Urgent help - RCD tripping

[gooders]

The RCD tripped out on Sat night for the ground floor ring main....usual chaos - freezer defrosted etc etc....currently running an extension lead to upstairs as there's nothing running on that ring main.

Strange thing was none of the MCBs had tripped. I tried to reset the RCD and it would only switch on it I switched off the MCB for the ground floor ring main. Since it has allowed me to reset the ring main several times and stays on for 20-30 mins before tripping out.

Thinking back this has happended 6 times over the past 4 years but each time the RCD re-set so I didn't think there was a problem.

I have tried unplugging everything on the ring main and switching off all the sockets. I also swapped out the MCBs but no change.

Its a 4 year old MK consumer unit with 30ma RCD.

It seems there is a problem with the ground floor circuit but as there's nothing new on the circuit, there's no obvious damp, no water leaking etc... why would it suddenly start tripping?....

What should I do to try to isolate the cause of the tripping?

All help gratefully Appreciates, Many thanks.

 

[plugwash]

disconnect the cuircuit from the the CU and use a multimeter to mesure resistnace between the different cores

if that gives infinite resistance all ways then you are going to need to get a sparky in with more specialist test gear to find the problem

if it reports a lower resistance then try to break the cuircuit into parts and test each part seperately until you find the bit that is causing the problem

 

[gooders]

Thanks,

I disconnected the ground floor RM for the CU and measured the resistance - 34 Ohms between L&N....which I take it is not good?

I tried to re-set the RCD and it came on first time and stayed on for 20 minsbefore tripping out.... does this suggest anything about the cause....seems to be about the same everytime?

How do I got about trying to find where the problem is on the circuit?

thanks

 

[ban-all-sheds]

Thanks,

I disconnected the ground floor RM for the CU and measured the resistance - 34 Ohms between L&N....which I take it is not good?

That depends on whether you'd also unplugged everything on the ring as well - 34 ohms could easily be a load (approx 1.5kW, if that helps to identify it).

For an RCD trip you're looking at an earth fault, so its L-E and N-E you want to be measuring.

PS - this is why you should not have the fridge/freezer on an RCD circuit...

 

[gooders]

thanks,

confused though ..... I thought the RCD measured the difference between the live and neutral and if the different was signficant it tripped?

When I measure the L-E , N-E its 40K Ohms ...... though I didn't have everything unplugged.... dooh, I'll start again.

 

[ban-all-sheds]

confused though ..... I thought the RCD measured the difference between the live and neutral and if the different was signficant it tripped?

Yup, and the way that difference arises is by current somewhere flowing to earth, i.e. a L-E or N-E fault. Current is supposed to flow between Live and Neutral, and say for argument's sake you had several appliances plugged in and switched on when you did your L-E test you could measure a resistance of 7 or 8 ohms...

When I measure the L-E , N-E its 40K Ohms ...... though I didn't have everything unplugged.... dooh, I'll start again.

40K is very badly wrong - either the cable is damaged, or there's a faulty appliance somewhere.

 

[fubar]

thanks,

confused though ..... I thought the RCD measured the difference between the live and neutral and if the different was signficant it tripped?

No, it measures the difference between phase-referenced-to-earth and neutral-referenced-to-earth (phase is the sparkies name for what you and I call live). That is, it trips when there is current leakage from phase to an earth or neutral to an earth. Another way to look at this is to say that the RCD trips when the current flowing in the phase conductor is not matched by the current flowing in the neutral conductor.

You can read more here:
http://www.tlc-direct.co.uk/Technical/Distribution/How RCD Works.htm

--
Michael

 

[plugwash]

a rcd basically monitors if the currents on the live and neutral cancel out

that is when current is flowing out of the live it is matched by the same amount flowing into the neutral and when current is flowing out of the neutral (remember mains is ac) it is matched by the same ammount flowing into the live

so there are basically 3 ways to trip a rcd
1: current flow between live and earth
2: current flow between neutral and earth (this needs a much lower resistance than the first one though as the volotage between neutral and earth is usually pretty small)
3: current flow from live or neutral after the rcd to live or neutral before it (ie cable in wrong neutral bar)

thanks,

confused though ..... I thought the RCD measured the difference between the live and neutral and if the different was signficant it tripped?

No, it measures the difference between phase-referenced-to-earth and neutral-referenced-to-earth (phase is the sparkies name for what you and I call live). That is, it trips when there is current leakage from phase to an earth or neutral to an earth. Another way to look at this is to say that the RCD trips when the current flowing in the phase conductor is not matched by the current flowing in the neutral conductor.

You can read more here:
http://www.tlc-direct.co.uk/Technical/Distribution/How RCD Works.htm

--
Michael

wrong earth has nothing to do with how a rcd works although the faults that trip a rcd usually involve it
(sorry had to correct typo)

 

[plugwash]

thanks,

confused though ..... I thought the RCD measured the difference between the live and neutral and if the different was signficant it tripped?

