Ir tests

[oldtimer]

Did you split the ring and then retest as 2 radials ?

 

[WayneL]

Did you split the ring and then retest as 2 radials ?

For what purpose? - That goes against everything I've just said.

Sorry, just realised which part of the thread you're replying to - we're back on the fault finding:smiley2:

 

[oldtimer]

Well old school says split the ring in the middle and test no a big job but it pins it down so 12 sockets on a ring cut it to 6 then 3 now you know what area you are looking at plus if its the live that is showing a lower reading then its just the live you break at each socket

 

[1shortcircuit]

I've had a couple of issues with brand new MK sockets recently. IR's the circuit and results were lower than expected, split the circuit and pin point to faulty new socket:thumbsup

 

[Landril]

The same thing can apply to CPCs - you IR test L-CPC with the conductors disconnected from the board - >299...........but what if the 'line' conductor has been snagged on some earthed metalwork somewhere - again, it's not gonna show up.

Surely if the Line conductor was snagged on the CPC/Earthed metalwork there would be an extremely low reading due to the fact no insulation is between the conductors and they are forming a circuit!!!! Technicaly giving you a R1+R2 reading??

 

[DaveyD]

In the 2391 I went on the ... RCD (i.e. the buzz bar) without any issues.

That's interesting, I did that on the practical and got picked up on it (verbally slapped)

 

[stonejenson]

OK chaps, here's a new one - Was doing a little light job for a mate, and thought I'd do some full on condition report practice while I was there (Got my 2394&5 exams in a couple of weeks - eek!) Anyway, IR test from L&N to E gave a poor result, so I looked at each circuit separately and it was the ring... [Well, it might be a ring, or two radials, as it is in old 4mm going into an old 3036 30A, and end to end gave no result on L or N but r1+rn tests all round all seemed similar, so further investigation is required] anyway, it's the IR I was most concerned about, L+N to E, or L to N both give an absolute 0 result! I've just had my meter calibrated, so it's not that, and all other circuits seem fine.. everything works OK, but how is that possible if there seems to be a dead short across all conductors? Again, I need to check around some more, but I'm a bit thrown by this, a low result is iffy, but zero???
Ring or radials aside, I unplugged everything I could find, and there's nothing special installed anywhere, Zs results and polarity good at all sockets, any ideas?

 

[soulman]

To the original OP question, i would suspect it is the dampness of the plaster as this happened to me some time ago IR was >80Mohms retested when plaster was dry a week later IR >500Mohms.

 

[Engineer54]

We tend to conduct first fix IR and continuity tests as an area is completed. These are not official record tests, just for conformation that installed wiring is all fine and dandy, before moving on to the next area. lol!!

Dampness of walls and ceilings will always bring your IR values down, so official recording of IR values should whenever possible not be taken in such conditions, if unavoidable, make a note in the remarks. As others have stated, those values will rise drastically when the area has fully dried out.... All official testing of an electrical installation, should preferably be conducted to fully completed installations. Barring any temporary wiring disconnections for testing purposes.

 

[Guitarist]

It still remains that everything should be terminated before IR testing - so you couldn't go on the bare ends of stripped conductors.

Bearing in mind the regs have only just been ammended, you'ld think they'ld have altered the above regulation (sorry, 'guidance'), if it was a problem.

In the 2391 I went on the outgoing side of the RCD (i.e. the buzz bar) without any issues.

The problem you've got is, if you do as you(jimmyboy) and telectrix suggest, by testing each circuit seperately and disconnected from the CU, you could still have issues between circuits.

I'll just give you a very quick example so you know what I'm on about :

Conduit install, you IR test L-N on a socket radial, (disconnected from board) - 500v >299 M everything rosy.

Unknown to you, though, the 'line' conductor you've just tested is shorting out to the neutral of a lighting circuit (within the conduit).
It won't be picked up because you had the socket radial circuit's conductors out of the board.

The same thing can apply to CPCs - you IR test L-CPC with the conductors disconnected from the board - >299...........but what if the 'line' conductor has been snagged on some earthed metalwork somewhere - again, it's not gonna show up.

This is the reason you test with all conductors terminated.
If you have sensitive equipment in circuits, then, as the regs state, you have to compromise some of the testing, but the fact remains you should IR test as much of the installation in it's finished (terminated) condition as possible.

It's only the protective conductor which has to be connected when testing.
Took me a while to figure out what a buzz bar was. Assume you mean busbar?

 

[Jimmy Boy]

Lol@ Guitarist must be where the Bees nest..and I thought what you said re earths

J

 

[huntos33]

Some interesting views cheers guys i will re check ir readings when i return

 

[Rob Smith 643]

Just dropped into this one. Hope to resolve an odd L-E IR reading by using the aforementioned old school method of dividing into two, then four. Much better than disconnecting everything an running seperate leg IR tests. All other dead/live tests are fine on the circuit. Just the puzzling IR reading (L-N >1000, N-E >1000, L-E >725). All well within guidelines/regs but just a tad puzzling! I don't like unsolved puzzles :dizzy2:

 

[Guitarist]

Just dropped into this one. Hope to resolve an odd L-E IR reading by using the aforementioned old school method of dividing into two, then four. Much better than disconnecting everything an running seperate leg IR tests. All other dead/live tests are fine on the circuit. Just the puzzling IR reading (L-N >1000, N-E >1000, L-E >725). All well within guidelines/regs but just a tad puzzling! I don't like unsolved puzzles :dizzy2:

Nothing odd about that if the installation is old.

 

[Rob Smith 643]

The house was built in 1998? Not that old surely?