Hager RCD weirdness

[brummydave]

Just got off the phone with Hager technical who said they'd "never heard of the fault I was describing and couldn't help". He didn't even suggest buy a replacement!

The story goes...

While testing an old split-load board, the RCCB stopped working. It's a 63A, 30mA one and the board was made in 12/99 - there's no labels or paperwork to say when it was fitted.

Firstly I pressed the Test button and it functioned correctly.

Secondly I plugged my Fluke 1653 into a socket circuit supplied from the RCCB and set the RCD time tests to Auto. This gave satisfactory results at 1/2x, 1x and 5x with the RCCB being reset just fine after each.

Finally I did the ramp test to see what current it was tripping at and got Err4 on 0 degrees, so decided to do it again and got 24mA.

I then did the ramp test again but at 180 degrees as I often find differing results. However, I went to the RCCB to reset it and although the dolly moved, no power went through. The Test button no longer functioned either.

Anybody else had this kind of experience?

Hager Tech did also say that ramp testing isn't a required test and so I probably shouldn't have done it. Maybe the ramp test broke the RCCB?

When I have a remedy I'll update this. My thoughts are to double check the wiring and then fit a replacement if still no joy.

 

[Sharpend]

It'll be a faulty connection, often had it where it doesn't sit on the bus bar correctly. Or hasn't been tightened to correctly (maybe this is why they introduced torque drivers???)

Other than that it'll be a faulty one. 

 

[Tony S]

I’d go with the faulty connection. Each successive test has jolted the terminals until one had given up.

 

[Adrian]

Tap the side with handle of  screw driver 

 

[Evans Electric]

That last post is worth noting ....   tap it with a screwdriver  is not so daft as it sounds .

I had at least one  brand new  Crabtree  RCD that tripped  for no apparent reason ....after two trips back to the job investigating to no avail  ..... I swore at the offending device then tapped it with a driver to teach it a lesson and it tripped out .     Reset it .... the front door slammed shut and it tripped again .    Cause identified , replaced RCD , problem solved .

I think the tapping test should be added to the recommended test procedures .

1)  Carry out Half rated current test .

2) Carry out  100% rated curent test .

3) Try test button .

4)  Hit device with large screwdriver.

 

[Sharpend]

How would you record the test result on the Niccy cert? Gentle, Firm, Bust ?? :facepalm:

 

[NozSpark]

That last post is worth noting ....   tap it with a screwdriver  is not so daft as it sounds .

I had at least one  brand new  Crabtree  RCD that tripped  for no apparent reason ....after two trips back to the job investigating to no avail  ..... I swore at the offending device then tapped it with a driver to teach it a lesson and it tripped out .     Reset it .... the front door slammed shut and it tripped again .    Cause identified , replaced RCD , problem solved .

I think the tapping test should be added to the recommended test procedures .

1)  Carry out Half rated current test .

2) Carry out  100% rated curent test .

3) Try test button .

4)  Hit device with large screwdriver.

I don't mind the tapping method as a test, but tapping a device to free it up and then leaving it as working is just plain wrong

and my order of tests is different

1x

5x

Ramp

Manual

1/2x

 

[Tony S]

Take a RCD apart and you’ll see the latch/trip mechanism. It’s delicate and prone to wear so vibration won’t help.

 

[Rob_the_rich]

I don't mind the tapping method as a test, but tapping a device to free it up and then leaving it as working is just plain wrong

and my order of tests is different

1x

5x

Ramp

Manual

1/2x

my fluke does auto test, so I just stand at CU resetting the rcd, then just scroll through the results. Is there any reason why I should do them manually with 1/2x test last?

 

[Evans Electric]

I don't mind the tapping method as a test, but tapping a device to free it up and then leaving it as working is just plain wrong

and my order of tests is different

1x

5x

Ramp

Manual

1/2x

Just saying ....the RCD was replaced .....   

