Daisy-chained, parallel or series? (failed outlets in home)

[berkeman]

That's always the hardest part of these jobs. You know and expect the "normal configuration", but that's not always what the guy before you thought "normal" was.

Definitely, which is why the inexpensive outlet tester that I mentioned is important to get and use on all of our outlets...

Do you have an outlet tester with GFCI test capability? They are inexpensive, and very handy for debugging problems like this.
https://www.electricalbasics.com/electrical-testers/receptacle-outlet-gfci-tester
View attachment 276311

 

[fog37]

I have been thinking: the GFCI outlet has two hot connections (top hot is for the hot line, bottom is hot load) and two neutral connections (top is return line, bottom is return load). I connected the black wires and white wires the way they were connected on the GFCI that I replaced...

However, is it possible that the new GFCI outlet is not working because of line-load reversal: the black line hot wire may be connected to the GFCI hot load connection instead of to the GFCI line hot connection?
Same for the load lines: maybe the load return is connected to the the GFCI line return connection and the line return is connected to the GFCI load return...

Is the line cable always on the left and the load cable always on the right on the inside of the receptacle like in the figure below? MY receptacle looks like that with a cable on lower left and a cable on the lower right...

How can I verify that line-load reversal is not happening using a voltmeter? I guess I can check the voltage between the hot and neutral for each cable with the breaker on and no GFCI outlet connected to identify the load and line cables...

 

[hutchphd]

It is typically marked on the GFCI. With the breaker on, do you get 120 VAC across one cable wire (black to white)? Then put those in the "toward box" (or whatever it says) connection and you should be good to go.

 

[fog37]

It is typically marked on the GFCI. With the breaker on, do you get 120 VAC across one cable wire (black to white)? Then put those in the "toward box" (or whatever it says) connection and you should be good to go.

Thanks!

Yes, the back of GFCI has line connections (lugs) on top and load on bottom. Will get voltmeter tomorrow to make the measurement between hot line and return line. If I get 120VAC, that is the line.

The GFCI outlet currently has a permanent small yellow light turned On and pressing set/reset button does nothing and does not engage...

 

[fog37]

Update. Problem solved :)

I got a non contact voltage tester and identified the live hot wire from the hot load wire. I realized I did a line-load reversal. So all works just fine now.

Thanks for the discussion. It surely forced me to think about the problem and all its facets more deeply.

In regards to the non contact AC tester, I thought it was quite a nifty tool since it can detect voltage from just one wire by using capacitive coupling since the voltage is AC.
First I thought "How is it possible since voltage is always between two wire" but then realized that the other "wire" is, at least I think, the tester and me with my body connected in some fashion to the ground...Is that correct? The change $\Delta V$ is the potential difference between the potential $V_2$ of the live hot wire and the potential $V_1$ of tester+me both poorly connected to the ground...Does that make sense?

 

[berkeman]

Glad that you are back up and running.

Do you still think that the same breaker controls the refridgerator and those outlets? At some point (home remodeling, inspection before the sale of the house, etc.), you may need to have that fixed:

https://www.thespruce.com/kitchen-wiring-circuits-1152911

EDIT/ADD -- Now I'm seeing some indications that it may not be an NEC requirement, maybe just a regional requirement. You can call your local city inspection department to ask.

 

[hutchphd]

Thanks for the discussion. It surely forced me to think about the problem and all its facets more deeply.

Yay

Your analysis of the voltage "sniffer" is pretty good. The key is I believe capacitive coupling between the hot wire and the tip of the wand. One the other side effectively is you (a large capacitor). So AC voltage is induced in the wand and a FET let's us see it.

 

[fog37]

Glad that you are back up and running.

Do you still think that the same breaker controls the refridgerator and those outlets? At some point (home remodeling, inspection before the sale of the house, etc.), you may need to have that fixed:

https://www.thespruce.com/kitchen-wiring-circuits-1152911

View attachment 276367

EDIT/ADD -- Now I'm seeing some indications that it may not be an NEC requirement, maybe just a regional requirement. You can call your local city inspection department to ask.

My understanding is yes: the same breaker controls the refrigerator (upstream of the GFCI) and the GFCI which controls the other two regular outlets. With the GFCI out, the refrigerator kept working but GFCI and the other two outlets didn't.

