Out of the loop

It is an unfortunate fact that the voltage potential between any two distant points in the earth is more than likely going to be different, and when we connect our low resistance copper bridges across these points (formed by the video signal cable screen), currents will inevitably flow. The problem is illustrated in Fig 1.

The effects of these currents are familiar to many CCTV engineers; dark shadows rolling up or down the screen, intermittent picture jumping or rolling, and sometimes horizontal picture pulling. On the other hand, I do come across other fault conditions, particularly EMI/RFI, being blamed on ground loops when in fact the symptoms themselves indicate that this could not possibly be the cause.

I have also encountered some interesting methods of eliminating ground loops or symptoms allegedly appertaining to them. In this article we will look at the problem, the true symptoms, the cures, and some interesting effects produced by ground loop correction equipment.

Ground loops can be present in any electrical circuit (my introduction to them came many years ago when I was involved in the sound industry), but in CCTV they are predominately associated with co-axial cable (unbalanced) systems.

Although it is possible for them to occur in twisted pair (balanced) systems, the impedance matching equipment required at each end of the cable usually serves as a means of isolation. Of course, they will never occur in a fibre-optic link because there is no form of electrical conductor.

The unwanted effects on the pictures occur because the 50Hz current passing through the screen is able to superimpose itself onto the video signal passing through the core. The reason why this occurs is illustrated in Fig 2 which looks at the a.c. equivalent circuit of a length of co-axial cable.

From this circuit you will note that there is a considerable amount of inductive and capacitive coupling between the screen and the core, resulting in the effective addition of the two signal voltages as shown.

A study of the output voltage reveals the reason for the dark shadow effect on the picture; the black level is being shifted by the 50Hz component.

The sync pulse level is also affected, and this is the reason for the vertical jitter and horizontal pulling effects. VCRs can also be affected because, in the record mode, the video head drum servo is referenced to the vertical sync signal.

The presence of ground loops is not difficult to prove. First of all, once you are familiar with the symptoms, a quick glance at the monitor screen will tell you all you need to know.

However, should there be a need to confirm the presence of a ground potential on a cable, disconnecting it from the equipment at one end and measuring between that end and a good earth point (usually the socket from which the connector has been removed) using a multimeter set to the a.c. voltage range should tell you exactly how many volts there are present. Just a few volts is sufficient to present a problem.

Although not an approved test method (!), sometimes the problem is confirmed before you get to apply the multimeter test leads by the electric shock you get as you remove the connector. Because of the high impedances involved this is usually nothing more than an annoying tingle but, joking aside, always be aware that if there is a loss of earthing at the other end of the cable and other electrical fault conditions also exist, then it is quite possible to receive a lethal electrical shock as the connector is removed.

If you have any reason to suspect a problem, take the necessary precautions and perform the appropriate tests. Looking at the problem, the solution is obvious; we need to find a safe method of breaking the circuit formed by the cable screen, without affecting the screen's ability to remove to earth any unwanted EMI/RFI.

One method would be to simply remove the earth at one end of the installation. This would definitely eliminate the earth loop problem, but in the previous paragraph we did specify that the method had to be safe! Taking the installation in Fig 1, if the electrical earth were removed at the camera end, the only earth to that part of the installation would be the one provided by the monitor plug connector via the co-axial cable screen.

Should the co-axial cable then be disconnected from the system for any reason, the camera installation would be left without any means of earthing, presenting a major safety hazard. Needless to say, removal of any electrical earthing is not an option!

The correct solution is to couple the video signal using either an isolation transformer or an opto-coupler (see Fig 3). With this method the integrity of the electrical earth is maintained at both ends of the installation, the cable screen will still be able to function to remove any interference, but it will not be possible for ac currents to pass through the screen.

From a technical point of view, it does not matter which end of the cable you install the isolation equipment. However from a practical point of view it is usually easiest to locate it at the control room end. This is especially true where optical isolation is employed because this equipment requires a 230v mains supply.

The important point to note is that you only require one unit. If the inclusion of an isolator on a cable does not remove the interference from that cable, then either you have a faulty isolation device (which is almost unheard of) or the fault is not being caused by a ground loop.

I have come across situations where an engineer has installed a second device at the other end of the cable, 'for good measure'. This is not only unnecessary because (as you can see from Fig 3) a single device is sufficient to break the earth loop, but it actually introduces other problems into the system.

First of all, remember that the function of the co-axial cable screen is to remove to earth any RFI and EMI signals that are induced into the cable. If you fit an isolator at each end of the cable you have in effect removed the earth connection from the screen (Fig 4).

Thus, high frequency interference, which usually takes the form of radio waves, will form standing waves in the screen and will 'bounce' back and forth along the screen until their energy is dissipated into the core via the coupling shown in Fig 2. In other words, you have created a situation where the only place for any interference to go is into the core (and onto your pictures).

Secondly, inductive ground loop correctors introduce a degree of signal attenuation (typically 3-6dB), so the last thing you should want to do is use them unnecessarily.

RFI/EMI produce significantly different fault symptoms than do ground loops. Being of a much higher frequency (usually in the order of several MHz), the on-screen effect is usually strong patterning flashing over the picture.

Where the effects of high frequency interference have been incorrectly interpreted as being caused by a ground loop, the inclusion of a transformer based correction unit can sometimes reduce the effect on the screen. This can lead to some confusion as to why the unit did not completely cure the problem.

The reason for the reduction in the fault effect is because of the filtering nature produced by the inductive/capacitive effects of the isolation transformer. However the device will never cure the trouble, and ground loop correctors should never be employed in the role of an rf filter.

Where RFI/EMI is the problem, there are other methods of dealing with this, but perhaps that is the subject of another article.

Correction
In Joe Cieszynski's article on oscilloscopes in the March edition of CCTV Solutions two symbols were mis-translated in the printing process.

The Ohms symbol W was printed as the letter "W" (on page 40, col 2 line 14, page 41, col 2 lines 17 & 22, page 42, lines 15, 18 & 45 and page 44, col 1, line 13).

The mu symbol for micro seconds m was printed as the letter "m" (Page 40, col 2, line 12 and page 41, col 1, line 24)

Our apologies to readers and Joe. We reckon that most readers will have picked up on the Ohms error which was more obvious but changing micro seconds to milliseconds was another matter! Needless to say we've sent the person responsible ohm …Ed.