Odd Symptoms.

[securespark]

Hi guys.

I have had to deal with two faults recently that rather stumped me, but both solutions turned out to be incredibly simple. I shall explain each problem:

Fault 1.

Incomer RCD (30mA) on PME tripping when any upstairs light is switched on, but never immediately. Sometimes it takes several minutes. Water ingress and cable damage eliminated as cause. EFLI's at fittings very low.

Fault 2.

"16th edition" split-load board with one 30mA RCD on TT supply with Ze of 58.5 Ohms.
Tested hall light EFLI: 0.99 Ohms.

Tested EFLI at lounge light switch: 20.72 Ohms.

Turned hall light on, Metrel now showing RP.
Turned hall light off, Metrel now showing correct polarity.

Lighting on incomer side, not RCD protected.

 

[Spark123]

Fault 1
Something to do with something electronicky in series shorting down to earth??

Fault 2
Live-earth fault on non RCD side? Odd reading caused by N-E fault in the circuit, switching causing L-E link via some DIY random wiring of the circuit??

 

[securespark]

Fault 1.

This was very odd. The customer had her son put up a light fitting where he had effectively wired the main ftg cpc (in the 3 core flex) to the cpc's in the T&E, then wired the cpc attached to the metal ceiling cup to the neutral....

Now, I did not break down the fitting to see what was what, but I would have thought this would have tripped the RCD immediately on powering anything on that circuit up....however, the reality was that it could stay on for anything up to 13 minutes.


Fault 2.

This was even odder, but again, very simple to rectify.

What immediately struck me as odd was the Zs on the hall light. Why was it so low, when the average reading elsewhere was late teens/early twenties?

I got a sub 1MOhm reading to earth on the lighting. Traced the short to the hall light itself.

LE fine, NE fine, LN down.

It transpired the MCB would pop, but not for a good few minutes each time, when the hall light was on. So I guess this was a short that, although reading almost zero with the megger, did not present a dead short circuit until 5-10 minutes had elapsed.

However, I presume that short was in circuit (high resistance SC?) as soon as the hall light was energised, hence the false RP reading by the Metrel.

Back to the 0.99 Zs reading. After taking down the fitting and replacing with a pendant, the Zs went up to match the others round the house.

With the old fitting connected, could the neutral have been "supplementing" the CPC, through the short circuit, thereby reducing the Zs from the higher reading to the lower one?

Definitely weird.....

 

[mattylad]

Did the customer say at the start that they had recently fitted a light fitting themselves?

I'd have thought that it would have been the first thing mentioned
but I bet it wasn't.

 

[securespark]

No, did they heck.

Before I began fault-finding, I asked (and always ask) the questions, have you had any work done, decorating putting up pictures, drilling putting up light fittings, changing switches etc... and she said no.

I then asked her how long it had been tripping for. A few days, was the answer.

The conversation after the discovery went like,

"Somebody has wired up a light fitting incorrectly."

"Oh, that was my son. Funny he's done it loads of times and we've never had an issue before..."

[Conjures up image of many many light fittings with NE short on non- RCD protected circuits. ]

"So, how long has the tripping been happening?"

"About three months."

"And when did your son idiotically butcher your installation, I mean, put up the light fitting?"

"Oh, about three months ago....Oh!"

Give me strength.