6.03 TELECOMMUNICATION NEWS • TELECO Secret code phone The Indian Telephone Industries Ltd., is developing a pilfer-proof public call of- fice which may turn out to be the first of its kind in the world. The PCO prototype would be ready by October-November this year and it would consist of a dual tone multiple memory push button tele- phone which would be operated only through a subscriber code, according to Mr. U.D.N. Rao, chairman and manag- ing director of the ITI. The PCO would approximately cost Rs. 2000 as against Rs. 20,000 for the magne- tic card or coin-operated phones. The secret code will be allotted to subscribers who deposit certain amount of money with the telephone department. Misuse of this PCO would be difficult as the sec- ret code would be known only to the sub- file ITI has also developed a low-cost satellite communication equipment for data transmission using technology cal- led Spread Spectrum Multiplex Access. The equipment is being manufactured by the ITI Equastorial Satcom Ltd., a joint venture of the ITI and the US-based Equatorial Pacific International Com- pany. The earth station for this system would cost about Rs. 400,000. Free of in- terference, the station would be able to receive facsimile data. Attempts are being made to add compressed voice channels to the station to enable voice transmission too. During 1988-89, ITI would launch the production of 34 MBPS and 140 MBPS fibre optics equipment, a 2000-line elec- tronic integrated local transit switch at a cost Of Rs. 3000 per line, and a wide band digital microwave system of 140 MBPS with foreign collaboration. ITI would also clear the backlog of spares required for the department of telecommunica- The ITI crossed the Rs. 500 crore pro- duction mark for the first time. In 1987- 88, its production was estimated at Rs. 509.34 crores against Rs. 452.49 crores achieved in 1986-87. Profits before tax for 1987-88 have been estimated at Rs. 11.59 crores. ITI’s production during 1987-88 in- I eluded 770.000 telephone lines and j 260,000 lines of electronic switching equipment as against 752,000 telephones and 175,000 lines switching equipment made in the previous year. The value of transmission equipment made in 87-88 was Rs. 107.68 crores compared to Rs. 999.28 crores in 86-87. The company made a capital investment of about Rs. 75 crores last year. Produc- tion figures for various units of the ITI last year were: Bangalore Rs. 187.19 crores; Mankapur Rs. 1 18.69 crores; Ral Bareili Rs. 81.59 crores; Naini Rs. 64.98 crores; Palghat Rs. 32.76 crores; Srinagar Rs. 8.22 crores. Sathe's communication Mr Vasant Sathe, Union minster for communication, is known for his uncon- ventional utterances on matters of public importance. His forthnight articles on the ills of Indian public sector have be- come legendary. Inaugurating a digital electronic ex- change in Bombay recently, Mr Sathe said that “self-reliance, indigenisation and appropriate technology are the equi- valent of obsolescence in communica- tion and information technology”. Tele- phone companies should be on a par with international technological standards, he added. The newly formed Mahanagar Tele- phone Nigam should take advantage of its autonomy and issue contracts to private parties for manufacture and mainte- nance. On behalf of the MTNL, Mr Sathe announced a Rs. 930 crore moder- nisation plan for the next three years. The plan includes doubling of Bombay’s existing 5,000 public call points and con- version of the coin-operated phones to use credit cards. The MTNL should allow liberal participation of the private sector in fulfilling the target while the Nigam could concetrate on key areas like the import of technology, Mr Sathe said. Mr. M.P. Shukla, managing director of MTNL said despite the 270,000 new lines commissioned in the last two years in Bombay, the waiting list had increased from 350,000 to 420.000 subscribers. Switching labour Rae Bareili appears to be a strong con- tender for the third Electronic Switching Systems (ESS-III) unit. This is likely be- cause over 6,500 workers at Rae Bareili would become redundant with the phas- ing out of electromechanical switching systems by 1990 and a new unit alone can absorb the workforce. To cope with the redundant labour, ITI proposes to begin production of elec- tronic transmission equipment im- mediately at Rae Bareili. The strategy would be to establish several new pro- ducts vital to the telecommunication net- work in India so that no worker would be forced to remain idle. The ITI chairman, Mr. Rao, dispelled any doubt over the siting of ESS-II by stating that this unit would be set up at Bangalore. Work was already in prog- ress on the first 128 lines Rural automatic exchange at the model plant set up by the ITI and the C-DOT. The first production model would be stabilised within six months and transferred to Bangalore complex of the ITT. The 512-line port C- DOT switching system would then be made at the model plant and then trans- ferred to the Bangalore plant in another 10 months, according to Mr. Rao. The ESS Max type would also be produced at the model plant first but commercial pro- duction would commence at another specially designed factory, adjacent to the existing model plant. Regarding the slogan “An Rax A Day”, scheduled to begin from April 1, 1988, Mr Rao pointed out that the ITI was al- ready producing a RAX a day but the delay was in shifting and installing these exchanges. By the end of May 35 ex- changes had been installed and progres- sively a higher rate would be achieved in later months. Agreement with NEC Bharat Electronics Ltd. and the ITI have signed a major technology transfer ag- reement with the NEC of Japan for indi- genous production of Digital Microwave Transmission System. A steering com- mittee comprising representatives of ITI, BEL, ECIL, department of elec- tronics and telecommunications examined various offers and chose the offer of NEC. Under the agreement, NEC will transfer knwo-how for the manufacture of 6 GHz, 140 megabits and 13 GHz, 34 megabits digital microwave system. The 6 GHz equipment will provide long haul trunk circuits while 13 GHz equipment will be used for connecting intra-city ex- changes. The equipment will be man- ufacted by BEL and ITI using monolithic microwave integrated circuits, dielectric resonator devices, LSI and VLSI circuits and thick and thin film microcircuits. 6.15 TELECOMMUNICATION NEWS • TELECO BEL and ITI will establish facilities to make 200 transreceivers of 6 GHz and 100 transreceivers of 13 GHz. Each com- pany will make an investment of Rs. 10 crores. The transfer of technology charges will be shared equally. Manufac- ture from raw materials and components will start by the middle of 1990. Data communication A countrywide, regular public data net- work called “Vikram” will be commis- sioned in 1989. Since June, 1986 an ex- perimental packed switching data net- work has been in operation, linking Madras, Bombay and Delhi. Mr. M.C. Venkatram, general manager, MTNL, Bombay, speaking at a seminar on data communications organised by the Victoria Jubilee Technical Institute, pointed out that MTNL was fully geared to providing all types of data communi- cation facilities. Bombay can boast of possessing the largest number of data cir- cuits anywhere in the country. Apart from 450 leased circuits, 33 dial up cir- cuits and five public data network cir- cuits, MTNL has provided 128 facsimile circuits on public switching telephone network. A large number of data circuits were provided to the railways, for their computerised reservation system, cus- toms and the ONGC to name a few. Sev- eral banks, LIC, RBI and a number of private sector units are in the line for get- ting data circuits. Organisations like Air India, Indian Airlines and Citibank are among the users of dedicated data cir- cuits. MTNL is also planning to provide videotext and electronic mail services. ELECTRONICS NEWS • ELECTRONICS N Russian Computers India and the Soviet Union have signed a protocol on computers and electronics which envisages supply of Soviet super- computers to India and setting up of joint ventures. The Soviet Union has agreed to provide the required specifications of two super- computer models, Elbrus-3-I and EC- 1068, before the end of July. India would examine the issue and depending on its response, the Soviet Union would make its commercial offer, according to Mr. K.P.P. Nambiar, secretary to the Depar- ment of Electronics, who signed the pro- tocol on behalf of the government of India. An Indian delegation will visit the Soviet Union in November, 1988 to see the developments in computers and as- sociated products. At the end of the sixth session of the Indo-Soviet working group on comput- ers and electronics, both the sides discus- sed the possibilities of production of full sets and systems for upgrading telephone lines in cities and villages. These systems were to be produced in India and deli- vered to the USSR. Also production of special telephone apparatus for mines, based on the Soviet know-how was con- sidered. Cooperation in production of colour TV sets and other consumer elec- tronic items in India and the USSR for supplying to third world countries fi- gured in the joint session, attended by Mr. V.M. Neiman, deputy chairman of the state committee for computer techniques and informations of the USSR. The Soviet Union has also shown in- terest in electron guns for colour picture tubes. A joint venture unit may be set up in India for producing the electron guns with the technical assistance of the USSR. Hungary Follows India and Hungary have signed an agree- ment for joint manufacture and market- ing of electronic goods and components. Joint venture units to be set up in India would be mainly for producing television sets and TV picture tubes, according to Mr Logis Koveskuti, president of the Council of Industrial Cooperatives, Hungary. In the next five years, Hun- gary would require at least 1.5 million TV picture tubes, it was stated. Electronics constitued Rs. 10 crores in the total Indo-Hungarian trade turnover of Rs. 110 crores. The target is to in- crease the share of electronics to about Rs. 60 crores. At present, Hungary is supplying microwave equipment to Doordarshan and telecommunications equipment to Indian Railways and the Department of Telecommunications. India is also interested in Hungary's medical electronics. Software tie-up Under an emerging cooperation bet- ween India and Australia, India will be exporting computer software to Au- stralia while Australian computer hardware firms will commence opera- tions in India to develop software for worldwide marketing. A delegation of the Associated Cham- bers of Commerce and Industry which attended the India-Australia joint busi- ness council meeting at Melbourne re- cently gained this impression. The coun- cil took note of India’s readiness to enter world markets in computer software through international collaboration. Au- stralian computer software houses evinced keen interest in India’s prog- ramme introducing computers in clas- srooms. The joint business council, which was ad- dressed by the Australian Prime Minis- ter, also identified opportunities of busi- ness growth in diverse areas such as min- ing and mining equipment, telecom- munication, food and manufactured food products, wood chips and pulp. The consensus at the council meeting was that the double taxation agreement between the two countries should be speeded up; for Indian software industry to become a major partner in global software trade, the legal protection av- ailable in India should be clearly de- monstrated; Cooperation would be vital in advanced computer software research projects; electronic components which are now imported by Australia from other Asian countries may be bought from India; and airlines should review freight costs and availability of space for high value, low weight cargo products. 'CRIS' goes abroad Indian Railways, having laid the tracks 6.16 ELECTRONICS NEWS • ELECTRONICS N and built the carriages for a number of foreign countries, now extend their ex- pertise in computerised reservation sys- tem to others. Indonesia and Thailand have sought the help of the Centre for Railway Informa- tion Services (CRIS) in computerising passenger reservation services. CRIS will help in setting up facilities, along with the software package to be de- veloped specifically for the two coun- tries. CRIS was set up two years ago to look after the entire computerisation prog- ramme of the railways. Encouraged by the queries from foreign clients, CRIS is evolving a marketing strategy for consul- tancy exports. Besides this, CRIS is now launching the phase II of the computeri- sation project which will bring smaller stations with light traffic on the compu- ter map linking the four metropolitan cities. Already experiments had been carried out successfully by linking the terminals in small stations to the main computer in a metropolis. For example, Delhi and Amritsar were connected by a terminal and in the east, Dhanbad was connected to Calcutta. More terminals would be set up in stations not having dense traffic. Another aspect of the sec- ond phase is to link Madras, Bombay, Delhi and Calcutta for two-way com- munication so that return tickets can also be confirmed from these four cities. At present, only journey from the originat- ing station is cofirmed. Liquor to Computer United Breweries Limited is a well- known liquor company but this firm has diversified its operations into many areas. The latest is in computer printers. United Breweries well set up a joint ven- ture with Genicom of US. The project will be set up in Bangalore. Genicom is believed to be the largest manufacturer of computer printers in the world. Last year’s turnover of Genicom was estimated to be $ 325 million. Under t..e tie up, Genicom will buy back 20 to 25 per cent of the printers manufactured in India. Unitel Communications, a sub- sidiary of the United Breweries, had al- ready entered into a technical collabora- tion with ATEA of Belgium for man- ufacturing EPABX systems. Chairman of the United Breweries, however, as- sures customers that the liquor units will not be closed down despite the diversifi- cations. After all, liquor provides the bread and butter for the rest of the busi- ness! RIBBONS TIE UP Vinkas General Carbon Limited, New Delhi, has entered into a technical and fi- nancial tie up with General Company Li- mited, Osaka, Japan for making the EDP ribbons in India. General Company has provided techni- cal know-how to Olivetti, Italy and Texas Instruments, USA. Vinkas' pro- ject is located at Bhimtal in Nainital dis- trict of Uttar Pradesh. The unit will pro- duce a wide range of nylon correctable and multistrike ribbons for use in most of the computer printers and electronic typewriters used in India. EDP ribbons are now imported. Vinkas will have exc- lusive export rights to West Asia and South Asia. The company will also explore the East European market. OPTO Electronic Lab A modern opto-electronic laboratory has been opened at the Machilipatnam unit of the Bharat Electronics Ltd. The sophisticated facilities being introduced will help BEL turn out precision optical lenses and prisms of international standards, 30 times faster. The facilities include high speed lapping, polishing, laser centering and edging, ultrasonic cleaning and multilayer thin film coat- • ing, besides testing and inspection of op- tical components and systems. The opto-electronic unit inaugurated by Gen Harbans Lai, director-general of quality assurance in the ministry of de- fence, marks the first phase of moderni- sation initiated by BEL five years ago, after acquiring the Andhra Scientific Company. BEL has invested Rs. 5 crores so far and with a further investment of Rs. 6.50 crores in the next two years. Tanker Laser Sights and Infra red Sear- chlights will be added to the product pro- file of the Machilipatnam unit. By the end of the decade, the unit would have the capacity to produce optical and opto- electronic goods worth Rs. 50 crores. Besides increasing the capacity, the modernisation will enable the manufac- ture of new types of high quality military opto-electronic products like passive night vision binoculalrs and goggles, weapon sights, Tank Laser Sights and TV camera zoom lenses. A dark tunnel of 60 meters has been constructed to simulate field conditions of Star nights for measurements and calibration of night vision instruments. A computer- aided optical design facility is also being added. Most of the products, with the exception of Tank Laser Sights, are in- dienously designed. Development and engineering work is being done in col- laboration with the Instruments Re- search and Development Establish- ment, Dehradun. Mobile Transmitters BEL recently delivered to the All India Radio a 10 KW, medium wave mobile transmitting station and the first solid state frequency modulated (FM) broad- cast transmitters, two units of 5 Kw and two units of 3 Kw each. The mobile station, designed and de- veloped by BEL, is a complete MW broadcast transmitting station cum studio on wheels. Portable taperecor- ders and audio mixers, radio networking terminal for receiving programmes from satellites and a rebroadcast are among the equipment provided. A trailer mounted diesel generator provides un- interrupted power supply. A transporta- ble radiating mast which can be erected in a very short time and rest facilities for operators are among the other features. The OB van is also air-conditioned. Each mobile station costs about Rs. 75 lakhs of which the foreign exchange component is worth Rs. 3.80 lakhs. The AIR placed orders for four such mobile stations in March, 1987. The first two were delivered in March 1988 as scheduled and the next two will be delivered by BEL in March, 1989. The AIR had projected a requirement of 168 FM transmitters of 3 KW and 31 FM transmitters of 5 KW in the seventh plan. While the 3 KW transmitters would be supplied by BEL and GCEL in a proportion of 2 : 1 , all the 31 5 KW transmitters would be supplied by BEL. The FEM transmitters will carry Hi Fi music without interference from other stations and noise. GCEL Circuits A thick film hybrid microcircuit laborat- ory was inaugurated by the DOE secret- ary, Mr K.P.. Nambiar recently at the Gujarat Communications and Elec- tronics Limited (GCEL). 