Do's and don'ts in home lab with no earth ground

[bottledwater]

Hi, there's no earth ground in my plugs at home. Driving a stake into the ground is not an option and nulling in the box outside is not allowed.

If this is what you had to work with:

1) How would you connect test gear and power supplies that ought to be grounded to earth? What would you do with their chassis grounds?

2) What are some practices you'd adopt around handling static sensitive components from the moment they arrive in the post?

3) Are there some types of projects you wouldn't do in this earth-deficient "lab", like mains connected power amp repairs?

My concern is the safety of my components, my gear, and myself. I have a relatively nice power supply and oscilloscope, and I'm in the market for a good function generator, but I'm not sure how much I should try to do here. Failing logic ICs I can handle, but I'd be sad to see any of my gear blow up.

I mostly play around with breadboards and the occasional PCB. My main interest is basic battery powered amplifiers, CMOS logic, playing around with MCUs. For a year I haven't had anything fail on me so I wonder how much is theory and how much is practice.

Thanks!

[bdunham7]

Hi, there's no earth ground in my plugs at home. Driving a stake into the ground is not an option and nulling in the box outside is not allowed.

Do you simply have two-pin sockets or is there an earth/ground terminal that just isn't connected?

[Grandchuck]

Can you access a metal water pipe?  You would need to verify that it is indeed safe to do so!

[themadhippy]

I'd be jumping for joy if i had a totally earth free location to work in

[bottledwater]

Hi, there's no earth ground in my plugs at home. Driving a stake into the ground is not an option and nulling in the box outside is not allowed.

Do you simply have two-pin sockets or is there an earth/ground terminal that just isn't connected?

Yes, 2-pin sockets everywhere and no earth coming into the building anywhere, I've looked. The building is old so it's technically legal.

Can you access a metal water pipe?  You would need to verify that it is indeed safe to do so!

I do have heat and water pipes coming in but it would be a long wire. In a pinch maybe I could access the main water pipe downstairs. But I heard it isn't safe to get your earth ground this way, at least for test gear? Maybe it would be safe/good for my ESD mat?

[bottledwater]

I'd be jumping for joy if i had a totally earth free location to work in

Hæhæ, nice

[CatalinaWOW]

Remember that all voltage is relative.  Draw a diagram, in your head or on paper, of the gear, devices and components you are worried about and how they are connected electrically.  This diagram will show how potential differences can arise between various parts.  And give you ideas how to control them.

Example:  For static safety the problem comes when your body is at a different potential than your parts, and the current that arises when you touch the part and force them to go to the same potential.

Solution:  Put a grounding mat on your bench.  It doesn't have to be grounded.  Now touch the grounding mat before you touch your parts.  Touching the grounding mat forces your body and the mat to be at the same potential.  Current has flowed between them to achieve this condition.  But both the mat and your body are robust to this current and the only damage is a mild tingle you may have felt.  After this, you and your parts are at the same potential and you can handle them freely.

The same principal applies everywhere, but there is an added problem.  Grounding on equipment is often as much for safety purposes as it is for noise control.  A fault in the equipment could allow the chassis or other touchable places to be connected to either the hot or neutral lines.  When grounded this failure either results in a blown fuse or non hazardous situation depending on which side the fault comes from.  Without the ground you don't have this safety backup.

Another problem arises if the floor surface is partially conducting (not open but high resistance).  In cases like this it is possible for there to be a voltage between neutral and the floor, which is the potential your body will most likely end up at.  Measure the voltage and be aware of this in the mental map identified in the first paragraph.  A little thought will also suggest how you can evaluate the impedance between the floor and wherever the ground and neutral actually are connected, which can give you more information on precautions you can take.  The danger zone lies in the middle.  Very low and very high impedances will not usually be a problem.

While your situation does expose you to some additional risk, it is not inherently terribly hazardous, and the risks can be controlled by thoughtful use.   There are literally millions of people who have worked successfully and safely under these conditions.