Thanks

 

[sophiecentaur]

Summary:: which connection is used for household electrical outlets
I am trying to identify which of the three outlets is actually broken

So this daisy chain connection is not a true parallel neither a series connection.

Incorrect. It IS a true parallel connection. Any current through one connected appliance will never pass through another. Don't ever be misled by the route that the actual cables take (easy done). It's the connections that count. All the L cables are at the same potential and all the N cables are at the same potential; that is termed Parallel.

The caveat, stated in some previous posts, applies. If you are not familiar with all of the following then you should really get paid help. (Your series / parallel confusion makes me uneasy.) Fault finding often involves a lot of things in one session and it can be hard to do it through Q and A on a forum. Experienced Electricians tend to do it 'all at once', without needing to think. Solving some problems may take a lot of time. for those with not much experience.

I am late to this thread so I may be repeating the following - in which case, apologies. There is almost certainly no need to examine the buried cables but it's perfectly possible to examine each outlet and the wire connections. You need to isolate everything first - best to turn off at the mains input to the distribution box, perhaps.

Is your GFCI on one outlet or at the breaker box? Or are there two? An outlet GFCI (certainly a replacement one) will not affect other circuits or even other outlets on the same circuit. Your information, so far, is a bit confusing. Is the GFGCI a 'plug in' or is it inside a wired outlet?

Remember that a GFCI can trip even if there is a small current to Earth via the Neutral side. So you need o disconnect both legs of the circuit it feeds. It sounds to me that you could have a Neutral - Earth short somewhere along the neutral side of the circuit. Just to disconnect the live side will not sort that out.
If the apparently faulty outlet is the culprit then (a photo first, if you're not confident) First disconnect the downstream cables. What effect does that have on the situation? The leak could have existed for a long time but may have increased slightly without human intervention.

Disconnect the input cables to the faulty outlet and examine them in detail. With the downstream circuit isolated, a Meter will show any continuity (it doesn't need a total short) between LN,NE and NE wires.

GFCI outlets typically are wired to interrupt both the attached outlet and all downstream outlets.

That sounds very unlikely - although I can't speak for US installations. An outlet with GFCI it there to protect the user of appliances plugged into that outlet alone. Any other arrangement could be very inconvenient. How would you know which other important appliance could be affected if your garden tool chops through its cable? But at the name "Outlet" tells you its purpose. How would the cables at the back be connected? You'd need an extra pair of terminals with a specific IN and OUT function.

A GFCI in the breaker box will, of course switch off the whole of that circuit and it doesn't have an outlet connector at the front.

 

[hutchphd]

That sounds very unlikely - although I can't speak for US installations

It may sound unlikely to you (I don't know why) but it is most certainly true in the US. The outlet box interrupts both sides of downstream connections (including the socket itself). Knowing this fact has allowed me to heroically aid many do-it-yourselfing friends.

 

[sophiecentaur]

It may sound unlikely to you (I don't know why) but it is most certainly true in the US. The outlet box interrupts both sides of downstream connections (including the socket itself). Knowing this fact has allowed me to heroically aid many do-it-yourselfing friends.

Over the years I have come across many differences between US and UK wiring practices. This is one that strikes me as 'inconvenient' (other words come to mind lol). Using that system means that, unless you know the details of your wiring circuit, you won't know whether or not an appliance on that circuit is protected. That strikes me as a serious drop-off. You would need to know whether it is upstream or downstream of the GFCI and that would involve actually activating the breaker and then checking on all (unknown) appliances. In UK, of course, the RingMain system is used so there is no daisy chaining of outlets and there are two paths to any outlet. Any socket on the ring is either protected or not, according to what sort of sub-breaker' is used. But it's a different world and you have to view systems as a whole. If that system is accepted then fair enough but what actually is the advantage of it?

PS While we are on the subject of US mains connectors, why do the pins all seem to have a hole drilled in them, near the end??

 

[hutchphd]

I don't know about the "why holes?" except that occasionally I do some very ad hoc temporary test wiring by putting 18 gauge stranded through them and twisting it on. Also probes and test clips.
All GFCI I know have a test button so you can trigger them and thereby see which things get shut off. It is also quite nice for instance in my garage the first thing I wired incoming is a GFCI outlet and everything else is then automatically protected including lights and switches. In the bath it also means the wall sconce lights can usually be protected easily without a dedicated GFCI breaker in the box.

