New lab - no grounded outlets

[Emil]

I'll move into a larger house soon and I will finally have a lab instead of just a table in the living room to use for electronics.

The house was built in the late 80's when electrical code in Norway only required grounded outlets in the kitchen and bathrooms. It's an IT-system with single GFI for the entire house.

The room that will become my lab only has ungrounded outlets.

Will the lack of grounding be a real problem? Would it be worthwhile to have a ground wire and new outlets installed?

[mikeselectricstuff]

Yes - you will need grounding for a lot of testgear, PCs etc. as their input filters will have caps to ground - without a ground connection the cases will float at half mains - not dangerous but you will feel it, and it will be enough to damage chips.
 

[Paul Price]

When people are talking about voltage potentials between well-functioning electronic equipment, such as oscilloscopes and PC computers, for instance, they are frightened because  they can measure a high voltage with a sensitive multimeter. There can be a considerable voltage between their respective chassis. However, it is only a very small AC current, too small to cause any injury, it can be somewhere in the high micro-amps, but yes, it sometimes it can cause a mild electrical shock that should be considered more unpleasant rather than dangerous. If there were any dangerous AC current, the GFI in my house would instantly kick open the AC circuit in the room.

When using my digital  and analog oscilloscopes, voltmeters and the likes under these situations, (which I have been doing for over 25 years doing electronic design and repair work.), I always connect the ground connection of the oscilloscope or other test instrument first to the D.U.T.  This means that the probe tip of the oscilloscope or signal generator, etc. is at the same ground potential before I connect a signal or a meter or an oscilloscope probe and can do no harm to the D.U.T.

I am not a lone maniac who chooses to work in this way, hundreds of other technicians in service shops in this tiny country operate in this very same way.

In fact, almost every house in this country has 2-wire AC plugs and people connect things together all the time without any injuries to themselves or their equipment, thanks to GFI. Once in a while people bring in equipment that they have burned out an USB port  because of connecting equipment where there are power ground loops. I always remind people to unplug at least one of the devices when connecting USB. But even different AC grounded circuits in the same room can have a nasty low-voltage AC but with a high-current that can also damage equipment and cause a tickle.

I haven't had any damage to any equipment that I've serviced, nor to myself, nor to the oscilloscopes and other test equipment in my work area.  I have been occasionally been tickled a bit by the bite of this inter-chassis AC  potential.

Before I plug anything in that I might be asked to repair or modify, I always make visual check of the equipment for rust or corrosion or other water damage or just check if the equipment shows any dampness, water drops, etc.  If I am unsure of the equipment, I use a DVM to check the AC power cord for resistance to the chassis ground before plugging it in. That is just common sense.  Electronics is about working with electricity and can be dangerous, so caution must be a thing to  learn and practice. I am not against grounding equipment if it is something like an electric washing machine, a refrigerator or a high-powered circular saw that is not double-insulated.

On the other hand, having a grounded oscilloscope on my workbench could put me in great danger. It could allow me to get a serious shock from some other piece of equipment that has a AC mains short to the chassis if I am touching the metal ground of the scope and the chassis or some live AC mains wire in the circuit of some piece of correctly working or even defective or unsafe equipment. I wouldn't be in mortal danger if the oscilloscope was ungrounded.[/b]

For safety's sake I would not allow any three wired grounded outlets around my work bench test area.

In servicing thousands of devices, I can recall only two or three that ever had a AC mains short to ground. These few items could have been really dangerous, but a visual inspection and a DVM resistance check revealed the danger of the problem before I even turned them on. One was a miswired 3-wire AC EU-style line cord the other was due to someone trying to repair the item and connecting one side of the AC line to the chassis.

[M. András]

those floating chassis cases can hurt like this: m**********r. but if everything is grounded it has no problem. just dont connect things to a live conductor

[Neilm]

On the other hand, having a grounded oscilloscope on my workbench could put me in great danger. It could allow me to get a serious shock from some other piece of equipment that has a AC mains short to the chassis if I am touching the metal ground of the scope and the chassis of some piece of defective or unsafe equipment. I wouldn't be in mortal danger if the oscilloscope was ungrounded.

For safety's sake I would not allow any three wired grounded outlets around my work bench test area.

If you are in the UK that statement could get you prosecuted under health and safety legislation. If I have to use a scope to measure something on mains I either use a scope that has isolation built in or a small isolation circuit. The bandwidth of the latter is pretty rubbish but it works.

Neil

[madires]

On the other hand, having a grounded oscilloscope on my workbench could put me in great danger. It could allow me to get a serious shock from some other piece of equipment that has a AC mains short to the chassis if I am touching the metal ground of the scope and the chassis or some live AC mains wire in the circuit of some piece of correctly working or even defective or unsafe equipment. I wouldn't be in mortal danger if the oscilloscope was ungrounded.

For safety's sake I would not allow any three wired grounded outlets around my work bench test area.

Why aren't you using an isolation transformer for the DUT and a differential HV probe for the scope? What about the ground of the mains HV transformer at the pole or substation?

[Paul Price]

Differential voltage probes are delicate, expensive and clumsy to use in most situations.

High power isolation transformers are expensive and take up a lot of bench space but are an excellent idea if you can afford them.

[ddavidebor]

I'll move into a larger house soon and I will finally have a lab instead of just a table in the living room to use for electronics.

The house was built in the late 80's when electrical code in Norway only required grounded outlets in the kitchen and bathrooms. It's an IT-system with single GFI for the entire house.

The room that will become my lab only has ungrounded outlets.

Will the lack of grounding be a real problem? Would it be worthwhile to have a ground wire and new outlets installed?

Absolutely yes.

As a people that has moved several times, i suggest you to fix the electrical system before yu get in the house.

[MasterOfNone]

Emil,
Also think about how you would like to dress in your new lab. Yeah that’s right, not because the women are going to come flocking now that you have your own lab (They may or may not I don‘t know). But because if you wear Polyester/Nylon plus rubber soul shoes you could find build up quite a bit of static electricity.
Having an ESD mat on your workbench that’s connected to ground could help discharge you safely before you damage any equipment/components.
Anyway this may not be a problem because you may wear decent clothes

[Bored@Work]

On the other hand, having a grounded oscilloscope on my workbench could put me in great danger.

"Hello, my name is Paul Price and I am an idiot."

[vk6zgo]

"On the other hand, having a grounded oscilloscope on my workbench could put me in great danger. It could allow me to get a serious shock from some other piece of equipment that has a AC mains short to the chassis if I am touching the metal ground of the scope and the chassis or some live AC mains wire in the circuit of some piece of correctly working or even defective or unsafe equipment. I wouldn't be in mortal danger if the oscilloscope was ungrounded."

GFI/RCDs and grounded equipment are not mutually exclusive.
An RCD would still protect you in the situation you described.

Even if you don't have the instrument earthed,there may be objects in the room which are incidentally grounded,which you may touch.

[Alana]

I had such setup as OP but after one electric shock too many i rewired it to include earth connectors. Thing is that my workbench is near central heating pipes and it was easy to touch both central heating and ungrounded PC case - and that resulted in painful electric shock - skin on other body parts that inner side of palm and fingers is damn conductive...

