'high security' ring main

[Dave Plowman (News)]

But how long do these take to trip?

[js.b1]

About 5 minutes given wiring accessory & wiring standards :-)))) Sorry, could not resist.

Thought GFCI were 10mA? Elsewhere 30mA was chosen as the lowest practical for economic mass production outside of Japan after WWII, may be a myth but perhaps logical re sweet spot on getting viable protection and market uptake.

10mA is available, but I think that tends to be used for situations like swimming pools or hazardous areas etc.

[John Williamson]

I was told when they fisrt came out that 50mA for no more than 50ms was a survivable shock for approaching 100% of the populace.

30/30 gave anacceptable margin for error. Of course, it may have just been a load of marketing hype......

[ARWadsworth]

30/30?

[John Rumm]

and livestock protection... horses etc are more susceptible to injury than us in certain circumstances.

[Steve Walker]

We have three PCs, three mains (not separate PSUs) powered printers and a whole lot of other stuff. Assuming each has a similar mains filter, that could be as much as 21mA already - we have suffered the occassional nuisance trip, so I can see why someone might well have a separate cicrcuit!

SteveW

--- Posted via news://freenews.netfront.net/ - Complaints to [e-mail]

[John Williamson]

Disconnecting after 30 milliseconds at 30 milliamps of current imbalance between live and neutral cables.

[John Rumm]

30 ms is an unlikely trip time on 50Hz - RCDs will normally trip within two mains cycles - or 40ms. You might manage 30ms in the US on 60Hz supplies.

[ARWadsworth]

A 30mA RCD must trip within 200ms at 30mA. The 40ms trip time must be met at

150mA.

[John Rumm]

Indeed. Probably also worth pointing out that in reality many RCDs will exceed the requirement and trip within one cycle, so real world trips in < 20ms are quite common IME. Also a 30mA trip RCD can in fact trip on as little as 66% of that rating - so 20mA (but not less than 50% IIRC)

[Andrew Gabriel]

Don't think I've ever found a working one which takes longer than 20ms to trip, and at 10ms resolution, I sometimes see a trip time of zero.

[ARWadsworth]

Usually between 20 and 28 ms at 30mA in my experience. 20ms or below for 5x testing.

My new test case says that ALL RCDs trip at zero:-(

To be fair it also claims that the test probes have a resistance >500Mohm.

[NT]

I've never understood how using 2 earth terminals per socket was supposed to improve anything. If an earth terminal comes undone, with a single earth terminal you've then got 2 CPCs that are twisted tgoether, which connects, but with 2 terminals you've lost the cpc entirely.

NT

[Dave Plowman (News)]

Think twisting conductors is frowned upon. But not here.

[Andrew Gabriel]

Agree - it seemed slightly silly to me, although people don't twist today's single core conductors.

I try to put in a ring circuit such that the conductors aren't cut through at any wiring accessory, so that there's no adding up of resistances of terminal resistances if they start going high, and the ring is one long continuous length of cable with the ends joined at the CU, with just the insulation stripped where required.

I always wire ring circuit earths at the CU according to the old High Integrity Earthing rules in 16th Ed, where each end goes into a separate terminal. That does seem sensible.

[John Rumm]

Remember with high integrity earthing, the CPC is a ring, even on a radial circuit. Hence there are two paths to earth - and both connections would have to fail to lose the earth at the socket. A complete failure at one socket would not affect the earthing of the others.

(As an aside, the wires should not be twisted).

[Dave Plowman (News)]

Is the reason simply they may break - at some point in the future if not at the time?

[ARWadsworth]

It makes fault finding difficult.

[dennis]

An interesting idea.. but flawed. Unless you test the ring periodically it will degenerate into a radial without the ring at the first fault and the user will never know. In fact unless its tested you may never have the ring and the user wouldn't be able to tell.

It really is a problem designing systems with redundancy that actually work, and is impossible if you can't tell a fault has occurred.

[NT]

If you dont understand anything, why comment