Can I put a low current (<5 mA) 230V coil, NC (normal close contact) relay between one wire of mains and the ground (earth) wire in order to check mains polarity?

In theory if the wire is line (hot) the relay will switch and prevent the equipment to power on.

On the other hand, if the wire is neutral (correctly) the relay will not switch, as supposed to be at same potential of the ground and the equipment can be turned on.

Is this insane?

Of course there be a LED circuit to show if ground connection is good.

  • \$\begingroup\$ Why does your polarity matter? It suggests that you are trying to compensate for a design flaw somewhere such as having the equipment chassis directly connected to neutral. \$\endgroup\$ – Transistor May 30 '19 at 16:09
  • \$\begingroup\$ I want to have the central tap in the secondary of my balanced isolation transformer connected to the neutral an not to live. \$\endgroup\$ – Gianluca G May 30 '19 at 18:27
  • 1
    \$\begingroup\$ That's the sort of thing I was afraid of. See my answer below. \$\endgroup\$ – Transistor May 30 '19 at 18:31

You could but it is not a surefire technique since you could trigger an GFCI device (Ground Fault Circuit Interruptor).

Moreover this could be really UNSAFE.

The fact is that whatever current you draw from the live wire it ends up to earth ground through the protective conductor (earth ground wiring system - green/yellow conductor).

This adds up to whatever leakage current is already present in the system. If the total leakage toward ground reaches the GFI trigger threshold (30mA usually) you have the whole system disconnected.

You really need a coil with an extremely low current to avoid that spurious trigger in any case.

This test is what is done by mains testers screwdrivers.

The safety problem comes because mains lines are subjected to voltage spikes that requires adequate safety countermeasures designed into whatever you connect to them. If, for example, your coil is connected to the mains when a 2000V spikes happens, the coil insulation could break down, giving rise to high-energy arcs, fires and explosions, if the coil is not designed with overvoltage protection in mind (a coil rated for 230V AC is not necessarily able to stand an overvoltage event on 230V mains).

This is the same problem you have with capacitors: you cannot place a generic 230VAC-rated capacitor across live and earth ground, you need a special Y-class rated capacitor which is guaranteed not to fail spectacularly during an overvoltage event.

  • \$\begingroup\$ In theory the live connection should last only few moments: the time to rotate the plug in the socket. For the rest of the time it will be connected to the neutral without consuming any current to ground. And in any case a 2000v spike will cause much severe disaster on all the other equipments connected at house with a ground connection. But i appreciate you comment. How to protect the coil from such a rare event? \$\endgroup\$ – Gianluca G May 30 '19 at 18:23
  • 1
    \$\begingroup\$ A 2 kV spike doesn't cause a severe disaster. Actually EN 61000‒6‒1 standard requires to test surge with that voltage, if you would like to sell your product in EU, so most likely the spike causes nothing at all (glitches during the spike are allowed, but the equipment must recover afterwards). The event is not that rare, lightning strikes cause that kind of voltages. You protect the equipment typically with a MOV. \$\endgroup\$ – TemeV May 30 '19 at 20:09
  • \$\begingroup\$ If I put a 275 v varistor across the coil that should be enough? \$\endgroup\$ – Gianluca G May 30 '19 at 20:21
  • \$\begingroup\$ @GianlucaG: No. What you are proposing is not safe. It does not fail safe. If your relay fails to pick up you have no protection. \$\endgroup\$ – Transistor May 30 '19 at 21:01
  • \$\begingroup\$ Please check this. Sligh changed; I don't see anything unsafe here, or at least less safe than the normal electronic we have at home. static.wixstatic.com/media/… \$\endgroup\$ – Gianluca G May 31 '19 at 11:35

From the comments:

I want to have the central tap in the secondary of my balanced isolation transformer connected to the neutral an not to live.

That is not the correct way to do this and is potentially dangerous as well as presenting an earth leakage which will contribute to tripping of your RCD / GFCI.

In Europe where 230 V phase to neutral is the norm portable tools are powered on 110 V, centre-grounded transformers. This gives a 55-0-55 secondary (with 110 V between the outer terminals) effectively limiting the shock hazard to 55 V with respect to ground. If either phase shorts to ground its fuse will blow. In the event that the ground connection is lost (which might not be unlikely on a building site) some reduction in risk is provided by the reduced voltage of 110 V.


simulate this circuit – Schematic created using CircuitLab

Figure 1. European tool transformer wiring. Note that the grounded centre-tap limits the phase voltage to 55 V AC.

You could take the same approach to your application.

  • \$\begingroup\$ It is safer connecting the central to neutral than to ground: in your scheme if you touch a grounded chassis with one of the two phases a current more than 30 ma may kill you because the Rcd will never trip. If you connect the central to the neutral the power is still balanced because ground and neutral are joint together at the source but the Rcd will trip for any fault current from one of the two phases to any grounded chassis >30 ma. \$\endgroup\$ – Gianluca G May 30 '19 at 19:56
  • \$\begingroup\$ douglas-self.com/ampins/groundloops/grndloop.htm#bal \$\endgroup\$ – Gianluca G May 30 '19 at 20:14
  • \$\begingroup\$ "It is safer connecting the central to neutral than to ground ..." No, not unless you use a second isolating transformer as shown in your referenced article at C1/C2. If you need RCD protection then add that after XFMR1 in my Figure 1. "... but the Rcd will trip for any fault current from one of the two phases to any grounded chassis >30 ma." Again, no. There will be no affect on the supply neutral. Current will only flow in the transformer secondary circuit and the RCD will not trip. \$\endgroup\$ – Transistor May 30 '19 at 20:58
  • \$\begingroup\$ Please simulate in Spice your circuit attaching a resistor of 2K between the right end point of F2 (or F3) to ground, and connect ground to neutral at the beginning in point N as it is in real life. The resistor is simulating your body touching one of the two phases of the balanced transformer and any grounded chassis in your home. There will be a current across the resistor of nearly 60mA which will kill you (the fuse will not blow as supposed for much higher current), but the RCD protection between L and N in yr home will not trip as the current circulating in point L and N stay the same. \$\endgroup\$ – Gianluca G May 31 '19 at 4:18
  • \$\begingroup\$ Now connect the central tap to Neutral, and again a resistor of 2K between the right end point of F2 to ground. You still have a current of 60mA, but in this case the RCD protection between L and N will trip! I have simulated and built, and it works as I just described. \$\endgroup\$ – Gianluca G May 31 '19 at 4:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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