GFCI 2-wire Experiment

[K8MHZ]

And I can't duplicate your experiment with my water, so asking me to show you by demonstration is really of no value.

 

[FionaZuppa]

Is the most important piece of information, that you believe you understand electricity, and GFCI's, better than most of the forum participants and moderators?

nope. but a hint, pick something else from the text that is in post #1 before pic #1.

And I can't duplicate your experiment with my water, so asking me to show you by demonstration is really of no value.

i never asked you to dupe what i did. i asked you to run your own experiment, even jump in if you like, do whatever experiment you see fit to prove a point, or disprove something i have done, etc etc.

 

[mbrooke]

i dont say this too often, its just one of those things i keep to myself, very obvious that many on this forum, including some mods, dont really understand electricity all that well.

here's an experiment showing you a possible scenario that a 2-wire GFCI will not protect you, and you are possibly dead.

the cord is a non-polarized nema 1- (i polarized it with silver marker). the hot has an inline ~18k resistor. the end of the cord has eyelets soldered on to attempt to mimic real terminations of an appliance. the sink is all plastic so no amps can flow out to a EGC, the water is warm tap water (my water is heavy in minerals and filtered via a water softener, more on this at the end). from the data, the "short" in the water (w/o the inline ohms) would not trip a 15 or 20A OCPD and the GFCI is as happy as a virgin for the 1st time. the stray amps across the unattached amp probes in random location in the water indicates that amps will still flow from wire --> across water --> into the body --> through body --> back into water, and return back to wire !! so as much as some of you think amps only take path of least resistance (WRONG), it takes all paths.

i dont care to explain this elementary experiment, so you'll need to understand it for yourself. in this experiment the amps through the unattached probes would not have been enough to kill, but the scenario shows the possibility exists.


voltage across end of wire no-load

short circuit amps

wires ends in water

amps of wire in water

amps probes just in water, not attached to anything

amps flowing through the unattached amp probes

shortest distance between the ends of the wire

as for water chemistry, the cleaner the water the more hazard there is. and depending on how the end of the wire terminates into the item, stray amps will vary (the closer the connections are to each other the less stray amps, the farther apart and bigger the connection then more stray amps).

hence, the reason why 2-wire GFCI does not cover all hazards, some potentially deadly. hence, the reason why all GFCI should be 3-wire. hence, the reason why all "bathroom" appliance should be NEMA 5- !!!

And as I said, GFCIs were mandated at first to take care of open EGCs, not dropping a hair dryer into an insulated sink with both hands in it.

If immersion is that much of an issue there is an LCDI, and in the case of pools code requires you to "bond" the water.

 

[K8MHZ]

If you do get a resistor for testing, please, show both directly across the leads (putting the water in parallel, like when a finger single finger is stuck in) and in series ( like when the current path is direct like it would be grabbing a conductor while in the water).

None of the above will show anything about the 2 vs. 3 wire GFCI, but may give an idea of the current range involved, say if the cat fell in the sink while the experiment was being performed.

 

[FionaZuppa]

a schematic of the experiment showing all possible current paths

i leave the challenge of illustrating all possible currents paths in a liquid conductor to you, but let us know when you have it completed :thumbsup:

And as I said, GFCIs were mandated at first to take care of open EGCs, not dropping a hair dryer into an insulated sink with both hands in it.

If immersion is that much of an issue there is an LCDI, and in the case of pools code requires you to "bond" the water.

are you saying that immersion is not one if the hazards NEC intends to try and protect against, especially in the bathroom area of resi ???
i didnt speak about pools.

 

[JFletcher]

Your own measurements, even if perfectly accurate, show one obvious reason the GFCI may not have tripped: they are at the cusp of the GFCI trip threshhold: 6ma. (6.79, 6.60)The last reading of 0.15ma is far below 6. Ofc its not going to trip then.

Not every GFCI that rolls off the assembly line is going to trip instantly at exactly 6.00000000000000000ma.

