Quality Solutions for the Video Professional
Ground Loops:What They Really Are
Ground loops are a mystery to many people. Even college-trained electronic engineers may not know what ground loops actually are. Engineers have either concentrated on power distribution (for the electric company) or on equipment that happens to plug in to the power distribution system. Not much thought has been given to power distribution and equipment as a single entity where ground loops arise.
Myths. A couple of myths need to be dispelled before ground loops are examined. Here are the facts:
First, there is no absolute ground. There is a certain amount of resistance to electrical current between all grounding points. This resistance can change with humidity, temperature, connected equipment and many other variables. No matter how small, the resistance can always allow an electrical voltage to exist across it. The ground wires between wall sockets and power company transformers are not perfect conductors and neither is the shield of your coaxial video cable. If they were, ground loops would not be a problem.
Second, percent coverage, a coaxial cable specification, refers to how well the shield conductor protects the center conductor from light, not electrical interference.
An Ideal Electrical World. A simplified drawing of an electrical distribution system is shown in figure 1. The path that the 12,000 Volt AC power should follow is shown by the arrows, flowing from the power line through the transformer primary coils and back to the power line. The transformers transform the 12,000 Volts AC to the 115 Volts AC to which we're all accustomed. Everything works well when the power follows the path it's supposed to.
Figure 1. The power company's 12,000 Volt distribution is supposed to follow the arrowed path. The ground symbols indicate where a copper stake has been pounded into the earth. In many cases the earth is a poor electrical ground because of high resistance.
60 Hz Hum-Bars. Figure 2 shows an alternative path for the power flowing through the right-most transformer. Unfortunately, the alternative path is right through your coaxial cable shield. Because the shield conductor is carrying an extraneous current that the center conductor is not, there is a voltage difference between the two conductors at the monitor input that will be displayed as interference. The percentage of the power company's current that you're carrying for them is determined by the ratio of their ground line's resistance to your cable's shield resistance. The power company takes solace in the fact that if their ground line breaks and the earth ground is poor, your video cable and equipment will keep the neighborhood from suffering a black-out.
If both the video source and the monitor are connected to the same power company transformer, this type of interference won't occur unless something is defective. The defect could be excessive resistance in the power cord ground line to either the source or the monitor. Such resistance is usually caused by loose or oxidized connections in the equipment or the building wiring.
The defect could also be within some other electrical device that is producing excessive ground current. The device could be leaking current to its safety ground (that's why it's there) or the neutral and ground wires could be reversed. Such a device can work normally, but it can cause interference in video systems.
Figure 2. By running a coaxial cable a fair distance, you risk setting up an alternate path for the power company's use, right through your system! Hum and hum-bars, if not worse, result.
RF Interference. Sixty Hertz hum-bars are not the only symptom of ground loops. "Herring Bone" or RF interference is also a common problem in long cables carrying composite video. Herring bone interference is caused by a ground loop (that includes your coax shield) acting as an AM radio antenna (see figure 3). Any large loop of wire makes a good AM antenna. These antennas are especially adept at picking up AM broadcasts if most of the loop is vertical. A video installation in a tall building with cables running between floors makes a good antenna. As figure 3 shows, this loop does not involve the power company transformers. Therefore, whether one or more transformers is involved has no effect on the likelihood of encountering this or any other type of interference following this particular path.
A very common misconception is that the center conductor of the cable picks up the RF interference and that a cable with a high "percent coverage" of the center conductor by the shield will alleviate the problem. It's not true. The shield actually picks up the interference and introduces it to the video system. Any shield whose coverage gaps are much smaller than the wavelength of the potential interference provides the center conductor with adequate shielding. The wavelength of AM radio is about 180 meters at the shortest. Percent coverage in even the cheapest coaxial cable is adequate to protect the center conductor against this RF interference. The best shield is one that offers the lowest resistance per unit of length (Ohms/kM) in order to minimize the voltage potential that can develop across it.
RF induced currents in the shield are the problem. As with hum-bars, the shield conductor is carrying an extraneous current that the center conductor is not. Hence, there is an undesirable voltage difference between the two conductors at the monitor input and it will be displayed as interference.
Light Dimmers and other Obnoxious Devices. Interference caused by light dimmers and similar electrical equipment is also a symptom of ground loops. Light dimmers operate not by reducing voltage, but by actually turning it on and off fast enough that the light bulb can't keep up and operates at some point of compromised brightness. Each time the power is turned on or off, a large, fast-changing, 'spike' of current is created in the building's AC wiring.
Electricians, by running 115 VAC cables with parallel conductors (commonly called romex) over long distances, create capacitors. The power and ground conductors in the AC cable are capacitively coupled to each other. High frequency AC current flows through capacitive coupling, the higher the current's frequency, the better it flows.
The current spike, because it is fast changing, has a high frequency component that couples from the AC power conductors to the ground conductor. The ground conductor is part of a ground loop that also includes your coaxial shield conductor. The path of the capacitively induced extraneous current is the same as that shown for broadcast interference in figure 3. Again, the video cable shield conductor is carrying an extraneous current that the center conductor is not. Hence, there is an undesirable voltage difference between the two conductors at the monitor input and it will be displayed as interference.
Figure 3. A ground loop serving as a loop antenna, introducing RF interference to your video.
Cross-Talk. Ground loops can cause one signal to interfere with another. Figure 4 shows two long coaxial cables whose shield grounds are connected together at each end (through one path or another). The signal shown traveling through the center conductor to the "Input B BNC" should ideally return through the corresponding shield conductor, but there's an alternative path through the other shield conductor. That causes the shield conductor connected to the "Input A BNC" to carry an extraneous current that the center conductor does not. Therefore, there is an undesirable voltage difference between the two conductors at the monitor input and it will be displayed as cross-talk interference.
The video and power industries have each designed their systems and equipment independently. As a result, there's a degree of incompatibility. Ground loop interference is a consequence.
Figure 4. Two long coaxial cable shields each provide a return path for the signal in either center conductor. In this case the signal in Cable B is causing an extraneous current in cable A's shield conductor. Cross talk is the result.