# How does a ground loop and a floating ground work?

1. Nov 18, 2016

### fog37

Hello Forum,
In DC circuits, I understand that a ground is a point or an entire conductor where the electric potential is set to be V=0 and the potential V at any other point in the circuit is referenced to that.

How does the the concept of ground loops work? Why are they not desirable?

Thanks,
Fog37

2. Nov 18, 2016

### analogdesign

It's pretty simple actually.

A "ground loop" is when you have two (or more) grounds in your system that are actually at different voltages so current can flow between them. In other words, you have two "grounds" but they aren't both at V=0. For example, imagine you have a preamp and a power amp each connected to ground, but there is a significant resistance between them (inadequate cabling, or whatever). Then you can have a hum because the preamp and power amp can communicate through their ground connections. This is not something you want.

A ground is "floating" when the common reference voltage is not connected to "earth" with a very low impedance. Sometimes you want to float a piece of equipment because you want it to be able to interface with something at very high voltage (which would typically be connected to earth with a very low impedance) without actually needing to withstand the full voltage. On the other hand, it can be really dangerous with high-voltage equipment because if you short the "floating" ground and the "earth" ground together through your body you will conduct all the current of your floating equipment (unless it has been specifically designed for ground faults). You can see how this is related to a ground loop but in this case the impedance between the grounds is your body.

3. Nov 19, 2016

### fog37

Thanks analogdesign. Let me see if I get things right:

Consider a 9V battery (constant voltage source), ideal copper wires (zero resistance) and a load (small lightbulb with internal resistance R). That closed circuit is a DC circuit. The positive battery terminal is 9V above the negative terminal potential. We don't know what the absolute potentials and we don't care. We only care about the electric potential difference (aka voltage).
The wire connected to the positive terminal will be equipotential at every point and have the same potential as the positive terminal. The wire connected to the negative terminal is equipotential to the same potential as the negative battery terminal. That implies that the potential difference between two different points on the same wire is zero. There is an electric potential difference across the load. We can choose to assign the value 0 to the potential of the negative battery terminal. That mean the potential of the positive terminal becomes V = 9V. All good so far, right? Real wires have finite small resistance so on the same wire we would measure a small potential difference between any two points (the larger the difference the farther the points are). That means that the positive battery terminal and the light bulb terminal are not equipotential anymore (the bulb terminal is slightly less than 9V).

Do the negative battery terminal and the wire connected to it represent the the floating ground? The electric current will return to the battery through that ground conductor.

We could connect that wire conductor to planet Earth and convert the floating ground to an earth ground, correct?

I am hope I am on the right track. I will write more later...

Thanks!
Fog37

4. Nov 19, 2016

### CraigHB

Yes it would be floating with respect to Earth ground, but most DC devices (battery powered) do not connect to Earth ground. Typically a connection is considered floating when there is no current path to ground with respect to zero reference within the circuit itself. You have to qualify the case when a connection is floating with respect to Earth ground. There is some ambiguity when using the term ground. It can mean Earth ground or the zero reference within a circuit. They may or may not be the same thing.

5. Nov 19, 2016

### fog37

Thanks CraigHB.

So, in my example, current goes through the ground wire. If we connected both ends of the ground wire to planet earth, would the current exiting the load go through the (return) wire, or pass through planet earth to get back to the negative terminal of the battery. I guess we can model earth another wire with low resistance. So earth is in parallel with the original ground wire and current should pass both through the earth and through the return ground wire.

6. Nov 19, 2016

### CraigHB

If you were to use sufficiently long enough wires, then yes you could have a ground loop where the potential is considerably different at two different ground connections. The current path through Earth ground can be modeled in part by a series resistor. Typically the wire connecting the two points would have much lower resistance than the Earth. Current flow would mainly occur in the wire with only a negligible amount travelling through Earth ground. Typically ground loops are not a big consideration for battery powered devices unless ground points are separated by some amount of resistance, like a very long wire. It can be a major consideration for utility power systems where human safety is involved. For example the plumber that got electrocuted upon disconnecting the metal water main pipe for a house.

7. Nov 19, 2016

### analogdesign

Hi fog37,

You are on the right track. In your last question, if you connected both ends of the ground wire to planet earth, the majority of the return current would be going through the earth to get back to the negative terminal of the battery. If you know the equivalent resistance of both the wire and the earth (tough, I know) you could use the current divider rule to see how much current is going through each path.

One note. While CraigHB is entirely correct in his definition of floating, I should say that in my experience when you here someone say "floating ground" they aren't talking about battery-powered equipment like CraigHB is. Usually I hear the term referring to interfacing between voltage domains. For example, if you have a sensor diode that needs a 200 volt bias but have to read it out using a 5 V op amp, what do you do? One thing you can do is connect the 200 V bias to the 5V of the op amp, and then connect the ground terminal of the op amp to 195 V, so the potential different across the op amp is 5 V. If you do this kind of thing, you have to be very careful not to short the grounds of the diode and the op amp because if you do you will have 200 V across the op amp and you will blow it out.

