Insulation resistance test on a EICR/PIR

[richard1972]

insulation resistance test on a EICR/pir

have read in 'unite the union guide to inspect and test'

test carried out with L/N joined to Earth and no need to disconnect loads/vunerable devices

pic1 to test whole insulation, shows link between L/N (incoming) at main switch with supply isolated.

Q. if linked on incoming side of main switch and switch is off, then i wouldnt be testing pass the switch, would i? also if linking on incoming side then i am working live?

Pic2 to test individual circuit shows link between mcb and N bar

Q if linked in this way can other circuits affect reading because they will still be connected via N bar.

i have tried method in pic 2 and getting strange readings

also if possible would someone please explain the best way to test whole insulation for IR on a EICR/PIR with vunerable device/loads still connected (switches closed)
eg how/where they would make link etc

very new to this, so would appreciate help, thanks

 

[jamesBoater]

insulation resistance test is a dead test, are u sure you are understanding the picture properly?

 

[worzol99]

Plenty of posts on the forum regarding correct procedures to carry out ir tests, have a look

 

[baazab]

[h=2]Re: Insulation Resistance Testing (Whole board at the same time).[/h]

You wouldn't need to disconnect the supply cable, if the switch isolator is double pole.
You may have problems with the earth conductor, if the system is TN-C-S.
Either disconnect the earth conductor to the board you are testing, or the main earthing conductor for the installation.
Then isolate the double pole supply switch at the board you are testing, link between the busbar and the neutral bar using a short piece of cable with a croc clip at each end.
Connect one probe to either the busbar or neutral bar, and the other probe to the earth bar, then press the test button for however long your particular tester requires.
Obviously, if the reading is too low, then you will have to test the circuits individually.
I would disconnect the live conductors for the individual circuit, place them in a large connector block, with one probe in the other side, connect the other probe to the earth bar and test.​

 

[richard1972]

this is pic 1 as per my original post

test carried out with L/N joined to Earth and no need to disconnect loads/vunerable devices

pic1 to test whole insulation, shows link between L/N (incoming) at main switch with supply isolated.

Q. if linked on incoming side of main switch and switch is off, then i wouldnt be testing pass the switch, would i? also if linking on incoming side then i am working live?

 

[derek]

You do not link the supply side, you are effectively short circuiting the tails. Seems a bit dodgy to me but I assume they mean isolate the supply before the feed to the CU.

 

[jamesBoater]

your tester will not let you carry out insulation tests if a voltage is present, hence it is a dead test, the picture looks slightly misleading to me. read baazab, s reply that explains more.

 

[tom79bmth]

You do not link the supply side, you are effectively short circuiting the tails. Seems a bit dodgy to me but I assume they mean isolate the supply before the feed to the CU.

Well spotted
that would go with a bang

 

[richard1972]

pic shows probe on incoming side of main switch with same switch locked off and isolated, therfore i wouldnt be testing pass that switch

 

[jamesBoater]

probably best to test by using your knowledge and not misleading pictures, like one of the post says there are more threads on here regarding IR testing, have a read.

 

[richard1972]

reason i was asking is because pic2 in the same book, shows how to test individual circuit with equipment connected, by linking L&N test to E,

was testing my sub board in outbuilding,

main switch rcbo 63amp
1) mcb type b 32a - ring with 3 sockets (no load)
2) mcb type b 32a - ring with 8 sockets (no load)
3) mcb type b 6amp with 2 fluorescent lights (is this the problem)
4) mcb type b 6amp with 3 lights (GU10 4 lamps on)

now when i test (250v, just as precaution) circuit 4 with all conductors pulled out get >99M
but when i test via pic 2 description i get 45-50M

is it FLU affecting readings via N ?

how would you IR whole install on PIR

 

[Octopus]

View attachment 17645

this is pic 1 as per my original post

test carried out with L/N joined to Earth and no need to disconnect loads/vunerable devices

pic1 to test whole insulation, shows link between L/N (incoming) at main switch with supply isolated.

Q. if linked on incoming side of main switch and switch is off, then i wouldnt be testing pass the switch, would i? also if linking on incoming side then i am working live?

Is that PIC really from a professional publication? Try that and you'll get a big bang when the DNO fuse goes, then you will be testing live before the main switch, to neutral before the main switch to the CPC within the property and outside.

Not good!

 

[wiganspark]

do your self a favour mate and buy a copy of GN3 because following unite the unions little drawing youve just blown your fingers off and you wont be able to turn the page to see what test your going to do wrong next

 

[Andy78]

I have an old copy of the amicus 16th edition testing book. It shows a similar drawing but does state that the supply is to be isolated. Your drawing pictures a lock over the CU main switch which is misleading. The method shown can only be used if the supply tails to the CU are isolated.

 

[Andy78]

The previous drawing in the 16th book.

EDIT: I did rotate the pic before uploading but it somehow flipped again. Sorry for any whiplash suffered.

 

[oldtimer]

View attachment 17645

this is pic 1 as per my original post

test carried out with L/N joined to Earth and no need to disconnect loads/vunerable devices

pic1 to test whole insulation, shows link between L/N (incoming) at main switch with supply isolated.

Q. if linked on incoming side of main switch and switch is off, then i wouldnt be testing pass the switch, would i? also if linking on incoming side then i am working live?

This drawing is wrong as you say you are on the incoming side of the switch where 99% of the time is live due to no local DNO isolator being available in this case I have made up a jump lead with croc clips on it and yes the main switch should be off and even then you need to test the dead side to confirm it is safe

 

[richard1972]

thanks for everyones help with this, i was doubting what i knew, because i am new to testing and inspecting.