[Siwastaja]

Are you going to build or repair any mains powered designs? If not, any of this does not matter the slightest. What comes to lab equipment, it's no different from normal home electronics: the same regulations apply, the products must be safe to be legally sold. Beware of Ebay China imports or very old antique gear, though.

For ESD, if you find any kind of steel pipe etc., you can attach the 1Mohm resistor to that and use it to ground your ESD mat. (Mat is really all you need; as you rest your hands on the mat, a special strap offers little additional help.)

[bdunham7]

Yes, 2-pin sockets everywhere and no earth coming into the building anywhere, I've looked. The building is old so it's technically legal.

OK, are there RCDs (GFIs) at the outlets or in the service panel?

[Ian.M]

Look up "equipotential bonding".  If your work area has full equipotential bonding, and you take precautions against coming into contact with unbonded conductors and potentially conductive surfaces (e.g. bare concrete floors reinforced with rebar), and you take precautions against contact with the bonded conductors from outside the area, its at least as safe for YOU as if you had a proper PE conductor ('earth' ground).

However do not ground to metal water pipes or other metal pipes if it passes through any area outside your control between its point of entry into the building and your proposed grounding point, as if any part of the pipes gets replaced with plastic, and a fault subsequently occurs in your work area there is a high risk of the pipes becoming mains 'live', risking electrocuting other occupants of the building. 
Exception: If *ALL* mains supplies to your bench pass through isolating transformers, and you do not extend their secondary side circuits outside your equipotential bonding zone, its impossible for them to make the bonding conductor 'live', so you can ground the bonding conductor to water pipes that pass outside your control.  If you do so, remember that an isolating transformer removes all upstream RCD (GFCI) protection from its secondary side.

[bottledwater]

Yes, 2-pin sockets everywhere and no earth coming into the building anywhere, I've looked. The building is old so it's technically legal.

OK, are there RCDs (GFIs) at the outlets or in the service panel?

I have hot and neutral coming into a RCD in the service panel. One phase then goes into the apartment via old cloth-style wires. That's all. I could/should get some more to have things separated but I wanted to know about the earth issue first.

... However do not ground to metal water pipes or other metal pipes if it passes through any area outside your control between its point of entry into the building and your proposed grounding point, as if any part of the pipes gets replaced with plastic, and a fault subsequently occurs in your work area there is a high risk of the pipes becoming mains 'live', risking electrocuting other occupants of the building. ...

I don't have any pipes that fulfill this criterion so I guess connecting to one of them is not an option.

[bottledwater]

Thank you, this is very helpful. Let me summarize what I think I understand:

My test gear and I will be fine, the same way my stereo is "fine" in a thunderstorm, as long as I don't do mains repairs.

To keep my components safe I can stick them in an ESD box which I keep at the same potential as the ESD mat that I work at. That way, any time I sit down by my mat, my components and I will slowly be brought to the same potential through the 1M resistor.

What if I then come in with my un-earthed oscilloscope ground to probe my sensitive component? Since chassis ground of my oscilloscope (and thus its input ground) floats, I'll zap the component. I guess if I touch the oscilloscope ground to my mat before I probe the component, the component will be ok when I do.

The chassis grounds of my test gear will be connected to each other ("equipotential") through the power strip that I connect them to. But they will be floating. Would I connect my ESD mat and ESD component box to this floating power strip as well?

[Ian.M]

If you 'ground' your ESD mat and other ESD prevention equipment to your bonding conductor (through the usual 1Meg safety resistors), and *ALL* your test equipment with three pin power leads have their PE ground connected to the *same* bonding conductor, along with all exposed metal nearby of significant size (i.e. don't bother with draw handles etc. but a metal bench or desk frame should be bonded), your parts and boards will be at no higher ESD risk than if you did have a PE conductor to true 'earth' ground.