But we have TN-C with central heating pipes connected to power line neutral somewhere in basement and Norway is IT so there may be a big difference there. Personally I'd leave it as it is - IT is like powering your entire workbench from big insulation transformer - or asked a local sparky for advice - IT power network has its differences.

[Marco]

Differential voltage probes are delicate, expensive and clumsy to use in most situations.

Are you talking about high frequency differential probes? They are delicate and expensive yes ... but it's not like you are generally going to get good high frequency measurements after you put the stray capacitance of your oscilloscope case into your circuit ... unless it's a high power RF circuit which doesn't mind the extra antenna.

HV differential probes generally are able to take >1000 volt differential voltages, hardly delicate ... you certainly wouldn't want to float your oscilloscope much more than that. They use two separate probe clips as well, which to me seems less clumsy than an ordinary probe with ground clip. Still expensive, unfortunately yes.

[Neilm]

Differential voltage probes are delicate, expensive and clumsy to use in most situations.

High power isolation transformers are expensive and take up a lot of bench space but are an excellent idea if you can afford them.

I'm not talking about differential voltage probes - I'm talking about a small battery powered circuit that is capable of probing a circuit then outputting a signal to a grounded oscilloscope with mains isolation between the two. I don't know if I still have the circuit for the one I built at work - will look tomorrow. To be honest, I haven't really used it since we bought an Tek TPS2000 with isolated inputs.

Neil

[duskglow]

On the other hand, having a grounded oscilloscope on my workbench could put me in great danger.

"Hello, my name is Paul Price and I am an idiot."

This is a topic that I am very hesitant to bring up, but I have noticed that this particular person seems to spread deliberate misinformation on occasion.  Is my BS meter faulty?

[Emil]

Emil,
Also think about how you would like to dress in your new lab. Yeah that’s right, not because the women are going to come flocking now that you have your own lab (They may or may not I don‘t know). But because if you wear Polyester/Nylon plus rubber soul shoes you could find build up quite a bit of static electricity.
Having an ESD mat on your workbench that’s connected to ground could help discharge you safely before you damage any equipment/components.
Anyway this may not be a problem because you may wear decent clothes

My clothes are ESD safe, but may or may not be considered decent by other standards

I probably should have mentioned it in my first post but I have some static dissipative mats for the work surfaces and a few liters of ElectraThane anti-static paint for the floor. Both of these should be grounded too.

[ejeffrey]

Equipment that is designed for use in a grounded circuit can have up to a milliamp or two of ground leakage (not microamps).  I don't know what the regulations are, this is measured by me on multiple pieces of commercial T&M and vacuum equipment.  If you lift the ground connection and you have an available path to ground, that current will go through you.  It is not enough to trip an RCD, and usually not enough to be dangerous (unless you have and existing heart condition) but it will hurt. It also is definitely enough to destroy semiconductor devices -- especially if you are working on a prototype in the lab that doesn't have ESD protection on all the inputs.

This is a bigger problem in 230 volt countries as the mid-rail voltage is 115 volts instead of 60 VAC.  I have never felt a shock from ungrounded equipment in the US, but I have in Europe.  Also, obviously equipment with universal supplies will have twice the ground leakage current when operated at the higher voltage.

[SeanB]

But we have TN-C with central heating pipes connected to power line neutral somewhere in basement and Norway is IT so there may be a big difference there. Personally I'd leave it as it is - IT is like powering your entire workbench from big insulation transformer - or asked a local sparky for advice - IT power network has its differences.

It may be isolated but if your neighbours have a fault to earth then it suddenly becomes a grounded system with no visible indication of the fault. Then you have a 50/50 chance to connect the wrong way and fry yourself. It is only isolated if there is a single load connected to the transformer, and that means a single item, no lamps, no PC power supplies with Class Y caps, no Tv sets and alarm systems, no sat receivers, no other electrical items other than a really short low capacitance lead to a secondary with low capacitance to the core and ground.

[Rufus]

On the other hand, having a grounded oscilloscope on my workbench could put me in great danger.

"Hello, my name is Paul Price and I am an idiot."

This is a topic that I am very hesitant to bring up, but I have noticed that this particular person seems to spread deliberate misinformation on occasion.  Is my BS meter faulty?

You shouldn't believe everything you read on the internet regardless of who says it

[duskglow]

Agreed, but some sources are more trustworthy than others.

[IanB]

Grounded three pin and ungrounded two pin electrical outlet systems are each individually safe as long as they are used consistently and not mixed in the same room or workplace. The two pin system is (was) considered safe as long as the outlets are in a dry room with no exposed metal pipes or other earthing points within reach of the sockets. As soon as you introduce one earthed or grounded item into the environment the whole setup becomes unsafe.

Since every piece of test equipment I have seen has been fitted with a grounded three pin plug, it follows that these devices must be used in a grounded, three pin socket environment. Any two pin device plugged into these sockets must satisfy the "double insulated" safety standards so that there is almost no chance of any live parts becoming exposed.

So the short answer to the opening question is yes: an electronic work bench or lab environment must be fitted with properly grounded three pin sockets for the test equipment to plug into. It is grossly unsafe to disconnect the earth pin of a device fitted with a three pin plug.

This will also provide a convenient place to connect the earth lead of static dissipating mats and wrist straps.

[NiHaoMike]

You might be able to derive a ground from something like a metal water pipe or cable TV connector. Use a 100W or so bulb to check the ground impedance and if that's good, add a GFCI/RCD. (If you don't like the idea of doing DIY mains wiring, an off the shelf solution is a plug in GFCI and a surge protector with coaxial connectors.)

[madires]

Differential voltage probes are delicate, expensive and clumsy to use in most situations.

High power isolation transformers are expensive and take up a lot of bench space but are an excellent idea if you can afford them.

Just checked ebay. 400W for US$10, 3KW for US$115 and 5KW for US$200. Is that high power enough? Doesn't seem to be really expensive, especially if you take care about your safety. I got a small one (350W) with a built-in variac which serves me well. What kind of equipment do you repair that requires high power?

[Thor-Arne]

Emil, absolutely get your lab grounded.

You might check if there is a unused ground wire inside the cable tubing, it was quite common to wire a house with ground and just install the ungrounded sockets.

Also, your fusebox is properly grounded....

[Alana]

But we have TN-C with central heating pipes connected to power line neutral somewhere in basement and Norway is IT so there may be a big difference there. Personally I'd leave it as it is - IT is like powering your entire workbench from big insulation transformer - or asked a local sparky for advice - IT power network has its differences.

It may be isolated but if your neighbours have a fault to earth then it suddenly becomes a grounded system with no visible indication of the fault. Then you have a 50/50 chance to connect the wrong way and fry yourself. It is only isolated if there is a single load connected to the transformer, and that means a single item, no lamps, no PC power supplies with Class Y caps, no Tv sets and alarm systems, no sat receivers, no other electrical items other than a really short low capacitance lead to a secondary with low capacitance to the core and ground.

Good to know, thx.

