| 1311 ^ j 1 (|H\!DIA| \ 8 mm video System of the future? Intercom tfc Photonics ' VHF filters 3 MSX extensions 4-03 Unable to understand complicated books? Unable to join classes? Now you can Learn Electronics Through Exciting Experiments Q:Can a beginner hope to become an expert in electronics? A: Certainly! Q:How? A: By experimenting! Q: What else does C he require? A: An excellent introduction and some confidence! Q: Who will show the way? AiDynatron! So hurry up, write to: Dynatron Electronics 3, Chunam Lane, Bombay 400 007. 4-05 The Challenger Scientific's New HM 605 60 MHz Oscilloscope provides a real challenge to other oscilloscopes in this range with its outstanding features and performance. See for yourself how HM 605 scores over the leading brand: HM 605 OSCILLOSCOPE Manufactured by SCIENTIFIC MES-TECHNIK PVT. LTD. B-14 Industrial Estate. Indore 452 003 Phone: 31777/8; Cable: SCOPE. YOUR DEPENDABLE SOURCE... TESTICA T-3 THE ONLY MULTIMETER WITH PROMPT SERVICE AFTER SALES ACCURATE! ROBUST! ECONOMICAL! AVAILABLE AT ALL COMPONENT SHOPS ELECTRICAL INSTRUMENT LABORATORIES, 339/68, RAJESH BUILDING. LAMINGTON ROAD, BOMBAY-400 007. PHONE-36 07 49. Learn Electronics Now !!! This Is The Age Of Electronics The Worlds Fastest Growing Industry Important Information: 1 . The price includes packing and postage. 2 All despatches will be by Regd. Past. 3. Please send full amount by D.D. or M.O. No Cheques or V.P.P 4 Allow four weeks for despatches. 5. The kits contain the PCB and components that go on to the PCB. 6. All these projects are from elektor india. Circuits/Data are not provided in the kits as the magazines contain this information. All issues are available at the cover price. precious® ELECTRONICS CORPORATION 1 1, Shamrao Vithal Marg (Kiln Lane) Off Lamington Road, Bombay-400 007. KITS ... Career ... Future 4-13 8 mm VIDEO Since its introduction to the public in 1984, the 8 mm video system has been seen, tested, and liked by thousands Already there are voices saying that this wit! become the video system ot the future. Ever since the develop- ment of the first commer- cial video cassette recorder —VCR— manu- facturers have been trying to reduce the width of the video tape without sacrificing picture and sound quality. In the early days, video studios used 2 inch wide tape and, to 4-18 eleklor India april 1986 achieve the required bandwidth ot up to 5 MHz, a recording speed of 120 ft/s. Current domestic VCRs use half-inch wide tape and a recording /playback speed ot 2.4 cm/s (15/16 i-p-s-)- The three existing domestic video systems. VHS, Betamax, and V2000, suffer from the great dis- advantage ot being totally incompatible with one another. Moreover, the video cassettes of all three systems are too large to construct a manageable camcorder (camera and recorder in a single hous- ing) around them. These disadvantages have always been considered serious enough by the various manufacturers to cause them to invest heavily in the develop- ment ot a new system. Most development was centered around 8 mm wide tape, which would, incidentally, also compete Fig. I. General view of the 8 mm video cassette. A number of holes at the lower end provide infor- mation to the recorder as to type of tape, tape thickness, and tape length. Fig. 2. Composition of the signals as recorded onto the tape. Note that here the NTSC standard is used; for the PAL system, the fre- quencies are slightly dif- ferent. with 8 mm film. Fortunately, the manufac- turers of the 1980s are more sensible than those of the 1950s and 1960s, and therefore put their heads together about the new system to try and avoid the mistakes of the past. The first 8 mm video conference in 1982 was attented by no fewer than I 122 manufacturers from all over the world. Progress was so rapid that within just over a year a pro- posal was made to the IEC (International Elec- trotechnical Commission) for a standardized 8 mm video format. This format has been accepted in principle, although some "peripheral” points have yet to be agreed. There is, of course, some reluctance on the part of certain producers, particu- larly those manufacturing VHS hardware and video cassettes to go hard on the new system. However, Just over a year ago, Sony surprised all and sundry with a complete 8 mm video system. The Sony designers are to be con- gratulated on succeeding in bringing out a perfectly working new system in such a short time. It is, of course, too early to say how the consumer market will take to the new 4-19 Fig. 3. Signal layout on the tape. At the two sides, space has been reserved t or the stationary heads, the use of which is, how- ever, restricted to special applications. Fig. 4. Photograph of a typical rotary head drum, which has a diameter of only 40 mm. system, particularly bear- ing in mind its incon- stancy as regards current video systems and the video disk. If the new system fails, it will not be because of its technology: this is already near perfect and is even now still being improved. uses a tape cassette that is about the same size as current audio cassettes. This cassette has a number of features which can be seen in Fig. 1. Various holes at the lower end give information as to tape length, type of tape, and tape thickness to the recorder. Noteworthy also is the protective "bridge" at the front of the cassette This consists of two parts so that the tape is protected at its front and back, which is a vast improve- ment over the current VHS and Betamax cassettes. Furthermore, the cassette has been simplified as compared with current types. For instance, it no longer has tape guides. Another novelty is the type numbering: there are dif- ferent cassettes for PAL/SECAM and NTSC (as with current VCRs). This is necessary because in the PAL and SECAM systems a field frequency of 50 Hz (25 Hz frame frequency) is used, whereas the NTSC system uses a 60 Hz field frequency. This means that there is a difference in the length of tape used per minute between the PAL/SECAM and NTSC systems. And since the recorder is given infor- mation by the cassette as to the playing time, the correct type must be known. For instance, in a Type P5-90 cassette, the P indicates metal powder tape (an E would indicate metal evaporated tape); the 5 indicates a 50 Hz The new video format video system Technical characteristics signal width field frequency (a 6 would indicate 60 Hz field fre- quency); and the 90 gives the playing time in minutes. It should be noted that there is, as yet, no agree- ment as to the SECAM video system (no video recorders are produced in France, the home of the SECAM system). Current thinking is along the lines of a SECAM-to-PAL trans- coder for use in SECAM countries (France and Eastern Europe). It may Fig. 5. Strictly speaking, the tape needs to travel an angle of only 180° along the drum for the recording of the composite video signal. The additional 41 ° are needed for the recording of the PCM Fig. 6. Auxiliary fre- quencies superimposed on the tracks enable exact alignment of head and track. If these are not aligned properly, a differ- ence frequency is generated, on the basis of which the degree and direction of the required correction are determined. also be that the new MAC system now being studied by a number ot European administrations will eventu- ally offer a solution to this problem. One ot the most note- worthy points of the new system is that the video, audio, and tracking signals are recorded onto the tape together: Fig. 2 shows how. Note that this is an NTSC layout, but the PAL system is virtually identical. Ot interest here is the FM-modulated audio signal, which is a great improvement over current video systems. Since the pilot signals are recorded at the same time as the video signals, separate synchronization heads are no longer necessary. Where the various signals are located on the tape is shown in Fig. 3. It is seen that the combined video and audio signal takes, as would be expected, the larger part of the tape. Guard bands are pro- 4-2" SONY M » Video 8 P5-60 vided at both sides ot the thing is that the picture tape, but these are not yet quality at LP is not used. Also provided is a significantly degraded as PCM (pulse code modu- compared with that at SP. lated) audio track, which is intended for stereo PCM signals The rotary Fig. 3 also shows a feature that has not been men- hGOO drum tioned before, namely that the track width is depen- Fig. 4 shows that the drum dent on the selected tape contains only two video speed. This width is 34.4 heads and one erase for single play (SP) and head. The advantage ot a 17.2 /j m for long play (LP). rotating over a fixed erase The playing time for LP is, head (as found in current therefore, twice as long as systems) is that recordings that for SP. The remarkable can be linked together without the occurrence of depends on a number of visible loops in the picture, frequencies superimposed This is because the part of on the video tracks. The the track that is erased is general operation of this immediately re-recorded. system is shown in Fig. 6. If The tape does not, as in the video head is not current systems, travel an aligned with its track, the angle of 180° along the ditlerence frequency is drum, but one of 221° detected by the head. On —see Fig. 5. The the basis of that fre- additional 41° are quency, it is determined in intended for the PCM which direction the head audio must be moved to bring it The video tracking uses into alignment with its an automatic track find- track, ing system (ATF), which has The 8 mm system can also been derived from the be provided with dynamic V2000 system. Its operation track following (DTF), 4-22 Fig. 8. Since the PCM signal is recorded a t the end of the video frame, the audio data must be col- lected and stored during a frame period and then recorded all at once. Dur- ing playback all data words are stretched over a frame period with the aid of a delay line. Fig. 9. It is possible to use the entire tape width for PCM audio, when a total of six stereo channels can be accommodated. another Philips develop- ment. In this, use is made of small pieces of piezo- ceramic material, onto which the video head is mounted. By applying a voltage to this material, the head can be moved up or down to a small extent, so that it can be perfectly aligned with the video tracks. At the moment. DTF is not yet pro- vided in the Sony equipment. Audio The audio signal is re- corded by the same heads, and at the same time, as the video signal. This guarantees high quality sound, which is, however, monaural. The audio frequency band lies between 30 and 15 000 Hz, and the signal+noise-to- noise ratio is a respect- able 90 d& The Sony PCM system Apart from the standard- ized FM signal, Sony has provided the possibility ot adding PCM audio onto the tape as already dis- cussed with Fig. 3. Sony has opted for an 8-bit system, probably in view of the available tape space. Note that the com- pact disk (CD) system uses 16 bits. However, non-linear quantization results in a dynamic range that is equivalent to a 13-bit Apart from non-linear quantization, a compan- dor circuit is used to com- press 10 bits to 8 bits. The sampling frequency in the PAL version is 31.25 kHz, so that the frequency range extends to about 15 kHz. There are also pre- and de-emphasis circuits for high frequencies, just as in the FM audio recording section. See Fig. 7. The number of data words to be recorded tor each frame amounts to 1250 (625 lines; /2x2(channels) x 2). The final factor ot 2 derives from the error cor- rection system, which adds for each recorded data word one word with the correction code (cross interleave code). The PCM signal is always recorded at the end ot a video frame. This means that all PCM data are col- lected and stored during a frame period and then recorded all at once. Dur- ing playback, the reverse happens: a delay line then stretches the data words over a frame period. In practice, the sound is of much better quality than would be expected of an 8-bit system. See also Fig. 8. Apart from the separate PCM part ot the tape, it is possible (but, so far, in the Sony system only) to use the entire tape width for PCM as shown in Fig. 9. This enables the recording of six stereo signals: a vast quantity ot audio infor- mation, indeed, which gives 18 hours playing time on one tape. What ot the future? The technology on which the 8 mm video system is based leaves little, if anything, to be desired. The system appears to meet all the requirements the ever more critical user will demand from this type of equipment. And, don't forget that however good the Sony system already is, there is room for improvements and exten- sions without the loss of compatibility. Much will depend, how- ever, on the buying