"Double insulated, do NOT earth." Why?

[max.wwwang]

Earthing is a seemingly trivial but actually very complex topic. I used to think it was simple and I knew about this, but I knew virtually nothing, which was dangerous. Only recently, I got to know a little bit about this, including the topic of isolation transformer, through discussions in this forum and watching YouTube videos. But there are still some questions I don't have an answer to. Here is one.

On the casing of double-insulated appliances, often we see this on the label: "Double insulated, do NOT earth". I understand the reason why there is no NEED for earthing due to double-insulation, which means earthing does not provide more protection than without. But I don't understand why we should NOT (not need not) earth a double-insulated device. What's the RISK of doing this, beyond no more protection?

I have googled this topic but have not found a clear explanation.

[brabus]

Earthing a double-isolated appliance is a no-go, simply because the producer has not defined a proper earthing contact for it. One could find a screw terminal, a hole in the metal case or even an actual earthing pole used in a different, earthed variant of the same appliance: please do NOT use them as earthing contacts. This would violate the manufacturer's specification regarding AC grid connection, potentially voiding the warranty and posing a risk for the end user.

[max.wwwang]

Earthing a double-isolated appliance is a no-go, simply because the producer has not defined a proper earthing contact for it. One could find a screw terminal, a hole in the metal case or even an actual earthing pole used in a different, earthed variant of the same appliance: please do NOT use them as earthing contacts. This would violate the manufacturer's specification regarding AC grid connection, potentially voiding the warranty and posing a risk for the end user.

Thanks. But I'm interested in what exactly the risk to the end user is.

[brabus]

The risk is given by the fact that the manufacturer has not defined an earthing strategy for the product, so the consequences of forced earthing are unpredictable. In other words, the manufacturer deemed the risk of failure of the double insulation as far more unlikely than the risk of failure of the earthing system.

The most pesky circumstance is the failure of the earthing system on the AC side. If the main building earthing fails and another device is discharging current through the earth, the metal casing (that is supposed to be floating in a double-insulation device) would become energized, posing a risk to the end user.

A practical example: suppose you decide connect the secondary side ground of a phone charger to earth, although the charger is isolated. Everything goes well, you hold the cellphone in your hand and nothing happens. But if the building's main earth connection fails (it's not that unusual!), you WILL become the discharge path of any current flowing through the earth path. So, it is better to stay floating (i.e.: double-isolated) than becoming an active part of the earthing system.

[SeanB]

The risk is failure of the internal parts, as often the internal insulation is not designed, or intended, to act as primary insulation, and can break down destructively with applied voltage, as the double insulation designation means the entire external insulation is designed to withstand mains, so internal parts do not need this.

For small power supplies however the output can be grounded by other methods, this just means your ground path may now have currents flowing, that were normally suppressed by the internal filtering of the power supply.

[max.wwwang]

That's what I wanted to look into. Thanks.

The risk is given by the fact that the manufacturer has not defined an earthing strategy for the product, so the consequences of forced earthing are unpredictable. In other words, the manufacturer deemed the risk of failure of the double insulation as far more unlikely than the risk of failure of the earthing system.

This makes good sense. It's all about likelihood and assumptions.

A practical example: suppose you decide connect the secondary side ground of a phone charger to earth, although the charger is isolated. Everything goes well, you hold the cellphone in your hand and nothing happens. But if the building's main earth connection fails (it's not that unusual!), you WILL become the discharge path of any current flowing through the earth path. So, it is better to stay floating (i.e.: double-isolated) than becoming an active part of the earthing system.

The person will indeed become the "discharge path of any current flowing through the earth path" but only if there is another earth fault in another device connected to the same AC source (such as the live wire inside touches its earthed metal chassis). In this scenario, it would require two faults to really make earthing this double-insulated device a hazard (because with the earthing fault of the house, which means the path from the earth wire to the ground rod is broken, but without the second fault, which otherwise would connect the mains live wire to the house earth wire, it will only be that the person becomes a part of the path for any current that may flow to the ground, but this is not going to happen unless the second fault exists). Is this correct?

[max.wwwang]

The risk is failure of the internal parts, as often the internal insulation is not designed, or intended, to act as primary insulation, and can break down destructively with applied voltage, as the double insulation designation means the entire external insulation is designed to withstand mains, so internal parts do not need this.

For small power supplies however the output can be grounded by other methods, this just means your ground path may now have currents flowing, that were normally suppressed by the internal filtering of the power supply.

Thanks. But I'm not sure if I've understood this. We know there are double-insulated devices that have metal external cases. Are you referring to those with fully insulated (non-conductive) external cases?