[staxquad]

live, neutral and grounded cable (~$30/10M?) from it's own breaker (or glass fuse?) in the fuse (breaker) box  (if there are spare) to that lab room, add as many sockets as required, voila (Bob's your uncle), or revamp your electrical   

done already, and while you're at it, add a few outlets in the house here and there as required (and some indoor and outdoor GFCI)

power trip

don't fart around, do it once correctly

Our ground (Ontario Hydro dictated) is the 1/2" copper water pipe from the city supply (now 3/4") and comes through the basement wall at floor level (8 ft deep outside).   A clamp secures a thick bare copper ground wire on that pipe that goes back to the breaker panel.  Only 2x hot and a neutral wire from Ontario Hydro connect to the breaker panel after passing through the meter.   

[edavid]

Has anyone tried using a neutral ground resistor (or some other clamp circuit) to absorb the Y capacitor leakage?  Does this violate code?

[alm]

Yes, I'm pretty sure this is a code violation in most countries. The neutral should have the same insulation and creepage as the live wire, because wires can get switched or neutral wires can become open.

Hint: what's the voltage on a neutral wire if the neutral connection to the power distribution transformer gets cut somewhere? This is a resistive divider with an extremely high resistance between equipment ground and earth, and a relatively low resistance of all the parallel equipment connected between live and neutral.

Grounded three pin and ungrounded two pin electrical outlet systems are each individually safe as long as they are used consistently and not mixed in the same room or workplace. The two pin system is (was) considered safe as long as the outlets are in a dry room with no exposed metal pipes or other earthing points within reach of the sockets. As soon as you introduce one earthed or grounded item into the environment the whole setup becomes unsafe.

I guess that they didn't ground water pipes or radiators back then either?

You might be able to derive a ground from something like a metal water pipe or cable TV connector. Use a 100W or so bulb to check the ground impedance and if that's good, add a GFCI/RCD. (If you don't like the idea of doing DIY mains wiring, an off the shelf solution is a plug in GFCI and a surge protector with coaxial connectors.)

This is a code violation in most places. What happens if a metal pipe gets replaced by plastic during maintenance? What if some poor plumber disconnects a pipe and then gets shocked from the leakage current of all Y caps in the equipment you have connected? Keep in mind that he/she is most likely working in a wet environment. I wouldn't trust a TV cable either, the ground connection should be able to carry enough current and have a low enough impedance to trip a breaker in case of a short, or at least until the GFCI trips. I expect the coaxial shield to burn up very quickly if you connect it to 115 V / 230 V (which is what happens in case of a fault it's supposed to protect from).

[Stonent]

Some of the long outlet strips you can buy have an external grounding screw, you could try to run that to a safe location.

[IanB]

I guess that they didn't ground water pipes or radiators back then either?

I imagine they did, but you would not be allowed to have exposed metal pipe in a room with two pin outlets. It must be impossible to come into contact with any kind of earth in a space where two pin outlets are used.

[madires]

I guess that they didn't ground water pipes or radiators back then either?

I imagine they did, but you would not be allowed to have exposed metal pipe in a room with two pin outlets. It must be impossible to come into contact with any kind of earth in a space where two pin outlets are used.

Then the people living in that house need to float 5cm above the floor :-) The building itself is earthed! It's not a good conductor but most times a few mA can flow. If the wall socket has no ground the few mA won't maybe trigger the RCD but could be enough to hurt you. A grounded wall socket makes sure that there's a good ground connection for a broken device, i.e a hot metal case would trip the breaker or trigger the RCD quite fast. As I wrote before we have also to consider the outside world. One tap of the pole/substation transformer is earthed/grounded (this is the neutral wire).

[IanB]

Then the people living in that house need to float 5cm above the floor :-) The building itself is earthed! It's not a good conductor but most times a few mA can flow.

It's not that bad. In a dry indoor room with wooden floor there is no path for a dangerous leakage current to follow. In my misguided youth I worked quite obliviously with a live 240 V bare wire while kneeling on the concrete floor of an outdoor garage and felt nothing at all (I forgot to unplug the circuit before working on it). I only remembered it was live after I accidentally touched the wire to earth and tripped the breaker.

The resistance of dry building materials is megohms and no current large enough to feel will flow through that. (For instance with 1 megohm you would have about 240 µA.)

[Monkeh]

You might be able to derive a ground from something like a metal water pipe or cable TV connector. Use a 100W or so bulb to check the ground impedance and if that's good, add a GFCI/RCD. (If you don't like the idea of doing DIY mains wiring, an off the shelf solution is a plug in GFCI and a surge protector with coaxial connectors.)

Please nobody listen to this guy.

[duskglow]

It's not that bad. In a dry indoor room with wooden floor there is no path for a dangerous leakage current to follow. In my misguided youth I worked quite obliviously with a live 240 V bare wire while kneeling on the concrete floor of an outdoor garage and felt nothing at all (I forgot to unplug the circuit before working on it). I only remembered it was live after I accidentally touched the wire to earth and tripped the breaker.

When I was young, I accidentally touched a live 120V prong while standing on a linoleum floor while wearing shoes.  I felt it.  It was not pleasant.  I suffered nothing more than feeling like my finger was buzzing.

Obviously, though, this is playing with fire.  I am astonished that with the number of televisions I messed with with the power on I never got thrown across the room.

[IanB]

When I was young, I accidentally touched a live 120V prong while standing on a linoleum floor while wearing shoes.  I felt it.  It was not pleasant.  I suffered nothing more than feeling like my finger was buzzing.

I suspect you also were making contact with the neutral prong with another part of your finger. That would explain why only your finger was buzzing and not your whole arm.

An electric shock traveling up your arm makes your arm muscles contract involuntarily in a way that is highly unpleasant, like having all your muscles cramped at once.

[nukie]

Now I understand why I got zapped by the perfectly working 2 mains pin SMPS 12v. Bloody painful but not as shocking as 240v, stupid design!

[jmacqueen]

I have felt the result of screwed up ground wiring or lack of grounding on a customers brand new house while doing plumbing when I unfortunately discovered a 240v potential between the wet garage floor and the water inside/case/chassis of the 240v water heater by becoming a quite decent path to earth for it with just a touch, put my lights right on out in an instant and felt like a truck just hit me. Scared me to death when my vision came back, scared the customer even worse, and after I got off the phone with her building contractor there was an electrician on site before I left.

Personally I would drive a couple of 8 foot copper clad grounding rods outside the house and get a proper ground wired into the house or at least the lab from there. I just got finished wiring up my own home several months ago , a 1950's farm house I bought and had moved to the property I bought to retire to. First thing I did was drive a couple of ground rods outside by the incoming power and tied neutral and the breaker box chassis to good earth ground, and made sure I had a good earth ground run to all outlets in the house.

[staxquad]

I guess that they didn't ground water pipes or radiators back then either?

I imagine they did, but you would not be allowed to have exposed metal pipe in a room with two pin outlets. It must be impossible to come into contact with any kind of earth in a space where two pin outlets are used.

My house was build in 1939, only had 2 pin outlets with copper water pipe ground (coming from the town supply, 8 ft underground), exposed copper piping throughout the basement with outlets nearby.  Nobody knowingly would stick metal things into a socket, holding it with their bare hands while touching a grounded metal pipe with the other.  If they did they would win a Darwin award.  The wiring was under code.  The house is now 3 prong, with copper water pipe ground still, (upped amperage, new meter, panel, outlets, etc)

[duskglow]

The house I grew up in was built in 1894.  It was not only under code, it was dangerous.  When the furnace or washing machine would start, the lights would dim appreciably.  Sadly, we were renting so we couldn't get it fixed.