pat- tern in the consumer market, as well as on the marketing of the system (the lack of good marketing is almost cer- tainly one of the main causes ot the relative failure of the V2000 system and of the total failure of the video disk). Because of the world-wide standard- ization and consequent compatibility, it is to be hoped that the new system be accepted soon and readily to the great benefit of the consumer. HB 4-23 PHOTONICS Photonics is the technology of using photons to convey information in a controlled manner. A photon is an elementary particle of tight in the frequency range from 3x10 s MHz to 6x10 w MHz (corresponding to wavelengths from 1000 nm —upper limit of infra-red region— to 5 nm —tower limit of ultraviolet region. Photonics must not be confused with opto-eiectronics —in which photons and electrons interact— or with e/ectro-optics, which is a study of the relation between the refractive indexes of certain dielectrics and the electric fields in which they are situated. Photons may not replace electrons in data process- ing and storage this cen- tury, but there are reliable indications that they will be used increasingly in data communications via optical-fibre cables. And, of course, they are already in use in the remote con- trol of countless hi-fi and television sets; they are also indispensable in the Strategic Defence Initiative (Star Wars). There is also the photonic computer now being developed at Heriot-Watt University, Edinburgh, and at the Bell Laboratories in Princetown, New Jersey. These computers use transphasors, the optical equivalent of transistors. Their main attraction is that they can work thousands of times faster than electronic ones because although elec- trons, under ideal con- ditions, move almost as fast as light, they are slow- Fig. 2. Depending on the angle of incidence of the light ray, the transmission path is called low- or high- order mode: the greater the angle, the lower the mode. Fig. 3a. Multi-mode fibre. Fig. 3b. Mono-mode (or single-mode) fibre. partial refraction of the light. For instance, in Fig. | 2a, the angle ot in- cidence is so small that a large part ot the incoming light is refracted. In optical fibre, this would mean that a large part of the light would be lost in the clad- ding of the cable. Fig. 2b shows the critical angle of incidence: the refracted light here is at an angle of 90° with the normal. At the critical angle, the refracted light may cause interterence. It is. therefore, essential that the angle of Incidence Is greater than the critical angle — see Fig. 2c — when total reflection takes place. The condition for total reflection is that the ray of light travels from an optically dense medium with a relatively large refractive index to a less dense one with a smaller refractive index. Rays of light that fall upon the media separation at an angle smaller than the critical are called high- order modes: they take eleklar India aprtl 1986 4-25 relatively longer to reach the end ot the cable. Rays of light that travel almost parallel to the optical axis, i.e„ at an angle greater than the critical, are called low-order modes. Low-order modes travel faster because they are reflected less often than high-order modes. Low- order modes are tar less prone to losses than high- order ones. The sine of the angle of in- cidence of the ray of light is called the numerical aperture: this is the prime factor where two optical waveguides are to be linked. The numerical aperture is also an indi- cation of the difference between the refractive in- dexes of the core and the cladding: the smaller it is, the wider the bandwidth of the optical signal. Optical-fibre cable In multi-mode fibre (see Fig. 3a), the ray paths of the different modes are ot different lengths and have, therefore, different trans- mission times. Because the modes are divided by a pulse, this is subject to pro- gressive spreading as it travels along the fibre, causing it to interfere with adjacent pulses. In mono- mode (also called single- mode) fibre (see Fig. 3b), the core diameter is com- parable with the wavelength of the light, so that there can be only one electromagnetic propagation mode and spreading of the pulse (called multi-path disper- sion) is eliminated. Its small core size makes mono-mode fibre more dif- ficult to use, but it can be made with an attenuation of less than 0.4 dB/km at a wavelength of 13 tim (as against 2 to 10 dB/km for multimode fibres). A typical bandwidth of mono-mode fibre is 10 GHz. A typical design of opfical-fibre cable is de- tailed in Light work for submarine cables elsewhere in this issue The fast rate of incremen- tal improvements in op- tical fibre technology, which is due mainly to AT&T's Bell Laboratories, British Telecom, and Japan's NTT (Nippon Telegraph & Telephone), have made multimode fibres already obsolescent as far as long distance cables are concerned. (Multi-mode fibre cables need repeaters every few miles, whereas with mono- mode fibres distances be- tween repeaters are of the order ot 50 to 100 miles). The three organizations are already researching new core materials which, they hope, will eventually enable repeaterless trans- oceanic cables. Currently, the central core of optical-fibre cables is made of doped silica sheathed in pure silica. New core materials now being studied include oxide-based and halide- based fibres. These could be from 2 to 1000 times more transparent than silica. They would also disperse less light than silica, which would result in cables with much greater capacities than present ones. Data transfer Optical-fibre networks need, of course, other than the cable a sender, receiver, coupler, and repeaters (see Fig. 4). There are two main types of optical sender: the infra-red diode and the laser diode (laser=light amplification by stimulated emission of radiation). Both enable light energy at wavelengths from 0.8 to 1.5 urn to be injected into the fibres. The most com- monly used type is the infra-red diode, however, because It is relatively cheap, reliable, has a long life (10* to 10 7 hours), and is easy to use. Further- more, they have only little drift with temperature, and their current can be modulated readily. For wavelengths from 0.7 to 0.9 urn silicon diodes are used, but in the range 1.1 to 1.5 fim AIGaAs (aluminium-gallium- arsenide) types are necessary, as the energy transfer of silicon diodes at those frequencies drops sharply. Infra-red diodes have the disadvantages that their bandwidth is limited and that they can- not emit parallel beams of light. The latter means that the light emitted must first be passed through an op- tical system where it is converted into a parallel beam. Laser diodes do not need such an optical network and also have a higher output (up to 650 mW). Pulsed lasers may deliver up to 100 W bursts. Further- more, the attainable bandwidth is much wider than possible with infra- red diodes. Because the are very noisy, and this is light rays in a laser beam are to all intents and pur- poses parallel, a larger part ot the available energy is injected into the fibre. Unfortunately, lasers also have some draw- backs: they are difficult to manufacture; they are, therefore, expensive; their life at 10 s hours is much shorter than that of IR diodes; and they drift with changes in temperature. The latter means that the relatively high current through them (in pulsed types a tew amperes as compared with about 150 mA In IR diodes) must be regulated. As laser diodes take 4 to 8 ns before they emit intra-red light (at the onset they start emitting visible light), their quiescent current must also be regulated to make controlled operation possible. Summarizing, laser diodes require auxiliary elec- tronic circuits, whereas IR diodes require additional optical networks (lenses). For the receiver there are also two possible devices: p-i-n diodes and ava- lanche diodes. A p-i-n di- ode is a photodiode that contains a region of almost intrinsic (i-type) semiconductor between the p-type and n-type regions. P-i-n diodes combine fast reaction times (shorter than 1 ns which makes them very suitable for op- eration with laser-type senders) with small supply voltages, simple electronic circuitry, and relatively low prices. Unfortunately, they f6 the more troublesome since their low output must be amplified by a so- called trans-impedance amplitier: this device also acts as a current-to- voltage converter. Avalanche photodiodes provide a substantial gain (40 to 60 dB) and are also very sensitive Furthermore, they are far less noisy than p-i-n diodea They are however, more expensive have a small demodu- lation bandwidth, and are only suitable for use with digital signala Moreover, they require a very high supply voltage of 100 to 1000 volta which, inciden- tally, shortens their life as compared with p-i-n A description of a typical repeater is given in Light work for submarine cables. A range of couplers is commercially available and some typical seen in the photographa These are used where the connection is not perma- nent. For permanent con- nectiona the two cables are spliced under the microscope with the aid of a small electric welding tool. If the splicing is car- ried out properly, the joint attentuation will be less than 0.15 dB. T-junctions are also poss- ible: a typical optical coupler for this purpose is shown in Fig. 6. This device ia incidentally, also suitable for use as a duplexer. Another interesting possi- bility is wavelength multiplexing as illustrated in Fig. 7, which can greatly increase the ca- pacity of the fibre cable. The light from the sender diode is paralleled and then projected via a lens onto a retlection filter that is inclined with respect to the axis of the lena This filter reflects the light rays into a direction that depends on the wave- length. The lens converts the change ot direction into a positional shift so that the reflected rays of all wavelengths converge, after which the compound ray is injected into the transmission fibre. Fig. 5. Optical- fibre cable with couplers. Fig. 6. Optica! T-junction il- lustrating the duplexing concept. Fig. 7. Construction and mode of operation of the optical multiplexer. 4-27 RF CIRCUIT DESIGN - 2 Do you get annoyed from time to time (or more often) by your favourite FM radio programme being interrupted when a strong out-of-band signal blocks the receiver? if so, read this article and find out how to design and construct a fitter that may ban this irritation for ever. VHF FILTERS by A Bradshaw & J Barendrecht Photograph I. Practical realiz- ation of the VHF roofing filter LPF section. Though its characteristics are not ideal, it provides a good starting point for more advanced filter construc- tion. Note the two capacitors at the input; their total capacitance should be about 44 pF Elektor Electronics has presented its | readers with comprehensive articles 1 on theory and practice of VHF aerial j amplification before; see. for in- j stance, the February 1980 (UK) issue of this magazine. The conclusions | reached in those articles may be i summarized as follows: 1. A well-designed aerial amplifier can only compensate for cable loss if it is mounted in the immediate ' vicinity of the aerial (masthead mounting). 2. To be of any beneficial use at all, this booster must have appreci- ably lower self-generated noise than the receiver. 3. The first active device in the re- ceiver RF signal chain determines I to a large extent the total receiver system noise figure and thus its sen- sitivity for weak signals. 