LEARIU-BUILD- PROGRAM to a powetful sY s ' ern The Junior Computer book is for anyone wishing to become familiar with microcomputers, this book gives the opportunity to build and program a personal computer at a very reasonable cost. The Indian reprint comes to you from elekt©? Send full payment by M.O./I.P.OJD.D. No Cheque Please Packing & Postage free to: eIeI4.17 ELECTRONICS NEWS • ELECTRONICS N will make energy shortages a thing of the Weather can be modified, hurricanes de- fused, cloud covers generated over scorched areas and rain brought to de- Geneticists will be reshaping forms of life amazingly, yielding tastier foods and eradicating some of the afflicting dis- American business will spend hundreds of billions of dollars on automation that will help them stay in competition. Robots will aid in performing everything from shipbuilding to heart surgery. Au- tomation and other changes may elimi- nate about 50 million US factory, office and retail jobs in two decades. More jobs will be opening up in services and elec- tronics. The birthrate in the world now is 150 a minute and the world population has crossed five billion. By the turn of the century, it will cross six billion and could reach 10 billion 30 years later. What will a typical retailing operation be like 10 years from now? A manufacturer of say, video recorders will buy advertis- ing time on cable television in a different city every day. Families in the city will phone in tollfree orders, paying by credit card. Orders will go straight to the fac- tory which will run round the clock. As video recorders roll off the assembly line they will be boxed, automatically label- led and loaded on the delivery trucks for delivery in the target city the next day. Communications will undergo tremend- ous changes. Computers and other de- vices will be driven by and respond in voices. Marketing surveys indicate 60 per cent of US homes will have comput- ers by 1995 as against 18 per cent now. Fastest growth in services is expected to be in the information sector. Medical services will climb up. There will be a surplus of physicians in 1990s. Doctors will pay more attention to pa- tients’ time. But, they won’t go back to the home visit, abandoned decades ago. Instead families will receive much medi- cal advice through home communication centres. The expert system in which spe- cialty knowledge is packed into a compu- ter programme, will find many applica- tions in medicine. Home computers will have phone and video connections with medical expert systems as well as with live physicians. The concept of doing office work from home may not catch on as the cultural value of the office, the opportunity to meet and work with other people in a di- verse environment, will remain strong. Thanks to automation, a 20-hour week may become common by the turn of the century. People will be working much more from cars fitted with computers that recognise speech and can talk back! Whether these predictions come true or not, the British author John Galsworthy’s words justify such an exercise “If you do not think about the future, you cannot Assam Electronics The government of Assam is giving spe- cial attention to the development of elec- tronics industry in the state by extending increased rate of different subsidies. Under this policy, electronics industry gets power subsidy at 60 percent, infras- tructure subsidy of 35 percent and 75 percent subsidy on the cost of drawing power lines. Mr Digen Bora, Assam's industries minister, stated that this subsidy was far higher than the subsidy offered to other industries. He was inaugurating the first top of the line colour television, Amtron Connoisseur, produced by the Assam Electronics Development Corporation. According to Mr D. Kakoti managing di- rector of Amtron, the company was en- gaged in the production of telecomunica- tion equipment too besides consumer electronics Amtron has been given the Rs. 1.50 crore EPABX project of 20,000 lines per year and another Rs.l crore rural automatic exchange project of 10,000 lines per year. Amtron has signed an agreement with ITI, Bangalore, to manufacture 300,000 push button telephones per year, under ITI brand name. Under this Rs.10 crore project, two plants will be set up at Guwahati and Silchar. Amtron will also set up in cooperation with the govern- ment of India a Rs. 2 crore electronic test and development centre near Guwahati. Regional Offices The department of electronics will set up its offices in different regions of the country soon, especially at Calcutta, Bombay, Bangalore and Madras. Ac- cording to Mr K.P.P. Nambiar, secret- ary, the department would finally estab- lish cells in all state capitals. Mr Nambiar also announced at Calcutta recently that the government was in the process of establishing “radio shacks” which would be a chain of retail outlets for electronic components, thus ensuring regular supplies to small scale industries. The Electronics Trade and Technology Development Corporation (ET & T) has been entrusted with the job of establish- ing these outlets. ET & T was also keen on setting up the - outlets in collaboration with the state electronic corporations. This has already been done in Orissa and Maharashtra. The government would also permit the private sector to set up such shops, ac- cording to Mr Nambiar. Besides, the government would also set up 10 design centres before the end of 1988. These centres would design printed circuit boards and make chips as well. Exit Kit Culture The planning commission has called for urgent policy measures to discourage the "kit culture” in electronics industry. This is to conserve foreign exchange and pro- mote genuine indigenisation. The import bill in the first two years of the seventh plan period exceeded Rs.2000 crores which is more than 60 per cent of total output in the electronics sec- tor in the country. The planning commis- sion has noted that the growth in elec- tronics has been confined to consumer and computer segments. However, there will be a shortfall in the production targets in the segments of communica- tions, components and industrial elec- tronics, which should form the backbone of the Indian electronics industry. A comprehensive national microelec- tronics programme has not yet been for- mulated. A good linkage between Semiconductor Complex Ltd. with con- sumers in the country is yet to be estab- lished. The user sector continued to rely on foreign suppliers for LSI and VLSI Boosting Export The commerce ministry has identified a number of products in the electronics sector to boost exports. The products fall in categories such as consumer electronic components, communication, software, consultancy . services, and test and measuring instruments. In 1986, export earnings from electronics amounted to Rs.240 crores against Rs. 154.50 crores earned in the previous year. In 1975, the exports were worth a meagre Rs. nine crores while by 1990, the figure is expected to touch Rs.1000 The commerce ministry and the depart- 4.18 .1, ELECTRONICS NEWS • ELECTRONICS N ment of electronics will coordinate their efforts to maximise foreign exchange earning through exports. The items identified on the basis of exist- ing export potential are: radio cassette recorders, colour and black and white TV sets; black and white TV picture tubes, ferrites, pre-recorded audio cas- settes, transformers, plastic film, mica and electrolytic capacitors, carbon film, metal film resistors, and cast alloy per- manent magnets; direct reception sets, EPABX and two-way radio communica- tion equipment, personal and microcom- puters, peripherals and eight-bit microp- rocessors. The seventh plan aims at achieving an annual production level of Rs. 10860 crores in value terms by 1989-90 with em- phasis on export promotion in elec- tronics. The break up of the production target is: components Rs.2100 crores; consumer electronics Rs.2000 crores; communication Rs.3100 crores; broad- casting Rs.240 crores; aerospace and de- fence Rs.540 crores; central instrumen- tation and industrial electronics Rs.2010 crores; and computers and office equip- ment Rs.870 crores. The estimated figures for electronics ex- port by 1989-90 are: electronic compo- nents and subassemblies Rs.300 crores; computer, control systems and instru- ments Rs.170 crores; computer software, systems engineering and con- sultancy Rs.300 crores; communication, broadcasting including consultancy Rs.65 crores; aerospace and defence electronics Rs.25 crores; consumer elec- tronics Rs.1000 crores, including video films and feature films worth Rs.140 Meanwhile, the Reserve Bank of India has sanctioned foreign currency loans to software exporters , thus giving green signal for the Exim Bank to launch its software export scheme, announced a year ago. The Exim Bank has so far sanctioned Rs.10 crores as loans to software exporters to import hardware and to make local purchases to meet the export orders. The bank has cleared six applications by January, 1988. Budget proposals The Union finance minister, Mr N.D. Tiwari, made the following proposals concerning electronics while presenting the budget for 1988-89: We have been using the fiscal mechanism for some time to give a boost to the entire electronics sector. As a re- sult of the government policy, substan- tial growth has taken place in this sector, giving employment to lakhs of young men and women. At present, concessional rates of cus- toms duty of 60 per cent or\70 per cent ad valorem are available in respect of specified items of machinery for the elec- tronics industry. With a view to provid- ing a stimulus and keeping in view the latest advances in technology, a uniform concessional duty of 60 per cent ad- valorem is extended in respect of 280 items of machinery for the electronics sector. Customs duty on moulds, tools and dies required by the electronics industry is being reduced further from 60 per cent to 30 percent advalorem. The coverage of the graded structure of duties for raw materials, piece parts and components for the in- dustry is being enlarged. Polycrystalline silicon will now bear a lower duty of 35 per cent instead of the existing-80 per Machinery and instruments required for the manufacture of Rural Automatic Ex- changes based on indigenous technology will attract a lower duty of 30 per cent. A uniform rate of 100 per cent is being pro- vided in respect of a large number of equipments for telecommunication transmission, satellite communication, switching, data communication termi- nals, television transmission, studio and sound broadcasting. Non-electronic components of these equipment will bear a lower duty of 80 per cent. With a view to encouraging production of high- tech items like LSI circuits, microproces- sors and other micro-electronic items, import of 22 items of machinery will be allowed at 15 per cent ad valorem. At present, computers, computer sys- tems and peripherals attract varying rates of duty from zero to 147.5 per cent. As a rationalisation measure, a uniform rate of duty of 80 per cent ad valorem plus countervailing duty is being pro- vided in respect of all computers, com- puter systems, computer peripherals and spare parts. Software will continue to at- tract the existing rates of customs duty of 60 per cent a valorem. As an export incentive, accompanying computer software and start-up spares imported under the policy on computer software export, software development and training will be allowed at the rate applicable to hardware. Computerised numerically controlled systems and their parts at present attract a customs duty of 80 per cent. This is being lowered to 55 per cent ad valorem. Excise duty on CNC systems is being reduced to 50 per cent ad valorem. Colour TV sets of screen size exceeding 36 ems and of assessable value exceeding Rs.5000 per set will now attract an excise duty of Rs.2000 instead of Rs. 1750. However, such sets of value not exceed- ing Rs.5000 will continue to attract a duty of Rs. 1500 per set as at present. Excise duty on audio magnetic tapes is being enhanced to Rs.4 per square metre. Blank audio cassettes are being exempted from duty. Excise duty on computer software is being reduced from 25 per cent to 10 per cent ad valorem. The proposed budgetary allocation for the year 1988-89 for the department of electronics is Rs.120 crores against the revised estimate of allocation for 1987-88 of about Rs.126 crores. Among the major recipients of funds are: National Informatics Centre Rs.31.50 crores; Centre for Advanced Studies in Electronics Rs.1.70 crores; Generation of special manpower for computer Rs.2 crores; Centre for De- velopment of Telematics Rs.5.50 crores; Indian Microelectronics Programmes Rs.4 crores; Manufacture of computer mainframes Rs.8 crores; Society for Applied Microwave Electronics En- gineering and Research Rs.4. 70 crores; Centre for Electronic Design and Technology Rs.4 crores; National Radar Council project Rs.2. 50 crores; and Fifth generation computer System develop- ment programme Rs.4 crores. SENSORS & ACTUATORS Radiant, mechanical, thermal, magnetic, and chemical effects in our environment are nowadays normally detected and measured by electronic means. The conversion of these (analogue) effects into (digital) electrical signals is invariably effected by sensors. These transducers have become so important that without them life on earth would almost literally come to a standstill. Sensors come in a wide variety: it is estimated that there are close to 20,000 different types produced by thousands of manufacturers all over the world. The most important types are used in the detection or measurement of tempera- ture, . pressure, gases, radiation, hu- midity, magnetism, acceleration, direc- tion, angle, flow, level, presence, pos- ition, displacement, and many more. The operation of most sensors depends on optics (lasers, optical fibre, infra-red emitters and detectors), semiconduc- tivity (photo transistors, photo diodes), thermoelectricity (thermocouples), or piezoelectricity. The demands made on most sensors are high: they must be sensitive, corrosion- resistant, inexpensive, precise, stable, easily integrated into a microelectronic circuit, and preferably have a linear in- put/output characteristic. Optic sensors Fibre optic sensors can be regarded as comprising three parts: the optical ' transmitter, the optical modulator, and the optical receiver. Each of the three parts has one major “active” compo- nent. The transmitter has an emitter (such as the LED or a laser); the modu- lator has the stimulus sensor mechanism (such as a diaphragm or a specific op- tical property material); and the receiver has a photodetector. The emitters employed in fibre optic sen- A sensor, the popular name for transducer, is a device that converts a non-electrical parameter into an elec- trical signal or vice versa. The variations in the electrical signal parameter are a function of the input parameter. Most transducers provide a linear, analogue output, but some provide a digital output in the form of discrete values. Most transducers are linear devices, i.e., they provide an output that is a linear function of the input. Like many networks, transducers may be considered as quadripole devices, but one pair of terminals is not necessarily electrical. Most transducers require external elec- trical excitation for their operation; ex- ceptions arc piezo-electric, photovoltaic, and electro-magnetic sensors. An actuator is a device that converts an electrical signal into another form of energy, normally mechanical. It is thus a special type of transducer. Typical examples are loudspeakers, electronic switches, and many measuring instru- ments. sors can be classified as broadband (in- candescent), narrowband (LED), coher- ent (lasers), or blackbody radiators (emitting from inside or outside the fibre). The choice of which one to select depends solely upon the modulator mechanism being employed. For in- stance, fluoroptic thermometers use tem- perature dependent fluorescence of materials at the end of a fibre optic probe. Many sensors for high tempera- ture measurements rely upon blackbody radiation for ranges from 300 to 2000 degrees Celsius. However, the vast ma- jority of applications use their own ex- ternal light sources in the form of an LED or a laser, primarily because of the specific need to accurately control the emitter wavelengths, power outputs, and modulation frequencies. Fibre optic sensors have been helped significantly by developments in LEDs, super luminescent diodes (SLDs), and lasers used in the fibre optic communi- cations and optical disc industry. Semi- conductor LEDs can emit from either their surfaces or their edges, depending upon their design. Surface emitting LEDs (SLEDs) have a wide solid angle on the output beam, and the beam inten- sity is Lambertian. Edge emitting LEDs (ELEDs), on the other hand, have a waveguide mechanism inherent in their structure (as do lasers) and thus have a narrower Gaussian intensity beam. An SLD lies midway between an LED and a laser. It possesses only a single pass gain. As the current density is in- creased, even though an SLD shows a. greater (super) luminescence than an LED, it still does not reach the threshold for multiple pass gain. However, because light is designed to undergo a single pass in the active area of the SLD, its spec- trum is narrower than the LED’s. 4.20 The amount of power an emitter needs to generate