6.17 ELECTRONICS NEWS • ELECTRONICS N The Laboratory has been set up at a cost of Rs. 45 lakhs of which the DOE pro- vided about Rs 30 lakhs through the Na- tional Microelectronics Council. GCEL has completed work on 12 designs of hyrid microcircuits while 10 more cir- cuits are in the pipeline. Hybrid microcircuits have advantages like small size, less weight, high reliabil- ity, lower costs and higher capabilities. They have a variety of applications in communications, instrumentation, au- tomobiles, television, audio and trans- mission equipment. These circuits can also be used for liquid crystal displays. Racing Via Computers A computerised betting system, be- lieved to be the first of its kind in the country, has been set up at Malakpet race course, Hyderabad. The system is being used for races held at Hyderabad as well as for inter-venue betting. The system is monitored and controlled by two ORG supermini computer systems. The two computers are connected by a local area network. In the event of fai- lure of one system, the other system au- tomatically takes over. The system sup- ports over 230 specially designed betting terminals and has a provision for sup- porting 200 more terminals. Security checks, validation of data, re- start procedures, and fast response to the betting person (punter) are among the notable features. The system enables the punter to obtain tickets on any pool, of any combination and of any denomi- nation at any of the selling windows. It also offers encashing facilities for win- ning tickets at any of the payout win- dows. In addition to betting functions, the system also provides cash tally state- ments, analytical statements of each pool and race and on line display of odds at regular intervals. The complete on- line software has been developed by ORG Systems and the Hyderabad Race Club. Computer Syllabus The Union government is finalising a standard syllabus for computer educa- tion programmes, especially in the area of programming. The syllabus, being examined by the All India Technical Education Council, will be announced shortly. Many recognised computer education courses would become availa- ble after the formulation of the syllabus. At present, most of the computer courses conducted by private institutes are not recognised for want of a stan- dard syllabus. Guided Vehicle The Project and Consultancy group of the BEL has designed and manufac- tured an Automated Guided Vehicle for industrial application. AG V is an intelli- gent, driverless, three-wheel transport vehicle moving at a speed of 1.5 km per hour with a maximum load of 100 kg to a distance of 10 to 15 km continuously. AGV is powered by two rechargeable maintenance-free batteries which may be used continuously for eight to ten hours. AGV is being used in the communica- tion equipment division of BEL’s Ban- galore factory for transporting high fre- quency transreceivers from assembly to testing. It moves on a pre-determined path, in response to commands given by the in-built microprocessor-based con- troller. Steering signals arc retrieved from a signal generator through a wire buried in the ground. Sensors detect any obstruction and the vehicle automati- cally stops at such instances. Recovery in chip industry Semiconductor companies fared better in 1987 than they have since the market slump, according to Dataquest. Many companies grew faster than the market, some showing growth rates that doubled and even trebled the overall industry growth rate. Dataquest estimates that semiconductor revenue grew 24.3 per cent over 1986 to reach $36.6 billion. Japanese companies had 48% of the world market; North American companies 21.7%; and European companies 11%. BALANCED LINE DRIVER AND RECEIVER These high-quality audio circuits are intended to overcome all the problems caused by noise picked up by long unbalanced signal lines between signal sources and amplifiers. Applications can be found in public address systems, studios, active loudspeakers, mixer desks and intercoms. The principle of balanced transmission of audio signals is relatively simple as shown in Fig. 1. The unbalanced signal from, for example, a preamplifier is ap- plied to an unbalanccd-to-balanced con- verter, which drives two output lines. One of these carries the inverted, the other the non-inverted signal. Noise picked up by the cable between the line driver and the receiver is superimposed on both AF signals. The complementary phase AF signals are added in the line re- ceiver to give an unbalanced output signal, which is a copy of the signal fed to the line driver. In this process, noise is effectively eliminated because its phase is identical on both input lines of the re- ceiver. In studios, practically all lines for inter- connecting equipment are of the balanc- ed type. Balanced-to-unbalanccd conver- sion and vice versa is usually effected with the aid of high-quality trans- formers. Unfortunately, these are hard to obtain and relatively expensive devices, and for this reason an alterna- tive based on semiconductors is offered Line driver The circuit diagram of the line driver is given in Fig. 2. The unbalanced input signal is applied to buffer Ai. This drives a non-inverting amplifier, A2, and an inverting amplifier, Aj. Both opamps are configured for an amplifi- cation of about 2. The amplification is I + R1/R2 in the case of A2, and -[(R4+R 5 )/R 3 ] in the case of A3 (no- tice that the minus sign denotes inver- sion of the input signal, not attenu- ation). Resistors Rs and R7 correct error voltages caused by the quiescent input currents of inverting opamps Ai and A}. Capacitors C7, Cs and C9 ensure very low distortion and stable gain up to the -3 dB roll-off frequency of 350 kHz. The opamps of the type stated in the circuit diagram give an output noise level of about 20 /iV «. This per- formance can be equalled by more com- monly found opamps such as the Type NE5534 (instead of the OP-27) and the Type NE5532 (instead of the OP-227), but only if all resistors in the circuit are ultra-low noise types with a tolerance of 0.1% or better to guarantee equal amplitudes of the balanced output signals. Line receiver: introducing the AMP-01 Special attention should be paid in the receiver design to low overall distortion. There are, however, awkward constraints to take into consideration. The most im- portant of these are the common-mode rejection of the opamp used, and cable capacitance. It is, therefore, necessary to use an amplifier that is geared to com- pensation, not amplification, of these sources of distortion. The Type AMP-01 precision instrumen- tation amplifier from PMI should meet with this requirement. The AMP-01 houses 4 interconnected opamps that amplify the potential difference between the input pins by a factor that can be ac- curately defined. The internal structure of the AMP-01 is shown in Fig. 3. Since the device is essentially an instrumen- tation amplifier, it rejects signals com- mon to both inputs. Unlike the com- plementary AF signals, noise induced on the balanced line between driver and re- ceiver is of the same phase and ampli- tude at both inputs of the line receiver. Hence, it is common to both amplifier inputs, so that it is effectively suppressed at the output. In contrast to an operational amplifier, an instrumentation amplifier requires precise internal feedback. In the AMP- 01, current feedback is used. This ap- proach has significant advantages over resistive feedback: • High common-mode rejection (CMR): approx. 130 dB at a gain of 1 , 000 . • Closed loop amplification, A vcl, can be set by the ratio of only two exter- nal resistors: A vcl=20Ris/Ri6. This allows any practical gain to be set with high precision and low gain— temperature coefficient. • The current feedback design is im- mune to CMR degradation when series resistance is added to the reference input. A small (trimmable) offset change results from added resistance, e.g., a printed circuit track. Close tolerance low-drift thin-film resistors are integrated on the AMP-01 substate to minimize output offset drift with temperature. Input transistors Qi and Q2 feed active loads, so that the amplification of this stage is about 4,000. Output amplifier Ai is a 2-stage circuit offering an ampli- fication of 50,000 in a 100 Q load. The open-loop gain of the AMP-01 is about 2 x10 s . Stability and linearity of the device are excellent, also at relatively high closed-loop gains. Ion-implanted super-beta transistors are used in combination with a patented bias-current cancellation circuit. Input quiescent current remains below 15 nA over the temperature range -25°C to + 85°C. A new geometry is used for the input transistors, resulting in an input noise of only 5 nV/j/Hz at a gain of 1,000. This noise includes contributions from the gain-determining and overload protection resistors. The input stage achieves an offset voltage drift of less than 0.3 J uV/°C. 6.20 elector India June 1988 The AMP-01 uses a special circuit for compensation of the load capacitance, ruling out any likelihood of instability oyer a wide range of practical gain. The high output current capability of 90 mAp allows the slew-rate of 4.5 /zs to be maintained with load capacitance as high as 15 nF. The circuit diagram of the balanced line receiver is shown in Fig. 4. Resistors Ris and Ri6 are dimensioned for an amplifi- cation of about 940. The value of Ris may be increased to reduce amplification as required for a particular application. The input offset voltage is set to nought with the aid of Pi. The symmetrical supply rails to the AMP-01 are decoupled with parallel combinations of a solid and an electrolytic capacitor. The potentials at the differential inputs of the chip are fixed with Rn-Rn-Ru. The balanced line receiver has a 3 dB bandwidth of about 30 kHz. Noise level at the output was measured at 5.3 mVm with inputs not connected, and 3.5 mVrms with inputs briefly connected to ground. Power supply The power supply shown in Fig. 5 should be familiar to constructors of previous high quality audio projects car- ried in this magazine. A number of readers have queried the use of the Type LM325 in this supply, and a short description of this device is, therefore, given below. The LM325 can supply equal symmetri- cal output voltages whose absolute value is accurate within 1%. Without external series regulator transistors, the device achieves a load regulation of 0.06 Vo at a maximum output current of 100 mA. On board the IC are a current limiter and an overheating protection circuit. The onset point of the current limiter is defined by an external resistor. Quiesc- ent current consumption of the LM325 is only 3 mA, while maximum input voltage is ±30 V. This makes it possible, in many cases, to feed the regulator direct from the existing symmetrical supply in the power amplifier. Bridge rectifier D1...D4 incl. (Bi) and rattle suppression capacitors C10. . .Cu incl. may then be omitted, but due attention should be paid to the working voltage of C» and Ci6. 1N4001 Fig. 5. The symmetrical power supply is a design based on precision voltage regulator Type LM32S. Construction The three circuits discussed arc ac- comodated on a single printed circuit board, whose track layout and compo- nent overlay are shown in Fig. 6. De- pending on the application of the balanced line driver and receiver, the PCB may be cut in two or three to enable fitting the circuits in the relevant lo- cations. The screened, balanced, cable between the line driver and receiver is connected as shown in the lower drawing of Fig. 1. It is recommended to use high-quality cable and XLR (Canon/Neutrik) con- nectors. Gb The functional description of the Type AMP-01 is based on information pro- vided in Linear and Conversion Prod- ucts . 1986/1987 Data Book. Precision Monolithics Incorporated. COPPER-ON-CERAMIC MICRO- ELECTRONIC TECHNOLOGY by Harry Cole, CEng, MIERE As the technology of modern micro-electronic circuitry advances, so too does the need to convey digitally coded signals at ever increasing rates. This requirement assumes considerable importance in the multi-layer type of circuit board where many widely spaced ICs have to be interconnected with negligible loss of amplitude. Traditionally, gold has been used for in- terconnection purposes in chip carriers intended for military and aerospace ap- plications where high reliability is of prime importance. Unfortunately, the relatively high electrical resistivity of gold (0.02 /uQ m) and the need for thin- ner interconnections makes this material unsuitable for the interconnection of densely packed ICs. Copper, although lacking some of the desirable properties of gold, has a resis- tivity of 0.016 /jS2 m which is con- siderably lower and has good solder- ability. It is much cheaper and has good adhesion properties when bonded to cir- cuit board materials. The Micro-electronics Technology Centre of British Aerospace’s Air Weapons Division at Hatfield has devoted considerable research into the use of copper interconnections laid down on substrates formed of alu- minium oxide ceramic. It has developed a fully documented repeatable process that can produce substrates in a variety of flat rectangular sizes up to 152 mm x 183 mm. Glass sealing glaze After being processed, the substrate can be machined by laser beam to virtually any shape, complete with access holes as required. The circuit interconnections are laid down on the ceramic substrate using screen printing technology, and circuit tracks as thin as 0.18 mm can be produced. Each circuit board may have up to six separate conducting layers, including gridded power and ground (earth) planes, the top layers containing the electrode attachment pads for the com- ponents to be fitted. Electrical isolation between the copper conducting layers is achieved by printing from two to five layers of dielectric material, the final thickness being tailored to suit the re- quired insulating properties of the circuit 6.22 elettor India iune 1988 Double sided copper-on-ceramic electronic module. being constructed. Interconnection between various metal layers is achieved by the printing of cop- per connector ’’slugs” in an isolation window cut into the dielectric insulation. This form of connection is known as a The final printed layer of each board takes the form of a high glass content glaze that effectively seals all preceding conductor layers from environmental hazards during subsequent manufactur- ing processes. A complete circuit may undergo as many as 50 screen printing operations and 30 separate firing cycles. During each firing cycle the printed substrate is subjected to a temperature profile that peaks at about 900 °C. Special furnace In the atmosphere of a conventional fur- nace such a temperature would cause the printed copper to oxidize, with conse- quent degradation of its electrical per- formance and solderability. For this reason, copper printed substrates are fired in an inert nitrogen atmosphere containing a critically controlled doping level of oxygen. The special furnace was designed in-house by the Micro- electronics Technology Centre at Hat- field. When manufacture is complete, the printed substrate is subjected to a rigor- ous programme of bare-board electrical tests to verify the correctness and con- tinuity of its circuit and connection pat- terns. A wide variety of components can be ac- commodated on the printed substrate, including ceramic and tantalum chip capacitors and leadless chip carriers con- taining up to 68 connection pins (this capability will shortly be expanded to ac- commodate larger chip carriers with up to 84 pins). The chip carriers referred to here are rectangular in shape and have connection pads located along all four sides spaced at pitch intervals of 1 mm or 1.27 mm. A particularly valuable advantage of packaging ICs inside chip carriers is that it enables them to be fully tested and qualified prior to being mounted on the ceramic circuit board. Easier flux clearance Once the ICs and components have been assembled on the board they are re- strained by an elastomeric fixative and then soldered into position by the tech- nique of reflow soldering. Because the body of the chip carrier is made from a ceramic material similar to that from which the circuit substrate is made, it has similar thermal characteristics and the soldered joints are not subjected to significant thermally induced stresses. The process developed by the Micro- electronics Technology Centre for the at- tachment of components to circuit boards results in a controlled stand-off (board clearance) height for the mounted components of about 380 mm. The advantages that come from such a clearance are that they ease the clearing of flux residues from under the compo- nents and make possible the close in- spection of solder joints by either nor- mal visual means or new techniques such as X-ray microfocus. This type of inspec- tion is not possible without stand-off height. The complete clearance of flux residues is highly desirable since the presence of such contaminants can