That was not the most dangerous thing.  The most dangerous thing was when the solenoid on the furnace gave out, and the landlord wouldn't fix it.  So my father (genius) stuck a jar and block of wood under it to hold it open.  To start the furnace, he'd open the valve and stick a flaming paper right in the gas stream.

When the landlord finally got around to fixing it, the repairman was astonished that the house hadn't blown up yet.  He ended up replacing the furnace.  That original furnace, I think, was 1940s vintage.  Or earlier.

I got the electrical parts from it when they scrapped it, so not all was lost.  That was a hefty transformer.  I probably would have killed myself if I'd had the nerve to power it up.

[Tepe]

In fact, almost every house in this country has 2-wire AC plugs

Assuming "this country" is the same one that I live in, I will add that of the eleven homes I have lived in, only the latest one has some outlets with PE and it only has them in the kitchen. As Paul wrote, this is very common for the 220V (now 230V) outlets. The 380V three phase outlets (now 400V) do have PE, but you don't tend to have those all over the house

Historically, PE seems to have been frowned upon in this country. In "Elektriske installationer", Eriksen & Hansen, 1960, the authors wrote that opinions varied strongly on the subject of PE in household installations  and while some countries like England and Norway held that all metal parts of an installation that was not live had to be grounded, others, including Denmark, have been more reluctant in making such a requirement.

If you go back in time in the literature, you will sometimes encounter rather strong opposition to grounding and discussions of whether the use of it makes installations safer or more dangerous.

Like it or not, this is the reality we live with here. And, yes, the typical mini tower pc in a metal box with an SMPS can tickle you if you touch it and a radiator at the same time. That is also something many of us have to live with.

[Paul Price]

Consider these points:

It might be so much easier and cheaper to install a single GFI interrupter in you fuse box of your building than to spend tens of thousands of bucks to rewire your house.

Tepe says it clearly, "Assuming "this country" is the same one that I live in, I will add that of the eleven homes I have lived in, only the latest one has some outlets with PE and it only has them in the kitchen. As Paul wrote, this is very common for the 220V (now 230V) outlets. The 380V three phase outlets (now 400V) do have PE, but you don't tend to have those all over the house "

You might have different opinions about this topic but reason must replace sophomoric and rude name calling as a way of establishing your opinions as the only correct authority.

You must consider that there are different ways to practice safety and in fact, these different ways do work well and are used by not just me but thousands of electronic service people in this big world. And the same non-grounding, GFI wiring standards are the standard for electrical connection  and the safety of household wiring in some countries, other than the one you are living in.

If you go about and ground everything in your work area with a good earth AC mains power ground...you might think you've eliminated any chance of shock hazard from improperly wired or properly wired equipment. But if you have one hand resting on the metal 3-wried grounded power strip on your work bench or holding anything else grounded with one hand and then accidentally touch a live AC mains somehow connected wire in a piece of equipment that you are working with, then you are exposing yourself to a possible lethal shock. If your grounded equipment or grounded metal power AC outlet strip was not grounded you would not be in such mortal danger, especially if I have a GFI interrupter in the AC power circuit in your building or house.

The best approach I have to this safety problem is to use an isolation transformer only on the  equipment D.U.T. and stay away from using dangerous earth grounds anywhere else on the workbench. This way neither I nor the equipment can be connected so as to get a shock from the AC mains.

Sometimes assumed safe practices may in fact create more danger in some situations.

 I say think about what I have said and remember  that you have a choice and a responsibility, as I have, to carefully decide why and under what conditions some safety methods are safe or not.

[Paul Price]

Maderes, You ask what kind of equipment...please see https://www.eevblog.com/forum/beginners/lc-power-supply-capacitor-value-question/msg251423/#msg251423

[westom]

When I was young, I accidentally touched a live 120V prong while standing on a linoleum floor while wearing shoes.  I felt it.  It was not pleasant.  I suffered nothing more than feeling like my finger was buzzing.

  Some building materials including linoleum, concrete, and some wall paints are electrically conductive.  Just another reason why two wire power circuits may need a GFCI/RCD.

  However the OP is talking about a lab.  Those unexpected electrically conductive material can create workbench confusion.  Just another reason why a (third wire) safety ground (and not just a GFCI or RCD) is important when doing lab work.  To avert confusion created by stray currents.

[madires]

If you go about and ground everything in your work area with a good earth AC mains power ground...you might think you've eliminated any chance of shock hazard from improperly wired or properly wired equipment. But if you have one hand resting on the metal 3-wried grounded power strip on your work bench or holding anything else grounded with one hand and then accidentally touch a live AC mains somehow connected wire in a piece of equipment that you are working with, then you are exposing yourself to a possible lethal shock. If your grounded equipment or grounded metal power AC outlet strip was not grounded you would not be in such mortal danger, especially if I have a GFI interrupter in the AC power circuit in your building or house.

Metal power strips with the metal gounded? Never saw any over here, i.e all power strips are plastic. I think it's designed that way to stay safe with very old buildings without a dedicated PE. BTW, in any case we should know what we're doing if dealing with HV. 

The best approach I have to this safety problem is to use an isolation transformer only on the  equipment D.U.T. and stay away from using dangerous earth grounds anywhere else on the workbench. This way neither I nor the equipment can be connected so as to get a shock from the AC mains.

The isolation transformer isn't fool proof either. Don't close the loop inside the DUT :-)

You must consider that there are different ways to practice safety and in fact, these different ways do work well and are used by not just me but thousands of electronic service people in this big world. And the same non-grounding, GFI wiring standards are the standard for electrical connection  and the safety of household wiring in some countries, other than the one you are living in.

The goal is to know how the local mains wiring works and to adapt your safety measures accordantly. But a proper grounding might be required for an ESD-safe working space, measurements of low values or connecting different T&M gear.

[jpb]

Metal power strips with the metal gounded? Never saw any over here, i.e all power strips are plastic. I think it's designed that way to stay safe with very old buildings without a dedicated PE. BTW, in any case we should know what we're doing if dealing with HV. 

Metal sockets with metal conduit are quite common in workshops and labs in the UK. In fact I have used them (via an RCD) on my workbench. I did so to cut down on the amount of mains electrical noise, but I think the idea generally is that they are tough against mechanical damage such as accidentally putting a power drill through one and because they are visible you know where the cable runs are.

[vk6zgo]

You must consider that there are different ways to practice safety and in fact, these different ways do work well and are used by not just me but thousands of electronic service people in this big world. And the same non-grounding, GFI wiring standards are the standard for electrical connection  and the safety of household wiring in some countries, other than the one you are living in.

Indeed,but you are falling into the same trap----many thousands of people throughout the world have used "PE" wiring & the associated equipment without any problems.
In fact,back in the day,many homes had domestic radios with "figure 8" power cables sitting in the kitchen in close proximity to earthed refrigerators,kitchen sinks,etc.

