I hope some of you sparkies can tell me what’s wrong.
40-year old SoCal tract house. Copper wiring, NM 12-2 with ground, steel boxes, ground wires clipped to edges of boxes (apparently run outside the boxes to clip on the front edge), duplex non grounding outlets (except in kitchen, baths and laundry room of course).
HO wants to replace outlets with grounding type.
Voltage tested on one of the outlets between hot and grounded slot = 115V; between hot and grounding clip on box = 95V. Tested box is neither first nor last on circuit; don’t know how many outlets there are upstream.
What is the likely cause of the low potential?
What problems might there be if I were to install grounded outlets where ground voltage is just 82% of normal?
Replies
The 95V is "phantom" voltage, meaning the ground isn't really connected. If you use regular outlets you'd be violating code and common sense safety rules installing 3-prong outlets here.
You can use GFCI outlets (or one GFCI at the head of the run, with the others "daisy chained" through it), but then you're supposed to apply a label to each outlet saying "ungrounded", and the outlets are no good for surge protectors (though perfectly safe for humans).
I am trying to get a grip on your current installation.
>>Copper wiring, NM 12-2 with ground, steel boxes <<
OK - cable containing black, white and bare conductors all contained in a non-metallic sheath. Steel boxes.
>> ground wires clipped to edges of boxes (apparently run outside the boxes to clip on the front edge),<<
Very confusing - the original installer brought 12-2 with ground inside of the steel box, stripped the outer sheath back and then threaded the the bare grounding conductor outside of the box and ground clipped it to the front edge of the box??????? This makes no sense. I find it hard to believe that someone would do this --- way too much work.
>> duplex non grounding outlets <<
OK
>>(except in kitchen, baths and laundry room of course).<<
Meaning that these outlets are already 3 prong grounding type? Or that when the existing 2 prong outlets are replaced, they will be replaced with GFCI type?
>> Voltage tested on one of the outlets between hot and grounded slot = 115V;<< meaning between the hot and neutral slot of an existing 2 prong duplex? Or between hot slot and the ground slot on an existing 3 prong duplex - see above question.
>> between hot and grounding clip on box = 95V.<<
Understood.
Jim
Never underestimate the value of a sharp pencil or good light.
>> ground wires clipped to edges of boxes (apparently run outside the boxes to clip on the front edge),<<I've seen this before. The ground wire is doubled-back out the hole for the cable, then around the outside of the box and held in place with one of those metal edge clips. Kinda made sense when the ground wire had no useful application inside the box. Probably not the ideal way to do it, but not in itself a problem.The net of course is that the ground wire is not solidly connected somewhere, and therefore it cannot (in good conscience) be used to ground 3-prong outlets.
If your view never changes you're following the wrong leader
OK, that's a new one on me.
I've seen most of the stuff Mike Hennessey mentions in his .6 post, but have never encountered a ground wire clipped to the outside of the box. I wonder if that was a regional practice which never caught on here.
Seen plenty of ground wires which enter the box and are then clipped to the box edge, but the conductor was inside of the box. From pre-threaded ground screw hole days. Also encounter a fair amount of cloth wrapped romex with a down-sized grounding conductor -- 12-12-14 and 14-14-16.
Jim
Never underestimate the value of a sharp pencil or good light.
Current code requires that ground wire be brought inside the box and connections made there.
But I have heard that in the early days of grounded circuit that was not always done. In some places it was just wrapped around the screws on the romex connectors.
And I don't think that all of the box clips are UL listed.
95 volts is kind of high for "phantom" voltage. My guess is that there is a (very) bad connection at one of the places with a lot of oxide under the clip.
Or there is a run in he cable where there is a nail in it.
If you want to do the work I you could pop the clip and if the wire terminate inside and then loops out use some needle nose plier to pull it in.
In some places it might require a cutting out the old box, pulling it out to get access to the wires and then putting in an old work box.
However, the wires will probably be short and a pain to work with.
.
A-holes. Hey every group has to have one. And I have been elected to be the one. I should make that my tagline.
I mostly work on old stuff, and I've seen lots of ground wires wrapped around the screws on the romex connector. (Back when ground wires in Romex first came into wide use, I don't think boxes even had tapped holes for ground screws. I've also seen it wrapped around outlet mounting screws, crimped to the side of the box without a clip, and many times just cut off at the Romex connector.) In the cases were the box was grounded from the outside, the ground wire was rarely routed to the inside and then back out -- it looked like the installer used the ground wire like a zipper, pulling it back to open up the sheath, then putting only the black and white into the box.
