Non-Earth Grounded, Oscilloscope Observation [video]

[IanB]

RCD and GFCI are not the same thing.

RCD and GFCI are just different acronyms used in different parts of the world. If you took an RCD to North America it would be called a GFCI, and if you brought a GFCI to Europe it would be called an RCD.

A "residual current" -- a current remaining after you balance the current in all the conductors -- is the same thing as a "ground fault" -- a current escaping to ground and not remaining within the power conductors.

[Spunky]

In europe we mostly use voltage independent residual current circuit protective devices for domestic use (they have to be I believe to meet EN61008), whatever your country calls it this is what I call an RCD. They are only required to detect current imbalance, whereas according to NEC regs residual current circuit protective devices must detect downstream neutral-earth faults without current flow and live neutral reversal, which as far as my knowedge goes traditionally uses an energised circuit, which leaks current from live to neutral, and a 2nd coil. This is what I (and the country that invented it) call a GFCI. Maybe some newer models can do this passively I don't know, it would only be small number of the GFCIs in use.

Some places in the world may refer to either technology as GFCI or RCD and have different requirements for their devices. So, I regard EN61008 voltage independent residual current circuit protective devices as RCD, and US type devices as GFCI, maybe this causes some confusion, especially in european countries that use the term GFCI. They may or may not be voltage dependent, depending on how their required operation is defined in local regs.

And no I don't have all these regs with me in writing. Download them if you really want to, I don't have the spare time.

[Monkeh]

In europe we mostly use voltage independent residual current circuit protective devices for domestic use (they have to be I believe to meet EN61008), whatever your country calls it this is what I call an RCD. They are only required to detect current imbalance, whereas according to NEC regs residual current circuit protective devices must detect downstream neutral-earth faults without current flow and live neutral reversal, which as far as my knowedge goes traditionally uses an energised circuit, which leaks current from live to neutral, and a 2nd coil. This is what I (and the country that invented it) call a GFCI. Maybe some newer models can do this passively I don't know, it would only be small number of the GFCIs in use.

None of which is in any way relevant to the trip current or time of the device. The fundamental operation is identical.

[Spunky]

You debated if they were the same device, not what their trip characteristics were. They are not the same device.

[Monkeh]

You debated if they were the same device, not what their trip characteristics were. They are not the same device.

And you separated them based on trip characteristics, your point being, Mr. Pot?

They are fundamentally the same device with the same purpose, and I would quite happily argue that zero-current neutral-earth fault detection is nothing more than a trip characteristic.

I'll admit to being unaware of the added function. I can't really decide if it's useful or a waste of parts.

[Spunky]

RCD and GFCI are not the same thing.

RCD and GFCI are just different acronyms used in different parts of the world. If you took an RCD to North America it would be called a GFCI, and if you brought a GFCI to Europe it would be called an RCD.

A "residual current" -- a current remaining after you balance the current in all the conductors -- is the same thing as a "ground fault" -- a current escaping to ground and not remaining within the power conductors.

Actually if you took an RCD to the US the building inspector would (or should) fail it. There are additional requirements on a GFCI in the US that an RCD does not meet. They are indeed very close, but different. Although some RCDs do incorporate more features, they are not a requirement here.

[IanB]

according to NEC regs residual current circuit protective devices must detect downstream neutral-earth faults without current flow

Which is fascinating, and Today I Learned Something New.

So when I get home I am going to try connecting neutral to ground on one of my GFCI outlets and see what happens (presumably it will trip).

[Spunky]

You debated if they were the same device, not what their trip characteristics were. They are not the same device.

And you separated them based on trip characteristics, your point being, Mr. Pot?

They are fundamentally the same device with the same purpose, and I would quite happily argue that zero-current neutral-earth fault detection is nothing more than a trip characteristic.

I'll admit to being unaware of the added function. I can't really decide if it's useful or a waste of parts.