In Australia,these were all quite safe,as they used transformer type power supplies,so the Mains were pretty much restricted to the on/off switch & the transformer primary.

If they failed,they may well have been tested using bench test equipment such as VTVMs & Signal Generators,all of which were earthed.

The incidence of fatal electric shock to Radio servicemen was  vanishingly small.

Of course,these days,Test equipment is still earthed,but much domestic stuff is not.
What we DO have is RCDs as a mandatory fitment to any new wiring,so I reiterate,GFI/RCDs & PE wiring
are not mutually  exclusive.

Most of the panic about Electrical safety is overdone,at least as far as Engineers & Techs are concerned.

Over some 40 years in Radio & TV Broadcasting,I have known  about 4 people who died accidentally__
One crashed a "Pitts Special" stunt plane, two others died in falls,& one was washed off the rocks while fishing.

Electrocution?---not one!

[alm]

I would not want an ungrounded metal conduit with mains wires in unknown condition (how often do you open it to inspect the wires?) near my bench. A grounded metal conduit will trip the breaker or GFI/RCD if a wire ever makes contact. I don't like to rely on my carpet (which incidentally was not UL tested for mains isolation) to limit the current through my body.

How do countries like Denmark, which apparently consider using equipment with a three-prong plug in a two-prong socket, treat the difference between IEC protection classes? In many countries where grounded outlets are common, equipment either needs a ground connection (class I), or has to apply to more stringent double insulation standards to reduce the chance that any single fault will bring the user into contact with dangerous voltages. Double insulation is getting increasingly popular because preventing contact with mains wires is considered safer than blowing the fuse in case of contact.

Can you just design equipment to the lower class I creepage/clearance standards without providing a ground connection in countries like Denmark? Do manufacturers like Agilent or Rigol test their equipment for safety and EMC/EMI without ground connection before marketing it in Denmark? Are there building code requirements about a minimum impedance to ground, just like the maximum earth ground resistance required in other countries?

[david77]

What exactly are we talking about here? Does the OP's house really use electric sockets without PE connection like this one



Or is it just an older installation using sockets with PE contact but only two wires (L and PEN, TN-C) to connect them?

In the first case I'd say get an electrician in and have the house/flat rewired ASAP.
I'm not going to say that the second variant is state-of-the-art but there is usually no immediate danger - if the installation is in good condition and nobody practiced their DIY skills on it. There must be millions of homes across Europe still wired like that.

In both cases I fail to see what a GFCI could do to improve the situation as there is no separate (TN-S/TN-C-S) PE connection back to the breaker panel. Here in Germany a GFCI is not allowed in a TN-C installation as it would cut the PEN conductor, thus cutting the earth return path wich is strictly forbidden.

One could consider splitting the PEN (2nd case scenario, TN-C system) directly at the workbench and installing a GFCI there powering dedicated sockets just for use on the electronics workbench, while leaving the rest of the house/flat alone.

All this depends very much on local regulations, though. And it is certainly allways a good idea to have a GFCI at least for the bathroom and kitchen.

[vk6zgo]

The things we call RCDs in Australia are really core balance relays.
They compare the currents flowing in the Active & Neutral lines.
These currents should be equal in value.

If they are not,current is flowing back to the Neutral line some other way,(via something connected to Earth,& hence the Neutral at the entry point),so the device becomes unbalanced & trips.

It does not require any connection in the Earth circuit,so can be used with sockets lacking an Earth connection--although such sockets are not used in this country.

They would not work with a Mains supply which was totally floating from the power station to the power socket on your wall(inclusive)-----but I don't think such supplies exist,otherwise Paul's GFI would not provide him any protection either.

[somlioy]

OP already stated that his house has a RCD. Tho if the house has a bad local earth it might need to be improved. The requirement is a cutoff within 0.4s, max 30mA. RCDs are there to protect you from contact voltages below 50VAC.
Splitting PEN at the workbench is NOT allowed i Norway. TN-C is only allowed up to the first fuse box. After that, TN-S and nothing else. Anyway OP already stated that his house is fed by a IT-system (i.e no neutral/PEN).

If your house has a open wiring you could be unlucky. Electricians usually snipped the earth wire if they didnt use it.
Or maybe the wiring is hidden? THere might be enough room in the pipes for a PE-wire.

[Emil]

This turned into a very interesting topic

After reading all the replies, I think I need to provide some more additional information about the installation.

The house has a ground connection. Several rooms in the house have grounded sockets (bathrooms, kitchen, laundry room), however the living room and the bedrooms do not. I will use one of the bedrooms for my lab.

The electrical installation is in compliance with code as they were at the time it was built. (All grounded sockets would be required for new construction.)

The sockets in the rooms without ground does not have the PE connector. The sockets looks like david77's picture.
There is an GFI/RCD in the fuse box. It should trigger in <=100ms at 30mA.

It's an IT, not a TN installation. There are two conductors with 230V between the conductors and 115V between either conductor and ground. There is no neutral conductor.

Thor-Arne mentioned that ground wire may be installed and that they just used ungrounded sockets. I have been told this is not the case, but I have not opened any sockets to check. If I'm wrong about this, installing grounded sockets will be very easy.

[IanB]

I would not want an ungrounded metal conduit with mains wires

Surely an ungrounded metal conduit with mains wires would violate every electrical code on the planet?

I don't think you would ever come across that in a properly installed system.

[Tepe]

How do countries like Denmark, which apparently consider using equipment with a three-prong plug in a two-prong socket, treat the difference between IEC protection classes? In many countries where grounded outlets are common, equipment either needs a ground connection (class I), or has to apply to more stringent double insulation standards to reduce the chance that any single fault will bring the user into contact with dangerous voltages. Double insulation is getting increasingly popular because preventing contact with mains wires is considered safer than blowing the fuse in case of contact.

That's where it starts to get fun. The current "Stærkstrømsbekendtgørelsen" incorporates the IEC classes and also states that class I appliances must be equipped with a special plug with a PE pin or have other means of establishing a connection to the installation's PE. Yes, the PE that you may or may not have in your house wiring. To make matters even more fun, the Safety Board actually writes that you may connect such a three pin plug to a two pin outlet with an extension cord having a three pin female connector at one end and a two pin plug at the other. Voila! But only if your outlet is of the non-PE variety, of course.

RCD (called HFI and HPFI here) is required in all homes but not in, for example, kindergartens. New installations are required to use a PE wire; except if the residence is from before April 1, 1975. You may replace a non-PE outlet with a new non-PE outlet. That doesn't count as a new installation.

You have no less than three choices for your new three pin outlet should you decide to get additional outlets installed, just to make sure the future will be even more fun:

The one to the right is the traditional type.

Can you just design equipment to the lower class I creepage/clearance standards without providing a ground connection in countries like Denmark?

I don't know.

Do manufacturers like Agilent or Rigol test their equipment for safety and EMC/EMI without ground connection before marketing it in Denmark?

Highly unlikely. They don't sell home appliances, so why should they?

Are there building code requirements about a minimum impedance to ground, just like the maximum earth ground resistance required in other countries?

I don't know about that either.

[madires]

The things we call RCDs in Australia are really core balance relays.
They compare the currents flowing in the Active & Neutral lines.
These currents should be equal in value.