pose a serious risk to long-term reliability. External connections to the circuit board are made by a surface-mounted connec- tor with soldered joints formed by reflow soldering. There are essentially two methods of as- sembly adopted by the Micro-electronics Technology Centre. The first makes use of a single sided board fitted with metal edge supports so that the ceramic board can be inserted directly into a standard Eurocard rack. The second method takes advantage of the good thermal conduc- tivity of the aluminium oxide ceramic substrate and uses it as a heat sink for two fully assembled ceramic boards mounted back-to-back on either side of an aluminium core plate. Punishing tests The plate functions as both a structural support and a very efficient heat sink. This form of assembly is used where space is restricted. The heat-removing properties of the assembly are further enhanced by the use of high thermal conductivity material for the elastomeric adhesive employed for component place- ment and for attaching the boards to the core plate. Circuit board assemblies using the back- to-back method of mounting have been subjected to an independent series of tests carried out by the Components Evaluation Department of British Aerospace’s Air Weapons Division. The tests, which are punishing to any electronic assembly, have included operating temperatures ranging from -55 °C to 125 °C; damp heat storage at 85% relative humidity at 85 °C for 1000 hours; a one minute acceleration of a gravitational force (g) of 1000 (9806 metres per second per second); 400 tem- perature cycles ranging from -55 °C to 125 °C with ten minutes dwell time and five minutes transfer time; and 100 cycles of ambient power cycling for 15 minutes each side when dissipating 130 mW/cm ! . As if this was not enough, low frequency power cycling was imposed at 130 mW/cm', and switched on and off repeatedly for 20 temperature cycles. Automated manufacture In addition, high frequency power cycle was superimposed on 70 temperature cycles while power at 130 mW/cm ! was switched on and off at one minute inter- vals. In addition to its design and manufac- turing capabilities at Hatfield, the Micro-electronics Technology Centre can also undertake the modification and repair of fully assembled boards. It can, for example, remove and replace all sizes of leadless chip carriers, chip capacitors, chip resistors and flat conductor cable, and cut and isolate copper track. Work is under way to commission an automated production facility dedicated to the manufacture of copper-on- ceramic multi-layer modules using the operating experience acquired from the company’s existing design and manufac- turing service. Micro-electronics Technology Centre, British Aerospace PLC, Air Weapons Division, Manor Road, HATFIELD ALIO 9LL. BREAKTHROUGH IN SUPERCONDUCTING MATERIALS by Peter Hartley, MlMGTechE One of the technological sensations of the past eighteen months has been the race towards the first effective room temperature superconducting material. Apart from achieving reliability, one of the major problems is how to fabricate useful products on a commercial basis from the new breed of ceramic materials. Basic Volume a specialist sensors and electronics materials company, claims to have produced the world’s first superconducting solenoid in a ceramic material. This development could start an avalanche of applications for super- conducting ceramics. Superconductivity, the phenomenon in which a material loses all its resistance to electric current, was until comparatively recently observed only at temperatures below -250 °C. This required the use of liquid helium as a coolant. In February 1987, however, Dr Paul Chu at the University of Houston in the United States discovered a ceramic com- pound that superconducts at -183 °C, consisting of yttrium, barium, copper and oxygen. This opened up the possi- bility of using liquid nitrogen, which boils at - 196 °C and is much cheaper to use than helium as a coolant. Single-turn solenoid Basic Volume, which manufactures solid-state chemical sensors and signal process electronics, was actually produc- ing some of the materials used for super- conductivity researches enabling the company’s Dr Tim Tavares and his team to experiment with materials that were available immediately. In March 1987, the company was able to produce samples of Yi iiBao.iCuOix with a superconductivity transition tempera- ture of -213 °C, and this was rapidly improved upon with its so-called YBC0123 compound in April. On 24 April, after three previously un- successful attempts, the company managed to produce a ceramic supercon- ducting single-turn solenoid. This was made of YBC0123, measured 90 mm long with a 14 mm outside diameter and a radial thickness of 3 mm. The device’s material was also, unlike some other similar materials, stable in water. This development means that many commercial applications of the new nitrogen-cooled ceramic supercon- ductors are months rather than years away. Among the uses immediately en- visaged are: nuclear magnetic resonance equipment, magnetic mineral separa- tion, magnetic bearings, high torque dc electric motors, spin resonance spec- troscopy equipment, and electronic im- aging apparatus. Metal matrix composites Dr James Watson of Southampton Uni- versity, a specialist in the magnetic separation of minerals, has been sup- plied by Basic Volume with tubes of YBA2 CUj O 6.S-7 superconducting cer- amic. These tubes, which are 37 mm in diameter, 90 mm long and 1 mm thick, superconduct reliably at temperatures up to -196 °C. He finds that the material supplied to him by Basic Volume provides a much higher density than competitive products — 90% of the theoretical maximum — and allows current carrying capacities up to 5 x 10 4 A/cm ! . Superconductors may have made the technological headlines during the past year, but an equally exciting engineering materials development has been that of the commercialization of metal matrix composites. To most engineers, the word composites conjures up the image of plastics rein- forced with glass or carbon fibres. In these materials, the fibres confer strength to an otherwise mechanically weak material. The fibres take the load and the polymer matrix serves to distrib- ute this load equally between them. Large capacity production The same principle is used in a family of new materials — metal matrix com- posites (MMCs) — in which a metallic phase (the matrix) is reinforced by very strong ceramic or metal fibres, whiskers or particles. In 1984, the British Collyear Committee Report on new engineering materials and processes states that the key areas for development in MMCs were con- cerned with the technology of producing them in tonnage quantities. This is an aim towards which the Metals "technology Centre at the Harwell Laboratory (2) is deeply committed. The so-called MMC Club, organized by, and centred on, Harwell, is carrying out research into titanium-based MMCs within the framework of the European Community’s BRITE (Base Research in Industrial Technologies) programme. Cray Advanced Materials D), backed by the Cray Electronics Group, was formed some 18 months ago to exploit MMCs commercially. The company operates under a licensing agreement from Britain’s Ministry of Defence and uses the patented liquid pressure forming (LPF) process, a new technique for mak- ing components from fibre-reinforced metals. Producing complex shapes Commercial applications of MMCs, cur- rently at the feasibility demonstration stage, include: components for the automotive industry such as pistons, connecting rods, brake callipers, and wheels; gas cylinders; marine propellers; armour plate; lead battery plates; bicycle frames; robotic arms; overhead pan- tographs for electric trains; and specialist tools. The LPF process is a technique for the production of ceramic fibre-reinforced metal components to net shape or near net shape, with excellent dimensional tolerances and exceptional mechanical properties. Various types of ceramic fibre, such as silicon carbide, alumina, boron and car- bon, can be used with metals such as aluminium, magnesium, lead zinc or copper alloys. The Cray LPF process can be used to produce complex shapes that vary in size from a few centimetres up to 2 m x 1.5 m x 1.5 m. It is essentially a single- batch process that produces a net- shaped reinforced component with toler- ances in the region of ± 0.2%. Pro- duction times are relatively short and die costs low for the volume output. Glass ceramics Ceramic Developments (Midlands) Ltd (CDML) I 4 ) i s carrying out a range of investigations into the engineering appli- cations of glass ceramics. These differ from true glasses in being polycrystalline ceramics resulting from the crystalliz- ation of glasses. They differ from traditional engineering ceramics in that the starting material is nearly always completely amorphous and not the product of the liquid phase sintering of ceramic precursors. CDML has carried out an internal devel- opment programme aimed at producing a range of photomachinable glasses and glass ceramics with differing expansion coefficients in the range 7 x 10“ 6 /°C to 11 x 10 V°C. Augmented by CDML— funded research at Sheffield University this has led to the development of a useful range of materials with potential for: • micro-electronics substrates where high densities at fine holes are needed for interconnection; • plasma display panels; • competition with low volume pro- duction of ceramic components but giving higher precision without pro- hibitive tooling costs. Potential applications The Department of Trade and Industry has recently awarded the company a grant of £28,000 to further this project. Glass ceramics developed at CDML show abrasion resistances comparable to that of boron carbide. Potential appli- cations for this material are pipe linings, the coatings for moving parts operating in abrasive environments, and possibly the plasma spraying of large metal com- ponents in situ. According to Dr Ronald Jones, CDML’s managing director and founder, the firm is now at the stage of being able to cast glass ceramic pipes centrifugally. One of the most impressive results of work at CDML has been the successful development of glass ceramic armour that shows a similar ballistic perform- ance, thickness-for-thickness, to that of alumina, when used as a protection against 7.62 mm calibre rifle bullets. The real advantages, however, arc weight saving — since the ceramic glass has a density of only 2.4 g/cm ! as opposed to 3.8 g/cm 1 for alumina — and relatively low processing costs. Screening equipment A further useful development at CDML has been the production of very high- quality glasses and glass ceramics by Sol- Gel technology. This involves polymeriz- ing the silicate networks from ethoxy- silanes by condensation polymerization. The resultant glasses have a very high surface area tension and can be densified by heal treatment at about 500 °C. This enables glass to be made for catalyst supports, barrier layers and coatings, which are almost impossible to produce by conventional fusing of oxides. CDML is also researching the use of glass ceramic materials for use as strong, heat-dissipating substrates for thick-film circuitry. The company has formed a consortium with Thorn-EMI, Lucas, Wade and Engelhard, to develop and ex- ploit its work in this area. Another area of United Kingdom elec- tronics materials development has been in the field of screening equipment against electromagnetic interference (EMI) and radio frequency interference (RFI). Thus, in anticipation of new and tougher European Community legis- lation on allowable levels of elec- tromagnetic noise emissions from equip- ment, Shipley Europe * 5 * has introduced a new EMI shielding technique based on the firm’s well established electroless plating technology. Longlasting adhesion The chemical process involved deposits uniform thicknesses of copper and nickel coatings on all component sur- faces to give, it is claimed, a 40 dB im- provement in attenuation of EMI over previous methods such as arc spraying of zinc or the use of conductive nickel Since copper is a better electrical con- ductor, it provides maximum protection with a fraction of the thickness used with other methods, resulting in lighter fabrications and greater cost savings. Shielding effectiveness is greater than 80 dB with a copper thickness of only 625 nm. Total immersion in a series of chemical treatment baths ensures that all surfaces, no matter how complex, receive a uniform coating of metal. Pre-cleaning and etching give long lasting coating adhesion and ensure, it is claimed, that electroless shields will not crack or flake the way some arc sprayed zinc coatings The process can be used for both solid injection moulded plastics parts and structural foam plastics components, made from ABS, polycarbonate, polypheylene oxide, polystyrene and many other polymers. A final coating of electroless nickel protects against cor- rosion, abrasion and provides a suitable base for cosmetic finishing. Elastomeric gaskets Dowty Seals * 6 * claims to have achieved a breakthrough in EMI/RFI screening with its new Dowshield range of conduc- tive elastomeric seals and gaskets. This range incorporates seven different com- pounds and three types of seals — Dowprint, moulded seals and extruded profiles. Four of the compounds are used specifically for the production of flat elastomeric gaskets by the company’s Dowprint screen printing process. For this, there is a choice of silicon or cross- linked vinyl polymers, loaded with con- ductive silver or silver-plated nickel par- ticles. They provide a volume resistivity as low as 0.0004 Q/cm and a signal at- tenuation as high as 106 dB by the American Society for Testing Materials (ASTM) test method. The moulded seals and extruded compo- nents employ silicon or fluoro-silicon materials loaded with silver-plated nickel. Radiation dosemeter Another interesting innovation on the plastics front is the development of 20% and 30% glass fibre-reinforced, nuclear radiation resistant, polyethersulphone — Victrex PES — by ICI* 7 *. The British custom-moulding company, Jarzon Plastics* 8 * has added these materials to its range of engineering plastics. It has collaborated with the GEC company to design and produce nuclear radiation dosemeters — worn on the wrist — in one of these materials, which allows gamma rays to penetrate. After exposure to radiation, the dosemeters are slotted into a drawer of the same material, which in turn is slot- ted into a reader to obtain the radiation level readout. References: 1. Basic Volume Ltd, 13a Cotswold Street, London SE27. 2. Harwell Laboratory, Harwell, Didcot, Oxfordshire OX11 ORA. 3. Cray Advanced Materials Ltd, 6 Ar- moury Road, Lufton Trading Estate, Yeovil, Somerset BA22 8RL. 4. Ceramic Developments (Midlands) Ltd, St Marks Road, St James In- dustrial Estate, Corby, Northampton- shire NN18 8AN. 5. Shipley Europe Ltd, Herald Way, Coventry CB3 2RQ. 6. Dowty Seals Ltd, Ashchurch, Tewkesbury, Gloucestershire GL20 8JS. 7. Imperial Chemical Industries PLC, Petrochemicals & Plastics Division, PO Box 6, Bessemer Road, Welwyn Garden City, Hertfordshire AL7 1HD. 8. Jarzon Plastics Ltd, Golden Crescent, Hayes, Middlesex UB3 1AQ. VLF CONVERTER Build this lO MHz up-converter if you ore interested in receiving signals from time standard, FAX, RTTY and other utility stations operating in the kilometric bands. The frequency range from 30 kHz to 150 kHz is generally referred to as the Very Low Frequency (VLF) band. It is used relatively little, because high transmitter powers and large aerial systems are required, which generally give a relatively small coverage (typically about 300 to 1000 km). For a number of applications, this is not considered a dis- advantage, however. Propagation of VLF waves is highly predictable, since there is virtually no atmospheric reflec- tion: the transmitter range is, therefore, fairly accurately defined. VLF signals travel almost exclusively via the so-called ground wave, while the ground- ionosphere space acts as a waveguide. Thanks to this property of the VLF band, received signals are usually free from phase shift and amplitude vari- ation (fading), so often found on the shortwave bands. The VLF band is well suited to medium-range, one-way data communication, such as time standard transmitters (Rugby GBR, Rugby MSF, Mainflingen DCF77, Prangins HBG), meteorological facsimile, submarine communication, and telex networks. One disadvantage of the VLF band is the huge aerial system required at the transmitter side. Aerial systems of several square kilometres, and with multiple transmitter feed points, are not uncommon, yet attain a radiation ef- ficiency of only a few percent. At the re- ceiver side, due account should be taken of the high level of man-made noise (computers, neon tubes, TV sets and other electrical appliances). In most cases, the so-called long wire is the only feasible aerial at the receiver side. Thirty metres or more of sloping or horizontal- ly running insulated wire, mounted well away from the previously mentioned sources of interference, is recommended for serious experiments in receiving VLF transmissions. Generally, the lower the frequency, the rarer and more interesting the stations. Not every communications receiver can be tuned as low as 15 kHz, but this is made possible by the up-converter de- scribed here. It effectively converts the frequency range from about 15 kHz. . .300 kHz to 10 MHz, so that the CW, RTTY, FAX, AM or SSB 6.26 oleklor India june 1988 facilities of the communication receiver tuned between 10.000 and 10.300 MHz can be exploited to receive VLF transmissions. Circuit description The VLF converter is an application of the Type NE602 active double balanced mixer and oscillator, whose block diagram is shown in Fig. 1. The chip re- quires only a handful of external compo- nents to make a good-quality up- converter. 