Finally!!! I saw the light!

[Guido]

I have good news and great news.
The good news is I finally finished running power to my shed... a couple days ago I flipped the light on for the first time (noticed how messy the place really is!!!) :)

I took the power from an existing GFCI outlet box at the back of my house and wired it as in the diagram. Please let me know if it's OK. My circuit tester indicated everything was fine, and the shed is still standing.

And for the great news...
Inside the box I found an ExTRA cable of 14/2 wire!! This HAS to be the extra cable I saw in the breaker box that I had no idea what it was for. Seems someone ran this extra cable and didn't get around to connecting it - so now I can run TWO 15A circuits to the shed. Thank God to the person who suggested I run and extra wire in the conduit "just in case".

[Guido]

And when I get the chance I plan on wiring the extra circuit like this diagram.
I don't want either circuit going to the shed GF protected, BUT, I plan on having a GFCI receptacle as the LAST item on ONE of the circuits. Any problems with doing this?

[Mr Fixit eh]

I'm glad you've met with success!

Great diagrams, too. What program are you using to draw them?

A couple comments...

1. Your switched light will be very dark (aka non-luminous) when the GFCI receptacle is tripped. If this is the only outside light, this could be inconvenient.

2. 14AWG cable from the panel means that even though you've run 12AWG to the shed, you must use a 15A breaker, not a 20A -- can't remember -- this was probably your intent, anyways, with the 12AWG giving you room to expand in the future when you can run new cable from the panel to your box.

3. It's always pleasant to get a good surprise! When you activate the "spare" wire, the two circuits will need to be on a tied breaker, so that both circuits will trip if any one of them trips.

You must have a little post-traumatic stress after making all those connections in 12AWG......congrats on a job well done.

[Guido]

I'm glad you've met with success!

Great diagrams, too. What program are you using to draw them?
Thank you!
The program is called Interleaf - and is extinct.


1. Your switched light will be very dark (aka non-luminous) when the GFCI receptacle is tripped. If this is the only outside light, this could be inconvenient.
I can't decide between having the light go out, or taking the risk of getting a shock whenever I flip the switch if there's water in it or if I'm changing a bulb and there's moisture trapped in the socket. Although the light is about 3 feet above the outlet box, I plan on extending the wire up into the soffit and horizontal about 10 feet to my porch. I was wondering what would happen if an animal chuws into the wire - I don't want it grounding on the soffit (assuming the breaker doesn't trip). So I'm stuck deciding between convenience and safety.




2. 14AWG cable from the panel means that even though you've run 12AWG to the shed, you must use a 15A breaker, not a 20A -- can't remember -- this was probably your intent, anyways, with the 12AWG giving you room to expand in the future when you can run new cable from the panel to your box.

Yup, I'll be using a 15A breaker. My original plan was to run another 12AWG cable from the box to the panel and install a 20A breaker, but now that I found the extra cable I'll just run two 15A circuits. And since I'm doing this I'll return the 12/2 cable I bought to wire the shed interior, and buy 14/3 instead.
Also, the 12AWG wire is 16m long so I don't mind having it a little larger.


3. It's always pleasant to get a good surprise! When you activate the "spare" wire, the two circuits will need to be on a tied breaker, so that both circuits will trip if any one of them trips.
This is a code rule because both circuits enter the same box right? I would like to have them separate so if I trip one breaker I can still run the lights on the other circuit. But if it's a big no-no then I'll do what you said and use a double breaker.



You must have a little post-traumatic stress after making all those connections in 12AWG......congrats on a job well done.
Thanks again.
It was actually not bad making the connections because the 12AWG is multi-strand. The worse part is tightening the #12 under a screw but Homer suggested to use spade wire terminals and I bought some. But with the particular connection above I didn't need any because the #12 wires are spliced to the #14 using wire caps.

[Homer]

Any problems with doing this?It looks like you're trying to create a multiwire circuit from 2 individual circuits. A multiwire circuit should begin at the panel, not at a device box.

How difficult is it to pull 12/3 or 14/3 from the panel to that box?

