Never trust the test button on an RCD

[Mbrooke]

Came across one that finally taught me otherwise:

https://www.youtube.com/embed/M8phlIi42qY?feature=oembed

 

[Tony S]

Hello Martin, as you know an electrician in the UK would carry out a full test procedure on a suspect RCD or while doing post installation certification testing.

BTW, is this how you damaged your test leads (MH).

 

[Mbrooke]

Hello Martin, as you know an electrician in the UK would carry out a full test procedure on a suspect RCD or while doing post installation certification testing.

BTW, is this how you damaged your test leads (MH).

Yup, which is what I've been calling for all along. This is the reason why. The test and reset buttons never actually test the time, ma threshold or correct deneregization of the device as seen here. My best guess it that the live contact welded shut at some point. The mechanism inside can operate but the contact itself does not part.

Ahhh the leads lol, good memory. No, not here fortunately. It was on a switch I was trouble shooting when I accidentally shorted the lead to the metal body tripping the circuit breaker. As you know we don't do earth fault loop impedance and our breakers have a very high trip threshold so the short circuit took its time...

 

[Blue Duck]

Sounds like you need a "17th edition" multi function tester!

:lol:

 

[Mbrooke]

I do, I really do LOL.

 

[Blue Duck]

Or one of these perhaps?

http://m.ebay.co.uk/itm/ROBIN-KMP5404DL-Digital-RCD-ELCB-Tester-/162174155320?nav=SEARCH

 

[Tony S]

Sounds like you need a "17th edition" multi function tester!

:lol:

US standard is 6mA @120V

I’ve gone it to this before with Martin

 

[Mbrooke]

At least simple lights will do in the mean time:

https://www.youtube.com/embed/v3r6Go3mWeU?feature=oembed

 

[Blue Duck]

belive in the lights! :slap

@Tony S

6mA that's a bit low.

Must keep electricians over there busy!

 

[Mbrooke]

Its better than test/reset buttons, which is all thats required mind you. But yes, 5ma is does not work well on long runs of cable. Or treadmills.

@Tony S

6mA that's a bit low.

Must keep electricians over there busy!

Also keep in mind that unlike the UK GFCI were first mandated to take care of appliances without an earth pin, which was near 100% 50 years ago. 5ma was a good idea, especially if a child was to come in contact with something energized. 

 

[Blue Duck]

I was under the impression that GFCI'S were a new addition to the American regs but what would I know!

So the whole problem with earth leakage it dealt with by not having an earth! :lol:

 

[Tony S]

They’ve been in the NEC for yonks, all be it limited applications.

Under the 14th RCD’s were only a “recommendation” even for TT supplies. They were hellishly expensive back then, my wife played merry hell with me for spending what would now be £600 on a “box” that didn’t do anything useful.

 

[Mbrooke]

I was under the impression that GFCI'S were a new addition to the American regs but what would I know!

So the whole problem with earth leakage it dealt with by not having an earth! :lol:

Nothing new, they started in the 70s and then spread each code edition. Earthing type sockets were not mandated until the mid 60s, which by then (as a result) every tool/appliance in the US was 2 prong. Manufactures kept making 2 prong equipment for existing installations. To make up for that they started mandating GFCIs in areas where an electrified appliance was most likely to kill.  Here is the official history:

http://www.necconnect.org/resources/gfcis/

FWIW GFCIs are still used today to cover for a missing earth which is not uncommon with our plugs:

http://forums.mikeholt.com/showthread.php?t=172466

UK plugs avoid this problem by incorporating shutters which reject a plug with a missing earth.

 

[Mbrooke]

Ok, Update! Got this from Leviton:

Good Morning,

Tuesday, we received the GFCI from the YouTube video (link below) and performed an analysis of the product. The sample is identified as a Phase 7 style GFCI, with date code 1A10I (January 2010).

https://www.youtube.com/watch?v=MGQXsCc11qc

Using an Ohm meter, the GFCI was initially found to have welded Phase side face (receptacle) contacts. When the sample was opened, the weld broke. In accordance with our safety certification and in production, multiple tests are performed to assure this does not occur. On a weekly basis Endurance, Low Resistance Ground Fault, Extra Low Resistance Ground Fault, Short Circuit and Overload tests are performed. These tests evaluate the GFCI’s ability to open a circuit under extreme conditions. Therefore this weld is an extremely rare occurrence. This typically indicates the GFCI has seen significant use/abuse and/or an overcurrent condition (see image below). In this instance, we believe an overcurrent condition is the most likely cause.

The video posted on YouTube demonstrates the user connecting a light bulb from the Phase face contact to ground. With the GFCI contacts welded, the device will provide power from the Phase side even when the device is tripped, as the video shows. However, when the bulb is connected from Phase to Neutral on the face, the bulb does not light up. This is because the Neutral side was not welded and those contacts were open preventing the flow of electricity. The sound that can be heard in the video when the light bulb is connected from Phase to ground, is the GFCI firing the trip solenoid in attempt to remove the fault. Since there is welded contacts, the device will continuously fire when there is a fault until it burns the coil open. It appears that the end user continued apply a fault even after the video ended since the trip coil is burned open (large red circle in image below.)

http://forums.mikeholt.com/attachment.php?attachmentid=15792&stc=1&d=1475165157

 

[Sidewinder]

Mbrooke,

Hopefully there are a few things here that would not be an issue in the UK.

1.  We use earthed socket outlets.

2.  We verify that the earth is present and adequate after installation.

3.  We have to verify by design and measurement that any device is capable of interrupting the fault current at the point of installation.  Therefore contact welding would not occur due to fault current overload.

4.  The higher voltage we use results in lower load currents.

5.  The higher voltage we use assists in the self cleaning arc for the contacts during operation, thus, the contact resistances are lower therefore less Joule heating across the contacts.

6.  The EN standard requires linked contacts thus the mechanism of both contacts is self assisting.#

7.  The user must test their RCD every 3 months.

Therefore this occurrence should be extremely rare in the UK

 

[Andy™]

7.  The user must test their RCD every 3 months.

and we all know everyone tests the RCD every 3 months...

 

[Sharpend]

and we all know everyone tests the RCD every 3 months...

That's what he just said? 

 

[Mbrooke]

Mbrooke,

Hopefully there are a few things here that would not be an issue in the UK.

1.  We use earthed socket outlets.

2.  We verify that the earth is present and adequate after installation.

3.  We have to verify by design and measurement that any device is capable of interrupting the fault current at the point of installation.  Therefore contact welding would not occur due to fault current overload.

4.  The higher voltage we use results in lower load currents.

5.  The higher voltage we use assists in the self cleaning arc for the contacts during operation, thus, the contact resistances are lower therefore less Joule heating across the contacts.

6.  The EN standard requires linked contacts thus the mechanism of both contacts is self assisting.#

7.  The user must test their RCD every 3 months.

Therefore this occurrence should be extremely rare in the UK

Our sockets are earthed as well, but we lack most of the other. BTW, wont high voltages result in contact pitting thus more chance of contacts welding shut?

 

[Sidewinder]

Our sockets are earthed as well, but we lack most of the other. BTW, wont high voltages result in contact pitting thus more chance of contacts welding shut?

That depends on how the contacts are designed, they should be designed to be self cleaning by switching arc.

 

[Mbrooke]

I see. Your RCDs and socket RCDs, how much fault current are they designed to interrupt?