Neutral Earth Fault ??

[Gazzza]

Decorating the hallway, so Saturday morning I started what should've been a simple job of changing the hall/landing lights.

However (before I'd hung the new lights) I noticed when the landing light was switched on, the earth returned positive when I tested it with a mains testing screwdriver. Very very faint, but the bulb in the tester was definitely on.

A quick reccy round the other light switches upstairs, I've found one other that seems to do the same in one of the bedrooms. ie. very faint glow on the mains tester against the earth when switched on. Interestingly the other lights lights do not appear to do the same (though it could just be the mains tester was so faint it was not visible).

I completely unwired the hallway and landing lights at the ceiling roses and both the switches (2 way, top and bottom of stairs) so they were completly isolated and using a battery operated circuit tester, there were no problems between any of the cables running from the switches to the roses or between the switches. ie, I tested on every cable, live to earth, neutral to earth, live to earth. all fine none of them completed a circuit.

However at the ceiling rose, the incoming loop feed, from the direction of the consumer unit, returned a completed circuit if I tested neutral to earth - circuit obviously isolated by the MCB at this stage. This problem traces all the way back to the consumer unit. ie. If I completely unwire the first rose in the circuit - the remaining roses in the circuit test ok - no apparent faults. But the incoming loop feed from the CU tests positive across neutral and earth. This is the same upstairs as well as down.


Am I right in thinking that testing across neutral & earth in the feed from the CU should not give me a compete circuit.
Am I right in also understanding that if neutral and earth is breached, as it seems, this will then affect every other circuit in the house given that at the CU all the earths are together and all the neutrals are together.
Does this also mean that the fault might not be in the light circuit at all and could be in any of the other circuits ?


Thanks,

G

PS. sorry that was a little long.

 

[JohnD]

There is always a tiny bit of leakage, but if you are using a neon tester you will be advised to throw it away and buy a test meter.

You can get an inexpensive multimeter for £5 to £10 which will tell you a lot more about what is happening.


BTW take a look at the earthing arrangements around the CU, and in the bathroom, and tell us what they are.

 

[Gazzza]

The intitial test on earth I mentioned at the start of my post was, as you say, just using a neon tester screwdriver. It was barely lit up but definately faintly glowing (but significantly less than testing live for example). n.b. i tried 2, both with the same effect.

You mention there's always a tiny bit of leakage - but would you ever expect this to be sufficient to return sufficent power to pick up on a neon tester. i'd have thought not, but would be happy to be corrected ?


So I then isolated the circuit via the MCB and then used a battery operated circuit tester. ie. one of these : http://www.focusdiy.co.uk/invt/288286

With one probe on the neutral, and one on the earth, it tests positive. This is on the cable from the CU to the first ceiling rose.


Re. earthing - non of the pipes in the bathroom are earth bonded.
The CU is in an odd location - basically its sunk into a hole in the wall in the dining room - its pretty tight so theres no room around it. The supply cables from the meter come straight through the wall behind from the meter box outside.
The earth block in the CU has an earth cable running down - but it disappears down the cavity between the walls so i can't see where it ends up. It doesn't re-appear anywhere visible. This is from recollection as im not at home at the mo. but will dbl check.


Cheers

 

[JohnD]

If you look in the meter box, you should see the incoming cable rising up from the ground (or down from a pole), It will have a big plastic or iron head on it with two insulated cables going to the meter, It will also very likely have a fat green-and-yellow earth cable (about as thick as a pencil) on the side that will go to the consumer unit. Have a look and see if this is the case. There may also be a label saying "Protective Multiple Earth".

If you have an overhead supply you might not have that, and may have an earth spike in the ground close to the meter box.

Look for evidence that there is green-and-yellow earth cable going to your incoming water main and gas.

About how old is your consumer unit, does it have fuses or MCBs?

An old installation might not have good earth conections.

The neon will not tell you how big the leakage is, a meter would be better and give you a voltage reading.

P.S. It is not a problem if the Neutral and the Earth give continuity. This is normal. There will also be a certain amount of resistance betwen them, hopefully small, and a certain amount of voltage, hopefully small.

 

[Gazzza]

Cheers - will check out the earth situation in the meter box tonight.

CU is an old style Wylex standard jobby. But its fitted with plug in MCB's, appart from the 45amp shower circuit which has a cartridge fuse in a plug-in holder.