Generally, there's a 'Conductors of the same circuit' provision that states all conductors must be in the same cable.

Wait for others to comment on this arrangement. It may violate NEC §300.3(B) and CEC 12-106.

Homer

[Guido]

It looks like you're trying to create a multiwire circuit from 2 individual circuits. A multiwire circuit should begin at the panel, not at a device box.
OK now when you say "should", is that because of the code, or is it a real safety hazard? I will have a 2-pole breaker so both circuits will be off at the same time.

How difficult is it to pull 12/3 or 14/3 from the panel to that box?
Extremely. The panel is at the front of the house and the box is at the back, and the basement is finished.

Generally, there's a 'Conductors of the same circuit' provision that states all conductors must be in the same cable.

Wait for others to comment on this arrangement. It may violate NEC §300.3(B) and CEC 12-106.

Tonight I'll check if my P.S. Knight's book mentiones this. I couldn't find any info on the 'net.

[Mr Fixit eh]

How hard would it be to run a second neutral wire through your conduit?

I was wondering what would happen if an animal chuws into the wire
Cable runs on the exterior of the building must be protected from physical damage. That would normally require sleeving with PVC.

I can't decide between having the light go out, or taking the risk of getting a shock whenever I flip the switch if there's water in it or if I'm changing a bulb and there's moisture trapped in the socket.

If the light fixture is wired with correct polarity, there should not be a risk of shock. For the few times you expect to change the bulb, you could shut off the circuit when it's time to change the bulb.

It seems odd that the switch is on the exterior. Is there any way to move it inside? It really should be protected from the elements, or you could look for a waterproof switch.

[Guido]

How hard would it be to run a second neutral wire through your conduit?
Physically it'll fit. It's not a piece of cake, but it's do-able if necessary. I have a feeling it's not allowed - something I read about you're not allowed to run two separate circuits to an outbuilding (a 3-wire with two circuits is considered one).

Cable runs on the exterior of the building must be protected from physical damage. That would normally require sleeving with PVC.

If the light fixture is wired with correct polarity, there should not be a risk of shock. For the few times you expect to change the bulb, you could shut off the circuit when it's time to change the bulb.

I am using PVC conduit on the exterior of the house, but where I run inside the soffit I was thinking of not using conduit (much easier) and it's protected from the elements anyway.
But you're right about changing a bulb. I never got shocked before but I figure since I'm wiring it "from scratch" I might as well make is safer than required. But it looks like I'll take the easy way out and do it the "traditional" unprotected way. Not a big deal.


It seems odd that the switch is on the exterior. Is there any way to move it inside? It really should be protected from the elements, or you could look for a waterproof switch.
Actually I just installed it last weekend. It's in a weatherproof 2-gang metal box. The cover has a GFCI flip cover on one side, and a switch cover on the other. Only $8 at HD (and $13 for the box).

[Roger]

Guido, as I see the diagram you would be in violation of NEC codes 300.3 b and 300.13 b and maybe the wiring methods for multi-wire branch circuits originating in the serving panelboard. You have joined the grounded conducters of two different circuits creating a neutral loop is in violation of the NEC. You are also serving the ungrounded (Hot) conductor of circuit one with the grounded (neutral) of circuit two, at the gfci. I dont see how this wiring design is going to be acceptable as a multiwire circuit serving your shed.

[Homer]

OK now when you say "should", is that because of the code, or is it a real safety hazard?Both.

When you split the circuit into more than one path the currents in the conductors of each cable have the potential to be unbalanced. By unbalanced I mean that the current flowing down the line in one conductor will not equal the current returning since some may take the alternate cable path.

When current is unbalanced in a cable it can cause inductive heating in ferrous alloys (like steel boxes) from induced eddy currents in the metal.

Remember that with a proper multiwire circuit the currents in the 3 wires always balance. Whatever is unbalanced between the Hot conductors is carried by the neutral so the sum in each direction is equal.

