36 replies to this topic

[Ali]

I was changing a CU today and didn't do an IR test before I started (that's a lesson learned). All went well until I started turning the new duel RCD board on.
I turned on the main switch and then the first RCD. I pressed the RCD test button like all good leccies should and nothing happened, the RCD didn't trip. All MCBs are off at this point.
I narrowed the problem down to one of the lighting circuits on that RCD. If both the neutral and the CPC were connected in the board, the RCD test button didn't work and the RCD didn't trip at or .
I disconnected the circuit and found that there was continuity between the neutral and CPC.
Looked in lots of the light fittings until I found the problem, rectified it and all was good with the world again.
My question is, why would a properly working RCD not trip if there is a neural to earth fault, I would have thought it would trip all the time and not reset.

It seems that I don't understand RCDs. Anyone able to enlighten me?
Thanks

[Zeespark]

Ali I have some paperwork on RCD's basic through to quite advanced. Want Me to try and track it down and send it to you in a PM?

AndyGuinness

[Ali]

Oooh yes please Zeespark that would be great.

[Zeespark]

You have mail Ali.

AndyGuinness

[PC Electrics]

Zee

is that something you could put in the downloads section? Sounds proper useful.

[Zeespark]

Possibly, but its written when the 16th was about, though it is just minor changes in terms of wording, the actual workings of an RCD havent changed.

AndyGuinness

[Badger]

id find that really useful zeespark aswell as others here

[SPECIAL LOCATION]

I was changing a CU today and didn't do an IR test before I started (that's a lesson learned). All went well until I started turning the new duel RCD board on.
I turned on the main switch and then the first RCD. I pressed the RCD test button like all good leccies should and nothing happened, the RCD didn't trip. All MCBs are off at this point.I narrowed the problem down to one of the lighting circuits on that RCD. If both the neutral and the CPC were connected in the board, the RCD test button didn't work and the RCD didn't trip at or .
I disconnected the circuit and found that there was continuity between the neutral and CPC.
Looked in lots of the light fittings until I found the problem, rectified it and all was good with the world again.
My question is, why would a properly working RCD not trip if there is a neural to earth fault, I would have thought it would trip all the time and not reset.

It seems that I don't understand RCDs. Anyone able to enlighten me?
Thanks

Hello Ali..
few observations and thoughts

1/ All MCB's off so even if power is on to the supply side of the RCD, there is no current flowing through the RCD yet cuz no loads switched on?

No current flowing so RCD wont trip even with a N-E fault present.


2/ If you had got some of your MCB's on with some appliances or lights switched on somewhere in the property I suspect the RCD would have tripped cuz current could now bypass the N half of the RCD through your N-E fault.


3/ Is the supply on the installation PME/TN-C-S..
i.e. Neutral and Earth joined together at the suppliers main cut-out?
cuz this will give you the symptoms you are describing:-

I will try & explain:-

Have a look at figure 11.1 on page 93 of the On-Site-Guide.

IF PME/TN-C-S, you can draw an connection between the N & E on the bottom left hand side of the drawing..

Now also put another connection between the N and E just before the load on the bottom right hand side of the drawing. (your N-E fault)

Remember the brown live wire from the RCD to the load would be open circuit because your MCB's are off..


Now lets consider a possible test current path...

• From Live
• Up the black wire to the test button
• Through the test button
• Through the test trip load resistor

(This is where it can now take two paths)

• (i) Along the neutral through the RCD coil
  
  or
• (ii) Along the neutral wire toward the load
• (ii) Through your N-E fault
• (ii) Along the Earth wire
• (ii) Back from the Earth to the Neutral at the PME/TN-C-S connection.

If this is a PME/TN-C-S earth arrangement the two paths will probably be of a very similar resistance so the fault test current will most likely divide equally down the 'N' and the 'E'.

So the little bit of current flowing through the N coil of the RCD is probably too small to cause it to trip!


If Earth arrangement is TN-S or TT the resistance via your N-E fault down the earth conductor will be higher so the test fault current going down the E will be less..

So more will go down the N coil of the RCD and provide enough to trip the RCD with the test button!




hope that makes sense????

[Ali]

Thanks Specs, that's a great description. I've had all my felt tips out and been drawing RCD diagrams and it's begining to make sense.
It it TN-C-S by the way.

I was trawling the internet for ages trying to get an answer, I should have know that this was the best place to come.

Thanks for the RCD info Zeespark, I'll sit and have a read through as soon as I've poured my G and T.

[Manator]

I was trawling the internet for ages trying to get an answer, I should have know that this was the best place to come.

This is the only place I think !!

[Zeespark]

It is , I have found that other paperwork if you require it Ali, just say the word.

