2 months ago I moved into a house from 1959 (1 220V phase and 1 neutral coming in). I noticed on the 2nd floor bathroom that there was stray voltage on the faucet (-+ 30V). No socket outlets in the house have a ground pin. When I connect a power bar (which has internal grounding), I notice a stray voltage on its sockets too.

I installed an earthing rod (260cm / 8,5feet in the ground). I tested it while connecting a light bulb (40W) between a hot wire and the ground, and it bright up. Next I connected the earth cable to the breaker panels earth bar. There were already 2 ground cables from somewhere in the house connected to this bar, and I did not remove them.

I figured this would solve the problem of my faucet, but it didn't. Whenever I would stand on the bathroom floor and touch the faucet I could feel the stray voltage.

I then connected a new ground cable from the breaker panel's earth bar to the faucet. Using a volt meter I got these 3 readouts:

3 measurements

1) I'm standing on the bathroom floor, and hold a Volt Meter's pin1 in my left hand, while I hold pin2 against the faucet

2) Same situation, only this time I connected the new ground cable to the faucet

3) This time I hold pin2 directly against the ground cable.

I have no idea what's going on. How can my ground cable have a voltage of 30V, when I'm measuring between the bathroom floor and the ground cable (20m / 65feet) to the earth rod?

Update 7 months later: This bathroom has been fully renovated now. I have a suspicion about what was going on initially. Under my feet, inside the flooring there was a metal tube which held insulated electricity wires. At the endpoint of that metal tube was a light switch 1 floor underneath the bathroom, and I think it probably was a faulty device leaking current (a bad light switch can act as a sort of capacitor I think). (= fault 1) The metal tube must have been connected in some way to the light switch. (= fault 2) By standing directly above this metal tube I put myself at 30V potential, and while touching the bath faucet (which I believe is a natural ground in any house) I felt this.

  • Your bathroom floor may well be 'more grounded' than tour ungrounded faucet, but it's nowhere near as grounded as your real ground rod.
    – brhans
    Commented Aug 31, 2017 at 13:40
  • but then why are my bathroom floor and ground rod at a different potential?
    – andy
    Commented Aug 31, 2017 at 13:48
  • 2
    Your house doesn't have a 30V source. What you're feeling is 240V impeded by high resistance, which is the only reason you are still alive. Whatever conditions are causing that fault and resistance are not engineered for safety, they are happenstance. If those conditions drift somewhat, they will kill you. Or trip your whole house GFCI. Commented Aug 31, 2017 at 14:23
  • 1
    take a look at this thread Commented Aug 31, 2017 at 14:33
  • @UlugToprak: Thank you, I understand that thread. But for me, it does not answer my last question (situation 3).
    – andy
    Commented Aug 31, 2017 at 15:32

2 Answers 2


One of your appliances has a ground fault

It often comes as a shock to people, but sometimes their appliances have a defect. Its internal electrical insulation isn't up to snuff, and it is leaking current (typically from the "hot" wire as neutral is usually harmless). If the leak is small, it may self-limit current to feelable but non-lethal amounts. Such leakage tends to get worse, so what tickles today could kill tomorrow.

Essentially, the insulation failure and the human become a resistor ladder as both are resistors and both are in series. So the amount of the insulation leakage, and the quality of skin contact, determine resistance, and thus the amount of current. Do it again with wet hands and now your resistance is much lower and you're dead.

As such, any tickle should be considered extremely dangerous and repaired with extreme prejudice.

DVMs (Digital Voltmeters) are not 100% reliable. Their resistance is so high (in theory they should read a fault as full line voltage) that they often misread ordinary capacitive coupling as relevant voltage.

Don't use yourself as a voltmeter. Hunt down a testing device that successfully indicates the fault, e.g. a non-contact voltage tester, mechanical meter voltmeter, neon light tester, etc.

GFCI/RCD is a starting point

If you want to get personnel protection in place tout suite, which I most definitely recommend, then install a GFCI aka RCD device powering either the offending circuit, all nearby circuits, or the whole house (which is the usual approach in Europe). An 8ma threshold GFCI/RCD will provide personnel protection. A 30ma threshold will provide limited personnel protection, it can still kill, especially by a stun causing a drowning or fall.

Hunt it down, one appliance at a time

Now you need to move through each of your loads, and see which load or device is causing this. You can start by shutting off one circuit breaker at a time, and see which one extinguishes the effect.

Once that's the case, note which devices are offline with that breaker off, and move through them one at a time.

Very rarely, it can be something totally out of left field. I have no idea how to advise patrolling for that.

  • "It often comes as a shock to people" I hope this pun was intended.
    – Hari
    Commented Mar 16, 2018 at 19:43
  • 1
    @HariGanti Very much so. It's a trope around here, people wanting to know why their <X> fault detector is tripping, taking for granted that it couldn't be any of their devices... Commented Mar 16, 2018 at 20:40
  • @Harper, could this situation be the result of two simultaneous failures? First that the appliance has faulted to it's chassis, then also that perhaps the ground (water pipe?) is not bonded to neutral. In this case the appliance energizes the ground and then does not return to the source (i.e. neutral). I know Europe uses 220(240?)V but not sure how Europe does things with split phase and all that. Commented Mar 16, 2018 at 21:16
  • 1
    @KerryThomas His house does not have a grounding system (and even then, Europe is a little peculiar about ground bonds) so his appliance has no ground pin to return current on. This is the picture postcard use-case for GFCI/RCD, except he may have Euro 30ma RCD, which won't trip on a tingle. Commented Mar 16, 2018 at 21:51
  • @Harper thanks for your answer, I have updated my question
    – andy
    Commented Mar 22, 2018 at 23:16

I had this problem IT was caused by an electrician that installed the neutral wire without removing the insulation on the wire. He tightened the connection very tight just making a resistant connection. I removed the insulation, reinstalled and solved the problem

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.