80 volts on earth !!!!!

[johnnyuhtar]

Hi All

Wonder if anybody can give some advice

I tried to do a RCD test - with a Multifunction tester - Seaward Powertest 1557 on my house

It done the half test no problem

Tripped straight away on x1 but displayed "RCD" on the screen and would not give a time

Tripped straight away on x5 and displayed "RCD " on the screen again and would not give a time

The button on the RCD unit is working and tripping it

Tried to do a loop test - the tester would not as it said the fault current would be over 50volts

After an age I narrowed it down to the Down stair ring circuit it had a Live to Earth fault when I tested the Insulation Resistance

I replaced the cable between the 2 sockets concerned - HAPPY DAYS - or so I thought !!

Decided to test the whole lot again from scratch

Disconnected the Main Earth in the CU to do a Ze test 0.20 ohms - It is a  TNCS supply 

Just happened to probe between main disconnected Earth and the Earth bar in the CU  I got   80 volts !!! 

The 80volts reduces as you switch the circuts off but NO circuit in particular stops it

The CU was a total mess and Was not Identified so I stripped it all out so I could make sure everything was where it should be

I have now stripped the board wiring out and tested all the circuts seperatly they all passed IR and R1+R2

But the 80volts is still there

And now the RCD will NOT trip on x1 or x5 and is displaying more than 300ms for the x1 and more than 40ms on the x5 and will not ramp test BUT the RCD test button will still trip it if pressed

Im going to now start testing between the Earth & Neutral on each circuit as I only done the end to end  and the R1+R2

Has anybody ever come across this before ????

Is it possible to check the supply cable to see if it has a broken Neutral or Earth coming in _ just so I can rule that out ??

They are putting up a housing estate across the road and Im wondering if they could of damaged the cable some how

Ant help or thoughts welcome

 

[Murdoch]

Hum ............. sounds like you may not have an earth ......... somewhere ............... or a floating earth ............ or have missed something

A N-E fault is likely to affect a RCD test but unlikely to affect a Ze test .......

I would start by connecting ALL CPC's to the earth bar, then running basic dead tests to confirm continuity (and in this case I would do Rn + R1 to make 100% sure that the neutral has continuity), then IR, then live tests.

Doing this 1 circuit at a time will help you BUT make sure you have ALL CPC's connected to the earth bar at all times when doing IR

Hope this helps

 

[phil d]

If you have a TNCS incomer then a broken cable would affect both the neutral and the earth, since the earth is taken from the neutral in the service head, I assume polarity is correct at the head? (This is one time a neon screwdriver comes in useful) Has this fault only just occurred, or is this the first time you've done any testing on this install, has there been any work done on the property? 

As Murdoch said, do all your tests again and check each circuit thoroughly, these faults can be murder to locate, but when you do they are usually something stupid, a damaged cable on a spur is a good one for throwing up tricky faults, or something connected wrong. I remember having one fault and it was a real pain, it turned out to be a combination of two faults on the same circuit, IIRC one was a missing earth at a junction box and the other was a nail through a cable that had gone between the live and earth. Sometimes doing the figure of eight test and going round each point will show up a lot of things that are not shown up just by normal testing from the board, although quite often people don't bother with this as it takes time. A fault on a spur can be a real mare to find too, one possibility would be to split the ring in half and then identify which half has the fault, then keep splitting that side down until you find it, the old knife and fork method. Fault finding is the one that separates the men from the boys in this game, virtually anyone can wire a house, the skill is in finding the fault, be prepared to think outside of the box, Never assume that something couldn't happen, because anything is possible, especially in an older property that has had Mr DIY working on it.

 

[ProDave]

If I am reading this right, your "floating earth" is the earth bar of the CU WITH the TNC-S earth disconnected.  So it WILL be floating and could be at any induced voltage even with very small leakage.

The fact it is not close to 0V suggests to me things like water and gas bonding are missing? or are those disconnected as well?

If the earth is disconnected then it is not a surprise the rcd tests fail.  The internal test button works because it does not put a fault to eath but a resistance from L one side of the sense coil to N the other side to create an imballance so will work without an earth.  Your proper tester will create a real current to E to trip the rcd.

