Can local voltage drop be developed over ground plane?

[goodphy]

Hello.

The ground plane is sometimes used as current return path. If current is low frequency and amplitude is small then voltage rising due to current flow can be ignorable.

However, what is return current is actually very high in both frequency and amplitude? In our lab, gas discharge results in ~240 A discharge current of single cycle for 400 ns, which corresponding frequency is ~2 MHz.

You can take a look at the attached image. The points A and B are the electrical wire contact points thus I believe current should choose this straight and shortest path to flow from A to B. If this is true, I guess there is severe local voltage drop just between A and B in such a transient time and after then all points on the ground plane becomes equipotent later.

Is my reasoning making sense?

 

[stedwards]

Certainly, there will be a voltage drop. The ground plan is resistive. If no voltage difference is developed, no current will flow.

The path of current from source to sink (A to B) spreads out over the ground plane with the maximum occurring in a direct line.

This is what the the current flow field looks like.

The greater the spacing between lines, the lesser the current density.

 

[goodphy]

Certainly, there will be a voltage drop. The ground plan is resistive. If no voltage difference is developed, no current will flow.

The path of current from source to sink (A to B) spreads out over the ground plane with the maximum occurring in a direct line.

This is what the the current flow field looks like.

The greater the spacing between lines, the lesser the current density.

Oh thanks. this is really great picture! This picture clearly confirms that in transient time voltage drop along straight path from A to B is the most strong. Thanks! I'll refer this for my future electronics enclosure design. Anyway...Could you tell me how you get this image?

 

[stedwards]

Oh thanks. this is really great picture! This picture clearly confirms that in transient time voltage drop along straight path from A to B is the most strong. Thanks! I'll refer this for my future electronics enclosure design. Anyway...Could you tell me how you get this image?

You're welcome. I did a quick search for the picture. If you look a little harder than I, you might find one that also shows the electric potential lines--lines connecting points of equal voltage. I searched on something like "flow from source to sink". Try searching images instead of URLs like I did [Edit: you will using google]

It's quite a universal diagram. It can depict two dimensional fluid flow from a inlet pipe to an outlet pipe. It can represent the electric field lines between two charged particles (in two dimension). The general field of study is called conformal field theory.

[Edit: By the way, I was speaking of DC conditions. A ground plan is capacitively coupled to to everything around it; notably a power plan. The inductance is relatively low.]

 

[tech99]

You're welcome. I did a quick search for the picture. If you look a little harder than I, you might find one that also shows the electric potential lines--lines connecting points of equal voltage. I searched on something like "flow from source to sink". Try searching images instead of URLs like I did [Edit: you will using google]

It's quite a universal diagram. It can depict two dimensional fluid flow from a inlet pipe to an outlet pipe. It can represent the electric field lines between two charged particles (in two dimension). The general field of study is called conformal field theory.

[Edit: By the way, I was speaking of DC conditions. A ground plan is capacitively coupled to to everything around it; notably a power plan. The inductance is relatively low.]

During the AC conditions of the step function, I suppose that the fields spread out from each point A and B like ripples from a stone in a pond. It seems to be like the generation of a surface wave from a grounded antenna. Perhaps when the two sets of ripples meet, they cancel, giving the steady state flow condition.

 

[stedwards]

I don't know how this works. You have a 400 ns. pulse load or it's probably shaped more like a spike with roll-off. At a wavelength of a bit less than 2.5 feet, I don't think you will get a lot of high frequency garbage reflecting all over the place. There are some sharp guys haunting the Electrical Engineering Forum in these Physics Forums that could do better than I.

 

[tech99]

I don't know how this works. You have a 400 ns. pulse load or it's probably shaped more like a spike with roll-off. At a wavelength of a bit less than 2.5 feet, I don't think you will get a lot of high frequency garbage reflecting all over the place. There are some sharp guys haunting the Electrical Engineering Forum in these Physics Forums that could do better than I.

I agree with you on that; the rise time is probably slow enough to use the steady state picture.

 

[CWatters]

One way to reduce or eliminate problem due to high current pulses is to add capacitors between the power and ground planes physically close to the switching device. This can work well but there are limits, particularly as large capacitors are far from being "ideal" capacitors.

 

[goodphy]

I agree with you on that; the rise time is probably slow enough to use the steady state picture.

Thanks for replying.

Thus..if wavelength of the signal is very long compared to the ground plane, we can say it is steady-state picture and all points on the ground plane are to be the same at a time? means voltage of all plane is varying in time without spatial variance like quasi-steady state solution?

 

[goodphy]

One way to reduce or eliminate problem due to high current pulses is to add capacitors between the power and ground planes physically close to the switching device. This can work well but there are limits, particularly as large capacitors are far from being "ideal" capacitors.

Thanks. You mean capacitor is to be added between wire of A (or B) and ground plane? capacitor is usually passes for high frequency like high frequency filter thus...what feature I can expect to this job? And do you think adding capacitance lower down pulse rising time that is actually not what I intend?

 

[CWatters]

capacitor is usually passes for high frequency like high frequency filter

Capacitors also try to maintain the voltage drop across them, like a battery. The capacitor would be connected between the power and ground planes at point A. When used like this they are usually called "smoothing capacitors", "decoupling capacitors" or "Bypass Capacitors"

Suppose we define point B as being "true ground". Without a capacitor when the return current flows the voltage at point A will rise wrt B (due to impedance in path AB). Likewise on the power plane the voltage at point A will fall wrt point B. A capacitor connected between power and ground at point A would tend to maintain a constant voltage between power and ground at point A by delivering current to the load.

Perhaps see also para 2.4..

http://www.ti.com/lit/an/scaa082/scaa082.pdf

There are other tricks you can use to eliminate problems. If the voltage drop across the ground plane is causing problems for other circuits then it might be better to partition the ground plane into sections.