Electrical problem: 120V between GND's (FIXED!)

[0dbfs]

I've moved into a new house recently and just discovered some electrical issues as a result of letting the smoke out of some gear. Luckily it was just gear and not my wife, son, or myself that the smoke came out of.

I measure 120V between two different GND prongs on different outlets yet my "plug-checker" shows that each outlet is correct.
Additionally, each of these two outlets measure 120V between line&neutral, line&gnd, and 0V between gnd&neutral.
Here's some more measurements:

---- SNIP ----
L1, N1, G1 = Line, Neutral, GND circuit #1
L2, N2, G2 = Line, Neutral, GND circuit #2

L1 to G1 = 125V
L1 to N1 = 125V
G1 to N1 = 0.xxV

L2 to G2 = 123V
L2 to N2 = 123V
G2 to N2 = 0.xxV

N2 to N1= 123V!
G2 to G1 = 123V!
N2 to G1 = 123V!

L1 to L2 = 123V
L1 to N2 = 0.xxV
L1 to G2 = 0.xxV

L2 to N1 = 0.xxV
L2 to G1 = 0.xxV
---- SNIP ----

Basically, when I patch some audio I/O's together that are connected to the different electrical circuits I get 120V flowing between the GND's which would also be present between the two chassis!

Fortunately, this caused the smoke to come out of a nice expensive piece. That however can be replaced.

Any ideas on how to localize and correct this fault?
Is this a symptom of having different phases on different circuits? I wouldn't think so because the L & G's shouldn't be affected by the phase of the L. or should they?

We do have two services hooked up to the house due to a renovation and "adding" an apartment at some point in the past. So these circuits are probably connected up to different services.

Should I call the electric company? An electrician?

It's not our house so I can let the owner know but regardless of that this is a huge safety issue that probably affects different circuits in the house that I have not yet identified and I need to get it fixed asap. Plus, I can't have my gear or others gear smoking like this.

Help is much appreciated!

Best,
jonathan

 

[nielsk]

Possible a Line & a Neutral reversed somewhere, or a problem with the neutral / ground bond in the panel. Have you tried one of those little yellow plug in testers with the 3 lights? They work very well.
Other than ripping everything apart, the best way to sus it all out is to shut off the main disconnect (hope there is one) and continuity check every outlet to the buss bars in the panel. The circuits can be properly labeled at this time (seldom done...).
Either way, best left to the owners responsibility to hire a licensed electrician. They should be happy to pay for it as opposed to a liability law suit.

 

[0dbfs]

Yes. I used the (tan) three-LED-plug-checker and it shows both circuits as "all-good" (nothing reversed or floating).

-j

 

[JohnRoberts]

I recall an example of this, that killed a musician when he was jamming with another guitar player and grabbed the other guys guitar, effectively putting 120V through him with fatal result. I know about this because I was working for the maker of the guitar amps. We got sued, but the amps were safe and the fault was his house wiring, as UL testified in our defense. 

In his case he had one amp plugged into a mis-wired wall outlet and the other into an outlet on his stove that was properly grounded.

Apparently this can happen when some less than competent electrician tries to convert an older house that was wired with two circuit floating wiring to 3 circuit grounded outlets and gets sloppy and swaps hot for neutral. 

Note: in the above case the house was condemned and the residents forced to move out until the wiring was fixed.

JR 

 

[mjk]

It's been a while since I've been an electrician and I don't normally give out safety advice but your situation is scary.

You have at least two wiring faults; a reversed hot/neutral and a broken ground.

Your 3 light tester has a list of the faults it detects, and the fault you have is not detectable by measuring at a single outlet. Since AC has no absolute polarity, the tester can't tell that the N+G are reversed from the L. "locally" everything looks fine.

You haven't provided enough information to speculate further, but it's not necessarily the 2 services causing the problem. I wouldn't assume anything.

I can only recommend not using the AC in that house until the problem is corrected. You don't know at this point which or how many outlets have energized ground and neutral connections.

I wouldn't have my family exposed to that level of unknown risk!

Cheers,

Michael

 

[JohnRoberts]

+1...

If compelled to approach this DIY... You can meter your outlets with a long ass piece of wire connected to a cold water pipe (if metal), or connected back to ground at your breaker box to serve as your true ground reference, and a simple VOM to meter voltage between ground and all outlet pins. 

Yes, this is very dangerous, as all properly ground bonded 3 wire line cord products, with exposed metal could threaten users with full mains voltage. 

