RCD spurs that trip faster than the MCB RCD

[JohnW2]

It would not stop the leakage but it would stop tripping as a result as the neutral to earth leakage is not going through any RCD device.

Indeed - but I said nothing about trips, only that leaving an N-E fault 'uncleared' (by a SP RCBO) could lead to "potential problems". For example, if the N-E fault were of very low impedance, then a substantial proportion of the return current from all the loads in the installation could try to go via the E, which could be a 'threat' to CPCs.

Not quite, 3.5 mA is the limit for most things, so at least 3 sockets per circuit, ....

Well, for a start, if you really wanted to try to stick to a "9 mA maximum", that would mean a limit of two outlets (e.g. one double socket) per RCBO - which would obviously be daft in most situations. In any event, whilst you say that "3.5 mA is the limite" that certainly doesn't mean that no loads (particularly older equipment) will have greater leakage than that - so you really would have to restrict yourself to one single socket per RCBO (and not use any double sockets) - which would be even more daft!

But the 10 mA MK socket we had a work increased the number of trips of the 100 mA and 300 mA trips, as people would press the test button, and it would trip all three every time.

That is, of course, inevitable, given the way that test buttons on RCD sockets work (by introducing an N-CPC 'fault'). I suppose the idea is that the test button tests for an adequate earth connection, as well as RCD functioning. which is no bad thing.

But what you need is for back ground leakage to be below 5 mA, as a 30 mA should trip between 15 - 30 mA so if there is over 5 mA leaking then it could still trip the house one first.

Sceptical though I can be about the 'new fangled', I suppose there is potential scope for technology to address this whole issue. If residual current devices could be made which tripped in response to a sudden increase in residual current (by 30 mA, 10 mA or whatever), then all these issues/discussion about 'residual leakage' (hence some of the arguments in favour of RCBOs) would go away.

The other is knowing it has tripped, my freezer has a light on when powered up, but I tend not to notice lack of a warning light, my emergency touch lights with a power cut, I notice that straight away, so I investigate and turn power back on ....

As I've often said, whilst I have had freezers cease functioning because they have died (sometimes tripping a protective device in the process), I have personally never had one stop functioning for an appreciable period due to an unnoticed loss of power - at least partially because I would never power a freezer from a dedicated circuit' or even from a circuit whose failure would not be rapidly apparent to occupants.

As you split the home into smaller and smaller circuits, the chance of not realising when the power has tripped increases. I did at one point consider a dedicated circuit for freezers, but that would mean less likely to notice when it trips.

Exactly. Well-intentioned though it may be, the practice (and regulatory encouragement) to split installations into ever-increasing numbers of circuits considerably increases the risk of loss of power to one circuit going unnoticed. That is why I would always encourage peopleto put things like freezers and alarms (e.g. smoke alarms) on 'widely used' circuits, and never on dedicated ones.

I was surprised how quickly a freezer defrosts. I was in the kitchen when our freezer went bang, it had clearly just stopped, and I had a spare in the garage, so turned on the spare and waited for it to cool down before starting the transfer of food, maybe ¾ hour, when I came to move food, found food at top of freezer had already started to defrost. I would assume it went bang as the motor tried to restart after the defrost cycle, and the heat had risen to top only from the compartment with the cooling bits in, and since no fan running it only affected food at top?

I would imagine that effective cooling of the freezer must have stopped long before it 'went bang'. My experience, and that of others, is that, provided the freezer is not opened, a freezer will usually keep things frozen for at least 24 hours, probably more.

Kind Regards, John

 

[JohnW2]

A few years ago, I bought and installed a cheap, digital thermometer, with a settable temperature alarm, and a remote wired sensor. The idea was one of the alarm being triggered, if the freezer door were left ajar, but it would also sound if the supply, or compressor etc. failed.

As I've often said, I have over-temp alarms on all my freezers. They are generally set to alarm at around -8 to-10 °C - and, from what I can gather, it ought to take 12 hours or so after 'failure' for a (closed) freezer temp to rise to that extent.

Over the years, these alarms have 'saved' me (well, the freezer contents!) on at least three occasions - none of which were due to a failure of the circuit powering the freezers.

Kind Regards, John

 

[JohnW2]

Yes, I've been lucky.

If it's a matter of luck, that means we've both been lucky.

I inherited a house with a Hager RCD incomer CU. It had been added to so that there are now multiple circuits fed by one breaker, but over the nearly 24 years we have been here, I have been able to trace the reason for every tripping occurrence. Faulty dryer, water in junction box, those kinds of issues. There have been very few, from memory, I think you can count them on one hand.

As I've said, my experience with dual-RCD CUs - like you, in the past 30+ years I have experienced extremely few, if any, unexplained (i.e. 'nuisance') RCD trips. My daughter has, like you, a CU with a single RCD incomer and, to the best of my knowledge, has suffered no 'nuisance trips' in the ~17 years she's had it.

