Transmission of Video Signals
This is not meant to be a text book on transmission but is intended to remove some of the mystery associated with various methods of transmission. Many approximations and simplifications have been used in writing this guide. This is to make the subject more understandable to those people not familiar with the theories. For general application in the design of CCTV systems it should be more than adequate and at least point the way to the main questions that must be addressed. The manufacturers of transmission equipment will usually be only too keen to help in final design.
This first part deals with the transmission of video signals by cables. Part 2 deals with the transmission of video signals by other methods such as microwave, telephone systems, etc.
Diagram 1 illustrates the many methods of getting a picture from a camera to a monitor. The choice will often be dictated by circumstances on the location of cameras and controls. Often there will be more than one option for types of transmission. In these cases there will possibly be trade offs between quality and security of signal against cost.
The video signal
A field of video is created by the CCD being scanned across and down exactly 312 1/2 times and this reproduced on the monitor. A second scan of 312 1/2 lines is exactly 1/2 a line down and interlaced with the first scan to form a picture with 625 lines. This is known as a 2:1 interlaced picture. The combined 625 line is known as a frame of video and made up from two interlaced fields. The total voltage produced is one volt from the bottom of the sync pulse to the top of the white level, hence one volt peak to peak(p/p). The luminance (brightness) element of the signal is from 0.3 volts to one volt, therefore is 0.7 volts maximum. This is known as a composite video signal because the synchronising and video information are combined into a single signal.
In the case of a colour signal, further information has to be provided. The colour information is superimposed onto the video signal by means of a colour sub-carrier. A short reference signal, known as the chroma burst, is added to the back porch after the horizontal sync pulse to detect the difference in position or phase.
The transmission system must be capable of reproducing this signal accurately at the receiving end with no loss of information.
Note that the imaging device is scanned 625 times but the actual resolution is defined by the number of pixels making up the device.
The video signal from a TV camera has to provide a variety of information at the monitor for a correct TV picture to be displayed. This information can be divided into: Synchronising pulses that tell the monitor when to start a line and a frame; video information that tells the monitor how bright a particular point in the picture should be; chrominance that tells the monitor what colours a particular part of the picture should be (colour cameras only).
The composite video output from the average CCTV camera covers a bandwidth ranging from 5Hz to many MHz. The upper frequency is primarily determined by the resolution of the camera and whether it is monochrome or colour. For every 100 lines of resolution, a bandwidth of 1MHz approximately is required. Therefore, a camera with 600 lines resolution gives out a video signal with a bandwidth of approximately 6MHz. This principle applies to both colour and monochrome cameras. However colour cameras also have to produce a colour signal (chrominance), as well as a monochrome output (luminance). The chrominance signal is modulated on a 4.43MHz carrier wave in the PAL system therefore a colour signal, regardless of definition, has a bandwidth of at least 5MHz.
Requirements To Produce A Good Quality Picture
From the above it will be obvious that to produce a good quality picture on a monitor, the video signal must be applied to the monitor with little or no distortion of any of its elements, i.e. the time relationship of the various signals and amplitude of these signals. However in CCTV systems the camera has to be connected to a monitor by a cable or another means, such as Fibre Optic or Micro Wave link. This interconnection requires special equipment to interface the video signal to the transmission medium. In cable transmission, special amplifiers may be required to compensate for the cable losses that are frequency dependant.
All cables, no matter what their length or quality, produce problems when used for the transmission of video signals, . the main problem being related to the wide bandwidth requirements of a video signal. All cables produce a loss of signal that is dependent primarily on the frequency, the higher the frequency, the higher the loss. This means that as a video signal travels along a cable it loses its high frequency components faster than its low frequency components. The result of this is a loss of the fine detail (definition) in the picture.
The human eye is very tolerant of errors of this type; a significant loss of detail is not usually objectionable unless the loss is very large. This is fortunate, as the losses of the high frequency components are very high on the types of cables usually used in CCTV systems. For instance, using the common coaxial cables URM70 or RG59, 50% of the signal at 5MHz is lost in 200 metres of cable. To compensate for these losses, special amplifiers may be used. These provide the ability to amplify selectively the high frequency components of the video signal to overcome the cable losses.
There are two main types of cable used for transmitting video signals, which are: Unbalanced (coaxial) and balanced (twisted pair). The construction of each is shown in diagrams 2 and 3. An unbalanced signal is one in which the signal level is a voltage referenced to ground. For instance a video signal from the camera is between 0.3 and 1.0 volts above zero (ground level). The shield is the ground level.
