RCD / GFCI finger test?

[Stonent]

RCDs tend to trip in about 20ms at about 30mA of current. Has anyone ever tripped one via direct contact with a finger?  Did you actually feel it?

I've seen videos demonstrating a hot dog being used as a test but surprised YouTube has not provided anyone doing it for real.

[Artlav]

but surprised YouTube has not provided anyone doing it for real.

I guess it's because you don't test emergency saw blade retraction system by sticking your arm across it. 

[electr_peter]

RCDs tend to trip in about 20ms at about 30mA of current. Has anyone ever tripped one via direct contact with a finger?  Did you actually feel it?

I have not done that on purpose and would not want to do that. I imagine what it would feel like, though (from experience...). It is dangerous lottery you are playing (depending on many factors). Also, remember that current is not limited in any way during those 20ms.

I've seen videos demonstrating a hot dog being used as a test but surprised YouTube has not provided anyone doing it for real.

They are either cowards or not stupid enough to try it.

[dexters_lab]

i'm surprised there is no video of it either... maybe drop a note over to Mehdi Sadaghdar, maybe he would try it?

knowing what uk mains feels like, 20ms is just going to be a quick jolt, still... i dont fancy trying it!

[amyk]

but surprised YouTube has not provided anyone doing it for real.

I guess it's because you don't test emergency saw blade retraction system by sticking your arm across it. 

Funny you say that, there's a video on YouTube of the inventor testing his sawblade retractor with his own finger.

[BradC]

RCDs tend to trip in about 20ms at about 30mA of current. Has anyone ever tripped one via direct contact with a finger? 

Yes. Not intentionally of course (thats what test buttons are for!)

Did you actually feel it?

Absolutely.

30mA across a finger is a *lot* of current. It hurts just like any other electric shock, it's just a lot shorter as it cuts off the juice faster than your normal retraction reaction. More of a stab than a good belt.
I've been thinking of picking up a couple of 10mA medical grade units for my bench, but the best policy is to try not to get whacked in the first place. We don't have RCD's in the house yet. I have them in the shed (new sub-board), and I have plug in units for my bench.

[Stonent]

but surprised YouTube has not provided anyone doing it for real.

I guess it's because you don't test emergency saw blade retraction system by sticking your arm across it. 

Funny you say that, there's a video on YouTube of the inventor testing his sawblade retractor with his own finger.

The only one I've ever seen with that was using a hot dog.

[Stonent]

RCDs tend to trip in about 20ms at about 30mA of current. Has anyone ever tripped one via direct contact with a finger?  Did you actually feel it?

I have not done that on purpose and would not want to do that. I imagine what it would feel like, though (from experience...). It is dangerous lottery you are playing (depending on many factors). Also, remember that current is not limited in any way during those 20ms.

I've seen videos demonstrating a hot dog being used as a test but surprised YouTube has not provided anyone doing it for real.

They are either cowards or not stupid enough to try it.

Well I've been socked at least 3 to 5 times from 120V AC but it's always been limited to one hand.

More times than that though I've been shocked by some other amount of electricity that was a reduced form the 120V AC.

At a house I used to live in, I would get a shock any time I touched the cable modem line.   I think I tested it and it was between 40 to 60 volts AC, not sure how it happened though.

I've also been shocked when some current was leaking across a computer's chassis and I unplugged a cable that was grounded on the other side.

[RJFreeman]

RCDs tend to trip in about 20ms at about 30mA of current. Has anyone ever tripped one via direct contact with a finger?  Did you actually feel it?

Yep, I have tripped one (not deliberately)  and yep, damn right you feel it  ....

[Stonent]

RCDs tend to trip in about 20ms at about 30mA of current. Has anyone ever tripped one via direct contact with a finger?  Did you actually feel it?

Yep, I have tripped one (not deliberately)  and yep, damn right you feel it  ....

In the US, they are typically found on outlets that are near sinks, but rarely elsewhere.

I never really understood what they did until more recently in life.  I had always thought they had something to do with tripping if you were to drop a hair dryer into the sink, but unless the hair drier has its own built in rcd, it won't do anything, the hair drier will just put load on the live and neutral and as long as it doesn't exceed the normal 15 to 20 amp breaker rating, nothing will happen at all.

I just assumed they were like a normal circuit breaker, but now understand they measure imbalance between live and neutral, indicating you may have come in contact with ground.

This one here says they trip at 5mA.

http://www.thisoldhouse.com/toh/article/0,,249536,00.html

[codeboy2k]

Yeah, they should trip at much much less than 30mA.   5mA seems about right.

When I lived in the US, my house had them in the bathroom, in the kitchen, in the garage, and outdoors. Each single RCD/GFI device controlled a downstream daisy chain of about 3-4 ordinary sockets.

I'm in Canada now, and surprisingly there is not much in this current home.  I actually put them in myself here at the kitchen and bathroom, and one on my bench.  This house is older though, about 1970's so it's probably understandable.

[Monkeh]

5mA is fine for single-outlet protection of devices or circuits with no electronics or contact with water (tumble dryer, anyone?). It's hopeless for protecting general purpose circuits.

