RCD - Why does that happen?

Earth to Neutral even when not powered up makes RCD trip. Bloomin PITA when you have your sparky in the loft and you're trying to connect wall sockets. (and the loft light is a temp one powered by kitchen ring)

Reply to
Pet
Loading thread data ...

Pet formulated on Friday :

RCD's are designed to be sensitive to any imbalance of current comparing the L to the N. As there is almost always some small voltage on the N when measured against earth, shorting the two will cause some current to flow in the N which will not be on the L.... Hence the reason it trips.

Reply to
Harry Bloomfield

Thanks.

Reply to
Pet

I have found that just touching the N against something with enough free electrons (i.e. lump of metal like a chair leg, filing cabinet etc not even connected to "earth" as such) can be enough to cause a trip

Reply to
John Rumm

In article , Pet writes

Very simple. The current going out of the live terminal and coming BACK into the neutral must be in BALANCE within the tripping or un balance level of the RCD if they are not, then it will disconnect. This can happen most usually with current on the live being leaked to Earth or current on the Neutral finding a leakage path to earth as well.

So current OUT must be the same as current back IN or within the tripping level.

However there must be sufficient current above the trip level. You could connect the neutral and earth together with no current flowing anywhere in the system as this will be below the trip level, but as soon as you turn something on then there is sufficient trip current, and off she goes!.....

Reply to
tony sayer

Ah, this was my point,

Even with the the circuit disconnected (fuse pulled) there is enough electrical stuff on the wire to trip it. Though insufficient voltage to trip it if connected through high resistance media like skin.

Reply to
Pet

But you had the loft light on (sounded like it). Any current flow anywhere in the house will cause the neutral potential to rise slightly above the earth potential. Any contact between the neutral (at raised potential) and earth will cause a current to flow to earth, causing an imbalance that will trip the RCD. Disconnecting the fuse/mcb doesn't disconnect the circuit neutral.

Reply to
fred

..or anywhere on the same substation, assuming that's where your neutral is earthed.

Reply to
Mike Harrison

On mine, Earth & Neutral are bonded where the cable enters the property, which I think that is quite common. I'm sure there are others who could tell us which designation that makes my supply :-).

Reply to
fred

Just so. Most of the answers so far have focused on the potential difference between E and N. This isn't the most helpful way of explaining the trip, IMNSHO. As Tony Sayer wrote, your RCD is a current balance device: what goes out on the L side must, to within say 20mA, come back on its N line. When you short the N side to E, you provide an alternative return path for all the current making its way along the N conductor. How much of this return current flows down this new path is determined exactly by the resistance of this new path compared to the existing one - so, if the resistance along the various E conductors is, say, twice the resistance of the existing path, fully 33% of the current will choose this nice new way back home. Inbalance. RCD goes pop.

"But", you say, "there's no effing return current flowing in my N wire, cos I popped the MCB on the circuit I'm working on!". Ah yes, little flower; but, your sage asks, how many poles in that MCB? "No bleedin' Polaks in my MCB, mate; not a moustache or a pair of crimplene trousers in sight!" Patiently, your sage berates you for perpetuating outdated East European efnic stereotypes, and points out that the MCB, being in sparky-speak Single Pole, interrupts only the L side. Your circuit's N wire is still connected to all the other Ns on the RCD side of your split-load consumer unit. So, the busy little electrons which are bumping into each other to make the current flow in all *those* Ns now get a new way to send the current back Home, bypassing the current-balance sense coil in the RCD. Relatively few of them will take this longer, higher-resistance path - here the resistance ratio is that of the relatively hefty N busbar and meter tails from the CU to the supply point, versus the path back out along your only-L-side-disconnected circuit's N conductor, back down its E conductor, and reuniting with the supplying N through a maze of your bonding conductors, the N-to-E bond you have an a PME installation or The General Mass Of Earth in a non-PME one, and the like. But, since this is a sensitive trip, your new return path can have a resistance

1000 times bigger'n the "right" return path, and still make your RCD trip if there's 20A flowing in the other circuits (or, 100 times and 2A).

(I keep suggesting that 20mA is the trip threshold, because a nominal-30mA RCD's actual spec is "will not trip at 15mA, will trip fer-sure within 40mS(?) at 30mA", so in practice (handwave) most of the nominal 30mA RCDs shipped will trip at 20mA for multiple-mains-cycle-times).

And yes, this means the "touched-N-to-E-and-house/loft-went-dark" hassle will indeed *not* happen if you (a) have individual *double-pole* RCBOs

- common in The Rest Of Yurrup, though single-pole seems to be still dominant here in the You Kay; or (b) disconnect the N side of the circuit you're working on; or (c) use lights which aren't on the RCD-protected side of the circuit, Obviously!

Aye - PME, in oldster-speak; or, TN-C-S in the New Uniform Dialect.

HTH - Stefek

Reply to
Stefek Zaba

Given the choice of explaining that a potential difference will cause a current flow and giving a lecture on Kirchoff's laws on current sharing, I'll chose the former :-). A few bright sparks there are here, but I like to keep it simple. The potential difference approach is something the punters can grasp easily and see that it will result in a current imbalance (hopefully).

Cheers,

Reply to
fred

Ah well - there's the glory of Usenet; this way our dear readership gets both, right? ;-)

Cheers, Stefek

Reply to
Stefek Zaba

IIUC, the requirement is that if your RCD has a trip current of I, it will trip within 200ms for a fault current of 100% of I, and within 40ms for a fault current of 5 x I or more. It ought to cope with 66% of I indefinitely.

Reply to
John Rumm

Thinking about it in the bath, I think the reason I preferred an explanation in terms of opening up a different way for the current to flow is preceisely that an RCD is a *current* balance device - so the explanations are all in terms of the one concept of "current", leaning on the semi-intuitive notion of "resistance" as the thing which determines how much current flows once it "has a choice". Introducing the idea of a potential difference - however much one can argue that it's essential to make current flow, of course! - introduces another idea, which your own post mentions will only "hopefully" cause t'punter to realise a current flow will result.

But like I said, multiple explanations (provided they're just different ways of phrasing the same physical reality, rather than requiring direct intervention of little green men from the planet Zog, or the inherent superiority of two combis ;-) all help...

Cheers, Stefek

Reply to
Stefek Zaba

Waaaaaaaaaaaaaaaaaaaay too much time on your hands ;-)

Reply to
fred

I thought it was 50%. At least, they are tested at 50%. I suppose the design may (will) be different.

Christian.

Reply to
Christian McArdle

Yes, that is the spec - John quoted the appropriate times-to-trip at nominal current (200mS) and 5 times that (40mS), but the last part of the spec is "will not trip at 50% of nominal applied indefinitely". Thuz-and-thurfor, the mfrs use the band between 50% and 100% of nominal as the target band for tripping; with the spec requiring definite tripping at the nominal current, yer smarter mfr maximises yield by designing for actual tripping at, oh, 70% of nominal, so that few will fail in service on either end of the spec...

Stefek

Reply to
Stefek Zaba

Sorry, should have made that clearer. I was not intending to suggest that the spec was for 66%, it is in fact 50 as you say.

The 66% came from Stefek's post (regarding a 30mA RCD tripping at 20mA). I was only commenting that chances are you will be able to run at a 20mA leakage without difficulty in most cases. Manufacturing tolerance will come into play however, and YMMV.

Reply to
John Rumm

HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.