8 Signal transmission

 8.2 Pneumatic transmission

In recent years, pneumatic transmission tends to have been replaced by other alternatives in most new implementations of instrumentation systems, although many examples can still be found in operation in the process industries. Pneumatic transmission consists of transmitting analogue signals as a varying pneumatic pressure level that is usually in the range of 3–15 p.s.i. (Imperial units are still commonly used in process industries, though the equivalent range in SI units is 207–1034 mbar, which is often rounded to 200–1000 mbar in metric systems). A few systems also use alternative ranges of 3–27 p.s.i. or 6–48 p.s.i. Frequently, the initial signal is in the form of a varying voltage level that is converted into a corresponding pneumatic pressure.

However, in some examples of pneumatic transmission, the signal is in varying current form to start with, and a current to pressure converter is used to convert the 4–20 mA current signals into pneumatic signals prior to transmission. Pneumatic transmission has the advantage of being intrinsically safe, and provides similar levels of noise immunity to current loop transmission. However, one disadvantage of using air as the transmission medium is that transmission speed is much less than electrical or optical transmission. A further potential source of error would arise if there were a pressure gradient along the transmission tube. This would introduce a measurement error because air pressure changes with temperature.

Pneumatic transmission is found particularly in pneumatic control systems where sensors or actuators or both are pneumatic. Typical pneumatic sensors are the pressure thermometer (see Chapter 14) and the motion-sensing nozzle-flapper (see Chapter 19), and a typical actuator is a pneumatic cylinder that converts pressure into linear motion. A pneumatic amplifier is often used to amplify the pneumatic signal to a suitable level for transmission.

 

8.3 Fibre-optic transmission

Light has a number of advantages over electricity as a medium for transmitting information. For example, it is intrinsically safe, and noise corruption of signals by neighbouring electromagnetic fields is almost eliminated. The most common form of optical transmission consists of transmitting light along a fibre-optic cable, although wireless transmission also exists as described in section 8.4.

Apart from noise reduction, optical signal attenuation along a fibre-optic link is much less than electric signal attenuation along an equivalent length of metal conductor. However, there is an associated cost penalty because of the higher cost of a fibre-optic system compared with the cost of metal conductors. In short fibre-optic links, cost is dominated by the terminating transducers that are needed to transform electrical signals into optical ones and vice versa. However, as the length of the link increases, the cost of the fibre-optic cable itself becomes more significant.

Fibre-optic cables are used for signal transmission in three distinct ways. Firstly, relatively short fibre-optic cables are used as part of various instruments to transmit light from conventional sensors to a more convenient location for processing, often in situations where space is very short at the point of measurement. Secondly, longer fibreoptic cables are used to connect remote instruments to controllers in instrumentation networks. Thirdly, even longer links are used for data transmission systems in telephone and computer networks. These three application classes have different requirements and tend to use different types of fibre-optic cable.

Signals are normally transmitted along a fibre-optic cable in digital format, although analogue transmission is sometimes used. If there is a requirement to transmit more than one signal, it is more economical to multiplex the signals onto a single cable rather than transmit the signals separately on multiple cables. Multiplexing involves switching the analogue signals in turn, in a synchronized sequential manner, into an analogue-to-digital converter that outputs onto the transmission line. At the other end of the transmission line, a digital-to-analogue converter transforms the digital signal back into analogue form and it is then switched in turn onto separate analogue signal lines.

 

8.3.1 Principles of fibre optics

The central part of a fibre optic system is a light transmitting cable containing at least one, but more often a bundle, of glass or plastic fibres. This is terminated at each end by a transducer, as shown in Figure 8.5. At the input end, the transducer converts the signal from the electrical form in which most signals originate into light. At the output end, the transducer converts the transmitted light back into an electrical form suitable for use by data recording, manipulation and display systems. These two transducers are often known as the transmitter and receiver respectively.