Homer

[Guido]

How about, to eliminate the neutral loop and the unbalanced neutrals I do the following:

120V from cable 1, black wire
Neutral from cable 1, white wire
Bare ground from cable 1
120V from cable 2, black wire that shares the neutral and ground of circuit 1

Basically it's a 3 wire circuit (shared neutral and ground) except the extra power wire comes from another cable. The wire itself doesn't know what cable it's in so it should work fine. And the unused neutral and ground in cable 2 gets taped and tucked away. This should keep everything balanced right?

I thought I hit a gold mine when I saw the extra cable in the box - there has to be a way that I can use it in my shed (safely).

[Mr Fixit eh]

Ok, Homer, you're testing the limits of my knowledge of electrical theory :eek:

When you split the circuit into more than one path the currents in the conductors of each cable have the potential to be unbalanced.

I'm not understanding the "split into more than one path" part. This is two circuits trying to share, right?

Are you saying that if current from both circuit hots, share 1 neutral wire in the conduit, the neutral wire may not be capable of carrying the current? E.g. 15A on one hot, 15A on 'tuther, 30A trying to return on the neutral?

Us Canadians, kinda slow on the uptake (well, speaking for myself, anyways. :D

[Guido]

I'm not understanding the "split into more than one path" part. This is two circuits trying to share, right? :D
Actually what I'm trying to do is run two 2-wire circuits from the panel and convert them into a single 3-wire circuit by capping all the neutrals together. So in the 3-wire circuit the current will flow down the single neutral, but on it's way back to the panel it'll hit a wire cap and split into two neutral wires in different cables. This is where Homer started talking about potential unbalance, and inductive heating, and cross-pollination and stuff. :D

So my plan is to only use one neutral and leave the other neutral unused to confuse the hell out of the next home owner. I'll leave him a note with this website address to help him figure out what's going on. ;)

[Homer]

Basically it's a 3 wire circuit (shared neutral and ground) except the extra power wire comes from another cable. The wire itself doesn't know what cable it's in so it should work fine.I think that you've missed the entire point here. Of course it will still 'work' but it will not be 'fine'.

If we look at this sketch it should be clear that you are creating an inbalance even if you don't use one of those neutral wires.

Consider situation #1. Equal loads on the multiwire legs of 10A.

Cable C is balanced with 10A on the black wire, 10A on the red wire, and 0A on the white. Remember that the red and black are on opposite legs so the currents go in opposite directions at any given instant.

Cable A is unbalanced. The black wire carries 10A and the white carries 0A.

Cable B is unbalanced. The black wire carries 10A and the white carries 0A.

Now, consider situation #2. Loads of 10A and 5A.

Cable C is balanced with 10A on the black wire, 5A on the red wire, and 5A on the white. Remember that the red and black are on opposite legs.

Cable A is unbalanced. The black wire carries 10A and the white carries 2.5A.

Cable B is unbalanced. The black wire carries 5A and the white carries 2.5A.

In cables A & B, most current flows in one direction so the magnetic field will be strong. In a balanced case the field would be relatively weak. Remember how a transformer works. All the current flows in one direction around a ferrous core.

Homer

[Mr Fixit eh]

*Finally I saw the light!*

Amazing schematics, Homer. Thanks for your patience.

I'm assuming that the potential for cable overheating multiplies with the amount of load(s) applied and the net imbalance. Is there the possibility of causing a fire?

In the OP's scenario there is the possibility that the two hot supply wires might be on the same phase. If this is true, what would happen? No imbalance in the sense you described, but the shared neutral might then be overloaded. E.g. 10A on each hot leg, and 20A returning on the shared neutral?

[Homer]

I'm assuming that the potential for cable overheating multiplies with the amount of load(s) applied and the net imbalance. Is there the possibility of causing a fire?Yes, the higher the load imbalance the greater the problem. Although I wouldn't expect a true fire hazard, the result is an unexpected and unmeasurable derating of the conductor ampacity.

In the OP's scenario there is the possibility that the two hot supply wires might be on the same phase. If this is true, what would happen? No imbalance in the sense you described, but the shared neutral might then be overloaded. E.g. 10A on each hot leg, and 20A returning on the shared neutral?Yes, you've got it figured perfectly. The main concern with multiwire circuits in making sure that the Hots are on opposite legs.

