Spark from ground to neutral???

[benaround]

Other than possibly higher neutral current how would non-linera loads contribute to this?

Why would you need more than one possiblity ?

 

[KP2]

OK sorry for not getting on yesterday.

The voltage to ground is 2.46 volts at 60 hz.

The main bonding jumper looks good and the GEC to the water main has 3.6 amps on it.

The whole service had 40 amps on the neutral at this time.

P.S. can anyone tell me how to put pics on the post and not as an attachment.

Thanks
Kevin

 

[LarryFine]

The voltage to ground is 2.46 volts at 60 hz.

The main bonding jumper looks good and the GEC to the water main has 3.6 amps on it.

The whole service had 40 amps on the neutral at this time.

All typical real-world numbers. The test would be to see if the N-G voltage goes away with the circuits unloaded.

 

[KP2]

All typical real-world numbers. The test would be to see if the N-G voltage goes away with the circuits unloaded.

The voltage is still thier with the breaker off feeding that circuit. Also someone asked about ungrounded to EGC voltage and it is the same 2.46 volts
Also ungrounded to grounded is .2mV.
All these last readings are with the ciruit off.

 

[KP2]

I just need to add that the SPARK I,m talking about is a little spark.

 

[LarryFine]

The voltage is still thier with the breaker off feeding that circuit.

Make sure you've deenergized all circuits sharing that neutral. If the voltage is still there, then it must be present at the source panel. Check there next.

There should be some point where the voltage zeroes out. Definitely at a proper MBJ point, and maybe at sub-panel(s) farther downstream.

 

[LarryFine]

I just need to add that the SPARK I,m talking about is a little spark.

That means it's a relatively small voltage.

 

[steve66]

During a project in our lab we saw a tiny spark between the ground and neutral conductors.

Any thoughts?

Does this mean you touched the neutral to the ground? If so, I would expect a spark. If you have 40 amps flowing back to the panel on the neutral, and you touch that neutral to a ground, a lot of that 40 amps is going to start flowing back to the panel over the ground wire, and any other ground paths available.

Steve

 

[KP2]

Does this mean you touched the neutral to the ground? If so, I would expect a spark. If you have 40 amps flowing back to the panel on the neutral, and you touch that neutral to a ground, a lot of that 40 amps is going to start flowing back to the panel over the ground wire, and any other ground paths available.

Steve

No, this is happening on a branch circuit that we use for an old work lighing lab assignment. The wires we are testing in the first post are the two that spark.

 

[KP2]

Make sure you've deenergized all circuits sharing that neutral. If the voltage is still there, then it must be present at the source panel. Check there next.

There should be some point where the voltage zeroes out. Definitely at a proper MBJ point, and maybe at sub-panel(s) farther downstream.

I will try this, but not till Friday.


Thanks
Kevin

 

[weressl]

During a project in our lab we saw a tiny spark between the ground and neutral conductors. This building has a 3 phase 208 / 120 service and at least 70% of the load today is flourescent, (no A/C). When the students asked I felt it had to do with the non-linear loads in the building. We measered it to be 5 mA.

Any thoughts?

Ground potential differences and other causes of circulating ground current can often manifest in such discharges. They are POA's, but IMO the Code itself is insufficient with its regulations to completely eliminate it.

 

[mivey]

P.S. can anyone tell me how to put pics on the post and not as an attachment.

Go to photobucket. Open an account (it's free). Upload your pictures to photobucket. Use the link to the photobucket pictures in your posts.

 

[don_resqcapt19]

The voltage is still thier with the breaker off feeding that circuit. Also someone asked about ungrounded to EGC voltage and it is the same 2.46 volts
Also ungrounded to grounded is .2mV.
All these last readings are with the ciruit off.

Is this circuit fed from the service panel? Is the neutral part of a mutliwire circuit?

 

[don_resqcapt19]

I just need to add that the SPARK I,m talking about is a little spark.

You have created a crude switch. Any time you "switch" a circuit that is carrying current, there will be a spark. When the grounded conductor is touched to the grounding conductor you have created a parallel path for the current that was flowing on the grounded conductor. The size of the spark that is created when you switch a circuit is related to the amount of current being switched and the design of the switch.

 

[KP2]

Sorry I haven't gotten back on this sooner.

Make sure you've deenergized all circuits sharing that neutral. If the voltage is still there, then it must be present at the source panel. Check there next.

There should be some point where the voltage zeroes out. Definitely at a proper MBJ point, and maybe at sub-panel(s) farther downstream.

I have not been tried this yet. I'm having second thoughts about turning off a 400 amp main breaker to kill all power to see if I still get the same readings.

Is this circuit fed from the service panel? Is the neutral part of a mutliwire circuit?

This circuit is feed from a sub panel off one of the main panels. the service is a 400 amp 208 / 120 volt with 2 200 amp panels.

