11

I am in Indonesia and the local electricity company is bad and there is no actual electrician around. People just connect wires together.

So with this in mind I recently ripped out my wires (unearthed) and put in new earthed ones:

  1. I have a digital meter, ungrounded.
  2. This connects via 50m long NYY 2x2.5mm cable to my consumer unit (self-installed)
  3. I have a ground rod which I buried recently in the ground, connected to my consumer unit.
  4. I have a plug-in receptacle tester which shows 11V neutral -> earth
  5. The correct voltage for Indonesia is 220V, but I have 240V-250V (a few years ago it was about 195V, because there was about 50kVA connected across neighbouring properties to a 20kVA transformer, so I got them to install a new one after some lobbying, and I believe there is a knob on it which they turned up too high).

So I have a fault but I am not exactly sure where.

The earth rod has not been tested using any tool, but it's rainy season and the ground is extremely wet right as I speak, so it doesn't seem like there would be a problem here.

Can anyone suggest how to resolve?

Edit: I just tested one circuit (which uses 2.5mm wiring), and plugged in a 1800W kettle. With the kettle OFF 251V L->N, and 11V N->E. With the kettle ON 237V and 8V respectively.

Edit2: I found some info on Youtube which seemed fairly credible, and checked my own meter. https://www.youtube.com/watch?v=ImdN08dHiYQ

This is the diagram on YT

enter image description here

You can see there are five terminals.

Here is my meter:

enter image description here

The labels are not clear in my photo but they are printed on the connector: 1 (phase in), 2 (phase out), ⏚, 3 (neutral in), 4 (neutral out). So the bottom ⏚ screw is supposed to be connected to a ground rod, but there was never one installed, and then the top ⏚ is supposed to bond to the top screw of neutral in.

Now, according to the video if the ground had been bonded at the meter (which is impossible as no ground exists there), then if I additionally bond at the consumer unit, this would cause the tamper detection on the meter to trip and it would stop working.

In this case where there is no connection between ground and earth at the meter, then the video suggests that if we instead bond at the consumer unit, then if the neighbour has a current leakage then that can make its way to ground at our consumer unit, and be read by our meter. For this reason it is recommended to ground at the meter.

So it seems like I have three possible solutions:

  1. connect the ground and neutral in my consumer unit with the possible cost of paying for the neighbour's leakage
  2. install a ground rod at the meter (because it is 50 metres away then this seems more practical than grounding it to the existing ground rod at the house) and then ask the electricity company to open up the meter and ground it for me
  3. have the electricity company move the meter nearer the house and connect to my ground rod. This is probably better than option 2, but likely to be troublesome and time consuming, so I will go for option 1 for now.
16
  • 4
    Yes, and the answer is usually meter anomaly or phantom voltage. You may not actually have a problem.
    – isherwood
    Nov 10 at 16:04
  • 2
    240V is not necessarily a problem - see en.wikipedia.org/wiki/Mains_electricity_by_country shows 230V, and in general most equipment for 220V systems is really designed for a range typically 220V - 250V. Where/how are ground and neutral connected? Nov 10 at 16:04
  • 8
    "there is no actual electrician around. People just connect wires together." I feel for you, my friend. I really do...
    – FreeMan
    Nov 10 at 16:18
  • 5
    @Ecnerwal based on the first sentence of the question, it seems that "local practices" are "it hasn't killed me (yet), so it must be right".
    – FreeMan
    Nov 10 at 16:32
  • 1
    If the meter is ungrounded, then you need (exactly one) bond between neutral and your grounding electrode. That would be "at your consumer unit" to the extent that I understand non-USA power terminology (="Main breaker panel" where the power comes into your property.)
    – Ecnerwal
    Nov 10 at 16:53
7

I have a 5/8" 8' copper-plated ground rod connected with 6mm^2 copper wiring to the ground bus bar in my Schneider consumer unit

This may be the key. In the US (different system, but hopefully close enough), the neutral and ground are supposed to be bonded by connecting them in one and only one place. That can be near the meter ("meter main") or in the main panel. (In years past, there were subpanels with neutral and ground bonded as well.) Because of the need for the bonding in only one place, many panels will ship without neutral bonded to ground. When that is the case, there is normally a place for a screw to connect the neutral bar to the case, or it can be done by connecting a wire from the neutral bar to either the case or the ground bar. Instructions for the panel should describe the approved method.

