RCD/MCB help

[Joe Shmoe]

The RCD has tripped a couple of times today, but none of the MCBs have.

By process of experimentation, I managed to replicate it twice when I turned a particular socket switch OFF.

I didn't manage to get it to happen on anything else, so I'm guessing that socket in particular is the issue.

My questions are:

A) why does the RCD trip but not the MCB for the ring main?

B) what is happening in the socket that causes this to happen with clicking the socket switch off? And why only on a couple of occasions and not every time?

 

[Joe Shmoe]

Sorry, forgot to add......

The socket had my xmas tree lights plugged in, so thinking about it, I'm guessing the lights are the issue and not the socket? But why did it happen when turning off? I'd expect it to happen when turning on?

 

[nickds1]

RCD and MCB. Different things entirely.

An MCB is just an electronic resettable fuse.

An RCD trips if the difference in current flowing in the live vs. neutral exceeds the set limit, typically 30mA (simplified version).

 

[Spectric]

Firstly an MCB disconnects on a fault current, an old fashioned ring main will have a 32 amp protective device. An RCD will disconnect on a residual leakage to the protective conductor of 30mA, or another way to think of it is a difference in current flow between live and neutral of 30 mA. Now with your lights I would say you do not have a protective connection or "earth" because of the insulation class. This would lead me to look at the socket as the fault and conduct further test. Another item to check is the RCD, but you can only fully test these with the right instrument.

 

[MikeJhn]

A) why does the RCD trip but not the MCB for the ring main?

B) what is happening in the socket that causes this to happen with clicking the socket switch off? And why only on a couple of occasions and not every time?

Put simply:

A) An RCD will trip if the current flow is irregular, an MCB trips on overload.

B) A loose connection can cause the symptoms you quote, thermal expansion, heating and cooling can loosen screw connections.

 

[clogs]

just to add, RCD's can get tired and trip to early.....I had one in the workhop....prev owner must have had a lot of trouble....
replaced it for a new QUALITY (SEimens) unit, no more probs....
but this will prob not be ur problem......

 

[Old.bodger]

Short answer is……you need an electrician and his test gear !

 

[woodieallen]

It's not a socket made by/branded Plus by any chance ? If so then ditch it.

 

[HamsterJam]

As others said an RCD checks the current in live and neutral legs is the same (within a certain tolerance - usually 30mA). The assumption is any imbalance means current is going to earth via a path it shouldn’t (maybe via a human).
The socket switch probably only switches the live conductor, with the neutral being permanently connected.
It is likely therefore you have an issue with the lights that causes some current to continue flowing in the neutral after the live is disconnected and the RCD is picking up this imbalance.
Remember if this is the RCD in the fuse box it will be measuring the sum of all the imbalances in the house. In other words you may have other issues that are adding up to say 25mA and the lights are enough to tip it over the 30mA required to trip.

 

[guineafowl21]

But why did it happen when turning off? I'd expect it to happen when turning on?

When you turn a switch off that’s under load, there’s often a momentary arc between the contacts as they separate. Your socket might have some damp or even carbon tracking that’s letting some current leak to earth from this arc.

You could try the socket without lights connected, or with another load attached, but I would just replace the socket.

 

[Richard_C]

An important thing about RCDs, and the reason they are so useful, is that they trip far faster and on much lower current than an MCB. Take a 5 amp MCB, it will only trip if there is an overload above 5 amps. 5 amps applied in the right place can kill you. They protect the wiring from overload (and in turn overheating) and any appliances that are plugged in. They don't do a good job with humans.

An RCD, as others have said, sense an imbalance between + and - current and trip really quickly. That 5 amps going through a human which wouldn't trip an MCB will trip an RCD because some, a very few milliamps, is going to earth. Possibly through you. You need both, to protect against both situations. When RCDs first became 'mainstream' in domestic situations - early 1970's perhaps - they were called ELCBs, earth leakage circuit breakers which in many ways is a better name because it explains itself.

Some things to think about - from a layman's perspectve, I'm not qualified:

As its a physical action, might there be a loose wire in the socket which momentarily touches the backbox when you wiggle the switch? Nothing to do with the lights. Try turning it on an off a few times with nothing plugged in.

