paul8f
Member
- Joined
- Jun 17, 2014
- Messages
- 5
- Reaction score
- 0
+++++++++++++++++++++++++++++++++++++++++++++++++++
Originally added to old discussion;
http://talk.electricianforum.co.uk/topic/17562-rcds-in-series/
Split off as new question due to age and length of old thread.
Doc H.
+++++++++++++++++++++++++++++++++++++++++++++++++++
Hi all. Just joined this forum so take it easy on the newbie....!
Hope nobody minds if I re-open this old can of worms. I'm in the middle of installing a bathroom fan (220v, ceiling mounted) in my house. Since the bathroom is the most dangerous room in the house to receive a shock, I was about to wire the fan up through an unswitched 3A spur outlet with RCD built-in, along with a 3 pole isolation switch. All the lights and sockets in the house are protected by the one RCD, and I thought there was no problem with wiring up another RCD for the fan..... that was until I stumbled on this discussion.
Does anyone know if there has been an update to the UK regulations since this discussion finished? I'm in Ireland by the way, and must try get my hands on the latest regulations here. Also, the earthing system here is mostly TN-C-S, if that makes any difference.
Anyway, I wondering if people's opposition to having two RCDs in series might be somehow down to the phenomenom of "bouncing contacts" within the RCD itself? No switch is ideal, and when the contacts open in the RCD during a Live/Neutral imbalance, the circuit will make-and-break a few times before it finally stays in the open position (all happening in the fraction of a milli-second). I suppose the worst case scenario would be when two identical RCDs get connected in series. Here, the trip times of each RCD will be very close. When an insulation problem occurs (or when someone touches a live wire), the two RCDs will be in a race with each other. Current Ib induced from the toroidal core of each RCD will begin to activate the magnetic switches. As soon as the "leading RCD" contacts begin to open, the flux on the toroidal core of the slower "trailing RCD" will be affected. Could this result in a sort of "contagious flip-flopping" of contacts, and end up affecting the overall trip time (i.e. two RCDs, each with an average trip time of 28mS, could end up taking 34mS before one of them finally overcomes the contact bouncing to stay in the open position)?
If this is the case, a workaround might be to ensure you never install two similar RCDs in series. If RCD1 trips at an average leakage current of 29mA after 27mS make sure that RCD2 trips at an average of 20mA after 17mS. That way if a fault ever occurs downstream of RCD2, then RCD2 should trip before RCD1 in 99% of cases. Bouncing of the contacts would be minimised, and the overall trip time would be equal to (or marginally longer than RCD2). If RCD2 ever fails, RCD1 will still give you 30mA protection.
Overall my logic is telling me that in most systems the RCD is one of the weaker links due to their high failure rate. So having two in series will add redundancy and increase safety. If one doesn't trip, the other one will. My background is Electronics, so please forgive my ignorance of the figures quoted in my examples! I'd be interested to see if anybody agrees or disagrees with my theory of bouncing RCD contacts.
(...as a side note... if RCDs in series is such a no-no, then why do so many DIY stores sell extension leads with RCD protection built into the plugtop???)
Originally added to old discussion;
http://talk.electricianforum.co.uk/topic/17562-rcds-in-series/
Split off as new question due to age and length of old thread.
Doc H.
+++++++++++++++++++++++++++++++++++++++++++++++++++
Hi all. Just joined this forum so take it easy on the newbie....!
Hope nobody minds if I re-open this old can of worms. I'm in the middle of installing a bathroom fan (220v, ceiling mounted) in my house. Since the bathroom is the most dangerous room in the house to receive a shock, I was about to wire the fan up through an unswitched 3A spur outlet with RCD built-in, along with a 3 pole isolation switch. All the lights and sockets in the house are protected by the one RCD, and I thought there was no problem with wiring up another RCD for the fan..... that was until I stumbled on this discussion.
Does anyone know if there has been an update to the UK regulations since this discussion finished? I'm in Ireland by the way, and must try get my hands on the latest regulations here. Also, the earthing system here is mostly TN-C-S, if that makes any difference.
Anyway, I wondering if people's opposition to having two RCDs in series might be somehow down to the phenomenom of "bouncing contacts" within the RCD itself? No switch is ideal, and when the contacts open in the RCD during a Live/Neutral imbalance, the circuit will make-and-break a few times before it finally stays in the open position (all happening in the fraction of a milli-second). I suppose the worst case scenario would be when two identical RCDs get connected in series. Here, the trip times of each RCD will be very close. When an insulation problem occurs (or when someone touches a live wire), the two RCDs will be in a race with each other. Current Ib induced from the toroidal core of each RCD will begin to activate the magnetic switches. As soon as the "leading RCD" contacts begin to open, the flux on the toroidal core of the slower "trailing RCD" will be affected. Could this result in a sort of "contagious flip-flopping" of contacts, and end up affecting the overall trip time (i.e. two RCDs, each with an average trip time of 28mS, could end up taking 34mS before one of them finally overcomes the contact bouncing to stay in the open position)?
If this is the case, a workaround might be to ensure you never install two similar RCDs in series. If RCD1 trips at an average leakage current of 29mA after 27mS make sure that RCD2 trips at an average of 20mA after 17mS. That way if a fault ever occurs downstream of RCD2, then RCD2 should trip before RCD1 in 99% of cases. Bouncing of the contacts would be minimised, and the overall trip time would be equal to (or marginally longer than RCD2). If RCD2 ever fails, RCD1 will still give you 30mA protection.
Overall my logic is telling me that in most systems the RCD is one of the weaker links due to their high failure rate. So having two in series will add redundancy and increase safety. If one doesn't trip, the other one will. My background is Electronics, so please forgive my ignorance of the figures quoted in my examples! I'd be interested to see if anybody agrees or disagrees with my theory of bouncing RCD contacts.
(...as a side note... if RCDs in series is such a no-no, then why do so many DIY stores sell extension leads with RCD protection built into the plugtop???)
Last edited by a moderator: