 feet of the sink but would it even matter if the refrigerator is within 3 feet of the sink and not GFCI protected? I would think that a big metal item so close to the sink should be GFCI protected even if the receptacle is in the front yard.

Up to and including the 2020 Code, the closeness of the refrigerator to the sink is irrelevant. It‘s the receptacle to sink distance that counts. In the 2023 Code this is a moot point for residential as ALL kitchen receptacles have to be GFCI’d (and AFCI’d).

 

[mtnelect]

More information to follow. My initial research has focused on the refrigerant that flows in the system. A hermetic compressor has refrigerant circulating around the system together with the motor. The refrigerant has some conductivity. At this point, I think this is the problem.

 

[ptonsparky]

More information to follow. My initial research has focused on the refrigerant that flows in the system. A hermetic compressor has refrigerant circulating around the system together with the motor. The refrigerant has some conductivity. At this point, I think this is the problem.

Meg the dang thing!

 

[tortuga]

Sounds like these little residential VFD's need drive isolation transformers put back in their likely 'cost reduced' product designs.
Why do new fridges trip GFCI's?
Answer: Because fridges are designed for RCD's not GFCI's.

Total hack 'void the warranty' suggestion (dont really do this) but
I bet if you open up the back of the fridge and installed a 1:1 isolation transformer on the VFD input leads the problem would go away. Then you'd still have GFCI protection to the appliance including light and water dispenser and the just compressor would be isolated.

The SPGFCI has a 20 mA trip, but uses the same time to trip formula as does the Class A GFCI. One issue is that these are stand alone type devices and not available as a breaker or a receptacle.

Its strange to me that UL and manufacturers are going thru all the R&D and meetings and whatever with another new type of gfci that basically is re-inventing the RCD.
If it floats like a duck, quacks like a duck ...
I mean why do we need a 6 letter acronym SPGFCI when its a designed to be the same as a RCD?

And how did they arrive on 20ma?
This SPGFCI that trips 4X higher than a normal gfic at 20ma but 30ma is still unsafe?
The reason I wonder this is all major manufacturers offer 30ma GFCI ( I think they used to be called class B GFCI) breakers that cost about the same as a regular GFCI breaker ( as Garbo used on his hospital fridge ).
Its odd how the UK and EU standards for GFCI allow for 30ma of protection (RCD) instead of 5ma.
It does seem like 5ma is well proven safe level for a GFCI, I wonder if the UK members on here know the rationale behind BS-7671 & 30ma RCD's? I'd love to see studies and data on this.

 

[Eddie702]

Follow the $$$$ that is what it's all about. The MFGs win everyone else loses.

 

[winnie]

If it floats like a duck, quacks like a duck ...
I mean why do we need a 6 letter acronym SPGFCI when its a designed to be the same as a RCD?

And how did they arrive on 20ma?
This SPGFCI that trips 4X higher than a normal gfic at 20ma but 30ma is still unsafe?
The reason I wonder this is all major manufacturers offer 30ma GFCI ( I think they used to be called class B GFCI) breakers that cost about the same as a regular GFCI breaker ( as Garbo used on his hospital fridge ).
Its odd how the UK and EU standards for GFCI allow for 30ma of protection (RCD) instead of 5ma.
It does seem like 5ma is well proven safe level for a GFCI, I wonder if the UK members on here know the rationale behind BS-7671 & 30ma RCD's? I'd love to see studies and data on this.

I don't know the pricess/politics that arrived at the 5mA trip level.

But once they had that 5mA threshold, they had a problem when they wanted to apply similar protection to 480V industrial devices.

My understanding is that the SPGFCI is different from an RCD because it incorporates both the residual current detection and EGC detection. Lose the EGC and the device trips.

Jon

 

[kwired]

Sounds like these little residential VFD's need drive isolation transformers put back in their likely 'cost reduced' product designs.
Why do new fridges trip GFCI's?
Answer: Because fridges are designed for RCD's not GFCI's.

Total hack 'void the warranty' suggestion (dont really do this) but
I bet if you open up the back of the fridge and installed a 1:1 isolation transformer on the VFD input leads the problem would go away. Then you'd still have GFCI protection to the appliance including light and water dispenser and the just compressor would be isolated.



Its strange to me that UL and manufacturers are going thru all the R&D and meetings and whatever with another new type of gfci that basically is re-inventing the RCD.
If it floats like a duck, quacks like a duck ...
I mean why do we need a 6 letter acronym SPGFCI when its a designed to be the same as a RCD?

And how did they arrive on 20ma?
This SPGFCI that trips 4X higher than a normal gfic at 20ma but 30ma is still unsafe?
The reason I wonder this is all major manufacturers offer 30ma GFCI ( I think they used to be called class B GFCI) breakers that cost about the same as a regular GFCI breaker ( as Garbo used on his hospital fridge ).
Its odd how the UK and EU standards for GFCI allow for 30ma of protection (RCD) instead of 5ma.
It does seem like 5ma is well proven safe level for a GFCI, I wonder if the UK members on here know the rationale behind BS-7671 & 30ma RCD's? I'd love to see studies and data on this.

View attachment 2569463

Last time I looked into a QO EPD breaker (30 mA) it was at least double maybe even three times the cost of a GFCI breaker. There really nothing majorly different with construction materials, just that the GFCI are more mass produced is my guess on the price difference.

