I live in a home that was built in the late 1930s with some old electrical wiring and some new wiring. As I was identifying which outlets and lights are controlled by which wires I discovered a brainteaser:

A garage outlet that my outlet tester identified as properly wired and grounded:

enter image description here

However, it only had 2 wires going into it: a red one and a white one:

enter image description here

Furthermore, for some reason, the box that holds the outlet is attached to another box with thicker gauge wires that are black, red and white which are not connected anywhere.

enter image description here

I suspect the other box with thicker (12AWG) wires used to have the outlet for a dryer, while the first outlet pictured used to power the washer (they have since been moved to another location).

But my question remains -- how can my tester (which has been pretty good at showing missing ground in other outlets) be showing this outlet as ALL GOOD in terms of ground even though it only has 2 wires going in?

P.S. Making sense of the source of these wires at the panel is not an option for me -- it's a mess.


3 Answers 3


The conduit is the grounding path

Note that the wiring in your case is run not using sheathed cables, but as individual wires inside a metallic conduit (aka the pipe-like stuff you see heading off to the left in your picture). As a result, the conduit is a serviceable grounding path in its own right, connecting the receptacle grounds and boxes to the grounded panel enclosure without any need to run an extra ground wire through it.

  • 1
    that makes total sense, the conduit is directly connected to the metal panel box which itself is grounded. THank you Commented Jan 13, 2020 at 5:23
  • 1
    Just to be clear though, the fact that the conduit is conductive does not mean that it is conductive enough. That cable tester simply looks for some kind of low resistance, but it would probably report continuity just as well for a single-hair-thick strand of wire. Wiring code (dependent on where you live in the world) may require a thick earth wire, so that in the event of a short to earth, the current drawn will definitely trip out your RCD/GFCI and will not heat up and start a fire. So whilst that tester is a good first clue to miswiring, it isn't the whole story.
    – Graham
    Commented Jan 13, 2020 at 14:51
  • 2
    @Graham: If there is a GFCI/RCD, a grounding conductor would not need to be very thick to trip it without heating up appreciably. Electrical codes, however, are designed to ensure that a grounding conductor would be able to trip an ordinary over-current breaker before overheating. If a grounding path had a resistance of 3 ohms at 120VAC, a ground fault could take many seconds to trip a breaker, during which time the grounding path would pass 40 amps and dissipate 4.8KW.
    – supercat
    Commented Jan 13, 2020 at 16:48
  • 2
    The key (and this applies to both wires and conduits) is that an outlet tester showing OK is not proof that the ground path is compliant, only proof that it is not nonexistent. Commented Jan 13, 2020 at 19:37
  • 1
    @PeterGreen: so in theory the pipe could be better, or in the real world it could be worse, but a known wire is, well, known. gotcha.
    – dandavis
    Commented Jan 13, 2020 at 20:59

The weak link in the grounding may be the cover. Your fully-raised cover only contacts the (grounded) box at the edges and through the screws. A proper ground requires a cover with flattened corners, to make solid contact with the corners of the junction box. The flat contact between the receptacle's strap and the cover completes the ground path.

crushed-corner cover

  • 2
    Interesting notion, but not one I'm familiar with. I'll trade you a code cite for an upvote. Commented Jan 15, 2020 at 1:36
  • 2
    I saw it in the Hubble-RTB [Raco] catalog; 250.146(a), copied from the internet: A listed exposed work cover shall be permitted to be the grounding and bonding means when (1) the device is attached to the cover with at least two fasteners that are permanent (such as a rivet) or have a thread locking or screw locking means and (2) when the cover mounting holes are located on a flat non-raised portion of the cover. Commented Jan 15, 2020 at 2:08
  • Just to be clear, the cover used, and the screws used do not accomplish a 250.146 compliant installation. It would at least need a ground wire from the receptacle ground screw to a ground screw or clip. Commented Jan 15, 2020 at 4:27

I've come across a few boxes that were like this, have accidentally contacted live wires to them (with enough oomph behind the arc to weld it to the box) and I will say wire everything securely and insulate or trim down any exposed wire if you do any work in them. In my experience, the discharge to the box didn't trip the breaker and it was blind luck that I jumped back fast enough to not eat a lethal dose of mains current. e: Insulated tools!

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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