Rules and Guidelines
Want to participate in the discussions? Please read our Community Rules and Guidelines. Need some help? Visit Support and Answers.

  • State Not Answered
  • Locked Locked
  • Replies 33 replies
  • Subscribers 33 subscribers
  • Views 14645 views
  • Users 0 members are here
Options
You may also be interested in
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

EICR and IR Testing

Former Community Member
Former Community Member
Oh the bane of IR testing on EICRs

Now its just impracticle to disconnect all loads to carry out IR testing, I think everyone can agree on that. And that Phase and Neutral are connected then tested to earth.


This is what I do, and then test @250V as to avoid damaging any equipment in the installation. And unless the IR value was <.5 Meg, would not bother me. What BS7671 states, but BS7671 is maybe somewhat unclear in that if that voltage (250V) was used then .5meg value should used, but that value should only be used on SELV, and PELV. Now experience also tells me that even if you had a return 0.0, and then did Kohms by using the ohms setting, even values of 100Kohms are fine. (Actually can be much lower than this) and we are pretty clear that its not the cable reurning these values.


So question is when doing an EICR @ 250V, at what point would you recommend an FI? For me the value would have to be <.5Meg
  • 0
    It rather depends what failure mode you are expecting, actually neither 230V nor 500V will 'jump' very far without assistance - much like a welding rod, contact has to be made either mechanically or by a signifiant over-voltage, and then an arc can be drawn out to a volume of hot gas limited only by the available power. Usually when this finally goes out the conductors have burnt back well beyond any re-striking distance. Equally for finding nails in the wire, trapped wires in backboxes or similar, a battery and bell set would probably do.

    I agree a 500V test will find a few things a 250V test will not, and a 1000V test will find a few more.

    In some ways the standard 500V test looks a bit feeble, given the overvoltage  categories defined in IEC 61010-1

    1500V for Cat I, (operation with protective low voltage), battery-operated devices, car electrics)

    2500V for Cat II  (household appliances, portable electrical appliances) and

    4000V for Cat III (loads with direct fixed connection, distribution, fixed installation appliances in the distribution system)

    Perhaps we should be testing at 2500V on new work and on retests when we are sure that all loads really are removed.
  • 0
    The old textbooks used to quote the example of uninsulated live parts only being some fraction of a hair's breadth apart - so the test voltage was chosen to be more demanding than normal service to reveal such faults before the 'bang test' did - the worry being that lower test voltages may not arc across small gaps and so give false pass results.


    Given that 230V nominal can be anything up to 253V and being a.c. those are r.m.s. values, the peak voltage in normal service is going to be something like √2 higher - say about 360V - a 250V d.c. test alone perhaps isn't that reassuring.


    Personally I do a 250V test first - just in case there's something still connected that shouldn't be (which usually isn't an issue with L+N - PE tests) - and if that shows reasonably clear, then try the proper 500V test.


      - Andy.
  • 0
    I suggest you are saying (and  I agree) a single figure is always a compromise, and rather depends what issues we are trying to detect.

    A small flat wired in PVC  where we can be sure no wiring goes outdoors should pass a higher limit.

    Large rambling installations in 'pyro' wire such as are often found in churches and other old public buildings, may well read low on the megohms test, and yet have no real fault - it is just a feature of mineral insulated cables. Indeed if enthusiastic one can test the circuit in sections and find nothing.

    Arguably looming problems with failing equipment or the wrong kind of load, that may  soon give tripping issues with RCDs are better found with a clamp meter with the system energised.


    What is of interest is the sort of fault that if it were all in one place would be hot enough to cause a fire - and that is probably a dissipation of a watt or two - so even the 100k figure (2.3mA on 230v,  about half a watt dissipation) is no more heat than a small torch bulb of the kind not used in modern torches, and worse could be tolerated.


    Your approach may not be the official technique, I suspect it is just fine in reality, and a lot better than a row of 'lim' or ''n/a"

Community Rules & Guidelines