Assuming the romex connector screws were tightened at all, I'd be surprised if the OP could get the ground wire to pull back into the box without breaking it at a bend or trashing the rest of the cable at the connector, leaving a worse problem than he started with.
I suspect new boxes will be the only solution. Probably easier in the long run anyway.
Mike HennessyPittsburgh, PA
All of the 50's to early 60's era houses I have seen in California had a separate ground run from the nearest cold water pipe to the outside of the box. The ones in my own house had a split bolt type connecter that had been thru bolted in the box.
In my own 1950's house, only the bath, kitchen sink, and laundry had a ground wire to the box. The 3 prong plugs were added to these outlets in the late 60's.
Could it be that maybe your ground wire might be corroded where it connects to a water line?
"All of the 50's to early 60's era houses I have seen in California had a separate ground run from the nearest cold water pipe to the outside of the box. The ones in my own house had a split bolt type connecter that had been thru bolted in the box."Note that the code still allows a wiring to be upgraded by fishing a separate ground wire.HOWEVER, CONNECTING TO THE NEAREST COLD WATER PIPE IS NOW FORBODEN. But it did follow code at the time. The prolbem is that pipes often get replaced with plastic.There are several problems with it and when you run into it you need to evaluate it and see about fixing it.The wire can connect to the ground bus in panel, to any point on the ground electrode system, or to another box with a ground wire that connects back to one of the first 2 places.
.
.
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 11/3/2007 12:52 pm by BillHartmann
HOWEVER, CONNECTING TO THE NEAREST COLD WATER PIPE IS <NOT> FORBODEN. But it did follow code at the time. The prolbem is that pipes often get replaced with plastic.Did you mean <now> forbidden but not at the time?The new added ground can be bare?
Yes, it should be NOW FORBIDEN.Thanks I corrected it."The new added ground can be bare?"AFAIK, yes. With out cross reference everyplace in the code where "ground" is used the section tht allows the use of a separate wire uses "equipment grounding conductor" and bare or insulated, solid or stranded, copper conductor is one of many types of equipment grounding conductors.
.
.
A-holes. Hey every group has to have one. And I have been elected to be the one. I should make that my tagline.
Thx.
When the ground wire was first introduced, there was a brief period when ground wires were run separately from the circuit conductors. Even when the grounds entered the boxes, they were often folded over the lip, and spliced outside the box. This method was also used to add a ground wire to existing 2-wire circuits. If you look closely, you're almost certain to find that the ground wire is quite a bit smaller (#16) than the circuit conductors.
95v is pretty high for a "phantom" voltage. It is more likely that their is a loose splice, poor bond at the panel, or some other high resistance connection.
The whole point of the ground wire is to clear faults ... that is, make the breaker trip ASAP. A high resistance path pretty much defeats the whole point.
The best solution is to completely rewire the entire circuit with modern wiring methods.
Perhaps you can find, and fix, the bad connections. I doubt it, though.
That leaves you with treating the system as if it were 'without a ground wire,' and GFI protect the circuit. With an old house - and assuming your panels are of a current make - the easiest way to do this is to get GFI-type breakers. Shared neutrals, and inter-mixed circuits (a major problem when there is DIY work done) can make that impractical. If you have problems installing GFI-type breakers, you probably have other problems present, and really ought to get a Sparky.
Edited 11/3/2007 10:37 am ET by renosteinke
95V isn't at all high for a phantom voltage, measured from hot to ground. Keep in mind that especially if the ground is run separately, it's much more strongly capacitatively coupled to ground than to the hot wire.
If your view never changes you're following the wrong leader
Please explain what you mean by "much more strongly capacitatively coupled to ground than to the hot wire."
I mean there's more capacitance to "ground" (which includes most of the surrounding structure) than to the hot wire. The apparent voltage of a disconnected wire, when measured with a high-impedance voltmeter, will be determined by the ratio of the capacitances to ground and hot.
If your view never changes you're following the wrong leader
Thanks for your insight.You wrote, "...you're almost certain to find that the ground wire is quite a bit smaller (#16) than the circuit conductors."
Right you are, they're 16ga; part of the original NM cable.You wrote, "It is more likely that their is a loose splice, poor bond at the panel, or some other high resistance connection."