for a fibre optic sensor appli- cation is a function of several design fac- tors. The power from an emitter must first be transferred into an optical fibre. In many cases, it must be tailored ap- propriately (e.g., through a polarizer, as in the case of a fibre gyroscope) prior to such introduction. LEDs, SLDs, and lasers are not only made of the same semiconductor materials but also have the same basic device structure. In principle, these semi- conductor devices have a p-n junction, which upon being forward biased leads to a recombination of holes and elec- trons with the simultaneous emission of photon energy. The wavelength of this emitted light is in turn governed by the composition of the semiconductor ma- terial. Thus, the amount of aluminium determines the center wavelength of the emitted light. LEDs, SLDs, and lasers are in an as- cending order of sophistication (see Table 1). An SLD can be regarded as an emitter that -is half-way between an LED and a laser. An LED produces spon- taneous emission in its “active” region and thus has a wide spectrum about a central wavelength. A laser has a built-in mechanism in its structure so that light produced in its active region is made to oscillate between its specially designed front and back facets, thus leading to a primary wavelength or mode of oper- An important application of the optic sensor is in robotics, since it makes poss- ible artificial vision, without which robots can not reach their full potential. Another application of the optic sensor is in seam tracking and process control in arc welding. The sensor is inherently insensitive to the arc light. An interesting application is the oxygen sensor that measures oxygen saturation in the human blood so as to control the rate of a pacemaker. The sensor is inte- grated in the stimulation catheter and located in the right ventricle of the heart. A new line of intelligent sensors prom- ises to rid cars, buildings, aircraft, and factories of most of the increasingly complex wiring. One of these sensors uses a multiplexable optical encoder chip produced for Honeywell by its Optoelec- tronic Division in Richardson, Texas. This chip combines sensors and ana- logue and digital circuits on a single wafer. The on-chip sensors can deter- mine direction or rotation, rotational velocity, and angular position. A new technique to measure physical correlations in multi-use fluid transpor- tation systems has been developed by the Berg Akademie Freiberg in Federal Germany. In this, fibre optic probes are used to accurately measure particle con- centration, fluctuation, speed, size, and cross-sectional distribution — all critical in process control and regulation. Among the measurement said to be possible with the system are impurities in water, organic or inorganic liquids, gas bubbles in liquids, crystals in saturated solutions, and flocculants in pipes and other apparatus. Japan’s Sofia University has developed an optic sensor that controls on-off switching for use in optical computers. The optical switch needs no electric cir- cuits, since the optical signals are con- trolled by light beams. Remote control and information exchange are possible. The development is expected to ac- celerate the development of optical in- formation processing technology, which forms the basis of optical computers and optical communications. Semiconductor sensors Semiconductor sensors have two import- ant advantages over other types: they are invariably produced from silicon, which is a plentiful and well-researched material, and they can easily be inte- grated with amplifier and logic circuits onto a single wafer. These sensors are normally encountered in the form of photo transistors or photo diodes. A photo transistor is a detector that consists of a bipolar junction tran- sistor operated with the base region floating. The potential of the base region is determined by the number of charge carriers stored in it. The elec- tromagnetic radiation to be detected is applied to the base of the transistor and produces the base current. The transistor is operated essentially in a common- emitter configuration. A photo diode produces a current when it is illuminated. There are two main classes of photo diode: depletion-layer and avalanche. Depletion-layer diodes consist commonly of a reverse-biased p- n junction operated below the break- down voltage. The p-i-n and Schottky photodiodes are versions of the de- pletionlayer type. Avalanche photo diodes are reverse-biased p-n junction diodes that are operated at voltages above the breakdown voltage. Sensors for the detection of gases are normally manufactured from other semiconductors materials, such as tin oxide, zinc oxide, titanium oxide, and Thermocouple sensors Thermocouple sensors depend on the phenomenon that when two dissimilar metals are joined at each end and the two resulting junctions are maintained at different temperatures a voltage is devel- oped between them. Copper-constantan or iron-constantan thermocouples can be used up to 500 °C. Temperatures up to about 1500 °C may be measured with the aid of a platium/platinum-rodium alloy thermocouple, and even higher temperatures may be measured with an irridium/irridium-rhodium alloy ther- mocouple. Piezoelectric sensors When certain materials are subjected to mechanical stress, an electrical polariz- ation is set up in the crystal and the faces of the crystal become electrically charged. The polarity of the charges reverses if the compression is changed to tension. Conversely, an electric field ap- plied across the material causes it to con- tract or expand according to the sign of the electric field. Piezoelectric sensors are important since they couple electrical and mechanical energy and, therefore, are used as gramophone pick-ups, loudspeakers, microphones, to name but a few. Practical applications Temperature sensors. As already men- tioned, many temperature sensors are based on the Seebeck effect that occurs in a thermocouple. They are normally produced in the shape of a probe: a wide variety of such probes is shown in Fig. 1 . Fig. 1. A selection of Iherniocouple tempera- ture probes. (Photograph courtesy Omega International Inc.) Another well-known type of temperature sensor is the thermistor. This is basically a resistor, made from semiconductor material, that has a negative temperature coefficient. This means that when the ambient temperature rises, the element becomes more conductive (its resistance decreases) and the consequent change in voltage across it is a measure of the tem- perature rise. It should be noted that there are also thermistors with a positive temperature coefficient, whose resist- ance, therefore, increases when the tem- perature rises. Temperature may also be measured by measuring infra-red (heat) radiation, for which an infra-red sensor as shown in Fig. 2 is used. This technique, called Fig. 2. Typical infra-red sensor. thermal imaging or thermography, is based on the property that each body or object radiates heat. The technique, for which a camera with a suitable lens system may be used, does not require any external source of illumination. It is used, for instance, in production tests to determine whether any component heats up too quickly (and is, therefore, almost certainly faulty). It is also used in medicine for diagnostic purposes to determine whether any areas of the body have an unusual temperature distri- bution. Pressure/force sensors. Although there are various methods of measuring pressure and force, the most common one makes use of the piezoelectric effect as discussed earlier in this article. The most widely used material for the manufacture of pressure/force sensors is quartz. This material has some import- ant advantages over others: (1) it is strong; (2) it is cheap; (3) it is a good electrical insulator so that the electric charge caused by the pressure collapses only slowly. l-'ig. 3. Constituent parts of a piczo-electric pressure sensor. (Photograph courtesy Telefunken AG). The parts making up a typical piezoelec- tric sensor are shown in Fig. 3. It con- sists of a wafer of silicon only 1 mm in diameter, onto which a tiny piezoelectric 4.22 ■+****.** ,988 crystal and four resistive tracks have been etched with the aid of ion implan- tation. When pressure distorts the crystal, the resistance value of one or more of the legs of the resistance bridge changes. This type of sensor is versatile: it can be used for measuring absolute or relative pressure, overpressure, and pressure difference. It is suitable for pressures up to 40 MPa. This type of sensor is, of course, widely used in all sorts of weighing machine. Other areas of use are hydraulics, water works, refineries, filter plants, pressure chambers, and loudspeakers. Pressure sensors are also used in ac- celerometers, but there they operate somewhat differently. Such a sensor for measuring mechanical vibrations or im- pact contains a freely moving seismic mass and a piezoelectric element (nor- mally quartz)— see Fig. 6. When the seismic mass is accelerated in the direc- tion of its axis, it exerts a force onto the quartz element that is proportional to the acceleration. The element is then dis- torted and the consequent piezoelectric voltage is used to charge a capacitor. This charge can be measured, but this has to be done quickly as otherwise some of the charge leaks away. At fre- quencies below the resonant frequency of the sensor, the seismic mass follows the vibrations faithfully. This type of sensor usually contains an integrated preamplifier. Humidity sensors. Humidity sensors are used almost exclusively in hygrometers. Fig. 5. A selection of typical pressure sensors. (Photograph courtesy Bruel + Kjacr). i.e., instruments for measuring the humidity of air. In the past, these sen- sors used a human hair, or a strand of silk, but nowadays they use a capacitor, a dew point mirror, or optical means. The dew point mirror sensor depends on the effect that when a smooth surface is cooled it mists up. The moment this misting up starts is determined optically. Since it is accurately known at which pressure and temperature gases con- dense, this technique yields very accurate Another type of dew point sensor con- sists of a very small wafer of resistive material which has been coated with a hygroscopic chemical. The wafer is fitted with two electrical terminals. When mist forms on the coating the resistance of the wafer increases. This type of sensor is quite vulnerable, but because of its very small dimensions, it is used in Sony’s 8 mm Camcorder. Optical humidity sensors make use of the property that gas molecules absorb energy at certain frequencies: water vapour does so in the infra-red region. It is thus possible with the aid of an infra- red sensor to determine how much energy is absorbed. The higher the humidity, the more energy is absorbed. This technique has the disadvantage that the infra-red sensor soils up easily and then becomes unusable. Nowadays, the most important and best- value-for-money type of humidity sensor Fig. 6. Construction of accelerometer sensor. M = seismic mass: P = piezo-electric element; B = underside; R = initial tension. (Courtesy Bruel + Kjaer). is based on a capacitor. This is, of course, a special capacitor which as a dielectric that is sensitive to humidity. In the Valvo sensor— see Fig. 10— the rtk dielectric is in the form of a foil that has f he been coated at both sides with gold. I which forms the electrodes. Humidity _ f changes the dielectric constant of the foil and thus the capacitance of the ca- \ pacitor. Since this capacitor forms one of the legs of a capacitive bridge, the ' J change in capacitance can be readily converted into an electrical voltage. Fig. 11. Some typical gas sensors. (P graph courtesy Driigcrwerk AG). Gas sensors. As stated before, gas sen- sors are normally based on a variety of semiconductor materials. Such materials have the property that their resistance decreases when certain gases arc present in the surrounding air. This effect is caused by adsorption of gas molecules on the surface of the semiconductor material. The consequent layer of gas molecules influences the conductivity, and thus the resistance of the element. These sensors are very sensitive: concen- trations of only 1 ppm of the relevant gas in air are readily detected. A variant of this type of sensor is Telefunken’s 1SFET— see Fig. 12. Basically, this is a modified MOSFET in which the usual metal gate has been re- placed by a layer that reacts to the ions of certain gases. ISFETs are un- breakable, small, have a low-impedance output, have a large linear range of oper- ation, are temperature compensated, and provide an output signal that is suitable for driving a microprocessor. Many gas sensors still depend (and will continue to do so) on a chemical reac- tion to generate an electrical voltage, current, or resistance change. Yet other sensors use the heat generated by the combustion reaction when a gas hits the surface of the sensor. This heat is ap- plied to a platinum wire whose resistance then changes. There are also optical gas sensors and these are used particularly for the detec- tion of fire or smoke. They normally use a photo diode or photo transistor to monitor the light absorption behaviour of the surrounding air. When smoke darkens the air, the photo transistor switches off and this operates an appro- priate actuator. Fig. 7. These twin axis gyros belong to a range of inertial sensors that includes rapid start coasting displacement gyros, rate gyros. Dynamically Tuned Gyros, and linear ac- celerometers. (Photograph courtesy British Aerospace). Fig. 12. The ISFET is a modified MOSFET used as a gas sensor. (Courtesy Telefunken AG). Fig. 8. Construction of a combined pressure and temperature sensor. (Courtesy Sen- sortechnik Widemann). Fig. 9. Selection of EPI Serit lies. (Photograph courtesy ICi Light sensors. Popularly probably the best known type of sensor is the light sensor. This can be based on a photo di- ode, photo transistor (see Fig. 13), p-i-n M diode, photo varistor, or solar cell. All j °f these are made from the same material, silicon, and function in similar 3 fashion at wavelengths from about For wavelengths below 400 nm (ultra- 400 nm to around 1000 nm. violet light), photomultipliers are used. Photons enter the silicon and cause a These are normally constructed as a number of electrons to jump to a differ- valve and have the usual advantages of humidity ent energy level. This in turn causes a electron tubes: good bandwidth, low y Valvo photo current which can be used to noise factor, high amplification. operate an actuator. Primary electrons, emitted from the el.ktor indil •pril 19BB 4.23 13. Some typical photo transistors le centre an infra-red photo diode. Fig. 10. Some tiny semiconduc sensors. (Photograph corn Philips). cathode as a result of photon bombard- ment, are accelerated by the field be- tween anode and cathode and arrive at the anode with great energy, causing a current in the anode circuit. The anode current is much greater than the original cathode current, whence the name photomultiplier. Photo detectors as described were also used as light sensors in camera tubes, but nowadays they are largely super- seded by CCD (charge-coupled device) sensors, particularly in video cameras, medical cameras, and robotics. CCDs are small: typically, one the size of a 28- terminal IC has a resolution of 60,000 Fig. 14. Photo diode used as a gas sensor. (Photograph courtesy Plessey Semiconduc- elements. Each of these elements con- sists of a photo diode with a light- sensitive area of only a few micrometres. Each photo diode is accompanied by a MOS capacitor that stores the generated photo electrons. The charge on the capacitors is read at regular intervals by a shift register. Suitable circuits inte- grated on the device process the infor- mation to a suitable level for driving a microprocessor. Biological sensors. During the past few years, a new type of sensor has entered the fields of biology and medicine. These so-called biosensors consist of biological molecules, such as enzymes and antibodies. When such sensors react to other substances, a small electric signal is generated that can be detected with the aid of a suitable electrode (probe). Remote sensing. Another interesting new field where sensors are indispensable is that of remote sensing. This exciting new technique has been made possible by the routine availability of satellite infor- mation for the entire surface of the earth. Remote sensors on board satellites provide digital data in seven wavebands of visible light, reflected infra-red radi- ation, and thermal infra-red radiation. Different surfaces reflect different amounts of radiation, which is why they appear in different colours and light in- tensities to us. In the same way that we distinguish objects by their appearance (but in a more sophisticated manner), these remotely sensed images can be used to identify the land-cover types which exist in an area. Since they record in the infra-red region, many things we cannot normally see are shown. Crop condition and the thermal properties of buildings or water can be ‘seen’ and displayed, for instance. Remote sensing enables scientists to study the earth’s surface on a scale which was until recently only dreamed of. For a very small part of the time it would take to survey a large area by con- ventional methods, digital information can now be used to identify and measure the extent of crop types, major land uses, soils, properties of water bodies, geological structures, and vegetation conditions. In sparsely populated areas, the existence of certain surface features is being established for the first time, and over all areas of the world, what were previously partial surveys can now be completed. A great attraction of remote sensing is the relatively low cost of large-scale surveys. For instance, im- ages with ground resolution down to Fig. 15. The DART (Dual Axis Rale Trans- ducer) is a miniature gyro particularly suited for stabilization of laser, infra-red. and radar seeker systems. The sensor, which is only 40.6 mm long and has a diameter of 18 mm, uses a rotating piezo-electric crystal to sense the applied rate. (Photograph courtesy British Aerospace). 