 

[tortuga]

Last time I looked into a QO EPD breaker (30 mA) it was at least double maybe even three times the cost of a GFCI breaker. There really nothing majorly different with construction materials, just that the GFCI are more mass produced is my guess on the price difference.

Thats a shame I like OQ, we happen to stock Siemens so for us its QE2 in single pole 15 amp and 20A.

 

[kwired]

Thats a shame I like OQ, we happen to stock Siemens so for us its QE2 in single pole 15 amp and 20A.

Well depending on who/what you and your typical annual purchases are it may be something you can get for less.

I can normally get QO loadcenters and majority of common components for less at my supplier than at big box. After I take a markup I end up selling for more than big box but not by any extreme margins.

Apparently the EPD breakers are not one of those common components? I wanted one for de-icing cable one time, ended up using the cable manufacturers ground fault device as it was less expensive (and on hand) than the EPD breaker.

 

[don_resqcapt19]

Sounds like these little residential VFD's need drive isolation transformers put back in their likely 'cost reduced' product designs.
Why do new fridges trip GFCI's?
Answer: Because fridges are designed for RCD's not GFCI's.

Total hack 'void the warranty' suggestion (dont really do this) but
I bet if you open up the back of the fridge and installed a 1:1 isolation transformer on the VFD input leads the problem would go away. Then you'd still have GFCI protection to the appliance including light and water dispenser and the just compressor would be isolated.



Its strange to me that UL and manufacturers are going thru all the R&D and meetings and whatever with another new type of gfci that basically is re-inventing the RCD.
If it floats like a duck, quacks like a duck ...
I mean why do we need a 6 letter acronym SPGFCI when its a designed to be the same as a RCD?

And how did they arrive on 20ma?
This SPGFCI that trips 4X higher than a normal gfic at 20ma but 30ma is still unsafe?
The reason I wonder this is all major manufacturers offer 30ma GFCI ( I think they used to be called class B GFCI) breakers that cost about the same as a regular GFCI breaker ( as Garbo used on his hospital fridge ).
Its odd how the UK and EU standards for GFCI allow for 30ma of protection (RCD) instead of 5ma.
It does seem like 5ma is well proven safe level for a GFCI, I wonder if the UK members on here know the rationale behind BS-7671 & 30ma RCD's? I'd love to see studies and data on this.

View attachment 2569463

The 20 mA trip and the time to trip requirements is about preventing heart fibrillation and the 5 mA is based on the letgo threshold.
The 30 mA equipment protection devices have no time to trip requirements like GFCIs and SPGFCIs.
The RCD standards are also about heart fibrillation, but have an even shorter trip time at 30 mA than a SPGFCI has at 20 mA. One of the reasons that they use the 30 mA is that the 240 volt system will have more normal leakage current than our 120 volt system.

 

[tortuga]

Thanks Don.
It would be interesting to see how these fridge VFD's configured and what type they are.

 

[kwired]

Thanks Don.
It would be interesting to see how these fridge VFD's configured and what type they are.

For household refrigerator they aren't all that big, but these compressors are only like 1/4 or 1/3 HP in many cases They may possibly be located in the air stream for the condenser fan to cool them. They are a definite purpose component unlike the bigger universal drives we are used to talking about here so probably has little that is easily configured in the field or even much if any kind of user interface that is a permanent part of it. Control signals to it all come from the motherboard of the appliance, drive possibly could even be part of the motherboard in some cases.

 

[don_resqcapt19]

Thanks Don.
It would be interesting to see how these fridge VFD's configured and what type they are.

Most of what I have been reading is that there is high frequency leakage current which is causing some GFCIs to trip, even where the high frequency leakage current is less than 6mA. GFCIs are only designed and tested for 60 hertz current.

The UL 943 panel is working on a new standard for a GFCI-HF to address the high frequency current issue, but it will be a few years before there will be any GFCIs built to the new standard.

 

[kwired]

Most of what I have been reading is that there is high frequency leakage current which is causing some GFCIs to trip, even where the high frequency leakage current is less than 6mA. GFCIs are only designed and tested for 60 hertz current.

The UL 943 panel is working on a new standard for a GFCI-HF to address the high frequency current issue, but it will be a few years before there will be any GFCIs built to the new standard.

Maybe they should find ways to eliminate the HF leakage in the appliance?

I've seen this addressed on clothes washers before. Had one that worked fine then suddenly was tripping the GFCI it was plugged into. Appliance repair guy was called first and of course told owner the GFCI was the problem. So now I have to prove something is wrong with the appliance.
This particular one had a "ground switch" in the EGC going to the motor that was otherwise isolated from other conductive components. That switch was NCHO and would be closed anytime the back cover was installed on the appliance, leaving the motor frame isolated from the EGC, yet not subject to anyone contacting it as the door was installed. The plastic mounting means ended up breaking so the switch falls out of place and is allowed to close. I opened that circuit and tried the machine and the GFCI held. Ordered a new switch and has been fine since.

I see no reason they can't do something similar with other appliances to make them compatible with GFCI's.

I also think outside of maybe swimming pool equipment anything not cord and plug connected should not need GFCI protection.