Search the forums

Loading

Floating neutrals

prz's picture

Can someone clarify the term "floating neutral" in relation to portable generators and what consequenses using a floating or loose neutral pose when hooked into either a transfer switch or as an improper "hot wire" directly to a main panel.

(post #114611, reply #1 of 29)

Electricity wants to 'go home.' That is, it will try to return to where it was made. The electricity from the PoCo tries to return to the PoCo; the power from a generator tries to get back to the generator.

A "Floating neutral" is a neutral that has no direct path home. So, rather than follow a defined path, it takes any and all paths it can find.

When this happens, there can be wild voltage fluctuations. The ground paths can act as a return path, setting the stage for a shock. The earth itself will carry some current, with the result that there will be 'voltage gradients' around the house ... touch a door, get zapped.

The end result: things break and people get hurt.

(post #114611, reply #2 of 29)

I've been wondering for years what the electricians were talking about with that phrase. Thank you.

Now is that what happens if the neutral and the ground are bonded in a sub-panel? A floating neutral?

What other conditions (specifically) can cause this?

I got shocked pretty good leaning against a metal raingutter downspout at a cheap motel in North Texas one time. Maybe that's what was happening.

(post #114611, reply #3 of 29)

If the neutral and ground are bonded at more than one place, they will divide the load. That is, the ground will always be carrying current. Open the ground ... say, by taking a receptacle out of a box .... and one side of the ground will be 'hot.' Bad things can happen then. These are the 'parallel paths' we try to prevent.

This is why we bond them where the PoCo comes into the house ... and nowhere else.

The shock you got probably had a simpler cause ... like a siding nail piercing a wire.

(post #114611, reply #4 of 29)

Thanks.

There is so much to know about electricity.

I know an electrician's phone number, that's it. <G>

(post #114611, reply #6 of 29)

If the neutral and ground are bonded at more than one place, they will divide the load. That is, the ground will always be carrying current. Open the ground ... say, by taking a receptacle out of a box .... and one side of the ground will be 'hot.' Bad things can happen then. These are the 'parallel paths' we try to prevent.


This is why we bond them where the PoCo comes into the house ... and nowhere else.


This is absolute nonsense. In the first place the neutral is tied to ground at every  transformer along the line and their are multiple reasons to have more than one ground connection and  they are called for in the NEC.


Jack

(post #114611, reply #7 of 29)

Please enlighten me as to the NEC requirements. I am not aware of them.

I speak only to the 'consumer' side of the electrical system. The PoCo is an entirely different animal. For example, transmission lines typically do not have a neutral. There are other details as well ... details that, if you tried them at home, would kill you.

(post #114611, reply #10 of 29)

One of the cases which you can have more than one ground if the sub panel is in a completely separate building it is permitted to have a separate ground rod and ground connection at the sub panel and you would then what the two to be bonded.


Ground is the only conductor that is not suppose to be carrying current unless there is a fault of some kind.


There is no such thing as a neutral wire in AC circuitry , this is a misnomer. The so called common wire carries just as much current and voltage as the hot wire, we generally don't see that because the common  is referenced to ground by the ground connection.


Jack

(post #114611, reply #11 of 29)

The example you cite - of a detached building - goes away in the 2008 NEC. Even before then, the safety of the system was dependent upon there being absolutely NO metallic paths between the buildings - no conduit, no water pipes, etc.

I'll not enter the 'define neutral' debate. As uses with household power, my use conforms to the IEEE definition, as well as common usage. Moreover, 'neutral' is also used in the NEC.

Nice try.

(post #114611, reply #12 of 29)

"There is no such thing as a neutral wire in AC circuitry , this is a misnomer. The so called common wire carries just as much current and voltage as the hot wire, we generally don't see that because the common is referenced to ground by the ground connection."

You don't have voltage on "any wire".

Voltage is the potenial between 2 points in a circuit. So waying a wire has voltage is meaningless.

But the term is often misused and the mean is implies. That is saying a wire has voltage means that is when measured against the NEUTRAL.

So measure the neutral against the NEUTRAL would always be zero or near zero if there is current in the wire.

