Phase to Earth short and RCds

[Toby]

Hi,

I have been helping a mate paper his hallway, during which we managed to short the live cable to earth at a light switch as it was removed (it came lose as the switch was unscrewed)

There was a flash and the 6A MCB tripped, but the 30mA RCD did not trip.

His consumer unit is three way, 2 x 30mA RCDs and an non RCD section. The downstairs lights are on the same RCD as the upstairs sockets and it trips when the test button is pressed.

I was thinking of getting a Fluke SM300 as it is only £27 has a basic 30mA RCD tester, so will be able to at lest do a basic test on it, but is this behaviour expected?

I didn't think to see what earthing he has, but I am fairly sure it is TN-S (the armour of the supply cable)

On another point, some of his light switch back boxes are wood, I assume these must be changed?

Thanks!

[ARWadsworth]

I have seen that happen before. And all the tests confirmed that the MCBs and RCDs were in perfect working order.

I am not familiar with the SM300 but a quick look suggests that it will only test that an RCD trips in 300mS so the actual result would not mean a lot.

A trip to Maplins might save a few quid to see if the RCD does trip at the required current imbalance.

I have a TN-S supply and I have a lower Ze reading than most people with a TN-C-S supply. Don't worry about the supply (unless it is a TT).

Before or after you have papered the walls:-)?

[John Rumm]

Basically you have a race condition here - The fault current will trip the MCB on the magnetic part of its trip response (i.e. < 0.1 sec) and the RCD would also trip due to leakage > 30mA. Its unlikely that both will trip since either opening will clear the fault seen by the other.

Chances are its ok then.

The behaviour is not abnormal.

There is no requirement to change them (in the same way there is no requirement to bring anything up to current standards unless modifying it anyway). In the grand scheme of things, a wooden backbox (especially if made from something relatively self extinguishing like oak) is not much to worry about - especially with plastic faced switches.

[js.b1]

If you have a wall bare in a hall, and are staying, it is worth doing some basic maintenance.

#1 Channel the wall for oval conduit cable drops. So making future cable replacement easy, pre c.1966 lighting may lack a CPC (earth conductor). If the plaster is deep enough, and often is on older houses, this is a trivial tasks for a lighting circuit.

#2 Replace wooden backboxes. If the circuit lacks an earth you need to use a Class-2 (plastic) backbox such as MK ESU/9/ML or use a Class-1 (metal conventional) backbox with a 4mm 6491X earth back to the consumer unit. You can use plastic M3.5 screws with a conventional metal backbox, if you can find them.

#3 Check the light fitting for wear. Both old & new light fittings can get a bit shabby - and dangerous.

If the old hall lighting circuit uses 2 core cable it may well borrow a live off the downstairs light circuit, bad practice, usually indicated by both up & down light circuits being supplied by the same RCD (otherwise it will trip when either hall light switch is cycled).

Regulations are not retrospective, however maintenance is worth considering if it avoids costly redecoration in the future. Maintenance such as installing oval conduit, replacing a faulty light switch or backbox is permitted anywhere under SI2006 and well worth doing if the wallpaper is/will be expensive.

[ARWadsworth]

Toby's post is suggesting that there is an earth on the lighting circuit and that there is a "17th edition" CU in place. So someone has done some modifications.

Now on a positive note it is quite easy to swap the wooden back boxes for metal ones. My prefered tool for such a job is a wood chisel (an old knackered one is ideal). Use this chisel to implode the existing wooden back box and the metal one can usually by pushed into place (sometimes with the help of a hammer).

I have such a job coming up soon. I might take photos if I remember:-)

[NT]

A plug with a resistor in is way cheaper.

its a case of whichever gets there first.

no need. If whats mounted on them is 1950s though, I'd definitely consider changing the lot.

NT

[Toby]

Yep, that's right, the one that shorted is already a metal backbox, with a CPC, there are others that are made of wood in other rooms.

Thanks for the advice :-)

[Toby]

Thanks NT,

I think I will build a resistor box, with various switches on it to ramp up the leakage, although, for the sake of £12.75, I might just buy this and save the time and effort...

[NT]

NT

[alexander.keys1]

I use a lampholder with a pair of safety test leads connected, and a clear 60 watt candle bulb. This instrument, if you know what you are doing, can test for:

Circuit live, L-N present & correct (probes to brown & blue wires, bulb illuminates) L-E present & correct (probes to brown & green/yellow wires, bulb illuminates if no RCD, if bulb dim check earthing, RCD may need to be fitted) RCD present & functional (probes to brown & green/yellow wires, filament glows briefly before RCD trips) N-E correct (probes to blue & green/yellow wires, filament does not glow, RCD trips)

One thing it won't detect is a high earth loop impedance, due to the low current drawn by the bulb, but it gives you a better idea than those plugs with three neons that are claimed to check correct wiring.