1 Fig. 1. NE602 integrated active double- balanced mixer/oscillator. The circuit diagram of the converter is given in Fig. 2. The function of the cir- cuit is to convert the frequency range from 15 kHz to about 300 kHz to an equally large band starting at 10 MFIz. The SSB/CW/FAX/RTTY receiver con- nected to the output of the converter is tuned between 10.015 and 10.300 MHz. A VLF station such as Rugby MSF is, for example, ’’received” at 10.060 MHz. The VLF aerial signal is passed through low-pass filter L 1 -C 2 -L 2 -C 3 -L 3 -O that defines the input frequency range (15 kHz. . .300 kHz). Transistor Ti forms an impedance transformer be- tween the filter output and one RF input of the active mixer in ICi. The NE602 is set up in an asymmetrical configuration here. RF input pin 2 is bypassed to ground with Ct, while Pi is used for set- ting equal direct voltages at the RF in- puts to optimize mixer balance. The out- put frequency of the local oscillator on board the NE602 is set to 10 MHz with the aid of an external quartz crystal, Xi. Trimmer Cm provides a means for ac- curately setting the LO frequency to 10.000 MHz, so that the tuning scale on the receiver corresponds to the true re- ceived frequency, ignoring, of course, the ”10” preceding the kHz digits in the read-out The output of the active double balanc- ed mixer is a single-ended configuration. The up-converted frequency band is filtered in pi-section C12+C13-L4-C14-C15 to suppress spurious mixer products. The low-frequency roll-off point of the filter is set to about 10 MHz by trimmer C13. It should be noted that the mixer also generates an image band between 9.985 MHz (10-0.015) and 9.700 MHz (10-0.3), but this of little consequence. The converter is fed from a regulated 12 V source, either via separate supply wires (do not fit link A — B), or via the downlead coax to the receiver (fit link A — B, and make sure that the receiver output voltage is between 10 and 15 V). The mixer/oscillator and preamplifier transistor are fed from a 6.8 V rail created with stabilizer Rj-Di. Construction and alignment The VLF converter is a simple to build project. The printed circuit board is a double-sided, but not through-plated, pretinned type — see Fig. 3. Commence the construction with fitting 15 mm high brass or tin metal sheet screens as shown on the component overlay. Component leads shown without a small circle are soldered at the track side of the board, and to the ground area provided on the component side. Radial inductors Li, L2 and L3 are ferrite encapsulated types from Toko. The mixer/oscillator, ICi, is fitted direct onto the PCB (do not use a socket). Drill a 2 mm dia. hole in the screen to give access to the spindle of multiturn preset Pi. Finally, fit soldering terminals for the input, output and supply connections. Install wire link A — B if the converter is powered via the coax cable to the receiver. Set Pi, Cu and C13 to the centre of their travel. Apply 12 V to the circuit, and check the presence of 6.8 V on pin 8 of ICi. Measure the direct voltage on pins 1 and 2, and adjust Pi until both are held at an equal potential of about 0.8 V. Connect the receiver, and tune this to 10.000 MHz. Mode: CW, BFO off, or to the centre of its travel. Switch on the input attenuator, or select re- duced RF input gain. Lower the fre- quency of the beat note heard to nought by adjusting Cu (zero beat). Connect the aerial to the VLF converter, and tune to a relatively strong transmission at elektor India june 1988 6.27 Frequency Station Power Service 16 kHz GBR Rugby (UK) 60 kW time signals during the 5 minutes preceding 03.00h, 09.00h, 1 5.00h and 21.00h. 50 kHz RTZ Irkutsk (USSR! 50 kW 60 kHz MSF Rugby (UK) 50 kW standard frequency & BCD time 71 kHz not identified 75 kHz HBG Prangins (SUI) 20 kW 77.5 kHz DCF77 Mainflingen (FRGI 50 kW standard frequency & BCD time 117.4 kHz DCF37 Mainflinqen meteoroloqical facsimile 134.2 kHz DCF54 Mainflinqen 139.0 kHz DCF39 Mainflinqen photofacsimile. Some stations that can be received below 150 kHz. The screening fitted onto the component side prevents oscillation and spurious mixer prod- about 200 kHz (10.200 MHz on the re- ceiver), e.g. Droitwich (AM). Reduce the input gain of the receiver, and peak Cu for optimum reception (this adjustment is relatively uncritical). Stations and services It should be noted that the VLF con- verter has some conversion gain, so that every care should be taken not to over- drive the communications receiver. It is, therefore, strongly recommended to make use of the fixed or variable RF at- tenuator provided on most receivers. The connection between the VLF converter and the unbalanced, low-impedance (50—100 Q) receiver input must be made in coaxial cable to prevent breakthrough of strong signals in the 10 MHz band. The frequency assignment used in the VLF band is roughly as follows: 15 — 100 kHz: submarine communi- cations (CW), beacons and time stan- dard transmitters: 100 — 150 kHz: RTTY (radio teletype), and meteorological facsimile services; 150—300 kHz: long-wave broadcast ser- vices and, occasionally, RTTY. SIGNAL PROCESSING AND ELECTRONIC ENCRYPTION by Brian P. McArdle This article examines the effect of encryption operations on the usual understanding of signal processing. Electronic signals are made secret by en- cryption operations. The original signals can be analogue or digital. For the pur- poses of this article analogue signals are complex waveforms made up of different frequencies. Digital signals are a se- quence of pulses where each pulse can be identified as a particular bit (logic state 1 or 0) by reference to the voltage level, polarity, etc. They are also called mark and space pulses respectively. I do not intend to become involved in a detailed description of encryption, but to provide an over-view which should assist an elec- tronics engineer or technician. For simplicity, the various encryption oper- are considered to turn a signal into another signal of the same category, i.e., analogue signals after a scrambling operation are still analogue signals. Elaborate systems where analogue signals are sampled, turned into digital signals which are encrypted and trans- mitted in digital form are not con- sidered. The paper explains two simple examples which can be altered as re- quired. The comments and conclusions are of a general nature and may require amendment according to particular cir- cumstances. Digital signal processing Consider the encryption process il- lustrated in Figure 1. This represents a typical arrangement for the encryption and transmission of confidential infor- mation between message centres, such as 1 Fig. 1. Encryption of digital signals, two embassies. If the teleprinter uses the CCITT Number 2 Code, each character pressed on the key-board will be represented as 5 bits plus start and stop bits. The electronic word is transmitted in bit serial mode (one bit at a time) to the encryption unit. For simplicity, we will assume that the start and stop bits are not encrypted which is the procedure used in most cipher machines. The 5-bit block can be encrypted as a single block or one bit at a time. These are called Block and Stream Encryption respect- ively. (Usually blocks are made up of 64 or more bits.) The reader is referred to Ref (1) for an analysis of the various en- cryption operations. These need not be examined in this article. The actual elec- trical connection between teleprinter and encryption device is a 20 mA current loop illustrated in Figure 2. The start pulse has the same duration as a data pulse. The stop pulse is 1.5 times the duration. Hence, the code is referred to as a 7.5 bit code. There are 31 possible combinations (2 5 -l) because the state 00000 is not used. The normal speed is 50 or 75 baud. There are other codes with other interfacing arrangements but the overall concept demonstrated by this example holds. Fig. 2. Interfacing arrangement using 20 mA current loop. The level of secrecy depends on the com- plexity of the encryption operation. This is usually varied by adjusting internal settings inside the device. Obviously, the same setting should not be used continu- ously. Otherwise the effect of the en- cryption operation would be cancelled or reduced considerably. An unauthor- ized listener (hacker) on the channel would probably know the arrangement but not the actual settings in use. There is always a very large number of possible settings in order to avoid deduction of a particular setting by trial and error. In cryptographic terms, the setting is gener- ally called the key because it unlocks the information from confinement by a secrecy process. The reader is referred to Ref (2) for an analysis of the secrecy re- -quirements of an encryption system. Basically it should not be possible to deduce the key by any method other than by trial and error. Hence the need for a very large set of possible keys. However, cryptosystems are not discussed in this article. If the channel is a HF link, the output from the encryption unit becomes the audio input to the transmitter. This con- sists of 2 different tones with a frequency difference of 850 Hz between them. The upper frequency usually represents the ”1” or mark pulse but this represen- tation is sometimes reversed. This is known as Frequency Shift Keying (FSK) Fig. 3. Communications over a HF radio link. and is explained in most modern text books on Telecommunications. It can be detected on most standard receivers by using Amplitude Modulation (AM) on Single Sideband (SSB). Obviously, the encryption unit must have the appro- priate outputs to interface to the trans- mitter and not the smame arrangement as the interface to the teleprinter. The re- verse procedure is applied at the receiver where the audio output is fed to the decryption unit. There are many other possible arrangements depending on the equipment and type of channel. Analogue signal processing This is usually used to encrypt voice communications, such as over tele- phones and radio links. The encryption operation conceals the information con- tained in an analogue signal but the resulting encrypted signal is still in analogue form. A typical arrangement is shown in Figure 4. Usually, the Fig. 4. Encryption of analogue signals. scrambler unit cannot scramble and descramble simultaneously and conse- quently half-duplex operation must be used. The output from the scrambler is an analogue signal in the same frequency range as the original audio signal (refer to the assumptions in the opening paragraph of this article). A common method uses frequency inversion where parts of the analogue voice signal undergo an inverting operation which is controlled by a definite procedure inside the scrambler. The descrambler uses the same procedure and simply reverts the frequencies back to their original ar- rangement. The full operation is its own inverse which in turn facilitates im- plementation and use. However, analogue signal encryption is not con- sidered to provide the same level of secrecy as digital signal encryption. This is because the cascaded substitution- permutation operations that provide real secrecy are more easily implemented with groups of bits. This statement re- quires further explanation in order to tie down the full problem. Consider a digital system again as shown in Figure 5. The permutation operation re-arranges the order of the bits. For example, bit 9 moves to the 4th position and so on. The substitution operation divides the permuted block into sub- blocks of 4 bits each which is replaced by another block (e.g. 0100 becomes 1010). Fig. 5. Permulalion-substitution operation. Different replacement tables can be used for each sub-block. If these operations are repeated a number of times with dif- ferent permutation and substitution tables, the entire process has a high level of secrecy. The complete operation is varied by changing the tables. The theory behind this procedure is ex- plained in Ref (3) but need not be con- sidered in this type of overview. If the channel is a VHF radio link, the scrambled audio signal is used to modulate an RF carrier using the same methods of modulation as for un- scrambled voice communications (AM and narrowband FM). There are other possible arrangements but they do not differ substantially from this example. Communications channels The quality of a channel can have a major effect on encrypted signals. Con- sider the arrangement in Figure 1 again. If the channel is noisy, a pulse could be corrupted sufficiently for a ”1” to be detected as a ”0” or vice versa. This means that a bit in an electronic word is incorrect and consequently the decryp- tion operation will produce a wrong character. If this type of corruption does not happen too often, the person who ultimately reads the decrypted message will notice the errors and be able to alter the text accordingly. However, a serious problem does occur where a decryption operation uses successive bits of a mess- age in some inter-dependent fashion. This means that a sequence of words elektor india jane 1988 6.29 would be reproduced incorrectly because of a single error. A simple arrangement is illustrated in Figure 6. The output bit from the encryption operation becomes part of the operation to encrypt the next input bit. Thus, successive output bits are linked together. A single error could have a disastrous result at the decryption stage. This example only uses one bit in the feedback loop but some systems could be using as many as 64 or 128 bits. Fig. 6. Inlcr-dependende between output bits (the + sign refers to addition modulo 2, which is an Exclusive OR logic operation). The reader is referred to Ref (4) on Cipherfeedback Mode which is a good example of this effect. If the channel is very noisy with a high level of corruption of the digital signals, a very reliable error detection/correction system must be in- stalled between the encryption/decryp- tion unit and the channel. The only alternative would be to transmit the in- formation in plaintext without being en- crypted which may not be satisfactory for the users. Consider the example in Figure 4 again. Although it was not stated in the original explanation, a synchronization signal would probably have to be transmitted at regular intervals in order that the descrambler can reverse the scrambling process in the correct sequence. If the channel is noisy, a noise pulse could be interpreted as a synchronization signal by the descrambler such that the output signals turn into unrecognizable rub- bish. This may not actually happen too often, but when it does communications are totally blocked. Alternatively, the scrambled signals may be corrupted but this may only affect certain frequencies. Voice communication has a very high level of redundancy and even noisy messages can sometimes be understood by an experienced operator. However, in general analogue signal encryption re- quires a good channel for reliability. Otherwise it may simply not work. Conclusions There is an obvious problem with analogue signals. Remember that these are essentially little packets of different frequencies that together form the signals, which in turn become the infor- mation. The various operations of en- cryption, transmission over a channel and decryption must reproduce the original signals as accurately as possible. In reality, these are difficult re- quirements to satisfy. For example, to compensate for a noisy channel, the en- cryption operation may have to be simple and straight-forward, which in turn reduces the level of secrecy. This could mean that the effect of encryption is to provide privacy rather than secrecy (in this context, privacy means secrecy against members of the general public rather than code breaking organizations, such as the National Security Agency in the U.S.). However, encryption oper- ations that require bits, such as the Data Encryption Standard, are much more flexible. They only require a single bit or block of bits as the input. The key or control to vary the operation is generally also a block of bits. The full operation can be described with Boolean Logic Operation(s) which are now known by 1st year students. Thus the whole area of encryption and secrecy seems to favour the use of digital rather than analogue signals. (6) References 1. Denning, Doroty E.; Cryptography and Data Security. Addison-Wesley: U.S.A. (1982). 2. Heilman, Martin E. and Diffie, Whit- field; ’’Privacy and Authentication: An Introduction To Cryptography”. Proceedings of the IEEE , volume 67, number 3 (1979). 3. Shannon, Claude E.; Mathematical Theory of Communications. Univer- sity of Illinois Press: U.S.A. (1962). 4. "Data Encryption Standard”. FIPS PUB 46, National Bureau of Stan- dards, Washington D.C., U.S.A. (1977). morse trail Mr For those who don't know, Morse is a little like binary without the logic. Understandably, learning the Morse code is a long process. In practical use one has to know all the signals by heart, there is no time to even think about it when listening to an actual transmission. Learning them is therefore very much like reciting multiplication tables in school. This is the idea behind the morse trainer. 