And system neutrals (not just the neutral of a single phase circuit) the currents don't equal what is in the hot. The currents are additive. And the phases are such as they usually only show the difference in currents (except for odd harmonics).

"One of the cases which you can have more than one ground if the sub panel is in a completely separate building it is permitted to have a separate ground rod and ground connection at the sub panel and you would then what the two to be bonded."

Actually you can have any number of "grounds" (ground electrodes) and the code often requires multple ones.

The only limitation is where the "grounded conductor (aka neutral) is bonded to the ground electrode system.

And your statement that you can have additional bonding in a separate structure is not completely true.

It is only allowed if there is no equipment grounding conductor run and also no other metallic pathes between the buildings.

Also the separate structure requires a ground electrode system no matter regardless of whether the neutral is bonded at the structure or not.

.
.
A-holes. Hey every group has to have one. And I have been elected to be the one. I should make that my tagline.
. William the Geezer, the sequel to Billy the Kid - Shoe

(post #114611, reply #13 of 29)

Back in the ancient times of discrete component electronics when I used to be an electronics tech, we often ran into what was termed "floating ground" in which the chassis of a radio or other electronics device was not connected directly to ground but, however, was used as a connection point for the common connections of various electronic components. The resulting voltage (in reference to ground) was below line voltage, but was still often enough to give one a bit of a buzz...

Wikipedia has an article that mentions floating ground (they also have one on floating neutral) but I only glanced at it:
http://en.wikipedia.org/wiki/Ground_(electricity)

(post #114611, reply #14 of 29)

Bill as there is no neutral or common in a 220 circuit, every one knows that both wires are "hot", but one wire in a 120volt circuit is labeled neutral or common, my point was that it is actually a return carrying the same current as the hot wire of the circuit not that it carries the same amount of current as all the hot leads in a house just like each hot wire in the different circuits don't carry the same current as the main.


There exists a difference in potential between all wires of an AC circuit and ground this potential is called voltage, I did error in using the term "on the wires" and not referencing it to ground. This difference is removed of potential from the common wire by bonding or referencing it to ground. Without this bond you would be just as likely to be electrocuted by the common wire as you would by one of the hot wires.


My other point was that there are times that other grounds are used.


 


I may not have been clear in my description but the concept I believe was correct.


The comment I made about his post being nonsense was when he said if you you remove an outlet you could cause the ground to be hot. I guess I should have been clearer and apologize for any misunderstanding


Jack


Edited 11/29/2007 8:51 pm ET by JLMCDANIEL

(post #114611, reply #15 of 29)

What the heck did I stir up with this topic? I still am looking for a simple confirmation of my basic premise regarding the type of portable generator to be used ("floating neutral" vs "neutral bonded to ground" - yeah I know, somebody is going to jump on me for using the term "neutral" instead of grounded conductor, but I think the term neutral is universally understood, and will not get mistaken for grounding conductor, or do I have it all backwards now???) when connected through a generator transfer switch, or as many people are doing in hot wiring their main panels directly (and by many people, this includes trade professionals, even though the practice is illegal). Refer to message #10 for my assumptions.


Sorry to cause all the trouble I stirred up but thanks for the good discussion anyway.


Edited 11/29/2007 10:33 pm ET by prz

(post #114611, reply #18 of 29)

Please correct me if I'm wrong.  An AC circuit is 60 Hz, reversing current direction 60 times a second with a sine wave rise and fall.  To me this means that 60 times every second the neutral is the hot, so there is really not a hot or neutral per say.

(post #114611, reply #19 of 29)

Actually it reverse every 120 times a second. And have 2 reversal it is back to the orginal. that 60 times a second.

But the instanous polary has nothing to do with which is the not and which is the neutral (grounded conductor).

Look at auto/tractor/small airplane wiring.

The use the chassis for the return wiring. Today all of them (at least 99.93%) tie the battern negative termina to the chassis and the hot lead is the positive connection.

30-50 years ago some used what was called a negative polarity system. That is the battery positive was connected to the chassis (ground) and the hot lead was the negative connection.

The unground lead was still the hot, regardless if it was positive or negative with respect to the common ("ground" or chassis).