Any pattern you discern in the miswired outlets may suggest where the problem is, but it's apparent you do not have a real ground path between the outlets and your power drop. This could be really expensive to pull new wire.

First order of business is to identify and tape over any killer outlets, even if you move out.

JR 



 

 

[0dbfs]

wow.
Yikes!

Any info you guys can notate that will help to understand the fault here is greatly appreciated. I don't have an issue with DIY'ing almost anything and in this case, taking some careful measurements to understand the extent of the situation adds knowledge to the pool.

If it's a reversed and/or open conductor and I can verify the extent and location fairly easily (as well as understanding and verifying the implemented solution), it's a no brainer to just fix it and move along.

I tend to stay away from breaker boxes (especially the source side of the box). However I will watch and take notes when someone more experienced with those items goes to work.

Your replies are much appreciated.

Best,
jonathan

 

[0dbfs]

A carefully worded email to the landlord using terms like liability, AC-line-voltage potentially present on fixtures/pipes/exposed metal/etc has resulted in a returned phone call within minutes and an electrician to be dispatched within the hour.

Thanks guys!

I'll update when I know more.

Happy new year!

Best,
j

 

[JohnRoberts]

I just noticed where you said there are two services coming into the house. That sounds a little odd.

Step one is to measure (ac voltage) between the two service panel grounds with VOM.

Typically I think they bond to earth at the panel or near the power drop. In general it is possible for there to be some modest potential between different earth grounds but I would expect multiple panels in one building to be bonded to each other. Even if not bonded to each other, there shouldn't be full mains voltage, unless one or both are floating. If it's an older building, there should be a reliable earth ground where the cold water pipe comes into the building.

I am not an electrician or expert on the electrical code, sounds like you may need one.

Unless you were doing something really stupid, it seems your landlord may owe you restitution for your damaged gear, and for putting your family in harms way. 

JR

 

[0dbfs]

The only thing I was doing that I feel was "stupid" is messing with an RCA-to-Mini-Plug adapter to connect up a "line-out" from a hardware-codec-device to a control-room-monitoring unit. I think mini-plugs and RCA's should just be abolished. Further testing has revealed that the mini-plug was not the issue, rather it was the 120V on the sleeve/gnd conductor between the monitor-station and the IP-codec.

The two electrical services are because of the prior renovation and part of the house was rented as an apartment. Apparently the electrical service may be mixed-and-matched to various outlets through the house by now. Even if that is the case, I don't see their GND being 123V apart. That would be an electric company issue no (if their G/N are 120V apart)? I've got plenty of broken gear (ie; spare-parts & doners). Maybe I could send them bills?

I'm not seeing the possibility of a fault in my equipment that could cause these symptoms. At least not without tripping the breaker. I certainly hope not.

Best,
j

 

[0dbfs]

Photo is sort of dark but you can see the service entrance is just a split off the L,N,G connections from the residential step-down (up-the-pole). So each of our two meters/demarcs receives the same signal from the pole WRT to L/N/G. From there, the two distro-panels in the crawl-space have no cross-connects. At least at the panel. So it appears from what I can tell so far that L/N/G are bonded on the service-provider side of the service entrance, not in the house.

Maybe I can take a measurement between the distro-panel GND or Neutral bars fairly easily. There is exposed building copper in the crawl-space with at least earth wire bonded which is connected up to the coax shield where our comcast service demarc's outside the house.

Romex everywhere in the joists with quite a few junction boxes. I can also measure about 250V between HOT of a GFCI outlet in the kitchen and HOT of a non-GFCI outlet in the kitchen but the GND's/N's there measure near 0V.

Best,
j

 

[crazydoc]

I can also measure about 250V between HOT of a GFCI outlet in the kitchen and HOT of a non-GFCI outlet in the kitchen but the GND's/N's there measure near 0V.

This is normal - the two circuits come from different sides of the breaker box.

http://hyperphysics.phy-astr.gsu.edu/hbase/electric/hsehld.html

 

[JohnRoberts]

Ok, that looks like both panels are tied into the same low voltage drop (same actual pole transformer secondary).

That doesn't mean there can't be funny business with how the service panels are wired up.

The two insulated wires coming from the pole are opposite polarity 120v legs, so 240v across both legs. The non insulated wire, is the transformer center tap and power pole ground, both service panel neutral bus should attach to this pole ground.

Since you are pulling from the same transformer there should not be any strange issues between different outlets hots, there are only two possible legs total.