I'm a bit more relaxed now you've told me that its normal for neutral and earth to offer some continuity.

G

 

[Gazzza]

Ok checked earth arrangements last night - looks like earth is supplied by the electrricity company.

Earth from CU goes straight out through the external wall into the exterior cupboard housing the meter and electricy company's fuse. The earth terminates at an earth block, where it's picked up by the thick earth cable running straight into the main supply coming up from the ground from the electricity company. It's all very obscured - i can't see whether or not there is a protective multiple earth label.

 

[JohnD]

Any chance you can lay your hands on a digital camera and post a pic of the insides of the meter cabinet and the CU?

 

[Gazzza]

Will try tonight - wasn't home last night.
Is there anything specifcally I should be looking for.

Just wondered what your thoughts were.

 

[JohnD]

I will be interested to see how up to date the installation looks, if it looks like TNC-S, if the main earth bonding looks right, anything that looks strange or unusual.

 

[Gazzza]

No luck getting photo yet.

Some extra info. in the meantime:
With the main Consumer Unit (ie. the whole CU and not just an individual circuit) switched off, the earth / neutral does not test positive for continuity with the circuit tester. Flick it back on, and the earth & neutral gives continuity. Would this imply it is TNC-S ?


Secondly, with the CU back on i turned off every circuit at the MCB's (apart from the upstairs lighting circuit). Then, at the two relevant light switches, tested the earth connection just with the neon tester again.

With only the upstairs light circuit on, one of the switches no longer appeared to show the original problem (or possibly it was insufficient to light the neon tester). The other switch still seemed to show the fault but it was barely visible or intermittent.

Then, turning on either the downstairs lighting circuit or the upstairs main circuit, had a notable effect that both light swtiches now definitely lit up the neon tester from earth when switched on.
With both those additional circuits on, the neon tester was brighter still. (still nowhere near as testing a live, though i appreciatte this is crude)

Turning on or off any other circuit had no effect.


Not sure if that helps any further?

 

[JohnD]

I really think you need to splash out that fiver on a meter!

As to the main switch; it is quite likely that you have TNC-S as it is most common in the UK. the main switch on the CU is dual-pole so isolates both the phase, and the neutral. The TNC bit is then isolated from your household Neutral by the switch. That bit is no surprise and does not indicate a fault.

Whether your other effects are caused by induced currents or capacitance or what is beyond me, but when you can measure it, it will become clearer whether its of any significance.

Slight flickering on a lighting circuit with no resistive load is very common and often the current is so low that it doesn't matter. It is only visible on a neon or some compact fluorecents and is a puzzle when you first see it. It will probably disappear if you put a light bulb in the fitting.



You are lucky that you haven't yet had abuse hurled at you from the many people on here who hate neon testers

 

[Gazzza]

I see what you mean, i just did a forum search for neon tester

Well i have just invested a whole tenner and am now the proud owner of a digital multimeter !
(don't think i've used a multimeter since doing my a levels ! That was longer ago than I'd care to admit )

So i'll get measuring tonight.

Quick question - given my issue is with earth, how do i test that earth.
I understand the basic principles of using a mulitmeter on a simple circuit (ie. testing for current in series, for volt diffences in parallel).
If the i was testing voltage in the live, and i had a good earth, or i was measure voltage across the bulb, that would be easy.
But in the case of my earth problem at the switch, what do i need to test ?

 

[JohnD]

You were worried about voltage on the earth conductor of the lighting circuit. For a start, measure voltage P-N, N-E, P-E

Then put a bulb into on that circuit lighting circuit, switch it on, and measure them all again.

Then take out all bulbs, turn off at the main switch, and measure (lowest ohms range) resistances between them all

Then main switch back on, and measure resistance N-E only.

I bet we will be able to see what's happening then.

 

[securespark]

But its fitted with plug in MCB's, appart from the 45amp shower circuit which has a cartridge fuse in a plug-in holder.

If it is an older style Wylex board with no mention on the front of the main switch rating, you should not fit any device greater than 32A.

Only boards with 100A main switch can accept devices in XS of 32A.

 

[Gazzza]

Yes, I recall this coming up a couple of years ago. Had an electrician check this and it was ok.


Will do the various tests on the lighting circuit tomorrow.

Cheers