How to properly ground ESD mat.

[Eklutna]

I'm new to the hardware side of electronics.  While setting up my work bench, I purchased a ESD mat from Jameco (link found at the bottom of this post).  The product didn't come with any directions for how to ground the mat, so I did a search on Google.  What I found was there are a lot of varied opinions on how to properly ground the mat, and some options seem to be potentially dangerous according to some people.  Since I know better than to trust just anyone one the internet, I was hoping Dave might do a blog about how to properly ground an ESD mat.

The most common options I found on Google were:
a)  Attach the ground wire to an electrical outlet via the small screw that holds the plate cover on.  (This option had the most strong opinions for/against.)
b)  Attach the ground wire to a conductive stake driven into the ground.
c)  Attach the ground wire to some large metal object.

Note:  The ESD mat I purchased gives off a REALLY bad rubbery odor even after 2 days out of it's packaging.  I have now placed it in outside (protected from the elements though) and hope that after a week or two of airing out the odor will go away.  I came across someone else on Google that had a similar issue.  Maybe another blog could compare various ESD mats and point out the ones that give off a bad odor (among other differences).

http://www.jameco.com/webapp/wcs/stores/servlet/ProductDisplay?freeText=2152075&langId=-1&storeId=10001&productId=2152075&search_type=jamecoall&catalogId=10001&ddkey=http:StoreCatalogDrillDownView

[Baliszoft]

I got a vermason rubber mat. Had some bad smell, but was OK after 2-3 days.
It is grounded via the wall outlet ground and a 1meg resistor. I cannot imagine what could be so dangerous thru 1M, when all the appliances are connected directly to ground.

[alm]

a)  Attach the ground wire to an electrical outlet via the small screw that holds the plate cover on.  (This option had the most strong opinions for/against.)

As long as your wall outlet is wired correctly, this would be the best solution. You could also take a grounded plug and connect a wire to the ground pin. Put a 1 Mohm resistor in the plug in series with the wire to reduce the maximum current that can flow to safe levels.

b)  Attach the ground wire to a conductive stake driven into the ground.

This is overkill and may be inferior (if done improperly) to just using the ground of your mains connection. Unless the ground connection in your lab is bad, in that case I would recommend getting an electrician to fix the ground connection, since this is critical for electrical safety.

c)  Attach the ground wire to some large metal object.

What good would this do? It's either useless or a roundabout way of connecting it to the mains ground pin.

[Uncle Vernon]

It's also essential to regularly touch the tip of your tinfoil hat to any mat to equalisethe  body's aural capacitance. 

[gregariz]

a)  Attach the ground wire to an electrical outlet via the small screw that holds the plate cover on.  (This option had the most strong opinions for/against.)

As long as your wall outlet is wired correctly, this would be the best solution. You could also take a grounded plug and connect a wire to the ground pin. Put a 1 Mohm resistor in the plug in series with the wire to reduce the maximum current that can flow to safe levels.

b)  Attach the ground wire to a conductive stake driven into the ground.

This is overkill and may be inferior (if done improperly) to just using the ground of your mains connection. Unless the ground connection in your lab is bad, in that case I would recommend getting an electrician to fix the ground connection, since this is critical for electrical safety.

c)  Attach the ground wire to some large metal object.

What good would this do? It's either useless or a roundabout way of connecting it to the mains ground pin.

A couple of years ago I was on a job where I was probing a design and noticed noise throughout all the measurements I was making. The QA department had the whole factory fitted with anti-static mats and all the grounds then connected to the 'Building ground' which meant that they wired it to the building trusses. After a while it dawned on me thats where the noise was being coupled from ie from the mat to the job. All the test instruments were of course grounded through the mains outlet. Reconnecting the mat grounds to the instrument ground solved it.

My recommendation would be to ground the mat to the same ground as the test intrument power ground.

[AlphZeta]

Actually, all three method would work. The idea of dissipating static electricity is to let the built-up charge go somewhere.

a)  Attach the ground wire to an electrical outlet via the small screw that holds the plate cover on.  (This option had the most strong opinions for/against.)

The reason people have strong opinion on this is that the ground wire isn't always wired properly. And if the ground wire isn't actually tied to the ground via very low resistance connection an out-of-balance three phase could create a high voltage on the ground. While highly unlikely, this could be a very dangerous situation.

b)  Attach the ground wire to a conductive stake driven into the ground.

This one will work for the same reason as above. And it will work sufficiently well as long as your ground wire has reasonable contact with the earth ground. The resistance doesn't have to be as critical as in a) since you are only dissipating static electricity.

c)  Attach the ground wire to some large metal object.