[SeanB]

Double insulated with external metal means the insulation between inner mains times, and the metal part, is provided by 2 layers of insulation rated for mains use, or by having robust insulation, which is thicker.  Grounding internal parts means the inner insulation may fail, or may flash over, as it is designed to float, while the casing is there to provide insulation, by using double layers, or robust insulation.

[soldar]

On the casing of double-insulated appliances, often we see this on the label: "Double insulated, do NOT earth". I understand the reason why there is no NEED for earthing due to double-insulation, which means earthing does not provide more protection than without. But I don't understand why we should NOT (not need not) earth a double-insulated device. What's the RISK of doing this, beyond no more protection?

Maybe things are different in your part of the world but never in my many decades of tinkering and repairing have I come across such a label.

In any case, a user is supposed to NOT modify a product in any way. An appliance passes a test of conformity for the market where it is sold and users are supposed to respect that because most users do not know what they are doing because they are not qualified to judge any changes. This goes without saying in any labels.

Grounding is not a matter of just connecting some part to ground. All parts which need to be grounded need to be correctly connected and bound.

If I saw such a label I would have to think why it might be there.
It may be for sale in some part of the world where things are weird. 
It may be that the Asian manufacturer got a bad translation and really meant to say "Double insulated, does not need earth".
It may be that some yahoo is under the erroneous but widespread impression, (including some on this board) that if a device has double insulation it is prohibited from using ground protection. This is not so.

If done right earth plus double insulation is better protection than only one of the two. But the important thing here is that it has to be done right. You don't want some yahoo getting electrocuted because he (or she) thought he (or she) knew what they were doing.

[max.wwwang]

Maybe things are different in your part of the world but never in my many decades of tinkering and repairing have I come across such a label.

This is not a bad answer! I've attached another example -- not a label but a part of the user manual. Essentially the same message.

In any case, a user is supposed to NOT modify a product in any way. An appliance passes a test of conformity for the market where it is sold and users are supposed to respect that because most users do not know what they are doing because they are not qualified to judge any changes. This goes without saying in any labels.

You are missing the point, or I have not made it clear. I have no intention whatsoever to add grounding to such a device. I only wanted to understand why is this.

If I saw such a label I would have to think why it might be there.

Man that's exactly what I'm doing here!

It may be for sale in some part of the world where things are weird. 
It may be that the Asian manufacturer got a bad translation and really meant to say "Double insulated, does not need earth".
It may be that some yahoo is under the erroneous but widespread impression, (including some on this board) that if a device has double insulation it is prohibited from using ground protection. This is not so.

Good point, that's a possibility -- seriously. But unfortunately, be assured this is not the case. The example below is from a European product, so it's not Asian (without suggesting Asian is bad in any way of course!).

If done right earth plus double insulation is better protection than only one of the two. But the important thing here is that it has to be done right. You don't want some yahoo getting electrocuted because he (or she) thought he (or she) knew what they were doing.

Again, be assured -- this is wrong. Follow this thread, for your safety's sake! (At least reply #3 above makes some sense, though I'm still challenging/testing that view.)

[soldar]

In any case, a user is supposed to NOT modify a product in any way. An appliance passes a test of conformity for the market where it is sold and users are supposed to respect that because most users do not know what they are doing because they are not qualified to judge any changes. This goes without saying in any labels.

You are missing the point, or I have not made it clear. I have no intention whatsoever to add grounding to such a device. I only wanted to understand why is this.

Speaking of "missing the point" ... Read carefully what i wrote. I did not say or imply you were intending to add grounding to any device. I explained why a user manual might tell the user to not make any modifications... which is what you were asking: "why does it say this?"

If done right earth plus double insulation is better protection than only one of the two. But the important thing here is that it has to be done right. You don't want some yahoo getting electrocuted because he (or she) thought he (or she) knew what they were doing.

Again, be assured -- this is wrong. Follow this thread, for your safety's sake! (At least reply #3 above makes some sense, though I'm still challenging/testing that view.)

You say it is wrong. I disagree. You cannot explain why it is wrong. You seem to interpret a vague statement to say what you want it to say and you extrapolate from there to imply what you are saying. If you assert that having double insulation plus grounding is dangerous then it is up to you to prove it.

What is dangerous is not double insulation plus grounding. What is dangerous is having idiots tinkering with things they do not understand. Double insulation plus grounding is double protection if designed correctly by a competent engineer.

[blauerscharik]

What is dangerous is having idiots tinkering with things they do not understand.

What if an idiot doesn't want to stay an idiot and tries to learn something? Everybody has to start somewhere.
I never understood such comments.

[soldar]

What is dangerous is having idiots tinkering with things they do not understand.