6.30 Figure 1. Schematic diagram of the morse trainer. The morse trainer constantly repeats a certain signal which has been chosen by a few switches. A letter is represented Tn morse code by a series of dots and dashes, a dash lasting three times as long as a dot. The interval between two dots (and the dashes too) is determined by the clock generator (N1) in figure 1 . The clock frequency can be varied for differ- ent difficulty factors (DF's?). When S5 is depressed, the outputs 'O', '1', '2', '3', etc. of IC1 (a decade coun- ter) are high in series according to the clock frequency. (The counter switches on the positive slope of the clock squarewave.) By using the outputs at '1' (pin 2), ’3’ (pin 7), '5' (pin 1 ) and '7' (pin 6) only, an equally long logic 0 follows every logic 1 of IC1. If all the switches SI . . . S4 are on 'c', four short signals are given which enable the low frequency oscillator /amplifier to produce four 'dots' through the loudspeaker. This is the morse code for the letter 'H'. As long as switch S5 is depressed, the decade counter (through the low frequency oscillator) will repeat this signal over and over with short pauses in between. If a switch is in the 'a' position however, the output of the corresponding pin of IC1 will be connected to an extra diode and an electrolytic capacitor C2. This prevents the clock signal from reaching the counter clock input (pin 14). The capacitor is discharged by R2 and Plb. The setting of Plb determines the time it takes to dis- charge C2. A dash is the result. .6.31 R2 - 39 k R7= 10 k Pla + Plb- 2x 100 k log. Capacitors: Cl = 1 p/IOV C2 = 20 THEN PRINT A: GOTO 1 ELSE 1 Instruction A = USR0(0) fetches the number of the actuated key. Its effect is similar to statement AS = INKEYS. The routine has a buffer with a holding ca- pacity of 128 key actions. The buffer is cleared by pressing keys [shift] and [clear] sequentially, while just [clear] empties the buffer until the last ”go” command. The [shift] key must always be pressed individually to ensure that the keypad routine assigns a different number to the next key pressed. The key numbers returned by the routine corre- spond to the numbers of the switches (1. . -23 inch; 24 is the [shift] key). The number is increased by 24 when the previously pressed key was [shift] (25... 47 inch). The keypad routine returns a nought to indicate that no key was pressed when it was called. R References: 121 16-key input for MSX micros. Elektor India, September 1987. Frequency doubler using 4011 This frequency doubler uses one CMOS quad, two-input NAND gate package type 401 1. The frequency doubler proper consists of an inverter N2. two differentiating networks R I / Cl . R2/C2 and NAND gate N3. N I and N4 function as input and output buffers. The incoming signal is buffered and inverted by N 1 , giving waveform ( A ) ( it is assumed the waveform has I : I mark- space ratio). (A) is inverted by N2, giving waveform B, which is the comp- lement of (AHi.e. it is in antiphase). The negative-going edges of waveform (B) are differentiated by R2 and C2. giving waveform (C), while (A) is differentiated by Rl and Cl, giving waveform (D). Waveforms (C) and (D) are fed into N3, and every time one of these waveforms is negative-going a positive-going pulse appears on the output of N3. (waveform E). The out- put of N4 is an inverted version of (E). The switching threshold of CMOS logic is about 45% of supply voltage, so the switching point of N3 on the rising exponential portions of waveforms (C) and (D) will occur at this point. The time taken for the waveform to rise to this voltage is just less than the time constant RC, so the pulse duration of waveform (E) is approximately equal to the time constants R1C1 and R2C2. For reliable operation these time constants should be chosen to be much less than the shortest possible period of the input waveform. The reason for this is that the width of the positive-going pulses (E) is constant, but the length of the spaces between them diminishes as the input frequency increases. If the pulses are not of short enough duration they may overlap at high input frequencies. 6.37 PLOTTER (part 1) mechanical design by J. Arkema Many owners of a personal computer and associated peripheral equipment will at some time have wished that graphics information available on screen could be sent to a plotter instead of a slow, noisy printer operated in the dot-matrix mode. But then, even the simplest of plotters is often more expensive than the computer itself. Not so the plotter described here, which is a unique project: fairly simple to build, complete with a versatile and power-efficient stepper motor interface driver, available in kit form, and offering a good price/performance ratio. The final accuracy of the plotter should be adequate for a host of graphics applications, and depends mainly on the constructor’s mechanical skills. are fine for text appli- it possible to keep the mechanical con- 1 not handle graphics in- struction relatively simple, while allow- well. They are invariably ing many paper sizes to be used. The the bit image mode pix- platen is driven direct by a stepper )ne line at a time. In ad- motor; the pen carriage indirect by a string and a second stepper motor. Arguably, in the absence of an absolute X-Y reference point, this arrangement has the disadvantage of being subject to accumulative positioning errors. For- tunately, deviations caused by these er- rors can be kept small in practice by en- suring that the paper and the carriage are not obstructed in their respective movement. Small electro-magnets are used for lif- ting and lowering the three pens. These are simply refills available from bookshops and warehouses, and come in various colours for many types of inex- pensive drawing pen. The pen carriage on the plotter is shown in the photo- graph of Fig. 1. The plotter described here is not of the X-Y type commonly found in the feast expensive class of commercially available plotters, but is similar to a standard text printer in that it has a pen carriage for the horizontal (X) direction, and a platen (paper roller) for the ver- tical (Y) direction. This approach makes former, is accomodated on the interface board to make for a compact and simple to connect unit. Mechanical construction The plotter is essentially a beam con- struction as shown in Figs. 2a and 2b. Two aluminium support plates (60 x 70x 2 mm) at either side hold the complete assembly in between them. The stepper motors are secured to the outside of the plates. Three round bars (dia. 6 mm solid aluminium or stainless steel tubing), and one square bar (10x10 mm aluminium) are fitted between the sup- port plates. The length of the bars deter- mines the maximum paper size, and can be dimensioned to individual require- ment. A length of 508 mm, for example, enables sideways drawing on A2 size sheets of paper, often used in the graphics industry. Details on the construction of all the mechanical parts for building the plotter are shown in the working sheet of Fig. 3. The paper roller is a round aluminium rod (spindle) of 12 mm diameter fitted immediately behind the square bar. Good grip on the paper sheet is ensured by reducing the diameter of the platen a few tenths of a millimeter over two lengths of 30 mm by turning in a lathe, and covering these areas in very fine sandpaper, wound spirally and glued onto the aluminium surface. Every pre- caution should be taken to prevent the total diameter of the platen increasing where the. sandpaper is secured. Two pressure rollers, fitted on a movable axle, rest on the sandpaper grips (see Fig. 4). To insert or remove a sheet of paper, the axle can be lifted by means of two small tilt levers made from aluminium U- beam. The rollers are firmly pressed onto the paper by the pull of two small springs. The platen is driven direct by a stepper motor with a step size of 200 per revol- ution. At the indicated platen diameter of 12 mm, this results in a resolution of 0.19 mm/step. Half-step operation is also supported by the driver board, in- creasing the attainable resolution to slightly less than 0.1 mm. The pen carriage is in essence a short length of aluminium U-beam. The guide rod runs through nylon slide bearings (Skiffy) secured in holes drilled in the legs of the U. Tilting of the pen carriage is prevented by its rear side resting on another rod. Carriage movement on the guide bar is effected with the aid of a string. This is wound one and a half turn around a shaft, up to the height of the securing screw, and then a further six to ten turns upwards. The shaft is made from the same material as the platen, and is fitted onto the spindle of the step- per motor. For optimum accuracy, the total diameter of the shaft plus string should equal that of the platen (12 mm). Fig. 1. The movable carriage on the plotter holds three pen refills in different colours. Not visible in the photograph arc the associ- ated electromagnets for pen up/down con- trol. This ensures equal pen travel per step in the X and Y direction. Each pen is guided through a hole drilled in the top of the U-beam, while the tips are kept firmly positioned on or above the paper with the aid of a support plate. The square aluminium bar in front of the Mechanical 1. side plate; left; aluminium; 60x70x2 mm. 2. side plate; right; aluminium; 60 x 70 x 2 mm. 3. angled support bracket for X motor; L-shaped aluminium; 20 x 40 x 4 mm; length 60 mm. 4. square connection bar; aluminium tOx 10 mm; 400 mm long' 5. round connection bar; aluminium/stainless steel rod; dia. 6 mm; 400 mm long'. 6. round support bar for pen carriage; dimensions as 5). 7. round guide bar for pen carriage; dimensions as 5). 8. round bar for pressure rolls: aluminium/stamless steel rod; dia. 6 mm; length 393 mm'. 9. platen; round aluminium bar; dia. 1 2 mm; length 400 mm 10. shaft; aluminium; dia. 11.2 mm Isee textl; length 25 mm. 10x10x1 mm; length 48 mm. 12. see 111. 1 3. angled support bracket for string wheel; U-shaped aluminium beam 15x16x2 mm; length 1 5 mm. 14. pen carriage; U-shaped aluminium beam 25x50x3 mm; length 60 mm. 15. pen positioning plate; aluminium; 8x50x2 mm. 16. pen carriage support; nylon; dia. 8 mm; length 23 mm. Miscellaneous parts: 6 off slide bearings; nylon; Skiffy 08-6. 1 off bushing for platen; nylon; Skiffy 08-4 or 08-6. 2 off washer rings; internal dia. 3 mm; thickness 2 mm. 2 off rubber pressure rolls le.g. cable grommet). 101. 6 off bolts M2. 6 x 5 (for fixing pen lift magnetsl. 1 6 off hexagonal nuts M3. 2 off springs for pressure rolls spindle, string; wound fishing line; 1 100 mm', fine grade sandpaper (for securing on platen). Electromechanical parts: 2 off stepper motors; 200 steps/rev.; dual-phase bipolar; 200 mA/phase (e.g. Berger as used in disk drives). 3 off pen lift electromagnets; 12 V; e.g. Binder Magnete Type 4O031-09B00. ’ Length in accordance with required size of plotter. Distributor of Skiffy products in the UK is Salterfix Fasteners e Salter Springs & Pressings Limited • Spring Road e Smethwick e Warley e West Midlands B66 1PF. Telephone; 1021 563) 2929. Telex: 337877. platen functions as a flat surface onto which the paper rests as the lines are drawn on it. An electromagnet to each pen arranges for this to be lifted from the paper when its colour is not required at a particular co-ordinate position on the sheet. The drawings and photographs in this article, in combination with the mechan- ical parts list, should give sufficient details on the basic construction of the plotter, which is reverted to below. Fig. 4. Close-up of the sandpaper grip on the platen, and the associated pressure roller plus clips on the movable axle. Also note how the 4-way Rateable to the carriage is wound on the rear rod to make a flexible connection with the plotter interface board. Circuit description of the plotter interface board The control circuit developed for the plotter is composed of a power supply, two stepper motor drivers, three buffers for energizing the pen lift solenoids, and an 8-bit digital interface to the Cen- tronics standard. The diagram of Fig. 5 shows that the cir- cuit is based around integrated stepper motor drivers Type MC3479 from Motorola or SGS. Three inputs of each chip, clock, full/half step and direction, are driven direct by the computer via the input connector. The fourth input, OIC, enables selection between high or low impedance termination of the energized stator winding during half step oper- ation. This selection is used for optimiz- ing the dynamic response of the relevant motor. The resistor connected to the SET input of the driver IC determines the stator current. In the non-activated condition, Ti and T; are turned off, so that the resistance between the SET in- puts and ground is relatively high. This effectively keeps the stator current be- tween 60 and 70 mA, ensuring a modest total dissipation of the motors and the driver ICs, whilst maintaining sufficient torque to keep pens and paper securely in position. A stepper motor is energized when the interface board receives a positive pulse transition on the relevant clock input (clc 1/2). The associated MMV is triggered, switches on the tran- sistor (T4 or Ts), and causes the stator current in the motor to rise to about 200 mA per phase. This rush-in current flows only briefly due to the inductance of the stator, and depends on the step rate. The driver IC, however, will counteract this reduction— within the practical limits of the supply voltage— to force a current flow of about 200 mA. Opening switch S2 disables the stepper motors to allow manual positioning of the carriage and/or the paper on the platen. In addition, opening S2 resets the logic circuitry internal to the driver ICs to the initial state, as indicated by the illuminated LEDs. This state occurs at each fourth (or eighth) step, and the LEDs will light correspondingly. The circuit for controlling the pen lift solenoids is relatively simple. Dvo-to- four decoder IC3 selects one of the three pens. When both IC inputs are held logic high, or are not connected, all three pens are lifted. The electromagnets are ac- tuated via darlington transistors and R- C networks. In these, the capacitor en- sures a relatively high pull-in current, while the resistor limits the hold current to a level that is just high enough to keep the electromagnet actuated. Flyback diodes are fitted across the coils to sup- press induced voltage peaks. Cutting, drilling, filing. . . The mechanical parts are made or pur- chased as indicated in the Mechanical Parts List and the working sheet of elektor India june 1988 6.43 Fig. 3. Each of the 16 parts that must be made to size is shown separately and with the relevant dimensions. The side panels are cut from a 2 mm thick sheet of aluminium. They are preferably clamped together and drilled simul- taneously to ensure accuracy. Use a centre punch for precise positioning of the drill, and lubricate this every now and then with methylated spirit to avoid burrs, and to clear aluminium shavings. A countersink drill should be used for sizing the upper hole in the right-hand support plate, which receives the head of the countersunk M4 screw bolt. The head must not protrude from the plate surface because this lies flat against the side of the stepper motor. Two ad- ditional 3 mm holes are drilled in the left-hand plate for securing the Y motor. Rods 4, 5, 6 and 7 are cut to equal size as required for the width of the plotter. The centre of both ends of each rod is prc-drilled with a 3.3 mm drill before tapping an M4 thread. Be sure to drill exactly in the centre of the axle: a lathe is, of course, ideal for this operation, but not strictly required for acceptable accu- racy. There is, however, no way to go round the use of a lathe for reducing the platen diameter where the sandpaper is to be secured. Unfortunately, a lathe is neither easy to control nor a commonly available tool. It is, therefore, rec- ommended to have the platen turned to the required local thickness in a mechanic’s workshop. Also remember that drilling an off-centred hole in the platen where it is secured to the motor shaft will wreak havoc by causing fric- tion in the nylon bushing at the other end of the rod, and, more seriously, ir- regular paper motion. The free end of the platen is turned to a diameter of 4.2 mm over 4 mm with the aid of a lathe to enable it to revolve in the nylon bushing. The shaft fitted onto the spindle of the second stepper motor is made by cutting off a short length of the platen tubing. The remaining compo- nents, 11 ... 16 inch, are relatively simple to make and require no further dis- cussion. Construction of the interface The plotter interface board is a single- sided type which is available ready-made through the Readers Services. Construc- tion is straight-forward by reference to the parts list and the component overlay of Fig. 8. Resistors Rio, Rn and Ri: are 1 W types mounted slightly off the board surface to aid in their cooling. Drivers ICi and IC2 require fitting with a DIL clip-on or glue-on heat-sink— see Fig. 6. It is recommended to solder the ICs direct onto the board, so that the ground area can aid in convecting dissi- pated heat. ICj is a type from the 4000 series, and has the disadvantage that its input is not TTL-compatible. In prac- tice, however, pull-up resistors Rj and R20 ensure correct operation of the in- terface in conjunction with virtually any Centronics port. None the less, when in- compatibility problems are suspected (the motors then appear not to start properly), ICj may be replaced by an equivalent of the HCT type. Jumpers JPi and JP2 are not fitted. As already stated, the input header on the board is wired in a manner that facilitates connection to a Centronics port via a length of flat ribbon cable. Handshaking is not used in this arrange- ment. of the plotter. This is a relatively wide version (508 mm) mounted on . plate for improved paper stability (A2 sideways; A3 lengthwise). It has four paper grips to ensure that lines of minimum thickness are drawn reliably and accurately. Fig. 6. The stepper motor driver ICs on the interface board require cooling by heat-sinks of a type that can be glued on top of a DIL package. A note on stepper motors The stepper motors used for building prototypes of the plotter were Berger types salvaged from discarded disk drives. Similar types may be used if these have the following specifica- tions: • 200 or 100 steps/rev. (ai. 8° or 3.6° per step); • current consumption: approx. 