8. Nov 19, 2016

### fog37

A copper wire resistance is based on the wire length L, copper resistivity, and wire's cross-sectional area.

Why are you so certain that the majority of the electric current will choose the path through earth instead of through the copper wire?

9. Nov 19, 2016

### CraigHB

If a wire is long enough and the ground posts in the Earth are properly installed, it is possible the return path could have lower resistance through Earth ground than through the wire itself. However, that's typically not the case. Most of the time a wire connecting two ground points has considerably lower resistance than a path through the Earth itself. You do have to differentiate applications. Obviously dealing with this sort of thing for low voltage electronic circuits comprises an entirely different set of considerations compared to high voltage or high power applications like those used for industrial power systems.

10. Nov 19, 2016

### fog37

Ok, my understanding is that in a simple DC circuit like the one I described (9V battery, copper wires, load), the current will indeed return to the negative battery terminal through one of the copper wires (not the earth).

But if something wrong happens to the return ground copper wire, the alternative earth path can provide the return path for the current instead of passing through the human body which may be touching (i.e. be connected to the ground). The safe alternative path is available only if the appliance is earth grounded and if the negative battery terminal is also grounded. If the appliance is not earth grounded, the current will choose to go through the human body, then through earth to reach the negative battery terminal....

If we don't connect anything to ground (negative battery terminal or the appliance), we should still have no problem even if the return copper wire is broken and we touch the appliance: the human body is somewhat connected to ground but the current will not run through the ground/earth to go back to the battery because the battery itself is not connected to earth...

11. Nov 19, 2016

### rbelli1

There is no such thing as absolute potential. Potential can only be measured between two points.

Current will flow (assuming the person is indeed connected to earth in some way) to charge the capacitor formed by the negative battery node and earth. Now you have -9V on the battery negative terminal relative to earth. For something like a 9V battery and any device able to be powered by it you will have negligible current in this case. For larger installations you will have a much larger capacitance to earth. This can still cause dangerous currents even with no resistive connections to earth. Especially with AC.

BoB

12. Nov 24, 2016

### fog37

Thanks a lot to everyone.
• A DC voltage source like a 9V battery separates equal amount of positive and negative charges at the terminals. For instance, if we connected it to a capacitor, the capacitor plates would collect equal and opposite amounts of electric charge. That would be true even if we connect two wires (instead of plates) to the terminals. What if we connected to the battery terminals plates of different size, one larger than the other? I think the charge on both plates would still be equal in amount.
• If we connect the negative terminal of the 9V battery to planet Earth, the negative charge would spread of this large sphere (Earth). The electric potential of Earth, whatever it may be, will not change much. The earth has a huge capacitance.
• When the voltage source is a time-varying voltage source, there is no positive and negative terminal. The instantaneous potential difference between the terminal changes in time. If we grounded one of terminals (either one), the electric potential on other ungrounded terminal would vary relative to the "stable" electric potential of the grounded terminal. the potential of the grounded terminal does not change much (or at all) because the charge we move on each will not affect its potential in a significant way.
• The terms "signal conductor" and "ground conductor" don't seem to make much sense in AC circuits where current oscillates back and forth. In PCB there are many different conductors and traces and the electric current moves back and forth (AC). Is that correct?
Thanks

13. Nov 30, 2016

### jim hardy

Quite so.
My old physics course described 'absolute potential' as the work to bring a charge from infinity to wherever you are measuring.
Since it's impossible to go out to infinity and make that measurement, the impractical concept of 'absolute potential' becomes just a thought tool that's handy for planting the very practical concept of 'potential difference' into students' thinking process .
Doing so makes them less likely to abuse the terms.

14. Nov 30, 2016

### fog37

Thanks Jim.

Do you know distinguish between the forward and return conductor when dealing with AC circuits?

15. Nov 30, 2016

### jim hardy

"forward" conductor is a new term to me.

Return has specific meaning in ac power circuits,
in US house wiring we have two "return " paths and they are both at a low potential with respect to earth . They're called "Neutral" and "Earth" .
"Neutral" is a white wire connected to the transformer winding that powers the house. It is the normal path for the load current back to the transformer .
Neutral is connected to Ground at a single point in the building, the location where power enters the building,

you might enjoy this thread from a week or two ago. We get a lot of questions similar to this.

https://www.physicsforums.com/threa...t-live-and-neutral-wires.892309/#post-5613568

old jim

16. Dec 5, 2016

Hi Jim,