What if an idiot doesn't want to stay an idiot and tries to learn something? Everybody has to start somewhere.
I never understood such comments.

I am not calling anyone an idiot and a person who wants to learn prudently is not an idiot. A person who jumps off a 100m cliff because he wants to "learn" if a 100m fall might hurt is an idiot.

Consumer products are full of warnings to the consumer and they are there for a reason. A qualified technician or someone with adequate training and knowledge can, at his own risk, disregard the notice. A person who is not qualified does well to ask those who know, which is exactly what the OP is doing and well done.

Those who think sticking scissors in power outlets is a way to learn anything are idiots learning about Darwin.

[Zenith]

What is dangerous is having idiots tinkering with things they do not understand.

What if an idiot doesn't want to stay an idiot and tries to learn something? Everybody has to start somewhere.
I never understood such comments.

Yes and you will make mistakes, sometimes expensive ones, and get the odd shock in the course of it. It's sensible to take appropriate steps.

I think the point here is it's unwise to modify a consumer appliance, subject to all sorts of regulations, in spite of a clear warning saying not to modify it in that way, with a chance of putting others at risk. You might have to explain your actions to a court.

[blauerscharik]

I am not calling anyone an idiot and a person who wants to learn prudently is not an idiot. A person who jumps off a 100m cliff because he wants to "learn" if a 100m fall might hurt is an idiot.

Consumer products are full of warnings to the consumer and they are there for a reason. A qualified technician or someone with adequate training and knowledge can, at his own risk, disregard the notice. A person who is not qualified does well to ask those who know, which is exactly what the OP is doing and well done.

Those who think sticking scissors in power outlets is a way to learn anything are idiots learning about Darwin.

True you didn't call anybody an idiot. But it seems that you are now paddling back a little by giving very dumb examples that have nothing to do with electronics.
If someone wants to fix their broken mixer then I would say go ahead. Stupid would be having it plugged in and trying to unscrew the blades with your fingers or something like that. I know common sense is not that common   
I just don't like reading things like "If your are not an engineer with 40 years of experience don't even think about repairing anything" (I'm exaggerating)

But anyway this going off topic and let's not clutter this post unnecessarily.

Have a good day. 

[soldar]

My point is that consumer products carry warnings and disclaimers directed to the average consumer.  That's all.

If a CRT monitor says do not open it because it is dangerous that warning is directed to the average, non-technical user. It does not apply to qualified technicians.

But the world is full of idiots who disregard safety notices and that is why every other day we see news of idiots who, disregarding warnings, got too close to a cliff and fell off or got too close to a wild animal and were attacked. And they record it on video for others to see. And others go and do similar things.

Getting back to the original question, the warning is there to stop someone who thinks they know what they are doing but is not an idiot because idiots will disregard the warning and knowledgeable people will know what they are doing.

[mikeselectricstuff]

An example : a power drill.
User drills into a live cable - if double-insulated, probably nothing bad happens.
If earthed via an untested/unspecified route, unspecified currents can pass though parts that are not designed for it - possible arc flash & physical damage that could be a risk both at the time, and causing damage to safety-critical parts.


 

[Gyro]

Sometimes you don't get the option of deciding whether you appliance gets grounded or not. I've designed and taken to market a bunch of metal cased, double insulated, consumer A/V gear. This may or may not get eathed via the A/V connectors, antenna connectors etc. depending on what other equipment or environment the end customer has. It's not a problem safety wise (the double insulation happens in the internal SMPS board, spacings, insulation barriers, restraining where internal wiring might stray etc, but it does significantly complicate EMC testing because you are required to test in typical usage scenarios - interconnects attached, cable shields grounded and ungrounded.

[soldar]

An example : a power drill.
User drills into a live cable - if double-insulated, probably nothing bad happens.
If earthed via an untested/unspecified route, unspecified currents can pass though parts that are not designed for it - possible arc flash & physical damage that could be a risk both at the time, and causing damage to safety-critical parts.

This could make sense at first glance but I disagree and I explain why.

The whole point is to consider the external conductive housing is either (1) grounded via a conductor or (2) "grounded" via the user.  If the housing becomes live, for any reason whatsoever, it doesn't matter if it's a drill or a refrigerator or a thingamagig, if it becomes live, what is better for the user: that (1) the current flow to earth via a cable or (2) that the current flow to earth via the user?

Grounding external conductive parts that can be touched by persons always enhances safety. It may be that double insulation is considered sufficient but (correctly) grounding the external parts will never decrease safety.

The problem might come when someone read that "grounding always increases safety" and he goes and grounds the device in the wrong (internal) place which should not be grounded.