200 mA/phase; • two bipolar phases; • resistance for each phase (stator): 30. . .40 Q. Unfortunately, many motors may not meet with the above specification in respect of the stator resistance. These types exhibit much lower values (e.g. 1.33 £2), and require driving from current sources. According to the manufacturers, the driver ICs Type MC3479 on the plotter interface board have current source output stages, but in practice these may be damaged when connected direct to a very low stator resistance. It is, therefore, good practice to check the stator resistance of the motors en- visaged for use in the plotter. If necessary, add a suitably dimen- sioned series resistor to ensure that each stator output on the driver ICs is loaded with 30. . .40 Q. Example: use a 33 Q; 4 W resistor when the stator alone has a resistance of 1.33 Q. 6.45 The completed board is secured onto the safest and easiest alternative in this mains transformer and the plotter. The respect. Indicators Ds and Ds need not two switches are connected as external be visible when the interface is fitted in controls, together with the mains switch, an enclosure, although this may prove Do not forget the mains fuse, which useful during testing and setting up. should be connected ahead of a suitably rated DPDT mains switch. A Euro-style Part 2 of this article will deal with mains entrance socket with integral general condiderations on control soft- switch and fuseholder is, of course, the ware for the plotter. 6.47 ARTIFICIAL INTELLIGENCE by M. Seymour, BSc, University of St. Andrews Man is intelligent, but his intelligence Is often thwarted (or worse) by his environment. That realization has given rise to a dream- thini, ?£o* t?aV ?J av . b S possible to builb a machine that can think, that is, need not be programmed to perform its functions Machine intelligence was first thought of by Charles Babbage (1792-1871). This century, Alan Mathison Turing (1912- 1954) has achieved immortality through the Turing Machine, which purports to show that machines are mathematical objects,, and his proof of the theorem that the set of mathematical tasks that is computable is exactly the same as the set that can be executed by the machine. He also formulated a theory about what questions could in principle be answered by an intelligent machine. Artificial intelligence grew out of the work on digital computers during the Second World War and was officially recognized as a branch of computing science in 1956. Since those early days, artificial intelligence has given rise to a number of myths, particularly, but not only, in the popular press. However, claims of computers achieving this and that, without human intervention, always prove, on close examination, to be mere illusions of intelligence. These illusions are created by the fact that com- puters work so extremely fast. Fortunately, such illusions are now recognized as such and the true science of Artificial Intelligence is taking shape. The aim of this science is not to create machines that are as intelligent as humans (it is doubtful whether this will ever be possible), but rather to create machines that are ever more capable of meeting human needs. Such machines need to be able to learn about their users and to do that, they will have to see, hear, and understand. Moreover, they must not make demands on their users as far as programming is concerned. Ideally, this would mean that the ma- chine responds to normal spoken language. Only limited progress has been made in that direction. The theory of artificial intelligence says that if you analyse the world in symbols and put the right rules (= program) in the machine it will have a mind and understand the world as we do. In other words, it will be a thinking machine. Some researchers say that any computer can simulate any other computer, be it serial, parallel, digital, or analogue. Assuming the human brain to be a com- puter, they assert that it must be possible to find the program that makes the human computer function. Once found, this program can then be used to run any other computer, which will consequently be intelligent. But is there such a program and will it ever be found? Other researchers feel that because of the world we live in, we suffer from the illusion that every substantive problem should have a technological or scientific solution. Because such answers are not forthcoming, we believe that a better technology or a more advanced science is required to solve the problems. It is easy to think of artificial intelligence in this context. These researchers do not believe that it is simply a matter of calling our brain a computer and saying that it has been ‘programmed’. Human intelligence is not just the ability to think logically: there is memory, experience, background, emotion, and so on. It is, perhaps, experience that creates the greatest distinction between artificial and natural intelligence. The relation- ship between human mental events and experience of the real world is called in- tentionality. It is these researchers’ con- tention that no existing machine or program has intentionality. Artificial in- telligence is, at present, proceeding along the lines laid down by these re- searchers, that is, with the aim of con- structing a machine that is most capable of meeting human needs. The most publicly visible applications so far are so-called Expert Systems. These arc programs that arc used to give advice on medical diagnosis and prescription, genetic engineering, chemical analysis, and geological prospecting for minerals and oils. Although most of these pro- grams are very limited in what they can do, some give more reliable advice than all but the very best human experts. Computers developed for Artificial In- telligence are called fifth generation computers. The first four generations are defined in hardware terms: machines based on valves; transistors; integrated circuits, and VLSI. The fifth generation is defined in terms of parallel operating hardware and artificial intelligence. Although the governments of all western nations, and, no doubt, that of the USSR also, are pouring money into the research and development of fifth gener- ation computers, workers, in the field have found that the difficulties involved have been grossly underestimated. One of the important facts that has emerged is that human common sense plays a far greater role in our daily lives than hith- erto generally imagined. Common sense enables people to cope with the fact that a statement assumed to be true at one time can later be found to be false. Computers, which depend on traditional logic, wherein truths are proved once and for all, can not come to grips with this — at least not yet. Everyday abilities like talking, seeing, or' sensing friendliness from a facial expres- sion, do not normally need conscious ef- fort. Nor can we say how we do them. As long ago as the 5th and 4th centuries BC, Socrates and Plato concluded that although early on in life we learn the rules (i.e. are programmed) for these functions, they are quickly forgotten and retained only in our subconscious mind once we grow up. None the less, these functions are far from simple. Indeed, their complexity — and subtlety — was The Chinese Room John Searle, Professor of Philos- ophy at the University of Califor- nia, defines two forms of arti- ficial intelligence research. Weak Al merely says that the principal value of the computer is that it gives us a powerful tool in the study of the mind. Advocates of Strong Al, however, maintain that the digital computer is not merely a tool, but rather, if cor- rectly programmed, a mind that can literally be said to understand and to have other cognitive states. Searle believes that the ‘equation’ mind/brain -pro- gram/hardware is invalid. Imagine yourself (it is assumed you are not a Chinese speaker) in- side a closed room. Your only contact with the outside world is through a small hatch. In through the hatch comes a large batch of Chinese symbols. Some time later, another batch arrives, along with a set of rules in English (which, it is assumed, you do understand) correlating the first set with the second. You carry out the instructions, no doubt wondering what you are doing, and another batch is sub- mitted, with more rules in English. These rules tell you to check this third set of symbols against the first two, and to send back out certain symbols from the first large batch as a result of this checking process. After a time, you get so proficient at this process that it seems as though you are fluent in Chinese. The third lot of symbols, you see, are questions and those you send out are answers to these ques- tions. The point here is that, although you understand the English rules, you have no idea whet the Chinese symbols mean. Searle asserts that the Chinese Room is analogous to the way the digital computer works. Alan Turing showed that any computer can be reduced to a Turing ma- chine, which merely manipulates symbols according to a set of rules. A computer running a program does precisely, and only, that. It can not, therefore, be said to understand what it is doing. In his original paper, Searle gives I six replies to his argument and Outside the Room 'the programmers’ p3=«ttOT^V| |@U,'00N<£ p3=» ROT/V ®U,'®0N/E 14. Il3appeais«- Buiiwwtewi third batch p3=* XOTZV @U,‘©0N/€ -THE PROGRAM' — —'QUESTIONS' — set of rales in English 34. Ilgappua SSKISSlZ &\aoussi — 'MORE PROGRAM' — ► . 'ANSWERS TO THE QUESTIONS' p3-*xorzv; ®U,'®0N/€ Searle thinks that the computer is ! simulating only part of the brain: the neural structure, but that is not enough. You could make a brain out of beer cans in a similar way and it would not think. In the context of the Chinese Room, connect the input and the output of the room via a system of water pipes and valves that are struc- tured in the same way as the neurocortex. Yet again, I do not understand Chinese. The Combination Reply. Com- bine the three previous responses so that we have a robot with a computer in its head, pro- grammed with all the synapses, etc. of the human brain, and think of the whole thing as a unified system. Wc would have to ascribe intentionality* to the sys- Schematic diagram of the main features of Searle's Chinese Room argument. The symbols used are, of course, not Chinese, but hopefully convey the point that taken together they mean nothing Searle thinks it is not sufficient just to have something that looks and behaves like us. The Other Minds' Reply. How do we know if anyone else understands Chinese? Well, by their behaviour. According to Searle, presuppose the reality and knowability of the mental in the same way as we presume the re- ality and knowability of the physical world when we do phys- ics. Obviously, this is another contentious issue in the light of counters them. In the journal The Robot Reply. Maybe the somc interpretations of quantum in which the paper appeared, computer on its own does not theory, for instance, but Searle twenty-seven further replies were understand, but put in the cranial gives it short thrift, given. We shall only deal here cavity of a robot, give the robot The Many Mansions Reply, with the original six, but it may manipulating arms, sensors, and Although wc arc working with be noted that although the paper a television eye, and it will under- analogue and digital computers was published in 1984 the argu- stand as we do. now, in future we could well ment is still hotly debated today. No, says Searle, for just imagine build devices that have the causal The Systems Reply. It is wrong to yourself and the room inside the power of the brain, of which in- think of me understanding robot's head. I can see the sym- tcntionality is a product. Chinese: the whole room is the bols through the camera, I can Searle agrees wholeheartedly, but understanding agent. manipulate them with my mcch- q j s no argument here, because This seems at first glance anical arms, but, of course, 1 still suc h a machine would not fall somewhat incredible in terms of don’t understand. within what we now define as Al, the Chinese Room, but it is The Brain Simulator Reply. If which, in its strong form, is the motivated by the feeling that my you get the computer to actually on |y thing with which he is con- brain understands nothing, but simulate all the neuron firings at cerned. that / do. the synapses of the brain (in par- Searle says that all one needs to ticular, of the brain of a native * Intentionality, do is "internalize” the rules, i.e., Chinese speaker), how can the LVwhiduh'cT a rejected at o memorize them. But, one still computer fail to understand (s anli y slates 0 f affairs does not understand Chinese Chinese — it seems as if wc wor | d Thus, beliefs, desires, then; merely following a set of would have to deny that the tenl i 0 ns (as we commonly u rules one has memorized is not to Chinese speaker understood, if word) are intentional statt understand the meaning of what we thought otherwise. directed forms of anxiety, they are about. sions, and so on, are not. not appreciated until researchers tried to model them on computers. Although the activities that involve our common sense have so far proved an in- surmountable obstacle to artificial intel- ligence, there are as already stated expert systems that can give advice on a num- ber of problems faced by genetic engin- eers, physiologists, geologists, and math- ematicians. The widespread ignorance of this, perhaps not so surprising, paradox may present a danger because most people lack a reliable sense of which questions Al systems can at present be expected to cope with. Current development work indicates that one of the important aspects to be tackled is to give fifth generation com- puters a much better grasp of natural language. Most people are not prepared to learn a special pogramming language. Moreover, the machines would not be able to learn about their users through normal conversations. However, the dif- ficulties are great and at present it is ex- pected that future expert systems will have a command of the natural language only within the subject for which it was designed. A technology that was originally developed in the 1950s, but was aban- doned after ten years, has recently been revived. It is called neural computing and could be of inestimable value in the creation of true artificial intelligence. Neural computers attempt to copy the human brain and are quite unlike con- ventional computers, because they are not programmed, but can learn by example. The brain is a complex network of in- terlinking neurons. It is the interlinking that is the key to solving problems quickly, but it is a problem in computer engineering: to create a 1 million-node network with 1 billion ‘hardwired’ inter- connects would require 92 m 2 of silicon. The Oregon Graduate Center’s Com- puter Science and Engineering Depart- ment is planning to build a neural com- puter with 10,000 nodes linked by five million interconnects. However, by using frequency-based encoding for the inter- connects instead of hardwiring only 0.8 m 2 silicon is needed. Neural computers may help in solving problems that still defeat conventional computers in spite of the enormous in- crease in processing power made poss- ible by VLSI technology. These prob- lems include pattern processing tasks, such as speech recognition, and the cre- ation of content addressable memories. References: 1. Minds, brains, and programs. Behav- ioural and Brain Sciences 1980, 2. By Prof. J. Searle. 2. Principles of Artificial Intelligence. By N.J. Nilsson (1980). 3. Artificial Intelligence and Natural Man. By Prof. Margaret A. Boden (1977). 4. Expert Systems in the Microelectronic Age. By Prof. D. Michie (1979). 5. Intent ionaUty: Essays in the Philosophy of Mind. By Prof. J. Searle (1984). 6. Alan Turing, The Enigma of Intelligence. By Andrew Hodges (1983). 7. What Computers Can’t Do (A Critique of Artificial Reason). By Prof. H.J. I Dreyfuss (1972). 8. Expert Systems Technology: A Guide. By I L. Johnson and E.T. Keravnou (1983). 9. A Perpective on Intelligent Systems. By L.J. Kohout (1985). 10. The Future for Artificial Intelligence. Elcktor Electronics, November 1985. By Prof. Margaret A. Boden. 1 1 . How Does The Human Computer Work? i By Dr. Kevan Martin. Elektor Elec- I ironies, May 1987. 12. Is the Chinese Room the Real Thing? I Philosophy, 62, 389-393. By D. Ander- son (1987). 