PUT the TNC-S earth back on and I bet all will be fine.

Try to explain it a bit better if my understanding of the situation is wrong?

 

[phil d]

If I am reading this right, your "floating earth" is the earth bar of the CU WITH the TNC-S earth disconnected.  So it WILL be floating and could be at any induced voltage even with very small leakage.

The fact it is not close to 0V suggests to me things like water and gas bonding are missing? or are those disconnected as well?

If the earth is disconnected then it is not a surprise the rcd tests fail.  The internal test button works because it does not put a fault to eath but a resistance from L one side of the sense coil to N the other side to create an imballance so will work without an earth.  Your proper tester will create a real current to E to trip the rcd.

PUT the TNC-S earth back on and I bet all will be fine.

Try to explain it a bit better if my understanding of the situation is wrong?

Excellent point Dave, I must admit, the possibility of him doing tests with the main earth disconnected had not occurred to me, but you never know.

 

[Murdoch]

Excellent point Dave, I must admit, the possibility of him doing tests with the main earth disconnected had not occurred to me, but you never know.

The op did give a Ze so you would have thought that the next test would be a. Zs 

 

[ProDave]

The op did give a Ze so you would have thought that the next test would be a. Zs 

He said

Disconnected the Main Earth in the CU to do a Ze test 0.20 ohms - It is a  TNCS supply 


Just happened to probe between main disconnected Earth and the Earth bar in the CU  I got   80 volts !!!

Clearly says 80V on the CU earth bar with the TNC-S earth disconnected.   

I asked for further clarification. Where did he put his earth probe for that Ze test?  On the TNC-S earth cable? Or on the CU earth bar?

 

[johnnyuhtar]

Hi Thanks for all your reply's

Ok - the original Ze test was done with the test lead clamped direct to the MAIN EARTH wire at the CU and the other leads on the LIVE / PHASE  = 0.20 ohms

The Gas & the Water bonding were still connected

I have noticed that the 80 volts ONLY appears when the MAIN EARTH is disconnected

If the Main Earth is connected and I then disconnect the Gas Or Water bond and then check either of them to each other or the earth bar in the CU (Main Earth connected) I get 0v as expected

I have checked the Water Bonding wire for continuity as It is not visible all OK

The Gas bond can be seen on its full length but I checked it anyway all OK  

If the MAIN EARTH wire is DISCONNECTED and I then have a meter between that (Main Earth) nd the earth bar in the CU then switch off the main switch the 80v is gone

If I switch the MCB's off one at a time the voltage goes down slowly but NO particular MCB is causing the full 80v

That's why I was wanting to check the main CABLE coming to the TNCS head to see if it was a break in the earth or Neutral before it gets to the main head and is separated to the earth block in the meter cupboard

I read a post somewhere saying that some testers will not do a loop/ rcd test with certain makes of RCD … So I tested the RCD Direct  Green& Black leads on incoming Live/Phase Red lead on Out going Neutral and It tested NO problem !!!!

Could it be the TESTER not compatible ????

Then again I want to know why ?how there is 80v when the main earth is disconnected that does not seem right

Like I said they are building a new housing estate over the road and further along they are altering the road layout and I noticed that they are messing on with the electrical  cables as I had a nosy down the hole mainly to see how they join them back together

Thanks so far

 

[ProDave]

Well if you have disconnected the earth form the Cu then all the earth wires are floating so can float to any voltage with only a very small leakage.

And nit the gas and water bond are to plastic incoming services they won't give you much earth either.

 

[poni]

If the MAIN EARTH wire is DISCONNECTED and I then have a meter between that (Main Earth) nd the earth bar in the CU then switch off the main switch the 80v is gone

so if there is a fault it lays with your CU and the circuits within, because when it off the fault goes away 

and its like Dave says 

Well if you have disconnected the earth form the Cu then all the earth wires are floating so can float to any voltage with only a very small leakage

 

[Rob.]

Why are you powering up the installation with the main earth disconnected??

 

[Sharpend]

Why are you powering up the installation with the main earth disconnected??

That is the Question!! Looking forward to answer.

Locked to prevent resurrection (again)