[BradC]

This one here says they trip at 5mA.

http://www.thisoldhouse.com/toh/article/0,,249536,00.html

And it's half right. Like a thermal circuit breaker, it's a cumulative system. The more currently leakage, the faster it will trip. So at 30mA you will get your 20ms. At 5ma it might take 5 seconds. Your average household RCD is a 20-30mA device. To go lower than that you start getting into medical grade devices for exactly the reason Monkeh states. All domestic devices leak to some degree, but some more than others. Outside of realy leakage (like moisture or degrading motor insulation), most issues with tripped RCD's seem to be badly designed mains input filters that leak on switch-on. Usually a big enough spike to trip the rcd at start up, but not steady state. I've seen really bad CFL's do this enough that having more than 2 or 3 on the same switch will trip the RCD every time they are switched on.

Nothing worse than coming home from holiday to a tripped RCD and a stinky fridge/freezer.

[amyk]

but surprised YouTube has not provided anyone doing it for real.

I guess it's because you don't test emergency saw blade retraction system by sticking your arm across it. 

Funny you say that, there's a video on YouTube of the inventor testing his sawblade retractor with his own finger.

The only one I've ever seen with that was using a hot dog.

Seek to 4:00 to see his finger go in.

[RJFreeman]

This one here says they trip at 5mA.

http://www.thisoldhouse.com/toh/article/0,,249536,00.html

RCDes (GFCIes) come in different sensitivities, between 5 and 30mA are usual for for protection against electrocution but are considered to be in the  'High sensitivity' group.

Fixed household ones in Australia are typically 30mA

[pickle9000]

RCDs tend to trip in about 20ms at about 30mA of current. Has anyone ever tripped one via direct contact with a finger?  Did you actually feel it?

I've seen videos demonstrating a hot dog being used as a test but surprised YouTube has not provided anyone doing it for real.

I have tripped one (120v) and while not painful certainly got my attention and felt it though my entire arm. This was a bench mounted one and I use it for safety.

Compared to a regular shock it was mild to nothing. I have done that twice at 120v and once at 240 (very painful). Even though I was taking precautions the events happened a long time ago, no injury. Not recommended though.

[electr_peter]

I will stress this again - fault current is not limited in any way to 5mA or 30mA until RCD trips in few tens of ms (hopefully).

Without an RCD, it would feel like a jolt going through an arm and leave tingling sensation (and your muscles/reaction play an RCD role - meaning 100-500ms of joy). If RCD trips in 20ms, you will feel very short pulse and have some seconds to think about life.

Still, it is not a recommended way to test RCD/GFCI circuits. Not a good way to find a stuck contact.

[Monkeh]

Still, it is not a recommended way to test RCD/GFCI circuits. Not a good way to find a stuck contact.

Or the various other methods of failure.

RCDs are great devices to have, very worthwhile. But they are meant to be the last line of defence.

[SteveyG]

It is dangerous lottery you are playing (depending on many factors). Also, remember that current is not limited in any way during those 20ms.

The other thing to remember, is there's nothing to say the current through you will reach 30mA. You may end up gripping onto a conductor, but because of what you're standing on or what footwear you have on the current through you gets no higher than 20mA for example, leaving you unable to let go. 

[Monkeh]

It is dangerous lottery you are playing (depending on many factors). Also, remember that current is not limited in any way during those 20ms.

The other thing to remember, is there's nothing to say the current through you will reach 30mA. You may end up gripping onto a conductor, but because of what you're standing on or what footwear you have on the current through you gets no higher than 20mA for example. 

Which is generally accepted not to be lethal for an adult, and won't prevent you from detaching from the source of the shock.

[SteveyG]

It is dangerous lottery you are playing (depending on many factors). Also, remember that current is not limited in any way during those 20ms.

The other thing to remember, is there's nothing to say the current through you will reach 30mA. You may end up gripping onto a conductor, but because of what you're standing on or what footwear you have on the current through you gets no higher than 20mA for example. 

Which is generally accepted not to be lethal for an adult, and won't prevent you from detaching from the source of the shock.

Depends very much on the individual. It's not considered lethal in terms of being unlikely to cause ventricular fibrillation, but over a relatively short period of time will kill you and is incredibly painful.

[Zero999]

There's a small risk the RCD may fail to trip.

This is probably why isolation transformers are used for shavers here in the UK, rather than RCDs because they're passive components and don't rely on anything mechanical.

Neither isolation transformers nor RCDs are perfect but I'd feel more comfortable putting my finger between an isolation transformer and earth than an RCD and earth.

[madires]

5mA is fine for single-outlet protection of devices or circuits with no electronics or contact with water (tumble dryer, anyone?). It's hopeless for protecting general purpose circuits.

Yep, and you also got the Y class filter caps (SMPS for example) which pass some current to earth. A few SMPSs and you'll have more than 5mA loss by the Y caps. Therefore there's a capacitance/current limit for the Y caps over here, to keep the current loss reasonable. IIRC, the maximum current allowed is 3.5mA for usual devices (some special stuff got different limits).