Poor bond at the panel - would that be found by testing voltage between hot post on the circuit breaker and the ground wire for that circuit?
loose splice - would I have to test voltage between hot and ground clip at each box working upstream until I find 115V?Presumably I could fix the splices by pulling the grounding clips off and twisting/crimping the ground wires together? Re-wiring is problematic due to the fact that one wall outlet in the middle of the circuit is in the center of a 10-foot-wide floor-to-ceiling mirror and the other side of the wall is stucco exterior.If simple fixes don't work, I think a referral to a sparky will be the next step.
BruceT
Have you tried to trace the problem down to a specific connection. Find the earliest point in the chain where there's a problem and see if tightening the connections there (or at the box prior to that) does the trick.
If your view never changes you're following the wrong leader
(You can even temporarily ground one of the problem boxes and see if the others become grounded, to see how wide-spread the problem is.)
If your view never changes you're following the wrong leader
So, how does one further identify a problem like yours? Well, first, let's eliminate that 'ghost voltage' issue. Ever have a car battery read 12v, yet fail to start the car? Notice how the headlights dim almost instantly? The 'true' voltage is read when there is a load applied.
Same with a 'ghost' voltage. Rather than use an electronic meter - which can be fooled by a lot of things - measure using a "Wiggy," a meter that uses a solenoid to move a bar across a scale. The load of the solenoid requires some power .... not much, but something .... so it will not read 'ghost' voltages. Still, a "O" voltage reading does not mean that the problem has gone away. There is still the question of having a good ground path. Let me illustrate the problem with an example:
We 'know' that the ground rod has no role in clearing faults (that is, making a breaker trip) because the resistance is too high; there is no way for that rod to carry more than a few amps. Our breakers can work, because the ground wire is connected to the neutral at the main panel. Installed properly, the ground wire provides a path of essentially no resistance back to the PoCo. No resistance = lots of amps flow = tripped breaker.
You need to find where there is a lot of resistance in the ground wire part of the circuit. While this can require some special equipment and training to find the problem, there are some things you can do yourself. The first is to realize that the problem can be at any point in the circuit; it may even be the accumulation of several small problems. You can check EVERY connection back to the panel. Is the screw on the buss tight? Are the wires securely twisted / crimped / wire nutted together in each box? Is there a little wire that directly attaches the receptacle to the ground wires from the Romex? It's not unusual for one of these wires to be 'almost broken' or just resting within a wire nut. That's a problem to fix. If you're taking the voltage measurement at the receptacle, using the same slots that plugs use .... maybe your meter probes are not making god contact. Or, the only 'grounding' connection it through the screw that's holding it to the box. Not a good way to do it. You need to open the box, and you need that little pigtail. It's also not impossible that the fault is within the receptacle. Beyond this, we're well out of the training of homeowners, and -sadly- far too many electricians.
I really appreciate your thorough replies to peoples' questions.You have expanded the possibilities for my low-voltage-to-ground problem beyond the specifics of my situation to also cover 3-prong outlets with ground screws on them and grounding wires that are actually inside the box - helpful for other situations as well. Thanks."Rather than use an electronic meter - which can be fooled by a lot of things - measure using a "Wiggy..."
Apparently analog is sometimes better than digital."...the problem can be at any point in the circuit; it may even be the accumulation of several small problems."
16ga copper wires pinched against galvanized boxes by steel clips in a daisy-chained circuit; add 40 years of oxidation. What could possibly go wrong?Looks like I can improve the ground connections if I pry off the clips and twist/crimp the ground wires of each leg of NM together and make sure that all is tight in the panel. If that fails, I'll call in a pro.BruceT
I lived and worked in Northern California for a while. Sometimes during demo I'd see #16 bare wire run around near the old NM and attached to the outside of the boxes, usually by wrapping around one of the screws on top that holds on the side of the box.
Boxes were often finished black. Anyone know why they used to make'm that way?
Ed
> Boxes were often finished black. Anyone know why they used to make'm that way?Actually, it was the soot from the fires.(Black wouldn't show as much if the box wasn't entirely covered by cover plate, etc.)
If your view never changes you're following the wrong leader
"Boxes were often finished black. Anyone know why they used to make'm that way?"Coming from an industrial paint mfr background, I would suggest that boxes in the early days would have been painted to prevent rusting because painting was cheaper than galvanizing and they were probably painted black because black is the cheapest paint to make.BruceT