The RCDs over here do exactly the same, no PE is connected to them.

[alm]

[complex story about the Danish electrical system snipped].

Wow, sounds like they are trying to combine international standards with several different local conventions. I guess you can't very well go back and rewire thousands of old houses, but combining two very different safety strategies seems a bad idea to me.

Highly unlikely. They don't sell home appliances, so why should they?

Fair enough, but it would be useful for the people using them at home. They may include a statement like 'Only for use in labs by trained engineers', though.

[Tepe]

Wow, sounds like they are trying to combine international standards with several different local conventions. I guess you can't very well go back and rewire thousands of old houses, but combining two very different safety strategies seems a bad idea to me.

While something fishy does seem to be going on (what's the sense in going from having one type of outlet with PE to having the choice between three different ones?), I think most residential installations are TT with the mandatory RCD and adding PE to the outlets should be okay.

[megajocke]

How do countries like Denmark, which apparently consider using equipment with a three-prong plug in a two-prong socket, treat the difference between IEC protection classes? In many countries where grounded outlets are common, equipment either needs a ground connection (class I), or has to apply to more stringent double insulation standards to reduce the chance that any single fault will bring the user into contact with dangerous voltages. Double insulation is getting increasingly popular because preventing contact with mains wires is considered safer than blowing the fuse in case of contact.

I can not answer for Denmark nor Norway, but it seems the situation there is similar to Sweden. In Swedish installations an area (usually a room) has either grounded outlets or ungrounded outlets, but according to code is not allowed to have both. New installations are required to use grounded outlets and RCD/GFCI.

Class II appliances use plugs that fit in both types of outlets and are allowed in both grounded and ungrounded areas.

Class 0 appliances have plugs that only fit in ungrounded outlets and are only allowed in ungrounded areas. The thinking is that there are two safety barriers: one is the internal single isolation barrier of the appliance, and the other one is the relative inaccessibility of ground in the environment, preventing a closed current path. Such appliances may no longer be legally sold.

Class I appliances may be used in both kinds of areas. In an area with grounded outlets, the protective earthing is relied upon as the second safety barrier, while two-layer safety is provided in exactly the same way as for Class 0 appliances if used in an area with ungrounded outlets.

So basically you are allowed to use an appliance in an area if its plug fits. Of course there might be other restrictions, such as appliances designed for indoor use only.

Can you just design equipment to the lower class I creepage/clearance standards without providing a ground connection in countries like Denmark?

That would be a Class 0 appliance, and selling them are no longer allowed in Sweden. Among other things, it would not be possible (without illegaly modifying the plug or similar) to use such equipment in a modern installation which has only grounded outlets. A Class I appliance needs to have the grounded cord and plug, regardless of where the user is going to plug it in.

Do manufacturers like Agilent or Rigol test their equipment for safety and EMC/EMI without ground connection before marketing it in Denmark? Are there building code requirements about a minimum impedance to ground, just like the maximum earth ground resistance required in other countries?

Sweden allows equipment designed to the Harmonized European Class I and Class II standards to be sold. There is no additional requirement. The safety of Class I equipment in ungrounded areas is no worse than Class 0 items in the same areas. As for ground impedance, equipment is governed by the same standards. Electrical installations have requirements that the fuse/breaker must trip within a certain time if there is a phase/ground short at the load, setting a limit to how high the grounding impedance may be.

Anyways, I think that for an electronics lab grounding is preferred, even if not having it would at least be relatively safe especially as an RCD is installed. Just as Mike wrote in the second post of this thread, there is a risk of damaging sensitive electronic devices when making connections. The ground connection is also preferable to have to be able to effectively discharge static electricity for ESD protection reasons.

Having your local "ground" floating (usually about 110 V AC due to those mains filters and other capacitive coupling in equipment) also tends to couple hum into high-impedance nodes of circuits being measured/tested. So there is also a signal quality problem.

[megajocke]

While something fishy does seem to be going on (what's the sense in going from having one type of outlet with PE to having the choice between three different ones?), I think most residential installations are TT with the mandatory RCD and adding PE to the outlets should be okay.

That looks like a mess.  Here in Sweden only the leftmost outlet in the picture is used and allowed. I guess we were lucky to use the one that were to become the most common right from the start. The two other types are only used in a few countries (including Denmark) as far as I know.

[westom]

Thor-Arne mentioned that ground wire may be installed and that they just used ungrounded sockets.

  Maybe.  But why?  If the ground wire exists, then why spend the same money for the inferior two wire socket?

  Equipment designed for three wire connections (ie Agilent) contain many layers of human safety.  Eliminating one (ie using it on a two wire circuit) only eliminates one protection layer. The user is still protected.  That is about human safety.

 Second, another factor are stray currents.  Strange currents cause confusion when stray milliamps do not have the common ground to leak away into.  A ground is also important for a more productive working environment.

  Third, that safety (protective) ground does nothing for ESD (a human static electric discharge).  ESD current must have another connection to a different ground.  ESD protection involves other solutions.

[Thor-Arne]

Thor-Arne mentioned that ground wire may be installed and that they just used ungrounded sockets.

  Maybe.  But why?  If the ground wire exists, then why spend the same money for the inferior two wire socket?

I suppose they did that so they could swap the fronts if the customers wanted grounded outlets.
The internals for the socket is the same and can be changed according to the needs. I believe the ones I've seen is the ELKO brand.

I'm not qualified to say what the code say about this, but I wouldn't be surprised it grounding is required in the wall tubing.

[madires]

  Third, that safety (protective) ground does nothing for ESD (a human static electric discharge).  ESD current must have another connection to a different ground.  ESD protection involves other solutions.

I don't agree. The ESD stuff (mat, wristband and so on) is connected via a 1M ohm resistor to PE, the same PE as connected to the power socket. The tip of my ESD compliant soldering station is connected to the PE of the IEC socket via a resistor. At the main mains distribution panel the PE is connected to a proper grounding system (ring armature). All house tubing is also connected to the same ground. There's only one ground!

[westom]

There's only one ground!

  Many grounds exist.  The chassis ground is different from the digital ground, is different from the ESD ground beneath shoes, is different from a floating ground inside a TV, is different from an analog ground, is different from the protective ground, and is different from earth ground.  Some grounds may be interconnected.  Others not.  But numerous and electrically different grounds always exist. 

  Your ESD iron connects to protective (safety) ground for human safety.  It might also perform ESD protection.  But only because of how is connects to the other and relevant ground for ESD protection.

  A benched UUT can be connected to protective ground.  And then be damaged by the resulting ESD discharge.  How?  Follow a complete current loop that is the ESD discharge.  And learn why grounds, even when interconnected, can be electrically different.

[tld]

I'm in a similar situation, also in Norway.

Wiring seems to be fairly decent, but for some reason PE wasn't included in the room where I've set up my lab, and I haven't gotten around to doing anything with it yet.

Estimates for having PE installed there are from $500 up.

Just thought I'd add in that Norway is fairly strict about people wiring stuff up themselves, and there are actually inspection from time to time, to make sure everything is according to code, documentation required that things were done not only by a certified electrician, but by one currently employed with a company, etc.