[HackedFridgeMagnet]

I will stress this again - fault current is not limited in any way to 5mA or 30mA until RCD trips in few tens of ms (hopefully).

Sorry I am just being pedantic but of course the fault current is limited.
It is limited by the source impedance.
I am not sure what the fault current would be after say 25ms though.

It is also generally less than the advertised tripping current. So in some cases the fault current may not even reach 30mA.

[electr_peter]

I will stress this again - fault current is not limited in any way to 5mA or 30mA until RCD trips in few tens of ms (hopefully).

Sorry I am just being pedantic but of course the fault current is limited.
It is limited by the source impedance.

I think my statement was misunderstood. Of course you could say current is limited by source and circuit impedance as in any other circuit. But that is not active or even passive current limiting in a real sense that is helpful for poor person on GFCI/RCD socket. In fact source impedance is made as low as possible for an equipment grounding purposes so I would say there is no current limiting what so ever.

RCD/GFCI monitors fault current and trips if its too high, but again, it does not limit current in any way until that happens.

[SeanB]

Current limit into a short with a RCD will trip it in under 2 cycles, same as an imbalance. The caveat though is the RCD can take 5 cycles after that to start to open, and if it is a short then arcing will occur across the contact gap until the arc breaker bars finally cool it enough to stop. This can take 15-20 cycles of the mains until the contacts are far enough apart to stop arcing. So if you are a ground path and the current drawn by the DUT is 10A or so the device can take a half second to disconnect, even though with a 30mA test current it will be open circuit in under 5 cycles. Being shocked for the half second will definitely hurt you, and if the current path through you is to low impedance ( wet skin on both contact areas) there is a high current that will flow, in the order of a half amp or so worst case. Definitely enough to stop the heart in many cases, so YMMV here.

Kind of like stepping off the pavement into traffic without looking and walking across the road. You might get across a good number of times but eventually you will be a road pizza. Test the RCD with a 10k 1W resistor between line and ground, not your finger. Otherwise use the test button, which does exactly the same internally, using a resistor connected so as to create the imbalance current across the main sensor coil.

[N2IXK]

In US terminology, the 5 mA units are referred to as GFCIs, or Ground Fault Circuit Interrupters. These are the ones most US folks are familiar with, installed in kitchens, bathrooms, basements, garages, and outdoors. Intended for protection from electric shock in wet/damp locations.

The 30 mA units are referred to as GFPE, or Ground Fault Protection for Equipment. They are used to limit ground fault current in certain industrial applications, or in places where the 5 mA threshold of a GFCI is too low (buried heating cables for driveway de-icing is one). They are NOT to be relied on for shock protection for personnel.

This one here says they trip at 5mA.

http://www.thisoldhouse.com/toh/article/0,,249536,00.html

RCDes (GFCIes) come in different sensitivities, between 5 and 30mA are usual for for protection against electrocution but are considered to be in the  'High sensitivity' group.

Fixed household ones in Australia are typically 30mA

[SteveyG]

Kind of like stepping off the pavement into traffic without looking and walking across the road. You might get across a good number of times but eventually you will be a road pizza. Test the RCD with a 10k 1W resistor between line and ground, not your finger. Otherwise use the test button, which does exactly the same internally, using a resistor connected so as to create the imbalance current across the main sensor coil.

You don't test between line and ground, you should test between line and a neutral conductor on the other side of the sense coil. You shouldn't dump current to earth as if there is a faulty earth, you then potentially give someone in the residence a shock.

[Zero999]

RCDs also use a fusible resistor for the test function. If a normal resistor is used and the RCD fails to trip, the resistor would glow red hot, smoke and may catch fire.

[SeanB]

Kind of like stepping off the pavement into traffic without looking and walking across the road. You might get across a good number of times but eventually you will be a road pizza. Test the RCD with a 10k 1W resistor between line and ground, not your finger. Otherwise use the test button, which does exactly the same internally, using a resistor connected so as to create the imbalance current across the main sensor coil.

You don't test between line and ground, you should test between line and a neutral conductor on the other side of the sense coil. You shouldn't dump current to earth as if there is a faulty earth, you then potentially give someone in the residence a shock.

True, but a typical house normally has no access to the supply side neutral other than by the connection between neutral and PE at the point of supply. Thus the use of a resistor and a connection ( just for under a second) to PE. If it does not trip then you have either a faulty RCD, or a faulty outlet.

Here in S Africa electricians all use a test plug, that both shows correct wiring for the socket, but which also shows up a floating earth, before you start to increase the current till the RCD trips out. You start at 0mA here and go up in 10ma increments to 50mA, which allows you to find those that are both faulty (do not trip at 30mA) and those where there is existing current or the RCD is too sensitive (trips at under 20mA) to test each socket outlet.

For industrial use in water heating you get a special 3 phase unit with adjustable sensitivity, which will trip at a preset current between 50mA to around 500mA. This is used so that a faulty element does not just keep on arcing in the water if the case fails, but which will not trip for things like condensed water on the terminals.

[briselec]

I've tripped RCDs a few times and have never felt it.