RCD discussion

GreeBo

<Mod Snip>

Mod note:
I have split this out from another thread as this is an important topic that comes up frequently.
Please keep it civil.

Thank you

Bruthal

Risteard81 wrote: »

They shouldn't. It's very basic stuff.

Imagine 3 phase rcds.

And besides that, he is right. Even if you believe you fully understand 3 phase to great depths, many don't.

Risteard81

Bruthal wrote: »

Imagine 3 phase rcds.

What about them?

Bruthal

Risteard81 wrote: »

What about them?

Will an upstream n-e short trip them?

Risteard81

Upstream neutral/Earth connection would be like a PEN conductor, i.e. a TN-C supply. So it won't be seen by the RCD.

Bruthal

Risteard81 wrote: »

Upstream neutral/Earth connection would be like a PEN conductor, i.e. a TN-C supply. So it won't be seen by the RCD.

I think you said before that the upstream n-e connection can desensitise the RCD. Can you explain this?

Risteard81

Bruthal wrote: »

Risteard81 wrote: »

Upstream neutral/Earth connection would be like a PEN conductor, i.e. a TN-C supply. So it won't be seen by the RCD.

I think you said before that the upstream n-e connection can desensitise the RCD. Can you explain this?

Faults in other installations and even supply network can cause issues depending on fault and circumstances. Faults can be imported through main bonding connections etc. so unusual things can happen which can make for interesting fault finding. Having a closed mind won't help resolve these issues.

Bruthal

Risteard81 wrote: »

Faults in other installations and even supply network can cause issues depending on fault and circumstances. Faults can be imported through main bonding connections etc. so unusual things can happen which can make for interesting fault finding. Having a closed mind won't help resolve these issues.

Almost politician like vagueness as usual when asked for a clear description. I like the "closed mind" one. You are some man at the avoiding questions if the answer is not in the rule book.

Anyway...

We know they are desensitized by a N-E short on the load side. If we take a 50/50 split of current, it will take 60ma for a 30ma imbalance to be seen if the L-E current takes the circuit CPC route. So the RCD is only half as sensitive in that example.

So, I have described how a load side N-E short can desensiitize an RCD. Clearly, concisely. No vagueness. If it is not clear enough, I can draw you a diagram.

So, with this open mind you have, can you describe in the same manner, how a supply side N-E short will desensitize a properly working RCD.

Bruthal

Risteard81 wrote: »

Having a closed mind won't help resolve these issues.

Im still amused by that closed mind one, but I suspect you dont actually understand the mechanisms of desensitized RCDs, or at least didnt until very recent discussions here. Possibly just read somewhere that the N-E fault can desensitize an RCD, and paste it in whenever a problem is posted.

If you did get it, you would easily see that a N-E short on the supply side is not likely to affect a properly working RCD circuits trip level sensitivity.

While it is true strange things can happen, (I had an RCD tripping on a switched off DB before caused by a N-E short at a socket), it would be strange indeed if a N-E short on the supply side of an RCD caused it to be less sensitive, since most installations here have a supply side N-E short.

Using terminology such as TNCS etc in a vague description of possible, but non described faults is meaningless. The load side of an RCD is a sort of 2 wire comparator setup, unaffected by anything until one of the load side conductors develops/finds a path other than the ones through the RCD.

This path will not occur by a supply side N to E short, as all load current through one pole of the RCD must still return through the other pole unless a load side conductor finds an alternative path, which will only happen on a load side fault, person receiving shock to earth etc.

2011

Risteard81 wrote: »

Faults in other installations and even supply network can cause issues depending on fault and circumstances.

Correct, but it does not explain how a neutral & earth connection upstream of the RCD can de-sensitise it. I assume that this was a typo?

Leaving that aside as you say external issues can cause a RCDs to trip, an example would be a transient over voltage condition. It is a topic worth exploring.

External transient over voltages can be caused by:
1) lightning strikes, even kilometres away can have a significant impact.
2) A fault condition in a neighbouring installation.
3) A sudden change in power requirements in a neighbouring installation.

Transient over voltages can cause RCDs to trip by:
• Causing a current of sufficient magnitude to flow to earth due to a failure of the insulation
• As the surge is high frequency it permits sufficient earth leakage current to flow due to the capacitance to earth.

Faults can be imported through main bonding connections etc. so unusual things can happen which can make for interesting fault finding.

Do you mean imported faults that would trip an RCD?

Bruthal

There are some situations where an RCD is undesirable probably.

We see it with fridge freezers now, using a spur to avoid it. A T bar socket would have been a good option there imo.

Or perhaps a submerged pump in a well or something also wouldnt benefit much from RCD protection.

But one thing is for sure, if someone believes a N-E fault on an RCD desensitizing it also means a shower load wont trip it because of this desensitizing, they are missing the paradox there.