Without bonding, neutral should be close to ground but will likely not be exactly at ground, and an 8V - 11V difference would be quite normal.

Bonding is important because if there is a short from hot to ground then you want to make sure current can flow properly in order to trip the breaker.

6
  • thanks I updated my OP after searching this on Youtube in Indonesian. It seems that because the meter itself is required to be bonded there is no specific bonding screw in the panel at the house, because this is wrong. But as you say, it needs to be bonded somewhere, which currently it is not. and this is the only option I have for now. So I will tomorrow try to run a wire from the neutral to ground bus bar and see if it eliminates the voltage.
    – thelawnet
    Nov 10 at 17:36
  • Typically in the US, the bonding is not at the meter unless it is actually a combination meter base + main panel = "meter main". Model # of Schneider panel? Nov 10 at 17:38
  • it is a DOMH12112F. IDK if you have digital prepaid (token) meters in the US, so I guess things are a little different.
    – thelawnet
    Nov 10 at 17:40
  • 4
    wow a different world indeed "prepaid (token) meters", "there is no actual electrician around", "people just connect wires together."
    – P2000
    Nov 10 at 18:08
  • 4
    I just bonded my neutral and ground and the earth->neutral voltage is now zero as expected
    – thelawnet
    Nov 11 at 7:15
10

What's up with all the voltage drop?

Easy. You have a 2.5mm2 cable (~13 AWG) going 50m (160'). That's a very small wire for such a long distance. In Canada you would not draw more than 15 amps on that. In North America I would tell you to use 4 or 2 AWG (25 or 35 mm2) aluminum wire for that run.

So you are (not surprisingly) having 14 volts of voltage drop (so you say). The hot and neutral are each having half of the voltage drop, so 7V each. Thus I would expect neutral to rise 7V above earth when experiencing 14V drop.

The cornerstone of your system's protection is the RCBO

If the RCBO is tip-top, all manner of other electrical sins can be tolerated. Which would explain why your country's code enforcement is so slack; the working RCBO removes most of the danger!

Bypassing it is madness, but very typically we hear this:

"I tried leaving it connected, but it trips immediately when I do. I cannot find any way to make it not trip immediately. Surely it is broken?"

The answer is always "No, it is NOT broken! There is an actual system fault, a real residual current leak: the very thing it is designed to detect - and it is tripping because it is doing its job correctly. What you actually need to do is hunt down and eliminate that fault."

And the response is always "Maybe, but a) I don't know how to do that, and b) I don't care because I don't understand the risks. Ignorance is bliss and I just want my power back on".

But bypassing RCBO on a known-faulty system is sheer madness.


So it would really be much better to raise your skills so you can fix the root problem. You can narrow it down by disconnecting circuits either one at a time or "half at once" - note to disconnect a circuit you need to disconnect both live AND neutral. Because the RCBO cares about leakage from neutral to earth, and that will trip it too.

Start by unplugging all your plugged-in loads. If it holds, plug things in one at a time, and wait to see if it trips. When it trips, it was the last thing you plugged in.

If that doesn't work, it's in the hard-wired circuits. Use process of elimination. Suppose you disconnect circuits 1-2-3 and it holds. So you know the problem is in circuit 4, 5, 6 or 7. You reconnect circuits 4-5. It trips. You disconnect circuit 5. It holds. Must be circuit 5!