10 m and covering 50 km x 50 km can be obtained for less than £1,000. Much pioneering work on remote sens- ing has been carried out in Britain by Salford University. A final thought. Although the science and technology of sensors and actuators has made vast strides in the past few decades, the most complex, reliable, and versatile sensor system remains man. Coupled with his intelligent data pro- cessing unit which almost certainly will not be emulated during the life of anyone alive today, he forms a for- midable system of intelligence. A pity we do not always appreciate it. We acknowledge with thanks the co- operation and help received from the fol- lowing organizations in the preparation of this article: British Aerospace; Bruel + Kjaer; Dragerwerk AG; Entran Ltd; Hawker-Siddeley; IGI Consulting Inc.; Omega International Corporation; Plessey Semiconductor; Salford Univer- sity; Sensortcchnik Wiedemann; Siemens AG; STC Mercator; Telefunken; Valvo Philips. 4.24 elektor India april 1988 THE REASONS FOR MINIATURE TRANSDUCERS by Mike Coope* A statement made by a senior engineer in the aircraft industry recently summed up some of the reasons why many in- dustries have turned towards the use of miniature transducers. ’’Years ago our aircraft engine control electronics were the size of a small suitcase but today the total package size has been dramatically reduced. This obviously means that when we are testing these components, particularly for important vibration tests, that the additional weight that we add by employing transducers on to the unit can, if the size, and hence weight, are not minimal, completely change the test performance. We must therefore look for transducers, in this particular case accelerometers, with the smallest possible mass”. The simple but important rule that ap- plies here is that F=m x a. Hence the additional force that is applied to the item under vibration test is dependent on the mass and the distribution of mass in the overall dimensions of the transducer. For instance, suppose that a conven- tional sized accelerometer is used for the vibration test. If it weighs 100 grams and is being used for 100 g acceleration, by calculation in the F=m x a formula, a further 10 kg is added to the weight of the component under test. This will con- siderably distort the results of the vi- bration tests. Entran designs and manufactures piezo- resistive semiconductor strain gage ac- celerometers which have the capability of measuring both steady-state dc and high response dynamic vibration inputs. The EGA range offers models with measuring ranges from 0-5 g up to 0- 5000 g within a housing as small as 0.140 X 0.140 x 0.270 in. (3.4 mm x 3.5 mm x 6.75 mm) and weighing as little as 0.5 grams. The EGA Series has the desirable feature of fluid damping to protect against resonant excitation. The EGAX model the additional feature of internal overrange stops which give the accelerometer the capability of accepting an overload of ±10,000 g in either the normal sensitive acceleration measuring axis, or in all directions. This feature is available even on the lowest measuring range of ±5 g. Not only will this over- range feature make the accelerometer suitable for many impact and guidance applications, but it will also protect a valuable measuring transducer from day-to-day mishandling on the work- shop floor where accidents will unfortu- nately happen. These particular features of low mass and high overrange of the EGAX ac- celerometers have proved beneficial to a particular area of research in the medical field. Various departments of Medical Establishments have the need to study the features of muscular tremours and diseases such as Parkinson’s disease. With patient involvement, the need is for an unobtrusive and small vibration measuring device to monitor the movements of the patient’s limbs that has the capability of withstanding several thousand 'g' should the acceler- ometer be inadvertently dropped to the floor. This specification exactly fits the Entran EGAX Series of accelerometer, which has been used for many years by several Medical Establishments. Entran designs and manufactures a var- iety of semiconductor strain gauges, of which the smallest is active over 0.020 in. (1.50 mm) length by 0.006 in. (0.150 mm) width. The distinct advan- tage of these resistive sensing elements is (a) their micro-miniature size and (b) their extremely high gauge factors. The term gauge factor (GF) is a measure of the incremental change in the resistance of the strain gauge for a given incremen- tal change in the active length of the gauge, i.e. GF = AR/A1 Hence, gauge factor is a measure of the sensitivity or output performance of the strain gauge. To give a comparison, stan- dard foil strain gauges typically have a gauge factor of 2.0, whereas that for a semiconductor strain gauge can be typically 150. This means that when semiconductor types arc used, we can ex- pect improvements of outputs in the region of approximately 75 times. En- tran takes full advantage of these properties of their strain gauges in their extensive range of miniature transducers. In the accelerometer (Fig. 1) strain gauges are bonded in pairs to the top and bottom surfaces of a single degree of freedom cantilever beam. A mass is attached to the end of this beam and the resulting deflection of the beam when experiencing a ’g’ force results in a linear signal output when the strain gauges are wired into a Wheatstone bridge and an excitation voltage i applied (Fig. 2). 2 [vt; ■■■■■ 1 - The advantages of strain gauge proper- ties are also used in miniature pressure transducers (Fig. 3) where the parameter is sensed by monitoring the deflection of a metal diaphragm. To achieve the highest possible dynamic response, the diaphragm must be small and its de- signed full scale deflection minimal. With their inherent high sensitivity and ultra-miniature size, semiconductor strain gauges can be used on a stiff, low deflection diaphragm to achieve this criterion. A further advantage of the miniature diaphragm is its low deflec- tion, resulting in low stress levels within the diaphragm material, ' and hence almost infinite fatigue life. Thus Entran pressure transducers offer measurement of both static and high frequency dynamic inputs. The smallness of pressure has an import- ant bearing on their performance. Size can also be important on the overall ef- fects of the test. For instance, in the testing of the aerodynamics of scale models of new military and civil aircraft, automobiles, aircraft components such eleMor India apill 1988 4.25 as helicopter blades, missiles, and gener- ally all transport where the efficiency of mdvement is important, the transducer must be as unobtrusive as possible so as not to alter the original shape of the test piece. Entran have ultra-miniature trans- ducers of low profile designs (EPL) with a thickness of 0.040 in. (1.02 mm), which are used in a recessed mounting to give original aerodynamic flow lines of the test model. Alternatively, all Entran EPI pressure transducers are available with diameters from 0.080 in. (2.03 mm) down to 0.050 in. (1.27 mm) and are used in many wind tunnel tests because they can be easily accommodated within the rivet head of an aircraft structure without affecting the structure and pat- tern of the normal air flow. Entran’s specialization is in the design and manufacture of miniature trans- ducers for the measurement of acceler- ation, pressure, load and strain, but many other models have been developed with the requirements of Entran’s cus- tomers in mind. Although standard models exist, it is accepted that many transducer requirements fall outside the normal specification and, for these situ- ations, Entran has special engineering facilities to provide the low-cost OEM style transducer or the ultra-sophisti- cated, latest-technology, quality-assured transducer. Within this framework of adaptation to market requirements, Entran offer a range of accelerometers and pressure transducers which are the outcome of long experience of transducer design. The new range of devices offers robust styling, both internally and externally, as well as the optional addition of internal miniature electronic circuits to give (a) amplified output up to 10 V FS; (b) supply regulation; (c) custom filtering. This short article emphasizes a few of the aspects that miniature transducers can play in the latest fields of industrial, research, medical, aerospace, chemical, automotive and many other industries. Further information on ENTRAN sen- sors may be obtained from ENTRAN Ltd • Sales & Technical Centre • 5 Albert Road • CROWTHORNE RGII 7LT • Telephone (0344) 778848 • Telex 847422 • Fax (0344) 777991. * Mike Coope is Technical Sales Director of Entran Ltd. SIGNAL DIVIDER FOR SATELLITE TV RECEIVERS by R. van Terborgh Our first application of a monolithic microwave integrated ampli- fier (MMIC) from Avantek is a wideband amplifier and splitter that makes it possible to connect two satellite TV receivers to a single outdoor unit. In other words: this is your chance to share the cost of a dish plus LNB with your next-door neighbour! As stated in the introductory article on MMICs (reference "’), these new devices are eminently suited to building high performance wideband amplifiers with only a handful of components. In the present application, a single MMIC is used for amplifying the IF output signal of a commercially available low noise block down converter (LNB or LNC). The standardized IF bandwidth of LNBs is 950— 1750 MHz, but it should be noted that these are not absolute band edges. Most LNBs have a relatively high conversion gain (55 dB typ.), but this is often found to decrease with frequency. Similarly, the attenuation of the downlead coax cable increases with fre- quency, so that the highest down- converted transponder signals are almost invariably of lower absolute amplitude than those further down in the IF band, while C/N figures are still roughly the same because reduced gain results in less noise (compare the S-meter reading of Super Channel to that of, say. Teleclub Switzerland). The foregoing observations have conse- quences for the design of a cable ampli- fier/signal divider for use in satellite TV receiving systems: 1. The amplifier should have a relatively high drive margin to prevent it being "blocked” by the high output levels sup- plied by the LNB. 2. The frequency response of the ampli- fier should be as Hat as possible across the entire IF band. Both requirements are met by the pro- posed amplifier/splitter based on the Type MSA0404 MMIC, which ensures an output power of +12 dBm at 1 dB compression, and a third-order intercept point of +26 dBm (0dBm=l mW in 50 Q). The 2-way signal divider described here provides a modest, but often welcome, insertion gain of about 4 dB on both outputs, and allows relatively inexpens- ive coax cable to be used for connecting the indoor units. Still, do not be tempted into using, say, 30 m RG-58, or ubi- quitious ”TV coax”: the massive at- tenuation such a cable introduces is im- possible to overcome by the best cable amplifier or IDU input stage. Stick to good quality COAX12, COAX6, or H43 cable (all 75 Q), terminated in BNC, N or F connectors. Green or yellow coax cable occasionally seen in TV and radio distribution networks is also suitable, but appropriate plugs may be difficult to obtain. Circuit description The circuit diagram of Fig. 1 shows the single-chip amplifier and 2-way signal divider. The LNB is connected to BNC, N or F socket Ki; the 2 indoor units (IDUs) to K2 and Kj. Amplifier ICi is a TVpe MSA0404, whose technical characteristics can be found in Fig. 5 in Gain of the MM1C is 8 dB; input and output impedance are 50 Q. The mismatch between the 75 Q cable and the 50 Q input of the amplifier (Ki; MMIC), is of no practical consequence (reference |!> ). The amplified wideband signal is ap- plied to a resistive splitter, R2-R3, for feeding to the indoor units. Given the amplification of the MMIC, and the loss in the splitter, the net gain on each chan- nel is still about 4 dB (do not forget that the MMIC has a 50 Q output, and that Kj and Kj are terminated in 75 Q). Mo.re gain is not desirable here because it would lead to overdriving of the input stage in the IDU. Output Kj on the amplifier/splitter ac- cepts the LNB supply voltage carried on the downlead coax cable to the IDU. This supply voltage is also used for powering the MMIC via Ri-Lj, and ap- pears on Ki after passing through chokes Lj and Li. Indoor unit 1 (IDU 1) connected to Kj powers the LNB and the signal divider. The supply voltage on the coax cable should not disappear when IDU 1 is switched off, because this would make make reception on IDU 2 impossible. The Elektor Electronics in- door unit (reference m ) causes no prob- lems in this respect: the LNB supply voltage is present on the RF input as. long as the unit js connected to the mains, i.e., irrespective of the position of the on/ off switch. Capacitors Cj, Cj, C<. and O ensure ad- equate supply decoupling, while Ci, Cj and Cs are DC blocking capacitors with a low reactance and stray inductance at the frequencies involved. Series resistor Ri should be dimen- sioned in accordance with the LNB operating voltage supplied by the IDU %on the downlead coax cable. The MMIC draws about 50 mA at the recommended supply voltage of 5.5 V, so that Fig. I. Circuit diagram of the wideband amplifier and splitter for satellite TV receivers. The value of 220 Q shown in the circuit diagram ensures safe operation of the amplifier when connected to the Elektor Electronics IDU (Vi.nb=15 V). Warning: some indoor units supply an LNB voltage as high as 18 V or even 24 V. Always measure Vlnb on the centre tip of the RF input connector of your IDU with the LNB connected to the downlead coax cable. Then calculate Ri as shown above. The dimensioning of Ri is not critical: calculate the theoretical resistance, select the next higher value from the E12 series, measure the voltage on the RF OUT/BIAS terminal of the MMIC, and decrease the resistor value if re- quired, until the operating voltage is be- tween + 5 V and +6 V. Be sure to use an ordinary 'A W carbon film resistor: its inductance is essential in this appli- cation. Construction Prepare the small Eddystone enclosure shown in the photographs prior to start- ing any soldering work. Study the lo- cations of Ki, Kj and Kj on the PCB before drilling the holes for these flange- type BNC sockets, which are fitted onto the lid as shown in Fig. 3. You may have to provide new mounting holes on the PCB if you intend to use Type N or F sockets instead of BNC. Mount a Fig. 3. Side view of a prototype of the signal suitable clamp to the box to enable at- divider, showing the PCB and the RF con- taching this near, or onto, the dish aerial nectors secured onto the lid of a diecast stand, if this is the most favourable lo- enclosure, cation for the splitter considering the lengths of the downlead cables. Provide lettering on the lid to rule out any likelihood of erroneous connections. Now proceed with the completion of the printed circuit board shown in Fig. 2. Less experienced constructors should note that all parts, with the exception of R i = (V lnb — 5 .5)/0.05 [£]. .4.27 the 3 sockets, are fitted direct onto the track side of the board. The leads of C7, Ri and the 3 inductors should be kept as short as possible. The MMIC is seated in a 04 mm hole, and its 4 leads are soldered straight onto the relevant copper areas. All capacitors, with the exception of tantalum bead type C7, are surface mount devices, secured in place by fast soldering with a low- power iron. The same goes for divider resistors R2 and Rj. The 3 home-made inductors in the amplifier are identical. The winding data are as follows: Li; L:; Lj = 12 turns 00.5 mm (SWG25) enamelled copper wire; close- wound; internal diameter: 2.5 mm. Use a sharp knife to remove the protec- tive PTFE collar around the centre pin where this protrudes from the BNC socket. The 3 sockets are held against the outside of the lid, and the PCB is pushed onto the centre pins that protrude at the inside. Press the PCB firmly against the lid, and push the twelve M2.6 screws through the holes in the PCB. Carefully tighten the screws, and solder the centre pins of the RF sockets onto the copper islands at the track side of the board. Finally, be sure to terminate output IDU 2 in 75 Q at all times. In most cases, the attenuation of the downlead cable is high enough to ensure proper termination of the amplifier if IDU 2 is disconnected