This will actually work. The physics behind static electricity is Q=CV where Q is the charge and C is the capacitance and V is the voltage. So in theory, if you have a large metal surface, the C will be significantly larger than your body capacitance and thus the Q will be redistributed to the two parallel connected capacitors (you and the surface) and V will drop significantly if metal object has enough capacitance. In this situation, the metal object actually does not need to be grounded at all! As long as the metal object has some way of dissipating its charge, there won't be any static electricity build up.

My personal favorite is b). Since it is rather simple (all you need to do is tie one end to your copper piping or something).

[alm]

Actually, all three method would work. The idea of dissipating static electricity is to let the built-up charge go somewhere.

It's about getting everything on the same potential. Compare birds on the high voltage lines. Tying an ESD mat to 110 kV would be fine when working on this high voltage line, since you would be at the same potential, no semiconductors would be damaged. It's just that a lot of stuff tends to be at ground potential due to things like grounded equipment and concrete, so ground is a safe and convenient potential.

The reason people have strong opinion on this is that the ground wire isn't always wired properly. And if the ground wire isn't actually tied to the ground via very low resistance connection an out-of-balance three phase could create a high voltage on the ground. While highly unlikely, this could be a very dangerous situation.

So all metal equipment cases (eg. desktop PCs) may be live, but you worry about a ESD mat with a 1 Mohm series resistance and a fairly high surface resistance to said ground connection? I'd say that you have bigger things to worry about.

c)  Attach the ground wire to some large metal object.

This will actually work. The physics behind static electricity is Q=CV where Q is the charge and C is the capacitance and V is the voltage. So in theory, if you have a large metal surface, the C will be significantly larger than your body capacitance

What's the capacitance of a single piece of metal? I seem to remember a capacitor requiring two plates, where's the second one?

and thus the Q will be redistributed to the two parallel connected capacitors (you and the surface) and V will drop significantly if metal object has enough capacitance. In this situation, the metal object actually does not need to be grounded at all! As long as the metal object has some way of dissipating its charge, there won't be any static electricity build up.

If the metal object has significant capacitance to ground, then it would be fine, but this is hard to guarantee (what if someone changes the distance of the object to the floor, for example?).

[AlphZeta]

What's the capacitance of a single piece of metal? I seem to remember a capacitor requiring two plates, where's the second one?

Not necessarily, only mutual capacitance requires two pieces of metal. A sphere has a self-capacitance of 4*pi*e0*R for instance. I am not saying just because it would work it is the best method

[alm]

You're correct, I forget about self-capacitance. Still, that has got to be a pretty big sphere. For ESD purposes the human body is often assumed to have a capacitance of 100 pF. For a sphere that has an order of magnitude more capacitance, it has to have a radius of about 10 meter. Flat surfaces would probably have to be even larger.

[SeanB]

I would ground via a resistor to the mains earth, provided that you get and use a mains tester, which will show incorrect wiring and faults. As well the resistor must be a 0.5W or higher power one, as it is the only thing that is preventing flashover in case of a fault. I prefer a 1W unit, as it is physically large, generally has thicker wire leads and is easy to heatshrink into the cable. Use a heatshrink with a hot melt layer in it to provide extra protection against the wires breaking loose, and even better is to shrink it in, rotate it 180 degrees so the input and output cables run next to each other and shrink a further layer onto it, so the wires are provided strain relief, and do not act directly on the soldered joints.

[bfritz]

You're correct, I forget about self-capacitance. Still, that has got to be a pretty big sphere. For ESD purposes the human body is often assumed to have a capacitance of 100 pF. For a sphere that has an order of magnitude more capacitance, it has to have a radius of about 10 meter. Flat surfaces would probably have to be even larger.

There is capacitance between any two bodies in space.  So, just choose a large sphere.  There happens to be one that is used almost universally for this purpose... It is called Earth!

In a lab, you really do need a grounded outlet.  So, the first order of business is to check that the outlet is properly grounded.  Then make sure the neutral and hot wires are wired correctly for that outlet.  Once those are done, using a 1Mohm resistor between the mat and ground (as the commercial grounding connectors for the mats do) provides a nice limiting current even if the ground at the outlet were to become energized.  Remember you would have the impedance of the mat, plus the 1Mohm, plus any impedance from your body to the mat, and your body to earth ground.  The risk is really quite low.  The other good thing to to is add a Ground Fault detection device to the power for your bench and equipment.