Homer

[mdshunk]

Be advised that this installation that you propose is a big red flag waving, flaming code violation no matter where you live. The inductive heating danger that Homer points out is a very real situation. You mention that these two 14-2's arrive in conduit. Is this metal conduit? You could make a pretty fine heating element over time. Are any of the boxes associated with this circuit metal? Do both of these 14-2's go through the same connector on the panel? If not, you could heat at this point also. The only good way to solve this problem would be to run one 14-3 from the switch/gfci to the panel, OR run two 12-2's to the shed instead of the current 12-3. While I fully support a man's right to do as he pleases at his own house, be advised that your proposal is wrong and a potential hazard to the occupants. While it probably won't cause a fire, it will certainly break down the conductor insulation pretty good due to this heating.

[Roger]

Guido if you wanted advice on investing money would you seek it from someone who is poor and has nothing or would you listen to someone who has been successful at it?
Use your common sense here and provide service to your shed correctly. Then people will say you did it right and knew what you were doing. Why have someone come behind you someday and wonder who the heck did this mess.

[Guido]

Sorry to be such a pain. Thanks for the diagram Homer. I understand it 100%, but I think I'm trying to do something different. Before confusing the matters more I'll start with the basics.

Is this the correct way to wire a split receptacle?

[Guido]

Be advised that this installation that you propose is a big red flag waving, flaming code violation no matter where you live. The inductive heating danger that Homer points out is a very real situation.
Thank you for the advice. That's why I'm here - to make my mistakes on the computer, not the house.

You mention that these two 14-2's arrive in conduit. Is this metal conduit? You could make a pretty fine heating element over time. Are any of the boxes associated with this circuit metal? Do both of these 14-2's go through the same connector on the panel? If not, you could heat at this point also.
Actually these 14-2's both poke out of the brick wall into the outside metal box that houses the GFCI (and a switch for a light). They are Romex and I assume they go to the panel inside the wall without any conduit protection.
Right now only one 14-2 is attached to a breaker in the panel (the original one for that receptacle). The other 14-2 is taped in the panel and not being used.

Going to the shed I have three #12 T-90 wires (plus ground) in a PVC conduit.

The only good way to solve this problem would be to run one 14-3 from the switch/gfci to the panel, OR run two 12-2's to the shed instead of the current 12-3.

I'm not able to run a new 14-3 to the panel because it's at the other side of the house and the basement is finished. I could always run it along the exterior wall - this was my intention (to be completed when the warm weather returns) but since I saw the unused 14-2 in the box I thought I was lucky.

Assuming I can fish another two #12 wires in the conduit to the shed (which I doubt), I thought it wasn't to code to run two separate circuits to an outbuilding.



Roger, please don't take offense to my basic questions. Every electrical project I did followed the advice and recommendations of people on this board who are obviously more knowledgable than I. Maybe it's because I'm an engineer, but even though I believe what's being said to me is correct, I would still like to understand WHY I'm doing it a certain way - and hence the basic questions. If I keep drilling at a point it's because I don't understand, NOT because I'm trying to outsmart anyone.

[mdshunk]

Assuming I can fish another two #12 wires in the conduit to the shed (which I doubt), I thought it wasn't to code to run two separate circuits to an outbuilding.That's probably true in Canada.

[Wgoodrich]

I opologize for jumping into a long discussion in the middle well after it has been discussed and rediscussed so much. However after reading along today from what has been said I am picking up that we have a house with a detached structure with a conduit running from the main house to the detached structure in the form of 2 nonmetalic sheathed cables inside a single conduit feeding as power to this detached structure from the main structure. There are several side discussions being worked in the same post at the same time. The entire post is hard to follow but I am getting the following understanding as to what is existing and what is trying to be done. Please correct me if I am way off base in my understanding. Not easy to jump in the middle of a discussion but just trying to help out in understanding of this subject posted.

I am picking up that 2- 14/2 Romex cables are going to this detached structure and it is being attempted to increase load capacity to that building by combining the two cables to serve that one structure.