So once again check that the neutral bus at both service service panels have no significant AC voltage between them. You can probably measure between screw heads on the panels. They should get a decent ground from the pole, but confirm to a cold water pipe if metal plumbing is handy.

Then the issue is to determine how your outlets are are actually wired. The problem is most likely "not" at the panel, but check it first just to be sure.

JR

PS: NO, you didn't do anything wrong... it looks like your landlord did either directly or indirectly with shoddy electrical work.

 

 

[JohnRoberts]

from NEC code-  http://www.international-electrical-supplies.com/electricity-nec.html


The NEC also has provisions that permit the use of grounded-type receptacles in nongrounded wiring (for example, the retrofit of 2-wire circuits) if a GFCI is used for protection of the new outlet (either itself or "downstream" from a GFCI). Art. 406.3(D)(3). The 1999 Code required that new 240-volt outlets be grounded also, which necessitates a fourth slot in their socket faces. U.S. 240 centertapped single phase has two of these slots being 'hot', with the neutral being the center tap. There is only one standard for these circuits, but 240 V sockets come in two incompatible varieties. In one the 'neutral' slot accepts a flat blade-prong. In the other the neutral slot accepts a blade with a right angle bend. These are officially NEMA types 14-50R (commonly used with number 8 wire for electric stoves) and 14-30R (commonly used with number 10 wire for electric clothes dryers), respectively, and differ only in current rating (50 A versus 30 A); previous installations would have used the 10-30 or 10-50 configuration. These changes in standards often cause problems for people living in older buildings.


A 120-volt GFCI socketreakers and fuses, which only open the circuit when the "hot" current exceeds a fixed value for a fixed time, a GFCI device will interrupt electrical service when more than 4 to 6 milliamperes of current in either conductor is leaked to ground (either directly or through a resistance, such as a person). A GFCI basically detects an imbalance in the electricity (that is, a different amount of electricity flowing in the "hot" side than what is flowing in the "neutral" side). Socket outlets with GFCI have the added advantage of protecting other sockets 'downstream' of them, so that one GFCI socket can serve as protection for several conventional outlets, whether or not they are grounded. GFCI devices come in many configurations including circuit-breakers, portable devices and outlet receptacle sockets.

A GFCI socket typically has a pair of small push buttons between its two receptacles: one labeled 'test' and the other 'reset' (or T and R). Pressing 'test' will place a small imbalance in the line sensor, which will trip the device, resulting in an audible "snap". Pressing 'reset' will allow the socket to function normally after a test, or after a faulty appliance has been removed from the circuit or insulated from ground.

Like fuses and circuit breakers, a GFCI socket has a finite number of uses. It must be replaced when a test fails to trip the device.

 

[PRR]

> Photo is sort of dark

There's your first problem: all 3 wires are black.

Conceptually, it would be nice if they were white, blue/black, and red.

Or at least white/bare and two black.

Leaving the "neutral" wire black is poor form. Actually banned on the property. This point is usually Utility Company jurisdiction, and an Identified Neutral may not be required.

Knowing that the Neutral was not Identified, the electrician who took this onto the fusebox buses should have been super-careful.

Clearly one of the hot wires has been landed on a fusebox common. (And not necessarily the "new" work; though the box feeding the stove is *probably* correct or the 230V stove would be 120V lame.)

There IS another problem. All the service commons MUST be bonded together. Very preferably with a solid copper wire. There are cases where each may only be bonded to dirt.

With a copper bond, landing the wrong black wire on fusebox common WOULD have blown either a main fuse/breaker or that copper wire.

If bonding is through wet dirt, you should have a "hot rod". In winter, the snow melts around it. In summer, barefoot, you get increasing shocks as you walk up to it.

If bonding is through dry dirt, or overlooked, maybe nothing happens until somebody touches (or runs a wire between) the cases of two appliances. As you and John say, this is Very Dangerous.

I'm thrilled the landlord responded super-well. This may be a "easy fix", swap two (live!!) black wires (but which ones?) (And the feeder should be disconnected, which involves the utility company.)

The electrician will get the black wires sorted. But please ask him to inspect the DIRT-ROD bonding, because clearly it isn't working (or there is some exceptional situation).

 

[PRR]

> permit the use of grounded-type receptacles in nongrounded wiring

Very off-point.

He's not reporting any "open ground" on his 3-lite tester. He probably has 3-conductor circuits.

And un-grounded outlets would be "safer" than what he appears to have: "neutral" and 3rd-hole 120V above dirt AND other circuits in the same room.