4. A good directional aerial is the best booster because it generates no noise, is absolutely intermodu- lation-free and functions as a selec- tive device at the same time. As evidenced by the article on the wideband aerial booster with Type BFT66 transistor, low noise, good in- termodulation ratio and high signal gain are generally appreciated characteristics of active devices in VHF aerial amplifiers. However, it was also pointed out that only one of these characteristics may be favoured over the others given a cer- tain transistor working point; the : three are never optimum for one bias setting. | It is for this reason that many wide- j band amplifier designs use two j identical, cascaded high fr type transistors; the first (aerial side) set I for low noise, the second (receiver j side) for high overall gain. It will be fairly obvious that intermodulation characteristics of such a design are far from ideal, simply for lack of suitable DC setting and appropriate filtering. To increase bandwidth and reduce the intermodulation prod- ucts, these transistors are usually direct-coupled, and every effort has been made to keep booster gain as high and constant as possible over a frequency range as large as SO. . ..800 MHz. It will stand to reason that this type of amplifier can not be used for recep- tion of weak FM band signals, because the odds are that a far stronger RF signal present outside the receiver tuning range will wreak havoc with the booster transistors. Even if the aerial features some at- tenuation for this out of band signal, booster input voltages may be as high as 100 mV with a powerful transmitter in close proximity. Even a very selective and intermodulation- free receiver can not do anything towards improvement of reception in this case, simply because it sees a mess of interference and inter- modulation products at its input. To keep strong out-of-band signals away from the base of the VHF pre- amplifier stage, some filtering device is called for. Conflicting design considerations contend for the upper hand, however, and a basic knowledge of filter operation and construction is required to find Fig. 1. Typical curves showing that a band-pass filler (BPF) pro- file is obtained from adding curves of con- stituent low pass (LPF) and high pass (HPF) sections. the right compromise for a given situation. VHF filters; a crash course For a basic understanding of filter operation, it is useful to think of it as a sieve; depending on the diameter of its holes, it will pass the desired liquid and block large particles, however many. In electronics, such a sieving device is generally referred to as a band-pass filter; it has a high attenuation for all signals outside its pass band. A typical frequency vs attenuation curve of a bandpass filter (BPF) is shown in Fig. la. The shaded area is referred to as the filter 3dB band- width. Note that Fig. Id also shows a BPF curve, but this time with lower skirt steepness than that of Fig. la, and a reduced 3dB bandwidth. From this comparison of filter curves it should be evident that the term filter selectivity it not related direct to 3dB bandwidth. Although the band-pass filter type is a suitable starting point for intro- ducing filter theory, it must be men- tioned here that it is basically a combination of two constituents; a low-pass filter (LPF) and a high-pass filter (HPF), the curves of which are shown in Figures lb and lc respect- ively. Note that skirt steepness of both LPF and HPF may be less, as shown in corresponding Figures le and If. It will be evident that the BPF curves of Figures la and Id may be obtained by adding Fig. lb to lc and Fig. le to If respectively. Tb define the 3 dB bandwidth of the BPFs, it will be seen that BPF A = HPF A and (1) BPF A = LPF A (2) where fc is the cut-off frequency of LPF or HPF, or the frequency at which the filter output, Uo. falls a>=0.708tt=W2«=3 dB attenuation (3) Thus, the 3 dB bandwidth of a BPF may be calculated from bW3dB=f2-fl (4) The curves shown in Fig. 1 are theoretical and therefore idealized; depending on component tolerance and construction method of the filter, it may feature far less smooth characteristics, as will be seen later. Neither need band-pass curves always be symmetrical like those of Fig. 1; depending on skirt steepness of constituting LPF and HPF, the low and high side roll-off characteristic of a BPF may have quite different profiles. To come to a conclusion about suitable electronic components for use in filters, the low-pass setup shown in Fig. 2a may be examined; it is also known as a 'pi type’ (note its visual similarity to n). Assuming that the circuit is at resonance, that R\-Z\=Z>=Ro=Z and that 0 (quality factor) is fairly high, then the basic design equa- tions for this filter are as follows: z- yfi ® 4-29 Photograph 2. Roll-off charac- teristic of the VHF roofing filter LPF section (2a) and response of the same when several hundred Megahertz (2b). Note that low at- tenuation cor- responds to a high point in the curve, as op- posed to the curves in Figures 1 and 2. Photograph 3. Band-pass profile of the VHF roof- ing filter, sweep centre frequency at 97 MHz. Fig. 2. Starting from the basic pi-LPF (Fig. 2a) m-derived sec- tions may be added (Fig. 2b) to obtain a low pass profile as shown in Fig. 2c. L= R Htc L=159.2R/fc (10) C=318000/Rfc (11) (7) R = filter termination resistance L = inductance in filter C = capacitance in filter f c - 3 dB cut-off frequency Z = filter impedance For VHF applications, these equations are adapted as follows to calculate with nH (nano Henry, 10~ 9 H), MHz (mega Hertz, 10 6 Hz), and pF (pico Farad, 10~ U F): Example: if a filter of this type were to be constructed for fc = 100 MHz, and Z = 50 Q, the following compo- nent values are found: C = 63.6 pF, L = 79.6 nH. To improve the filter skirt steepness, several of these sections may be cascaded provided they have been designed for the same termination impedance. However, so-called m- derived sections at both LPF input and output may be a more efficient way to get the desired curve shape; see Fig. 2b for the basic arrange- ment With L and C calculated from (9), (10), and (11), the component values for these additional sections are computed from Z- 03) Cz=mC (14) To understand how m is determined, refer to Fig. 2c which shows the fre- quency vs attenuation curve of the filter proposed in Fig. 2b. To be noted are the ‘humps' which appear above fc, at /<*>, the filter attenuation seems to be infinite, and this is repeated at regular intervals as / in- creases. The points of infinite at- tenuation are called poles and, generally speaking, the more filter sections, the more poles will appear; this also goes for high-pass sections, and, consequently, for band-pass filters which will feature poles at either side of the curve. The value of m is calculated from m = \J\-fc ! /f°° ! (15) where /°° is the frequency of the first pole. Most designers, however, use the value 0.6 for m, which gives us L\ = 0.6L (16) Ci = 0.27/7 (17) Cz = 0.6 C (18) 2 a for the three-stage LPF of Fig. 2b. There are several types of m-derived sections, and some of them are shown in Fig. 3. To go into the design calculations for the components in these sections would be beyond the scope of this article, and interested readers are referred to the numerous handbooks on this subject. VHF roofing fitter A practical example will no doubt be quite helpful at this stage; if only to get an idea of the practical problems involved in filter design and con- struction. Figure 4 shows the circuit diagrams of precisely calculated filters with m- derived sections shown in Fig. 3. If the proposed LPF and HPF are cas- caded, a band-pass filter may be ob- tained with suitable characteristics for selective VHF reception (85.. . . 110 MHz). Note the component values in LPF and HPF; they are, of course, theoretical. The term roofing filter is used to refer to the protec- 4-30 eleklor 1 4-31 Photograph 5. This is a 5-stage helical filter for use in the 400 to 500 MHz fre- quency range. Coils are inductive- coupled and tuned with brass precision screws. Note the low- impedance tap at input and out- put coil. The only known and stable im- pedance in the receiver RF chain is provided by the coaxial cable (50 or 75 Q). The undesired signal, then, will find the filter input as highly unmatched, and a large part of the signal will be reflected into the cable, only to be reflected again by the aerial. The delaying effect of the coax cable added to the unavoidable phase shift and reflection cause a so-called standing wave. It will stand to reason that the filter input must be as reflection-free as possible for the desired frequency band, simply because a large part of the RF signal would else be lost to the active device Furthermore, filter insertion loss must also be as low as possible, but, as we have seen, good band- pass profiles require many filter sec- tions and thus many components to pass the signal, and neither of these has ideal flow-loss) characteristics. A total filter insertion loss of 0.5 to 1 dB is already a good figure, but it should be kept in mind that any in- sertion loss adversely affects the op- timum noise figure of the active device coupled to the filter output. Fitter construction To conclude this article, some useful suggestions will be given for the choice of filter components and mechanical construction, because it ought to be clear by now that good filter calculation may be useless if the practical realization is not up to the 'VHF standard’. As these are mostly unwritten laws, it is very in- structive to have a look at some of the established VHF construction methods in, for instance, a discarded VHF/UHF TV tuner. Coils: Use 20 SWG or thicker silvered copper wire (CuAg) for the self-supporting, air-cored coils, and make sure that coils in separate filter sections can not ‘see’ each other to avoid unwanted stray coupling. In case the coils are PCB mounted, coupling can be avoided by pos- itioning them at an angle of 90°. There are, however, also filter types that are based on inductive or capacitive coupling of coils to achieve a suitable bandwidth, e.g. helix type narrow band slot-coupled filters, in which case the above rule does not apply. Capacitors. To arrive at the calculated cut-off frequency, the capacitors must be close tolerance types (1 or 2%) with good high fre- quency characteristics (NPO or silver mica). Keep leads as short as poss- ible to avoid introducing stray in- ductance in the circuit; where available, ceramic chip capacitors are the ultimate solution. Trimmers, if used, are preferably tubular glass or ceramic types with extremely low end capacity (lpF or less); older types of TV tuner still contain them in abundance, but they are not easy to get out intact. J Connectors. Use standard 50 Q plugs and sockets such as those in the UHF series (PL259-S0239), BNC or N I types are even better, however, and much to be preferred. Do not ask for trouble by using the cheap coax con- nections as used with modem TV sets and FM tuners. Housing. The filter should be fitted in a stable metal housing (diecast box) to prevent strong signals from bypassing. If at all possible, fit the amplifier in a separate housing and connect it to the filter output with a short length of low-loss coax cable fitted with BNC or N plugs; this also goes for the aerial-to-filter connec- tion. Phograph 4 shows some pre- ferred parts for VHF filter construc- tion, and, finally, Photograph 5 shows a UHF type band-pass filter for pro- fessional use. Next time | A further instalment in this series will i concentrate upon an up-to-date VHF preamplifier stage constructed on the universal HF board. JB:JB Literature: Radio Communication Handbook vol. 1, publ. the Radio Society of Great Britain (RSCB) The Radio Amateur's Handbook, publ. the American Radio Relay League (ARRL) Elektor Electronics, February 1980 issue (UK ) Reference data for Radio Engineers 4" 1 edition, pp 164-182, publ. ITT UKW Berichte 3-75, R. Lenz DL3WR: ‘Rauschen in Empfangsanlagen' 4-32 el. This third pari in the series presents an MSX busboard to overcome the limitations of that single slot on the computer; up to eight cartridges may be inserted and selected with keys or under software control. EXTENSIONS - 3 eight-slot bus board Any MSX user in possession of several cartridges must at some time have wished to be relieved of the cumbersome cartridge-exchange procedure: power off — remove car- tridge — insert cartridge — power on — test. Moreover, frequent cartridge exchanging may cause bad slot con- tacts after a while. Note that not all cartridges have an insert/remove protection fitted, so that it is sensible to always switch off computer power before exchanging any cartridges; it is better to be on the safe side! The present eight-slot MSX busboard offers an interesting solution to these problems, because cartridges are now constantly available to the user; he need only issue a slot (i.e. cartridge) select in- struction in MSX BASIC or press a single key to have the desired game or utility ready for use. Block schematic A functional diagram of the MSX busboard is shown in Fig. 1. All MSX computer signals have been buf- fered for safe use with the car- tridges; this is customary practice with computer expansion projects to avoid overloading the available com- puter output signals. One of eight slots is selected by the DECODE SELECT section; the active I slot (cartridge) is indicated by a lighted LED: Slot selection is effected either manually or by software; de- pending on the data transfer direc- tion set by the DATA SELECTOR, either three databus bits (software) or three bits from the manual slot selection circuit ENCODE SELECT are passed to the DECODE SELECT section which decodes the three bit combination into a relevant slot sel- ect signal. The DATA SELECTOR is set to databus transfer by a signal from the I/O SELECT DECODER section which compares the eight-bit ad- dress LSB (least significant byte) dur- ing CPU