I am seeing two power sources to a structure which would be a Code violation. Two power souces invites someone to mistake thinking they turned the main power off to that certain building and a second breaker also serves power to that same structure inviting a shock hazard. This is not allowed in the CEC or the NEC.

I am picking up the above to have been discussed then thinking about connecting the two cables together to make one multiwire feeder serving that one structure. This also would violate standards in both the CEC and NEC.
One violation is both Codes forbid power conductors to be paralleled in sized smaller than 1/0. These discussed being 14 awg violating a paralleling rule of no smaller than 1/0 conductors allowed.

Also there is a rule in both Codes forbidding an installation of smaller conductors than 1/0 not only being paralleled but being in different cable sheathings. Both Codes require all conductors of a branch or feeder to be contained inside the same cable sheathing whether contained in the same conduit or not.

When connecting two neutral conductors together at each end you created a parallel design however they are not ran together creating concern of harmonic heating because these neutrals are in different cables whether they are in same conduit or not this is a bigger concern with smaller conductors smaller than 1/0 such as the 14 awg conductors being mentioned.

The main reason of concern about paralleling smaller conductors is the variable tolorence allowed in larger conductors not being a problem as much as those same tolorences less forgiven in the smaller conductors. Much more reaction of imbalance and harmonic heating appear in the smaller conductors than the larger conductors. To reason this out take a torch and heat a 1/4" bolt and a 1" bolt and see how much faster that 1/4" bolt will turn read than that 1" bolt. Mass resists heat influx much better when larger mass than smaller mass.

Installing two neutrals connected together at each end yet not running them in the same sheathing is the cause of concern here especially being the smaller conductors than allowed to be paralleled which compounds the heating concern as explianed above.

Hope this helps

Wg

[Guido]

Thank you for your input Wgoodrich.
I know it's difficult to track what's going on, I'll try to summerize:

From the main panel to an exterior box attached at the rear of the house exists two 14-2 Romex cables. Currently only one cable is connected to the panel, the other is just taped up and unused.

From the same exterior box to the detached structure (shed) there is PVC conduit containing three #12 T-90 wire (and a #12 ground).

My goal is to bring two 15A circuits to the shed.

If I understand correctly from what is being said, there is NO WAY I can use both 14-2 cables AND the three #12 wires (in the conduit) to bring two 15A circuits to the shed. My only option is to install 14-3 (or 12-3) from the box to the panel (since code doesn't allow multiple circuits to the shed then there's no point in installing extra wires in the conduit).

Now I know what my options are to remain in compliance with the code.

I also understand what was said regarding the neutral loop and would happen if using two neutrals.

Now here is my "academic" question for strictly learning purposes...
What if I use ONLY ONE neutral? And leave the other unused. This is how I envision it:

Two hot WIRES coming from the same 2-pole breaker in the panel. Each wire is part of a different cable (I know this isn't to code). One single neutral wire that the two hot wires share (doesn't matter from which cable this neutral wire comes from, as long as only one is used). One single ground.
I don't see a neutral loop in there and I don't see any difference than your typical 3-wire circuit (hot/hot/neutral). Everything should be balanced no?

Below is a modified version of Homer's drawing of what I'm talking about. Basically, if you ignore the code - is there still any danger of wires heating up, etc?

[Wgoodrich]

First is must be said to ignore the minimum safety standards making you your own design is setting you up to losing a liable suit later because you did not meet minimum safety standards set by rules of law as minimum wiring design. YOu just lost the protection of meeting minimum standards and accepted all the risks on you.

We never advise to do less than meeting the minimum safety standards.

However if discussing theory of engineering only hypotetical.

By the seat of the pants the chances are slim heating would occur enough to cause a concern due to harmonics as long as both cables are inside the same conduit. They are in close enough proximity to help counter the affects of harmonics to a degree. However the limits are being pushed due to the resistance of communication between the two hot conductors and neutral conductor by the insulation of the wire normally there and the added noncurrent carrying resistance of the sheathing of both cables. Then the concern of proximity would be involved creating a larger distance between actual conducting metals of each conductors separated more due to the thickness of the two sheathings of the two cables. Normally all three insulated conductors would be touching commonly while inside that conduit the entire run. When within two cables sheathings of nonconductive material created by the two cables these conducting metals can never get close enough to touch. While they will try to communicate the farther apart the are from each other the more heat is produced trying to convey the harmonic affect creating more heat.