Funny how that works.  It doesn't actually matter how qualified you are, you're still not allowed to touch it unless you currently do it for a living.

Really funny when you consider that improving crappy work done by such an electrician would be illegal to fix or improve.

I think actually removing a socket to check for issues would be illegal.

- tld

[Thor-Arne]

I'm in a similar situation, also in Norway.

Wiring seems to be fairly decent, but for some reason PE wasn't included in the room where I've set up my lab, and I haven't gotten around to doing anything with it yet.

Estimates for having PE installed there are from $500 up.

$500 isn't too bad, work on electric installations tends to be insanely expensive here.

Just thought I'd add in that Norway is fairly strict about people wiring stuff up themselves, and there are actually inspection from time to time, to make sure everything is according to code, documentation required that things were done not only by a certified electrician, but by one currently employed with a company, etc.

Funny how that works.  It doesn't actually matter how qualified you are, you're still not allowed to touch it unless you currently do it for a living.

Really funny when you consider that improving crappy work done by such an electrician would be illegal to fix or improve.

I think actually removing a socket to check for issues would be illegal.

- tld

You can only touch what is unpluggable, all permanent installations is off limits.

As for checking the wiring one is supposed to call the DLE (Det Lokale Eltilsyn), which is probably going to cost an insane amount of money. And then you have to get a certified electrician to fix any problems they find, yet another insane amount of money.

On the otherhand, I've seen a few bad things done by these certified electricians; entire floors without grounding, wires not properly fixed to the socket, electric meter placed on the wrong circuit (two meters in series), ....

Strange how they can get away with these things.....

[c4757p]

It's almost as if red tape doesn't work....

[madires]

There's only one ground!

  Many grounds exist.  The chassis ground is different from the digital ground, is different from the ESD ground beneath shoes, is different from a floating ground inside a TV, is different from an analog ground, is different from the protective ground, and is different from earth ground.  Some grounds may be interconnected.  Others not.  But numerous and electrically different grounds always exist. 

I thought we're were discussing grounded outlets, so I meant there's one ground for the whole house wired in a star topology. Signal grounds and other grounds inside devices are a completely different story.

  Your ESD iron connects to protective (safety) ground for human safety.  It might also perform ESD protection.  But only because of how is connects to the other and relevant ground for ESD protection.

Since we don't float above the floor and are grounded by our feet or our butt when sitting the best ground for a save ESD potential equalization is the ground of your house, i.e. PE. 

  A benched UUT can be connected to protective ground.  And then be damaged by the resulting ESD discharge.  How?  Follow a complete current loop that is the ESD discharge.  And learn why grounds, even when interconnected, can be electrically different.

Yes, interconnected grounds got differences in their potentials caused by currents and the wire resistance. To minimize those effects a star topology is used. Before touching any insides of the DUT you would equalize your potential via the ESD handwrist (the resistor limits the current, also in case you touch a mains wire) which is connceted to the same ground as the DUT.

[Tepe]

That looks like a mess.  Here in Sweden only the leftmost outlet in the picture is used and allowed. I guess we were lucky to use the one that were to become the most common right from the start. The two other types are only used in a few countries (including Denmark) as far as I know.

The left one is indeed a Schuko outlet like you have (CEE 7/4), the middle one is a CEE 7/5, and the rightmost is a standard Danish grounded outlet (DK 2-1a).

All three are required to be tamper-resistant to you don't go around poking knitting pins or nails into them one at a time.

[Tepe]

It's almost as if red tape doesn't work....

It works. Somebody, somewhere, is making a fortune.

[eKretz]

It's almost as if red tape doesn't work....

It works. Somebody, somewhere, is making a fortune.

Right; it works great for the entities that lobbied for the red tape in the first place, it just puts the screws to everybody else.

[cthree]

Wow, I cannot believe some of the answers coming from so-called "electrical engineers", holy shit!

Hire a licensed electrician to come and inspect your house's wiring and panel. Get an estimate detailing what needs to be done to bring it up to code and then DO IT! It is an investment in your property. It only takes one shock to kill you and you will not see it coming, bam! lights out!

GFCI and arc-fault breakers are a last resort attempt at circumventing the process of your death. If you are lucky, you won't die. Don't bet your life on them. 100mA at mains voltage and you are dead.

Anyone with a "better idea" is a danger to themselves and, more importantly, others. I hope you didn't buy your house from one of them; another reason to get it checked by a licensed professional.

[eKretz]

I'm staying out of this debate. In my opinion there is nothing to debate. Homes and labs should have grounded circuits.

[duskglow]

In staying out of the debate, you just entered it.   

My take on it, from a decidedly hobbyist perspective, is that grounding was created to solve a real problem, and the fact that we're debating this shows that the problem was so successfully solved no one thinks about it anymore.  I know I don't want anything to do with equipment without grounding.

[c4757p]

For me, at least, it's a pretty easy question to answer. You can flap your lips all you want about the safety of floating equipment, but I'll take "I know this is grounded" over "I hope this is floating" any day. Case closed. It's easy to see that something is grounded, it's hard to verify that it's not.

[edavid]

For me, at least, it's a pretty easy question to answer. You can flap your lips all you want about the safety of floating equipment, but I'll take "I know this is grounded" over "I hope this is floating" any day. Case closed. It's easy to see that something is grounded, it's hard to verify that it's not.

Sure, if it doesn't cost anything, but if you have to open the walls, that is a huge hassle and expense, for probably zero benefit

[SeanB]

Different countries, different rules, and they are all correct in their own place.

However the modern consensus is that items must be double insulated such that any single failure which does not prevent it operating as desired will not cause a hazardous voltage to be accessible to the user of the appliance. If this is not possible, or there is exposed metal that is able to be touched, then it must be grounded with a conductor of appropriate cross sectional area to withstand the breaking current of a protective device.

As to the supply side, it is recommended that the conductors must be of appropriate cross sectional area for the current, and that there should be provision for an appliance with a protective earth to have that connected if that is able to be plugged in to the socket.

RCD protection is a good addon if it is not mandatory, and in the USA AFCI is a good recommendation to have in the panels as well as the GFCI. Whether the RCD is in the panel, or in the socket, or the connecting cord is a matter of convention, local regulations and local standards. No one set of rules really covers the whole planet other than equipment has to be safe for the end user to use.

In my country you have to have grounded outlets and you have to have RCD protection on all socket outlets that can take a standard plug in them. You also have to have circuit breakers on each circuit. Fuses are only allowed on older installations untill the house is sold, when you have to upgrade to have RCD and breakers.

[madires]

Wow, I cannot believe some of the answers coming from so-called "electrical engineers", holy shit!

Me neither!

GFCI and arc-fault breakers are a last resort attempt at circumventing the process of your death. If you are lucky, you won't die. Don't bet your life on them. 100mA at mains voltage and you are dead.

The problem with RCDs and breakers is that you can't trust them to be always working. They can fail too! And if the RCD is the only line of defense you got, you got a nice SPOF! Proper grounding, RCDs and breakers all together are the best insurance currently available. Add an isolation transformer for the lab and you're ready to EE.