I'm therefore not convinced that it's a matter of you and I being lucky, but perhaps more a question of the likes of eric having been very unlucky.

Kind Regards, John

 

[SUNRAY]

You may well be right. My earliest RCDs looked like that.

Again, you may well be right. However, from when they became available to install in domestic CUs, 'single-width' RCBOs invariably only had single pole switching (and are what most of us have), but single-width DP ones now seem to be available.

Kind Regards, John

No sorry I think I disagree with that, I was certainly installing double width RCBOs before the singles. Mid 90s I was doing nightclub installations and typically installing a CU at the DJ area with multiple RCBO's, these were standard domestic style CU's with a 2 pole bussbar. I've had a quick look at photos from the time but I have vanishingly few of that work. For display stands at NEC RCBOs were a requirement (Although most seemed to get away with "alternative arrangements") and we used 4 20A per CU.

 

[JohnW2]

No sorry I think I disagree with that, I was certainly installing double width RCBOs before the singles.

I don't think that we are disagreeing.

I'm pretty sure that what you say is right. If you look, I was talking about the time when single-width RCBOs became widely available (which was after earlier double-width ones) - and was saying that, until relatively recently, the single-width ones all had SP switching. Do you disagree with any of that?

Kind Regards, John

 

[ericmark]

Commercially I have used double pole RCBO's and used the converters which clamp to the MCB, but with domestic there are loads of double pole now, but early versions were single pole switching even if they detected all live wires.

 

[JohnW2]

Commercially I have used double pole RCBO's and used the converters which clamp to the MCB, but with domestic there are loads of double pole now, but early versions were single pole switching even if they detected all live wires.

There may well be "loads of DP 'domestic' RCBOs" theoretically available now, but I think that's a fairly recent situation, and I wonder how many have actually been installed in domestic CUs. There must be millions of RCBOs in domestic CUs by now, but I wouldn't mind betting that the vast majority have SP switching.

.... but early versions were single pole switching even if they detected all live wires.

I presume you are talking about the over-current sensing, since the residual current part obviously has to use "all live wires".

I have to wonder why one would really need over-current sensing in both L and N, since, at first sight, it would seem that one would probably have to scrape barrels pretty deeply to think of scenarios in which there would be over-current in N but not in L. Am I missing something?

Kind Regards, John

 

[davelx]

I have to wonder why one would really need over-current sensing in both L and N, since, at first sight, it would seem that one would probably have to scrape barrels pretty deeply to think of scenarios in which there would be over-current in N but not in L. Am I missing something?

Kind Regards, John

IAnd in that case wouldn't the RCD function trip the breaker anyway due to the current imbalance? So, as you say, no need (or benefit, as far as I can see) to having over-current sensing in the N line.

 

[JohnW2]

IAnd in that case wouldn't the RCD function trip the breaker anyway due to the current imbalance?

Yes, you're right. I hadn't even thought of that but, as you imply, that seems to make it 'doubly unnecessary' to have overcurrent sensing in both lives.

So, as you say, no need (or benefit, as far as I can see) to having over-current sensing in the N line.

Exactly.

Kind Regards, John

 

[SUNRAY]

I don't think that we are disagreeing.

I'm pretty sure that what you say is right. If you look, I was talking about the time when single-width RCBOs became widely available (which was after earlier double-width ones) - and was saying that, until relatively recently, the single-width ones all had SP switching. Do you disagree with any of that?

Kind Regards, John

No. That works perfectly for me.

 

[SUNRAY]

Yes, you're right. I hadn't even thought of that but, as you imply, that seems to make it 'doubly unnecessary' to have overcurrent sensing in both lives.

Exactly.

Kind Regards, John

Even now we have specs for control panels which still state 2P MCB mechanically linked to RCD, these consultants really do not listen.

I've lost count of the number of times I've constructed panels with a control voltage (typically 24V) transformer with 2 such combinations around the transformer. I've shown the earth in 3 different locations as different consultants/designers have different ideas but of course only one would be used:

So 8 modules of DIN rail space when 2 will do and a lot cheaper. and often for a transformer of only 25VA although the same arrangent would also be used up into KVA sizes.


Albeit this if veering well away from DIY again.

 

[JohnW2]

Even now we have specs for control panels which still state 2P MCB mechanically linked to RCD, these consultants really do not listen. ... I've lost count of the number of times I've constructed panels with a control voltage (typically 24V) transformer with 2 such combinations around the transformer.

I somewhat struggle to understand how there could ever be need or justifivation for an RCD (whether or not mechanically linked to an MCB) in a 24V system. Am I missing something?

I've shown the earth in 3 different locations as different consultants/designers have different ideas but of course only one would be used:

Hmmm if the only earth used was the right-most one you show, the RCD could never operate, could it (because the 'supply' is not earth--referenced)?

Kind Regards, John