Disconnected circuit appears live

We've just had the EICR done and they picked up a circuit that appeared live on the tester, but when the wiring was traced out the supposedly live circuit was not electrically connected to any other circuit. The only thing which I can think of is some kind of capacitive or electromagnetic pickup from adjacent live circuits. The wiring of the problem circuit ran alongside a couple of live SWA cables for about 1.5M and also crossed at right angles to a live T&E cable, apart from that it was not close to any other wiring.
Any thoughts on this, has anyone else come across dead circuits appearing live. The sparky who reported this is scratching his head, as he checked with a colleague using a different tester and the circuit appeared live. Problem has now gone away as the redundant wiring has been stripped out, but it would be nice to know how his might have happened

  • yes this is common, as is floating cores in a multi-core bundle looking live when not really connected. Assume something in the range of 50-100 pF/m for cables that are run in contact or very close proximity. On a volt meter with a high input impedance say 10 million ohms plus - which is common for a modern digital device, then the capacitrance bewen the cable under test and a live one forms the upper part of a voltage divider, and the meter forms the lower. The give-away is that the voltage falls away if any serious load is added in shunt with the meter, and that the voltages do not add up well - so it may read 200V from floating wire to earth, 230 from L to earth, and yet only a few volts from live to the floating wire.
    Mike

    PS to save you a sum, at 50Hz, 1uF is about 3jk ohms and 1000pF is about 3j M ohms, so 100pF, that meter or two length of wire,  could be about j30megohms - Taking a DVM of say 10 M ohms impedance you may expect to see 100V or so from no-where - enough to trigger a warning on some  test kit.

    If you  have a moment and fancy a self educational laugh, wire up a few metres length of scrap mains  flex or T+E and liven up one core (insulate the ends of course) leaving the others floating. Then measure the voltage from the floating cores to ground with different meters you posses. (wire the meter and stand back before plugging in) you may be surprised at just how much some meters actually read.

    Mike.

  • Agreeing with everything Mike said - this is a very common mis-interpretation of what a high impedance DVM is telling you.

    If in any real doubt a good test is to connect this 'live' circuit to a test load - a 60W filament lamp is ideal (NOT an LED lamp!). You'll almost certainly find that the lamp will not light & your phantom voltage will have just as mysteriously vanished!

  • Hi Mike,

    really appreciate the time and trouble you have taken to give such a helpful and detailed answer

    Kind regards,

    Peter