Silly things can trip an RCD - damp, self-immolating spiders, mice chewing the nice tasty insulation, the momemtary arc tracking to earth, is it all nice and dry and clean in the area?

Then get something dead simple, desk lamp say, plug that in and see if it trips at all.

Then put your lights in a different socket, see what happens. Do you have one RCD protecting the whole house? If its a split load fuse box try the lights in a circuit protected by a different RCD.

But quite likely its the lights, some odd bit of circuitry - once upon a time they had mains voltage and just went on and off, now they have complicated stuff inside and do all sorts. Assuming its only happens with those lights, and on the grounds that new lights cost a lot less than any of the alternatives, perhaps the bin beckons? Perhaps the bin beckons before you risk fiddling with anything.

OR, get an electrician in? ("Tricky chap, electricity" - to misquote some ancient BBC TV comedy)

 

[Spectric]

Some confusion here, a circuit breaker disconnects on fault current not overload. A type B MCB rated at 16 amps will require between 48 to 80 amps to disconnect(three to five times rating) and is why you should not just change a type B for a type C which requires between 80 to 160 amps to disconnect(five to ten times rating. Overloads occur in motor circuits where for some reason the load has become higher than normal and so an excessive current is being pulled which is not high enough to cause the MCB to trip but high enough to damage the installation, and is why thermal devices are used in DOL's and some motors have thermistors embedded in their windings which feeds back to the starter control circuit.

It is important to realise that even a 1 amp fuse will require more current to blow than that needed to cause death by ventricular fibrillation, anything over 10mA will be painful and 100mA you are probably toast, so 30 milliamp residual current protection is your best freind, test and check on a regular basis as they will only save you if they are working.

 

[HamsterJam]

When RCDs first became 'mainstream' in domestic situations - early 1970's perhaps - they were called ELCBs, earth leakage circuit breakers which in many ways is a better name because it explains itself.

The early ELCBs sensed voltage on the earth conductor. If this was too high they tripped as it was assumed the earthed part of some appliance somewhere was in contact with the live. This avoided people getting a shock off the faulty equipment. However a significant limitation is they did not detect a person coming into direct contact with the live conductor.
So the RCD was developed which senses an imbalance of the current flowing in live and neutral conductors.
Although these are both technically ELCBs, to avoid confusion convention has it that voltage operated ELCBs are referred to as ELCBs and current operated ELCBs are referred to as RCDs or RCCBs.

As Spectric said an MCB is an overload device and depending on the rating and characteristics should trip in a given amount of time for a specified overload.
Manufacturers will have graphs in their data sheets but generally they should operate at the rating without tripping, trip quickly with a large overload but take (much) longer to trip with a smaller overload.

These typical graphs show with a 10% overload an MCB could take an hour or more to trip, but a B MCB should be tripping in less than 10 seconds at 3 times its rating

 

[nickds1]

Just a quick addendum to why ELCBs were replaced with RCCB/RCD/RCBs and things to watch out for with RCBs.

If the earth impedance at an installation was high, a short to earth from live could result in a potentially harmful voltage on anything connected to the earth (or protective ground, PG). By tripping if the voltage on the PG got too high, this fault case was mitigated. However in cases where earth impedance was low, an ELCB might not trip.

RCDs work, as has been mentioned, by measuring the current difference between live and neutral so however current is leaking away (to the PG conductor or through your body to the physical ground etc.) the RCD will trip.

Testing RCDs is tricky. The "Test" button on them is next to useless - all it does is test that at some point the breaker will trip.

RCDs have very specific current/time response curves. To test an RCD you need a) special & quite costly equipment, and b) someone who knows how to use it.

RCDs can fail quietly over time. It may be worth getting an EICR done by a qualified electrician. I've had brand new quality manufacturer's RCDs fail certification out-of-the-box - this is stuff bought from CEF so not fake knock-offs.

So don't rely at all on those stupid "Test" buttons and get a proper test done by someone qualified. Your life or that if someone you love may depend on it.