When the hot/neutral swap AND the dirt-rod bonding are fixed, he "should" be OK. Not certain: there is always the trick of tying green to white at the outlet. If a neutral comes loose, "ground" is at 120V (and the outlet appears "dead", inviting investigation). It is to avoid such maybe-dangerous cheats that the ungrounded GFI exemption exists. You should NEVER use that clause until SURE that the entrance/fusebox is wired correctly, that the circuit is correct 2-conductor, and that you do NOT need the audio/data advantages of a solid ground, merely wish to avoid powerline electrocution.

> current in either conductor is leaked to ground

Strictly: when a 4mA difference exists between black and white "hots". When the books do not balance.

If you touch a GFI-protected black, and a black on the other side of the 240V/120V split, the GFI will trip, though no current flows direct through ground. Other more-contrived examples are possible.

With hs 120V-hot "common", a GFI outlet/breaker will NOT trip if he touches two appliance cases 120V apart: a US GFI does not sense ground current.

 

[JohnRoberts]

The black wires in the picture are still on the power company's side of the meter, so I am hopeful that the power company linemen, wired it up to the meter correctly.

Trust but verify. It should be pretty straightforward to ring out with VOM and long ground lead.

Lets hope the mistake is somewhere easy/cheap to fix.


JR

 

[0dbfs]

If bonding is through wet dirt, you should have a "hot rod". In winter, the snow melts around it. In summer, barefoot, you get increasing shocks as you walk up to it.

When the comcast guy installed our internet/cable service a couple months ago he mentioned that while crimping the f-connector onto the coax (from the street to our house/demarc) that he was kneeling on the ground and felt the tingle down through his knee into the damp earth. The cable modem that was connected up to the other end of that coax was plugged into one of these sockets. He bonded his shield to a gnd wire that runs to bldg copper in the crawl space. I didn't think much of it at the time. As a matter of fact, I was thinking something like: This cable modem is running of a 125mA 5V supply... He couldn't have felt current..

I'm thrilled the landlord responded super-well. This may be a "easy fix", swap two (live!!) black wires (but which ones?) (And the feeder should be disconnected, which involves the utility company.)

I was thrilled too but have not heard from the electrician yet. It is new years-day. If he feels like i do, I hope he stays home instead of messing with my "problem".

So once again check that the neutral bus at both service service panels have no significant AC voltage between them. You can probably measure between screw heads on the panels. They should get a decent ground from the pole, but confirm to a cold water pipe if metal plumbing is handy.

The panels don't have voltage between the chassis nor bldg-copper. Conduit (one steel, one sheathed/pvc) from the service entrance terminates directly into the distro-panels. Romex comes out for circuits. No apparent bonding between the distro-panels although I didn't measure resistance between the chassis yet.

He's not reporting any "open ground" on his 3-lite tester. He probably has 3-conductor circuits.

Most of the romex in the basement looks like 3-conductor flat-white'ish cable and the various junction boxes are mostly open. Lots of romex & junction boxes down there. Some junction boxes with up to five parallels. My guess is that this is where the source of the fault lies. The woven-aluminum-sheathed romex may be two conductor Hard to tell here but the woven-jacket should be capable of GND. Maybe that's how they intended it to be used. There is plenty of that installed. House is from the 30's.

Thanks a lot for all the input and have a happy and safe new years.
Best,
jonathan

 

[0dbfs]

You should NEVER use that clause until SURE that the entrance/fusebox is wired correctly, that the circuit is correct 2-conductor, and that you do NOT need the audio/data advantages of a solid ground, merely wish to avoid powerline electrocution.

Yes, I'm not in the corrections-institution-business. Just music, broadcast, audio, and studio business. However I have been talking about implementing one of those electric-dog-collars to get talent to perform on demand. Maybe I can just run an inter-chassis-connection through my N.O. momentary foot-switch and the talent-chair and back to earth.

Seriously though, I am preparing to install a full patchbay into this room and have been considering the best-grounding-scheme for terminating audio-cable-grounds. The potential for potential has me visualizing this aspect a little differently now.

Best,
jonathan

 

[JohnRoberts]

Another quick question... When your equipment released it's smoke because of the hot ground, did it pop a fuse/circuit breaker in your panel?

If your two panels are sitting at same ground/neutral, a hard short between the two suspect outlet ground pins should identify which branch is miswired by popping the fuse/breaker (or burning your house down).  8)

HNY

JR