Can a two-pole 240 V GFCI detect a ground fault that draws equal amounts of leakage current (e.g., 1.0000 A) from both poles?



[This message has been edited by Dspark (edited 05-08-2001).]

If you wired it as per the diagram, you'd probably kill someone.

But this was mentioned in the other thread...

I'm hoping that Bill fixes the image.
The diagram, I think, will show how the manufacturer wires it.

Wiring Diagram Fixed!

ok, thanks Bill...

now can someone explain this??

The way I understand it is that all three conductors pass through the coil, and the current should balance to zero under normal operation. When a fault occurs, there is an imbalance that is detected by the coil and opens the line.

I was thinking before that it would take one coil for each pair (Line "A" to Ground, Line "B" to Ground, Line "A" to Line "B") but I'm mistaken.

As far as getting deeper into how they work... Better let Scott handle it!

That's right about how it works in a nutshell.

But I am not clear on whether a symmetrical fault to earth would be detected.

Such a fault is improbable in a hot tub. But I want to know whether it would be detected, if, for instance, a 240 V floor outlet were filled with Pepsi causing a symmetrical fault to ground on both poles.

Dspark;
i never heard of a "symetrical fault" , but theoretically, if both conductors climbed the trip curve equally..???>>>

that would be the pepsi challenge

Well everyone, if the leakage current to ground was exactly equal from both Ungrounded conductors and at the same time, that would not trip the device!!

It would be similar to someone grabbing both lines [L1 and L2], then getting the ever loved 240 VAC barbeque. The device would not see that as an inbalanced current, so no trip.

Kind of sucks, doesn't it?

Scott SET

yeah, ouch....
but in the practical world, the probability is slim.
So a good point of discussion is the noodle, gotta have it , straight 240V or not right?

I'm thinking it operates along the lines of a current transformer. There is no neutral load other than the one at the sensing device. It doesn't appear to be manually adjustable to set the normal load current setting.

A neutral makes no difference.
I can't think of a way to detect a symmetrical (balanced) fault.

There is no neutral load at all in the hot tub scenario that spawned this thread.

GFCI is not an OCPD. Even if it were, I can't think of how a GFCI could tell whether the motor is starting or a human grabbed T1 and T2.


[This message has been edited by Dspark (edited 05-09-2001).]

So back to the hot tub installation...

It is imperative to ensure that it is impossible to touch both ungrounded poles in any 240 V installation. GFCI cannot protect against this.

I think this is a bit different from a 120 V installation with GFCI.
In 120 V, I can get a nasty shock between ungrounded and grounded conductors. However, enough current might leak through my feet to the earth and trip the GFCI.

With 240 V, if the amounts leaking to earth from each ungrounded conductor are equal, they cancel each other locally and don't trip the GFCI.

Envision the cable being cut and submerged into a mud puddle.

The 120 V GFCI will probably trip as current follows earth or ground rather than neutral.
The 240 V GFCI possibly won't trip.


[This message has been edited by Dspark (edited 05-09-2001).]

Dspark;
ok, this is assuming a perfect short circuit, short being different than ground fault.

I am not assuming a perfect short circuit at all.

I am assuming a symmetrical ground fault.

Take one ungrounded conductor to a mud puddle on the north side of the building and the other ungrounded conductor to a mud puddle on the south side of the building.
Say that both have a resistance from the main panel to their respective points of 25 ohms.

Energize the circuit.
Current flows indefinitely.
Please don't tell me that this is not a ground fault condition.

(It is a contrived scenario. But clearly a ground fault.)

To put a fine point on things, remember, this is not really a 240v GFCI Circuit Breaker. We should try to find a digram for one of those.

Bill

I think you are the first person in this thread to say "Circuit Breaker".

I was just calling it a (non-specific) 240 V (two-pole) GFCI (variations mentioned in the Hot Tub thread).

The CB wiring should be the same albeit with some components integrated within the CB.

I was looking at the schematic [more like a pictorial, but so what!] with a bit more detail today, when this came to mind:

I kind of thought that the GFCI elements in this type of circuitry would be connected L-L, not L-N!!
I figured this since they are mass produced items for either 240 VAC or 120/240 VAC use, plus some crackpot ideas of mine [loss of one Line or open neutral cannot trip device - told you it's crackpot!! ].

So much for me thinking, huh??

Now, all this talk about being fried from both lines to ground [leakage ground faults] at the same time and of equal intensity has me wondering if this has happened before, and if so, how many times??
It could happen!!
Wouldn't want to be that human conductor [not the ones that ride in cabooses] caught not only between both lines, but also to ground!!
OWWWWW---CCCCHHHH!!!
Meanwhile, the GFCI sensing items are confident that all is OK with the flowing currents, so why trip?!
Again - OWWWWW---CCCCHHH!!!!

On a non-related note , the "Conductor" and "Caboose" thing reminded me of something I say often in the field:

When someone asks if I am an Engineer, I'll say - Yes! I drive the Cannonball from Hooterville to Pettycoat Junction!!
[then after they are completely convinced that I am looney, I'll mention something towards EE].

This is only for fun, not thrown at any clients [unless they have sense of humor, then there's no holding back!!].

Got a lot of strange reactions from that one!!

Time for some meds [joke]

P.S. - what was the original pictorial setup like before Bill edited it?? was it a connection mistake or just difficult to read??

Scott SET - local Cannonball steam locomotive operator on silly 60's sitcoms.