output with a switch-set output channel code; when the two bytes match, ie. the computer selects the desired output channel, the DATA SELECTOR transfers the three-bit slot selection code sup- plied with the output instruction to the DECODE SELECT section, and the desired slot is selected. Similarly, the manual slot selection code may be passed to DECODE SELECT whenever the I/O SELECT DECODER is inactive. Practical circuit I Circuit diagram Fig. 2 shows how a I I number of integrated circuits realize the above mentioned functions. Databus buffer ICi is an octal bidirectional device enabled with MSX signal SLTSL (slot select), while direction of data transfer is deter- mined by the logic level of RD (read); this was so arranged because using WR (write) for this purpose more readily leads to bus contention prob- lems due to critical signal timing. Output pin 19 of 8-bit magnitude comparator IC< will only go low when two conditions are met; the ad- dress set with switch block S9 matches t he CP U-generated address LSB and IORQ is active (i.e logic low), which indicates that the eight bits are a valid output channel cur- rently addressed by the CPU. The SEL input of multiplexer ICe is consequently low and this device will pass D»-Di-D 2 from the databus to the A-B-C inputs of IC 7 . The multi- plexer may conveniently be com- pared to a four-pole two-position switch where the position of the switch is determined by the logic level applied to the SEL input. If SEL is at low level, data transfer nA-nY is effected (software slot selection), else nB—riY (manual slot selection). Thus, latching 3-to-8 decoder IC 7 may receive its slot selection code from two sources, and it activates the I output corresponding to the binary Fig. 1 Block schematic presentation of the MSX bus board. Any one of eight slots may be selected either by soft- ware or manually with a set of switches. code applied to the A-B-C inputs on the low-to-high transition of the logic level at the GL (latch enable) input. In this way, one of eight MSX slots plus relevant LED may be activated. The latching function of 1C? is essential to operation of the present circuit; the device holds the last applied binary code and activates the corre- sponding output until a further low- to-high level transition at its GL input signals the presence of a new slot select command. I n the case of soft- ware slot selection, WR supplies the la; - ling pulse, whereas for manual selection combination of gates N1-N2-N3 simulates a correctly timed WR signal. Manual slot selection is effected by IC9 and associated keys Si to Se; if the user wants to enable a specific cartridge by hand, he may simply press the appropriate key to override any previous slot selection com- mand. When one of the keys Si to Si is pressed, priority encoder ICs sup- plies the three-bit binary code rel- evant to_the number of the key, and output GS (group select) goes low. This pulse, together with Eo, triggers the WR simulator. Key Si selects slot ft which is also the default slot after power-up; thus, any cartridge pres- ent in slot 0 will be automatically selected when the computer is i switched on. Should any keys be pressed simultaneously, then the one with the lowest number has highest priority. In case the power supply inside the computer is not able to handle the current consumption of the car- tridges on the bus board, the supply voltage connections may be re- moved to connect an external power supply to the relevant pins (+5V, + 12V, —12V). However, check and double-check that external power is applied to the correct pin, i.e. the one that connects to the cartridge bus lines. If this precaution is not ob- served, irreparable damage may be j inflicted on vital (ie. costly) com- ! puter parts. Unfortunately, one 4-34 r 3 36 ........... !■■■■■■■■■■■■■■■■■■■■■■■■■! n666666'6 6 loaooooooo. s mistake may have detrimental conse- j quences, and it is therefore strongly suggested to mark the relevant soldering pin. j Construction j The ready-made PCB for the MSX ! bus board is shown in Fig. 3; its dimensions are mainly determined by the eight slot connectors K. . K« and the necessary space for the car- tridges. Construction is best started with fit- ting the wire links, followed by the IC sockets and the six soldering pins , for external supply connection. For the time being, these pins may be jumpered with wires. J The component mounting plan shows a block of eight DIP switches for output address selector S* and slot selection keys Si . . . Ss. The latter, however, may be replaced by eight small push-to-make keys, connected I to the bus board via a length of fiat j ribbon cable and a DIP header. In I this case, a normal 16-way IC socket I is fitted on the board instead of the | DIP switches. The constructor is free to make a nice looking slot-select j keypad with a LED to go with every | key. A noteworthy aspect of the present design is the application of high- speed CMOS ICs (HC or HCT types) which results in very low power con- sumption and a high degree of immunity to digital noise. More infor- mation on these novel devices may be found in Elektor India \ October 1983 issue. However, ! where 74HC(T) types are not yet j available, the well-known 74LS Fig. 5 Recapitu- lation of MSX slot signal functions. equivalents may also be used in this Computer connection Last month's article in this series on MSX add-on units introduced a car- tridge extension board which basi- cally consisted of an adapter plug and an EPROM section (see Elektor India. March 1986 issue). The present bus board may be con- nected to this 50-way plug with a length of 50-way flat ribbon cable equipped with suitable press-on type sockets — see Fig. 4. With the bus board completed and no cartridges inserted as yet, con- nect the extensions as indicated and verify that the computer still func- tions as normal. The LED with slot 0 should light at this stage. Set a slot selection output channel on the bus board with switch block Ss, for instance 3Ftwx. This is done as follows: first, establish the binary code of the channel, in this case 3Fnex = 0011 1111 (A?... A«). Next, set this code with the eight switches, but note that 'switch = ori corresponds to ‘bit = low (0)', and also remark the order of the switches as arranged on the board. Output channel 3Fhex corresponds to this combination of Ss (left to right): on-on-off-off-off-off-off-off (As- Ai-Ae-Ai-Ar-As-Ai-As). Set this combination and see if the LED with slot 4 (Ks) lights when instruction OUT &H3F.4 is issued in MSX BASIC. If this works, test the manual slot selection by pressing some of the keys to see whether the desired slot is selected as indicated by the corresponding LED Switch off power and insert car- tridges, but remember to plug them in with the front (label) side towards the computer connector Kn; it is sen- sible to mark slot pins 1 with a spot of white paint to avoid inserting car- tridges the wrong way about. Fig. 5 once more shows the MSX slot pin designation with signal functions. The universal I/O bus may also be connected to the present bus board (see Elektor Electronics. January 1986 issue). With this amount of com- puter expansion available, it would be fair to say that MSX interfacing is truly up-to-date and ready for almost any task that has to do with periph- eral control. CD.BL Fig. 4 Connect- ing the bus board to the car- tridge extension board with a length of flat rib- bon cable effec- tively enables MSX users to install a total of eight cartridges. 4 L 4-38 elektor i INDUCTORS IN PRACTICE in spite of their apparent simplicity, inductors none the less often pose problems, because invariably they cannot be obtained ready-made, he they have to be designed and wound by the constructor. This article aims at removing some of the obscurities surrounding this subject and showing that making an inductor is not such a daunting task as some think. An inductor is an elec- tronic component that possesses appreciable in- ductance. Self-inductance is the property of a circuit to oppose any changes in current (lowing through the circuit: this manifests itself by the production ot a voltage that tends to op- pose the change ot cur- rent. This voltage is called the back-e.m.t. Mutual in- ductance is the phenomenon whereby voltage is induced in one circuit by changing the current in another. The unit ot both sell- and mutual inductance is the same: the henry, but their respec- tive symbols are L and M (or Z 12). An inductor has an inductance of 1 henry if the back-e.m.f. In it is 1 volt, when the current through it is changing at the rate ot 1 ampere per second. Inductors invariably consist I of many turns of wire wound adjacent to one another on the same sup- port, called the former, but in high-frequency appli- cations they are often self- supporting (i.e., air-cored). The former may also be ot ferromagnetic material to increase the inductance j many hundreds of times. Unfortunately, so-called eddy currents are induced in the ferromagnetic material, and these in- crease the DC resistance in a practical inductor. Powdered-iron cores are, therefore, used at high fre- quencies because their high resistivity makes eddy-current losses negligible Such territe materials are not as useful as iron at low frequencies, however, because mag- netic saturation restrict the maximum power level ot the inductor. Inductors have a fre- quency -dependent resist- ance (called reactance) to AC currents, and an ohmic resistance, which is primarily due to the wire from which the inductors has been wound. The reactance, X\. is equal to a >L, where (2) for circular coils, and L=non 2 a(4/nlog e (1+n a/b))+ + 1/(3.64+2b/a+ 0.51 (b/a) 2 ) (3) for square coils, where a and b are the inductor sizes in metres as indi- cated in Fig. 2. L is in henries, and po is the ab- solute permeability stan- dard, defined as 4n10 rM . The three charts shown in Fig. 3 give inductor winding data for a number of popular wire and former diameters, but the computer program of listing 1 allows a great many more possibilities for fast calculation of induc- tor winding data, both for circular and square in- I ductors. The latter are perhaps less known among designers, but square inductors may be used as window mounted, multi-turn rhombic aerials for directive reception of medium- and long-wave signals. The computer program listed has been written in MBASIC, and may require a patch here and there to suit the specific screen and cursor commands of some computers. For spac- ed inductors, the program uses an iterative approx- imation routine, which sup- plies a start value (guess) to the main calculations and adapts the variables to step towards maximum accuracy. Obviously, the better the guess, the faster the program will come up with the result, since in that case less calculation time is required. It stands to reason that n-step itera- tion is practically not feasible with only a pencil and a cheap calculator, since far too much time would be wasted before a useful result is obtained. Therefore, the number crunching facilities offered by the computer are welcomed by many designers of air-cored in- ductors. Literature references: 11 Proceedings of the IEEE, Vof 70 no 12: December 1982: Wheeler, HA: ‘Induct- ance formulas for circular and square coils ' 2) Radio Engineers Hand- book, by F E Terman ; McGraw-Hill. 4-41 Light work for submarine cables by Kenneth Fitchew, British Telecom Research Laboratories, Martlesham Heath, Ipswich Submarine cables have entered a new era with the application of optical technology, making them compatible with digital networks ashore. During 1985 the world's first international optical- fibre submarine cable was laid between the UK and Belgium, bringing a new economy to this means of international communication. It is well over 100 years since the first submarine telecommunications cables were laid. The Atlantic Ocean was spanned successfully in 1866, an event that opened the first age of submarine cables, the telegraph cables. They were simply insulated con- ductors; on long routes they caused a great deal of attenuation and distor- tion of signals at even very low frequencies, so they could be used only for low-speed telegraph signals. And their gutta- percha insulation was often attacked by teredo worms. The development ot the thermionic valve provided a means of amplifying signals at Intervals along a cable; this allowed higher frequencies to be transmitted over long distances, and meant that a number of speech channels could be car- ried on one cable. In 1943 British Telecom (then the British Post Office) laid the world's first experimental submarine telephone 4-42 el«k:or Indio april 1 986 Sample of deep-water optical cable made by STC. repeater in a cable be- tween the Isle of Man