Now you follow circuit 5 for half its outlets and disconnect hot and neutral there. It still trips. So we know the problem is in the first half of circuit 5. Now we divide that half into two quarters. With the first quarter only connected, it holds. So we know the problem is in the second quarter. Repeat until you are down to one single appliance. GOTCHA!

Then, have that appliance repaired or replaced.

If your refuse to use whole-house RCBO, consider many RCDs

People laugh at Americans for having many individual RCD's (GFCIs) on a per-circuit or even per-outlet basis. But you are yourself experiencing the method to our madness. In America when a GFCI trips, it only takes out that one circuit or receptacle, not the whole damn house.

So you could try that. Replace each of your individual branch-circuit breakers (or even receptacles, if that option is available) with appropriately rated RCD devices. Now you will only lose the faulty circuit instead of the whole house.

This also gives you the option of removing well-earthed but GFCI-tripping appliances from GFCI/RCD protection altogether. For instance we don't protect furnaces and refrigerators, which are notorious for tripping GFCIs, because it is a well-earthed appliance with the electrical parts inaccessible.

Your earthing system needs to be absolutely tip-top

And European-style installations are notorious for bad, bad earthing - it is generally geared for European apartment buildings, where relying on the utility to "earth the neutral" is not unreasonable.

When you're spread out on a property like an American homestead, I recommend the American viewpoint of earthing:

  • An earth wire (green or yellow/green or bare) is carried to all buildings and all connected
  • Every building gets its own competent system of earthing rods - typically two 2.5m rods placed at least 2m apart (preferably opposite corners of building)
  • The earthing rods are bound to the earth wire at each building, in that building's panel/consumer unit

And then, the most important part: There must be only one earth-neutral connection. If you had two, then neutral current would detour via the earthing system. Generally we put this at the first disconnect point past the electric meter. However in an installation where there is a long feeder from the meter and that does not have an earth wire, we put this at the first place there is an earth wire.

The neutral-earth bond cannot be on the protected side of the RCBO! (that will instantly cause it to trip). So the RCBO must be "downstream" of the one neutral-earth bond.

The Youtube diagram shows the neutral bond at the meter (between 3 and 4) which is viable. Note the diagram shows a plain circuit breaker. If it were an RCBO, the neutral wire would be routed via the RCBO.

If you do not have an earth wire between meter and first building, then you will need to do something like this.

enter image description here

13
  • 1
    A sidenote: the arrangement Harper is describing in this post is called "TN-C-S" in the IEC earthing standards Nov 11 at 1:25
  • 1
    @thelawnet well the #1 thing about using RCBOs when that is unusual, is improperly installing them. Mainly, they are different from everything else: the neutral must go through the RCBO as well as the live. If the neutral does not, then the RCBO will instantly trip on any use. You definitely need one that fits. Beware of cheap ones, as they may contain nothing at all. BigClive has done a teardown on DIN rail "circuit breakers" that were nothing but a switch inside. Nov 11 at 18:18
  • 1
    @thelawnet Ok, so your RCBO tests aren't really sufficient. You need to disconnect all "Load side" wires from the RCBO (both hot and neutral) and then see if it will stay on then. If it does hold, then it is probably not faulty and you have a second fault in the wiring (and troubleshoot as discussed, remembering to disconnect neutral also). I would not spend US$75 on an RCBO until I tried that, even if there was a Wickes shop a mile away! Nov 11 at 18:24
  • 1
    @thelawnet The Test button on an RCBO will only work with power applied to the input terminals. It cannot trip otherwise.
    – Simon B
    Nov 11 at 21:11
  • 1
    Right, a meaningful test requires an imbalance of current, so tripping requires the GFCI itself to be powered! So power needs to be connected on the "Line" (input side) terminals. The purest test will have nothing on "Load". It is physically impossible to build a true RCBO for $10US. Even $30 seems a bit low ($40++ in the US)... but I don't really understand PPP macroeconomics. Someone should buy the $10 one and send it to bigclivedotcom on Youtube, and ask for a teardown video be made! Nov 11 at 21:56

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.