Insulation resistance testing

[nicsin02]

Hi guys, I obviously know how to conduct a IR test but I'm just looking for a few tips, I was doing a lighting d\b today and obviously I never did a L-N test as it was a massive building and not feasible to disconnect all lamps. It was a triple pole +N mainswitch so I isolated the board, it is just a report I was doing so no need to get deep into the problem, I started by testing between live to earth and considering it was a 12-way board (36 circuits) only 15 circuits were clear and the rest were either a dead short or around the 0.5M ohms .

1 . Is it more than likely a neutral to earth fault ( voltage up line through lamp and back down neutral ) rather than L to E as a L to E would trip the mcb.

2. Does a lamp effect L to E or N to E readings

3. If there is a fault to neutral on a lighting circuit does this have an effect on all circuits until this neutral can be traced and disconnected from neutral bar ( e.g if I find a problem with 1 circuit will it clear the other16)

4. I'm only new to testing so any advice and routine for finding a fault doing a IR test would be gratefully appreciated



THANKS GUYS

 

[johnboy6083]

I always test at 250 before doing a higher voltage test. This usually gives an indication of loads that may be connected. With lighting, I would always test L&N to E. even if I thought I'd disconnected everything, experience tells me that I'm bound to miss something.
make sure all switches are on, and operate 2way switches and test again, to ensure both strappers are tested.

 

[telectrix]

most likely a N-E fault on 1 or more circuits. to identify it, you'll need to trake the neutrals out of the bar 1 by 1.

 

[MDJ]

please don't feel insulted, but was it a TNC-S/ PME system? if so did you ensure all poles were isolated because as you are probably aware you will get a reading between neutral and earth unless all poles are broken.

 

[nicsin02]

please don't feel insulted, but was it a TNC-S/ PME system? if so did you ensure all poles were isolated because as you are probably aware you will get a reading between neutral and earth unless all poles are broken.

Yes m8 stated earlier 3-pole + neutral mainswitch, cheers

 

[sparkymalarky]

I agree with both telectrix and johnboy. I would always start IR testing at 250v unless you know the installation very well and you can be assured all lamps removed etc.

A great tip is to bring an apprentice to flick light switches on and off so you can stand at the board and IR to your hearts content!

 

[Brightspark2]

I think it will be the load giving you false readings. When I'm testing large lighting DB's I test each circuit seperatly rather than a group IR. Take the neutral out the bar and try again. L+N to E.

Ive had contactors at high level with all the neutrals going into one block causing me head scratching before....

 

[Smart Sparks]

Hi all,

I was testing an old db today, a per inspection and test before replacing db. I tested all circuits between L+N & E, I did try to disconnect everything I could find and kept getting strange readings. I then took all lamps out and tested again still no change. I then looked in every cupboard and found a fridge plugged in and cooker switch on. Removed all the loads and opened a switch to a fluorescent switch start light fitting and hey presto all was fine. So my query is with L+N connected together why was the load being picked up, thought this test bypassed this stuff.

Also although l-n and l-e were perfect n-e were around 100Meg ? Again any clues. As board is tiny I only removed live conductors and left earths and neutrals in their blocks?.

Reason for testing was trying to ascertain if db would go in without having to re-wire (1971 install with cpc).

Thanks in advance for your advice.

 

[HandySparks]

Removed all the loads and opened a switch to a fluorescent switch start light fitting and hey presto all was fine. So my query is with L+N connected together why was the load being picked up, thought this test bypassed this stuff.

Also although l-n and l-e were perfect n-e were around 100Meg ?

Your description suggests a L to E or N to E fault in one of the appliances or maybe the light fitting.

100M is close enough to "perfect" not to bother about.

 

[spark 68]

Some Items have filters and/or components to Earth from both L and N, it is wise to conduct this test @ 250V first, a so called 'soft test' , this usually shows anything that may still be connected, forgotten or otherwise, such as alarm panels, central heating controllers and aerial amplifiers (hidden in the loft) and the like.

These are usually things that can be damaged by the 500V test, so this may save you from damaging (usually expensive to replace items), a sort of look before you leap type test.