A balanced signal is a video signal that has been converted for transmission along a medium other than coaxial cable. Here the signal voltage is the difference between the voltage in each conductor.
External interference is picked up by all types of cable. Rejection of this interference is effected in different ways. Coaxial cable relies on the centre conductor being well screened by the outer copper braid. There are many types of coaxial cable and care should be taken to select one with a 95% braid. In the case of a twisted pair cable, interference is picked up by both conductors in the same direction equally. The video signal is travelling in opposite directions in the two conductors. The interference can then be balanced out by using the correct type of amplifier. This only responds to the signal difference in the two conductors and is known as a differential amplifier.
Unbalanced (Coaxial) Cables
This type of cable is made in many different impedances. In this case impedance is measured between the inner conductor and the outer sheath. 75 Ohm impedance cable is the standard used in CCTV systems. Most video equipment is designed to operate at this impedance. Coaxial cables with an impedance of 75 Ohms are available in many different mechanical formats, including single wire armoured and irradiated PVC sheathed cable for direct burial. The cables available range in performance from relatively poor to excellent. Performance is normally measured in high frequency loss per 100 metres. The lower this loss figure, the less the distortion to the video signal. Therefore, higher quality cables should be used when transmitting the signal over long distances.
Another factor that should be considered carefully when selecting coaxial cables is the quality of the cable screen. This, as its name suggests, provides protection from interference for the centre core, as once interference enters the cable it is almost impossible to remove.
Balanced (Twisted Pair) Cables
In a twisted pair each pair of cables is twisted with a slow twist of about one to two twists per metre. These cables are made in many different impedances, 100 to 150 Ohms being the most common. Balanced cables have been used for many years in the largest cable networks in the world. Where the circumstances demand, these have advantages over coaxial cables of similar size. Twisted pair cables are frequently used where there would be an unacceptable loss due to a long run of coaxial cable.
The main advantages are:
1) The ability to reject unwanted interference.
2) Lower losses at high frequencies per unit length.
3) Smaller size.
4) Availability of multi-pair cables.
5) Lower cost.
The advantages must be considered in relation to the cost of the equipment required for this type of transmission. A launch amplifier to convert the video signal is needed at the camera end and an equalising amplifier to reconstruct the signal at the control end.
It is extremely important that the impedances of the signal source, cable, and load are all equal. Any mismatch in these will produce unpleasant and unacceptable effects in the displayed picture. These effects can include the production of ghost images and ringing on sharp edges, also the loss or increase in a discrete section of the frequency band within the video signal.
The impedance of a cable is primarily determined by its physical construction, the thickness of the conductors and the spacing between them being the most important factors. The materials used as insulators within the cable also affect this characteristic. Although the signal currents are very low, the sizes of the conductors within the cable are very important. The higher frequency components of the video signal travel only in the surface layer of the conductors.
For maximum power transfer, the load, cable and source impedance must be equal. If there is any mismatch some of the signal will not be absorbed by the load. Instead it will be reflected back along the cable to produce what is commonly known as a ghost image.
Mixing Cable And Equipment Types
It is essential that coaxial cables and balanced cables should only be used with the correct type of equipment. Unpredictable results will occur if the incorrect cable type is used. For instance, if the intention is to use a balanced cable, this cannot be connected directly to a coaxial cable or an amplifier designed to drive a coaxial cable. Some form of device is required to be connected between the two cable types so that both cables are correctly matched. This piece of equipment may be an amplifier or video isolation transformer.
Every joint in a cable produces a small change in the impedance at that point. The mechanical layouts of the conductors change where it is joined. This cannot be avoided. However, the changes in impedance should be minimised by using the correct connectors. When in line joints are being made, ensure the mechanical layout of the joint follows the cable layout as closely as possible. The number of joints in a cable should be minimised, as each joint is a potential source of problems and will produce some reflections in the cable.
The Decibel (dB)
Cable and amplifier performance are usually defined as a certain loss or gain of signal expressed in Decibels (dB). The dB is not a unit of measure but is a way of defining a ratio between two signals. The dB was originally developed to simplify the calculation of the performance of telephone networks, where there were many amplifiers and lengths of cable on a network.