The circuit feeding the lighting does not share a neutral with another circuit.

I was just hoping you guys can help me rule out my original thought of this being caused by harmonics due to non-linear loads.

Thanks

 

[gar]

100325-1310 EST

KP2:

I have quickly looked back thru this thread. It is not clear where measurements are made and of what.

My suggestion is go to the main panel where neutral, EGC, and the earth grounding system are connected together. Using a good DVM, like the Fluke 27, make various measurements relative this point that I will list below.

One additional item you need is a long wire, the EGC lead of an extension cord or cascaded cords is convenient and these cords do not get as tangled as does a wire off of a spool. In the use of this wire there will be negligible current flow. The wire is to provide a long test lead to your meter. Connect one side of the meter input to the wire and the other end of the wire to the neutral buss in the main panel. All of the following voltage measurements are relative to the neutral buss of the main panel.

I am assuming that the inherent 1 turn loop that exists from the meter tests leads is not picking up substantial errors from stray magnetic fields. There are ways to test for this noise but I won't discuss them now.

1. Measure the voltage of the of the neutral wire itself at the main. This means the meter probe touches the actual wire and not the lug. This voltage probably should not exceed 1 MV.

2. At the main panel the same measurement to the EGC wire, and to the grounding wire.

3. Pick an arbitrary point outdoors some distance from where the grounding electrode is located. Put a 12" screwdriver in the earth and measure its voltage.

4. Next go to the sub-panel supplying the circuit that you are questioning. Measure EGC and neutral voltages here.

5. Along a circuit path to where you experience the spark problem measure the voltages of the EGC and neutral.

Somewhere along these paths you should see something that will give you an idea of the source.

Do this with normal loads on. If these measurements indicate a clear cut problem at some point, then you might change loading and see what effect occurs.

Report back.

.

 

[hurk27]

Sorry I haven't gotten back on this sooner.


I have not been tried this yet. I'm having second thoughts about turning off a 400 amp main breaker to kill all power to see if I still get the same readings.


This circuit is feed from a sub panel off one of the main panels. the service is a 400 amp 208 / 120 volt with 2 200 amp panels.

The circuit feeding the lighting does not share a neutral with another circuit.

I was just hoping you guys can help me rule out my original thought of this being caused by harmonics due to non-linear loads.

Thanks

Since this circuit is fed from a sub-panel, any load on this sub-panel can add or take away from the neutral load, this is because the sub-panel is fed via a multi-wire circuit that shares the neutral, and also the fact that the grounding and neutral are kept separate at this sub-panel, this allows the voltage drop of the neutral between the sub-panel and the main service panel or disconnect (the location of the main bonding jumper) to have a difference of potential, because there is no load on the grounding it will not have a voltage drop, so you are reading just like you are connecting one lead to the neutral at the main panel and the other lead to the neutral at the end of the circuit.
The voltage you are reading is the voltage drop in the neutral.
when you touch the neutral to the EGC at the load end, you have just added the resistance of the grounding path to the neutral thus lowering the over all resistance, this is the spark you see, but this allows dangerous current to flow through the grounding and can cause a shock or fire hazard, this is why the NEC does not allow the grounding to carry current.

 

[OTT2]

The voltage you are reading is the voltage drop in the neutral. when you touch the neutral to the EGC at the load end, you have just added the resistance of the grounding path to the neutral thus lowering the over all resistance, this is the spark you see, but this allows dangerous current to flow through the grounding and can cause a shock or fire hazard, this is why the NEC does not allow the grounding to carry current.

By theory you have just created a parallel path for the neutral current to follow, current will follow any path provided. Its impedence is much greater than the neutral, limiting the current.

Just my theory though.

 

[gar]

100325-1916 EST

OTT2:

When I built my house, 1965, the Anaconda NM (Romex) had an EGC of about #15 with #12 for the load conductors. Not too much after this time period the same type NM had a #12 for the EGC.

If hot and EGC are both the same wire diameter, and I place a dead short between them, then the voltage of the shorted end relative to earth rises to about 1/2 the voltage of the hot relative to neutral at the main panel. So if the source at the main panel remained at 120 V, then 60 V exists at the shorting point relative to ground or the neutral point at the main panel. Smaller wire for the EGC and the voltage is higher.

Assuming hot, neutral, and EGC are all the same diameter and the same material, and you place a dead short between neutral and the EGC, then current injected at the shorting point will divide equally between the neutral and the EGC.

If there is a high resistance short, then less current flows in the EGC.

.

 

[weressl]

By theory you have just created a parallel path for the neutral current to follow, current will follow any path provided. Its impedence is much greater than the neutral, limiting the current.

Just my theory though.

The impedance would roughly be the same since we are talking about EGC that would run parallel with the Neutral.

Why do you think the EGC wiould have higher impedance than the neutral?