Existing Rcd Sockets With Rcbo's

Talk Electrician Forum

Help Support Talk Electrician Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
TBH, I would leave the RCD sockets in situ.

Non-discrimination is not always a bad thing.

The best case scenario is that a socket will trip before the RCBO, which prevents nuisance trip on the rest of the circuit

The worst case scenario is that the RCBO trips before the socket, which is the same as switching to standard sockets.

I've been away from the Regs for a few years, but I don't think it conflicts with BS7671 guidance. I expect to be possibly shot down!

 
I often fit series RCD's, that is 30mA RCD s.o. with upstream instantaneous 30mA RCD, as redundant protection.

That is at least one of them will trip under fault hopefully, even if they have not been tested every 4 months.

It is not specifically against the regs, however there will be no discrimination as to which trips under fault conditions.

 
Hi Sidewinder,

My point precisely. Who cares which trips first, except that if it is the socket, then nothing else is affected.

Changing the RCD sockets would be a totally unnecessary expense.

 
I always thought that RCD sockets worked slightly differently to a standard RCD as when the earth is connected there is a low impedance to earth - hence instructions to disconnect when doing insulation tests.

My concern is that the RCD socket may make the tripping threshold of the upstream RCD just that bit closer.

Personally I would change them - and recycle them as appropriate!

 
No Dave,

The instructions to disconnect are because they are electronic devices in the main and could be "fried" by 500V to earth!

In the experience I have of fitting dozens of these in a redundant role as I explained above, testing @ the "new" upstream RCD & the downstream RCD s.o. the RCD s.o. has no affect on the disconnection time in reality.

Remembering that an RCD should be tested with all downstream loads disconnected, and a cable is seen as a load by an RCD...

Stupid that an RCD can pass test under BS7671 but still fail to disconnect in the event of a fault when in circuit, and this is OK, unlikely but OK!

 
Thanks - I've always appreciated the 'Frying Tonight' issue - but when I've done overall insulation tests as a health check before fitting a new CU (generally at 250v until I know what's going on with L+N connected) I've found that there is a low impedance to earth when there is an RCD fused spur fitted.

This wouldn't occur with a conventional RCD as there no connection to earth.

Whether this materially affects trip times - I don't know - I've never tested it - but I'll bow to your experience!

Must confess I never test RCDs with cables disconnected - surely when testing we're looking system rather than individual components.

No Dave,

The instructions to disconnect are because they are electronic devices in the main and could be "fried" by 500V to earth!

In the experience I have of fitting dozens of these in a redundant role as I explained above, testing @ the "new" upstream RCD & the downstream RCD s.o. the RCD s.o. has no affect on the disconnection time in reality.-

Remembering that an RCD should be tested with all downstream loads disconnected, and a cable is seen as a load by an RCD...

Stupid that an RCD can pass test under BS7671 but still fail to disconnect in the event of a fault when in circuit, and this is OK, unlikely but OK!
 
No mate, the cables act as a load & a capacitance, the meters we use are too dumb to know the difference between the disconnection time of the RCD on its own, or the disconnection time of the RCD plus the voltage decay from any connected cable capacitance.

Hence why RCD only.

RCD in circuit is not as per regs.

 
Blimey - love this forum - learnt something new!

Thinking about it - as soon as a very small resistive load is connected the cable capacitance would be completely swamped.

If the RCD is passing trip times with cables connected surely this is worst case?

 
Last edited by a moderator:
Top