13. Minds and Brains without Programs, by J. Searle in Mindwaves (1987) eds C. I Blakemore and S. Greenfield. SIMULATING SIGHT IN ROBOTS by Arthur Fryatt, CEng, MIProdE Although industrial robots have been in widespread use for well over ten years, their inability to respond intelligently to unexpected or rapidly changing situ- ations has limited their usefulness to tasks in highly ordered environments. The problem is the robot’s lack of awareness of what is happening around it. Attempts to solve this problem have led to the development of sensory systems that in some measure emulate human vision, touch and hearing. Most research has concentrated on the design of com- puterized vision systems which act as the robot’s eyes and brain to provide a basic form of artificial intelligence. The major parameters of robot vision systems are recognition, location and in- spection. With this information, a robot knows what components are present in its workspace, where they are positioned, and the extent to which they are dimen- sionally or structurally correct. Although it can be seen that the develop- ment of vision systems is extending robot technology into inspection and as- sembly, some of the most promising commercial developments have occurred in paint spraying, welding and colour quality control of items such as fruit and vegetables. Co-operation between British research institutions, universities and manufacturers is increasing the range of commercial applications. Practical Research One of the leading research institutions in the United Kingdom for the develop- ment of vision systems is the National Engineering Laboratory (NEL) (1) which 6.50 elektor indie iune 1988 has designed a considerable amount of software for manipulating and inter- preting images. For scenes that display high contrast between components and their backgrounds, a simple thresholding operation will convert the grey scale in- put array into a binary image in which each pixel has the value 0 (background) or 1 (component). Binary images can be efficiently stored in a computer memory and their simple format enables fast analysis to be carried out to determine dimensional and topological measurements. These values, along with other invariant features, can be used to build a simple component recognition and location strategy that will operate effectively on uncluttered scenes. Reliance on high contrast effectively precludes the use of binary processing techniques in most engineering appli- cations, which are typified by visually ’’noisy” conditions such as poor light levels, low contrast, or components lying jumbled together in bins partially obscured by other workpieces. In such situations an alternative ap- proach to recognition is based on match- ing local features (boundary segments, corners, holes and so on) rather than on matching global feature values (area or perimeter length, for example). The NEL has recently developed advanced techniques for the matching of local features involving the latest computer- on-a-chip device. A practical example of robot vision work at the NEL is a recent project under- taken for the National Nuclear Corpor- ation involving the development of a sys- tem for automatically locating fuel pins in a nuclear fuel assembly. Accurate information on pin position is communicated to a robot which grasps and removes each one in turn. The NEL system is ten times faster than manual dismantling. University collaboration A vision sensing system provides colour quality control for grading fruit and vegetables in the Autoselector, a joint de- velopment involving the Essex Elec- tronics Centre* 2 *, a department of the University of Essex and Loctronic Graders* 3 *. Their collaboration initially led to the introduction of the Autoselec- tor A, which employed a monochrome television imaging technique to detect differences in the grey scale. Subsequently, with the introduction of the Autoselector C, a very significant ad- | vance has been achieved with colour im- aging which enables up to 4096 colours and shades to be identified in areas as 1 small as 3 mm diameter at very high speed. Since the entire area of the product needs to be scanned, Loctronic Graders , has developed the Thrudeck which presents constantly revolving products such as tomatoes, onions, kiwi or citrus \ fruits at speeds up to 2500 per minute to ; the camera. Even though the products i are of irregular shape, the system can track, size and count each one as it I follows a meandering path down the deck. Another interesting technical achieve- ment is the way in which the three- j dimensional aspect of colour television i is handled. Since a colour camera has three channels — red, green and blue — the permutations possible could be handled only by a very large computer. In conjunction with the Electronics Systems Engineering Department of Essex University, the Electronics Centre developed a method simplifying this task so that it can be handled by hardware controlled by a Motorola 68008 micro- processor. By selecting the region of colour hue carefully — for example green and brown for potatoes, or red and green for tomatoes — and examining tone satu- ration in the chosen colour sector, di- mensions are reduced from three to two, which can be handled relatively easily. Sighted robot welding Founded at the beginning of 1984 with the help of an Oxford University research team, Meta Machines* 4 * is now accepted throughout the world as a leading commercial organization specializing in sensors for robot control. Its MetaTorch adaptive vision guidance and control for arc welding is designed to ensure that a robotic welding system achieves consistently high quality output despite components fit-up variations and inaccuracies. The aims are minimum downtime for reprogramming in response to component batch changes, and maximum flexibility to adapt to future changes through the fixing of simple and inexpensive parts. The company’s two most recent developments are the MetaTorch 200, a compact vision sensor mounted co- axially around an MIG or TIG welding torch and the MetaTorch 500, for higher current applications, on which the vision sensor is mounted external to the welding torch. The Metatorch system can recognize complex joint types, guiding the robot to locate, track and weld the seam in a single pass operation. The vision processing eleci'onics and powerful vision processor enable the sys- tem to analyse the position of the joint and communicate this information to the robot controller at a rate in excess of 10 Hz. As a result of its single pass oper- ation and fast vision analysis, the system has no significant effect on the robot Used in production environments, the MetaTorch requires no optical adjust- ment or alignment and is quickly inter- changeable. By combining a solid state laser light source and camera in a single unit, it is capable of withstanding harsh operating conditions. Precise spraying By combining closed circuit television with automatic paint spraying equip- ment, Lektrodesigns* 5 * has developed the Videospray system, which can assess separate paint stroke requirements. It controls spray patterns individually so that irregular shapes loaded on a con- veyor at random will be painted auto- matically with a minimum of paint. Mounted together on a single stand, the Videospray’s closed circuit television (CCTV) camera with built-in monitor and electronic module are easily in- stalled adjacent to existing spray equip- ment. Any reciprocating gun can be controlled and one unit can handle up to four spray guns. To establish the relative positions of the spray gun and the item to be painted, the camera is directed so that the reflector on the spray gun, and the workpiece as it enters the spray booth, are in view. From this relationship, timing instructions are generated and stored in the logic bank to control the spray stroke, ensuring paint economy. It is possible to achieve an accuracy of paint spray to 12.7 mm with the elec- tronic and mechanical time lag provided by the system, even to the extent of com- pensating for angular workpieces where the gun needs to be rotated through an arc. The complete Videospray instal- lation occupies only 0.8 m ! of floor space and on average rises to a height of 2 m. The company’s latest development is the Videospray II, a shape recognition sys- tem again comprising a CCTV camera linked to a microprocessor, which in turn can be connected to a painting robot to call up the appropriate part painting program. A particular feature is the method of lighting the part moving on a conveyor to give a strong silhouette for the camera to view. A microprocessor digitizes the outline shape and compares it with a pre-stored library of shapes to determine the part number. Outline recognition software routines have been incorporated to determine the attitude of parts on the conveyor, for example, higher, lower, tilted, retarded or advanced, compared with their mean positional attitude. References: 1. National Engineering Laboratory, East Kilbride, Scotland G75 0QU. 2. Essex Electronics Centre, Wivenhoe Park, Colchester C04 3SQ. 3. Loctronic Graders Ltd, Eves Corner, Danbury, Essex CM3 4AH. 4. Meta Machines Ltd, 9 Blacklands Way, Abingdon Industrial Park, Abingdon, Oxfordshire OX14 1DY. 5. Lektrodcsigns Ltd, c/o Market Op- tions Ltd, 75 Middle Gordon Road, Camberley, Surrey GU15 2JA. Polarity indicator circuit loading by this device. However, when checking circuits with high impedances, the loading effects must be taken into account. The opamp compares the voltage of the test point relative to common. If the voltage is positive so is the output of the opamp. As a result LED D1 will light. If the voltage is negative, LED D2 lights up. The pin numbers given on the drawing are related to a TO-5 case. 6.51 MAKING THE WEATHER WORK FOR YOU The past 10 years or so have seen what are probably the greatest advances ever in weather forecasting. Certainly they are the greatest in terms of potential value of forecasts to commerce and industry. by Dr John Houghton, Director-General and David Houghton, Marketing Director, UK Meteorological Office, Bracknell First to appreciate how big are the benefits from better weather forecasts have been the world’s airlines, tradition- ally the customers who work most closely with the meteorologists. This was highlighted last year when a team of scientists from the UK Meteorological Office received the prestigious Royal Society ESSO Energy Award, in recog- nition of their pioneering contribution to energy saving through developing the world’s best operational weather fore- casting model. By international agree- ment the world’s airlines have access to flight forecasting information for all parts of the world from the Meteorologi- cal Office headquarters at Bracknell, typically information on winds and tem- peratures at all the heights at which air- craft fly. Airlines using Bracknell data together spend over £5000 million every year on aviation fuel, so the saving of one per cent or more in fuel consump- tion which can be achieved through using the better forecasts is not incon- siderable. Forecasts contribute to fuel saving in a variety of ways. For instance, a forecast can indicate the position of the strongest tail wind so that the aircraft can be flown to take advantage of it and thereby achieve a higher speed relative to the ground, saving both time and fuel. Forecasts of temperature are also im- portant, for air temperature influences the efficiency of the jet engines. It is possible only to conjecture on the much larger savings achieved in comparison with using no weather forecasts at all. Global approach Airlines are interested only in short period forecasts, for up to 24 hours ahead at the most. However, recent im- provement in weather forecasting is even more evident in predictions for two, three, four and five days ahead. Until 1971 forecasts for more than two days ahead were of little practical use, cer- tainly not in commercial terms. Then a new 10-level forecasting model was in- troduced, and the quality of forecasts for days two and three rose dramatically. But the model was only hemispheric: there was neither the computing capacity 6.52 elektor India june 1988 nor the observing system capable of sup- porting a global model; indeed, it was thought at the time that over a few days the weather over one hemisphere was largely independent of the weather over the other. The latest global models of at- mospheric behaviour have demonstrated that it is not so. Accurate forecasting beyond a day or two demands a totally global approach, which has been made possible through parallel advances in computing, new observing systems using both geostationary and polar orbiting satellites, and mathematical modelling of the global atmosphere. The first illus- tration shows how all these elements combine to provide a global forecast. The result is that today’s forecasts for three to six days ahead are better by two days than were forecasts only 10 years ago. The acid test in any forecasting system is its ability to predict change. In the sec- ond illustration this test is applied to forecasts produced here for the area covering Europe and the North Atlantic. The element tested is surface pressure, the forecast surface surface pressure against the observed surface pressure. It can be seen that forecasts for day three are now as good as those for day one were 10 years ago, day four as good as day two, and so on. Many sectors of industry and commerce can also derive considerable financial benefit from these better forecasts. But this benefit is yet to be fully realised because the forecaster is still largely Steps from observation to forecast. Numbers indicate the average number of observations each day for use in the main 0000 utc and 1200 utc global numerical forecast model runs. unaware of the user’s requirements, and the potential user is still largely unaware of what information the forecaster can provide. What is required is a marketing dialogue similar to that which has been taking place between the meteorologists and the aviators for some 70 years. The aviator has not been slow to tell the meteorologists what information he re- quires, and the meteorologist has re- sponded to the best of his ability by deriving increasingly better methods of producing and communicating the re- quired information. Ship routeing The shipping industry are even older customers for weather information: in fact, the Meteorological Office was set up around 1855 specifically with the pur- pose of giving warnings of storms to ships at sea. Nevertheless, in contrast to airlines, shipping companies have been slow to appreciate the opportunities for saving time and money by using special- ist routeing advice. Only a small propor- tion of shipping uses a ship routeing ser- vice, although the benefits have been shown to be great. The diagram on the front page shows the routeing of a ship across the Atlantic for a minimum- time crossing. The saving achieved by avoiding adverse weather was 14 hours. Ships can also be routed on the basis of minimum wave height if, for instance, the comfort of passengers or animals is paramount. Offshore drilling operations, especially from floating platforms, are the most weather-sensitive activities at sea. High winds and swell can be particularly dangerous for drilling and diving oper- ations, or when platforms are being moved or towed. The cost of operating a platform is high, of the order of £1 million per week, and the value of ac- curate forecasts of weather and waves for a few days ahead is also high. Fore- casters working on oil rigs are a vital part of the operational team. One of the better known forecasting suc- cesses of 1986 was the record breaking achievements of the Virgin Atlantic Challenger. An accurate four-day weather forecast was essential, and the signal to go followed a favourable forecast from our Central Forecasting Office. The value of a weather forecast to the aviator and the sailor is almost assumed, because they are open to the elements. The farmer is another obvious customer, though traditionally he has been re- garded as so experienced a weather observer that he can rival the pro- fessional forecaster. But even for him things have changed. Now, for the First time, he is able to schedule many of his farm operations several days in advance. Armed with specialist forecasting and climatological advice, he can sow his seeds, apply his fertilizers and preventive sprays, and harvest his crops, all at times to get the best yields. Matching forecast to demand As with other marketing business, the First aim in marketing weather forecasts is that the provider and the user be brought together to their mutual benefit. The product is weather information, both historical and forecast, specified and assembled to meet the needs of the user. Experience has shown that the ma- jority of users of weather information cannot afford either the time or the ef- fort to glean the information required for a particular operation by attemping to interpret a general weather summary and forecast. The full beneFits which may come from weather information are reaped only when the information is tailored to the particular requirement. For example, anyone who sells umbrellas is interested only in whether it will rain at a time of day when people are likely to want to be out of doors. A manufac- turer of heating equipment is vitally con- cerned with temperature and, because it takes several days of cold weather before demand is stimulated, he wants a package comprising both historical and forecast information. The food retailer requires a particularly complex and com- prehensive package relating the expected weather to various factors, for instance to variations in demand for a wide var- iety of foodstuffs, to variations in their availability if they are grown in the field, and to the ease with which they can be transported and stored. To benefit most in such cases, forecast and historical in- formation have to be assembled so as to relate as closely as possible to the pro- cedures for making decisions throughout the industry. The second essential marketing con- sideration is how to convey weather in- formation efficiently to the user. Should it go by digital link, facsimile or tele- phone; to the company headquarters, the local office or the building site? If the means of communication or the destination are wrong or inappropriate the information may be of little practical value. Third, the price must be