Testing an RCD on a portable distro caused main RCD to trip

[numpty with a crowbar]

Evening folks,
I've built a couple of portable distro's (32A plugs to small CU's with RCD's and MCB's out to 13A sockets etc), and 3 of 4 are fine, the RCD tests ok and no bother.
The 4th one is causing a bit more of a strange 'un.... It's a proteus 40A RCD with a functional earth cable (as detailed here - //www.diynot.com/forums/viewtopic.php?t=88594 ) and as advised I connected the earth cable to the earth bar.... In order to test the distro I've plugging it into a socket in the house (via a 13A to 32A adapter), however when I test the proteus RCD (with the test button) it trips out, and also trips the main RCD in the CU in the house (and I get into trouble for the power going off ).

Is the RCD FUBAR'd or is this ok for that type of RCD (all the others test fine and don't have a functional earth). And should I chuck it???

Cheers

 

[RF Lighting]

This is normal. (But not good in a fixed installation setup)

When you press the test button the RCD leaks >30mA to earth which causes the RCD to trip, as this RCD type has the functional earth.
The other RCDs leak >30mA from the monitored side to the supply side which will cause the RCD being tested to trip, but this can't be detected by the RCD in your CU.

As (I assume) both RCDs are 30mA rated, and neither are time delay then there is no discrimination between the devices, so in the event of a fault (albeit a deliberate one) either device may operate first, or both at the same time.

 

[Spark123]

Most RCDs put an imballance across the coil by means of a resistor between live out and neutral in or vice versa operated by a momentary push button, these are probably what your other devices are. Does this device have a diagram on it? Can you post a pic, I can't make it out from the pic on your other post.

 

[numpty with a crowbar]

This is normal. (But not good in a fixed installation setup)

When you press the test button the RCD leaks >30mA to earth which causes the RCD to trip, as this RCD type has the functional earth.
The other RCDs leak >30mA from the monitored side to the supply side which will cause the RCD being tested to trip, but this can't be detected by the RCD in your CU.

As (I assume) both RCDs are 30mA rated, and neither are time delay then there is no discrimination between the devices, so in the event of a fault (albeit a deliberate one) either device may operate first, or both at the same time.

Yep - both are 30MA RCD's, so it looks like I've got nowt to worry about then - much appreciated....

Just out of curiosity - what would happen if I disconnected the functional earth and terminated it in a choc-block?? Would it render the RCD unsafe??

 

[RF Lighting]

Having done more research and following a technical discussion with a fellow spark, my first reply is incorrect regarding how the test button operates with the functional earth.

Do not disconnect the functional earth. That would make it an 'un-fuctional' earth.

What brand are the other RCDs?

 

[Spark123]

The functional earth is there for a reason, it is part of the RCDs design and hence should not to be disconnected. If need be, purchase an RCD without a functional earth.

 

[numpty with a crowbar]

Do not disconnect the functional earth. That would make it an 'un-fuctional' earth.

No problem

What brand are the other RCDs?

2 are MK's, the third's a proteus without a functional earth

If need be, purchase an RCD without a functional earth.

I'm thinking I might.... Don't really want to trip the main CU RCD every time I test the thing... I'll get into trouble from the better half

 

[Taylortwocities]

Don't you have circuits in your house that are on the non-rcd side of the board? Maybe you could use one of those to do the testing with...

If you dont and if you are going to do a lot of testing why not pop in a 16A cb on the non rcd side with a socket on it?

 

[sm1thson]

Don't you have circuits in your house that are on the non-rcd side of the board?

but presumably the distro will get used at other locations? last thing you want is someone pressing the test butting and killing the power at where ever you are using the distro!

 

[numpty with a crowbar]

Don't you have circuits in your house that are on the non-rcd side of the board?

but presumably the distro will get used at other locations? last thing you want is someone pressing the test butting and killing the power at where ever you are using the distro!

That's what I'm worried about... The distro should really be tested at every location it's used at, but I'm not going to be mr popular if I keep tripping RCD's wherever I go

Methinks I'm going to have to get another RCD... Rats....

 

[Taylortwocities]

It may not help. If the upstream RCD trips quicker than the one in the distro you'll be shutting down someone's power. Still, its energy saving, I suppose...

 

[sm1thson]

It may not help.

it will help if the RCD he replaces it with has a test button that shorts (via a resistor) 'live post RCD to 'neutral pre RCD', rather than live to earth as he appears to have. 'live post RCD to 'neutral pre RCD' will not effect any RCDs down stream.

 

[Taylortwocities]

Ahh yes

 

[plugwash]

you could connect the funtional earth to the incoming neutral, this would mean it wouldn't detect a raised neutral but the majority of RCDs in common use cant and it really doesn't seem to be a problem.

 

[bernardgreen]

it wouldn't detect a raised neutral but the majority of RCDs in common use cant and it really doesn't seem to be a problem.

Depends on where the disco might be used. At an outdoor event or a marquee the feed may be missing its earth, have phase and neutral reversed. At least using the device to trip on raised or missing neutral would reduce the risks involved with temporary ( and therefore sometimes bodged ) supplies.