By the way, lots of equipment can over time slowly start to leak a bit to earth, especially if it contains surge protectors and/or Y2 capacitors - "Y" capacitors go from live & neutral to earth; "X" capacitors go between live and neutral; surge suppression devices go between any of the conductors.

Surge suppression devices have a finite life that can be considerably shortened if the utility feed is subject to spikes etc. Every time such a device suppresses a spike, it changes a bit internally as the energy from the spike is absorbed. In noisy environments, lifetimes of just a few years, e.g. 3 to 5, are not uncommon.

 

[Spectric]

As Spectric said an MCB is an overload device and depending on the rating and characteristics should trip in a given amount of time for a specified overload.

a circuit breaker disconnects on fault current not overload.

RCDs have very specific current/time response curves. To test an RCD you need a) special & quite costly equipment, and b) someone who knows how to use it.

Not so much these days as the test is automated, but you do need to set the type of rcd being tested otherwise results will be incorrect and once results are downloaded you can even print out the appropriate cert . Most things are being simplified these days to allow deskilling and wage reduction, why can't someone come up with an easy method of plastering or brick laying as both have evaded me for decades !

 

[moosepig]

Can any of the electricians on here explain why MCBs only switch live?

I was replacing a socket the other day upstairs - no sweat says I, just flip the MCB for upstairs sockets and press on. Accidentally touch neutral and earth at the same time... residual current is enough to trip the RCD and the whole house goes off. "WTF?" says I, then it dawned on me... MCBs don't isolate neutral.

Is there a good reason that neutral is all bussed together in the CU and then not switched anywhere but the master breaker? Switching both live and neutral on every circuit would surely be safer, and much less prone to taking the whole house out.

Apologies to OP for thread drift...

 

[woodieallen]

Can any of the electricians on here explain why MCBs only switch live?

I was replacing a socket the other day upstairs - no sweat says I, just flip the MCB for upstairs sockets and press on. Accidentally touch neutral and earth at the same time... residual current is enough to trip the RCD and the whole house goes off. "WTF?" says I, then it dawned on me... MCBs don't isolate neutral.

Is there a good reason that neutral is all bussed together in the CU and then not switched anywhere but the master breaker?

The MCB is only there to protect against overload. It would be pointless to have one in the neutral feed.

Switching both live and neutral on every circuit would surely be safer,

No, an extra MCB would not proivide any extra safety. It is the RCD that protects you against shock.

and much less prone to taking the whole house out.

Apologies to OP for thread drift...

The RCD was doing what it was supposed to do. There is often a very slight voltage difference between neutral and earth. Accidentally touching these together can cause enough current to trip the RCD. Nothing at all to do with the MCB.

HTH

 

[guineafowl21]

You can get double-pole MCBs, apparently more popular on the continent, but single-pole ones are used more here for simplicity, cost, space-saving, force of habit, you name it.

DP ones must, of course, be designed to isolate both poles at the same time. You certainly don’t want the neutral side only off, so the circuit is still live but appears dead. Ditto fused neutrals.

Obviously to avoid the problem, disconnect the neutral as well, but I would suggest your wiring could do with an upgrade if the whole house is on one RCD.

 

[Spectric]

The MCB is only there to protect against overload.

The MCB is an overcurrent device, an overload will not cause it to disconnect. EG circuit designed to handle 30 amps with a 32 amp type B MCB, the motor is being overloaded due to say a bearing seizing and is now pulling 50 amps but the MCB needs at least three times rated to disconnect, the wiring now overheats and you would have a thermal event. This is why you have thermal overload devices in contactors and DOL's to trip on excess current.

The other thing to point out is that a disconnected neutral can be live, enough to electrocute you and has caught out many, just revisit Ohms law !

 

[woodieallen]

.... but I would suggest your wiring could do with an upgrade if the whole house is on one RCD.

I can't agree with that statement for several reasons. First, you have no idea on the size or type of his property. Second, you are applying 'current' best practice. The newer regulations suggesting, for example, split upstairs lighting from downstairs lighting is more a question of convenience IMO then electrical safety. Thirdly....rewiring a whole house ? You must be joking. The hassle etc. Replastering. Redecorating.