GFCI receptacles protecting branch crts

[Peter Wigle]

Apart from other problems in this picture, I'm curioius about these receptacles being used to protect downstream outlets. The downstream outlets are not grounded and, as they are 3-prong receptacles should be labelled as ungrounded, but my main concern is this: When I tested the downstream receptacles with a very basic circuit tester, I was not able to trip the GFCIs. I understand that the test button on the tester in not the same as a true ground-fault (?) but should it not still have tripped the GFCIs? Or is there something special about this arrangement.

I've already recommended an inspection and probable upgrade by an electrician for other reasons, but still curious about this arrangement. I expect there is a limit to the number of downstream outlets, but don't know.

[Rick Cantrell]

If the downstream outlets do not have a ground then you will be unable to trip the GFCI at that outlet with a tester. Press the test button on the GFCI then check the downstream outlets.
I don't know of a limit on the number of outlets on a circuit or connected to a GFCI.

[Jim Luttrall]

Ditto, what Rick said.

[Jerry Peck]

The reason Rick and Jim are correct is that a tester requires a ground to work, the GFCI device itself does not require a ground.

There is no code limit on the number of downstream receptacle outlets off a GFCI device, however, I believe someone posted (not too long ago) that they read an installation instruction which had a limitation to the number of downstream receptacles of that particular GFCI device.

[Michael Thomas]

Also (just to cover all the bases) according to Square D to limit nuisance trips you should observe at 250ft (one way) limit on the distance to the load from a GFCI breaker, which I would assume might also apply it loads connected to a GFCI outlet.

http://ecatalog.squared.com/pubs/Circuit%20Protection/Miniature%20Circuit%20Breakers/QO-QOB%20Circuit%20Breakers/QO-GFI/48840-079-01.pdf

[Markus Keller]

- Old cloth wiring has reached the end of it's safe and useful life cycle. Recommend replacement by licensed electrician
- Improper electrical box usage for GFCI installation
- No covers on GFCI receptacles
- Is that romex or lamp cord?
- No ground or downstream receps? Call licensed sparky
- What's that black wire? romex, lamp cord? I've never seen black romex. Is that something common up there?
- So somehow Joe DIY figured a bunch of GFCI's was better than proper wiring to breakers? Damn those Time Life books, where is mine anyway.

[Jerry Peck]

Michael,

Imagine the voltage drop one would get with 500 feet (out and back) of #14 or #12 AWG connected to a breaker.

Using #12 AWG, that would have 0.965 ohms resistance (DC resistance) for that circuit run of 250 feet (one way), for an approximate voltage drop of 15 voltage across the circuit conductors themselves at the maximum circuit rating of 16 amps on a 20 amp breaker.

120 volts - 15 volts = 105 volts at the load end of the circuit

With #14 AWG, that would have 1.535 ohms resistance (DC resistance) for that circuit run of 250 feet (one way), for an approximate voltage drop of 18 voltage across the circuit conductors themselves at the maximum circuit rating of 12 amps on a 15 amp breaker.

120 volts - 18 volts = 103 volts at the load end of the circuit

I wonder if the "nuisance tripping" is from leakage across the conductors through the insulation (should not be any) or from the voltage drop creating higher current draw at a motor at the far end, which would increase the voltage drop even more, increasing the current ... would the nuisance tripping be GFCI tripping or overcurrent tripping?

[Markus Keller]

250'? I find it hard to imagine a safe or reasonable scenario where having the load from a GFI anywhere near that distance away. If the load side tripped how would anyone but the installer even know what to reset? The distance doesn't seem reasonable for a home (possible in a McMansion of course). I doubt that anyone (legit) would install such a set-up, it's just dumb and open to problems.

[Corn Walker]

- What's that black wire? romex, lamp cord? I've never seen black romex. Is that something common up there?


I think black jacketed ROMEX 12/2 and 14/2 was used primarily in the late 60's and was rated at 60C. I don't believe it was required to be labeled along its length at the time.

Currently black jacket is mainly used for cables of 8ga and larger.

Corn Walker
Western Massachusetts

[Michael Thomas]

250'? I find it hard to imagine a safe or reasonable scenario where having the load from a GFI anywhere near that distance away. If the load side tripped how would anyone but the installer even know what to reset? The distance doesn't seem reasonable for a home (possible in a McMansion of course). I doubt that anyone (legit) would install such a set-up, it's just dumb and open to problems.


