IET logo
 
IET
Decrease font size
Increase font size
Topic Title: Earth path
Topic Summary: shower fault
Created On: 03 October 2010 11:54 AM
Status: Post and Reply
Linear : Threading : Single : Branch
<< 1 2 3 Previous Next Last unread
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 03 October 2010 07:19 PM
User is offline View Users Profile Print this message



alancapon

Posts: 5746
Joined: 27 December 2005

Originally posted by: ant1uk
. . . and because his 5x test was over 40ms . . .

You could be in trouble for that!

Regards,

Alan.
 03 October 2010 07:36 PM
User is offline View Users Profile Print this message



ant1uk

Posts: 2376
Joined: 22 June 2010

Originally posted by: alancapon

Originally posted by: ant1uk

. . . and because his 5x test was over 40ms . . .


You could be in trouble for that!



Regards,



Alan.


Not sure I get you there?

Regards
 03 October 2010 07:40 PM
User is offline View Users Profile Print this message



alancapon

Posts: 5746
Joined: 27 December 2005

Originally posted by: ant1uk
Not sure I get you there?

Writing ". . . because her 5x test . . . " would be more appropriate.

Regards,

Alan.
 03 October 2010 08:22 PM
User is offline View Users Profile Print this message



Zs

Posts: 2787
Joined: 20 July 2006

Ant, I don't mind assumptions that I am a man. it is par for the course. But believe me, I'm so not.

Geoff, well the first lot of testing was voltstick and a deal of amazement, which I know you won't like. The Fluke one and not one of the crazy ones which you can make flash just by rubbing them up your jumper. Although they too have their uses in my tool kit. JP, the latest kewtech one is pretty impressive and the new Fluke super sensitive one is a beauty with two different colours of lights to tell you all sorts of interesting things. However, I digress.

Once I had the shower unit apart and all the water had flooded out over the boots (grey, the pink ones are the toe-tectors, come on keep up). I put a much loved old test lamp between one of the taps and the earth of the cable incoming to the shower. The lamp lit fully. Bear in mind that it was soaking so I don't think any circuit rules apply. The only position it did not light up in was between tap and neutral but even then it gave off a really feint low glow just in the filament of the little lamp. Not much on the actual shower hose but plenty on the copper pipes inside the shower unit. Look, say what you will, I know it is old fashioned but as a starting point I think an old style test lamp/bell is superb. Lights and bells are so helpful at that stage, particularly in front of a customer who isn't remotely interested in milliAmps that he cannot see. Went downstairs to check the bonding and earthing in more detail. Decided on a danger situation and didn't go much further. As such I don't have a voltage reading for you.

For the sake of checking it for my letter and wondering why it hadn't gone off, I then tested the 30ma RCD which was for the shower only, using the Fluke multi and crocs, and using the Geoff Blackwell x 5 as a first test approach. I used the BOD method of not touching the test button before running this test. 60.6 mS. Now, any doubting Thomii out there can comment on the fact that the RCD might well work now under a x1 test but my view is very strongly that if it has stuck once then it could stick again. As such it must go in the bin. I would expect a 30ma @ x 5 to go off in 11-18 mS so anything much higher is not a welcome sight.

It isn't perfect. But I say not bad for 6 o'clock on a Friday evening while the clients were busy getting ready to go out and do a gig (Jazz musicians I am told...peavy amps all over the place). They are of the school that says 'it has been working perfectly for years' and seemingly good people. They have agreed to everything and the supply authority were coming out on Saturday morning. I'm so busy I've had to pass the work on which is a shame, I'd like to have done that one for them. But will find out what the supply guys did tomorrow.

But the science of this problem is annoyingly beyond me. I've never been good at earthing and bonding science. You have been helpful of course.

Zs
 03 October 2010 08:30 PM
User is offline View Users Profile Print this message



ant1uk

Posts: 2376
Joined: 22 June 2010

Hi Zs

I think I was been abit slow there now I understand what Alan was saying.