in the home. If, for some reason, the downlead cable to IDU 2 is temporarily disconnected from the amplifier output, this must be ter- minated in a 75 Q dummy load (a 75 Q resistor fitted in the RF plug). In general, the signal divider should be fit- ted as close as possible to the LNB.RGK References: 121 Loss encountered when interconnec- ting cables having the incorrect im- pedance. By Dr P. Brumm. VHF Com- munications, 3/1974. ,J) Indoor unit for satellite TV recep- tion, parts 1, 2 and 3. Elektor India, November 1986, January 1987, February 1987. Fig. 4. All parts, with the exception of the RF connectors, are fitted at the track side of the PCB. The small, greyish, rectangular, blocks are surface-mount resistors and capacitors (prototype). INFRA-RED DETECTOR FOR ALARM SYSTEMS As a follow-up to the design abstract on the Type PID-ll published last year this article describes a versatile infra-red sensor which will find many applications in security and alarm systems. The Type PID-ll infra-red sensor from Siemens introduced in reference (l> is a versatile component that lends itself to building a simple, yet effective and sensi- tive, transducer that detects heat emanating from mammals. To be able to understand the basic operation of the circuit shown in Fig. 1, it is rec- ommended to read the sections Appli- cation lips and Some suggested circuits in reference The infra-red sensor, IQ, is powered by a regulated 5 V supply. The reference voltage available on pin 4 (2.2 V) is ap- plied to opamp Ai for comparison with the voltage at pin 3. The voltage on pin 2 of the comparator is .held slightly below the reference with the aid of potential divider R1-R2. In the non- activated state of the circuit, the output of Ai is, therefore, high. When the sensor detects infra-red radiation, how- ever, the comparator supplies a short, low, pulse. Opamp A2 functions as a monostable multivibrator (MMV) and a buffer whose gain depends on the am- bient light intensity measured by phototransistor Ti. The trigger threshold of Ai can be adjusted with preset Pi. The preset in network Cs-Rs- P2 at pin 6 of A2 enables adjusting the Fig. I Circuit diagram of the PID-ll based infra-red detector for alarm systems. component in heat emanating from a mammal, the voltage on pins 5 and 7 of A2 drops abruptly to practically nought. The voltage on Cs is no longer main- tained by Di, and the capacitor discharges. At the end of the discharge period, the monostable reverts to its in- itial state. Opamp A2 supplies a digital (TTL compatible) switch pulse at output DIG. . The function of relay driver T2 and alarm indicator D3 is self-evident. The maximum on-time of the relay that can be set with P2 is about 1 minute after detection of any single alarm pulse from the detector. Construction, adjustment and applications The printed circuit board shown in Fig. 2 holds all the components in the circuit diagram, and so enables ready construc- tion of the compact detector unit. The completed board is shown in Fig. 3. mono time of the MMV, i.e., the hold time of the circuit. Incident daylight on phototransistor Ti effectively raises the trigger threshold for the MMV, and hence ensures automatic disabling of the alarm by reducing its sensitivity. Pins 5 and 6 of the monostable are logic high in the non-activated state of the alarm. When the PID-ll senses the infra-red Note that the top of the phototransistor, Ti, and that of the alarm indicator, Dr, is level with the top of the PID-11. Relay driver T2 and regulator ICj do not need heat-sinks. The completed board can be fitted in a water resistant, strong ABS enclosure, with suitable grommets, strain reliefs and sockets for the connection of the wires for the relay, the mains, and the digital output, if used. When the circuit is used as an automatic porch light controller, it is recommended to fit it in a sheltered position over the front door, paying due attention to safe and sound insulation. In many cases, it may be safer (and cheaper) to use a separate 8 V AC adaptor plugged into a mains outlet in the home, rather than the PCB mounted mains transformer, Tri. The adjustment of the circuit, i.e., the sensitivity and the relay on-time, is governed by the application in question. Initially, it is recommended to test the completed circuit by adjusting Pi such that the circuit is just off in the absence of an infra-red source. For this adjust- ment, it is necessary to temporarily cover the phototransistor against incident light. The value of tantalum capacitor Cs may be increased when the maxi- mum relay on-time of 30 to 60 seconds is too short for the given application. It should be noted that the PID-11 signals detection of an infra-red source by an output pulse of about 1.5 s rather than by a continous logic level. The pulse, which can be measured on the output of Ai and A2, is positive or negative, in- dicating a cold-to-warm or a warm-to- cold transition, respectively. When fitting the IR detection unit, it is important to ensure that it can not detect heat from external sources (sunlight, heating systems, etc.). Also note that the sensitivity of the IR detector depends on the ambient temperature. Strong mag- netic fields may cause interference in the sensor and hence spurious operation of the alarm. Finally, be sure to avoid overloading the relay contacts by switching too heavy loads. Si Reference: 111 Design abstracts: Passive infra-red detector Type PID-11. Elektor India, April 1987. 4.30 el, [ImNAONj PCBs & Set of COMPONENTS e®® Fig. 2 Track layout and component ing plan of the PCB for building (he IR detector. Bi =B40C400 or B40C1000 Oi;D2=tN4148 D3— red LEO Dd = 1N4001 Ti=BP103-3 ICi =TL272C or TLC272 IC2 = PID-11 (Siemens! IC3 = 7805 TRi = PCB mount transformer: 8 V: Rei = PCB mount SPOT relay: 5 V; . Siemens V23127-B001-A101. 2-way PCB mount screw terminal. PCB Type 87067 (available through Services! . Readers ABS enclosure. TOWARDS THE SUPERNODE COMPUTER by Dr. Chris Jesshope, CEng, FBCS, MIEE Department of Electronics and Computer Science, Southampton University Esprit is a research programme sup- ported by the Commission of the European Communities. It is currently funding a collaborative project to de- velop and exploit a low cost, high per- formance supercomputer, in which Southampton University* 1 ' is playing a major role. Unlike conventional supercomputers, such as the Cray 1 and Cray 2, which tend to use expensive ultra-fast circuits, the prototype being designed at Southampton University makes use of the latest microprocessor technology. The Supernode supercomputer is based on the revolutionary transputer, which is a modern microprocessor designed by INMOS (2) as a component of parallel processing systems. Parallel processing uses many processors to obtain increased system performance. For example, the Supernode supercom- puter may eventually contain several thousand transputers, all of which could be brought to bear on a single appli- cations problem. The research pro- gramme, as well as developing several prototype supercomputers, will in- vestigate programming and applications techniques for this novel computer archi- tecture. The exploitation of such large scale parallelism is by no means well estab- lished and the United Kingdom, like the rest of Europe, has an active research programme to ensure its information technology industry remains competitive during this period of major change. In fact Southampton University was in- itially funded by Britain’s Alvcy pro- gramme to investigate the feasibility of using the transputer as a basis for a supercomputer. Other partners in the Esprit collaboration also had prior funding for transputer research. Prototype production The collaborators in this project come from Britain and France, and include universities, small and large companies, and a government research establish- ment. The role of prime contractor with overall project management is filled by the Royal Signals and Radar Establish- ment (RSRE)* 3 '. The remainder of the plan is divided into work packages which involve the close cooperation of small groups of collaborators. The prototype designing is being led by Southampton with collaboration from the RSRE and the two industrial part- ners, Thorn EMI <4) and Telmat of France. These two will manufacture seven small and four large machines for the work, with commercial exploitation to follow. Provision for real time input and output to the supercomputer is be- ing developed by collaboration between RSRE and Thorn EMI. The major component of the Supernode supercomputer — also known as the ReconFigurable Transputer Processor — is the newly announced T800 transputer, which was developed within a work package under this collaboration by IN- MOS. IMAG at Grenoble University, in France, is working on the system software for the Supernode machine in collaboration with Southampton. They are also study- ing the implementation of high level languages such as Prolog on this novel architecture. The remaining collaborators are working on applications of the supercomputer, including image and signal processing, image generation by ray tracing, com- puter aided design (CAD) for very large scale integration (VLSI), computer aided manufacture (CAM), and appli- cations in science and engineering. Transputer development The T800 transputer is the major com- ponent in the Supernode. It is a derivative of the T414 transputer an- nounced by INMOS some two years ago. The major difference between these two chips is that the T800 contains an ad- ditional processing unit for handling floating point numbers. Floating point or real numbers may have fractional parts and have a very Wide range of values; most microprocessors handle only integers (whole numbers), with op- erations for real numbers being provided by software, or alternatively by an ad- ditional special chip. Southampton University identified the limitation of software floating point at an early stage in its feasibility study, based on its applications in science and engineering. This was basically a speed limitation, for it should be noted that all supercomputers provide support for rapid floating point computation and, inevitably, a software implementation will not provide a comparable perform- ance. The T800 chip is about 1 cm ! and con- tains about 250,000 transistors. Unlike more conventional microprocessors, it can be used entirely on its own as it con- tains a simple but efficient 32-bit integer processor, the floating point processor which handles numbers stored using up to 64 bits of information, some very fast random access memory (it can store 4096 characters), and four high speed in- put and output channels. It is the latter communication channels that provide the key to the transputer’s success as a parallel processing compo- nent. In any parallel processing system, the processors must be able to com- municate with each other, to share data and to synchronise their activity. The communication channels on the trans- puter provide both of these facilities. A single channel can transmit about two million characters per second in each direction over the three-wire circuit used to connect transputers together. The four links provide the ability for any transputer to talk to four others directly. This means that transputers could be connected in a regular two-dimensional lattice. The limitations of such networks are that communications between trans- puters that are not neighbours will have to be provided by software, with in- termediate transputers acting as sorting offices which forward data in con- veniently sized packets to a transputer they can talk to but which is closer to the data’s destination. Being provided by software, this mechanism is considerably slower than a direct connection. Powerful workstation The problems faced in designing the supercomputer all relate to communi- cation, for this is the key to all successful parallel processor designs. Transputers can only have direct connections to four other devices although the sorting office analogy could provide a solution to the difficulty of providing other channels. The problem is that the more transputers are included in the system the slower communication between distant trans- puters becomes. In programming the supercomputer, what is ideally required is the ability to realize the direct connec- tions between transputers specified by the program. One of the key features of the Supernode supercomputer is the ability to realize this aim. This is provided by switching circuits on the links. Each transputer has its links connected into a switch through which they may be connected to any other transputer in the system. To pro- vide an alternative analogy, this is similar to telephones when each user (transputer) is connected to an exchange from which he can make a call for a given duration to any other free user connected to the exchange. In practice, it is expected that the com- puter will be used with all of the connec- tions established at a given time, pro- viding a pattern of communication or network that reflects the flow of data in the applications program. One of the disadvantages of using these switching circuits is the cost of the switch, which grows as the square of the number of inputs to the exchange. This cost function is avoided to some extent by designing the computer in modules, each with their own local exchange and of course with lines to other main ex- changes. A unit of about 30 transputers can be constructed economically in this way. This supertransputer, the Supernode, can stand alone as a powerful work- station, or can itself be used as the basis of a super-supernode. One transputer acts as a supervisor, setting the switches on request. It talks to all other trans- puters within the node by means of a control bus which is used to synchronise many transputers to a common event, so that the switches may be reset. The prospects A single Supernode can contain 32 worker transputers, each containing storage for 256,000 characters, a con- troller, a memory server with storage for 16 million characters, and a disk server with capacity limited only by disk drive technology. Such a node could perform up to 50 million floating point oper- ations per second, the rate usually ob- tained from machines such as the Cray 1, a multi-million pound supercomputer dating from 1976. A Supernode supercomputer could be manufactured for tens or perhaps hun- dreds of thousands of pounds, and a col- lection of 32 supernodes in a single supercomputer could produce a propor- tionally higher performance, in excess of 1000 million floating point operations per second. This is in the same league as today’s supercomputers which sell for about £12.5 million. The Supernode supercomputer could be marketed for a fraction of this cost. References: 1. Southampton University, Department of Electronics and Computer Science, Southampton, United Kingdom, S09 5NH. 2. INMOS Ltd, 1000 Aztec West, Almondsbury, Bristol, United Kingdom, BS12 4SQ. 3. Royal Signals and Radar Establish- ment, St Andrews Road, Malvern, Hereford and Worcester, United Kingdom, WR14 3PS. 4. Thorn EMI Research Laboratory Ltd, Dawley Road, Hayes, Middlesex, United Kingdom, UB3 1HH. non-inverting integrator A drawback of conventional integrator cir- cuits (figure a) is that the R-C junction is at virtual earth; this means that C appears as a capacitive load across the op-amp output, a fact that may adversely affect the stability and slew rate of the op-amp. Since the non- inverting character of an integrator is of minor importance in many applications the circuit shown in figure b offers a viable alternative to conventional arrangements. This integrator, unlike that in figure a, is non-inverting. The time constants R,C| and R2C2 should be equal. If both R! and C t , and R 2 and C 2 are transposed then the result is a non-inverting differentiator. For correct offset-compensation R, and R 2 should have the same value. 