The bonded zone will probably have enough leakage to 'earth' ground to prevent the build-up of high static charge voltages, but if you can, a 1Meg resistor (actually 3x 330K in series to reduce the risk of it ever shorting out) between your bonding conductor and a waterpipe will discharge static charge without risking energising the water pipes with mains if anything faults to the bonding conductor.

[langwadt]

Yes, 2-pin sockets everywhere and no earth coming into the building anywhere, I've looked. The building is old so it's technically legal.

OK, are there RCDs (GFIs) at the outlets or in the service panel?

here that is a requirement is all residential installations no matter how old

[langwadt]

Thank you, this is very helpful. Let me summarize what I think I understand:

My test gear and I will be fine, the same way my stereo is "fine" in a thunderstorm, as long as I don't do mains repairs.

To keep my components safe I can stick them in an ESD box which I keep at the same potential as the ESD mat that I work at. That way, any time I sit down by my mat, my components and I will slowly be brought to the same potential through the 1M resistor.

What if I then come in with my un-earthed oscilloscope ground to probe my sensitive component? Since chassis ground of my oscilloscope (and thus its input ground) floats, I'll zap the component. I guess if I touch the oscilloscope ground to my mat before I probe the component, the component will be ok when I do.

The chassis grounds of my test gear will be connected to each other ("equipotential") through the power strip that I connect them to. But they will be floating. Would I connect my ESD mat and ESD component box to this floating power strip as well?

in my experience the biggest issue is when somethings are earthed and some not, the things that are not earthed will have their "earth"
floating on mains/2 realtive to true earth via the y capacitors in most supplies. That can give you a tingle or fry electronics

[JustMeHere]

If you're really worried, put a GFCI breaker in your panel. Other than that most of the time your DUT is on an ungrounded, isolated power supply anyway.

[jonpaul]

240v mains>>1:1 isolation transformer, of desired KVA.

We floated our entire benches and lab that way to safely design and debug KW SMPS and electronic ballasts

Never had any problems

Search for 240/240V isolation transformers

Jon

[TimFox]

Thank you, this is very helpful. Let me summarize what I think I understand:

My test gear and I will be fine, the same way my stereo is "fine" in a thunderstorm, as long as I don't do mains repairs.

To keep my components safe I can stick them in an ESD box which I keep at the same potential as the ESD mat that I work at. That way, any time I sit down by my mat, my components and I will slowly be brought to the same potential through the 1M resistor.

What if I then come in with my un-earthed oscilloscope ground to probe my sensitive component? Since chassis ground of my oscilloscope (and thus its input ground) floats, I'll zap the component. I guess if I touch the oscilloscope ground to my mat before I probe the component, the component will be ok when I do.

The chassis grounds of my test gear will be connected to each other ("equipotential") through the power strip that I connect them to. But they will be floating. Would I connect my ESD mat and ESD component box to this floating power strip as well?

in my experience the biggest issue is when somethings are earthed and some not, the things that are not earthed will have their "earth"
floating on mains/2 relative to true earth via the y capacitors in most supplies. That can give you a tingle or fry electronics

Precisely.  If you simultaneously come into contact with the metal case of an ungrounded box (with a normal IEC-connector line filter) and something that is grounded (radiator, faucet, water pipe, etc.) you will encounter 1/2 the line voltage through a medium-size capacitor (maybe 4.7 nF) and get a (normally) non-lethal shock.
If you are on a ladder, or live in a 220 V country, that shock can still be hazardous.

[bottledwater]

in my experience the biggest issue is when somethings are earthed and some not, the things that are not earthed will have their "earth"
floating on mains/2 relative to true earth via the y capacitors in most supplies. That can give you a tingle or fry electronics

Precisely.  If you simultaneously come into contact with the metal case of an ungrounded box (with a normal IEC-connector line filter) and something that is grounded (radiator, faucet, water pipe, etc.) you will encounter 1/2 the line voltage through a medium-size capacitor (maybe 4.7 nF) and get a (normally) non-lethal shock.
If you are on a ladder, or live in a 220 V country, that shock can still be hazardous.