Picture two unlike magnets attracted to each other when they are touching it is like they are married and happy. The more you pull those magnets apart the harder you must pull to break that marriage. Same is happening in laymen terms in the harmonic concern above.

YOu could have a melt down with this design you are thinking of yet you may last for years without experiencing anything. This design and predicting what would happen is much like sitting on a rodeo bull about to bust out of the gate. While it will burn one building down the next building may never experience anything all depends on loads, timing, and cycling of phase. That bull may have killed the previous rider never allowing anyone to ride him yet you jump out of the gate and you may ride the whole 8 second ride and win the contest with no trouble then next guy rides him again gets killed. Both are like a wild child all depending on the elements every 60th of a second the life of the bull or the life of those incorrectly installed conductors.

Harmonic heating while being aware of the affects it is only in the last decade the Code started realizing many shopping malls burnt down because of those harmonic affects in the past yet many malls did fine without occurrance. Harmonics has a lot of unknown just like space and the galaxy. We are not sure how far we can push the harmonic subject or when it will bite us.

You are thinking of an electrical design that is much like a wild child that may do anything at any given time. Too many variables in the life of its use concerning harmonics.

No one can accurately predict when harmonics will cause a meltdown due to the variables. Anyone stating otherwise is a person to doubt.

Hope this sheds some light on the subject of theory and why and if.

Wg

[Guido]

Thank you so much for that explanation - FINALLY I got it! The reason I was so confused was because I thought that the harmonic heating was caused by the neutral loop so I figured just get rid of the loop and all will be OK. But if I'm reading correctly it's actually caused by the excessive distance between the hot and neutral wires resulting in diminished "comminication" between these wires. I never knew such a thing existed.
I'm going to read up on that.

So I guess I'm back to square one - I'll wire a single circuit to the shed and if I need more power I'll just run a temporary extension cord from the house.

Thanks again to everyone for their patience and time in explaining this to me. I really appreciate it.

[Guido]

OK I did some research and this is what I found:

90% of me searching leaded to harmonics one 3-phase circuits. Not too much on single-phase, but it still is a problem.

Harmonics are created by nonlinear loads such as personal computers, printers, fluorescent lights, adjustable-speed motor drives, and certain solid-state equipment such as battery chargers and electroplating supplies.

One thing that opened my eyes was a few sites mentioned electronic ballasts as a producer of harmonics. I just bought two shop lights for my shed that have electronic ballasts. I also have variable speed tools. If it's all used on the same 14-2 circuit (no shared neutral) is there a concern for harmonics?

[Wgoodrich]

In my years of attempting to explain in very down to earth language that normal joes can understand I came up with the following that seems to help some.

Generically speaking when concern of harmonic and how it works picture two hot wires carrying 120 volt loads with electronics involved [the electronic {nonlinear load} tends to distort the sine wave if seen on an ascilliscope test equipment causing much of the heating concern.

When the above conditions occur it is much like rapidly rubbing your hands together tightly. This can heat your hands to the point of pain. Picture the two hot wires creating 220 volts of a single or three phase system as being your two hands rubbing together. This is generally speaking the heat produced being the two hot wires of three in three phase or two hot wires of single phase running opposite direction of flow in the cycling of AC voltage much like your hands rubbing together creating the heat.

Now picture grease applied to your hands that will stop the friction causing heat being produced.

Now picture the neutral wire being the grease for those two hot wires.

Generically speaking the above is what is happening when harmonic problems occur.

You will hear more about harmonics as the electrical industry grows and matures. Also you will hear much more about poison of health due to magnetic atmosphere around electricity as well as transient volatage concerns.

These four electrical subjects will become more in the lime light as more and more is discovered and understood in their affects.

Looking to the future of electrical industry

Wg

[Wgoodrich]

Reply didn't advance

Wg