4.32 .MU, Ml. UNIPHASE LOUDSPEAKER SYSTEM A loudspeaker system that is based on Audax drive units and uses a 12dB Linkwitz filter. The closed box design enables the drive units to be located in a straight acoustical line. The use of a Linkwitz filter (Ref. 1) in a loudspeaker system makes sense only if the drive units are positioned in straight acoustical line. The three-way system presents nothing new, but the drive units have some special characteristics as will be seen later. Total costs for two loudspeakers (drive units, wood, filter components, etc.) is of the order of £250. Listening tests in which the uniphase system was compared with commercially available products show that the quality is roughly the same as that of a commercial system costing twice as much. The most noteworthy aspect of the box is the staggered front, which is essential to get the drive units in a straight acoustical line. This means that the drive units are positioned in a manner which ensures that the acoustical output of each of the three drivers reaches the listener at exactly the same time. It might be thought that to achieve this it is sufficient to measure the depth of each cone to be able to calculate by how much each drivers must be displaced with reference to one another. It’s a good start, but unfortunately not suf- ficient. This is because the phase behaviour of each drive unit is far from ideal — see Fig. 1 for a typical phase characteristic of a bass driver in a closed box. Although W. Marshall Leach pub- lished a very interesting article on the phase behaviour of drive units in the Journal of the AES as long ago as 1980, it appears that in the practical systems of most manufacturers no notice has been taken of the findings of Mr. Leach. Fig. 1. Prototype of the uniphase loud- speaker system. In an ideal loudspeaker system d^/dc o must be a constant to obtain optimum pulse behaviour. With reference to the curve of Fig. 1, it is seen that this is vir- tually impossible to attain. Any box that contains more than one drive unit and a cross-over filter will have a phase behav- iour that causes impulse distortion. Even in a wideband system without filter, it is very difficult to attain optimum impulse behaviour. Is phase shift audible? During the past few years, there have been a number of investigations into the question whether phase errors are aud- ible or not. These investigations have failed to agree. It is probable that the sensitivity to phase errors varies from one person to another. And what about the test methods? Our own experience shows that serious phase deviations can definitely be detected in the reproduced sound. Particularly the pronounced phase jumps around the cross-over point of the filter seem to be the culprits. These jumps also cause the loudspeakers to produce a different sound pattern at different positions around the room. This is because the acoustic radiation pattern around the cross-over points shows large variations along its axis. We have the impression — shared with a number of researchers — that most listeners are not not so much sensitive to absolute phase deviations, but rather to sudden phase differences. A matter of less than an inch Above resonance, the loudspeaker behaves capacitivcly at first and then, at higher frequencies, inductively. This behaviour is caused by the voice coil. 2 rx ] B" I GS GS — □ | C50 cs ra ntr i ' EJ cm m cm ,■*"** | ■*/«<>• © K; . — t4 Hi lt If : • - w a i 1 * S» c M022 front panel. using the soft tuning facility for manual selection of the A timebase sweep speed. When the A timebase was placed in automatic mode, a maximum reliable trigger frequency of 35 MHz was ob- tained, this being for the triggering of both the automatic facility and the timebase itself. A trigger holdoff facility is also provided, which is very effective in providing accurate triggering on a wide variety of waveforms. Its perform- ance was particularly good on pulse and digital waveforms, providing a stable trace under almost any alteration of the frequency of the waveform and over a very wide range of timebase speeds (in- cluding automatic). External triggering is also good with a typical sensitivity of 400 mV and a maximum bandwidth ap- proaching 40 MHz, allowing the exter- nal synchronization of most events. Probably the main asset of the M022 is its fully automatic main timebase, which enables very fast and easy operation of the scope when waveforms of a fairly constant amplitude, but not frequency, need to be measured. Timebase speeds range from 220 ms/div to 500 ns/div, or 50 ns/div if the xlO magnifier is brought into operation, this being covered in 18 steps, either by automatic or manual soft tuning via a continuously variable control. Timebase range indi- cation is by 11 green LEDs. This takes into account the 9 ’range’ indicators, which are calibrated in the standard 1-2- 5 sequence, as well as two scaling LEDs, which indicate whether the range in- dicators are calibrated in micro or milli seconds. Auto mode is indicated by a single red LED, timebase speed still being given by the remaining indicators. When in soft tune mode, the continu- ously variable control gives a linear response: it is very easy to set the desired timebase speed. A large amount of hysterisis is provided between the range switching thresholds, preventing any Table 16 other 20 kV tubes. unstable timcbasc speeds over the con- trol’s full range. Overall, I found that this enables a very accurate, easy and convenient way of selecting the sweep speed, combining the speed of a conven- tional switched timcbase speed control with the ease of use of a two-way ’up- down’ rocker switch, which can be found in some other scopes with this facility. The LED display also gave a clearer, and in my view easier, to read display than the conventional switched system, although it is perhaps not as easy to read as would be a numeric 7 segment display, examples of which can be found on other digital timebase scopes. The auto- matic timcbase facility itself is selected by placing the soft tuning control in a fully anticlockwise position. Operation of this was very effective, locking effec- tively onto a wide range of frequencies and producing a very stable display. The effective frequency range (4 cycles div- ision) is 4 Hz to 800 kHz, although ob- viously the actual frequency range is what can be expected from a normal analogue timcbasc. Timcbasc switching occurs at between two and six cycles/ 10 divisions, depending on the range, and whether the timebase is being scaled up or down. Switching between the ranges is very fast on most sweep speeds, but it was noticeable that there was a small time delay when changing down in sweep speed. This is not, however, clearly noticeable until changing down from, for example, 200 /vs/div to 2 ms/div, when a delay of approximately 1 second is present. This increases to about 3 seconds when switching to 200 ms/div, the slowest sweep range. These are ‘worst cases’ and typically may be considerably faster, depending largely on the waveform and previous speed setting. Performance of the autoranging system was good, locking onto a large range of waveforms, from a sinewave to a complex pulse train. Overall, I also found the automatic timebase an extremely useful facility with few failings, although as a very minor point it might have been helpful to have an automatic xlO deflection speed facility to increase the maximum speed under automatic control to 50 ns/div in place of the 500 ns/div for higher frequencies. The second timebase/deiayed sweep is of the ‘coarse’ type, in that in most cir- cumstances it is not possible to carry out accurate timing measurements in situ- ations where a calibrated delay time multiplier would normally be required. Waveform expansions can, however, be carried out accurately by the second timebase. For unstable, or changing, fre- quencies, it is obviously advisable to use the soft tuning option of the main timebase, although where this is not the case, the autorange option can be used with good results. The analogue second timebase sweep speeds range from 2 ms/div to 0.5 //s/div, and can be used in one of two modes, either intensifying the trace to be magnified, or as the magnified sweep. It is worth noting that in common with most other ‘coarse’ delayed sweep systems only one timebase can be displayed at a time, i.e., either A or B, and not both, thus giving a maxi- mum of two traces on the screen at any one time. Jitter is one part in 10,000, and as such is just visible on high magnifi- cation ratios, although it can be kept to a minimum in some situations by using the triggered delay facility. The delay time is variable by a uncalibrated control over 10 horizontal divisions, and as men- tioned above can either be continuous or triggered. CRT performance is obviously compar- able to the MO20, since the same CRT and drive circuitry are used. Typical magnification ratios in normal (artificial lighting) conditions were 1000:1 without the x 10 magnifier in operation, or, with ELECTRICAL CHARACTERISTICS te voltage: 100,220.240 VAC ±10% •nally adjustable, le frequency: 45-65 Hz ■wer consumption 35 Watts MECHANICAL CONSTRUCTION Dimensions: W 375 mm, H 160 mm, D 430 mm Housing: steel sheet Weight: approx. 8.5 kg AMPLIFIER ETC. Operating modes: CHI alone, CH2 alone Inversion capability on CHI or CH2 Dual CHI, CH2 (alternate or chopped (250 kHz)) CHI CH2 requency response: 0. . .20 MHz (-3 dB). iRisetime: £ 17.5 ns, (23.4 ns x 5 Mag.) Deflection factor: 12 steps: V/div ±3%, vernier control [adjusts max. sensitivity on 5 mV/div range t 1 2 mV/div (fully ccw) aut coupling: AC, DC or Gnd. aut impedance: 1 MO/25 pF ax input voltage 400 (DC + peak AC) Y MODE 11 X-axis and CH2 Y-axis. Less than 3° iase shift at 50 kHz [Bandwidth DC to 1 MHz (-3 dB). ating Modes: A; timebase s. B: A brightened in rang :ed: B; timebase B only aquence. Vernier controlj ( up to 2.5:1; o 20 ms/div. ± 3% in Delay jitter: 1/10000 TRIGGERING er modes: Auto (p-p). Normal Trigger coupling: AC.DC.HF reject, LF reject, TV frame and line(auto). Trigger sources: CHI, CH2, Line, Ext. Triggering sensitivity: Internal si div at Hz, External S0.5 V p-p at 20 MHz, [Normal mode MISCELLANEOUS |cRT-make: Valvo, measuring area 80 mm x tim, accelerating voltage 2 kV [Compensation signal for divider probe; ampli approx. 1 Vpp (±3%), frequency 1 kH: dulation Sensitivity: 3 V (complete Iblankingl Covered by 1 year warranty. this, approximately 100:1. For a 2 kV tube these figures are certainly above average, especially as the traces at these speeds were fairly well defined. Z modulation is provided as standard on the M022 and exhibited a very good sen- tilcklor inejia ap'il 1988 4.43 sitivity of +2 V for total blanking of a trace at maximum brightness. Negative voltages have no effect however, so the trace cannot be intensified by an external voltage. Internal construction, like that of the MO20, is based on two PCBs; the lower one, housing the Y-amplifiers, power supply, etc., is almost identical to that used in the MO20. The upper board is entirely different, however, and houses all the components for both timebases, as well as the trigger circuitry. Both PCBs are of fibre glass construction and screened with both component identifi- cation numbers, and, where appropriate, their functions. The PCBs are held in place by a steel chassis which should be extremely rugged under a very wide range of operating and environmental conditions. The outer casing is also steel; the absence of ventilation slots makes the M022 externally resilient to a range of conditions compatible with those out- lined above. The manual is very similar to that sup- plied with the MO20, the main differ- ence being a very brief explanation of delayed sweep operation CONCLUSION The Grundig M022 represents a con- siderable advance in terms of timebase technology for an oscilloscope in its price range. 1 found the automatic timebase facility genuinely useful, making the measurement of a wide range of waveforms significantly easier and faster. The second timebase provides a good range of sweep speeds and thus expansion ratios, although its use in delayed sweep applications is limited in some situations by the lack of a cali- brated sweep delay and delay line for triggered operation, and, of course, the 2 kV CRT. TheSc featues, however, can- not really be expected in this price range, especially when the high grade of con- struction and additional features are taken into consideration. This, along with the automatic timebase, should make the scope ideal for service work, particularly in the TV area as the TV triggering was very effective in both horizontal and vertical modes. To sum up the Grundig M022 is unique in its price range in having in fully automatic timebase. The advantage of this over a conventional analogue system obviously depends on the application, but from my own experiences 1 have found, it well worth while. This, coupled with the sec- ond timebase, should make the scope a good choice where ease of use and ver- satility are among the main re- quirements, such as in a servicing or educational environment. The Grundig M022 was supplied by Electronic Brokers Ltd., 140-146 Camden Street, London, NWI 3YP. Tel; 01-267 7070 Other scopes available under £1500 in the Grundig range. M053 Dual trace 50 MHz bandwidth. Full autoranging timebase with digital readout; trigger holdoff; TV line and field (I or 2) triggering: II kV CRT; calibrated delayed sweep facility. LOW-NOISE PREAMPLIFIER FOR EM RECEIVERS First in a short series of articles on simple to build RF preamplifiers is a tuneable aerial booster for the FM band. The RF preamplifier described here is in- tended for fining as close as possible to the FM band aerial. It is a tuneable, rather than a wideband, amplifier, which is fed and tuned via the downlead coax cable. The amplification and the noise figure of the FM aerial booster arc 25 dB and about I dB, respectively. All preamplifiers described in this series arc powered and tuned by a common supply/tuning unit installed at an appro- priate location in the home. Circuit description The circuit diagram of Fig. 1 shows that the preamplifier is a conventional design based on a VHF MOSFET tetrode Type BF981. The preamplifier input can be connected to unbalanced (60. . .75 Q) as well as balanced (240. . .300 Q) aerials or feeder systems. The balanced input 4.44 eletlo. India aprtl 1988 allows the preamplifier to be connected direct to the dipole clement. In this case, the preamplifier can take the place of the balun removed from the ABS, water- resistant, enclosure that houses the dipole terminals. This solution ensures the lowest possible input loss, and ob- viates the need for a separate preampli- fier enclosure on the mast. The balanced or unbalanced aerial signal is applied to winding a of tuneable inductor Li. Varactors Dt-D: form the adjustable capacitance across winding b, so that the tuning range of Li is about 86. . . 109 MHz. Gate 2 of DG MOSFET Ti is held at about +4 V by potential divider R:-R>. The bias voltage is effec- tively decoupled by surface mount ca- pacitor Ci. Blocking capacitor O takes the amplified RF signal from a half- impedance tap on drain inductor L 2 . The MOSFET is fed with a constant operating voltage of 12 V, supplied by regulatoi ICi. The direct voltage on the downlead coax can be varied between 15 V and about 26 V by means of the supply/tuning unit near the receiver. Zenerdiode Dj in the preamplifier en- sures that the tuning voltage for varac- tors D 1 -D 2 is the voltage on the downlead coax minus 12 V. Example: if the downlead coax carries +I8V, the tuning voltage at junction D 1 -D 2 is +6 V with respect to ground. The lowest downlead voltage is about 15 V to ensure the minimum voltage drop across regu- lator ICi. Choke L> forms a high im- pedance for the amplified RF signals superimposed on the luning/supply voltage. Construction Commence the construction of the pre- amplifier with winding the inductors. Input inductor Li is wound on the former in the Type 10V1 inductor as- sembly from Neosid— see Fig. 2. First, close-wind Lib as II turns, 00.6 mm enamelled copper wire. Study the com- ponent overlay on the PCB (see Fig. 3) to find the 2 pins on the base that con- nect to Lib. Close-wind Lu as 4 turns, 00.6 mm enamelled copper wire onto Lu., starting at the base of the plastic former. The tap is made after 2 wind- ings. Stretch the turns, and carefully scratch off a small area of the enamel coating approximately half-way the in- ductor. Solder a short wire to this point, and point it down towards the base. Press the inductor together again. Con- nect the end wires and the tap wire to the base pins, and verify continuity and orientation of the completed inductor. Drain inductor L2 is wound as 4 turns 00.3 mm enamelled copper wire through a small ferrite bead. The centre tap is made after two turns by twisting 2 Neosid 10V1 Fig. 2. The Type I0V1 inductor assembly from Neosid. 3 cm or so of the wire before making the third and fourth turn. The twisted wire is then cut to length, the enamel coating is scratched off, and the connection is carefully pre-tinned. Choke L3 is the simplest to make: it is wound as 6 turns 00.2 mm enamelled copper wire through a small ferrite bead. The three home-made inductors in the preamplifier are shown in the photo- graph of Fig. 4. The PCB for this project is a double- sided, but not through-plated, pre- tinned type. The four resistors are mounted upright. Ascertain the pinning of MOSFET Ti before fitting it on the printed circuit board: depending on the make of the device, it may be necessary to mount it with the type indication fac- ing the PCB. The ground terminal of R2, Ri, ICi, C2 , Cj, C4, Cs, the source terminal of Ti, the anode of D2, input pin 2, the 2 solder tabs on the shielding can of Li, and the output ground terminal, are soldered at both sides of the PCB. The only component at the track side of the board is SMD capacitor Ci. This is soldered direct across the gate 2 and source connections of the MOSFET. Fit a 15 mm high brass or tin metal screen with a small clearance for the MOSFET as shown on the component overlay. Supply/tuning unit The circuit diagram of the simple, regulated and adjustable, power supply for the downlead-powered preamplifiers is shown in Fig. 6. The output voltage of integrated regulator ICi is adjusted be- tween 15 V and 26 V with tuning control Pi. The tuning and supply voltage for each preamplifier is applied to the centre core of the respective downlead coax cable by a choke-resistor combination. The tuning/supply unit is built on the double-sided, pre-tinned printed circuit board shown in Fig. 7. Construction should not present problems: grounded component leads or terminals are soldered at both sides of the PCB. Fit ICi with a TO220-style heat-sink, but make sure that this is insulated from the ground area. ,4.45 Fig. 4. A close look at Ihe home-made inductors. Parts list FM BAND PREAMPLIFIER. CIRCUIT DIAGRAM: Resistors l±S%): Rt;R3= 100K R2 = 56K R4 = 1 0K Capacitors: Ct = InO (surface mount assembly) C2= InO ceramic (pitch: 5 mm) C3 = IjuO: 16 V: axial Ca= 1/rO; 40 V; axial Cs= 47/r; 35 V: axial C6 = 560p ceramic (pitch: 5 mm) Semiconductors: Di;D2 = BB405 D3 = zenerdiode 1 2 V; 400 mW ICt = 78L12 Ti=BF981 Inductors: Winding data are given in the text. Li = inductor assembly Type 10V1. Neosid part number: 15955100. (Neosid • Eduard House • Brownfields • Welwyn Garden City • Herts AL7 1 AN. Telephone: (0707) 325011. Telex: 25423). 