The calculations become extremely difficult, and often produce very large figures using ordinary ratios, when many of them have to be multiplied and divided to work out the signal levels of the network. However these calculations become relatively simple if the ratios are converted to the logarithm of the ratio, which can then be just added and subtracted. This therefore, is the reason for using the decibel, which in simple terms is:
10 x log (ratio)
This dB (power dB) is often used to measure power relative to a fixed level. It is not a measure in its own right. If the impedance at which the measurements are made is constant, the dB becomes 20 x log (ratio). This is the dB (voltage dB) which is normally used to define cable loss or amplifier gain in the CCTV industry.
The advantage of using this method becomes obvious when working out the performance of a network containing more than one or two items. Many people who do not use dBs all the time have problems relating them to real ratios. The key figures to remember are:
If the ratio is 2:1, then 20 x log 2= 20 x .310 = 6.021, e.g. 6dB.
If the ratio is 10:1, then 20 x log 10= 20 x 1 =20, e.g. 20 dB.
If the ratio is 20:1, then 20 x log 20= 20 x 1.3=26, e.g. 26 dB.
Similarly a ratio of 100:1 is equal to 40 dB.
Therefore, put in reverse, some common ratios are:
6 dB is a loss or gain of 2:1
20 dB is a loss or gain of 10:1
26 dB is a loss or gain of 20:1
40 dB is a loss or gain of 100:1
Diagram 5 illustrates the relationship between the measure of signal to noise in dB and as a ratio.
Example Of Network Transmission
The following example illustrates a typical network and how to calculate the losses and gains.
To work out the net loss or gain of signal on a network, add the amplifier gains and subtract the cable losses.
1st cable -- loss 12dB, 1st amplifier -- gain 6dB
2nd cable -- loss 20dB, 2nd amplifier -- gain 26dB
3rd cable -- loss 6dB.
The result would be: -12dB + 6dB - 20dB + 26dB - 6dB = -6dB
Reduction Of Signal To Noise Ratio.
When a video signal is amplified the noise, as well as the signal, is increased. If the amplifier were perfect then the resulting signal to noise ratio would remain unchanged. Amplifiers are not perfect and can introduce extra noise into the signal. The amount of noise introduced increases as the amplifier approaches its maximum gain setting. A typical amplifier or repeater operating at maximum gain may reduce the signal to noise ratio by about 3dB. Consequently, it is not advisable to run such equipment at the maximum levels. This is similar to the results of turning the volume up too high on a domestic HI FI. A lot of interference is evident and most units are only operated at up to about half their maximum rating.
In the same way as the net gain or loss in a network can be simply calculated by adding the dB values arithmetically, so can the reduction in signal to noise ratio. In the previous example if the original s/n ratio is 50 dB at the camera then after two amplifiers the s/n ratio could be reduced to 44dB. After four amplifiers this could be reduced to 44 - 12 = 32 dB. At this signal to noise ratio the picture would show a lot of 'snow' and be close to the limit of a usable picture. This then is the limit of the distance that a video signal may be transmitted using this type of transmission. Therefore, besides calculating the losses and gains of the network the reduction in s/n ratio must also be calculated. This example assumes that the worst case is considered. Manufacturers' data or assistance should be sought if equipment is to be used at maximum settings.
Misuse Of The dB
The term dB is very often misused as a measurement, which it is not. This practice is very common. However, the correct way of stating a measurement is +/- YdB's relative to a base level. It is a common, though technically incorrect, practice not to mention the base level, which can lead to the assumption that the dB is a unit of measure.
Examples Of Typical Configuration
Diagram 7 shows some typical configurations for cabled systems.
Overall cable performance is usually defined for its ability to pass high frequency signals. After selecting the correct type of cable with the desired impedance, the next most important factor is the cable transmission loss at frequencies within the video band. Most cable manufacturers provide figures at 5MHz and 10MHz. The 5MHz figure is the most important for CCTV use. The cable losses will be defined as a loss in dB at 5MHz per 100 metres. Care should be taken when dealing with cables of American origin as these are often defined as loss per 100 feet. Generally, the larger the size and the more expensive the cable, the better will be its performance. This holds true for most cables as larger conductors produce the least loss.
If the loss is given for a frequency but not the one required, the conversion is as follows. Assuming the cable is rated at 3.5 dB loss per 100 metres at 10MHz, then the loss at a frequency of 5MHz would be:
Note that before using this conversion the cable specification should be checked to ensure that it will transmit satisfactorily at 5 MHz. Some cables are designed specifically for high frequency transmission only, and will not be suitable for the lower frequencies used in CCTV.