right. Price set- ting is the part of the marketing mix which many scientists will try to avoid as being at best unscientific and at worst immoral. But in this context it is not only necessary to pay for the resources devoted to the provision of the service; the price must also relate directly to the perceived value of the service. The reci- pient of specially tailored weather infor- mation is much more likely to use it to advantage if the cost bears some sensible relation to the benefit which may be de- rived. It is not that the price needs to be elektor India juna 1988 6.53 high — just realistic. In practice, the price of most meteorological services is only a small percentage of the benefits derived, which makes a weather service a necessity rather than a luxury. This ap- plies not only to individual services but on the scale of national meteorological services, too. The contribution of every such service to a nation’s economy is many times the cost of the service. Public communications The value of weather forecasts for the public at large through newspapers, radio and television must not be overlooked. They enable people to make millions of small decisions that con- tribute significantly to the well-being of the community and to the efficient and effective use of its resources. For the man in the street the weather forecast may sometimes do no more than satisfy his curiosity regarding the future. On certain occasions it contributes signifi- cantly to his comfort and convenience. He can go out wearing suitable clothes and footwear, and not carrying an um- brella unless he really needs one. On at least a few days each year the fore- cast confers a measurable benefit. For example, he may save fuel by avoiding a leisure trip to the coast or mountains that would have been spoilt by bad weather. Just as for specialist users, the amount and quality of weather infor- mation useful to the average citizen has increased greatly in the last few years, so much so that new means have had to be found for him to reap the benefit and for the meteorological service to reap some rewards. In Britain, a new telephone in- formation service known as Weathercall has recently been introduced. It costs more than the average rate of charge for a call and a proportion of the charge to the subscriber is paid to the meteorologi- cal service. A similar service, Marinecall, is available to inshore sailors. It provides, detailed forecasts over 15 consecutive telephone numbers, each for a sector of the coast around Britain. There are, of course, many other areas of industry and commerce where weather forecasts properly applied can contrib- ute to profitability and efficiency. In the power industry weather information is essential to short and long term plan- ning; highway authorities make huge sayings during the winter by applying grit and salt only when there is advance warning that it is necessary. In building, construction, transport, manufacturing, maintenance and repair, many activities and processes are weather sensitive; and the demands for goods and services vary with the weather. The World Meteoro- logical Organization is working to ob- tain better estimates of the benefits of weather services throughout the world, both to communities as a whole and to individual sectors. In the UK, a conser- vative estimate is that the total benefit is well over 10 times the total cost of the service. Today, computer models give good de- tailed weather information up to five or six days ahead. As the models improve and as larger computers become avail- able, the period of forecasts will perhaps be extended to 10 or even 14 days. The question then will be how predicatable is the atmospheric circulation beyond two weeks ahead? Almost certainly it cannot be forecast in detail. Nevertheless, there is a good possibility that the average character of the weather a month or two ahead may be predictable. The economic value of such predictions, even if they are not perfect, would be very large. So the stakes are high and the world is waiting on the meteorologists to tackle the problem of weather forecasting at longer range, for which they will require at the very least the next generation of supercomputers. All told, one thing is certain, the weather service is no longer seen as a luxury but a necessity. Annual average correlations between 24, 48, 72, 96 and 120 hours forecasts and actual pressure changes at sea level Tor the North Atlantic Region since 1976. Liquid level indicator This circuit was originally intended as a water level indi- cator for use by blind persons, to give an audible indication when a cup, bowl or other container was full. It will function with any liquid that will conduct electricity, such as beer, tap water, tea, milk. It will, of course not function with distilled or de-ionised water. The circuit has other applications such as a rain sensor (when used with a suitable probe). The circuit is extremely simple. The in- put of N1 is normally held low by a 1M resistor. When the probes are im- mersed in a conducting liquid the input of N I goes high, so the output goes low and the output of N2 goes high, en- abling the astable multivibrator N3/N4. which switches T I and T2 on and off to produce a tone from the speaker. An open collector transistor output is also provided to drive a relay or other cir- cuit. Probe construction for level sensing and for rain sensing are shown in figure 2. The level sensor probes should preferably be made of stainless steel wire for ease of cleaning, and the circuit housing should be watertight in case of 'HI 6.54 Distances can be measured in three via gate basic ways : by using a ruler, a tape only be a measure, or by electronic means. The output ol first two are often time consuming and and N4, prone to inaccuracy and the third is the start t relatively expensive. A fourth possi- If the st< bility is now provided by none other clock pul than this canometer, its prime merit any read being simplicity. the displa It is, in fact, just an electronic meter will ther incorporated into an empty can and start key I which indicates the measured distance is rolling, on a digital display. circuit. The principle behind the circuit could All the ' hardly be more basic. When a cylinder the pulse is rolled along a flat surface, the distance multiplex that it covers during one revolution will displays) be the same as its circumference. If cathode the diameter of the cylinder, or can. decimal canometer revolutions the can has made, the dis- tance covered can be calculated very easily. Therefore, the electronics in- volved only has to record the number of revolutions, calculate the distance travelled and display the result. How- ever, with a little bit of ingenuity, we can dispense with the mathematical section altogether, thereby making the electronics even simpler. The diameter of the common 0.33 litre can is approximately 65.8 mm. This can be increased to 66.8 mm by adding a layer of sellotape around the can. This will then bring the circumference of the can up to 21 cm, a nice whole number. To detect how far it has travelled, the can utilises a disc which has 21 segments. Each segment consists of a dark and a transparent section. The disc can revolve on a central axis and is weighted with either lead or solder so that the weighted portion will remain face downwards when the can is rolling. A photo detector construction from an LED and a photo transistor, on either side of the disc, is incorporated so that the transistor generates a pulse each time a segment rolls past. Since there are 21 segments and the can has a circumference of 21 cm, you don't need a degree in mathematics to work out that each pulse from the photo transis- tor corresponds to 1 cm of travel! Effectively, the can rolls together with all its contents (circuit boards, battery etc.) but the disc stays still. measuring can S. Heilmann Construction The complete circuit can be mounted on two printed circuit boards as shown in figures 2 and 3. The round display board is mounted at right angles to the rectangular board after which the corresponding connections arc made with short wire links. The axle for the segmented disc is mounted on the rectangular board in the position marked by a dotted line. The space with the triangular points next to it has to be removed first with the aid of a fretsaw. This is essential so that the disc will be positioned correctly between the LED and the photo transis- The fact that a can has volutational properties was the inspiration behind this particular design in which the roll plays a very important role. The circum- ference of the can is used to measure distances with a fair degree of accuracy. The distance covered, or rather, rolled is indicated on a four digit display, the maximum readout being 99.99 metres. The segmented disc is illustrated in figure 5. In fact this can be carefully cut out and glued onto a circular piece of perspex with the same diameter. This is then weighted with lead or solder, bearing in mind that the completed disc must be able to pass through the slot in the board without touching the sides. Once it is mounted on the axle it must be able to turn freely. Prior to final assembly in the can, preset potentiometer PI should be adjusted so that the display increments by one each time a segment of the disc passes through the light beam (after S2 has been depressed of course!). The only other requirements are that the can needs to be 'screened' so that no light can enter to cause a misreading and Circuit diagram The circuit diagram of the canometer is shown in figure 1 . The photo detector mentioned previously is made up from LED D2 and the photo transistor T2. Whenever the light from the LED passes through transparent section of the disc transistor T2 will generate a pulse. This pulse is shaped by the schmitt trigger N1 and fed to the clock input of IC2 6.55 6.56 the axle for the segmented disc must be positioned in line with the central axis of the can. Operation The directions for use as given by the author are so clear there is no need to add anything to them: ' Switch on the canometer * Place the can at the starting point of the distance to be measured, press 'reset' followed by 'start' * Roll the can as evenly as possible until the 'finish', then press 'stop', pick up the can and read off the dis- tance measured from the display * Note that the unit will only work correctly when the can is rolled in one direction. If the canometer is put into reverse half way through a measure- ment an incorrect reading will result. * Interim measurements are also poss- ible: 'stop' is pressed at the required point, the distance indicated on the display is noted and the can is replaced in the same position. Then press the 'start' button and continue. * Do not forget to switch off the canometer after use. * Roll on! H PCBs & Set of COMPONENTS for projects are normally available with precious ELECTRONICS CORPORATION 52-C Proctor Road, Bombay-400 007 Phones: 367459/369478 | ,988 6.57 SWITCHING TRANSISTOR "... what is a switching transistor?" "A switching transistor is a transistor used for switching ! " "You mean a transistor can operate a switch? Like we do?" "No, that would be difficult for a transistor, but the transistor itself acts as a switch." "Oh, I see, and how are the contacts made in the transistor?" "No, there are ho mechanical contacts and springs etc. inside the switching transistor. It looks like a normal transistor, however, it does not amplify a signal uniformally like a normal transistor. It either conducts or does not conduct at all." "Like a tap which is either fully open or fully closed?" "Correct, and that is why it is called a switching transistor. It is either fully open for the current or fully closed." "Which terminals are the switch contacts?" "Obviously the collector and the emitter, because the main current of a transistor flows between collector and emitter." "And the base?" "Base is the control terminal. When base current flows, the switch is on" "But the collector-emitter junction can conduct only in one direction!" "That is right." "So the switching transistor is not really a switch?" "Yes and no! It is a switch, but it is a switch with a diode in series! Also it has another small disadvantage that the voltage between collector and emitter never really goes to "That means a switching transistor can never replace a real switch." "Yes, but this remaining voltage across the junction is just 0.1 and 0.2 V, and can be neglected in most cases. This voltage is known as the saturation voltage." "Saturation voltage? With what is this transistor saturated?" "When the base current becomes so high that any further rise in it cannot make a proportionally higher current to flow through the collector, the transistor is said to be saturated." "I don't quite understand." "Take an example. A switching transistor is to switch a voltage of 10V through a 1 0 fi resistance. When there is no base current, there is no collector current and the transistor switch is off. Now when base current flows, the transistor is switched on and current flows through the resistor. How much current will flow?" "10 V across 10 II , that gives 1 Ampere." “Yes, if you neglect the 0.1 V saturation voltage. Fora 1 A collector current, now you can calculate how much base current will be required?" "For that you must tell me the current amplification of the transistor." "Yes, you are right, say it would be 100 " "then you have 10 mA base current flowing." "And what do you think will happen if I force a current of 20 mA into the base?" "Well, then that extra 10 mA has to come out from somewhere!" "Exactly, and this is what is known as saturation. This extra base current is no longer amplified but it just comes out at the emitter. This happens because 10 V across 10 n cannot allow mo r e than 1 A current through collector." I THE TRANSISTOR An Electronic Potentiometer A transistor is often compared to a potentiometer, to explain how it works. Though the analogy is not correct in all respects, it certainly helps in a better understanding of the functioning of a transistor. Figure 1 shows the symbol of the transistor and the direction of current flow. The base current flows from base to emitter and the collector current flows from the collector to emitter. The analogy of a potentiometer is shown in figure 2, where the transistor is shown as a current controlled potentiometer. The sliding contact rises up towards the collector, as the base current goes on increasing. This can be imagined to happen in such a manner that the collector current is higher than the base current by a factor of B which is the current amplification factor of the transistor. This is just a theoretically simplified picture. It just helps to illustrate the fact that by increasing the base current, the collector current can be increased to a much larger extent, which is called amplification. Figure 1 : Two important current paths of an NPN transistor. The base current controls the emitter current. The ratio of Collector Current to base current is called the current amplification factor B. Figure 2: The potentiometer model of the transistor. Easy to understand but theoretically inaccurate. Collector Collector 6.60 NEW PRODUCTS • NEW PRODUCTS • N Weighing Balances The Bombay Company Limited in col- laboration with A & D Company Li- mited of Japan have started manufactur- ing, a range of nine models of High Pre- cision Electronic Weighing Balances. These balances are sold under the brand name of AFCOSET. The accuracies of the various models range fi;om 0.00001 gm for weighings upto 32 gm to 0.1 gm for weighings upto 20 Kg. The Afcoset Balances can be used as stand alone Balances for determina- tion of weight or with the use of the op- tional RS-232C interface can be con- nected to a printer or computer for on- line analytical work. Afcoset balances can be used for various applications like quality control to inven- tory control, stores and shop-floor weighing. The Bombay Company Limited • 334, A- Z Industrial Estate • G Kadam Marg • Lower Parel • Bombay-400 013. 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. M/S. PUNEET INDUSTRIES, • H-230, Ansa Industrial Estate, • Saki-Vihar Road, • BOMBAY 400 072. Milliohm Meter Economy offers a digital Milliohmmeter with a 3'/2 digit, 7 segment RED LED display, 5 ranges with lowest range of 200 milli ohm with O.'l 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 P V C 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 700016. ■ Phone: 290482, 29 5939. 6.62 Biii If your child v jfm is a chip of mq the OLD BLOCK, j He deserves j jFS RE&zxmis: 1 The Electronic Comic-Books in Electro Technolo3y RESI & TRANSI I BANISH THE MYSTERIES OF ELECTRONICS! In the course of the book, they explain how to build a few practical circuits, a continuity tester, a morse code, bleepes & an amplifier. A Printed Circuit Board is included to simplify the actual construction as well as an extremely useful sadset, The Resimeter RESI & TRANSI II Hands of my Bike!! Electronics & Cartoons? Why not? In this, their Second Book Resi & Transi set about buildins an Electronics Alarm with a powerful siren to protect a bike, motorbike, car or even a house. And, as usual they are totally indifferent to the normal code of practice of electronic components. But then, who ever said the electronics has to be dull? Resi & Transi make ELECTRONICS easy .... and fun. Price: Part 1 - Rs. 90/- , Part 2 - Rs. 75/- Packing and postage free. Send full payment by M.OAP.OVD.D. No Your ^■Kl The Champion Connectors! o Dual In-line 1C Sockets, o Single In-line 1C Sockets o Pin Grid Array Sockets, o Socket Adaptors, o Cage Jacks. Zlhampizin Bertramcs Pi A IM