Aside from what others have mentioned re: the meter, you have also failed to take into account sampling error, and assumed your GFCI will trip at exactly 6ma. You are not allowed to make that assumption.

It is akin to me deigning a new drug, giving it to one person, who then gets the flu. Thus, 100% of my subjects had the side effect of getting the flu. If my drug also cures that person, I can tout a 100% cure rate. Which would be awesome until it is independently and/or double blind tested on a large sample of people (say, 2000), they get very different (and more accurate) results, and my drug company is sued into oblivion.

All that GFCIs need to see is a 6ma differential between ungrounded and grounded and they *should* trip. Thus, if I stick my hands in the sink with your leads (or better yet, you do it ), and say 9ma finds their way into me and out to ground (or even a surface with a lower potential), the GFCI will trip.

 

[mbrooke]

What I am thinking is that we remove the resistor from the cord and place one in series with the leads. This way we have a scenario which IMO makes more reflection of the real world. Appliances submerged are likely to have the terminals with no impedance between them and the plug, and the resistor will mimic typical body impedance.

 

[FionaZuppa]

Your own measurements, even if perfectly accurate, show one obvious reason the GFCI may not have tripped: they are at the cusp of the GFCI trip threshhold: 6ma. (6.79, 6.60)The last reading of 0.15ma is far below 6. Ofc its not going to trip then.

well, i am sorry you had to write out all you did. so do this, rip out your wizzy-wig do-dad calculator, whatever that is, use the #'s i have provided, remove the inline resistor, run your math.

geez, i didnt think it would take this long for anyone here to figure out that the inline R is safety item !!! i perhaps didnt want to get bitten by 15A worth of juice with my hands in the water that has a 1- cap cord jammed into a useless GFCI !!!!! you get it now ???? the inline R is meaningless in the context of the experiment !!!!!

What I am thinking is that we remove the resistor from the cord and place one in series with the leads. This way we have a scenario which IMO makes more reflection of the real world. Appliances submerged are likely to have the terminals with no impedance between them and the plug, and the resistor will mimic typical body impedance.

you really dont need to. use math from here.

 

[mbrooke]

well, i am sorry you had to write out all you did. so do this, rip out your wizzy-wig do-dad calculator, whatever that is, use the #'s i have provided, remove the inline resistor, run your math.

geez, i didnt think it would take this long for anyone here to figure out that the inline R is safety item !!! i perhaps didnt want to get bitten by 15A worth of juice with my hands in the water that has a cap cord jammed into a useless GFCI !!!!! you get it now ???? the inline R is meaningless in the context of the experiment !!!!!

Safety? For what? Aren't we trying to replicate a real world danger where we will not have have an inline R?

 

[FionaZuppa]

Safety? For what? Aren't we trying to replicate a real world danger where we will not have have an inline R?

the inline R doesnt matter !!!

and yes, safety, the GFCI is useless in this experiment, with 18k inline i cannot get bitten by 15A. but again, the R is meaningless, you get same end results w/ or w/o the inline R there !!

 

[mbrooke]

the R doesnt matter !!!

and yes, safety, the GFCI is useless in this experiment, with 18k inline i cannot get bitten by 15A. but again, the R is meaningless, you get same end results w/ or w/o the inline R there !!

It does. It pulls the voltage down across those leads when immersed. Skewing your result in the end.


Also, if you get bitten its unlikely to be 15amps. That would require a body resistance of 8 ohms, not likely under normal conditions.

 

[FionaZuppa]

for some odd reason i knew this post would go south fast with lots of folks not understanding it.

anyways, feel free to experiment, jump in the non-GFCI'd water if you like to obtain some real human data. death is still real data, as long as someone is around to document it :lol::thumbsup:

 

[mbrooke]

for some odd reason i knew this post would go south fast with lots of folks not understanding it.

anyways, feel free to experiment, jump in the non-GFCI'd water if you like to obtain some real human data. death is still real data, as long as someone is around to document it :lol::thumbsup:

Explain in math why the inline Z will not matter.