Regards
 03 October 2010 08:30 PM
User is offline View Users Profile Print this message


Avatar for MrOther.
MrOther

Posts: 530
Joined: 08 June 2010

Zs - nout wrong ith the old tests as long as they are used a means of indication that their is a fault, rather then proof there is a fault - proof requires more solid facts. As you point out - and I really like the idea and will incorporate it in my next job - the Customer won't understand the facts nor interested in facts, but are happy to see/hear a simple indication of a fault.
 03 October 2010 08:32 PM
User is offline View Users Profile Print this message


Avatar for GeoffBlackwell.
GeoffBlackwell

Posts: 3492
Joined: 18 January 2003

Time for round 2 - BTW I am on a slow connection via a portable Wifi in a hotel in West Sussex - 11 lucky people have the pleasure of my company for three days on a 17th edition course I am delivering .

I hope the point is getting across that we need to establish if this really is dangerous. We can only be sure if we have reliable test data.

I would probably favour isolating the installation and carrying out an IR test on the DB to see if any faults are revealed. If so trace and rectify - and then deal with the earthing and bonding issues.

If not you may wish to retest your pipe work using a 'proper' voltage test instrument with two leads, and preferably with a relatively low input impedance - Fluke do one that you can 'dump down' by reducing its input impedance. If you measure a substantial voltage with an instrument like this you have a serious problem. Test between the pipes and a source of earth potential - you could use Alan's famous screw driver in the garden test for that.

The tests described above should clarify if this is serious or trivial.

Zs I have just seen your post above and I have not read it all yet so some of the above may no longer be applicable

Regards

Geoff Blackwell
 03 October 2010 08:47 PM
User is offline View Users Profile Print this message


Avatar for GeoffBlackwell.
GeoffBlackwell

Posts: 3492
Joined: 18 January 2003

I put a much loved old test lamp between one of the taps and the earth of the cable incoming to the shower. The lamp lit fully.


This is beginning to sound like a serious problem. If a test lamp is lighting you could have a 'real' fault that could deliver a lethal shock - if so it should be traced without delay. I recommend that the customer be advised that the installation might be dangerous and that it should be isolated for further investigation immediately. You might be able to just isolate the 'faulty' circuit.

It may all come to nothing but there could be a real risk.

Regards

Geoff Blackwell
 03 October 2010 08:57 PM
User is offline View Users Profile Print this message



alancapon

Posts: 5746
Joined: 27 December 2005

Originally posted by: Zs
. . . I put a much loved old test lamp between one of the taps and the earth of the cable incoming to the shower. The lamp lit fully. Bear in mind that it was soaking so I don't think any circuit rules apply. The only position it did not light up in was between tap and neutral but even then it gave off a really feint low glow just in the filament of the little lamp. . .

I agree with Geoff. If you can actually light a real test lamp (I am assuming a "Drummond" or something similar), then you have not only a missing earth, you have a fairly good connection between phase and earth - more than just water! That is a great concern and needs to be located and fixed or the faulty circuit disconnected.

Regards,

Alan.
 03 October 2010 09:03 PM
User is offline View Users Profile Print this message



John Peckham

Posts: 7375
Joined: 23 April 2005

I have a couple of the old drummond type test lamps. They are good at proving you have a neutral and more than just a few milli-amps of leakage. They will not indicate lower voltages that may be lethal.

GB

As staed above voltsticks come in various sensitivities. The Fluke Volt-Alert comes in 90-1000V and 200-1000V.

Zs

What is this magical VoltStick you speak of. I cant't have anyone having better kit than me.

-------------------------
John Peckham

http://www.astutetechnicalservices.co.uk/
 03 October 2010 09:17 PM
User is offline View Users Profile Print this message


Avatar for GeoffBlackwell.
GeoffBlackwell

Posts: 3492
Joined: 18 January 2003

John
I don't want to hijack this post as Zs may have a serious problem here.

Suffice to say that voltage sensitivity is not the main issue - input impedance is much more relevant. How many people spend ages chasing 'phantom' voltages because they are using instruments with high input impedance. If a low input impedance instrument indicates that a voltage is present then it really is, and it may be dangerous.

Regards

Geoff Blackwell
 03 October 2010 10:04 PM
User is offline View Users Profile Print this message


Avatar for sparkingchip.
sparkingchip

Posts: 5823
Joined: 18 January 2003

I'm on the Geoff Blackwell team.

I would also be doing a insulation test on the whole installation with everything still connected, then removing circuits whilst still under to narrow down the area of the fault.

Also I would want to do a loop test on the shower circuit before using it as a reference point for testing to extraneous metalwork or anything else, the lamp or what ever else you connect across it will not light up if both the taps and cpc are at mains voltage.