Now someone will come up with a scenario where the ground conductor was mistakenly connected to 400V and it would have been better if the device wasn't grounded. Or something like that.

[soldar]

Sometimes you don't get the option of deciding whether you appliance gets grounded or not. I've designed and taken to market a bunch of metal cased, double insulated, consumer A/V gear. This may or may not get eathed via the A/V connectors, antenna connectors etc. depending on what other equipment or environment the end customer has. It's not a problem safety wise (the double insulation happens in the internal SMPS board, spacings, insulation barriers, restraining where internal wiring might stray etc, but it does significantly complicate EMC testing because you are required to test in typical usage scenarios - interconnects attached, cable shields grounded and ungrounded.

Yes, A/V introduces other considerations about ground loops, interference, etc. But I think here we are considering solely the safety of users.

But you are right that often you do not get to decide that the device not be grounded.

If I have a double insulated drill and I am drilling into a grounded pipe or metal post or whatever which is grounded then the drill itself is grounded and it would be absurd to say that if you have a double insulated drill then you cannot drill into grounded metal objects because this would cause the drill itself to become grounded.

[mikeselectricstuff]

An example : a power drill.
User drills into a live cable - if double-insulated, probably nothing bad happens.
If earthed via an untested/unspecified route, unspecified currents can pass though parts that are not designed for it - possible arc flash & physical damage that could be a risk both at the time, and causing damage to safety-critical parts.

This could make sense at first glance but I disagree and I explain why.

The whole point is to consider the external conductive housing is either (1) grounded via a conductor or (2) "grounded" via the user.  If the housing becomes live, for any reason whatsoever, it doesn't matter if it's a drill or a refrigerator or a thingamagig, if it becomes live, what is better for the user: that (1) the current flow to earth via a cable or (2) that the current flow to earth via the user?

there is also the situation where there are external conductive parts which are not in contact with the user. E.g. the chuck of the aforesaid motor on a drill with a plastic handle.

Grounding external conductive parts that can be touched by persons always enhances safety. It may be that double insulation is considered sufficient but (correctly) grounding the external parts will never decrease safety.

The problem might come when someone read that "grounding always increases safety" and he goes and grounds the device in the wrong (internal) place which should not be grounded.

Now someone will come up with a scenario where the ground conductor was mistakenly connected to 400V and it would have been better if the device wasn't grounded. Or something like that.

In the UK at least, that is a significant enough possibility that regs require EV chargepoints (and hot-tubs I think) to include protection against it. Neutral and Earth are typically bonded togther where the service enters the house.
An upstream neutral fault can render the earth connection live relative to actual earth. This is not normally a big risk, as everything inside the house, e.g. water pipes, is bonded to the mains earth so there is no voltage differential, however an exposed outdoor conductor, like the body of a charging EV can be come live with respect to "real" earth.

[brabus]

(...)
Grounding external conductive parts that can be touched by persons always enhances safety. It may be that double insulation is considered sufficient but (correctly) grounding the external parts will never decrease safety.
(...)

Sorry, this is just plain wrong, at least in the EU.
Grounding the metallic chassis of a double-insulation appliance voids the EN60335 (I will provide the exact passage in the norm for completeness' sake).

[soldar]

There is also the situation where there are external conductive parts which are not in contact with the user. E.g. the chuck of the aforesaid motor on a drill with a plastic handle.

Note that the rule is not concerned with whether a part is normally touched by the user during use. The rule states that any conductive parts which can or may be touched by a person must be protected by an acceptable means (grounding or double insulation).

With a drill the user will touch the chuck while setting up the drill bit and at other times. The chuck is an exposed part and needs to be protected.

Even parts which are inside a housing but are reachable by small fingers (via vent holes etc) are considered "touchable" and therefore need to be protected.

We could also talk about sticking pointy knives into toasters in an attempt to fish out the bread that got stuck. Every time I do it I am admonished it is dangerous.

[soldar]

(...)
Grounding external conductive parts that can be touched by persons always enhances safety. It may be that double insulation is considered sufficient but (correctly) grounding the external parts will never decrease safety.
(...)

Sorry, this is just plain wrong, at least in the EU.
Grounding the metallic chassis of a double-insulation appliance voids the EN60335 (I will provide the exact passage in the norm for completeness' sake).

Please do. If it is what I am thinking I believe it does not say what you think is says and it has been misinterpreted. But let's have the text and see.

ETA: i find it incredible that you have to pay to get the text. It is legislation, the people are assumed to know it, it should be free.

[brabus]

Excerpt from the IEC61558:


(still searching for additional normative references, it takes some time to find the exact topic in the norm).