First of all I'm pretty sure I know the theory behind the effect. I know about the inductive and capacitive coupling, about 3-phase balance problems and about ground bounce, so this question isn't about normality of existence of voltage between ground and neutral line.

What I want to know is in general are there established levels at which the voltage is normal, at which I should start investigating the voltage and at which I should kill power and start panicking?


Due to recent answers, I added the following section to the question:

First, I AM AWARE OF OHM'S LAW! I hoped that it would be obvious from the second sentence of the question, but it isn't.

Second, the type of answer I'm looking for is something like: "The difference in a properly installed should be below XX V because it is written so in regulation YY". Bonus points for information on how number XX has been derived.

Alternatively acceptable answer would be: "There's no limit in any regulatory documents because of ZZZ." if it's accompanied by a good description of ZZZ.

I will down-vote answers which basically say: Ohm's law.

I know that this sounds extremely rude, but unfortunately I don't see any other way to explain exactly what I'm asking. If you don't know the answer, that's fine by me.

  • 3
    well if it's above what the hot should be carrying you should be panicking Dec 25, 2011 at 23:27
  • @ratchet freak Well I do agree on that. :) It's just that some sources for example say that the voltage should be below 0.5 V (source claimed that exposing cows to Vgn higher than 0.5 V can impact milk production), others say that below 2 V is fine, some say that some computers may work up to 30 V and so on and so on.
    – AndrejaKo
    Dec 26, 2011 at 0:56
  • Where are you measuring the voltage?
    – Tester101
    Dec 28, 2011 at 16:45
  • @Tester101 At the plug, but I'm looking for general answer.
    – AndrejaKo
    Dec 28, 2011 at 19:04
  • 2
    "Neutral to ground voltages at branch receptacles are directly proportional to circuit length and circuit current and inversely proportional to conductor size or cross-sectional area. In other words, making the circuit longer or increasing the circuit current will increase the neutral to ground voltage. Increasing the conductor gauge will reduce the neutral to ground voltage that occurs for any given length circuit at any given load." Source
    – Tester101
    Dec 28, 2011 at 20:31

6 Answers 6


The best I could do was to dig up a fine print note (FPN) in the National Electrical Code (NEC). In theory, there is no limit (as stated in other answers). In practice, the limit is when things stop working. In this FPN, the limit is 3-5%. In reality, the limit is around 1-2%.

NEC 2008

210.19 Conductors — Minimum Ampacity and Size.
(A) Branch Circuits Not More Than 600 Volts.
(1) General.
FPN No. 4: Conductors for branch circuits as defined in Article 100, sized to prevent a voltage drop exceeding 3 percent at the farthest outlet of power, heating, and lighting loads, or combinations of such loads, and where the maximum total voltage drop on both feeders and branch circuits to the farthest outlet does not exceed 5 percent, provide reasonable efficiency of operation. See FPN No. 2 of 215.2(A)(3) for voltage drop on feeder conductors.

This Fine Print Note (FPN) says that to provide "reasonable efficiency of operation", the voltage drop of the branch circuit to the furthest point should not exceed 3%. And that the total voltage drop, including the feeders should not exceed 5%.

120V * 3% = 3.6V
120V * 5% = 6V

Given this information. You should start to examine your wiring, if you measure anything near 3V neutral to ground.

Note: Fine print notes are informational only and are not enforceable as requirements of the National Electrical Code.

  • This is much closer to what I'm looking. Thanks!
    – AndrejaKo
    Jan 10, 2013 at 16:09
  • Could "reasonable efficiency of operation" simply be referring to power loss in the circuit resistance? As opposed to the tolerance of the load to the voltage between neutral and ground. Jan 11, 2013 at 8:37
  • @PhilipNgai The voltage between neutral and ground will be equal to the voltage drop across the wire.
    – Tester101
    Jan 11, 2013 at 11:59
  • Since the ground and neutral are tied together at the circuit breaker panel, causing 0 volt ground->neutral there, the metric you want is the 3%. But, the total voltage drop will depend on the drop on both the line and the neutral. So, for the neutral to ground measurement, you'll want half of the 3%. This drop should be much less with 240V loads since there isn't any current on the neutral line if the load is properly balanced.
    – Pigrew
    Jan 11, 2013 at 12:51
  • This answer makes even more sense in the context of your other answer, which provides a bit of background on why the difference arises. Maybe they could be combined?
    – mac
    Jan 11, 2013 at 17:02


Proper Installation.

In a proper neutral-to-ground installation, the voltage between the neutral conductor and any metal part of the electrical system will be equal to the voltage drop of the neutral conductor in accordance with the following:

  1. At service equipment, the voltage difference between the neutral conductor and the service equipment case will be 0 volts.

  2. At panelboards, the voltage difference between the neutral conductor and the equipment grounding conductor (panelboard case) will be equal to the voltage drop of the feeder neutral conductor.

  3. At branch circuits, the voltage difference between the neutral conductor and the equipment ground (ground contacts of the receptacle) will be equal to the voltage drop of the feeder and branch circuit neutral conductors.



NEC recommends (but does not require) that there is no more than 3% drop on a branch circuit. On a 120V circuit, that would be 3.6V, which would be split across the hot and the neutral, so you wouldn't want to see more than 1.8V on the neutral.

  • Please read the question before answering.
    – AndrejaKo
    Jan 10, 2013 at 12:58

It depend on load to load senstivity. It depent on voltage difference between Vnw.r.tg and Vlw.r.tg at any perticular time t. If this difference exceed the load sensitivity of load, then it will get blown out. As it is AC voltage Vnwrtg can be in synch or out of synch therefore can be addition or subtraction at time t and can vary. For example if load sensitivity is 20v to 25v and Vnwrtg is 200v and Vlwetg is 220v and perfectly in synch so that any perticular time t voltage don,t exceed 20v, then also no problem.

  • 1
    Hello, and welcome to Stack Exchange. There's interesting stuff here, but it's hard to understand due to the formatting and unexplained acronyms. If this is important to you, you probably should edit your answer to make it clearer. Jul 26, 2016 at 22:49
  • "Voltage Neutral Wire Referenced to Ground" and "Voltage Live Wire Referenced to Ground", maybe?
    – Tester101
    Jul 27, 2016 at 11:33

As Corporate Trainer for a Large HVAC manufacturer I can say that our equipment cannot reliably prove flame if there is more than a 2.0 volt difference between the neutral and ground

  • 1
    The question has nothing whatsoever to do with HVAC equipment, at least, according to my reading...
    – FreeMan
    Feb 13 at 22:53
  • 1
    @FreeMan -- it is a useful datapoint WRT "acceptable" at least Feb 14 at 2:19

Normally , it is zero.
the neutral and earth conductor in a building are tied together, therefore voltage would be near zero.

  • But I'm asking for exact values.
    – AndrejaKo
    Jan 8, 2013 at 9:35

Not the answer you're looking for? Browse other questions tagged or ask your own question.