The important factors when selecting a cable for a particular installation are:
1) Establish the type of cable to use, coaxial or twisted pair.
2) Select a range of cables of the correct impedance.
3) Select the correct mechanical format, i.e. normal cable to be laid in ducts or single wire armoured for direct burial etc.
4) Consider the distance the cable is required to run and calculate the length of cable required.
Do not forget to make allowances in this calculation for unseen problems in installing the cable. A minimum of a 10% allowance should always be made. This provides a safety margin to cover inaccurate site drawings, sections of the cable running vertically and other problems likely to be met during installation.
5) When the length of cable has been established, assess the high frequency loss from the cable data.
6) Once the cable loss has been estimated, then the equipment requirement can be established.
The data for twisted pair cables is not always easy to obtain. However, most telephone type cables are highly suitable for video transmission. Even the internal telephone subscriber cable can be used over quite long distances for video, with the correct equipment. (Typical losses at 5MHz are 4dB per 100 metres.) If in doubt about the suitability of a twisted pair cable, the general rules are that suitable cables will be unscreened and will have a very slow twist to the conductors, 1 to 3 twists per metre.
Many twisted pair cables are advertised as "Wide Band Data Cables." These are usually of American origin and are heavily screened. They are designed for use with computers and are generally unsuitable for video use. If a cable is to be used about which there is some doubt, it is worth testing the cable with the equipment to be used before installation. Although this may be considered as a waste of time, it can avoid a costly mistake in the installation.
Tests can be run with the cable on drums as the performance will improve when the cable is taken off the drums and installed. When faced with using existing cables on a site, the only safe way to establish if they are suitable is to run an actual test with the equipment it is intended to use.
The problems that can be encountered when attempting to use existing cables include:
Cables that have absorbed water or moisture.
The cable route is much longer than it appears.
Other cables have been connected in parallel.
If in any doubt, run a transmission test.
Transmission Equipment and Methods
When considering the preceding details regarding cable performance, it is obvious that special equipment is required to transmit video signals over long cables. The type of equipment required is dependent on the length of cable involved and the required performance.
This equipment falls under two headings:
1) Launch Equipment
Launch equipment is designed to precondition the video signal for transmission over the cables.
2) Cable Equalising Equipment
Cable equalising amplifiers are designed to provide variable compensation to make up for the losses after the video signal has been transmitted over the cables.
Selection Of Cable And Equipment
When selecting the cable and equipment for a particular installation the following rules apply:
1) Select the cable to be used, noting the high frequency loss associated with the length of the cable selected.
2) Select the line transmission equipment required to compensate for the cable loss.
3) Sometimes it is possible to save on the installation cost by using a cheaper cable with more powerful equipment.
4) Determine the level of performance required.
5) For colour transmission, it is wise to allow a margin of 6dB extra equalisation in the equipment over the projected cable losses.
6) For high quality monochrome transmission no margin is required other than the 10% for variations in cable length mentioned previously.
7) An acceptable monochrome picture can be obtained with a net loss of 6dB over the transmission link.
Cable = 1000 Metres of URM70 = Loss of 33dB at 5MHz.
Equipment required for full equalisation = Launch Amplifier with +12dB at 5MHz + Cable equalising amplifier with +32dB of equalising at 5MHz.
This combination of equipment provides a total of +44dB at 5MHz against a cable loss of -33dB giving +11dB at 5MHz in hand.
This configuration will provide a first class colour picture. In fact it would work well up to a cable length of 1200 metres.
The normal transmission levels for video signals in the CCTV industry are:
Coaxial Cable:- 1 Volt of composite video, terminated in 75 Ohms, positive going, i.e. Sync tips at 0V and peak white at 1 Volt.
Twisted Pair Cables:- 2.0 Volts balanced, terminated in the characteristic impedance of the cable, normally between 110 and 140 Ohms.
Typical cable losses.
A selection of commonly used cable specifications is given below.
Principles Of Transmission
The object of using special transmission amplifiers is to be able to produce a video frequency response that is a mirror image of the cable loss. The net result is that the video output will be a faithful reproduction of the input and effectively the cable loss disappears completely. The above is a much simplified version of what happens in a correctly installed transmission link.
The example in Diagram 8 shows that the equaliser response is produced by being able to adjust the gain of the amplifier at different frequencies. In this case the amplifier has five sections operating at 1, 2, 3, 4, and 5MHz.