 

[JFletcher]

well, i am sorry you had to write out all you did. so do this, rip out your wizzy-wig do-dad calculator, whatever that is, use the #'s i have provided, remove the inline resistor, run your math.

geez, i didnt think it would take this long for anyone here to figure out that the inline R is safety item !!! i perhaps didnt want to get bitten by 15A worth of juice with my hands in the water that has a 1- cap cord jammed into a useless GFCI !!!!! you get it now ???? the inline R is meaningless in the context of the experiment !!!!!

"Useless GFCI" says it all. Glad you've uncovered the roots of a 55 year long conspiracy (check the 1962 NEC for first GFCI uses). Man, think of all the money you're going to make suing Leviton, Hubbell, P&S, etc for making unnecessary GFCI devices.

You have no idea the resistance value of your water. Also, 15A is a useless, arbitrary number. 15A will not trip any 15A breaker.

Here's a test idea for you: take a GFCI receptacle (connected to EGC) put a jumper between neutral and ground on the load side. Tell me if it resets or doesnt trip the second you put a load on it. Now, try the same jumper line side. Nothing happens.


Quite frankly, your one-off sink test is not going to convince me that 50+ years of testing by mfg with far more precise instruments and controls are wrong. Good day.

 

[Ingenieur]

let me get this straight:
you start off with mods/members don't understand electricity
and follow up with you won't explain it to us
ok

123 V / 18000 ohm ~ 6.8 mA
in line with your readings (6.79, 6.6 mA)
no EGC
so the current thru the gfci ct is balanced
did you expect it to trip?

why do you think if someone stuck their hand in the water (or same conditions in a bathtub) that the current would flow thru the water from the H, then into the body, back thru the water, to the N?

 

[growler]

geez, i didnt think it would take this long for anyone here to figure out that the inline R is safety item !!! i perhaps didnt want to get bitten by 15A worth of juice with my hands in the water that has a 1- cap cord jammed into a useless GFCI !!!!! you get it now ???? the inline R is meaningless in the context of the experiment !!!!!

Do you really have your hands in the water with power on? That's a good way to win a Darwin award.

 

[JFletcher]

let me get this straight:
you start off with mods/members don't understand electricity
and follow up with you won't explain it to us
ok

123 V / 18000 ohm ~ 6.8 mA
in line with your readings (6.79, 6.6 mA)
no EGC
so the current thru the gfci ct is balanced
did you expect it to trip?

why do you think if someone stuck their hand in the water (or same conditions in a bathtub) that the current would flow thru the water from the H, then into the body, back thru the water, to the N?

Because he has isolated his experiment from all grounds, thus N is the only return path. He has also ignored the extreme resistance of the water further from the probes.

 

[FionaZuppa]

Do you really have your hands in the water with power on? That's a good way to win a Darwin award.

is that not what i wrote ?? and even if i did not the experiment would still have a current limiting device inline.

do you grab cap cords with your bare hands ?? wow, why would you do that ???

Because he has isolated his experiment from all grounds, thus N is the only return path. He has also ignored the extreme resistance of the water further from the probes.

WRONG, those scenarios are not ignored, they can be categorized as being scenarios that may fall under the mA levels as obtained from the experiment, as i have already mentioned, please go back and re-read the 50 posts.

 

[Ingenieur]

the inline R doesnt matter !!!

and yes, safety, the GFCI is useless in this experiment, with 18k inline i cannot get bitten by 15A. but again, the R is meaningless, you get same end results w/ or w/o the inline R there !!

yes it does
with 6 mA flowing the Vdrop across it is ~ 0.0067 A x 18000 Ohm or 120 V
there is really no drop across the water

 

[mbrooke]

Do you really have your hands in the water with power on? That's a good way to win a Darwin award.

I actually should have addressed this a few posts back. Fiona, when doing this do not use your hands or any other body parts. If you have to, use at minimum a long plastic mixing spoon to manipulate anything in the water.