Legit?

About the only way I could see it happening is to a garage, I was just sorta' making the point that when I saw that one take was: "Essentially,that's functioning as a GFCI breaker", and my mind started to wander down my mental GFCI breaker check list.

----------

OTOH, not so legit?

Well... idiots are endlessly creative... just recently saw a picture on another site of a crawl space where someone had inserted what appeared to be an entire 100' coil of NM into the middle of a circuit,


True - that's still not 250' even if you include the rest of the circuit - but it is creating a substantial inductive effect.

And as it happened I found myself wondering "How would that affect an outlet at the far on that circuit and protected by a GFCI installed at the near end..?

[Corn Walker]

Well... idiots are endlessly creative... just recently saw a picture on another site of a crawl space where someone had inserted what appeared to be an entire 100' coil of NM into the middle of a circuit,

Wait, did they leave it coiled up? You could do some pretty cool magnetic tricks with something like that.


Corn Walker
Western Massachusetts

[Michael Thomas]

Yup.... nice tight coil, just the way they found it at HD.

[Michael Thomas]

Here ya' go... found it:

http://img2.timeinc.net/toh/i/g/0608-home-inspection-12/10-home-inspection-12.jpg


It was credited to Corey Folsom, Ben Lomond, California

[Jerry Peck]

Michael,

The other end is just out of the photo, was it connected? It looks to 'loose and free' to be connected, but maybe not.

Either way, not only is that a waste of NM cable, but it adds a tremendous voltage drop to the circuit, not only in resistance but in inductive capacitance - that acts like a choke coil with enough current through it. I'm not trained nor smart enough to calculate out the gauss created (depends on the number of wraps, size of the coil, voltage across and current through, among other things), but that's just plain old *not good*.

[Jim Robinson]

250'? I find it hard to imagine a safe or reasonable scenario where having the load from a GFI anywhere near that distance away. If the load side tripped how would anyone but the installer even know what to reset? The distance doesn't seem reasonable for a home (possible in a McMansion of course). I doubt that anyone (legit) would install such a set-up, it's just dumb and open to problems.

I could see it happening outside, but you could just as easily put the GFCI at the far end instead of at the beginning, or have two of them.

[Corn Walker]

250'? I find it hard to imagine a safe or reasonable scenario where having the load from a GFI anywhere near that distance away. If the load side tripped how would anyone but the installer even know what to reset? The distance doesn't seem reasonable for a home (possible in a McMansion of course). I doubt that anyone (legit) would install such a set-up, it's just dumb and open to problems.

I was consulting to a school in SF a couple of years back (about software, not construction) and when I checked in at the desk the receptionist was having difficulty finding an electrician to fix a problem. They had lost power to a receptacle feeding a refrigerator but none of the breakers were tripped at the panel. I told him I wasn't a licensed electrician but I could take a look and see if there was anything obvious while they tried to find someone. While I was looking around a teacher down the hall asked, "do you think you could check out why the outlet in my classroom isn't working too?" Sure enough, the GFI in her classroom had tripped and it was supplying the refrigerator some 60 ft down the hall. I doubt the installer would have figured that one out.

[Jerry Peck]

Sure enough, the GFI in her classroom had tripped and it was supplying the refrigerator some 60 ft down the hall. I doubt the installer would have figured that one out.


Which is why I always told my clients: "GFCI devices are not that expensive, maybe $15 each, don't 'feed through' wire a GFCI, just install a GFCI at each location for that location, not 'fed through'."

When we bought our house up here, it was a 1978 house, while going back and forth in preparation for moving, I hired a local electrical contractor to come by and install a GFCI device at each receptacle location which required GFCI protection.

It took them 4 trips to understand what I was stating, and *they still missed* two receptacle outlets. Then they wanted to charge me for the additional three trips. I finally convinced them that if they had done what *I REQUESTED* the first trip, there would have only been one trip, and that as it stands there *are still* two they missed. They removed the additional charges for the additional trips, then added in the cost of installing the extra GFCIs. I said 'Okay, now come out and install the other two you still missed, if *I* need to do it myself, I will, but you owe me a credit for having to do it myself because your electricians did not do it.' - I got a credit off for the cost of adding two GFCIs because it was cheaper for them to do that than to come out and install the remaining two GFCIs.

Incompetence knows no bounds.

Then, when I went to remodel our kitchen, I found out that they had reverse wired (line-load) the kitchen GFCIs.