Waving a voltstick around may indicate a problem, but it is not going to identify it.

Andy
 03 October 2010 10:29 PM
User is offline View Users Profile Print this message



John Peckham

Posts: 7375
Joined: 23 April 2005

If a test lamp was lighting up between the CPC and the tap as GB says that is a very real problem and a danger to the occupiers.

Zs I hope you isolated the shower?? and anything else that dispalyed similar test results.

-------------------------
John Peckham

http://www.astutetechnicalservices.co.uk/
 03 October 2010 11:38 PM
User is offline View Users Profile Print this message


Avatar for tonyericsson.
tonyericsson

Posts: 485
Joined: 24 August 2010

Zs
In answering your OP question earlier, I suggested that had the earthing been OK but no main protective bonding, that you would probably find the RCD would have tripped.

You tested it at x 5 but the ELI is such that the unit wont trip satisfactorily....ie > 1667 ohm

In your shoes I would isolate, then R1 + R2 each circuit, if you skip straight to IR without R1+R2 you can find all circuits pass the IR because the cpc goes half a meter out of the consumer unit before it stops, so you could end up re-energising the fault.

How you deal with the earthing conductor, well a call to the DNO for an rapid response would be OK, Alan is your man there.

Tony
 03 October 2010 11:49 PM
User is offline View Users Profile Print this message



Zs

Posts: 2787
Joined: 20 July 2006

The new one is LVD2 Volt Light. Won't be any good for finding switch wires because it detects from a distance. But natty nonetheless.

To put your minds at a degree of rest, and thank you for your concern. The couple did realise that this is serious and told me they would not be home, because of work. Whether or not they were I don't know but no read receipt from an email I sent them. EDF(?) were due there on Saturday morning and another electrician is in there tomorrow to take out the DBs, replace with new and replace the shower circuit. And to add Gas and water bonding. As I said, I'd love to have done this job myself but couldn't let it wait.

I'll phone the electrician tomorrow and ask what EDF said or did.
Thank you
Zs
 04 October 2010 06:52 AM
User is offline View Users Profile Print this message


Avatar for GeoffBlackwell.
GeoffBlackwell

Posts: 3492
Joined: 18 January 2003

Zs - good to hear that you have initiated actions to resolve the situation.

So now that the dust can settle we can make some more general comments about this sort of situation.

1) A report of electric shock has to be taken seriously, but we must bear in mind that it could be due to something that is not necessarily inherently dangerous. Our first task is to try and establish the degree of danger. Make no mistake there could be a potentially lethal fault and it might kill you .

2) An electric shock is caused by current passing through the body and for this to happen there would have to be a potential difference (voltage) between two points that the body makes contact with. We need to be careful about the word contact here as it might not mean actually touching - 'contact' could be due to, say, capacitive coupling. This is how the dreaded volt stick works .

3) Good information is key so we need to use methods that produce reliable, unambiguous results to help us evaluate the situation. (That's the volt stick in the skip then ).

The first consideration is the safety of all those present - including ourselves. This might mean that the best approach is to isolate the area or, in small installations, the whole installation. To decide this we need to attempt to evaluate the risk - always bearing in mind that we may have misread the situation - so proceed with caution. Let's call this first step by some catchy name - what about 'Risk Assessment' .

So that is the first step - Risk Assessment.

This is going to involve determining the presence or otherwise in the area of a potential difference. This by its name indicates that it is between at least two points within the area.

Now before we rush in with volt meter in hand we need to be sure that we have a good chance of being able to rush out again . If there is any doubt - isolate and carry out some 'dead' testing first.

So hopefully that will start things off - more later if no one beats me to it .

Regards

Geoff Blackwell
 04 October 2010 10:48 PM
User is offline View Users Profile Print this message


Avatar for sparkingchip.
sparkingchip

Posts: 5823
Joined: 18 January 2003

Going off the plot as we do, several years ago I did a job for a customer involved in controlling motors on sites such as Blackpool Pleasure Beach fun fair who said he kept a AVO in the boot of his car as it can give much more valid results than newer digital meters.

Andy
 05 October 2010 06:32 AM
User is offline View Users Profile Print this message


Avatar for GeoffBlackwell.
GeoffBlackwell

Posts: 3492
Joined: 18 January 2003

I used to work with two FE college workshop technicians who didn't like the AVO Model 8 as its internal construction made it harder to repair than the Model 7 .