and mainland Britain, and in 1956 a repeatered tele- phone cable carrying 36 circuits was successfully laid across the Atlantic Ocean, firmly establishing the arrival ot the age ot the coaxial cable. This type of system has been developed over the years and systems with ca- pacities of up to 5520 Cir- cuits using frequencies up to 45 MHz have been sup- plied by Standard Telephones and Cables (STC). Coaxial cables are providing excellent service around the world and new ones are still being ordered. Developments in technology have now reached a point where we are on the threshold ot the third age, that of the optical-fibre cable. Op- tical systems offer several advantages. First, they I should be cheaper, , mainly because they use fewer repeaters; in the first generation of optical systems the repeater spacing is about 40 km. seven times greater than that of the latest coaxial systems. Second, more traffic-carrying circuits are available, because several fibre pairs can be in- cluded in one cable and the development of systems using higher data trans- mission rates is quite feasible Third, optical cables are better suited to digital operation, and so are more compatible with the worldwide move to digital communications. The world’s first reported trial of an experimental optical submarine cable was carried out in Loch Fyne, Scotland, by British Telecom and STC in 1980. This has been followed by trials from all those other countries with an interest in making submarine systems, namely Japan. USA and France. The world's first international optical-fibre submarine cable has been provided | between the UK and I Belgium by STC. The cable, code-named UK-Be/gium 5. will eventually carry I 11 520 circuits linking the UK lo Belgium, the | Netherlands and West | Germany. The system con- | tains six fibres, three | operating in each direc- tion and each carrying I data at an information i rate ot 280 Mbit s. The | system length is 113 km, ! with three repeaters. Optical fibres The basic principles of optical-fibre transmission were established by Hockham and Kao. work- ing at Standard Telecom- munications Laboratories in the UK in 1966. For most I submarine applications ! the high data-rates and i long spans make it essen- j tial to use monomode fibre rather than the more well known multimode ! fibre. Monomode fibre Core Cladding diameter | diameter 50|im Il25|im JL Multimode fibre 1 1 All power in single propagating mode _TL Monomode fibre _TL Output pulse In multimode step-index Fibre (a) the different propagation modes may be represented as ray paths (x, y, z) which are of different lengths and therefore have different trans- mission times (delays) Because a pulse divides between the modes, it is subject to pro- gressive spreading as it travels along the fibre, causing it to interfere with adjacent pulses. In monomode fibre (b) the core diameter is comparable with the wavelength of the light, so there can be only electromagnetic propagation mode and spreading of the pulse is eliminated. transmits light in only one transmission path (see diagram on page ) so eliminating multipath dispersion, but its small core size makes it more difficult to use Monomode fibre can now be made with an attenuation of less I than 0.4 dB/km at a wavelength of 1.3 pm; the manufacturing process needs to be well con- trolled to ensure not lust low loss, but good con- centricity of the core and | adequate strength. Design The design requirements for optical submarine cable are severe because of the physical environ- ment. The cable must be capable of withstanding pressures of 70 MPa in areas as deep as 7.3 km, and tensile loads of several tonnes when being picked up during repair j work. Owing to the move- ment of a cableship on the sea the tensile force | on the cable during recovery is a static load with a cyclic variation; this means the cable has to withstand fatigue. In most optical submarine cables the fibres are pro- tected within a tube which provides a pressure- resistant structure. It also J carries longitudinal steel | wires to provide the necessary tensile strength. J However, a cable with a ] single layer of stranded steel wires as a strength 1 member has a rather low I modulus of elasticity, so | under tension it stretches, ! partly through extension of the steel and partly through the strand un- twisting. But the silica fibres are far less elastic than steel, so they may break. Three possible ways round this problem are to make fibre that is very strong and can withstand the stress, to include 'slack' fibre in the cable, or to design a strong cable which can cope with high tensile loads without un- due strain. The third ap- 4-43 proach is used in cable made in the UK; its strength member consists of two layers ot steel wires wound with opposite lays so that there is no un- twisting under tension. With shallow-water cables a main danger is from (ishing trawls. One way to protect them is to surround the cable with steel ar- mour wires; they provide resistance to abrasion and add to the strength. Co- axial cables have been protected in this way for many years, using one or . two layers of armour wires wound with a relatively long pitch. However, tests here have shown that a much more resistant cable can be made by winding the outer layer ot armour wires with a very short pitch; this type of cable, known as 'rock' armour, has been used in the North Sea in areas known to be especially haz- ardous. Burial A second way of protec- ting the cable is to bury it in the sea-bed. Where possible the preferred method is immediate burial by means of a plough towed behind the cableshlp. For UK-Belgium 5 the cable was buried for most of the route using a | new plough being built in I the UK for British Telecom n conjunction with the Danish Posts and Telegraphs Department It differs from most existing ploughs for submarine telecommunications cables in that the plough blade has been designed to disturb the sea-bed as little as necessary, to give immediate, good cover with the added advan- tage of reducing the tow- ing force needed. In areas of the UK-Belgium 5 route where burial was not poss- ible. rock armour cable was used. Repeaters The repeater housings contain optical regenerators, one for each fibre, with equipment for power feed and remote fault-location. An optical regenerator has tour main parts: a receiver compris- ing a photodiode and an amplifier; an electrical de- cision circuit; a retiming unit; and a transmitter which includes a semicon- ductor laser. The regenerator examines every element, or 'bit' of received digital signal and generates it anew for onward transmission. Most of the regenerator consists of highspeed electronic circuits; the cir- cuit tunctions are far more complex than those of repeaters for coaxial systems which often use simple, three-transistor I amplifiers. So it has been necessary to progress from using separate transistors to employing integrated circuits (ICs). The re- quirements for the ICs are very demanding. First, they must handle data at rates in excess ot 300 Mbit/s, which means using emitter-coupled logic (ECL) circuits. ECL is suitable for high-speed operation because the saturation states of the transistors are controlled to avoid conditions requiring significant recovery times. Furthermore, the circuits allow signals to be passed between electronic gates using balanced ar- rangements that avoid the parasitic inductance en- countered when local earth is used as the return path. Second, the ICs must be very reliable: the target failure rate for a complete transoceanic system is a maximum of three failures in 25 years, and that in- cludes all type of compo- | nent. For ICs themselves, the target is better than one failure in 25 years from 15 000 devices. In the UK. the problem of 1C designs has been ap- proached by building on previous experience of semiconductors. For many years British Telecom Research Department has manufactured high re- liability silicon transistors for use in coaxial sub- marine systems The last j generation of these was I the Type 40 transistor sup- signal onwards. plied by BT for use in the 45 MHz repeaters manufactured by SIC The same fundamental technology has been used to construct ICs for use in optical repeaters, building on the proven re- liability of the Type 40. Pro- cessing technology for this includes the use of a special gold-titanium metallization process which is inherently very reliable; other processing techniques are substan- tially the same as for the Type 40. though certain additional steps are necessary in making a complete 1C In this way a series of ICs known as ECL40 has been developed for use in UK Belgium 5. The silicon- diffused transistors and resistors of the ICs are in the form of an uncommit- ted array, whereby various circuit configurations can be obtained simply by using different metalliz- ation patterns to connect the transistors and resistors. Because metallization is the last step in processing the silicon wafer, this ap- proach means that to pro- vide a new circuit to do a particular job only a new metallization pattern has to be made instead of new designs of diffusions as well as metallizations, which have to be made for fully custom-designed ICs. After metallization the individual silicon chips are mounted in ceramic chip carriers and are rigorously tested to ensure reliability. Other key components used in optical regenerators include the lasers and photodiodes, and Surface Acoustic Wave (SAW) filters, used in the retiming circuits to filter the clock frequency signals from the data stream. Lasers and photodiodes used in most optical systems at 1.3 are based on the com- pound lll-V semiconductor InGaAsP, but germanium avalanche photodiodes are proposed for receivers by some manufacturers The task of producing suitable lasers from 4-44 InGaAsP has proved far from easy, but suitable devices are now available and undergoing life tests Other planned cables In addition to UK-Betgium 5, other contracts for op- tical submarine cables are In hand. The TAI-8 cable is due to be laid in 1988, linking the USA with the UK and France, using a submerged branching unit on the edge ot the European continental shelf. The firm ATT ot the USA is to provide the main transatlantic link trom the USA to the branching unit; STC and the French company Submarcom are providing the links from the UK and France to the branching unit. A cable spanning the Pacific Ocean is planned on a similar time-scale, to be made by the USA and Japan; other systems both long and short, are under discussion. Short systems of up to 150 km otter in- teresting possibilities tor developments in optical technology means that spans of this length can now be contemplated without having to use in- termediate repeaters giv- ing a useful saving in cost. Such systems will offer an attractive option on routes which are too long tor satisfactory performance from microwave radio links Cable or | satellite Arguments ot the relative merits ot satellites and cables are very complex. Each case must be treated on its own merits bearing in mind techni- cal, economic and political factors Never- theless, there are a tew broad principles that may help in understanding some of the factors influ- I encing decisions about which to use First, satellite and cable links both continue to get cheaper as each new generation otters lower cost per circuit. Second, a signal crosses the Atlantic ocean by cable in about 30 ms whereas with satellites now in use the delay is about 260 ms because of the high orbit required for geostationary operation. This means that cable cir- cuits have advantages tor certain applications of speech and data. Third, cable tends to be cheaper on short routes j because cost is roughly j proportional to length. For satellite circuits the cost is independent of length, making satellites pro- [ gressively more com- | petitive on longer routes. Fourth, cable is essentially a point-to-point carrier and is especially ap- j propriate tor routes with a reasonable concentration ot traffic Satellites may be more appropriate where light traffic originates over a wide area. Filth, submarine cables have a design life of 25 years as opposed to be- tween seven and ten years tor satellites Cable therefore becomes more attractive (or a longer-term view of financial planning. Last, many telecom- munications administra- tions like to use a mix of circuits of different types tor diversity and security. It now seems certain that the introduction ot optical technology to submarine cables will increase the demand tor cable circuits and there is good reason to believe that circuit costs will continue to fall as the technology is developed. LPS HIGH-RESOLUTION COLOUR GRAPHICS CARD - 7 After the general programming