Intemittent tripping RCD.

[HappyHippyDad]

Hello everyone..

I have a job tomorrow tracking down a tripping RCD fault.

This would have panicked me a year ago, now I'm looking forward to it and cant wait to get there!

However, from the phone conversation I have discovered the following:

1. No recent DIY work or professional work has been carried out recently.
2. One of the RCD's (dual board) trips every 5-40 minutes, this RCD feeds the kitchen ring, all lights and one other circuit (forgotten which)
3. It does not trip at any regular time of the day (so possibly fridge or other irregulary timed power source)
4. It started approx 3 days ago, nothing 'different' happened at that time. No new fridge, cooker etc.
5. Sometimes trips when different appliances are switched on, sometimes just spontaneously.

Basically the prelimanary detective work has given me little insight into the problem so I will be relying on IR testing and earth leakage clamp meter. I expect/hope that the global IR test will show a poor reading so then it will be a simple job of narrowing the fault down. However..

I do like to try and have a plan for any eventuality, so this is my question:

1. If the IR tests show no fault then where do you go from there?

I know I'm being a bit ----, but I like to play it a bit like a game of chess and think many moves ahead!

My thoughts are (given good IR results):

- homeowner could leave one of the affected circuits switched off and see if RCD stops tripping.
- Check all items that get plugged in for earth leakage.
- Visual inspection of O/S lights for poor IR.
- continuity tests between neutrals of differing RCD's (highly unlikely given fault has only just started occuring)
- RCD tests, including ramp.
- Check for loose connection upstream of RCD.

Have I missed anything that you guys would automatically do/think?

 

[Leesparkykent]

Ramp test the rcd and see what ma it's tripping at.

 

[R G]

I think with the recent posts about this topic , the most valuable test once the preliminary tests of the circuit have been done is the ramp test with appliances connected and without , that would indicate what appliances are faulty , you can keep trying it and unplugging appliances until the readings improve to narrow it down ..

 

[telectrix]

as above. ^^^^.

 

[Doomed]

Had this the other day, eventually traced to a twisted bit of cable on the kitchen ring main - took hours to find - had been happy for years, but after time it started to trip occasionally, then more frequently and then a couple of times an hour, thats when we were called.

 

[HappyHippyDad]

I think with the recent posts about this topic , the most valuable test once the preliminary tests of the circuit have been done is the ramp test with appliances connected and without , that would indicate what appliances are faulty , you can keep trying it and unplugging appliances until the readings improve to narrow it down ..

Thanks Ray,

I'm unsure how having the appliances connected, unconnected affects the Ramp test though? Doesn't the ramp test just show at what mA the RCD is tripping when you put the fault current through it?

 

[VoltzElectrical]

One that got me on an intermittent was not leaving the earth cables connected in the bar testing between cores on IR. The cable cores were clear to each other, but the live was down intermittently to the general mass of earth which I had eliminated by taking the earth cable out of the bar. Gggrrrr.

When the installing electrician stripped the cable back for the outside light, he had ringed the insulation on the live that was touching the brickwork as it came through the wall. Sometimes it would trip every couple of weeks. Sometimes every couple of days. Very frustrating trying to fault find a fault that isn't there when you attend, but get a phone call a week later to say it's tripped again!

You can have this little gem for free, as long as you let us know what it was when you find it dude.

x

 

[telectrix]

if you ramp test with no circuits connected and get, say, 25mA, then with circuits/appliances connected get 10mA, that says there's 15mA leakage.

 

[R G]

On pretesting,one that got me was some outdoor spike lights that are class 2 with twin flex in use , and someone had clipped twin flex around to adaptable boxes ,with the lighting circuit with CPC disconnected at the DB it was all clear , But testing down to the MET it was down , eventually found a buried adaptable box wrapped up in a carrier bag ,

 

[R G]

Thanks Ray,

I'm unsure how having the appliances connected, unconnected affects the Ramp test though? Doesn't the ramp test just show at what mA the RCD is tripping when you put the fault current through it?

Thats the most valuable part really because your testing with the leakage there , as Tel says the readings will tell you which appliance is registering the most when they are isolated ( i.e. the difference ) as you unplug items