Ground plane and antennas

[zachtheterrible]

I've always been curious about this but never actually needed to know. What exactly does a groundplane do in relation to an antenna? I've heard that it becomes the second dipole in a 1/4 wave antenna if the ground plane is connected to ground of the circuit.

Then I was just reading an article on the internet that says that the groundplane merely limits the downward radiation of the antenna, and that it doesn't even have to be connected to ground. Can someone shed some light on this? I searched google and couldn't find much of anything. I'm stumbling around in the dark!

 

[dknguyen]

From what I learned in electromagnetics (I forgot most of it but it was just a year ago!), a ground plane is conductive makes a "virtual charge" when a real charge is near. The charges in the plane rearrange to make it appear as if there is an opposite charge on the other side of the plane at the same position of the real charge (on the other side).


...............Conductive Sheet of Material (ground plane)
.............................|
Virtual charge (-)......|......(+)Real charge
.............................|
.............................|

The real and virtual charge have opposite fields and sort of cancel each other out. The closer the plane is to the real charge, the more closely the opposite fields overlap and the more emissions are cancelled out. From a very long distance, you cannot see the real charge because the virtual charge cancels out it's field. With infintesmal distance between the real charge and plane (or at infinite observing distance) you cannot see the real charge's field because it's being cancelled out. So you stick a ground plane very close to a source of EMI and it helps cancel out the EMI being radiated. The closer the plane is to the source, the nearer you have to be to detect more noise.

I also stumbled around the net for this, and extrapolated the above from the "virtual charge" method of solving electromagnetic problem in class with the mention of ground planes and virtual charges I saw noted very briefly on a website...I don't know about it much more than that. I would love to hear someone else's response...especially about the antenna part. I didn't quite understand that myself.

 

[Nigel Goodwin]

There's no need (and indeed no advantage) for a groundplane to be earthed, essentially the groundplane acts as a 'mirror' and the other half of the dipole is 'reflected' in it - crude and rough explanation, but it gets the idea over

The obvious use is in a car, where the body acts as a groundplane.

I used to have a Trio/Kenwood portable 2m transceiver, which worked off 10 AA NiCd's, and had a microphone on a curly lead. If you stood it on top of a car, the signal strength increase was considerable! - if you held it in the same place without the car, it peformed a lot worse.

 

[stevez]

I see a groundplane as Nigel does though dknguyen may have explained the physics in more detail. Look at it as the other half of a dipole. I see the term counterpoise applied to random wire antennas - that term may also apply here.

In discussions about antenna I'd guess that ground plane might actually have meant the earth. Quite often a vertical antenna or a random wire is one side of the antenna and the earth (ground) is the other - a flat plane. An efficient vertical antenna, such as a medium wave broadcast tower might appear to work off the earth ground however it most likely has conductive radials buried just below the surface - the number of them and the length contribute greatly to antenna system efficiency. The length of radial is related to frequency and as frequency increases - radial length decreases.

 

[JimB]

I think a lot of confusion is caused by using the word "ground" to describe the 0volt line in a circuit.
Sometimes the 0volt line is connected to ground, or earth as we say in the UK (and maybe other places as well).

In relation to an antenna, first consider a dipole.
If we connect a transmitter to a feeder cable (co-ax or twin feed) and connect the feeder to the dipole, RF current will flow in the feeder and in each leg of the dipole. The currents in each leg of the dipole will be equal and opposite.
If we disconnect one leg of the dipole, the current in the remaining leg will drop (theoretically) to zero, and the radiated signal strength will also drop to zero.
The current in the leg of the dipole has to flow somewhere, consider it as flowing through space to the other leg of the dipole so that a complete circuit is created.

If we have a situation where it is physically inconvenient to use a dipole, then a single quarter wavelength wire can be used, but the current in that wire has to flow somewhere, so we can use a "ground plane". The ground plane may be the body/chassis/circuit board of the transmitter, or for more predictable results, a sheet of conductive material under the single wire antenna, or often several lengths of wire a quarter wavelength long connected to the screen of the co-ax cable connecting to the antenna.

That is the simple explanation, there are many ifs, buts and complications to what I have just written, this is just the Janet and John version.

While I have been typing it also occurs to me that the expression "groundplane" is also used to describe a continuous layer of copper on a circuit board. Usually where high frequency signals are involved.
The groundplane is usually used as the 0volt return for the DC and signals, and because it is one sheet of copper, parasitic inductance low preventing unwanted coupling between various parts of the circuit. (Which is why high frequency circuits dont work on breadboards).

JimB

 

[dknguyen]

Is that's what has been going on? Ground plane is referring to two different things? That would explain a lot...unless they are both really just the same thing used in different ways.

 

[Nigel Goodwin]

Is that's what has been going on? Ground plane is referring to two different things? That would explain a lot...unless they are both really just the same thing used in different ways.

It's really the same thing, a flat piece of metal - as you say, just different applications of it.

 

[dknguyen]

There's no need (and indeed no advantage) for a groundplane to be earthed

But when making PCBS you shouldn't leave unconnected copper islands because they build up charge and can cause problems. Doesn't this also apply to ground planes?

 

[Nigel Goodwin]

But when making PCBS you shouldn't leave unconnected copper islands because they build up charge and can cause problems. Doesn't this also apply to ground planes?

The groundplane on a PCB should be conected to your reference point, usually 0V - but I see no need to do anything with unconnected copper islands, there's no where for them to build up charge from!.

In any case, the ground reference was for aerials! - but applies equally well to a PCB - do you have your portable radio grounded?, or your Walkman?.

 

[dknguyen]

The groundplane on a PCB should be conected to your reference point, usually 0V - but I see no need to do anything with unconnected copper islands, there's no where for them to build up charge from!.

In any case, the ground reference was for aerials! - but applies equally well to a PCB - do you have your portable radio grounded?, or your Walkman?.

And if I did? hehe

I have read that unconnected copper planes on PCBs can cause problems. I am unsure as to what those problems are, but in more than several articles it says that. Do you have to ground Faraday cages? I can't think of why you would need to, but I have seen that said, or were the articles saying that the Faraday cage IS the ground (from the perspective of things inside the cage).

 

[zachtheterrible]

Very very very interesting. Thanks for the explanations, it's starting to make a lot more sense to me now. I'm curious about dknguyen's question about whether faraday cages should be grounded or not.

 

[RadioRon]

No, Faraday cages don't need to be "grounded" to function at RF frequencies. However, it is common to ground them to the building ground system for the sake of safety. This is done so that faults from 60 or50 Hz mains to the cage won't make them "hot" at 120 or 240 V and cause a safety hazard.

 

[Nigel Goodwin]

Very very very interesting. Thanks for the explanations, it's starting to make a lot more sense to me now. I'm curious about dknguyen's question about whether faraday cages should be grounded or not.

Check out the guys who work on live 300,000V pylons, they wear a faraday 'suit' - no change of grounding that! (nor any reason to want to).