120V, 49V, 51V on outlet test and hot ground

[WSP]

How is the protective ground (grounding conductor) connected to the outlets from the breaker box?

Maybe one of the earlier posts said it, but I did not see it in a quick look back.

It's metal-clad (BX I guess) cable. I just added a ground wire from each outlet and switch to a screw on the inside of its box to get solid to the ground. All outlets and switches were, I supposed, ground by just the body of the receptacle/switch contacting the box edge—I'm not sure they actually all were making solid conductive contact...

Copper water pipe may or may not be a great ground, there could be a section of plastic pipe, a connector, maybe a water filter or water softener or who knows what in line before it can make it outside and be a decent potential grounding point. The only grounding point that is important is the one at the power panel. Now, you often bond things like the metal pipes to that ground, but generally, you should not rely on them as your grounding point. Even the soil may not be a very good grounding point, depending on its composition, the moisture content, and probably other factors. For example, dry sand makes a lousy ground point. Way back when I was in the military, on field exercises in the desert, they would requisition large quantities of table salt, mix that into the soil where they needed a good ground, and keep it wet. This improved both the electrical safety and the performance of things like antennae, since it then provided a decent ground plane as well. It can sometimes take numerous ground rods to get a decent ground in parallel to get the resistance down. Pull the breakers out of the panel and look at the bus bar and their contacts. If they come out easily, over time, the heating/cooling from changes in load may have made their spring contact pressure too low to maintain a good connection. There could be some arcing or corrosion there as well.

Have you checked the connections in the main panel for tightness, arcing, or corrosion? Sometimes, just a loose connection can be a source of all sorts of weird indications.

Interesting about that desert stuff. I have been in the box—I put an AFCI-GFCI breaker on this circuit—and all grounds and neutrals are quite tight. They were loose a few years ago and we got some heat/melting on the main neutral coming to the bar, so I keep an eye on that.

I am opening up this 90 year old ball of wires in the basement where many of these outlets join together under the floor tonight. Going to wire it all solid in a new junction box and I am hoping the weirdness goes away after that. If not, will probably call in a pro, as the cable that comes out of the breaker box disappears into the wall and goes who knows where, before it makes it's way across the house to the area it serves. I have a guess as to where it's coming in, but really have no idea.

 

[Bgard]

it sounds like you have the devices bonded to the boxes but there is no way to know if the bx connectors are tight in each box between the problem area and the main panel or if the bx is actually in connectors the entire way. it is the steel armored cable that would create the ground path back to the main box.it is not very good at doing this for the obvious reason of the problem you are having now. it relies on every connector to be tight and the set screws to be tight both can work themselves loose with heating and cooling over the past 90 years. the new armored cable has a bonding wire in it to prevent this from happening.

 

[Sponsor]

 

[WSP]

it sounds like you have the devices bonded to the boxes but there is no way to know if the bx connectors are tight in each box between the problem area and the main panel or if the bx is actually in connectors the entire way. it is the steel armored cable that would create the ground path back to the main box.it is not very good at doing this for the obvious reason of the problem you are having now. it relies on every connector to be tight and the set screws to be tight both can work themselves loose with heating and cooling over the past 90 years. the new armored cable has a bonding wire in it to prevent this from happening.

True. I tightened down every clamp in each box as I added the ground wires, but this makes me think of something: I bet there are untight cables in this junction I am about to tear into. It's not inside a proper box, it's inside the 100 year old light fixture base. Will find out soon.

 

[WSP]

And here's something interesting. Everything is suddenly back to normal. All readings are right, and the ground is showing no current except the 1V also seen from neutral to ground. I'm going to bang around on wires and boxes and see if the bad readings come back...

 

[WSP]

So update is that it appears that all ground paths are now solid in the circuit, except that there is a break in the ground path from the first outlet on the circuit back to the breaker box. When I read the resistance of the BX cladding coming in on that cable coming from the box, across to the neutral going back, I get infinite resistance, so clearly there is a break in either that neutral going back or the ground path going back. Further, if I clip a wire directly to metal cladding in this circuit and then run the other end to the breaker's ground cable, and then go back to an outlet and read neutral to ground, I get 1-2 ohms or so. This leads me to believe it is indeed the ground, not neutral that is discontinuous back to the breaker, and this discontinuity must be somewhere back in the walls between where the power hits my first outlet and where it comes out of the breaker box.

With that in mind, I am thinking of just running an 8 or 10 gauge green wire from a box on the circuit directly back to the ground bar in the breaker box. This would ensure a ground on the whole circuit, since as mentioned above I tested the idea. The other alternative would be to discontinue use of the old BX cable going from the breaker into who-knows-what in the wall, and then into my first outlet, and instead pull a brand new metal-clad cable to that first outlet from the basement below with a proper ground/green in it. This latter option would not be a small effort and cost, though, compared to just running a direct ground wire to the box. It does seem like the cleaner/better option—except I'd be leaving a dead old BX cable in the walls for a future owner to puzzle over should they come across it and not find the clipped off in near the breaker...

Any thoughts on the two options?

 

[WSP]

And here's something interesting. Everything is suddenly back to normal. All readings are right, and the ground is showing no current except the 1V also seen from neutral to ground. I'm going to bang around on wires and boxes and see if the bad readings come back...

One last interesting tidbit. The normal readings I was suddenly seeing were somehow happening because a laptop charger was plugged into the circuit. When I UNPLUGGED it I got wacko readings (120/50/50), and when I plugged it back in everything read fine again. No idea, what was happening there, but soon after I thought to check the ground resistance and see what I got on that outlet... Was strange though. Maybe the charger somehow looked like a ground to my crappy multi-meter...