information given in parts of this series (see Etektor Electronics — January 1986), the present article enables any user to get the video interpreter up and running on his computer system, j Hexadecimal patching Owing to lack of space, the video in- terpreter developed for the high- resolution card (see Elektor India. February 1986) can not be presented in the form of a source listing to assist users who wish to make their own modifications to it. None the less, the hexadecimal dump provided with this article, together with the information below, will be sufficient help to adapt the in- terpreter for any individual purpose. Here are the main points to observe for patching the program; • The address of the initialization routine is B000; that of CHROUT is B003, while that of the (optional) character reception subroutine is B006. When CHROUT is called, the character to be transmitted must be present in the CPU accumulator. • The graphics card selection ad- dresses are E150. . .E15F. Table IS. To facilitate modification of the graphics video in- terpreter. this hexdump of the program shows underlined and shaded bytes which require i • The 6502 zero page is used, but its : contents not modified; this goes for the stack as well. The video ; I buffer area is defined as 6000. I . . AFFF. • The scratch area for the in- j terpreter is BF80. . .BFFF. which | implies the presence of RAM at I these locations. • No external routines are required, except for BREAK testing, and for I keyboard character reception if j CHR1NP is called for. Except for ] these subroutines, the interpreter is a fully autonomous program. I The following are hints to effect the j necessary modifications to the given set of arrangements. If the graphics | card is intended solely for use as a | graphics terminal, enter an RTS in- i struction (opcode 60) in location B009; this is done if the BREAK and | character read routines can be dispensed with. If. however, the | BREAK test is to be retained, ad- dresses B00A and B00B must contain a vector, pointing to a routine that I detects depression fo the BREAK I key and returns with the ASCII con- trol character 03 present in the accu. | To select another character for this | purpose, modify the relevant byte at B00D. Next, the address of the BREAK handler, ie. the routine the ! CPU has to jump to when BREAK has been depressed, is entered at B016 and B017. As customary in 6502 pro- gramming, all vectors should be entered with their LSBs preceding the MSBs (reverse order). Using the interpreter character reception and cursor movement routines is slightly more com- plicated, but proceed as follows: B029 and B02A contain a vector poin- ting to a suitable keyboard scan routine, which may only be left when a key has been depressed and with the corresponding character available in the accu. Contrary to this wait routine, locations B023 and B024 contain the address of a routine that waits only when a key is depressed, in order to load its ASCII value into the accumulator. They keyboard strobe detection routine may be found between B018 and B01F in the video interpreter. All addresses rel- evant to these modifications appear shaded in the hexdump of Table 15. Relocation of the video interpreter may be accomplished by modifi- cation of the underlined address MSBs (B0...BA) and scratch area pointers (BF). If. for instance, the card should be decoded at D4xx, all underlined MSBs El should be re- placed by D4. Running the inter- preter from EPROM implies moving the BF80 . . BFFF scratch area to an appropriate RAM area; this may be effected, for instance, by replacing all underlined bytes BF with D1 to achieve a move to D180. . . D1FF. Mov- ing the entire interpreter to A000 . AFFF (RAM) is possible when all underlined bytes Bx are re- placed by Ax. Finally, adapting the size of the video buffer area (6000. . . AFFF) is straightforward because all underlined bytes 60 and AF may be changed into, say, 70 and 9F to locate the buffer into the ' 7000. . 9FFF area. Timing problems Now that a number of graphics cards i have been constructed, it has ap- peared that a few of these suffer from liming problems associated with | PROM Type 82S123 (IC.o, see the December 1985 issue of Elektor | India Some makes of this IC | are too slow (long access time), ' which results in effects like doubling i and instability of the image. Some of the cards appeared to be fully oper- ational in the monochrome setup, but caused trouble once the colour extension was added. A simple test may be performed to spot the trouble; put a dry finger on the RAM PCB connections to see if the picture quality is affected in any way; if this is not the case, the timing is correct. Do not run this test, how- ever, when the GDP is drawing, and watch for short-circuits caused by rings on your fingers! In case the pic- ture quality deteriorates, there are three possible remedies. The most simple solution is to use a PROM from a different batch (the year and week of manufacture are usually printed on the IC body; 8526, for instance, indicates the 26th week of 1985). The PROM programmer featured in the July/August 1980 issue of Elektor Electronics (UK) is emi- nently suited to try out the effects of a new PROM in the circuit. The other two solutions to the prob- lem involve minor modifications to the main graphics card. In certain cases, a pull-up resistor was suf- ficient remedy to put a stop to the undesirable effects; fit at 470 S type on the MUX line (for instance be- tween pins 7 and 16 of IC«). It may also be helpful to pass the MUX signal through gate N12, still available in IC??; this effectively delays MUX before it is applied to ICs and ICs . We would like to thank Mr S Lichtenberger of Hautot-Mer, France, for this suggestion. And yet, it would seem obvious that the very first cure suggested, replacement of the PROM, is the most preferable of the three. PL;DM INTERCOM A domestic intercom doesn't require a complicated circuit. The unit described here uses only a handful of components, even though an automatic gain control has been incorporated in the circuit. This intercom was designed to meet two requirements: low cost and high per- formance. To keep the cost down, the output stage is nothing more than a two-transistor class-A stage that can deliver 100 mW. To maintain good intelligibility in spite of this low power output stage, an automatic gain control is incorporated. This ensures that the output stage will nearly always be fully driven. To keep the cost of this gain control down, an OTA is used as the preamplifier - its gain is a function of a DC bias current. Small 1 50 Si loudspeakers are used both as microphone and as loudspeaker. The circuit Switch SI is the master ‘press-to-talk’ A button. When this switch is depressed, the loudspeaker in the main station (the upper loudspeaker in the diagram) is connected to the input of the ampli- fier. When the button is released, the other loudspeaker is connected. Resistors R2 . . . R4 and capacitor Cl produce a smoothed DC bias voltage for the OTA. R5 and C4 are included in the input circuit to reduce clicks during switch-over from 'talk’ to ‘listen’; they also reduce the effect of inter- ference pulses picked up by the (long) leads from the substation. The output signal from the OTA is fed to the ‘power’ output stage T2 and T3. This stage only gives a voltage gain of a factor 2, its primary function is current gain, to drive the loudspeaker. SI is wired in such way that the output is fed to the loudspeaker that is not being used as microphone at that moment. Understandably. . . . The automatic gain control is derived from T1 with its associated com- ponents. This transistor is connected as a current source. The maximum current it can supply to the bias input of the OTA (pin 5) is 0.6 V 47011 mA- This corresponds to a maximum gain of I the OTA of 2500 x. When the output I voltage rises above a certain level, current starts to flow through D3. This raises the I base voltage of T! towards the supply j level, thereby reducing the current through Tl. This in turn reduces the gain of the OTA. This automatic gain control action is ‘tamed' by including C5. j Installation In most cases, the wiring to the sub- R2 = 22 k R3.R4.R6 - 10 k R5.R10 = 1k2 R7 = 10 M 1 R8 = 470 SJ R9 = 33 S2 R12.R13 = 2M2 R14.R15 = 1 k C1.C3- 100 ji/16 V C2,C5= 10 m/16 V C4 = 100 n C6 = 2m2/16V C7 “ 100 m/25 V Semiconductors: IC1 - CA3080 Tl - BC557B.BC1 77B.2N 1711 T2 ■ BC517 (darlington) T3 ■ BC160.BC161.2N3553 D1... D3- DUS.1N4148 Cooling fin for T3 2 loudspeakers, 150 ill 1 Watt Switch or pushbutton, 2P2T power supply : Trafo 12 ... 1 5 volt, 100 mA Fuse 100 mA slow blow (and Bridge rectifier B40C400 1000 m.25 Velco station can be normal two-core cable. However, if the distance is too great or if the leads run close to mains wiring it may be necessary to use single-core screened cable. Since the output stage is running in class-A, the current consumption is too high for battery operation. It would be possible to use batteries if an on/off switch (s included in the main station, but this would mean that the substation cannot initiate the conversation. For IEIH Ip 'Mm SB ■I battery operation, the supply voltage can be reduced to 9 V, although this will reduce the output level. A better solution is to use a mains- driven supply. A transformer with a 12 ... 15 V secondary which can supply 100 mA is sufficient. Connect this to a bridge rectifier followed by a 1 000 (i/25 V smoothing electrolytic. The gain of the intercom will vary be- tween a maximum of 5000 x and a minimum of 150 x, depending on the I makes it much less effective. automatic gain control. The maximum is set by R8 ; increasing the value of this resistor will reduce the maximum gain. Do not decrease the value, as this can damage the 1C. The minimum is set by the value of R7; this should not be altered. SI can be either a switch or a push- button, according to taste. In either case, it should be a break-before-make tlakior India ap.il 1986 4-49 MAGNET1SER During the second ‘Bioclimatological Colloquium' which took place in September 1976 in Munich, a report was presented of a series of experiments carried out initially by Professor Dr. R. Mecke of the University of Freiburg and continued by several researchers of the University of Tubingen (Dr. W. Ehrmann. Dr. W. Ludwig et al.). 920 patients who complained of psychosomatic ailments were treated with a device which is the model for the ‘magnetiser’ described in this article. Of these 920 patients, 220 received a placebo, i.e. the device was a dummy. The complaints of the patients included insomnia and chronic headaches; since 1975 patients suffering from such ailments as migraine, neuralgia, extra- articular rheuma, damaged joints, neck and back pains, skin allergies, bronchial asthma, travel sickness and fear of heights have also been treated.lt is significant that during the above exper- iment, the patients required approx. 50% less medicament than normal. The overall results of the experiment (shown in table 1 ) are quite remarkable, particu- larly when one bears in mind that they are far better than the results obtained by the use of pharmaceutics. The figures given are all from a report released by W. Ehrmann, W. Ludwig, and their colleagues at Tubingen University. Our thanks go to Dr. Ludwig for his co-operation in the preparation of this article. The device which is described in the remainder of the article, is of the same type as that used in the above exper- iment. It should be stressed that, although F.lektor cannot offer any guarantees as to the efficacy of this treatment, the device is by no means to be considered in the same light as copper bracelets and old potatoes, but rather is a scientifically based approach which merits serious medical consideration. The effect of magnetic fields The penetration of an alternating electromagnetic field is determined by its frequency. As long as the frequency is in the ELF (Extremely Low Fre- quency) range, the electric field can be ignored. The alternating magnetic field Recent medical experiments have lent weight to the idea that magnetic fields are of therapeutic value in the treatment of psychosomatic complaints and rheumatic ailments. The following article, which is preceded by details of a controlled | experiment into the efficacy of this method of treatment, describes a device which will produce an alternating magnetic j field of the type suitable for medical use. on the other hand, will induce eddy currents throughout the entire organism, thereby causing shifts in the charge of the cell membranes. This stimulates the nervous system, removing any blockages which may exist. For example, it was noticed that at frequencies below 8 Hz, a widening of the blood-vessels occurred, whilst at fre- quencies above 1 2 Hz the blood-vessels Experiments have also shown that the sensitivity of an individual to magnetic fields can be quite frequency-dependent. It is at a maximum at the frequency which coincides with the alpha-rhythm of that person’s EEG. This is readily explicable in the light of the fact that externally induced pulses will obviously have the greatest effect upon pulses with which they are synchronous. Steep pulses which have a large number of harmonics produce better results than sinusoidal fields of similar ampli- tude. However the rise time need not be shorter than the response time of the The therapeutic ELF-frequencies lie between approx. 