2 off small, single-hole, ferrite beads (length: Suitable waterproof enclosure. PCi3 Type 880042 (see Readers Services page). 5 off soldering pins. The winding data for the 3 chokes on the board are as follows: Li;I_ 2 ;Lj = 6 turns 00.2 mm enamelled copper wire through a small ferrite bead (length: approx. 3 mm). The tuning control, Pi, is conveniently fitted onto the 3 soldering pins to go round a 3-wirc connection. The as- sembled board, the 24 VAC power trans- former, mains switch and fuse arc hous- ed in a small enclosure with a sloping front panel. Omit Di...Dj inch, and C1-C2, when a 24 V DC source, such as a mains adapter, is already available- connect this to the points marked + and 0. The tuning control can be fitted with a vernier and a scale for the frequency range of each preamplifier. Testing Connect the AC input of the completed tuning/supply unit to the secondary of the 24 VAC mains transformer, apply power, and verify the presence of the ad- justable direct voltage on the three DC/RF terminals. Check whether Pi sets the voltage between +15 V and +26 V. Fig. 5. Prototype of the completed low-noise Fig. 6. Circuit diagram of the tuning/supply unit. 4.46 PCBs & Set of COMPONENTS for projects are normally available with CORPORATION 52-C Proctor Road, Bombay-400 007 Phones: 367459/369478 Tline the FM receiver to a relatively weak transmission at about 108 MHz, and make a note of the signal strength. Con- nect the input of the completed pre- amplifier to the aerial, not a cable net- work outlet. The preamplifier output is connected to the appropriate soldering terminal on the tuning/supply board via a short length of coax cable. Similarly, connect the unbalanced (75 Q) input of the FM receiver to the RF (RX) output on the tuning/supply board. Set Pi to +26 V on the cable to the preamplifier. Verify the presence there of +12 V on C3, and + 14 V on Cs. Use a plastic trim tool to adjust the screw core in Li for optimum reception. Vary the supply voltage between 22 V and 26 V to check that this tunes the preamplifier. Set the supply to +15 V, and tune the re- ceiver to a signal at the lower band edge, i.e., approximately 88 MHz. Check that Li is still adjusted for optimum recep- tion by carefully adjusting the core. Tunc to a number of stations at regular fre- quency intervals in the FM band, op- timize reception by adjusting Pi in the tuning/supply unit, and make notes of the downlead voltage. If necessary, redo the adjustment of Li to ensure that the span of the tuning voltage covers the en- tire FM band. For optimum tracking of the resonance frequency with the tuning voltage applied to the varactors, the core in Li should be turned to about half- way the aerial winding. This completes the initial adjustment of the FM band preamplifier, which is ready for fitting into a waterproof enclosure. The next instalment in this series will deal with preamplifiers for the short- wave, VHF and UHF TV bands. B s4.47 sinewave generator A sinewave generator is a virtually indispensable tool for anyone engaged in the testing or measuring of electronic equipment. It is commonly used when measuring the frequency response or dis- tortion characteristics of audio equip- ment. In particular, harmonic distortion is still considered to be one of the im- portant parameters in performance of audio amplifiers, and in order to measure this accurately, it is obviously imperative that the input test signal itself have as little distortion as possible. In fact the distortion of the input sinewave must be at least an order lower than that intro- duced by the amplifier. Furthermore, it is important that the frequency of the sinewave be extremely stable, if one is to avoid having to constantly retunc the notch filter in the distortion meter (see the circuit for a distortion meter pub- lished in Llektor 27/28, July/August 1977), The amplitude stability of the sinewave is of secondary importance in distortion measurements, however it is often a critical factor in a number of other test applications. of commercially available, continuously tunable sinewave generators of high quality can be counted on the fingers of The basic problem with continuously adjustable sinewave generators is ampli- tude instability. In almost every case, the sinewave output is produced by an oscillator circuit. 0 ) An oscillator is essentially an amplifier with positive feedback, whereby the feedback loop contains suitable frequency-selective net- works of capacitors and resistors. In the example of the Wien bridge oscillator shown in figure 1 , positive feedback is applied via the RC network to the non- inverting input of the op-amp, whilst negative feedback is applied to the inverting input via the voltage divider network formed by R 0 and the negative temperature coefficient resistor (ther- mistor). If the negative feedback exceeds the positive feedback the oscillations will not be sustained and the output of the amplifier will fall; if the positive feedback predominates, however, the output of the amplifier will rise until the latter There are a number of measurement jobs which require an AC test signal, which, as nearly as possible, is a perfect sinewave. Not only must the amplitude of the signal be absolutely stable, but the hum, noise and harmonic distortion components must be reduced to a minimum. The spot frequency sinewave generator described here will provide a sinewave output with harmonic distortion of less than 0.0025% and whose amplitude is constant to within 0.1%. Continuous or 'spot' If all three of the above-mentioned demands on a sinewave generator, viz. amplitude stability, constant frequency, and extremely low distortion, arc to be satisfied, then unfortunately it more or less precludes the use of a sinewave generator with continuously adjustable frequency. It is true that such devices do exist, however they are exceedingly complex and expensive, and the number 1 saturates. The circuit is prevented from lapsing into either of these two con- ditions by the thermistor, which stabil- ises the output , amplitude as follows: should the output voltage rise, the current through the thermistor will increase, causing its temperature to rise and hence its resistance to fall. This causes an increase in the proportion of negative feedback, thereby automati- cally reducing the gain of the op-amp. The opposite, occurs when the output voltage tends to fall; the resistance of the thermistor is reduced since it dissi- pates less heat, thus also reducing the amount of negative feedback. Assuming that the resistor and capacitor values in the two arms of the bridge arc identical, the proportion of output voltage which is fed back round the positive feedback loop at the resonant frequency, f„, of the oscillator is 1/3. The output voltage of the oscillator settles at the value which ensures that the resistance of the NTC resistor is equal to 2R 0 . It is obvious that the frequency of the oscillator could be continuously adjusted by using a stereo potentiometer or twin-ganged trimmer capacitor to vary the RC time constants in the arms of the bridge. However, in practice it is impossible to obtain stereo pots or trimmers in which both gangs are perfectly matched. Variations in the resistance or capacitance values between the two arms of the bridge have the effect of altering the positive feedback factor, k, the result of which is a change in the resistance value of the thermistor ,4.49 4 r l > 600 a (see figure 1). Thus varying the fre- quency of the oscillator has the effect of also varying the amplitude of the output signal. What is more, the ampli- tude of the output signal at the new frequency (after the balance between positive- and negative feedback has been re-established ) differs from that obtained before the change in frequency. The op-amp is not the only source of distortion in the sinewave output (this can be counteracted by a high open- loop gain); a further contributory factor is the fact that the voltage-current transfer characteristic of the thermistor is not completely linear. Other ampli- tude-stabilising components such as filament lamps, diode-resistor networks or voltage-controlled FETs can be used, but these are also by no means perfect. For a large number of applications the above-mentioned failings are not par- ticularly critical; however, for measure- ment purposes where accuracy is important, they represent an unaccept- able source of error. For this reason, the most common solution is to do without the admittedly attractive facility of continuously adjustable frequency an I to settle instead for an oscillator with . number of switched output frequencies. Basically this amounts to a series of individual oscillators each designed to produce a single optimal frequency. This elegantly solves the problem of amplitude stability which bedevils con- tinuously variable oscillators. If one considers that a high-quality continu- ously variable sinewave generator will cost in the region of £ 500 - £ 600, whilst a ( simple ) spot frequency sinewave oscillator on the other hand can be constructed for under £ 10, and further- more, that only four or five test fre- quencies are normally required in harmonic distortion measurements, then it is clear that a spot frequency generator represents a highly cost-effective ap- proach. The distortion meter published in the Summer Circuits 1977 is also designed for spot frequency measure- Spot sinewave generator The basic principle of the spot sinewave generator described here should be familiar to a number of readers, since it was used in the design for a simple spot sinewave generator published in last year’s Summer Circuits issue ( circuit 25). The operation of the circuit is illus- trated by the block diagram shown in figure 2. A symmetrical squarewave signal is fed to a number of cascaded selective filters (in figure 2 two such filters are used). These remove the harmonic content of the squarewave, leaving the more or less pure sinusoidal fundamental. The resulting sinewave is in turn used to trigger the squarewave from which it is derived. The amplitude of the sinewave is clipped to ± u, before being fed back to the input of the squarewave oscillator, so that the oscillations arc maintained. For this in fact to happen, two conditions must be fulfilled: the input- and output signals must be in phase; this means that the phase shift of the selective filters must be either 0°, 360° or a multiple of 360° (the phase shift introduced by the clipping circuit’ can be neglected). Secondly, the loop gain of the system at the oscillator frequency, f osc , must be greater than 1 . The former is the product of the gain of the clipping circuit plus that of the selective filters, and any damping introduced by an attenuator which may be included in the system. In figure 2 the centre frequencies of the two selective filters are identical, hence fosc = fo The output signal of the clipping circuit is not a perfect squarewave, since it does not have an infinite gain. Strictly speaking the output is a clipped sinewave, which has more in common with a symmetrical trapezoidal waveform. This is all to the good, however, since this type of waveform has fewer harmonics to filter out than a perfect squarewave. Figure 3a shows the amplitude response curve of the type of selective filter employed in the circuit, whilst in figure 3b we see the phase response of the filter. The overall response of a number of filters connected in cascade can be obtained by adding each point of the separate response curves for each filter. The resonant frequency of the system is that at which the combined phase response curve intersects the x-axis. With two selective filters whose centre frequencies, f 0 i and f 02 , are offset slightly, the resonant frequency f osc will equal Vfoi ' fo2 • The amplitude 6 values shown in figure 2 assume'that the output signal of the limiter is a perfect squarewave and that the resonant gain of each filter is 2. The harmonic sup- pression of the filters is discussed in Appendix 2 at the end of the article. Practical design The block diagram of the full spot sinewave generator is shown in figure 4, whilst figure 6 contains the complete circuit diagram. In contrast to figure 2, the block diagram of figure 4 contains a variable attenuator (in the shape of a potentiometer), a lowpass filter and an output buffer stage. In addition to varying the amplitude of the output signal, the potentiometer fulfils a second function. Without some kind of signal level control at this stage there is the danger that an excessively large input signal would overload the filters, causing their output to clip. The output buffer stage ensures that, even under heavy load conditions, the generator can provide a low distortion output signal. It is an obvious step to combine the output buffer with an 18 dB per octave lowpass filter - all that is needed is three extra resistors and capacitors. If the turnover fre- quency of the filter is calculated to roughly coincide with the oscillator frequency, the result is further sup- pression of harmonics without incurring too great a voltage loss or significantly affecting the amplitude stability of the output signal. This latter point may require further explanation: see figure 5. If one assuifies that the oscillator frequency can vary by a factor of - A fosc (the frequency stability is then — x 100%), then the amplitude of the output signal of the lowpass filter can vary by ± A A; the result is that in addition to variations in amplitude caused by the oscillator itself, the amplitude of the sinewave generator output can be affected by variations in the output of the lowpass filter caused by frequency drift. Fortunately, in view of the extreme stability of the oscillator and the relatively gradual roll-off in the lowpass filter's response at the 3 dB point, this effect is of little practical importance. The detailed circuit dia- gram of the spot sinewave generator is shown in figure 6. The clipping circuit is built round 1C1, (which has a gain of 11) R3. and T1 and T2, which are connected as sym- metrical zener diodes. The trapezoidal voltage at the junction of R3 and R4 is attenuated by R4 and P 1 , and fed to the first selective filter consisting of IC2, IC3, R5 . . . R9, Cl and C2. The second bandpass filter (1C4, 1C5, RIO . . . R14, C3, C4) is identical to the first; a more detailed discussion of these filters is contained in Appendix 1 at the end of the article. The frequency-determining components of the lowpass filter are R15. R16, R17. C5, C6 and C7, whilst IC6 is the associ- ated emitter follower, which also functions as output buffer. If desired, a symmetrical emitter follower (T3 . . .T6 etc.) can be connected to the output of IC6, allowing the generator to be used with load impedances as low as 47 £2. If load impedances as low as this are not foreseen, the emitter follower com- ponents can be omitted, points A and B are linked together and outputs I and II can be used with impedances of 600 £2 or greater. The frequency of the oscillator is determined by the choice of component values for Cl . . ,C7: Cl =C2=C3=C4 = ^^: Capacitances are in nanofarads, the' oscillator frequency is in kHz. Construction Figures 7 and 8 show the copper track pattern and component overlay respect- ively of the p.c.b. for the 47 £2 version of the spot sinewave generator. Figure 9 shows the component layout for the version without the emitter follower output buffer (into 600 £2 or above). As far as the choice of component values are concerned, the values given for R8. R9, R13 and R14 are nominally 33 k: possible alterations to these values are discussed in the following section describing the calibration procedure. The values of R6. R7. Rll and R12 j4.51 should be as closely matched as possible, response curves of a number of selective first time). The extent of the phase shift The best procedure is to measure their filters of differing centre frequency is a measure of the difference between resistance, but in practice it is sufficient whilst figure 10b shows three different the centre frequencies of the selective to take four successive resistors from response curves obtained: (1 ) when two filters. Thus the centre frequency of one the 'belt' in which they are packaged, filters with the response of curve I in or both filters should be adjusted until Although desirable, 1 or 2 % metal-oxide figure 10a are connected in cascade (i.e. the two signals are as nearly as possible types are not absolutely necessary. The both filters have the same centre frc- in phase; at the same time the amplitude values of Cl . . . C7 are calculated from quency); 1 2) when thecentre frequencies of the sinewave at the output of 1C4 the equations listed above. Room has of the two filters are slightly offset, as