AndyGuinness

[steptoe]

sorry if I am wrong on all this,
BUT,
isnt the test button on the RCD a mechanical device for inducing a false fault.?
so as long as the RCD has a proper incomer then it should trip otherwise it is a faulty unit.

[SPECIAL LOCATION]

sorry if I am wrong on all this,
BUT,
isnt the test button on the RCD a mechanical device for inducing a false fault.?
so as long as the RCD has a proper incomer then it should trip otherwise it is a faulty unit.

you are Incorrect there steps!...


see drawing on page 93 O-S-G.


If you stick a meter across the L&N on a dead RCD..
(e.g. not connected)

With the switch on and the test button pressed you typically get around 3400ohms.

Basically it just puts a diagonal path across the RCD to induce a trip current..

What it doesn't do is prove the accuracy of the trip current or the operating time..

It Just forces the RCD to mechanically operate by bypassing one half of the L & N coil windings.

It is a mechanical device in as much as the button is manually pressed to do the test and it is never automatically operated in any fault conditions.

But ALL of the symptoms I described in post#8
and the scenario of a N-E fault on the load side
with a PME/TN-C-S incomer
but no load current flowing will cause the Test button to NOT trip the RCD!


Go on try it yourself.. I can assure you its correct







I have tried it!!! :Blushing

[sellers]

I thought the test button switched on a load that takes the live from the outgoing live, and neutral from the incoming neutral, thus creating an imbalance. As the rcd will see the load being taken off the live but it not returning down the neutral as this is connect before the rcd on the incomming side.

[steptoe]

you are Incorrect there steps!...


see drawing on page 93 O-S-G.


If you stick a meter across the L&N on a dead RCD..
(e.g. not connected)

With the switch on and the test button pressed you typically get around 3400ohms.

Basically it just puts a diagonal path across the RCD to induce a trip current..

What it doesn't do is prove the accuracy of the trip current or the operating time..

It Just forces the RCD to mechanically operate by bypassing one half of the L & N coil windings.

It is a mechanical device in as much as the button is manually pressed to do the test and it is never automatically operated in any fault conditions.

But ALL of the symptoms I described in post#8
and the scenario of a N-E fault on the load side
with a PME/TN-C-S incomer
but no load current flowing will cause the Test button to NOT trip the RCD!


Go on try it yourself.. I can assure you its correct







I have tried it!!! :Blushing

I have never tried it, but I have known an RCD to fail a trip test then after a button test to pass with flying colours, my 2391 tutor told me it was because of a mech test,.!
obviously not then.
I stand[seated with beer actually] corrected.



ps, but with all MCBs switched off it will still trip, AKA no load, so I fail to understand your point just yet.

---------- AUTO MERGE Post added at 23:49 ---------- Previous post was at 23:44 ----------

I thought the test button switched on a load that takes the live from the outgoing live, and neutral from the incoming neutral, thus creating an imbalance. As the rcd will see the load being taken off the live but it not returning down the neutral as this is connect before the rcd on the incomming side.

that makes sense.

[NozSpark]

I had this argument with a tutor a good few years ago..

He said it's a mechanical test of the RCD

I said it's a test of the mechanics of the RCD

He said, that's what I mean't!

A RCD will trip if the test button is pressed without any load connected, but if you introduce a N-E fault with MCBs off (and PME - TNC-S supply) then it probably won't trip as some of the current will go down through the fault!

[steptoe]

turn your rcd on,
disconnect the N before the RCD and no fault, even thought he N is 230V,
BTW, only do this in a controlled enviroment.

[sellers]

Just to get my head around this, at 230v 30mA, means a 7666ohm short or less will trip the rcd.

now the test load say it be 7000ohm would trip with no loads connected. Then connect a N+E fault on a tncs system. With the n + e paths being of the same resistance the test fault would be taken down both conductors roughly equaly thus halving the fault current, not tripping the rcd?

[SPECIAL LOCATION]

Thought it time for a picture....





---------- AUTO MERGE Post added at 00:21 ---------- Previous post was at 00:16 ----------

Just to get my head around this, at 230v 30mA, means a 7666ohm short or less will trip the rcd.

now the test load say it be 7000ohm would trip with no loads connected. Then connect a N+E fault on a tncs system. With the n + e paths being of the same resistance the test fault would be taken down both conductors roughly equaly thus halving the fault current, not tripping the rcd?

Test load is typically 3400ohms so giving more like a 60+ma fault..

sometimes why your 30ma tester may not operate RCD then you do the test button..
it stuffs a bigger trip fault current down..
clears the cobwebs gets it working again..
after that it then works good with tester as well!

[sellers]

and the reason this doesn't happen on tns and tt, is due to electricity taking the easiest path. Which in that case would be down the neutral due to the higher resistance of the earth on those systems.