Why does my RCD keep tripping?

[RF Lighting]

Yep. In fact we were messing with this today.

It's been the same on several completely different jobs we've worked on. Always brand new CUs bought from a reputable wholesaler. Neither of the twin RCD's seem to trip as a result of a deliberate or accidental N>E short.

The RCD's have always tested out fine with both mine and my mates RCD testers. Ze has been within normal parameters too.

 

[JohnW2]

Yep. In fact we were messing with this today. t's been the same on several completely different jobs we've worked on. Always brand new CUs bought from a reputable wholesaler. Neither of the twin RCD's seem to trip as a result of a deliberate or accidental N>E short. The RCD's have always tested out fine with both mine and my mates RCD testers. Ze has been within normal parameters too.

Hmmm - as I (and bernard) have said, that really makes no sense, as I'm sure you'll agree. The appearance (and magnitude) of an L-N balance (which should trip the RCD) as a result of an N-E short is a function of the circuits, loads on them, and possibly other network issues, and is in no way dependent upon the RCD itself. Do I take it that you are certain that some of the other circuits were loaded when you experienced this non-tripping?

Kind Regards, John

 

[RF Lighting]

Yep, today's install had a live lighting circuit with a 5' pop pack, and a 3 or 4 60W GLSs. My mate was working on the sockets with the live isolated at the breaker, and noticed an accidental short didn't pop the RCD, so we tried a couple of deliberate shorts on both RCD's and neither tripped.

 

[JohnW2]

Yep, today's install had a live lighting circuit with a 5' pop pack, and a 3 or 4 60W GLSs. My mate was working on the sockets with the live isolated at the breaker, and noticed an accidental short didn't pop the RCD, so we tried a couple of deliberate shorts on both RCD's and neither tripped.

Hmmmm, again - I'm really lost for words, and can but repeat that it really doesn't make sense, particularly when the loads were on the same RCD as the N-E short

Kind Regards, John

 

[ban-all-sheds]

In my experience, the incidence of tripping when a neutral is in contact with an earth connection also depends on the supply type. ... I have observed that the problem appears to be less prevalent in TN-S supplies than TNC-S.

Mine is TN-S.

Trips every time.

 

[ban-all-sheds]

I suspect that, like many before me, I experienced a very unpleasant feeling in my stomach, assuming that (despite all the precautions) I must have somehow been working on a live circuit!

You were.

 

[JohnW2]

I suspect that, like many before me, I experienced a very unpleasant feeling in my stomach, assuming that (despite all the precautions) I must have somehow been working on a live circuit!

You were.

Oh dear ... you know exactly what I meant

Kind Regards, John

 

[JohnW2]

In my experience, the incidence of tripping when a neutral is in contact with an earth connection also depends on the supply type. ... I have observed that the problem appears to be less prevalent in TN-S supplies than TNC-S.

Mine is TN-S. Trips every time.

As I said, it would seem that, if there is a significant difference, TN-S ought to be more likely to trip than TN-C-S - at least, if there are no loads within the installation. If there are any loads, I really don't see that there should be any noticeable difference.

With my TT (with loads) it certainly trips every time.

Kind Regards, John

 

[ban-all-sheds]

Oh dear ... you know exactly what I meant

Do you mean you disagree with the classification of neutral as a live conductor?

 

[JohnW2]

Oh dear ... you know exactly what I meant

Do you mean you disagree with the classification of neutral as a live conductor?

No, I meant that you knew exactly what I meant.

Kind Regards, John

 

[bernardgreen]

The amount of load current in the Neutral that by passes the RCD sensor depends on the impedance of the by-pass caused by the Neutral to CPC fault. This impedance will be less when the CPC is connected to the incoming Neutral at the cut out ( TN-C-S ) than it would be if the CPC was connected to a Network Earth ( TN-S ) which in turn was connected to the Network Neutral at the substation.

The one with the highest by-pass impedance TN-S will require a larger load current to get 30 milli Amp difference in the RCD sensor.

 

[JohnW2]

The amount of load current in the Neutral that by passes the RCD sensor depends on the impedance of the by-pass caused by the Neutral to CPC fault. This impedance will be less when the CPC is connected to the incoming Neutral at the cut out ( TN-C-S ) than it would be if the CPC was connected to a Network Earth ( TN-S ) which in turn was connected to the Network Neutral at the substation. ... The one with the highest by-pass impedance TN-S will require a larger load current to get 30 milli Amp difference in the RCD sensor.

That's all true but, of course, in most installations there will be loads drawing considerably more than 30mA, so the proportion of neutral current flowing in even bypass paths of (relatively) very high impedance will still usually be greater than 30mA. The fact that N-E shorts invariably trip RCDs in my TT installation (with an impedance around 100 times that of a TN-S earth) illustrates that. If it happens consistently with TT, I find it hard to believe that there would be an appreciable difference between TN-S and TN-C-S.

There is, however, the other consideration I mentioned - that with TN-S (but not TN-C-S) there can be a pd between N and E at the origin, which means that, even in the absence of of any loads on the installation, an N-E short downstream of the RCD could could the RCD to operate.

Kind Regards, John