Interesting. So if I connect myself to the metal case of a mains powered piece of gear through my ESD mat, I'll be floating at 1/2 the line voltage, but this is not a problem?

[themadhippy]

If you're really worried, put a GFCI breaker in your panel.

and if the incoming supply is not referenced to ground?

[langwadt]

in my experience the biggest issue is when somethings are earthed and some not, the things that are not earthed will have their "earth"
floating on mains/2 relative to true earth via the y capacitors in most supplies. That can give you a tingle or fry electronics

Precisely.  If you simultaneously come into contact with the metal case of an ungrounded box (with a normal IEC-connector line filter) and something that is grounded (radiator, faucet, water pipe, etc.) you will encounter 1/2 the line voltage through a medium-size capacitor (maybe 4.7 nF) and get a (normally) non-lethal shock.
If you are on a ladder, or live in a 220 V country, that shock can still be hazardous.

Interesting. So if I connect myself to the metal case of a mains powered piece of gear through my ESD mat, I'll be floating at 1/2 the line voltage, but this is not a problem?

your connection to the ESD mat has a large resistor limiting current and anyway you wouldn't feel it unless you touch something that is really earthed

I suspect quite few TVs and sound cards have been zapped by plugging the audio of an unearth PC into a TV that is earthed through the cable TV antenna connector

[TimFox]

in my experience the biggest issue is when somethings are earthed and some not, the things that are not earthed will have their "earth"
floating on mains/2 relative to true earth via the y capacitors in most supplies. That can give you a tingle or fry electronics

Precisely.  If you simultaneously come into contact with the metal case of an ungrounded box (with a normal IEC-connector line filter) and something that is grounded (radiator, faucet, water pipe, etc.) you will encounter 1/2 the line voltage through a medium-size capacitor (maybe 4.7 nF) and get a (normally) non-lethal shock.
If you are on a ladder, or live in a 220 V country, that shock can still be hazardous.

Interesting. So if I connect myself to the metal case of a mains powered piece of gear through my ESD mat, I'll be floating at 1/2 the line voltage, but this is not a problem?

It's only a problem when you complete the circuit (e.g., by touching some heating or plumbing metalwork).
A bird can sit on an uninsulated high-voltage wire without damage.

[bottledwater]

in my experience the biggest issue is when somethings are earthed and some not, the things that are not earthed will have their "earth"
floating on mains/2 relative to true earth via the y capacitors in most supplies. That can give you a tingle or fry electronics

Precisely.  If you simultaneously come into contact with the metal case of an ungrounded box (with a normal IEC-connector line filter) and something that is grounded (radiator, faucet, water pipe, etc.) you will encounter 1/2 the line voltage through a medium-size capacitor (maybe 4.7 nF) and get a (normally) non-lethal shock.
If you are on a ladder, or live in a 220 V country, that shock can still be hazardous.

Interesting. So if I connect myself to the metal case of a mains powered piece of gear through my ESD mat, I'll be floating at 1/2 the line voltage, but this is not a problem?

It's only a problem when you complete the circuit (e.g., by touching some heating or plumbing metalwork).
A bird can sit on an uninsulated high-voltage wire without damage.

I understand, although it was suggested earlier to connect myself and my chassis grounds to a metal pipe through a current-limiting 1M resistor (or 3x330k). Is that still the recommendation if the chassis grounds bring us to 1/2 the line voltage?

[Doctorandus_P]

I'd be jumping for joy if i had a totally earth free location to work in

Maybe go practice your jumping skills beyond the Kuiper belt? 

[Ian.M]

The resistors I suggested between bonding conductor and grounded waterpipe if available are to discharge DC static charge build-up to prevent the (small) risk of ESD induced flashover inside any nominally grounded equipment's PSU.  They will do so successfully in the presence of AC leakage current driving the grounding conductor to approximately half the line voltage but wont ground out the leakage current, which would need a low impedance dedicated grounding conductor, marked as such where it passes through any common areas.