is should rise. The adjustments are realised been provided on the p.c.b. to make up the case with curves 2 in figure 10a; (3) by altering thevalueof one or more of re- the correct values by connecting two anc * when the centre frequencies of sistors R8, R9, R13and R14(see Appen- capacitors in parallel. Cl . . . C4 should > he ‘ wo filters are fair| y far a P a " D- Each resistor can be varied be- also be as closely matched as is possible, (curves 3 in figure 1 0a). The Q and res- tween 22 k and 68 k. Of course it is also If there are discrepancies in the values onant gain, A, of all the filters in possible to vary the value of other fre- of Cl C4 or R6. R7. Rll and RI2, figure 10a are identical. It is apparent quency-determining components (again it may slightly affect the quality of the that the greater the difference in the see Appendix 1). Once the frequencies output signal. However this can be centre frequencies of the two filters, the of the selective filters have been aligned rectified during the calibration pro- smaller the gain at the resonant fre- as accurately as is possible, the resist - cedure, which is described next. quency (it may even fall to the point ance bridge across R1 can be removed, where the loop gain of the system is less As described above, the more accurate than 1; see also Appendix 3), and also tuning of the two filters will have the the less filtering of higher frequencies effect of increasing the resonant gain of Calibration there is - i.e. less suppression of the the system; if as a result of this the An oscilloscope is a prerequisite for higher harmonics. output of one or both filters should correct calibration of the sinewave One should thus attempt to ensure that start clipping, PI should be adjusted generator. After the usual checks the the centre frequencies of the two until the loop gain is at a satisfactory generator is connected to the oscillo- bandpass filters are as close as possible, level. The calibration procedure is then scope and the power switched on. The at least enough to ensure that the complete, wiper of PI should be turned fully oscillator starts, towards R4, whereupon, hopefully, a If, during the calibration procedure, the sinewave signal should appear on the oscillator should initially refuse to start. In conclusion screen. If, however, nothing happens, the loop gain of the system should be The spot sinewave generator requires a then the circuit is failing to oscillate, a temporarily increased by bridging R I symmetrical stabilised supply of ± 15 V. state of affairs which is almost certainly with a resister of a couple of hundred The current consumption per oscillator due to the fact that the centre frc- Ohms. As soon as the oscillator starts, is a maximum of 50 mA for the 600 H quencies of the two selective filters the output signals of both bandpass version and 150 mA for the 47 S2 are too far apart, with the result that fillers should be displayed on the scope, version. The quiescent current of the the loop gain at the resonant frequency The signals at pin 6 of IC2 and 1( 4 will output stage of the latter.should be set is less than 1. The first thing to do, almost certainly exhibit a considerable to 1 00 mA using P2. The lower the therefore is tune in the frequencies of phase shift (if there was only a small amplitude of the output signal, the less these filters. Figure 10a shows the shift the oscillator would have started harmonic distortion. Thus the size ol 4.52 elrtlor indl» op'll .388 the output signal can be adjusted as desired by means of P 1 . There are two conditions attached to using PI as an amplitude control however: it should be set neither too high as to allow clipping to occur, nor too low as to cause the oscillator to stop. It is also possible to omit PI altogether. R4 and R5 are then joined and between this junction and earth a resistance of suitable value is inserted. In nine out of ten cases the value of a simple carbon resistor will prove stabler than that obtained using a potentiometer, the above step can there- fore only improve the overall amplitude stability of the generator. If several oscillator frequencies are required, then, in order to keep the com- ponent count down it would be logical to use a 9-pole switch (for Cl . . .C 7 and PI) with however many ways as one requires different frequencies. Although this represents the most elegant solution, whether it is the cheapest is another question. Spot sinewave generators are of course most commonly used in AF applications, however the model described here can also be used for high frequency work. It was with an eye to this type of appli- cation that the 50 SI output was pro- vided. Unless one possesses a tunable two-tone generator, measuring the inter- modulation distortion of r.f. amplifiers can be a difficult business. The two-tone generator produces a pair of signals of identical amplitude but differing fre- quency. If one feeds the output of the spot sinewave generator to a double- balanced mixer (DBM) (see figure 11) one obtains two output signals whose frequencies differ by twice the fre- quency of the original input signal. Of particular interest are the uneven harmonic distortion components, since their frequencies lie in the region of the desired signals. The IM distortion of the two-tone generator itself must be less than —60 dB for reliable measurement purposes - a specification which the spot sinewave generator easily improves Bibliography: 1. Spot frequency sinewave gener- ator; Elektor 27/28, July /August 1977. 2. Klein and Zaalberg van Zelst, A non-linear low output im- pedance AF oscillator with ex- tremely low distortion. Philips Technical Journal. 25.20.1963. Appendix 1. It can be shown that the centre frequency fo. the resonant gain. A. and the Q of the selective filter formed by 1C 2 and IC3 in figure 2 can be determined as follows: v R9 C2 R6- R7 If Cl = C2 = C, R8 = R9, RS = Rq and R6 = R7 = R, then: These equations are also true for the second filter (round IC4 and IC5). It is apparent from the expression for f 0 that (small) variations in centre frequencies of the two filters can be obtained by varying the value of one Or more of resistors R8, R9, R13 and RI4. 2. As far as the amplitude response of the selective filters used in this circuit is concerned, it can be shown that: input voltage and u 0 the output voltage of the filter, and n = ^ — . If the Q of the filter is sufficiently high, the above expression can be simplified to: EEo_- — EL_ uj 2 (n 2 - 1) Q for n > 1 A symmetrical squarewave contains exclusively odd harmonics (this is in addition to the fundamental, which ’ is - x the amplitude of the squarewave), i.e. n = 3, S, 7 etc. ■ The amplitude of the n-th harmonic is - x the fundamental. The ampli- tude of the third harmonic of a symmetrical squarewave is therefore 33 1/3% that of the fundamental, the fifth harmonic is 20% of the fundamental, the seventh harmonic is approx. 14%,. . . and so on. The Q of the filters shown in figure 2 is approx. 55. If the centre frequencies f 0 i and foj of the two filters are identical (and equal to the resona nt frequency, f osc = V foi ‘ foj). then a single filter will suppress the third harmonic by a factor of 146, the fifth harmonic by a factor of 264, and so on. With two filters connected in cascade, these factors should be squared'. In actual fact the filters are fed not with a perfect squarewave, but with a trapezoidal waveform, whose har- monics are less pronounced than those of a squarewave. 3. It can be shown that, with two bandpass filters connected in cas- cade, which have resonant fre- quencies of f 0 i and fo2 , respect- ively, but which have the same resonant gain and quality factor, Q, that, at the frequency \J f 0 | • f 02 , where f 02 > f 0 , , the gain will fall by a fact or of 1 + (Q — r — ), where x^\J ~S 1 .1 If, as a result of component toler- ances, f 0 i and f 0 2 vary from one another by more than 10% (x * 1.05, x 2 = 1 .1), and if Q = 55, then the gain of the two filters at the oscillator frequency will be reduced by a factor of 28.4. For this reason it is important that, as far as possible, one should attempt to match the components used in the two filters. M UNIVERSAL MULTIPLEXER A fast, analogue and digital compatible, 16-channel multiplexer with provisions for manual and computer control. The circuit is offered as a design idea, and should find applications in test, measure- ment and instrumentation equipment. The circuit described here is essentially an electronic 16-way rotary switch. It is composed of 2 functional sections: one takes care of the connection between the selected input channel and the ’’pole” of the 16-way switch', i.c., the output of the circuit, while the other provides the con- trol signals necessary to select a par- ticular channel from the 16 available. The control section accepts manual as well as computer or automatically gener- ated channel selection codes. Appli- cations of the universal multiplexer in- clude quasi-simultaneous temperature measurement in a network of sensors mounted in different locations, con- trolled capturing of signals from strain gauges, light sensors or transducers, and the routeing of command signals and voltages in automated test, measurement and production systems. Manual or automatic control In the manual mode, the desired channel is selected by the user pressing the chan- nel increment key as many times as re- quired. In the automatic mode, an oscil- lator provides the channel increment pulses. With reference to the circuit diagram of Fig. 1, the user selects be- tween manual and automatic channel in- crement pulses with the aid of toggle switch Si. This supplies a clock pulse to bistable FFi via set-reset bistable ICiib, whose Q and Q outputs toggle on each rising edge of the CLK signal. Thus, each time Si is pressed, the multiplexer changes between manual and automatic channel control, or vice versa. LEDs Do and Dis indicate the currently selec- ted mode. On power up, network Ri-Ci at the SET input of FFi selects manual channel increment pulses. These are generated by S: and S-R bistable ICiia, which functions as a debounce circuit. AND gate N«, blocks the manual chan- nel increment pulses when the unit is set to the automatic mode. Similarly, Nm prevents the channel increment pulses from oscillator IO being applied to pin 1 of N42 in the manual mode. The oscil- lator pulses are only used in the auto- matic channel increment mode, that is, when Q of FFi is logic high. The output frequency of the oscillator set up around IC4 is adjustable to enable the channel increment speed being set as required for the application ‘in question. The channel incement pulses are not routed direct to the switching section: AND gate Nis effectively blocks them when the microprocessor or micro com- puter holds circuit input //P/MAN logic high. In that case, LED D19 lights to in- dicate that the increment pulses originate from the computer, and are ap- plied to the CLK input of binary counter ICi via Nn and N*. The clock pulses re- ceived by ICi increment the 4-bit binary value at outputs QA...QD. LED D20 indicates the presence and the relative speed of the received or internally generated clockpulses. Computer control So far, the circuit description suggests that channel selection in the multiplexer is sequential and unidirectional. This means that if, for example, channel 3 is currently selected, the next channel can only be number 4, making it is imposs- ible to step back to, say, channel 2, or on to 5 with a single clock pulse. This restriction was found inacceptable, so that the circuit was extended to enable the direct selection of any 1 of the 16 channels via input lines DO. . .D3, which control counter inputs A . . . D direct. When the computer’s output port ap- plic s a logic high level to inputs ^P/MAN and LOAD, pin 1 of ICi goes high, so that the binary value on D0...D3 is transferred to QA...QD. The control of input lines D0...D3, //P/MAN and LOAD is within every pro- grammer’s reach when the appropriate data is sent to the computer’s parallel output port (c.g. the Centronics outlet) with the aid of a simple BASIC program or (machine language) subroutine. The use of computer control on the pro- posed multiplexer makes it possible to activate any 1 of 16 (2 4 ) channels at any time. This is in contrast to the sequential and unidirectional channel selection in the manual (automatic or switch- controlled) mode. The 4 DIP switches marked RESET in the circuit diagram determine the last (highest) channel that can be activated. Four-bit comparator ICj compares the number of the selected channel to the configuration of the DIP switches, i.e., to the number of the channel defined as the last one. Output A = B (signal RS) goes high when equal channel numbers are applied to the An and Bn inputs of IC3. The LOAD input of counter ICi is activated, and inputs A. . .D read 0000 thanks to pull-down resistors R2 . . . Rs inch This resets the counter to output state nought. The current channel number is indicated by 1 of 16 LEDs selected by 4-to-16 decoder IC2. The channel code is also applied to display driver IC10, which ar- ranges for the decimal channel number to be shown on a 1 'A -digit common cathode display. It should be noted that binary input 0000 on the SAB32U >4.55 Fig. I. Circuit diagram of the universal multiplexer. causes the display to read ”16”. The Relays or electronic switches SO that inverters Type 4049 are required type SAB3211 is manufactured by instead of non-inverting buffers Type Siemens, and may be a difficult to ob- Some applications of the multiplexer call 4050. Do not forget to fit a protective di- tain component. The more familiar Type for the use of relays rather than elec- ode across each relay coil as shown. The 9368 is suggested as a suitable alterna- tronic switches in IQ. . .IC 9 inch The R-C filters on the input lines may not be tive, but it should be noted that this modifications to the circuit to enable op- required in all cases, but are rec- causes the channel numbers to be eration with relays are shown in Fig. 2. ommended as a protective measure displayed in hexadecimal (0. . .F) rather The Type 4514 decoder is replaced by a against crosstalk and switching noise, than decimal. 74HC154, whose outputs are active low, . Electronic switches not only consume 4.56 eleUo* India april 19SS *seetexl 87159 -2a Fig. 2. Modifications to the multiplexer to enable the use of relays in the switching stage. little power, they also have the advantage of being fast, silent, small, and inexpens- ive. They do not, however, allow the safe use of different potentials in the channel selection circuit (i.e., the multiplexer) and the circuit(s) drivingthe input lines, in applications where this is expected to cause problems, it is recommended to ensure adequate insulation through the use of relays. The multiplexed signal is buffered in op- erational amplifier ICs. This guarantees light loading of the selected channel, and a relatively low output impedance for driving a wide variety of test equip- The analogue switches may be protected against static discharges by fitting 2 small-signal diodes on each input line. One diode is connected with its cathode to ground, and its anode to the input line; the other with its cathode to the positive supply voltage, and its anode to the input line. Depending on the type of relay used, and the multiplexing speed, the inertia of the contacts may give rise to er- roneous measurements owing to brief short-circuits between input channels. This is prevented by delay network R27 - Rjs-Ci. freezing the output state of decoder IC2 for about 10 ms in between 2 clock pulses. This ensures that the cur- rently energized relay has enough time to complete opening its contacts before the next channel is selected. The multiplexer can be fed from any supply voltage between 5 and 16 V, pro- vided the series resistors for the LEDs and the display are dimensioned accord- ingly. Relays, as well as the 74HC154, re- quire a supply of 5 V. Finally, it should be noted that the cir- cuit described is experimental: it can be extended as well as simplified to in- dividual needs. The RESET configur- ation around ICs can be simplified by hard wiring the relevant inputs An on the comparator; the IZ 2 digit read-out can be set up around a display driver other than the SAB3211; the LEDs and associated driver ICs are optional; the number of channels can be reduced, and the gates used for the microprocessor in- terface may be omitted if computer con- trol is not envisaged. room 'thermometer Using a National LM391 1 IC, a 1 mA meter and a few resistors it is a simple matter to construct a thermometer to measure over the temperature range -20° to +50°C, which should be adequate for all but polar climates! As the circuit is intended as a room thermometer the entire circuit operates at the temperature which is being measured, so the resistors used should be low-temperature coefficient types to maintain the accuracy of the circuit. To calibrate the thermometer the meter scale must first be marked out linearly from zero = -20° to full-scale = +50°. With P2 set to its mid-position the circuit should be placed in a freezer or the freezing, compart- ment of a refrigerator set to -20 C and PI should be adjusted until the meter reads -20. The circuit should then be placed in a temperature of +50°C and P2 adjusted until the meter reads 50. Of course