[ResR]

In Estonia, it is not controlled who replaces the wall sockets or cables. Electrical parts (cable, sockets, switches, RCD and circuit breakers) is freely available to anyone to buy and install, only electric meter and main circuit breaker is sealed into electrical box. Most of the installations in here is soviet era TL-C system with aluminium wiring done by plumbers .
For my opinion, grounding metal cases is a must, double insulation even better. You can't rely just on a RCD and circuit breaker to break you from phase, because any mechanical equipment can and will eventually get stuck/break/misfire.

[Tepe]

The problem with RCDs and breakers is that you can't trust them to be always working. They can fail too!

Indeed they can. As can a ground connection.

[alm]

Which is why you want to have both.

[Tepe]

Which is why you want to have both.

Or check the actual accident statistics before making a decision that might entail spending a lot of money.

[madires]

Which is why you want to have both.

Or check the actual accident statistics before making a decision that might entail spending a lot of money.

You got no idea what a DYI father is capable of :-) If a new safety device with some new technology would be available I would buy it immediately and throw it into my father's distribution panel.

[edavid]

Which is why you want to have both.

Or check the actual accident statistics before making a decision that might entail spending a lot of money.

That's true, you're much more likely to be killed driving to the store to buy the rewiring supplies

[westom]

If a new safety device with some new technology would be available I would buy it immediately and throw it into my father's distribution panel.

  GFCI as required in North America trips on 5 milliamps.  And will not reset if damaged.  That second condition was an upgrade from original designs implemented and required in the mid 1970s.

[Monkeh]

If a new safety device with some new technology would be available I would buy it immediately and throw it into my father's distribution panel.

  GFCI as required in North America trips on 5 milliamps.  And will not reset if damaged.  That second condition was an upgrade from original designs implemented and required in the mid 1970s.

So, they trip if you sneeze or dare to use an SMPS or two on them.

[westom]

So, they trip if you sneeze or dare to use an SMPS or two on them.

  Any SMPS that leaks even one milliamp is defective.  Any device that trips even intermittently on a 5 milliamp GFCI/RCD is defective.

  We expect leakage in our designs and equipment to be below 100 microamps.

[c4757p]

The only time I've ever tripped a U.S. GFCI was when testing them, and I have plugged many SMPSs into them. Sneezed on them, too.

[SeanB]

Here you find consumer RCD units trip at 30mA for single or three phase units, but you can get industrial units that are 100mA, 1A or 2A, mostly used for protection of 3 phase heating banks or industrial loads where you can get a high leakage current but want protection against a heater shorting to ground at the low end. They often have a delay built in to reduce nuisance tripping at the set current for a few seconds, but trip fast at high imbalances.

I had some old 1970's vintage units i took apart years ago. Inside was the core with the 2 mains windings on it wound so as to balance the magnetic field to zero in the nice permalloy tape core, and a sense coil of a few dozen turns of fine wire and a test wire of a dozen or so turns of fine wire with a resistor to limit the current through the test switch when pressed. The sensed current went to a small PCB with a bridge rectifier on it, made from 2 1N4001's and a pair of 12V 400mW zener diodes. One tantalum dipped capacitor for smoothing, then a trigger circuit made from a SCS to dump charge into the trip coil, a small coil with an Alnico magnet holding the latch mechanism in the closed position. When the imbalance current was enough the SCS would fire, dumping the stored charge on the capacitor into the coil, and opposing the Alnico magnet so it would be pushed off the coil by a light spring and then via a plunger pin tripping the main mechanism to switch the breaker to off. The newer ones do not have the PCB, the sense coil directly drives the trip coil, driving it into magnetic saturation and releasing the magnet irrespective of the pulse polarity.

On a side note these trip mechanisms were used on the first generation of prepaid electricity meters made by a local company, as they were already approved. These had a flaw in that they covered the existing test switch with a label, and the enterprising users found they could defeat the meter by pushing a pin through the label and then disabling the trip mechanism with it, thus getting free power. The next versions all were split, with the indoor unit only being a keypad that communicated via a 2 wire interface to the power switch outside up the pole, the 2 wires being built into the split concentric supply cable used for them as 2 0.5mm copper cores placed next to the earth strands.  A central 7 core 10mm line, 7 individually insulated neutrals making up 10mm and 3 cores making up the earth of 6mm with a blue and white 0.5mm control cores.

[IanB]

The only time I've ever tripped a U.S. GFCI was when testing them, and I have plugged many SMPSs into them. Sneezed on them, too.

Not that anyone should ever do this, of course, but if a friend of yours should happen to try testing a GFCI by placing finger and thumb between live and ground terminals to see how much it hurts before it trips, make sure that friend does not accidentally touch the unprotected live terminal instead of the protected live terminal 

Not that anyone would be dumb enough to try that of course... 

[c4757p]

Yeah, I know what 120VAC feels like, TYVM 

On the other hand, I wonder where 240VAC falls between 120VAC and 600VDC (damn oscilloscope had HV outside the HV warning shield...  )

[madires]

So, they trip if you sneeze or dare to use an SMPS or two on them.

  Any SMPS that leaks even one milliamp is defective.  Any device that trips even intermittently on a 5 milliamp GFCI/RCD is defective.

  We expect leakage in our designs and equipment to be below 100 microamps.

The local regulations allow that the Y filter caps in a moveable device may pass up to 3.5mA (<= 22nF). The standard RCD trips at 30mA to allow using several devices with Y filter caps (usually SMPSs) at the same time.   

[madires]

The only time I've ever tripped a U.S. GFCI was when testing them, and I have plugged many SMPSs into them. Sneezed on them, too.

The small wart warts don't got PE  :-) The fun starts with grounded SMPSs like the ATX PSUs for PCs since they got two Y caps besides the X cap and the common-mode choke. The Y filter caps are between hot and PE, also neutral and PE.

[madires]

Yeah, I know what 120VAC feels like, TYVM 

On the other hand, I wonder where 240VAC falls between 120VAC and 600VDC (damn oscilloscope had HV outside the HV warning shield...  )

We got twice as much fun 

[duskglow]

240V is halfway to arc flash territory.  At 480V you start needing PPE.

[Monkeh]

240V is halfway to arc flash territory.  At 480V you start needing PPE.

Oh, you need PPE with 240V.

[Emil]

Thanks to everyone who replied to my question.

It sounds like my lab should definitely be grounded.

I will see if there is a ground cable behind the sockets. If there is I will check that it is actually connected to ground and install grounded sockets. If not I will get an electrician to fix it.
 

[rsjsouza]

GFCI as required in North America trips on 5 milliamps.  And will not reset if damaged.  That second condition was an upgrade from original designs implemented and required in the mid 1970s.

Thanks for this remark; I could almost swear they were specified for 30mA, and had todouble-check to see if I was not crazy...

We got twice as much fun 

Yep... Fun indeed. My home town was 220V/380V and I lost count of how many shocks I got from my experiments as a kid...

[Monkeh]

GFCI as required in North America trips on 5 milliamps.  And will not reset if damaged.  That second condition was an upgrade from original designs implemented and required in the mid 1970s.

Thanks for this remark; I could almost swear they were specified for 30mA, and had todouble-check to see if I was not crazy...

They are. Everywhere else. But we protect circuits.