"22K Extraneous Part Test"

[Sidewinder]

Just been asked how to show someone how to do this, the "RIGHT" way!

Now I know how I do it, but, TBH, I'm not sure if it's right!!!

My MFT only goes to 2k hms  in Low R mode, my Fluke DMM, goes to M hms  in R mode, but the current generated does not meet the requirements for a tester under BS7671.

You can get higher resistance readings with an MFT on Insulation Resistance mode, depending on the applied voltage, now mine does 50, 100, 250, 500 & 1kV, with corresponding changes in the maximum resistance measured.

Now what is the "book" way, because I can't find it?  :C

 

[steptoe]

I do it as if Im doing an Ra test,

sort of,

but Im fairly certain that's probably not right either, now you have got me thinking about it,,,,,,,,,,,,,

 

[Sidewinder]

I do it as if Im doing an Ra test,

sort of,

but Im fairly certain that's probably not right either, now you have got me thinking about it,,,,,,,,,,,,,

Exactly!!!

Oops?

Dunno

I use a megger MIT 400 series 

So you do it as an IR test?

 

[Technician]

I stand to be corrected but is this the test that is conducted to prove

that there is no need to bond and the fault current is limited to no

more than 10mA  IF the result is greater than 22K?

So.....should the test be carried out with a test current up to 10mA?

I know this is a tall order for a lot of instruments but is this the only

way out?

Only a suggestion.

 

[Marvo]

My MFT only goes to 2k hms  in Low R mode, my Fluke DMM, goes to M hms  in R mode, but the current generated does not meet the requirements for a tester under BS7671.

Surely the test would need to be performed at the same voltage as the supply so you'd require a mega tester on 250v range.

The current flowing during the test would be governed by the resistance of the circuit up to the point where the tester capability is the limiting factor. I think the testing voltage would be a critical factor but I'm not sure why the testing current would have any bearing on the result seen. 

 

[apprentice87]

Ok....

So, what is the idea of bonding.... to create an equipotential zone..... Now, an extraneous part, is a Part liable to introduce a potential, generally earth potential.......

So there we are and we have an earth fault, and an exposed conductive part of whatever we are touching becomes live due to insulation failure, whatever.

We do not care, we do in fact, not even notice, because we are standing on an insulating floor covering. We touch the sink, we still do not care, as it is bonded to the MET [same as our conductive part], so even if the earth conductor is removed, we still do not care. No potential difference exists between anything we can touch....

Now sticking through the wall is an extraneous part, this IS at earth potential. We touch this and we have a terrible shock!!!! So how do we avoid this?? we bond extraneous parts to the MET. But how do we know they are extraneous???

We measure the resistance between the part in question and our MET. Everything else is connected to the MET so what we need to know is, if, in the event of a fault, can enough current flow between the MET and its associated "bits" [all the conductive parts etc] and the "extraneous" bit we are testing, to harm us.... [Remember this; in the event of a fault, if we are touching both a conductive part, and an "unbonded" extraneous part, WE form the earthing conductor!!!!]

Now, if we have an earth fault, for the duration of it, everything in the equipotential zone is going to be near, or at, line voltage. From this, we can see that it is line voltage that we need to apply to our "extraneous" part, in order to be able to measure how much current will flow, and, will this current be at a "safe" level. In practice, this would be impractical [eh!] so the nearest and best thing we can do, is to measure the resistance from the part to the MET using a tester that applies around about line voltage, so that would be about 230V, so 250V is used.

If we find the resistance is high enough, we can then be certain that not enough current will flow to harm us, and that we, from the point of "extraneousness!!" can forget about this object. What resistance do we need though??? Well, if depends on circumstances. In a DRY location, about 20K ohms is fine. In wet locations though, a person might be all wet, and wet [salty] skin conducts rather better than dry, so to compensate for this we would be looking at finding a resistance of more like 50K ohms

john....

 

[Marvo]

Now, if we have an earth fault, for the duration of it, everything in the equipotential zone is going to be near, or at, line voltage.

Can you explain what you mean by 'earth fault'?