So let's not get too nostalgic or we will soon be back in the 'dark ages' .

Modern digital instruments are generally very good - they have more than adequate basic accuracy and the better makes are robust, reliable, and safe to use as long as we are not too close to some big transformer .

They generally have high input impedance which means that they only draw a small current from the circuit when measuring voltage. This is a great advantage in electronic servicing and similar applications but it can be a disadvantage when the instrument is used on electrical installations.

High input impedance ensures that the instrument has minimum impact on the circuit under test - but when dealing with power circuits this is not an issue as even the old AVOs did not adversely affect power circuits .

The down side is because they don't 'load' the circuit they will indicate the presence of almost any voltage including those that have been induced by the local radio station .

Many of us have been caught out by this and have ended up chasing our tails trying to find the source of some stray voltage. Loading the circuit will usually eliminate this problem. You could do this by placing a resistor across the voltmeter input terminal (make sure it's resistance is in the thousands of ohms range, not just ohms ). However, this is tricky to do.

There are instruments available that have a 'Lo Z' voltage range to overcome this very problem - the Fluke 117 for example, this even has a 'Volt Alert' - non contact voltage detector built in - for those who haven't given up their dummies yet .

Regards

Geoff Blackwell
 05 October 2010 09:50 PM
User is offline View Users Profile Print this message


Avatar for MrOther.
MrOther

Posts: 530
Joined: 08 June 2010

Originally posted by: GeoffBlackwell

I used to work with two FE college workshop technicians who didn't like the AVO Model 8 as its internal construction made it harder to repair than the Model 7 .



So let's not get too nostalgic or we will soon be back in the 'dark ages' .



Modern digital instruments are generally very good - they have more than adequate basic accuracy and the better makes are robust, reliable, and safe to use as long as we are not too close to some big transformer .



They generally have high input impedance which means that they only draw a small current from the circuit when measuring voltage. This is a great advantage in electronic servicing and similar applications but it can be a disadvantage when the instrument is used on electrical installations.



High input impedance ensures that the instrument has minimum impact on the circuit under test - but when dealing with power circuits this is not an issue as even the old AVOs did not adversely affect power circuits .



The down side is because they don't 'load' the circuit they will indicate the presence of almost any voltage including those that have been induced by the local radio station .



Many of us have been caught out by this and have ended up chasing our tails trying to find the source of some stray voltage. Loading the circuit will usually eliminate this problem. You could do this by placing a resistor across the voltmeter input terminal (make sure it's resistance is in the thousands of ohms range, not just ohms ). However, this is tricky to do.



There are instruments available that have a 'Lo Z' voltage range to overcome this very problem - the Fluke 117 for example, this even has a 'Volt Alert' - non contact voltage detector built in - for those who haven't given up their dummies yet .



Regards



Geoff Blackwell


Geoff I haven't done my 2391 and I do no testing at my firm, so was interested by what you said about high input impedance meaning that the instrument can pick induced voltages from other sources - is there any offical paperwork to advice on your resistor idea because it would be handy out in the field.

PS - would say, a normal T1000 Fluke meter suffer the same problem?
 06 October 2010 05:33 AM
User is offline View Users Profile Print this message


Avatar for GeoffBlackwell.
GeoffBlackwell

Posts: 3492
Joined: 18 January 2003

Fluke don't list a T1000 - they do have the T100 and the T5-1000, so I am not certain which instrument you are referring to.

I have a T100 but I don't often use it. I only got it to impress the 'idiots' whoops sorry - I mean nice well meaning safety expects - who seem to think that everyone is incapable of selecting the correct range on a multi-function instrument when measuring voltage .

I suspect that both the T100 and T5-1000 have a relatively low input impedances but you would have to confirm that with Fluke.

The 'resistor idea' is just a method of loading a suspected voltage source to see it it can actually deliver a substantial current.

You can easily make one up but it could be connected across 400 volts or more so I am not going to publish a 'design' - your on your own there .

Regards

Geoff Blackwell
IET » Wiring and the regulations » Earth path

<< 1 2 3 Previous Next Last unread
Topic Tools Topic Tools
Statistics

See Also:



FuseTalk Standard Edition v3.2 - © 1999-2014 FuseTalk Inc. All rights reserved.