If the higher frequencies of the video signal are sent at an increased level, this will reduce the high frequency noise by reducing the amount of amplification required at the end of the cable. This method of changing the video signal is known as pre-emphasis.
A cable equalising amplifier acts rather like the audio "Graphic Equaliser" with which most people are familiar. It enables the gain of the amplifier to be adjusted independently at different frequencies within the video band. The object of this is to be able to produce a mirror image of the cable response.
Each amplifier requires setting up to match the cable with which it is to be used. Once set, it should never require readjustment unless a drastic change in the installation is made.
Test Equipment Required
Correct cable equalisation cannot be achieved without the use of special test equipment. This enables the various adjustments to be set to optimum. Some people claim to be able to set up this type of equipment "by eye". No matter how experienced a person is, the results obtained by attempting to use this method will be always inferior to those produced with the proper test equipment.
Pulse And Bar Generator
This produces a special wave form that is designed to show problems in a video transmission link. The timing and period of the chroma burst are especially important in the transmission of colour signals, particularly if multiplexing equipment is incorporated in the system.
This is required to observe the wave form from the pulse and bar generator and should have a bandwidth of at least 10MHz.
Object Of Adjusting The Equipment
The object of setting up the video line transmission equipment is to obtain a true replica of the Pulse and Bar wave form after it has been transmitted through the amplifiers and cable. If this is achieved, a satisfactory picture will be produced by the monitor.
Method Of Adjustment
The pulse and bar generator should be connected in place of the camera. The resultant wave form is viewed on the oscilloscope at the output of the amplifier before the monitor. If a launch amplifier is being used, the output level of this should be set first to 1 Volt with no pre-emphasis. The gain of the cable equalising amplifier should then be set to give 1 Volt output.
The equalising controls should then be adjusted in ascending order, i.e. low frequency (LF) lift first to obtain the best equalisation. Each control affects a different portion of the video signal, to obtain the best results. The controls may need adjusting more than once as there is a certain amount of interaction between them.
Once the controls are set to optimum in the equalising amplifier, the high frequency (HF) lift control in the launch amplifier should then be adjusted to give the required pre-emphasis. The HF lift controls in the equalising amplifier should then be able to be set to a lower level. Care must be taken to ensure that the launch amplifier output is not overloaded as this may produce peculiar results.
When a video signal has to be transmitted over extremely long or poor quality cables, it is necessary to use a repeater amplifier within the system. The distance along the cable at which it should be installed can be calculated from the cable loss figures. When using repeater amplifiers, an extra allowance of 3dB should be made for the cable loss. It is better to insert a repeater amplifier in a cable run before the video signal deteriorates too much, than to attempt to equalise a very poor quality signal. There is no actual limit to the length of cable and number of repeater amplifiers that can be used. The problem that occurs is that the signal to noise ratio deteriorates with each amplifier.
The practical limit is approximately 4 repeater amplifiers in cascade with a launch and equalising amplifier at the ends of the cable. This configuration can easily operate over cable lengths of 50 Km or more if the correct type of cable is used. This applies equally to coaxial or balanced cables.
Method Of Adjustment
The method of setting up a system with repeater amplifiers is identical to adjusting a single equalising amplifier. The pulse and bar signals are inserted in the cable at the position of the last repeater amplifier. This enables the final equalising amplifier to be adjusted. When this is completed, the pulse and bar unit is moved up the next section of cable to enable the last repeater to be set up. The procedure is then repeated working along the cable towards the camera position until the launch amplifier is reached. Great care should be taken when setting up a transmission link using repeater amplifiers. This is because once an error has been introduced into the video signal by an incorrectly adjusted amplifier it cannot be corrected by miss-setting another amplifier. Errors are normally additive and a slight mis-setting of several amplifiers will produce unacceptable results.
When installing TV cameras or other equipment on large sites, the potential of the earth connection provided for the equipment can vary by quite large voltages (up to 50 Volts). This can produce high currents in cables connected between different points on the site and will produce interference on the video signal.
Most video equalising amplifiers have differential inputs that can reject a certain amount of interference due to earth potential variations (up to 10 Volts). However, it is good practice, and a safe precaution, to break the earth connection using a video transformer or opto-coupled equalising amplifier on long cables. It is not safe or legal to remove earth connections from equipment and rely on the earth provided by the video cable.
This latter procedure, which is still common practice in the CCTV industry, is in breach of the electrical safety regulations and is extremely dangerous and should on no account be used
Please continue reading to part 2.