Fibre-optic cable consists of an inner cylindrical core surrounded by an outer cylindrical cladding, as shown in Figure 8.6. The refractive index of the inner material is greater than that of the outer material, and the relationship between the two refractive indices affects the transmission characteristics of light along the cable. The amount of attenuation of light as it is travels along the cable varies with the wavelength of the light transmitted. This characteristic is very non-linear and a graph of attenuation against wavelength shows a number of peaks and troughs. The position of these peaks and troughs varies according to the material used for the fibres. It should be noted that fibre manufacturers rarely mention these non-linear attenuation characteristics and quote the value of attenuation that occurs at the most favourable wavelength.

Two forms of cable exist, known as monomode and multimode. Monomode cables have a small diameter core, typically 6 µm, whereas multimode cables have a much larger core, typically between 50 µm and 200 µm in diameter. Both glass and plastic in different combinations are used in various forms of cable. One option is to use different types of glass fibre for both the core and the cladding. A second, and cheaper, option is to have a glass fibre core and a plastic cladding. This has the additional advantage of being less brittle than the all-glass version. Finally, all-plastic cables also exist, where two types of plastic fibre with different refractive indices are used. This is the cheapest form of all but it has the disadvantage of having high attenuation characteristics, making it unsuitable for transmission of light over medium to large distances.

Protection is normally given to the cable by enclosing it in the same types of insulating and armouring materials that are used for copper cables. This protects the cable against various hostile operating environments and also against mechanical damage. When suitably protected, fibre-optic cables can even withstand being engulfed in flames.

A fibre-optic transmitter usually consists of a light-emitting diode (LED). This converts an electrical signal into light and transmits it into the cable. The LED is particular suitable for this task as it has an approximately linear relationship between the input current and the light output. The type of LED chosen must closely match the attenuation characteristics of the light path through the cable and the spectral response of the receiving transducer. An important characteristic of the transmitter is the proportion of its power that is coupled into the fibre-optic cable: this is more important than its absolute output power. This proportion is maximized by making purpose-designed LED transmitters that have a spherical lens incorporated into the chip during manufacture. This produces an approximately parallel beam of light into the cable with a typical diameter of 400 µm.

The proportion of light entering the fibre-optic cable is also governed by the quality of the end face of the cable and the way it is bonded to the transmitter. A good end face can be produced by either polishing or cleaving. Polishing involves grinding the fibre end down with progressively finer polishing compounds until a surface of the required quality is obtained. Attachment to the transmitter is then normally achieved by gluing. This is a time-consuming process but uses cheap materials. Cleaving makes use of special kits that nick the fibre, break it very cleanly by applying mechanical force and then attach it to the transmitter by crimping. This is a much faster method but cleaving kits are quite expensive. Both methods produce good results.

The proportion of light transmitted into the cable is also dependent on the proper alignment of the transmitter with the centre of the cable. The effect of misalignment depends on the relative diameters of the cable. Figure 8.7 shows the effect on the proportion of power transmitted into the cable for the cases of (a) cable diameter > beam diameter, (b) cable diameter D beam diameter and (c) cable diameter < beam diameter. This shows that some degree of misalignment can be tolerated except where the beam and cable diameters are equal. The cost of producing exact alignment of the transmitter and cable is very high, as it requires the LED to be exactly aligned in its housing, the fibre to be exactly aligned in its connector and the housing to be exactly aligned with the connector. Therefore, great cost savings can be achieved wherever some misalignment can be tolerated in the specification for the cable.

The fibre-optic receiver is the device that converts the optical signal back into electrical form. It is usually either a PIN diode or phototransistor. Phototransistors have good sensitivity but only have a low bandwidth. On the other hand, PIN diodes have a much higher bandwidth but a lower sensitivity. If both high bandwidth and high sensitivity are required, then special avalanche photodiodes are used, but at a severe cost penalty. The same considerations about losses at the interface between the cable and receiver apply as for the transmitter, and both polishing and cleaving are used to prepare the fibre ends.

The output voltages from the receiver are very small and amplification is always necessary. The system is very prone to noise corruption at this point. However, the

development of receivers that incorporate an amplifier are finding great success in reducing the scale of this noise problem.