Low Ir Reading From A Lightbulb - Nuisance Rcd.

[brummydave]

Anyone had this before?
 
Called to investigate nuisance tripping - i'll get to the RCD stats later - and customer's main suspect was conservatory electrics "after all the recent storms and rain".
 
Checked out wiring, socket, outside lights and inside ceiling fan/light. Found no damage, short circuits or corrosion. Loop impedance was good (0.87Ohm) and IR readings nearly all >500MOhm. Until I came to the feed from light switches to the inside fan/light.

I got low IR readings of 2MOhm Neutral to Earth.

So I took down the dead-wasp-infested fan unit and tested the wiring again - IR all back up to >500. Suspect unit I thought. Removed the three 60w tungsten lamps (lightbulbs to all customers of course) and tested again. IR at >500.

Inserting one lamp gave an IR of about 18MOhm, having a second made it 11MOhm, and the third took it down to 5MOhm.

Replacing them all with new 60W tungsten lamps - ths customer has a shedload and doesn't like anything else - and my IR readings were all >500MOhm.

I was glad to get that mystery sorted, but never have I heard of this before in a lamp. After replacing heater elements where insulation has broken down I should've suspected the lamp earlier as it's effectively a heater!

Hopefully the nuisance tripping will have ceased. Although with RCD readings like these I think it may be hypersensitive rather than 'in-spec':

Time (0deg and 180deg)

x1: 11.7ms and 20.1ms

x5:  7.8ms and 17.6ms

Current:

15mA and 24mA

Comments and thoughts appreciated.

 

[SPECIAL LOCATION]

I got low IR readings of 2MOhm Neutral to Earth.

Comments and thoughts appreciated.

2Mohm is DOUBLE the minimum acceptable of IR...

230v/2,000,000ohms = 0.000115amps

which is nowhere near the 30ma, 0.03amps needed to trip an  RCD...

IR would need to be 7,666ohms or lower   (7.6K)

So can't really see that any of your readings would have caused an RCD to trip ?????

:C

 

[brummydave]

I know 2Mohm is double the minimum acceptable, but it's so much lower than normal readings, I thought it warranted further investigation.

I take your calculations on the RCD though, will keep hunting through the installation. thanks.

 

[Andy™]

2Mohm is DOUBLE the minimum acceptable of IR...

230v/2,000,000ohms = 0.000115amps

which is nowhere near the 30ma, 0.03amps needed to trip an  RCD...

but its N-E that has low IR. so voltage would be almost 0 and even less current flowing. and even less likely to trip an RCD

 

[ProDave]

odd that fitting a lamp can cause a low N-E reading. Lamps normally only have two connections, so no path to earth.

 

[brummydave]

odd that fitting a lamp can cause a low N-E reading. Lamps normally only have two connections, so no path to earth.

My point exactly, and the original reason for posting. Very odd.

Think the nuisance RCD thing is a separate issue and has clouded the post somewhat.

Another thought is that as the lampholder is earthed brass, the insulation in the bayonet cap must have degraded enough to cause the lower IR reading?

In addition,

Spoke to Electrium (Crabtree) Tech line today giving those readings....

"Oh we only have information about the milliseconds thing, we don't have information about the rest, but based on your readings the RCD is faulty. It should be 40ms for x5 and 300ms for x1. Did you do a 1/2 times test?"

"Yes, it read >310ms"

"Oh, DEFINITELY faulty then. If you want to speak to Trevor Day in Manchester he can help with the other numbers...."

I really wasn't sure what to make of their technical knowledge, but will call Mr Day in the morning.

 

[Andy™]

well a 1/2 test and a trip is a fail... but either way, id like to see an RCD that always trips at exactly 40 or 300mS

 

[batty]

The ones I test are usually around X1 18mS and X5 17mS. Some of the older ones are probably nearer those readings.

 

[SPECIAL LOCATION]

odd that fitting a lamp can cause a low N-E reading. Lamps normally only have two connections, so no path to earth.

Traditional bayonet fitting in a metal lamp holder will generally have a earthed metal housing which the bayonet gubbins mounts onto ..

and the L & N pins are on the bottom...

ES lamps wont though.

opcorn

Those quoted times are the maximum times,...

not the actual operating times...

Can't think that I have ever had an RCD come anywhere sneer the max operating time..

They are generally well under or fail altogether!

Guinness

 

[Dunx]

Agree with above, worsening readings when you put the lamp in is generally a neutral fault........none of the IR's measured would trip rcd!