 

[apprentice87]

Hi There,

A fault caused as a result of breakdown in insulation or whatever resulting in a L/E fault

From the regs;

"Earth fault current" An overcurrent resulting from a fault of negligible impedance between a line conductor and an exposed conductive part or a protective conductor. The regs of course talk about "negilgible impedance and all that, as what they are getting at for their purposes is PEFC We are more worried, [for the purposes of this] with the terrible shock we would get!!!!

john...

 

[apprentice87]

Just thought i would add that I know the above is obvious to you all, but i was bored so just thought i would "do a post!

john..

 

[soulman]

Apprentice87, that is explained really well.  You have explained this in much simpler terms than I&T book.

Cheers

 

[apprentice87]

Hi Soulman, Thanks for that!!!!!

john....

 

[Technician]

I have just turned something out about this, from an authoritative source.

"Determining if a part is extraneous, or just a piece of metal."

The text clearly states that an insulation tester, set to 500 volts

with range of Mohms and if the result is 0.02Mohm or greater,

then no supplementary bonding is required.

The text goes on to state that it should not be confused with a

continuity test and that the insulation tester is the correct instrument.

 

[Sidewinder]

Thanks Technician, can I ask where you found that please?

I did look but did not find this in GN3.

 

[Technician]

Sidewinder;  found it in a text by Chis Kitcher.

His Practical Guide to Inspection and testing.

I must confess when I read your original post

I really had to think about it and I realised that

the wrong test could be applied.  That was why

I looked everywhere at home and thankfully,

I found it.

 

[Sidewinder]

Thanks, shame there is no "real" "official" source!

 

[Technician]

Going on from Canoe......Page 100 of GN5 describes how the value of

22kohm arises. 

When Rb (resistance of the human body as determined by IEC data)

is added to 22k hms  (Rx) and divided into Uo (nominal voltage to earth) the

result should be less than Ib, the human body current that should not

be exceeded (taken as 10mA).

The note also goes on to identify why, if the test is done under "dry"

conditions it should be repeated under worst conditions if moisture

is to be expected and the "dry" test has produced an acceptable result.

I wonder if I should get out more..............

 

[Steve3948]

Maybe but thank you for sharing.

 

[Sidewinder]

[SIZE=11pt]Thanks for this, now I do what is basically an IR test.[/SIZE]

[SIZE=11pt]I started this thread because I was asked the "right" way to do this, and could not find anything in the IET documents etc. in the way of "formal" guidance on this test.[/SIZE]

[SIZE=11pt]I have GN's 1to 8, and several other documents, what I don't have are the text books suggested above.[/SIZE]

[SIZE=11pt]I don't however, doubt these sources.[/SIZE]

[SIZE=11pt]It would be nice for some sort of "official" test, or procedure.[/SIZE]

[SIZE=11pt]My Fluke DMM's (87, 87V, 189, 289, 867, 8060A) will measure up to 100's of MΩ but with low voltage & current behind the measurement, a couple of them also have low Siemens scales which are stupidly high resistances in reality, however, again measured with low voltage and current, so low that it is designed not turn on a PN junction.[/SIZE]

[SIZE=11pt]However, under BS7671, this sort of meter does not provide sufficient voltage and current for a low resistance test it seems, see also  BS EN 61557-1, -2 & -4.[/SIZE]

[SIZE=11pt]This is also covered in GN3 Para 4.3 where the meter must deliver 4 - 24 V & SCC of 200mA from 0.2 to 2Ω with a resolution of 0.01Ω.[/SIZE]

[SIZE=11pt]A typical say Fluke 289 DMM which does output 5V OCV, from 500Ω full scale upward, with a resolution which varies from 1mΩ @ 50Ω full scale to 0.1MΩ from 50MΩ upward, however the SCC across the resistance ranges are typically:[/SIZE]



[SIZE=11pt]I have been wondering about using something like this:[/SIZE]

[SIZE=11pt]http://isswww.co.uk/Earth-Testers/Fluke-Testers/Fluke-1630-Earth-Ground-Clamp-Meter/[/SIZE]

[SIZE=11pt]or this:[/SIZE]

[SIZE=11pt]http://isswww.co.uk/Earth-Testers/Chauvin-Arnoux/CA6415-Earth-Clamp/[/SIZE]

[SIZE=11pt]Which are two different meters that do the same job basically.[/SIZE]

