Rccb question

[oldskoolskater]

Hi All

I have a consumer unit problem that is driving me up the wall ! I installed a Crabtree split-load dual rccb unit in a house with a TT supply, but am having trouble with the split-load rccb (63A 100mA) in that it will not trip using the test button. It does trip however as soon as I put a load on any of the MCBs it feeds.

I naturally thought that the rccb was faulty and swapped it for another but this does the same?? The test button operation of the rccb is not rocket science - you press the button which puts a resistive load across live and neutral, the fault sensing winding then trips the rccb. So why doesn't this work??

I now have a 3rd rccb to try but I am confident this will not help things. This has never happened before and the supplier also confirms that repeated failures of crabtree gear is unheard of - it just doesn't happen (though its not impossible)

Any ideas?

The other rccb (80A 3mA) in the CU is working fine and trips as expected when tested.

[ARWadsworth]

Check that all the neutral tails are in the correct bus bars. I have had two Wylex CUs (made by the same people as Crabtee) where the neutral tails were in the wrong busbar and had identical symptoms.

Adam

[oldskoolskater]

Cheers for the thoughts

I have removed the links so that N1, N2, and N3 are separated and wired as per the tech diagram for this configuration.

Basically both rccb's get the common supply neutral (N1) into them and they feed the separate neutrals (N2 and N3) out the other side.

I really am scratching over my head over this one, and am on 1st name terms with Crabtree technical haha!

Cheers

Oldskool

[oldskoolskater]

Cheers for the thoughts

I have removed the links so that N1, N2, and N3 are separated and wired as per the tech diagram for this configuration.

Basically both rccb's get the common supply neutral (N1) into them and they feed the separate neutrals (N2 and N3) out the other side.

I really am scratching over my head over this one, and am on 1st name terms with Crabtree technical haha!

Cheers

Oldskool

[Harry Bloomfield]

[e-mail] formulated the question :

Something rather odd here.

That is not how it works!

The button connects between L and E via a resistor which causes an imbalance between the L and N (not passing an equal current through both conductors), the imbalance is what causes it to trip.

Should the button be between L and N then pressing it would not make any difference as the load would remain equally balanced through the two conductors.

Combine that with it tripping when ever you do put a load on via an MCB

- leads me to suggest that you have crossed or swapped the N & E somewhere in the board.

[Nick]

[ARWadsworth]

It is how it works! The test button is connected to the opposite sides of the sensing coils on the RCD between live and neutral. Most RCDs have a picture of this on them.

Adam

[Nick]

Adam, The resistor simulates leakage to Earth - by connecting a conductor (Phase or Neutral) to Earth.

Normally, current flows between the two conductors and this is when it should NOT trip.

Only when the current is different in each of the flow/return conductors ( i.e. some to Earth) should it trip. The coils for the trip circuit are contra-wound so equal current, in each, balances and no trip is caused

[ARWadsworth]

Crabtee RCDs have Lin, Lout, Nin and Nout. Four connections and NO earth.

Adam

[ARWadsworth]

You can test the 100mA RCD by swapping it with the 30mA one. Easy and cost free.

If the test button works when you have swapped them then assume the RCD works.

Adam

[Ron Lowe]

In which case, the only way the test button can work is to connect a resistor from the L output to the N input ( or vice versa ), so the test current flows through the L current sense , but not the N sense. ( or vice versa )

In this case, the test function is 100% within the RCD, and does not depend on external wiring. If the test button fails to trip it, it must be faulty.

So it revolves around exactly how the test button works. Does is actually attempt to flow some curent to earth, or does it just flow some current in one leg which remains un-accounted for in the other? I think we should be told :-)

[Harry Bloomfield]

Ron Lowe submitted this idea :

You are correct Ron, I have just checked my DB in the garage! So the test button must connect a resistor in series with one of the windings, as well as applying a small load to the output. Still amounts to the same thing - he has some how managed to get his connections wrong.

[ARWadsworth]

the best I can find at the monent. Notice the words

"A test circuit is also incorporated whereby connection is made from load phase to supply neutral via a TEST COIL and RESISTOR and activated by a TEST BUTTON."

Adam

[ARWadsworth]

Not if there is no neutral present. You need the potential difference to make the current.

Adam

[Allan Mac]

I now have a 3rd rccb to try but I am confident this will not help things. This has never happened before and the supplier also confirms that repeated failures of crabtree gear is unheard of - it just doesn't happen (though its not impossible)

Any ideas?

Last year I had 3 Wylex split load CUs in a row where the RCD was faulty straight out of the box. Different problem, would not latch closed. (before installation).

Allan

[oldskoolskater]

I have checked the technical drawings in the Crabtree catalogue and it is definitely correct.

It does appear that the test facility is entirely within the rccb and there is no connection to earth at all. I'm not sure how I could get my connections wrong as this should function work within the CU independant of any connections with regard to the domestic circuits.

Snippet from the rccb installation instructions:

1)The incoming cables to the rccb may be taken either to the top or bottom set of terminals, dependant entirely on the ease of wiring the device 2) As the rccb works on the current balance principle, all cables feeding the protected installation (live and neutral) must be connected to the device. If this is not complied with, continuous nuisance tripping of the breaker will result when power is drawn from the supply.

I know that no.2 above sounds like I haven't connected a neutral but if that were the case the other rccb wouldn't work either..

3) Having installed the unit and switched on the rccb, press the test button in order to test the correct function of the device. This ONLY tests the operation of the rccb and not the continuity or value of earth resistance path

Surely a rudimentary test would be for me to connect one set of the L & N on the rccb to a flying lead plugged into my socket ring main and hit the test button? This would save me trekking over to the installation yet again to find it not work again!

Thanks for all your views

[cynic]

Sounds as though the live input and output of the RCD is wired in opposite direction through the neutral input and output. The intended cross path provided by the test button will not cause imbalance. Any normal current will immediately trip the unit. Check the input live and neutral go to the right input terminals and the output goes to the load @(i.e. the MCB busbar and the protected Neutral rail) Most likely error is in the neutral connections If this is not the case try posting a photo on a suitable site.

[Andy Wade]

Yes, that should work.

How about providing a close-up photo of your CU internal wiring in-situ? Someone might then be able to spot something...

[Jim Michaels]

The source L & N must be taken AS A SET to either the top or bottom SET of terminals.

The output L & N is from the other SET (top or bottom).

Splitting the sets will cause the exact symptoms you describe.

Remove SPAMX from email address

[Dave Plowman (News)]

Assuming you've wired the CU correctly, I'd be looking for a neutral funny elsewhere. Something like a shared neutral.