Failure on IR

Is it really a "fail"?

Table 64 gives a minimum resistance of 1.0 MΩ but 643.3.2 does go on to say, "Insulation resistance values are usually much higher than those of Table 64. When measured values show evident differences between circuits, further investigation to identify the reasons is required.


Personally, I would say that 500+ MΩ is fine.

Are you testing L+N to E or also L-N or L-E  NE separately.

I suggest that unless you can remove all loads (including that light) only L+N to E is important, not th others and the fail limit is 1 meg, though much less than a few and I;d be wondering. I'd also stay off the 1000V setting unless you are very sure there is nothing that can be damaged by it. Electronics in particular does not like over-voltages.

Unless there is something else that makes it look dodgy - are the walls damp and peeling ? Signs of rat and mouse damage ?  I'd not get too hung up. Is there anywhere you can look under the floor  -  a loose board near where the water main comes in is a classic if you are lucky.

Another source of random leakages are the ceramic insulation in some designs of heaters and cookers - nothing like that left plugged in anywhere  - spur to a towel rail or something ? Again not a fail, just an observation.

Thanks for the input guys, Yeah your right, Im probably just getting hung up as I have never really seen such a slow ramp up like this.


Testing was done L+E and L+N, both showed same characteristics albeit L+E start much lower as it ramped up.

All the loads were removed. Outside lights disconnected at switch. all sockets unpluged.

Unfortunately there is laminate over the top of the floorboard so no where I can inspect. No signs of damp though.

Daft question but have you checked the batteries in your tester? MFTs do some strange things when the voltage starts to drop off

Yup, Had to change them as the fluke wouldnt allow me to zero the leads. I think it has to be above 5.9v to zero leads. Plus I carried out IR tests on other circuits after I tested the lounge one.

This subject has come up before here.

The general opinion is that there is a tiny amount of water somewhere in the circuit, zap it with 500 volts, and a tiny part of it evaporates, giving you a perfect >500M ohms reading. If you go back a hour later, you again get the slightly low reading, which gradually rises up to >500M again.

Something like a newly plastered wall can give a small amount of condensation in a socket or back box.

I remember seeing it talked about on the old forums and the mention of moisture. That why it was going through my mind that that could be the reason. 


no new plastering works, no sign of damp on walls. 

With cabling being under the floor boards of ground floor, Cables may have not have been attached to joists properly and resting on the ground (probably old bricks and crap from when the house was built). Looks a old house, late 60s, so might have no damp proof on the ground under the floor board. If a cable is resting on cold damp floor, could that introduce moisture in to the system.


I don’t see anywhere in the regs that say you can’t route cables in the void under floors at ground Level, but is there not a heightened risk under there of rodent damage or moisture?

Sounds to me as if there could be capacitance in the circuit. The reading rises as the capacitance charges up until it reaches a stable reading when the capacitance is fully charged. This is most often seen on very long cable runs due to the cable capacitance, but I have never heard of it happening in a domestic situation as there is not enough cable to cause a problem, so if this is the problem there must be some other capacitance introduced into the circuit.

HI Rusty H


When you say, "Running IR tests, no issues at 250v, but then at 500v it ramps from 280 m ohms till it reaches 500, then at 1000 same scenario, starts at 280 m ohms and ramps up but never really makes it to the full 1000 m ohms. It seems worse on the L-E, but does also fail on Lives."


Do you mean that the IR at 500 Volts starts at 280 MEG Ohm and rising? 


I'd not bother to test any further - there's no fault to find and there's nothing wrong with this circuit I'd say. Something is connected somewhere and the electronics cause this or there's some capacitance or something causing the rise. After 200 Meg Ohms there not much point in measuring Resistance. In an installed existing circuit I'd only do L&N together to earth too. 

How do you think you can possibly find a nonexistent fault?


The circuit passes the insulation test, so there is not a fault to find.