.
.
A-holes. Hey every group has to have one. And I have been elected to be the one. I should make that my tagline.
. William the Geezer, the sequel to Billy the Kid - Shoe

(post #114611, reply #21 of 29)

Why does polarity have nothing to do with hot or neutral?  I thought that was what it was all about.

(post #114611, reply #22 of 29)

A neutral is a point where you have currents sum from 2 (or more for 3 phase) loads so that the currents, at an instant of time tend to cancel (and ignoring reactive currents). For AC, as you go through the cycle You might have L1 positive to the neutral and L2 negative to the neutral. But /120 second later it is reversed.

For a simple 2 wire (120 volt in the US) circuit there really isn't true neutral. The more accurate term is grounded conductor.

But at the panel or for multiwire circuit, as some point you have the hots from both legs feeding loads and the neutral only carries the net difference in currents.

.
.
A-holes. Hey every group has to have one. And I have been elected to be the one. I should make that my tagline.
. William the Geezer, the sequel to Billy the Kid - Shoe

(post #114611, reply #23 of 29)

My head is hurting!  So if I understand, a two phase circuit does have a neutral ( they have to share the one neutral wire for this to happaen?), but my house circuits (120 - one phase) do not really have a neutral, just a grounded conductor? 


Are the two hots coming from the transformer to my house different phases and if so how far out of phase?  Why don't they drop voltage returning or fighting with each other over the others hot leg?

(post #114611, reply #24 of 29)

With convential 120/240 service the two legs are exactly 180 degrees apart. It is generated by the secondard of a transformer that is center taped to form the neutral connection.

(Note that is condo and apartment buildings each residence often gets 2 legs of a 3 phase service and the voltages are 120/208. But that gets into more confusing questions).

http://en.wikipedia.org/wiki/Split_phase

Note that the UK also has a 110 center tap system. The center tap is ground, but not connected. So no current flows through it in nomral operation. The loads are connected hot to hot leg.

That is basically what happens in the US for 240 volt loads. But the US has both 240 and 120 loads so that the neutral often carries current.

This might be a better discussion.

http://www.allaboutcircuits.com/vol_2/chpt_10/1.html

.
.
A-holes. Hey every group has to have one. And I have been elected to be the one. I should make that my tagline.


Edited 11/30/2007 3:43 pm by BillHartmann

. William the Geezer, the sequel to Billy the Kid - Shoe

(post #114611, reply #25 of 29)

You are right about the sine wave, but see my previous posting above - if the neutral is connected to the ground and the ground is the zero reference, then the voltage on the neutral at the bonding point is always zero, literally by definition.

(post #114611, reply #27 of 29)

Actually the neutral is "hot" all the time with the current flowing in one direction for half a cycle then in the other for the other half of a cycle and this is 60 times per second not 120.


Jack

(post #114611, reply #20 of 29)

The so called common wire carries just as much current and voltage as the hot wire, ....


For residential home wiring, that is often NOT the case.  The classic example is the split receptacle - given balanced loads, the current in the neutral is 0 and the current in both hots is the same.

(post #114611, reply #26 of 29)

The only time in residential wireing the current flow in the common is 0 is when the load is purely a 220 volt. Electricity can not flow in a circuit that has no return.


Jack

(post #114611, reply #28 of 29)

The only time in residential wireing the current flow in the common is 0 is when the load is purely a 220 volt. Electricity can not flow in a circuit that has no return.


It's true that electricity does not generally flow when there is no "return", but often that return is not what you might expect.


However, it is NOT accurate to say that the current in the neutral is 0 only when the load is 220 VAC.  In the example I cited, the loads are 110 VAC and there is no current flow in the neutral.

(post #114611, reply #29 of 29)

I hate to disagree with you but a split receptacle requires a common connection to work, unless you have an insight to a device I am not familiar with. Would you please explain.


Jack

(post #114611, reply #5 of 29)

A floating neutral is one that is not referenced to a common potential, usually ground. With a generator,if it is not properly grounded the neutral would be floating and if you tried to read the voltage from neutral to ground it may read a few volts or blow your meter up. That is why it is important to have the generator properly grounded before you hook it up.