Yes I did seperate them based on trip characteristics, GFCIs are generally 5mA, RCDs are generally 30 to 100mA. Sure you could get a 5mA RCD or a 1000mA GFCI. Both would be unlikely in an apartment don't you think?

As for the other features, I never said it wasn't a waste of parts. Personally I think it is, as an RCD would trip when loaded under the same fault anyway.

And my point? The original topic has already been dealt with. I think we're losing the point now.

[Spunky]

according to NEC regs residual current circuit protective devices must detect downstream neutral-earth faults without current flow

Which is fascinating, and Today I Learned Something New.

So when I get home I am going to try connecting neutral to ground on one of my GFCI outlets and see what happens (presumably it will trip).

Test first there's no potential between them. It should trip all on it's own once connected. That's the theory.

[Monkeh]

As for the other features, I never said it wasn't a waste of parts. Personally I think it is, as an RCD would trip when loaded under the same fault anyway.

I know for a fact they will. I can't think of a situation in which a combined N-E and L-E fault would desensitize even a 30mA device sufficiently, let alone a 5mA hair trigger.

And my point? The original topic has already been dealt with. I think we're losing the point now.

But we're learning!

[IanB]

according to NEC regs residual current circuit protective devices must detect downstream neutral-earth faults without current flow

Which is fascinating, and Today I Learned Something New.

So when I get home I am going to try connecting neutral to ground on one of my GFCI outlets and see what happens (presumably it will trip).

Test first there's no potential between them. It should trip all on it's own once connected. That's the theory.

My meter says there is potential between them--about 10 mV. That's how it can work, by trying to drive a current through any such ground-neutral connection. By comparison I measure 0 mV between ground and neutral on a regular socket.

And it appears it does trip, but only with a dead short. I first tried connecting a 15 W night light bulb, but apparently not enough current flowed to be detected.

[Tepe]

Some places in the world may refer to either technology as GFCI or RCD and have different requirements for their devices.

In this country we talk about HFI and in newer installations about HPFI. The latter can also deal with pulsating DC.

[Spunky]

HPI would be the same as our Type A, and HPFI would be our TYPE AC, probably all derived from the same European approval.

When I said it detects N-E faults without current flow that was wrong, what I should have said was that it does it without an external load by raising the neutral potential slightly so it will drive a tiny fault current N-E. This is presumably something to do with why in the US they don't stack them in series (1 on consumer unit, 1 on the socket), whereas in the UK it's fairly common.

[Tepe]

HPI would be the same as our Type A, and HPFI would be our TYPE AC, probably all derived from the same European approval.

I think it is the other way around: HFI is type AC and HPFI can be either A or B.

[Spunky]

HPI would be the same as our Type A, and HPFI would be our TYPE AC, probably all derived from the same European approval.

I think it is the other way around: HFI is type AC and HPFI can be either A or B.

Yep, quite likely. I only heard about those terms yesterday 

[madires]

As for the other features, I never said it wasn't a waste of parts. Personally I think it is, as an RCD would trip when loaded under the same fault anyway.

I know for a fact they will. I can't think of a situation in which a combined N-E and L-E fault would desensitize even a 30mA device sufficiently, let alone a 5mA hair trigger.

A too low mA value would cause problems if you got several SMPSUs or other devices with Y caps for filtering (between N and PE, and L and PE). The usual RCD for homes/apartments over here is triggered by 30mA and there's a current limit for the Y caps of 3.5mA per device (for 230V 50Hz it's max. 22nF per Y cap).

[Spunky]

A too low mA value would cause problems if you got several SMPSUs or other devices with Y caps for filtering (between N and PE, and L and PE). The usual RCD for homes/apartments over here is triggered by 30mA and there's a current limit for the Y caps of 3.5mA per device (for 230V 50Hz it's max. 22nF per Y cap).

Yes that and old motors in vacuum cleaners, washing machines, etc. Lots of things leak to ground, which is presumably why it's 30mA to protect a circuit and only 5 or 10mA on individual outlets.