0.5 Hz and 20 Hz, and can be subdivided into 4 treatment- specific groups: 1 ... 3 Hz, counteract infections; 4 ... 6 Hz, have a soothing effect, and counteract muscular spasm; 8 ... 1 1 Hz, act as an analgesic, as a tonic, and exert a stabilising influ- 13 ... 20 Hz, for patients who suffer from over-tiredness, these fre- quencies have the same effect as 8 . . .11 Hz have upon ‘normal’ patients. The last group of frequencies is only used when lower frequencies have had no result. The 4 ... 6 Hz range should not be used whilst the patient is engaged in activities which require increased concentration (e.g. operating machinery, driving etc.). Treatment with magnetic fields is not known to produce any side-effects, although persistent use may result in a lessening of its efficacy. It is therefore recommended that, for the time being, a treatment session should not last longer than 15 minutes. Patients with a heart pacemaker should not be treated with the lowest frequency range unless it is known for certain that it will not react to the magnetic field. For normal use, i.e. when not applied to a localised area of pain, the magnetiser can be carried in a jacket pocket or waist pouch. If used when lying down, it can be placed under the neck or beneath a cushion or pillow. The circuit Figure 1 shows the circuit diagram of the magnetiser. The circuit contains two astable multivibrators, one of which (N1/N2) oscillates at approx. 1.15 Hz, the other (N3/N4) at either 4.4 Hz, 9.7 Hz or 14.2 Hz, as selected by SI ... S3 respectively. Some further frequencies are obtained by closing more than one of the switches: SI + S2 = approx. 3.0 Hz; 51 + S3 = approx. 3.4 Hz; 52 + S3 = approx. 5.8 Hz; SI + S2 + S3 = approx. 2.5 Hz. Transistor T1 is turned on and off in time with the chosen frequency. The pulsed collector current magnetises the core of coil LI, which consists of 600 turns of 0.2 mm diameter enamelled copper wire (38 SWG). In the Elektor lab a normal 'steel’ bolt 40 mm long and 6 mm in diameter was used as the core. The coil may be scramble wound, i.e. the turns need not be wound in layers. The resultant field strength is com- parable with that obtained from commercially available devices. To prevent possible risks arising from a defect in the second AMV, it is rec- ommended that in devices intended for use by patients with a heart pacemaker, components Rl, R2, R5,C1 and C5 are not soldered onto the p.c.b., and that the free input of Ni be connected to the positive supply rail. Figure 1. Circuit diagram The device requires only Figure 2. Copper of the printed cin (EPS 9827). Bibliography: G. AUmann, 11969 1: Die physiologische Wirkung elektrischer Felder auf Organismen. Arch. Met. Geoph. Biokl. 17: 169-290. S.M. Bawin, L.K. Kaczunarek, W.R. Adey 11975): Effects of Modulated VHF Fields on the Central Nervous system. ANN. New York Acad. Set. USA, 24 7: 74-81. D. E. Beischer.J.D. Grisser, R.E. Mitchell, 11973): Exposure of Man to Magnetic Fields Alternating at Extremely Low Frequency. NAMRL-1180. Dokumenta Geigy 11968): 109-123 W. Ehrmann, H. i\ Leitner, W. Ludwig, M.A. Persinger, W. Sodtke. R. Thomas (1976): Therapie mil ELF-Magnet- feldern. Z. Phys. Med., 5: 161-170. b'. Ehrmann, H. v. Leitner, W. Ludwig, M. A. Persinger, W. Sodtke, R. Thomas (1 976): Entwicklung eines elektro- medizinischen Taschengerates. Acta Medicotechnica, 24: 282-285. N. Geyer, G. Fischer, 11. Riedl, H. Strampfer, (1976): The effect of an Artificial Electroclimate on Physiological Values. Arch. Met. Geoph. Biokl. Ser. B. 24: 111-112. E. S. Maxey, (1975): Critical Aspects of Human versus Terrestrial Electro- magnetic Symbiosis. USNC/URS1- 1EEE Meeting, Boulder, Colorado,. US A. ©*©*® = 25Hr 4-53 selex-ii Digi-Course II Chapter 5 We described the divider circuits based on Flipflops in the last chapter of Digi Course-ll. We have seen that by feeding a series of pulses at the input of a Flipflop. we get only half the number of pulses at the outputs of the Flipflop. The input pulses toggle the Flipflop ON and OFF for every pulse, alternately. By cascading many such Flipflops together, it is possible to obtain a division by 4, 8. 16 1 As can be seen from the sequence of binary numbers and the corresponding number of pulses, the group of 4 LEDs functions as a pulse counter. This pulse counter can count from 0 to 16. and is suitable for hexa decimal system. With the 1 6th pulse, the counter resets to 0000 and starts counting again. Our observation about the cascaded dividers also applies to the counter, each additional Flipflop increases the counting capability by a factor of 2. Thus a counter with 5 Flipflops cascaded together would count from 0 to 32. It will reset on the 32nd pulse. A counter with 6 Flipflops will count from 0 to 64 and so on. An adaptation of the hexadecimal counter to the decimal system is already known to us and it is given here again in figure 2. This counter resets to 0000 on the tenth 2 The decoder part using gates T, U, X and Y reacts to the binary combination 1010. We can reorganise the decode also to react to another binary combinations like 1 100, which, then will reset on the 12th pulse and will function as a Duodecimal counter. | This type of decoder circuits are very important to the Computer Technology. The Central Processing Unit. (CPU) | which is the brain of the computer works with various I peripheral devices like the Keyboard. CRT Screen. Printer | The CPU can select the desired peripheral by sending a binary code to the decoder circuit, which decodes the | binary code and turns ON the Interface to that particular peripheral device. Even in the Digital Technology, decoders have an important place. The most commonly encountered decoders digital circuits are the BCD-to-Decimal decoders and BCD to 7-Segment decoders. The standard ICs avaialble for these functions are 7442 and 7447. Figure 4 shows the pin connections of 7442 4 SN 7442 This 1C can convert the output values available at the outputs E.F.G.H of our decimal counter to the corresponding decimal number. The particular pin corresponding to that decimal number is made "0" by the decoder, whereas all other 9 pins remain ”1" This decoder is not suitable for driving a seven segment digital display, which is the most commonly used display device in digital technology. The seven segment LED display consists of seven tiny bar shaped LEDs arranged to make the figure of 8. The decoder 1C 7447 has seven outputs which are connected directly to these seven segments. (In practice, current limiting resistors are also used, one for each segment) 5 SN 7447 6 Let us go back to our decimal counter again. The 4 gates used in the decoder specifically react only to the 1010 combination. It will also be enough to decode only the 1st and 3rd binnary position to check if it is "1". Decimal ten is the first number which gives a ”1" at the 1st and 3rd binary position simultaneously. If this condition is used to reset the counter the remaining decimal numbers which also give rise to this condition, will never be encountered as the counter will always reset on the tenth pulse. This simplified decoding arrangement is shown in figure 7. The simplified decoder is called an "Incomplete Decoder' compared to the Complete Decoder shown in figure 2. However, even when using the incomplete decoder, the first combination that resets the counter is 1010, and remaining combinations like 1011. 1110 and 1111 are never allowed to reach. Another important point to note is the spurious triggering that many take place and affect the functioning of the circuit. To take care of this problem, connect all unused inputs to "1 " (Pins 4/9. 12/16. 2 and 7). selex 4-53 CAPACITORS selex "Guess whal I have got today, it looks like a candy, has two legs, has a low resistance in the beginning when measured with a multimeter but shows open circuit after some time. I have taken it out from my old pocket radio "Well, candy reminds me of Condensors! Show me what you have brought!" "Oh! It is really a Condensor.” "What kind of resistance is this? " "No. it is not a resistance, it is a Condensor. or a "But my multimeter showed a low resistance in the beginning and then it went on increasing to infinity "I will explain it later, first let us see what the Condensor does." "If it sort of a resistor that slowly opens up to a high "As a matter of fact. Condensors have nothing to do with resistance. Condensors. more commonly called Capacitors, store electrical charge. When current is allowed to flow into a Capacitor, it accumulates the charge flowing into it. The Capacitor is said to be charged. When we provide a conducting path to this charge, it can flow out of the Capacitor once again "Then why these Capacitors are not called "Capacitors and Accumulators do have similar function; but Accumulators are slower than Capacitors in charging and they can store much more charge compared to the Capacitors. Capacitors are quickly charged and discharged." "Which is the circuit symbol used for Capacitors ?" "The Capacitor symbol consists of two parallel bars, which represent the two Capacitor plates ' "What are these plates? I don't see any plates in this Capacitor I" "The symbol of plates has come because of the old Condensors which really consisted of two large metallic plates insulated from each other and placed parallel to "These two simple plates can store electrical charge? ' "Yes. however, they must be quite large and must be sep3rated by a very small distance. Such plates are not used any more. The plates are now replaced by thin aluminium foils, separated by a thin insulating plastic film. To increase the total area of the foil, this sandwitch of foils is rolled up to make the Capacitor, the two wires coming out from the capacitor are internally connected to these two aluminium foils." "Wait a minute, you just said that these two foils are totally insulated from each other?" 'That is right!" "Then how did the multimeter show a low resistance in the beginning?" "Yes. that happens only for a brief period, during which the charging current flows into the capacitor After that, the multimeter slowly goes towards infinite resistance. The battery inside the multimeter provides the charging current to the capacitor through the test terminals. When a multimeter measures resistance, what it really does is. it measures current flowing through that resistance and knowing the voltage available at the test terminals, the reading is directly calibrated in Ohms. When the charging current flows through the capacitor, an equivalent resistance value is shown by the multimeter. As soon as the capacitor is fully charged; the current stops and the multimeter shows infinite "This is exactly similar to the Accumulator. But what voltages do the these capacitors have?" "I don't understand what you are saying, what voltages can the Capacitors have?" "Like 1.2 Volts of the Nickel-Cadmium Accumulator batteries!" "Oh, if that is what you mean, the capacitors have no such voltages. The uncharged Capacitor has no voltage on it. and as the charge builds up. the voltage goes on increasing." "You mean the voltage on the Capacitor is an indication of the filling level of the Capacitor ?" "Yes. and one must always consider the storage capacity of a particular Capacitor In case of large Capacitors, the voltage increases slowly than in case of small "Like in a bucket the water level rises quickly than in a bath tub " "Quite right! But to be more precise a capacitor can be compared to a tube in the cycle tyre. The pumping of air is initially quite easy, because the tube is empty, and becomes more difficult as the pressure inside the tube builds up. The pressure inside the tube can be compared to the voltage on the Capacitor and the air can be compared to the charge in the Capacitor. "Can a Capacitor also burst like the tube if we try to force more charge onto it and raise the voltage beyond its limits?" "Even this is true in capacitors, they can rupture or even burst with a loud noise if a high voltage is applied beyond its specified value." "Well, then we can even design a voltage operated bomb using capacitors!" 