Jack

(post #114611, reply #16 of 29)

"That is why it is important to have the generator properly grounded before you hook it up."

?? not home to check, but I don't recall the little suit-case sized Honda I bought during the
blackout in Buffalo last year having any special grounding procedures....certainly none required if using it as a simple generator, i.e. extension cord to device. Are you referring to using a generator to feed house wiring?

Aha, read further, thanks.......


Edited 11/30/2007 12:57 am ET by jrnbj

(post #114611, reply #8 of 29)

The section in the NEC on grounding is one of the largest in the book and probably the most complicated.  The lines coming in from the power company to a house are two hot ones and a neutral (grounded conductor in NEC language, the white wire that is the return path for the current from the two hot conductors); this neutral may have been tied to a ground rod at the transformer, but you can't count on it being at ground potential (voltage) at the house.  At the service entrance panel (usually meter + main CB disconnect or fuses), the neutral must be connected to a ground electrode system, usually a ground rod and the buried water pipe if it is metal.  This insures that the house neutral is at ground potential (the voltage from the neutral to earth at the SE panel should be 0).  This ground reference is then carried throughout the electrical system by the grounding conductor (I know this is too close to the above, but that is what the NEC calls it and that's the bible for the electrical world); this is the green wire in conduits and the bare wire in Romex.  This grounding conductor is never supposed to carry current in normal operation and it should run like a tree - always branching out and not connected back on itself.


A portable generator does not have a reference to ground as long as its metal parts are not touching the ground; hence it's neutral is "floating".  If you use it to feed a panel that has the neutral connected to the grounding system, then the voltage from the generator is referenced to ground just as that from the power company is.  I've done that a number of time to supply power for lighting during construction when there is no other temporary power.  What you have to be careful of is that the system in the house should be grounded and there must be no connection to the power company lines (this would be illegal because it could back feed into their system).  There is a minor problem with feeding a sub panel because the neutral is not tied to ground there, but that hasn't caused me any problems in the past.  Also, the panels are designed to have two phases coming in, and there is only one 120V from most generators (there's usually no neutral on the 240V outlet), so only half of the circuits will be powered unless you do some creative temporary jumpering.

(post #114611, reply #9 of 29)

I think you have confirmed my side of the arguement we are having around the coffee table, but let me be sure I've got this straight. If I have a generator that has a  floating neutral, the neutral would not be bonded to the generator frame, so grounding the frame would not cause me to have neutral tied to ground in two places, it would still only be tied together at the main panel where the neutral would pick up a ground reference, and this is good. Having neutral bonded to the frame of the generator would be the same as having them tied together at a subpanel, and I know this is bad, because return current will be placed on the grounding conductors. Is that correct?


As a side note, I find it hard to believe that such an important technicality is so poorly understood by those that should know, such as generator salesman and electricians.

(post #114611, reply #17 of 29)

Your question did seem to stir things up! As mentioned in other posts, voltages are measured between two points.  Technically, a "floating" conductor is one that is not referenced to whatever the chosen point of reference is for a system.  In the normal AC power systems, the zero reference is the earth ground.  The grounded conductor (NEC designation) is commonly referred to as the neutral because it's voltage is zero, or close to, with respect to the ground reference. 


I can't answer your basic question because I don't know if the neutral at the generator is bonded to the generator frame.  If, you have the manual on it, see if it mentions that.  But what the generator manufacturer says is floating may or may not be what I described above.  You can check it for sure with an ohmmeter - with the generator off and all loads disconnected, measure the resistance between the neutral pin on the outlet and the frame; if it is close to zero then they are connected together there.


If the generator neutral is not bonded to the frame, then you are fine.  The frame will be referenced to ground through the ground wire to the panel, and the generator neutral will be referenced to ground at the panel.  If the frame and the neutral are tied together at the generator, then you will get current flowing through the ground wire because it and the neutral will be connected in parallel.  This will cause a voltage drop in the ground wire, so the generator frame will be at this voltage with respect to ground, assuming that the metal of the frame is not touching the ground.