 

[Evans Electric]

How would you record the test result on the Niccy cert? Gentle, Firm, Bust ?? :facepalm:

Well naturally yet another tool has to be purchased along with the torque driver......  a spring powered device held against the RCD  that ejects a steel button at a pre-determined force set by BS :innocent

 

[NozSpark]

my fluke does auto test, so I just stand at CU resetting the rcd, then just scroll through the results. Is there any reason why I should do them manually with 1/2x test last?

OK so lets firstly agree that an RCD is a device which (if working) will get a chance to save our lives if/when we do something stupid..

Now lets say that we do that stupid thing,,,, the RCD has one and only one chance to operate,,, if it doesn't then we're going to be put at a higher risk than we should be.

So... given that, we should never exercise a RCD before or during the tests to achieve the prescribed trip times  -  you won't get that benefit if you're getting an electric shock I can assure you!!

And now to my reasoning,, given the above

1x first - when I test a RCD I want to be sure that it'll trip the first time 99.9% of the time,,, if you carry out the 1/2x test first then you could be pre stressing/exercising the mechanism a bit... I know it won't be much, but the trip coil will have had some current passing through it and it will have had a mechanical effect on the mechanism

1/2x last - like I've said,, this is the order I do things... 1x 0deg, 1x 180deg,5x 0deg, 5x 180deg, ramp, manual and then 1/2x

so when I test I have already done 6 tests on the RCD before the 1/2x test,, that's 6 operations of the mechanism.... IMHO this is the most likely time that a RCD will fail the 1/2 times test

When I'm testing a RCD I am looking for it to fail,,, I'm not looking for it to pass................. mostly they pass though, and for that I'm happy as I know that I've thrown everything that I can at it

Oh... and I have a MFT 1502.... It doesn't have an auto test!

 

[Rob_the_rich]

Thanks Noz, I never really thought about the order of tests. just assumed the way in gn3 was correct, same as auto on tester, x1/2, x1 then x5 last then test button.

I have had a read around this evening and I am tending towards what Geoff Blackwell has been saying for some time, x5 first. For basically the same reason as you, if there is any stickiness it should be tested for, not negated/mitigated by doing a x1/2 test, and especially by not pressing the test button (should always be the last test). His reason for x5 first instead of x1 is that it is the more onerous test , 40ms instead of 200, and that a faulty rcd would be more likely to show up. Testing first at 30mA (x1) is not really a "life threatening" situation whereas testing at 150mA is, and it seems to make sense to do this test first.

brummydave, did you test the rccb with all loads disconnected?

Edit; that is pith-poor customer service from Hager BTW.

 

[NozSpark]

Personally I think that the 5x exercises the mechanism a bit too much,,, I'd rather make sure that it trips at the lower 1x current first

 

[Rob_the_rich]

Personally I think that the 5x exercises the mechanism a bit too much,,, I'd rather make sure that it trips at the lower 1x current first

It should be able to handle fault currents well in excess of 150mA, up to its breaking capacity.

 

[NozSpark]

you're right,,, it should

but I'd personally rather check that it trips with a minimal amount of force on the RCD mechanics as opposed to a much larger force (as the first test)

 

[Rob_the_rich]

Ah, I see what you mean, the tickle approach :innocent

 

[NozSpark]

thats kind of it

Give it a stroke, then a punch, and lastly a tickle

 

[kerching]

Well naturally yet another tool has to be purchased along with the torque driver......  a spring powered device held against the RCD  that ejects a steel button at a pre-determined force set by BS :innocent

Brace yourself...you,may not believe this ..........

i have an impact setter from when we used to,fit Viper trembler sensors to windows and doors. Basically a sealed Eclipse Auto Centre punch with a nylon pad on the tip

 

[phil d]

Brace yourself...you,may not believe this ..........

i have an impact setter from when we used to,fit Viper trembler sensors to windows and doors. Basically a sealed Eclipse Auto Centre punch with a nylon pad on the tip

Oh yes brilliant they are,especially if you don't know how to use one and don't read the instructions, it clearly states do NOT use direct on glass, I knew somebody who completely missed that bit, and yes it did break.lol