When I measure the L-E , N-E its 40K Ohms ...... though I didn't have everything unplugged.... dooh, I'll start again.

live-earth 40k

assuming 240V (which is fairly realistic and i can't work with 230 in my head)
240/40=6ma

this on its own should not be enough to trip a rcd though the fault may not be purely resistive so this calculation may be inaccurate and furthermore some of the problem may be on other cuircuits

older appliances (especially anything that involves electronics or motors or water and has an earth connection) may well be the source of your problem

it could also be failing cable insulation though your CU must be fairly recent to have a rcd so failing cable insulation probablly means one of the following
1: very poor quality cable
2: damage to cable by rodents diyers etc
3: sustained overloading
4: some arsehole left old vir or lead in desipite replacing the CU

 

[gooders]

thanks for the responses

could it be that something is sending it just over the limit as its only just starting happening - I guess I could check this by unplugging everything and seeing if it still trips.

If it does still trip does this then point at a definite problem with cable insulation.

We have solid wood floors, 2 year old kitchen etc etc so its going to a nightmare with everything buried in the walls and under solid wood floors.


..... if I or a sparky is unable to isolate the problem is it possibe to change to a 100mA RCD, or is that bad practice/unsafe.


thanks

 

[ban-all-sheds]

Bad practice. Unsafe (masking a problem which may be worsening instead of fixing it).

And non-compliant with the wiring regs (sockets which may reasonably be expected to supply equipment outdoors should have 30mA).

 

[gooders]

thanks ban-all-sheds

Its just that I've read other threads where tripping RCDs are a niusance - particularly as I've freezer on the RCD, and I don't have any outside lights, pond pumps etc, no electric shower or immersion heater I thought it may be okay ...... that said I think its looking like a cable problem so like you say its not a wise move.

Hopefully, I'll get to the bottom of it over the next few days.

 

[gooders]

thanks,

confused though ..... I thought the RCD measured the difference between the live and neutral and if the different was signficant it tripped?

When I measure the L-E , N-E its 40K Ohms ...... though I didn't have everything unplugged.... dooh, I'll start again.

live-earth 40k

assuming 240V (which is fairly realistic and i can't work with 230 in my head)
240/40=6ma

this on its own should not be enough to trip a rcd though the fault may not be purely resistive so this calculation may be inaccurate and furthermore some of the problem may be on other cuircuits

older appliances (especially anything that involves electronics or motors or water and has an earth connection) may well be the source of your problem

it could also be failing cable insulation though your CU must be fairly recent to have a rcd so failing cable insulation probablly means one of the following
1: very poor quality cable
2: damage to cable by rodents diyers etc
3: sustained overloading
4: some **** left old vir or lead in desipite replacing the CU

I've now re-measured making sure everyting is disconnected, unplugged and it all looks okay except the L-E connection measured - 7.5k upto 20k ohms, not sure why it varied, probably my connection from the multimeter leads to the cables.

230/20 = 11.5mA...so its getting closer to tripping (I read 30mA RCDs can trip out at 16mA?) and 7.5 would be just over 30mA.

It did stay on for over an hour last night without tripping.... does this indicate that its close to the trip level and something like the freezer going onto cooling cycle, or something like that, could then send it over the trip level? thanks.

Is the next step to test each socket on the ring main to identify where the fault is? Could you explain the best way to do this?

many thanks.

 

[ban-all-sheds]

Find the socket which as far as you can tell is half-way round the ring.

Disconnect it from the circuit.

Measure L-E on each half of the ring.

On the dodgy half, find the socket which is half way along that, disconnect it etc.

Shouldn't take too long to identify the run of cable to blame, always assuming that you don't find the problem at one of the sockets...

Or suddenly think, "Hang on - I wonder if I nicked the cable when I hung that picture the other week?"

 

[securespark]

have a similar problem which I am going to have to look at SHORTly.

I have installed a 16A radial for a toilet heater, which is not yet installed. The RCD tripped the other day, and elimination showed it as the fault. Yet this circuit has only been installed a few weeks. Guess damage has ocurred somewhere, but where? I suspect a chippie has put a nail or screw through the cable, but then the skirting was fixed before my cable went in.

Very odd. Since I put the cable in, the hall carpet (over the hatch) has been put down (brand new!) so I shall ask the customers to cut a new one in the adjacent room (no carpet). I have to go down and look, but I think it is cut and dries because the cable is in capping and now it won't move so i guess it has been pinned through. Lucky for me I always leave a loop so I can free the cable and pull a bit of spare up.

I'l let you know what I find, in case it is of use to others tracing RCD tripping faults.