Fluctuating IR reading.

[NDG Elecs]

Hi One and All.

Fault finding today. Narrowed it down to one leg of RFC between N&CPC.

What puzzled me was the fluctuation in the IR test. Normally I find the reading stabilises after a short period (5-20 seconds). Today the reading was constantly jumping between 20Mohms with an occasional flicker towards 0.00Mohms. The readings were up and down like a yo-yo with this intermittent jump to 0.00 every 10 seconds or so. If Lee Evans (manic comic from +15yrs ago) was an IR test result then this was it!!

I realise digital meters have their foibles and an anologue meter may have told a different story, but I am intrigued to discover any views or thoughts or knowledge on the above.

On this particular installation there are three points affected, all fed individually by a single T&E, so there is at least one hidden junction somewhere, possibly more. But that's an aside really.

Thanks!

Neil

 

[Richard Burns]

I might initially worry about your test meter and leads, but if this variation was only found on this circuit then I would be expecting some vibration at the point of failure allowing the fault to escalate at times or volatile conductive contamination that is flowing, evaporating or jumping from the applied voltage.
Possibly there is a failing capacitor that breaks down when fully charged.
Odd one certainly.

 

[Deleted member 9648]

I've had very similar readings before on faults, usually it has turned out to be a charred burnt section of cable due to mechanical damage, or charred terminations, both caused by long periods of arcing. Sometimes you can break it down into a 'proper' reading by zapping it with 1000v. Just be certain the affected part of the circuit is completely isolated from anything sensitive.

 

[Bob Geldoff1234]

Agree what the others have said but i would also be looking at the possibility of water getting into something.After the water has broken down the insulation then the arcing starts across the charred part.Are there any outdoor lights or sockets on the faulty circuit?

 

[SparkyChick]

I've had similar issues and I've normally found them to be puncture wounds in the insulation of the conductor being tested.

 

[NDG Elecs]

Just remembered what I forgot to put in the OP also; at 250V the reading stabilised and was clear (300Mohms ish). At 1000V the reading was like at 500V.

Thanks for replies so far. Pushed for time at minute but will respond later.

 

[Lucien Nunes]

Either a flashover between two bits of stuff (e.g. mangled conductors) very close but not quite touching, a capacitor likewise (maybe part of a delta suppressor cap in an appliance still connected) or perhaps some oddball like an earth-present indicating neon lamp in an extension (nasty things). There are possible failure modes of electronic appliances that might cause this but as you've presumably unplugged most or all, that is stretching the bounds of probability.