RCDs and the Importance of the Test Button.

[telectrix]

done a small job today. just a socket & SFCU ( for a light) in customer's shed, on existing feed. feed cable is 2.5mm on a 16A MCB, RCD protected. cable tested OK, installed bits. Zs 2.4Ω. ramp test of RCD, mft reads>51mS. prerssed test button on RCD (which the customer said had never been done or RCD tripped in 15 years). then re-tested. ramp 27mS. times... 32mSec@ x1. 9mSec@ x5. just goes to show that they can get sticky.

 

[Deleted member 72666]

Just think of all those wasted labels!

 

[GLENNSPARK]

I've come across many RCD's that fail on disconnection times and on a leakage current ramp test but then the pass after operating the test button a few times.

Surely if the customer isn't pressing the test button periodically then the RCD should be replaced if it fails tests before the button is operated because it's failing in real life conditions and even if it passes eventually after the test button is pressed, in reality it will not meet disconnection times again after another few months of nobody touching it.

if the RCD fails to operate 1st time using the manual test button then i was always told it should be condemned...but as has been pointed out....thousands of RCDs would then have to be condemned....a few goes usually loosens em up....
another one as well is an RCD should be replaced everytime it clears a fault....
suppose that puts testing them out to grass then....

 

[telectrix]

bloody hit and miss. either they work or somebody gets fried. who's responsible? sparky who tested it 6 months ago or the manufacturer?

 

[GLENNSPARK]

bloody hit and miss. either they work or somebody gets fried. who's responsible? sparky who tested it 6 months ago or the manufacturer?

well i suppose this would be covered by the same rules that apply to ECRs...in that it was good for continued service when the test was conducted

this i think would be down to type testing, recall etc...

 

[Adam W]

I was told on a recent 2395 course to test the rcd's before pressing the manual test button and one of the reasons amongst others of doing it in that order was that it could influence the test results. That argument could be difficult to prove or disprove I suppose.

I was told the same, I think it was on my 2330 - the sequence of tests (in order) should be:
x1, (both sides) x5 (both sides), x0.5 (both sides), push test button, optional ramp test.

As you say, then you can't 'influence' the results by warming it up by pushing the test button, and having already been tripped several times by the time you come to do the x0.5 test it's less likely to hide any 'nuisance tripping'.
I heard of an electrician who went to fail an rcd in a stage installation and was told "yes it will have failed - rcds need to be warmed up", suggesting you need to pre-empt when the fault will be and go in and warm up the rcd beforehand. I would have said 'warming it up' quarterly (or as a worst case scenario, annually) as recommended should be enough.

 

[Adam W]

stakes are for pinning apprentices to the workplace.

They're things you plant in the garden aren't they?
Stakes I mean, not apprentices.

 

[GLENNSPARK]

They're things you plant in the garden aren't they?
Stakes I mean, not apprentices.

thought this was something to do with racing...

 

[Adam W]

thought this was something to do with racing...

I like a nice Sirloin. Rare.

 

[GLENNSPARK]

I like a nice Sirloin. Rare.

softened up first with one of these:

View attachment untitled.bmp

 

[Sharpend]

This rcd fail issue is a clear explanation as to why relying on Rcd's is fundamentally wrong, the things should be banished, a better approach needs to be sort. RCDs only give a false sense of security. The manufacturers would dissolve all responsibility as they would say that their instructions had not been followed.
Whilst I agree that after several manual trips the things work within specified parameters, they should be replaced. Perhaps the manufactures should employ quarterly rcd testers. Either that or every one that fails should be returned for a free replacement.

 

[stuartcourt]

why are we not told to at least put down mcbs every now and again as they can be up for years also

 

[GLENNSPARK]

This rcd fail issue is a clear explanation as to why relying on Rcd's is fundamentally wrong, the things should be banished, a better approach needs to be sort. RCDs only give a false sense of security. The manufacturers would dissolve all responsibility as they would say that their instructions had not been followed.
Whilst I agree that after several manual trips the things work within specified parameters, they should be replaced. Perhaps the manufactures should employ quarterly rcd testers. Either that or every one that fails should be returned for a free replacement.

agreed....if you cant get a compliant Zs without RCD then theres something wrong....i`v seen several previous schedule of tests now when on an ECR with 1667 scrawled down the loop boxes.....not acceptable to me as its supposed to be additional protection on TN...i wouldn`t accept it for fault on a TN....which is effectively what it`s doing if Zs hasn`t been met without it...

 

[trev]

yes....
and how many customers ............. are shown functional operation on handover...

All of mine mate, whether they take any notice of it is another story

 

[stuartcourt]

my old boss would run people through and then say right now you tell me how it works/what to do no one ever could

 

[Brightspark2]

HI,
The Company I used to work for had us pressing the test buttons on RCD's on over 200 sites every 3 months ( as per manufacturers instructions ). Needless to say we integrated this with other maintenance work. Once a year we tested the tripping times of the RCD's and tested the button last. As Tel says they can get a bit 'sticky'.

Regards.

Ha I do this on a couple of sites...just walk round the site pressing the buttons, nice and easy...