In the upper left the three wires changed from horizontal to vertical at the wrong places, like LN to T1, L1 to T2, L2 to TN, or something like that.

Dspark;
L1 & L2 both to ground would be a 'ground fault'
a human holding both L1 & L2 ( assuming well insulated sandals) would be a 'short'

either situation is suspect, maybe the listing would serve as clarity,


from siemens GFI instructions;

"EL GFCI no protege a las personas contra choque de tension entre fases."

[oops.... wrong side....... ]

"The GFCI will not protect a person against a line to line shock"

makes the noodle look important !





[This message has been edited by sparky (edited 05-10-2001).]

Still might not be detected...

If my twin brother who has the same impedence as I were holding L1 while I held L2 standing on the ground in bare feet...

A GFCI won't protect against line to neutral either (120 or 240 V). What the listing doesn't mention is that it may also fail to detect symmetrical (240 V) line to ground faults.

A 120 V GFCI still may trip when a human hand shorts line to neutral while standing on the basement floor, touching the sink, etc.

those instructions are simply stating that the device lives only up to it's name.

i would wager that a perfect short across L1 & L2 , or perfect 'symetrical ground fault' from both to Grn that would leave a device set to trip @ 50ma untripped , while climbing into many 00000's of Amps would be hard even in a laboratory setting to create.

shall we continue into 3ph GFI's ????

Whoa!
An L1 to L2 short no matter how perfect or imperfect will not (almost never) trigger the GFCI, low impedance, high impedance, arcing, or whatever.

I could recreate this is my backyard (but preferably yours) any dry day.


As for the symmetrical ground fault, I could do it in a lab with two closely matched power resistors. I would guess that I could do it with two christmas light bulbs and two three-way switches with their handles tied.

But it is a pointless experiment to me since Scott can do it for me in his head.


[This message has been edited by Dspark (edited 05-10-2001).]

On the matter of three equal phases, that case has just a few combinations more than two-pole, but still the same issues.

Symmetrical ground faults and line-to-line and line-to-neutral faults won't be detected by the GFCI.

Doesn't that look a whole lot scarrier in Spanish??

Apparently, the GFCI is not going to protect anyone with tense faces [joke ].

I've gotta post my "Shock Hazard Sign" in this forum for review. It features a clip art image [cartoon] of a guy getting the Be-Geezes shocked out of him because he placed a finger into a receptacle [on the neutral side too!! so much for accuracy]. It looks so funny, but passes the point well.

On top it has a clipart sign "High Voltage", then some warning messages I made up [this WILL kill you, burn you, destroy you, etc.].

To be funny, I wrote some stuff, similar to the way Warner Brothers' cartoons goof with French words and phrases.

Mine are in Spanish, French and German and not one bit of the text means anything in those languages!!
Many laughs from the signs.

Scott SET

well boys;

were back to wondering how to protect a straight 240 piece of equipment here !


This thread does leave me wondering about my next installation, as well as inclined to do some experimenting.

I do have a few ornery chickens that just won't stop cra**ing on the porch, maybe they would volunteer ?

Mmmmm... I smell fried chicken!
Extra Crispy Too!

Mr. Webmaster;
I have invited Breaker manufacturers via e-mail to participate here, maybe they'll straighten us out?

better fried chix than fried customers

(they tend to not pay up!)

You protect a straight 240 V piece of equipment by making sure no one can contact the ungrounded conductors. You make sure that everything is bonded so that one conductor can't unintentionally energize one thing while another energizes something else.

With these as givens, the OCPD and GFCI should do the rest.

Just a little trivia here. We had a rep from UL at a meeting and GFCI's was the main topic.

GFCI protection for personnel is set to trip between 4ma and 6ma. An interesting thing mentioned was that the speed with which they operate depends upon the level of the fault. A low threshold 4ma fault could take up to something like 4.8 seconds to trip depending upon the sensitivity of the particular unit.

Bill

we probably induce more than that doing panel work....

Another piece of trivia. An Electrical product is permitted to have a loss up to .5ma and still be listed. The importance of this tidbit comes into play where there is a 'nusance' tripping going on. As the GFCI operates on a cumulative value of stray current it could seem to be tripping for no reason but it may be operating perfectly. There could be 8 items on the circuit with a .5 'loss' each (8 x .5 = 4ma) and tripping could occur and there is nothing wrong. It's something to think about when deciding on CB vs Receptacle type GFCIs' and the number of downstream devices and length of a circuit protected by a single GFCI.


Bill

If the ground fault was symetrical, it would mean that the human creating the short had three arms of identical impedance. With the current summing at zero in his chest, he wouldn't have cardiac stress, just a real bad feeling in the arms. Quite a party trick.

Bill,

Great point on the line charge Capacitance!!
A great example of when circuit charging becomes a nuisance would be a few extension cords connected together with normal NEMA 5-15 R/P connections, several power tools which use series wound commutator motors with open brushes, and a little bit of rain. This is normally followed by the quick removal of the GFCI breaker, swapped with a standard type frame

Another one would be with a whole gang load of cords running everywhere and each being a very long distance [upto 50 feet each], from there many power tools are connected to each cord. Some start and run together, some don't - most have dirty commutators, so the brushes spark like fireworks.
This stuff will trip even the most highest rated GFCI device!!
Long circuits, spark shows on brushes, cords running long distances on concrete/against bonded steel/etc.. many non-coincidental starts/runs on many circuits causes the circuit's charging Capacitance to be really high!

Just FYI. I'll return to my room now

Scott SET