 

[Jadnashua]

The more correct method to add the ground would be to run a new cable with a proper ground verses a separate ground lead, which, at least as I understand it, does not meet code.

Some switching power supplies essentially have a transformer on their input. That can throw off some readings to a high impedance DMM.

 

[WorthFlorida]

If you really want to locate where the wires go inside walls, this tracer tool is pretty good. I've been able to pick up the tone upto about 4" behind walls. One major thing is the power must be turned off. It was designed for low voltage telephone wire tracing over well over forty years ago. You clip on the tone generator on two wires and one side can be on a ground wire. Turn on the probe and set the volume to MAX. If any of the wires you're tracing goes to wall switches and outlets you'll easily pick it up at those locations. It is under $100 and I have this model. Not sure if lower cost unit work as well. This model is identical to ones I used forty years ago and I think it is the best there is.

http://www.homedepot.com/p/Standard-701K-G-Tone-and-Probe-Kit-701K-G/203495654

 

[WSP]

The more correct method to add the ground would be to run a new cable with a proper ground verses a separate ground lead, which, at least as I understand it, does not meet code.

Some switching power supplies essentially have a transformer on their input. That can throw off some readings to a high impedance DMM.

So, I opted to just get a good green wire ground straight to the house ground on this circuit via the last junction box that I fixed up, and my main circuit of concern is all set. I may pull new cable to it later and ground it the right way in romex, but for now everything reads perfectly safe.

I did find out that the other circuit that I was concerned with earlier in the thread was not properly grounded and it shared a switchbox with this circuit, so that one light and box still read "hot", but I pulled a newer grounded Romex from another circuit running right by the old, and just pigtailed a ground over to a junction box wall on the ungrounded and it's all set now too.

If you really want to locate where the wires go inside walls, this tracer tool is pretty good. I've been able to pick up the tone upto about 4" behind walls. One major thing is the power must be turned off. It was designed for low voltage telephone wire tracing over well over forty years ago. You clip on the tone generator on two wires and one side can be on a ground wire. Turn on the probe and set the volume to MAX. If any of the wires you're tracing goes to wall switches and outlets you'll easily pick it up at those locations. It is under $100 and I have this model. Not sure if lower cost unit work as well. This model is identical to ones I used forty years ago and I think it is the best there is.

http://www.homedepot.com/p/Standard-701K-G-Tone-and-Probe-Kit-701K-G/203495654

Thanks good to know this thing would let me tracer the wires. I didn't know such a thing existed. I am going to ask around and see if anyone that I know owns one and would let me borrow it. I heard about this thing, which is pretty cool: https://walabot.com/diy It's way cheaper than it was before, too. May think about buying one, but thought it was a longshot that anyone I knew had one. Either $100 would probably be worth it for either tool.

So for the first time ever, I assume, this old house is fully ground, and presumably safer. Pretty happy about that. Thanks for all the talk, guys!

 

[Jadnashua]

One thing to keep in mind...the electrical ground is NOT designed to normally carry current...it is there to provide an alternate path back to the power panel to trip the breaker or blow the fuse. If there is power on the ground, you will not be able to install a GFCI unit and expect it to not trip immediately. So, what you did may have masked the problem, it did not fix it. For a GFCI to stay on, ALL power that goes out the hot side, MUST return on the neutral (not the ground!), or, if that leakage amount exceeds 5ma, it will trip or disconnect.

 

[WSP]

One thing to keep in mind...the electrical ground is NOT designed to normally carry current...it is there to provide an alternate path back to the power panel to trip the breaker or blow the fuse. If there is power on the ground, you will not be able to install a GFCI unit and expect it to not trip immediately. So, what you did may have masked the problem, it did not fix it. For a GFCI to stay on, ALL power that goes out the hot side, MUST return on the neutral (not the ground!), or, if that leakage amount exceeds 5ma, it will trip or disconnect.

OK, so it sounds like if I put a GFCI on the circuit and it didn't immediately trip, I'd know I was in the clear. I'm pretty sure this will be the case, I believe that if I had a better meter, I'd not be reading the funny voltages. But I assume that if I just read the current on the ground wire that I installed to ground the circuit from that junction box, that if any current showed up I'd know I DID still have a problem, right? You're saying that if I am just bleeding off the rogue current down this new ground wire, there still could be dangerous current in the ground on the circuit somewhere? Wouldn't I get some kind of voltage read on an outlet, if this was true?

 

[Jadnashua]

Modern DMMs have a very high input impedance. This means that it is easy to develop a voltage since they are not a load on the circuit. This is good when measuring digital electronics, since they generally run at very small currents and a load would skew the readings and interpretation of what's going on...it's bad when you're dealing with power circuits, since you can get phantom voltages that have no current capacity. A typical DMM may have 100Mohm/volt input impedance...IOW, no load. An old, analog meter with a pointer may have a 20Kohm/volt resistance...while not a huge load, still 1000's of times more, and those phantom voltages tend to disappear.

If you were to remove the ground wire and run it through the meter when configured to measure current, you should not measure any. A GFCI will trip if the delta between hot and neutral is >5ma or so...it that's leaking into the ground system, the GFCI won't reset (i.e., won't turn on or stay on). Well, that power could be leaking through you, and not directly into the safety ground connection, but that's a different story. In a properly operating system, the ground wire should never have current on it, and all power that goes out the L1 or L2 lead, should return either on the opposite L-lead, or neutral (depending on if it is a 240vac or 120vac system) - nothing on the ground lead.