 

[sophiecentaur]

I do some very ad hoc temporary test wiring by putting 18 gauge stranded through them and twisting it on.

Wow. Sounds like the Wild West. In very nearly every respect (except for the initial choice of Voltage), the UK electrical philosophy seems more fussy about safety issues. The UK Ring Main system is as much a mystery to US and the 'split phase' US system is likewise a mystery to the UK.
The sayings about US and UK being divided by a common language also apply to Electrical matters, I think. It's taken me a long period of exposure to the US point of view to realize that it's really not all that bad. But, if I ever spent time as owner of a US property, I know I'd spend a long while going tut tut and shaking my head when I looked at the domestic electrical situation. And I'd bet the reverse would apply to you.

 

[hutchphd]

, I know I'd spend a long while going tut tut and shaking my head when I looked at the domestic electrical situation. And I'd bet the reverse would apply to you.

Let me assure you the wiring kludges I mentioned are strictly momentary test setups. I did mention above finding aluminum and copper wire twisted into a splice in the open in a wall cavity. No arcing after 30 yrs thank god.
Also 120 volts is much more forgiving than 220 (probably 4 times more forgiving!) It can kill you but you need to be very stupid. My only truly potentially disastrous electrical experiment was me at age 13 with my trusty Van de Graaf and a home-made leyden jar. Lots of energy there. I put it away.

 

[sophiecentaur]

probably 4 times more forgiving!

This is true but electrocution is still extremely rare. This is partly because people are actually 'scared of 240V electricity' (in the same way that they are scared of circular saws), I think. The regulations for using electrical appliances under risky conditions (building sites etc.) require the use of 120V and isolating transformers (and nowadays you find GGCIs everywhere.)

I am more than happy with the available equipment and regs for domestic electrical supply here. Those 'over-sized' 13A fused plugs are not always handy but the RingMain system means that you usually find a lot of outlets in any room so that extra dodgy 'adaptors' are not often needed. EU is also fussy about these things. I guess that my experience of US Electric Supplies is strongly coloured by the posts that you can read on PF, from Off-Grid enthusiasts and the like. Those guys are much less common in UK because (alas) we don't have the open spaces for them but their stories can be read here.

 

[dlgoff]

... but their stories can be read here.

Here'a a story for you. I was helping a lady install an outlet in her small house. She pointed out where she wanted it. Sounded simple enough, just had to cut a hole in some dry-wall, install a box and feed some romex wire down to the box between the wall studs. Well while cutting the hole for the box, fire flew. Turned out the lady had done some electrical work on her own. She had added a circuit to somewhere and buried her wire on the surface of the dry-wall with dry-wall mud. She did have it pretty smooth.

 

[fog37]

Reflecting back on this thread, I was thinking about the following:

In an oversimplified view, devices, appliances, etc. all electrical loads in the house are connected to the panel (which is connected to the electric grid) in a parallel fashion. There are, I believe, sub-circuits in the house. Each sub-circuit is connected to the panel and each sub-circuit comprises several outlets all inter-connected in parallel so the loads connected to those outlets are also all in parallel. A parallel connection is such that if one outlet and/or device connected to it fails, all the other devices continue to operate undisturbed.

In the case of a GFCI outlet, the situation is slightly different: the GFCI outlet is connected in parallel and so are the devices connected in parallel (daisy chained) to the GFCI itself. When a GFCI outlet fails, all the outlets downstream fail too even if they are in connected parallel. My interpretation: when the GFCI fails is it is as if it was disconnected from the panel and so are all the outlets parallel-connected downstream to the GFCI.

Any error in my thinking?

Also, the non contact pen that I used to see if which wire was HOT can only work with AC voltage. The pen would not be able to detect a DC voltage since DC voltage can only be measured considering two points (2 wires). AC potential difference $\Delta V_{AC}= V_2 - V_1$, on the other hand, is a time-varying difference that can be detected even using a single wire. However, since there must two potentials $V$ to take a difference $\Delta V$, The other "wire" can be any other objects with some finite conductivity in the environment.

Thanks as always.

 

[sophiecentaur]

Any error in my thinking?