it is also poss- ible to mark out the scale from 0°F to 120°F and calibrate zero and full-scale accordingly. PI and P2 interact to a small extent, so it may be necessary to repeat the procedure several times until both the —20 and +50 readings are accurate. As the IC contains its own stabiliser the R1 = (kS2). i4.57 selex-33 THE MEGAPHONE Whether it is a sports event, or a police operation, or a demonstration, or a rally-the small funnel shaped gadget can be noticed, the MEGAPHONE. The purpose of the megaphone is to make the human voice audible to a large crowd, upto a large distance. It replaces the public address system when importance. The sound quality is of secondary importance, as long expected to deliver Hi-Fi quality The circuit provided here is designed to deliver-maximum possible power, without much attention to the frequency response and Hi-Fi quality. It will be very useful during your next bicycle tour or family picnic, or even the football match How Does It Work? Let's have a quick look at the extreme left bottom corner, we have the microphone. Then the first stage of amplification- the transistor T1 in common emitter configuration. Its collector current flows from the plus pole of the power supply, over the loudspeaker. R5. D1 and D2. R5 is the collector resistance, which converts the current amplification to voltage amplification. The collector voltage becomes the base voltage of the next transistor T3. The base voltage of T2 is just shifted by about 1 .4 Volts by the two diodes D1 and D2. This is due to the fact that each diode drops about 0.7V. Thus the base voltage ofT2 is 1.4V higher than the base voltage of T3, T2 T3 form the output amplifier stage. Both the emitters of T2 and T3 lie on approximately half the supply voltage; which corresponds to the zero signal level. If the input is positive, then transistor T2 conducts and pulls the emitter terminal high. T3 remains blocked. When the input voltage goes negative. T3 conducts and T2 is blocked. In effect, what happens is that the negative parts of the input waveform are amplified by T3 and the positive parts of the input waveform are amplified by T2. The common emitter terminal of T2,and T3 is thus pulled high and low in the As this is connected to the supply voltage through the capacitor C2 current passes through the loud speaker, causing the input signal to be amplified and made audible through the loud speaker. In technical language, this circuit is called the single ended push pull output stage. T2 and T3 are operating alternatively and the common emitter terminal is alternatively pulled up or pushed down. Hence the name 'Push- Pull' stage. The two diodes D1 and D2 are very important for proper operation of the output stage. resulted i and D2 ir i that would have absence of diodes D' Figure 3: Many things which are requirec to build the megaphone can be found in the junk room. The things required are as follows: 1. Measuring Beaker 2. Loud Speaker 3. Plastic Pipe 4. Aluminium Clamp. 5. Switch. 6. PCB 7. Mounting Disc 8. Electret Microphone s of D1 and D2 solve by shifting the common emitter terminal equal to 4.6 V. approximately half of the supply voltage. This voltage is stabilised by T1. On the loud speaker, we have the AC voltage due to the coupling capacitor C2. This can swing between 0 to 9V depending on how strong the input signal is. It we connect on oscilloscope, we would effectively see that the voltage swing is between 4.5V at the common inals of T2 and T3 way of the supply s is ensured by the i R3 and R4. Theratic resistors is 180:27, le as 4:0.6 We know shold voltage of T1 is of this fact by connecting the collector resistance of T1 through 2. The PCB and battery are secured to the disc from i Part list: R1 = 22 K n R2 - 4.7 K f! R3 = 180Kf!R4 = 27K!IR5 = 1.8Kfl Cl =2,2 jiF/1 OV C2 = 1 00 /xF/1 OV C3 = 470 pF/IOV D1, D2- IN 4148 T1 = BC 549C or BC 550C T2 = BD 139 -10 T3 « BD 140-10 Other Parts : SI =On/Off toggle switch 1 Electret microphone 1 Loud Speaker 8flf0.5 W 1 SELEX PCB size 1 Various parts for construction. This gives a larger amplification for T1, and makes our circuit .more efficient. For the microphone, an electret microphone is used, which obtains its supply over the resistor R1. In case a carbon microphone from a telephone handset is to be used, R1 should equal the DC resistance of the microphone. R2 decides the volume level. The output transistors must be of the same specified number, that is BD 139-10 and BD 140-10, to get good results. 6. Solder the speaker wire the speaker and fix the speaker firmly into the measuring beaker. TIPS FOR CONSTRUCTION: Figure 3 shows a view of how the Megaphone can be constructed. A measuring beaker is used as the horn of the megaphone. (The manufacturer of the beaker would have never dreamt that it j cessary to follow these legaphone. legaphor for larger batteries a connecting cortf for i legaphone the simplest Figur possible constructic always be made m< professional. As usi is assembled on SE layout of figure 4. Saw the disc (7) and fix the microphone. The thickness of the disc should be about 5 mm. This will allow the fixing screws to take a firm grip. The Digilex-PCB is available! Price: Rs. 85.00 + Maharashtra Sales Tax. Delivery charges extra: Rs. 6.00 Send full amount by DD/MO/PO. Available from: precious ° ELECTRONICS CORPORATION Journal Division 52-C, Proctor Road. Bombay-400 007 Ph. 367459. 369478 Telex - (01 1) 76661 ELEK IN NEW PRODUCTS • NEW PRODUCTS • N Buzzer This buzzer introduced by ION Electri- cals is an audio transducer, driven by an external oscillator. When driven by a 9 volts dc peak to peak square wave with 1 KHz frequency, IPB27M will produce an audible tone with sound pressure of at least 75 dB at a distance of 30 cm. The frequency response is fairly even over the range 200 Hz to 4 kHz. The transducer is 27 mm in diameter and 5 mm in height. It is mounted on a plastic base measuring 30 x 30 mm, The IBP27M is ideally suited for small gadgets. Since the power required to Typical applications of IPB27M are: Electronic Toys, Hobby Projects, Melody Generators, Metronomes, Tele- phone Tone Caller, Tweeters etc. PROMOTION, • Blk # 4, Fir # 1, • 10, Subhash Cross Lane, • Bombay 400 057. STRIP CHART RECORDER With 32 channels every two seconds scan frequency, MOLYTEK'S 2702 is a fast multipoint strip chart recorder. The state of the art microprocessor technology has been used to pack a host of useful fea- tures into its compact size. The auto calibrating and auto scaling 2702 can ac- cept any combination of thermocouple, RTD, voltage and current inputs. Specific type of inputs can be assigned to individual channels thru the keyboard. The thermal paper printing coupled with dot filling features provides sharp print- ing. Various parameters like chart speed, margins, channel information, alarm setpoints, relay logic, math func- tions etc. are all field programmable thru a typewriter like slide-in keyboard. These parameters are retained in battery backed-up memory during the power down conditions. A battery backed up real time time clock keeps precise time and date for accurate records. 2702 comes with a standard RS-232 computer interface. Software is available to net- work upto 16 of 2702 units with one IBM-PC/XT giving a full fledged distri- buted data acquisition system. JELTRON INSTRUMENTS (I) PVT.. LTD., • 6-3-190/2, Road No. 1, • Banjara Hills • Hyderabad 500 034. • Tel: 222411 Power Supply Puneet Model DC-22 is an accurate and stable DC Power source with output var- ying from 0 to 30 V and current from 0 upto 20 Amps. It incorporates continu- ous current and voltage limiting. It has short circuit protection. This is a dual output supply which can be used in inde- pendent mode or tracking mode. Regu- lated power supplies with specific output voltage and current values are also made against specific enquiries. • H-230, Ansa Industrial Estate, • Saki-Vihar Road, • BOMBAY 400 072. Milliohm Meter Economy offers a digital Milliohmmeter with a 3‘A digit, 7 segment RED LED display , 5 ranges with lowest range of 200 milli ohm with 0:1 milli ohm resolution and highest range of 2 K ohm with 1 ohm resolution. Other ranges are 2 ohm, 20 ohm, 200 ohm. Selection is by inter- locked push button switches. 4 wire mea- surement avoids lead resistance error. Accuracy of ± .05% of range ± 0.1% of reading ± 1 digit is reported. Cabinet size: 247 mm Depth x 195 mm Width x 92mm Height. Weight: 2.5 kg. M/s. ECONOMY ELECTRONICS, • 15, Sweet Home, Plot No. 442, • 2nd floor, Pitamber lane, • Off. Tulsi pipe road, Mahim, • Bombay 400 016. Earth Sleeving Suresh Electrics & Electronics has de- veloped Earth Sleeving, for use in elec- trical and electronic industries. The sleeves are manufactured in yellow col- our with green stripe, and comply with international regulations for marking of single core cables and conductors. They are manufactured from plasticised PVC resistant to oils, ageing, light, ozone, acids and alkalies and can withstand temperature from -35°C to + 95°C. SURESH ELECTRICS & ELEC- TRONICS, • Post Box No. 9141, • Cal- cutta 700 016. • Phone: 29 0482, 29 5939. 4.62 NEW PRODUCTS • NEW PRODUCTS • N Contactors Larsen & Toubro Limited has intro- duced bar-type contactors which arc spe- cially designed for heavy duty under ad-” verse conditions. These contactors, de- signed series N, are suitable for both AC and DC applications up to 1000 V and can withstand up to 600 operations per hour (Continuous) and up to 1200 opera- tions per hour (occasional). Their DC2/3 ratings are equal to AC3 ratings and DC 4/5 ratings are equal to AC 4 ratings. They are available in AC3/DC3 ratings of II0A, 180A, 320 A, 600A, 900A in various NO and NC pole combinations up to 4 poles. L&T’s bar-type contactors are available with accessories like mechanical inter- lock and a mechanical latch. Larsen & Toubro Limited • Switchgear (S) Division • P O Box 278 • Bombay 400 038. Digital Multimeters Motwane have recently introduced, three handheld Digital Multimeters; model DM 452, a TRMS 4 1 /: digit and DM 352 & 350, both 3 1 /’ digit, These multimeters are heavy duty, top of the line, 7 function, 28 range handheld. LCD readout, with shrouded leads and terminals. All the three Multimeters have been de- veloped indigenously. Additional features are AC/DC currents upto 10A. Resistances upto 20M. ohm., continuity, conductance and diode test- ing, basic accuracy of 0.05% for the DM 452 with frequency response upto 10 KHz and 0.25% for the DM 352 and 350. these instruments, with help of probes can measure temperatures upto 1200°C. AC currents upto 600A, high voltages upto 40 KV AC/DC, frequenciesupto 20 KHz and RF upto 200 MHz. The Motwane Manufacturing Co Pvt. Ltd. • Gyan Ghar • Plot 434A • 14th Road • Khar • Bombay 400 052. Temperature Transmitter JNM's two-wire temperature transmitter is a design for 4-20mA current output suitable for various ranges of ther- mocouples, mV and RTD and has field adjustable zero and span with a turn down ratio of 3:1. The two wire concept helps to minimise the cable costs which is a great advantage. The output is current limited and upscale sensor fault indica- tion is standard. The RTD version features a built-in lineariser and the specified overall accu- racy includes linearisation errors and having the output being linear with the input temperature signal. The. transmitter has a removable PCB as- sembly housed in a weatherproof NEMA-4 enclosure. This arrangement reduces the installation, commissioning. maintenance and inventory costs. This unit can be conveniently mounted on a pipe in a horizontal or vertical direction . J.N. Marshall Pvt. Limited • P.B. No. 1 • Bombay - Pune Road • Kasarwadi • Pune-411 034. Transformers MAHAVIR INSTRUMENT TRANS- FORMERS are manufactured and de- signed by SHEPHERD TRANSFOR- MERS. Applications are in ELEC- TRONIC ELECTRO-MECHANICAL AND ELECTRICAL INSTRUMENT EQUIPMENTS. MAHAVIR' Transformers are tested as per l.S. Specification for resistance, Vol- tage, No Load. On Load Open and Short Circuit, H.V. Humidity Heat Run etc. These transformers can also be supplied with Electro Static shield. M/s. Shephered Transformers • Nityanand Nagar • Off Link Bridge • Ghatkopar (West) • Bomhay-400 086. Fasteners FTC offer a very wide range of fasteners to suit every requirement. Enquiries may be forwarded to. HTIilfil slllffffff Forged & Turned Components • Marine House • Shop No F • 11-A, Navroji Hill Road • 93 Dr Maheshwari Road • P.B. No 5153 • Chinch Bunder • Bombav-400 009. 4.64 NEW PRODUCTS • NEW PRODUCTS • N Wrist Strap This personal grounding wrist strap is an essential part of static proofing the elec- tronic work station for assembly workers and laboratory personnel. Protects sen- sitive devices from static charges gener- ated by the operator. This wrist strap is suitable for any size wrist. Coil cord has oxidation preven- tion, 6 ft., 360° swivel and high function resistor, strong grip banana plug and crocodile clip provide sure grounding. Davie Tech Inc. • 2-05 Banta Place • Fair Lawn • New Jersey-07410. Storage Scope With its maximum sampling rate of 100 MHz and a memory capacity of 10224 words, the Digital Storage Oscilloscope Bos is something special. A host of maths functions for signal processing and the IEC-bus interface (fitted as stan- dard), make the handy BOS indispensa- ble for both lab work and use in complex computer-controlled test systems. Rohde & Schwarz • Pressestelle • MuhldorfstraBe 15 • D-8000 Munchen 80 • West Germany. Power Mos SGS has added two new transistors to its N -channel POWER MOS family. The MTP3055 and MTP3055AP arc rated 60V (V DSS) and 12A (ID). The on-rcsistance, RDS (on) is as low as 0.15 ohm. Applications in- clude series regulators. DC/DC conver- ters and motor drives for industrial con- trol equipment. Both transistors are pro- duced using high density N-channcI en- hancement mode POWER MOS technology. Ultra high switching speed, low drive energy and low on-resistance are features which make these low cost devices ideal for very advanced, energy efficient, designs. SGS Semiconductors (Pte) Ltd • 28 Ang Ko Kio Industrial Park-2 • Singapore- 2056. Light Source ANDO of Japan offer the AO 4137 & AQ 4141. Both instruments are mul- tipurpose stabilised LD/LED light sources. The AQ 4137 can accomodate upto 2 channels while the AQ 4141 can accomodate upto 6 channels. Both LED and LD light source units of various wavelengths are available to allow many configurations with both AQ 4137 & AO 4141. The AQ 4141 can automatically sweep each light source unit, a conve- nient feature for loss measurment of multichannel or multicore fibers. The plug in units are available for wavelengths 1310 nm, 1550 nm. 1300 nm, 850 nm and 660 nm for LED light and 1310 nm, 1550 nm and 850 nm for LD. All the units can putout CW light or 270 Hz chopped light. A high optical out- put of 3-dBm is available from the LD light and an external modulation of 0.3 to 100 Khz at TTL level can be per- formed on LD units. GP-IB is available as standard in both the instruments. Both the instruments can operate on AC mains as well as battery. Murugappa Electronics Limited • ‘Parry House" 3rd floor • 43 Moore Street • Madras 600 001 . Switch Unique Electric Company introduces KS-30, a low profile snap action tactile feed back switch. This low stroke switch of 11.5 x 11.5 mm size is single pole, single throw, normally open type. Its design & low cost makes KS-30 a very economical switch ideally suited for ap- plications like keyboards. Telephones, Microprocessor-based CNC machines & Instrumentation, Toys, Electronic in- struments etc. The KS-30 snap action switch is also available with cap arrange- Unique Electric Company • B K Chhabra Compound • Vakola Bridge • Bulsroyce Lane • Santacruz (East) • Bombay 400 005. and tested circuit diagrams. PCB layout. Assembly process, Addresses of raw material suppliers, market prospects, Cost analysis on (1) Fire alarm systems (using smoke sensor) (2) Soft touch (feather touch) intercom (3) Electronic air purifier/freshner (4) STD Telephone lock (parallel/series) (5) Telephone amplifier (6) Solid State sine/square wave inverter (7) Mini emergency tube light (1 foot tube) (8) Colour TV using indigineous components. Each report cost Rs 290- only. Send M.O./D.D/P.O to S. DASGUPTA K. 2094 Chittaranjan park, New Delhi 110 019 Boards designed and m ENTERPRISES, P Bhag ram Nagar. Dombivali (E CORRECTIONS BASIC computer December 1987 p. 12.26 elekte* Subscription Name Old©! ABC ELECTRONICS 4.06 ADVANCED VIDEO LAB BALAJI ENGINEERING 4 06 BSM 465 CHAMPION ELECTRONICS COMTECH CYCLO COMPLTERS 4 04 DYNASCAN . 4 58 DYNATRON ELECTRONICS . 4 06 4 7C ECONOMY ENGINEERING . EFFTRONICS ELECTRONICA SALES EXCEL ELECTRONICS GH INDUSTRIES G S ELECTRONICS HCL 4 1' IEAP INDIAN ENGINEERING 4 63 JM ENTERPRISES 4 07 JR COMPUTER BOOK 468 JR COMPUiFR KIT 4 68 KAYSONS RADIO 4 7-, IEADLH ELECTRONICS 4 14 MLCO INSTRUMENTS 4 61 NLWAGF ELECIRONICS 4 06 OSWAL LLLC1 RONiCS 4 /? PIONEER ELECIRONICS 4 61 PROUUCI PROMOTERS 4 63 PRECIOUS BOOKS 4 2 PRECIOUS KITS 4 63 ROLAND El ECTRONICS 4 CS SAGAR INSTRUMENTS 4 73 SILICON AIDS 4 08 TANTIA ELECTRONIC CC 4 70 VASAVI ELECTRONICS 4 38 VIKAS HYBRIDS 4 15 VISHA ELECTRONICS 4 75 4.74 elektor India RN No 39881/83 MH BY WEST . 228 UC No 91 ZZhampmn £l£ctrziniz!s VxA Ltd S-17. M.I.D.C- Bhosari. Pune 411 026. India. (0212) 82682. 83791 Cable CHAMPION Telex: 0146-343 CHMP IN. 0145-333 MCCI IN. 0145-505 MCCI IN. u n