[SIZE=11pt]I "think" that this:[/SIZE]

[SIZE=11pt]http://isswww.co.uk/Earth-Testers/Fluke-Testers/Fluke-1623-Kit-Earth-Ground-Tester/[/SIZE]

[SIZE=11pt]Can to a clamp on test also, which would get a similar result to the two above, plus this:[/SIZE]

[SIZE=11pt]http://isswww.co.uk/Earth-Testers/Megger/Megger-DET4TCR2-4-Terminal-Earth-Ground-Resistance-Tester-100034/[/SIZE]

[SIZE=11pt]unit also has a clamp on ability.[/SIZE]

[SIZE=11pt]Not sure on the last 4 web links if they have the resolution and accuracy.[/SIZE]

[SIZE=11pt]I feel perhaps some experiments are needed.[/SIZE]

[SIZE=11pt]Say on a known extraneous part such as an incoming copper water pipe, to do a clamp on test, say a "Z" type test, both non-tripping and high current, an IR test at 50, 100, 250 & 500 V, and 3 stake earth rod test, then on say something in between, say a temporary earth rod, with an R of a few hundred Ω repeating the same tests, then on something which would probably be not considered extraneous, say a radiator supplied via copper pipe directly but, with plastic upstream with perhaps a value of several if not 100’s of kΩ.[/SIZE]

[SIZE=11pt]I don’t think that all the tests would need to be done in a short space of time, but I do think that the suite of tests on each earthed “device” would need doing as quickly as possible, to eliminate the possibility of changes in the characteristics due to external influences, plus one could always return to the first test, say a high current Z type test to establish any changes in the characteristics.[/SIZE]

[SIZE=11pt]There is also the possibility of using a low resistance extraneous part & a pot that would carry the high current Z test, that way several values could be dialled in perhaps.[/SIZE]

[SIZE=11pt]As an aside:[/SIZE]

[SIZE=11pt]Today I had to test some insulated & isolated comms cables.[/SIZE]

[SIZE=11pt]I was told that there was a fault to true earth on one of the cores.[/SIZE]

[SIZE=11pt]Supply 11kV/400 DNO Tx about 5m away.[/SIZE]

[SIZE=11pt]This is how clients waste your time, the client & their phone system guy were convinced it was an earth fault on one of the cores[/SIZE]

[SIZE=11pt]Did an IR to the supply circuit cpc in the enclosure > 500MΩ (500V).  I knew the circuit Zs was around 0.9Ω.[/SIZE]

[SIZE=11pt]Whacked one of my earth tester stakes in & did a R between the circuit cpc & the spike < 900Ω. Fine for a quick check.[/SIZE]

[SIZE=11pt]Checked between the comms cables & the earth stake > 500MΩ (500V).[/SIZE]

[SIZE=11pt]The actual fault has yet to be found, but I think it is voltage coupling inducing noise on the comms cables, either from the HV to the Tx, to the comms cable near that, or within a short section of unscreened CAT5E inside the premises, or a long length of screened CAT5E within the premises, both running near power cables, used for the analogue comms.[/SIZE]

[SIZE=11pt]They are experiencing noise on some remote call pads, 3 wire Tx/Rx.[/SIZE]

[SIZE=11pt]I wonder if one of those clamp on jobbies would have given the same indication?[/SIZE]

[SIZE=11pt]Perhaps I will try this the next time I’m on site, but, mine is a CA6412, one of these:[/SIZE]

[SIZE=11pt]http://isswww.co.uk/Earth-Testers/Chauvin-Arnoux/CA6412-Earth-Clamp/[/SIZE]

[SIZE=11pt]I'd never thought to use it for the "extraneous part" test![/SIZE]

 

[Technician]

I am glad that we have a workable satisfactory outcome to this.

It is all very well in training and discussion describing the various

differences between Exposed and Extraneous-conductive-parts.

What do you about the smart guy who asks how it might, in any

way, be possible to determine the difference?

 

[Technician]

[SIZE=11pt] so low that it is designed not turn on a PN junction.[/SIZE]

Are these the widgits you can use for fault finding at board level

so that (you could) measure a base/emitter voltage on a transistor

and not run the risk of overbiasing it to extent that collector current

hits the roof and the device goes pop?