It's fine if you restrict the description of a GFCI outlet in terms of the practice where you happen to live. Most 'Protected Sockets' in UK will only protect the appliances that are plugged into them. In UK, AFAIK, if you want a protected circuit, then the GFCI will be part of the distribution unit (the 'fuse box'). It worries me that, with the arrangement you describe, there could be one outlet to the right of a GFCI outlet and one on the left of it and you would have no idea which one was protected for a ground fault, without using the test button (or opening up the wiring to look at it).
But, of course, there is no 'downstream' in a ring main circuit and the GFCI would not actually protect any of the other appliances on the ring because the 'other path' would always be connected.
Different strokes.

 

[pbuk]

I have been thinking: the GFCI outlet has two hot connections (top hot is for the hot line, bottom is hot load) and two neutral connections (top is return line, bottom is return load). I connected the black wires and white wires the way they were connected on the GFCI that I replaced...

However, is it possible that the new GFCI outlet is not working because of line-load reversal: the black line hot wire may be connected to the GFCI hot load connection instead of to the GFCI line hot connection?
Same for the load lines: maybe the load return is connected to the the GFCI line return connection and the line return is connected to the GFCI load return...

Is the line cable always on the left and the load cable always on the right on the inside of the receptacle like in the figure below? MY receptacle looks like that with a cable on lower left and a cable on the lower right...

View attachment 276335

How can I verify that line-load reversal is not happening using a voltmeter? I guess I can check the voltage between the hot and neutral for each cable with the breaker on and no GFCI outlet connected to identify the load and line cables...

If you are this uncertain about what you are doing you should not be touching any wiring, call in a professional.

 

[pbuk]

It's fine if you restrict the description of a GFCI outlet in terms of the practice where you happen to live. Most 'Protected Sockets' in UK will only protect the appliances that are plugged into them. In UK, AFAIK, if you want a protected circuit, then the GFCI will be part of the distribution unit (the 'fuse box'). It worries me that, with the arrangement you describe, there could be one outlet to the right of a GFCI outlet and one on the left of it and you would have no idea which one was protected for a ground fault, without using the test button (or opening up the wiring to look at it).
But, of course, there is no 'downstream' in a ring main circuit and the GFCI would not actually protect any of the other appliances on the ring because the 'other path' would always be connected.
Different strokes.

In the UK we use the term Residual Current Device (RCD), not GFCI. The current standard for domestic electrical installation requires that all circuits are protected by an RCD in the consumer unit (distribution board); sockets with RCDs built in are available and can be fitted in a house with older wiring.

It worries me that, with the arrangement you describe, there could be one outlet to the right of a GFCI outlet and one on the left of it and you would have no idea which one was protected for a ground fault, without using the test button (or opening up the wiring to look at it).

Yes, this seems a very bad idea to me. The protected live and neutral on a UK RCD are only accessible through the appliance socket so daisy chaining is not possible.

Discussing wiring on an international forum is not a good idea: what is common practice in one country could be potentially lethal in another.

 

[hutchphd]

Any error in my thinking?

First let me reiterate the "always consult a professional" if you have any doubt. I think there are some US jurisdictions that require professional inspection of your work. This is certainly true if it is a rental property. And the killer issues for me are fire related. Shocks are unpleasant but seldom directly fatal.

That being said I believe your understanding is correct. And the existence of the "ring circuit" was unknown to me until a few years ago. I think it is virtually unused in US. Good luck.

 

[sophiecentaur]

Discussing wiring on an international forum is not a good idea: what is common practice in one country could be potentially lethal in another.

It has its downside but knowing something of what goes on 'over the pond' allows one to make sense of some of the comments that 'they' often make. (This applies in both direction!)
You will always find caveats inside these discussions and the 'compare and contrast' theme runs through most of these threads.
I have a feeling that the 110V system is regarded as much safer than the 240V system and, consequently, people are much easier about regulations and practices. US members always deny this strongly but I can only react to the images of US installations that are often posted on PF. Terrible bodges exist on both sides, of course, but I refer to postings from reputable US engineering types. I'm sure my reactions are based on my automatic assumption that "there's 240V about". Of course, there is a fair amount of 'double Volts' on domestic US. I wonder if there's a European-style worry about that, when they've had generations of 110V built into their instinct.