4-54 elektor i selex Different Types Of CAPACITORS The basic principle behind all capacitors is same, there are j two metallic surfaces parallel to each other, insulated from each other and each connected to a terminal. However there are many different types of materials and methods of construction being used in the manufacture of capacitors The insulating material between two metal surfaces is known as dielectric materials. A plastic film is generally used as dielectric Figure 1 shows one method of making a tubular capacitor. Two long strips of metal foil and dielectric foil are placed alternately and rolled up to make a tubular capacitor. Two wires are connected to the two metallic foils and brought Figure 2 shows the construction of a plastic film capacitor, which is made up of a number of aluminium and plastic foils stacked one above the other. These two methods are very efficient, because they practically double the metallic surface foil. Each foil has surfaces on both the sides. Many manufacturers have also developed techniques for depositing aluminium layer on plastic film which is then used to roll up and make the capacitors. Ceramic and Mica are seldom used as film capacitors. Both these materials are used in capacitors for high frequency applications like Radio and TV. Electrolytic capacitors are made by using a paper strip soaked in an electrolyte as the insulating material. After manufacturing, these capacitors are subjected to a voltage which makes the metal foil and produce a very thin oxide film, on the Electrolytic capacitors have very high capacitance, even upto several thousands of uF The highest capacitance manufactured uptill now is IF (1.000,000 uF). The polarity is very important in case of electrolytic capacitors, because reversal of polarity results in destroying the oxide layer. 1-55 The dielectric strength is generally low in these film capacitors. This results in comparatively low voltage ratings for the electrolytics. Tantalum capacitors is a modern development in electrolytic capacitors. The One of the recent developments in variable capacitance diodes. These are specially designed semiconductor devices posses a variable capacitance that depends c the applied voltage. Commonly used variable capacitors have very low values, generally around 100 pF (1 pF = 1 Pico Farad = 1 0’ 2 Farad) Most common application for these variable capacitors (Gang Condensors) is in the turning circuits of radio receivers. these type of capacitors. tuning circuit of large Radio sets use air as the dielectric insulating medium and have truly solid metal plates which are grouped alternately They are evenly spaced and one of the groups can be rotated around a common axis, so that the two groups of plates can be interposed decrease the capacitance. Miniature types of variable capacitors are made from thin metal foils and plastic film dielectric. These metal foils are grouped alternately together and one group is rotated around a common axis to change the effective interposed area. This, in selex CAPACITORS in series/parallel connection We have already studied the series and parallel adding up standard values through parallel Similar formula for capacitors in series is connections of two resistors. Two capacitors can also be connected in series or parallel “ i ]•■ connections For example, a 20 nF capacitor will not be available as a standard value, and to obtain 20 nF 1_ = J_ + L combination. However, the result is totally different Figure 1 shows a parallel combination of two capacitors. Cl and C2 are the values of individual T J we can connect two 10 nF capacitors in parallel The formula for parallel combination of capacitors is series combination of or - Cl • C2 ^ Cl +C2 effective value of the combination. The effective surface available is increased by the Rg = R1 • R2 The converse is also true. capacitors gives an effective both individual values of the capacitors. This type of combination is useful in parallel connection and as the total capacity is The formula for a series same as that for the parallel obtaining nonstandard increasing the voltage Vj Figure 1: the plates, the effective capacity is obtained by adding the two individual Cg = Cl + C2 This relation is quite useful in practice to obtain non standard values by just combination of resistors 1 _ 1 + 1 Rg R1 R2 or R - R1 x R2 9 R1 + R2 rating. In case of AC voltages, series combination of capacitors can also be used as voltage devider. THE FILTER CAPACITOR The filler capacitor used in battery eliminators is a simple form of stabilising circuit. It is just a single passive device which does quite an effective job. It stabilises the pulsating D C voltage coming out of the rectifier bridge by its charging and discharging characteristics. Figure 1 shows the circuu of a simple battery eliminator The mains A C. input is stepped down by the transformer Tr The bridge rectifier made of four diodes D1 to D4 converts the output voltage of the transformer to a pulsating D C voltage. The transformer output voltage wave form is shown in figure 2 a. During the positive half cycle of this voltage diodes D2 and 03 are forward biased and they ! start conducting. During the negative half cycle, diodes D1 and D4 are forward biased and start conducting. during both the half cycles, the voltage available at the output of the bridge (across the capacitor C) is always output voltage is shown in . figure 2 b. 4-57 selex Even though the output of the rectifier bridge is have a steady value. This shortcoming is corrected by. the capacitor c. the so called 'Filter Capacitor' When the voltage shown in figure 2 b is applied across capacitor c. (an Electrolytic capacitor) it accumulates the charge during the rising portion of the wave When the voltage from bridge rectifier output starts falling, the capacitor supplies some of its accumulated charge effective voltage to fall rapidly. The voltage does fall, but very slowly. By the time the voltage has fallen by a small magnitude, the output voltage from the bridge again starts rising and the capacitor again charges upto the peak value. This cycle continues and produces a voltage at the output which is shown in figure 2c. The process of charging and discharging of 3 are bridged by the filter capacitor which takes up charge during the 77 of the voltage voltage the capacitor is charged, during the falling part of the discharged. up of a steady D.C. voltage and an A.C. voltage superimposed on it. t r r r X- 83726X-3 the capacitor C is shown in figure 3. The minor fluctuation that still remains in the output voltage is called residual hum. The value of capacitance generally used is upto several thousand uF. The filter capacitor used here not only stabilises the output voltage, but it also serves another important function. Electronic circuits operating from this voltage may not always draw a steady current. The current requirement frequently varies over a wide range. Sometimes the circuit may draw a high current for a short period. This high charge to be supplied during, that period, and this charge is also supplied by the electrolytic capacitor. The voltage thus remains considerably steady even during peak loads. The function of the filter capacitor can also be explained in simpler terms. The capacitor can be considered as a short circuit for A.C. and an open circuit for D.C. If we consider the pulsating D C. voltage in figure 2b as a combination of a stable D.C. and an A.C. voltage superimposed on A.C. part and an open circuit for the D.C. part. The A.C part is thus prevented from going over to the output, and can be said to be effectively 'Filtered' out by the capacitor. t50v470*f 50 ' 30mF 50v330uF 50V I* 4-58 selex THE DARLINGTON PAIR Figure 1 shows a circuit which is known as the Darlington Pair, it is just simple connection of two transistors, but the result it gives is quite astonishing. It almost multiplies the gam of one transistor by that of A Darlington Pair of transistors having individual current gains (0) of 200 each will give an amplification almost as high as 40000 ! >r T1 simple. Transist amplifies its base current at B1 by a factor 0 . which is the current gain of transisto T1. The amplified output current is available at El. which is same as the base current of transistor T2 Transistor T2 also amplifies its base current at B2 by a factor of 0 . which is also the current gain of T2. Thus the total amplification available is 0' Assuming a typical current gam p- 200 for both T1 and T2. we have the effective current gain of the Darlington Pair as 200 x 200 = 40000. In this process, the effective base to emitter voltages of the two transistors are added up and we need a driving voltage which is threshold voltage of each transistor (approximately 0.6 volts each.) In practice, we need not connect the cirucit as shown in figure 1 The Darlington Pair is available as a single package with three single transistor, but has a threshold voltage of 1.2 to 1 4 Volts and gives a very high current gam. Some of the standard Darlington Transistors available are listed in table 1. along with the important specifications. In case of Power Darlingtons. feedback resistances are incorporated transistor to ensure a stable operation. This reduces the effective current gam of the than the theoretically expected current gain multiplication. A parallel diode is also incorporated across collector and emitter to protect the transistor Darlington transistor packages with three or more transistors connected in the Darlington configuration are emitter voltage increases by 0.6 to 0.7 volts with each additional transistor. Table 1 BCS16 BC517 B0675 BD676 BD677 BD678 BD679 BD680 PNP NPN PNP PNP S internal circuits 1. 59 check list for electronic fault finding or 'where and how to look for what that doesn't' Before soldering in components ■ Check that the components agree with the parts list (value and power of re- sistors, value and voltage rating of capaci- tors, etc . . .). If in any doubt, double-check the polarized components (diodes, capaci- tors, rectifiers, etc . . .). ■ If there is a significant time lapse between last reading an article and building the circuit, take the trouble to re-read the article; the information is often given in very condensed form. Try to get the most important points out of the description of the operation of the circuit, even if you do not understand exactly what is supposed to happen. ■ If there is any doubt that some compo- nents may not be exact equivalents, check that they are compatible. ■ Only use good quality IC sockets: ■ Check the continuity of the tracks on the printed circuit board (and through- plated holes with double-sided boards) with a resistance meter or continuity tester. ■ Make sure that all drilling, filing and other ‘heavy’ work is done before mount ing any components. ■ If possible keep any heat sinks well isolated from other components. ■ Make a wiring diagram if the layout involves lots of wires spread out in all directions. ■ Check that the connectors used are compatible and that they are mounted the right way round. ■ Do not reuse wire unless it is of good quality. Cut off the ends and strip it After mounting the components ■ Inspect all solder joints by eye or using a magnifying glass and check them with a continuity tester. Make sure there are no dry joints and no tracks short circuited by poor soldering. ■ Ensure that the positions of all the components agree with the mounting diagram. ■ Check that any links needed are present and that they are in the right position to give the desired configuration. ■ Check all ICs in their sockets (see that there are no pins bent under any ICs, no neighbouring ICs are interchanged, etc . . .). ■ Check that all polarized components (diodes, capacitors, etc . . .) are fitted correctly. ■ Check the wiring (watch for off-cuts of component leads); at the same time ensure that there are no short circuits between potentiometers, switches, etc . . ., and their immediate surroundings (other components or the case). Do the same with mounting hardware such as spacers, nuts and bolts, etc . . . ■ Ensure that the supply transformer is located as closely as possible to the circuits (this could have a significant influ- ence in the case of critical signal levels). ■ Check that the connections to earth are there and that they are of qood quality. ■ Check that any pins, plugs or other connectors used are making good contact. ■ Make sure the circuit is working correctly before spending any time putting it into And if it breaks down . . . ■ Recheck everything suggested so far. ■ Reread the article carefully and clarify anything about which you are doubtful. ■ Check the supply voltage or voltages carefully and make sure that they reach the appropriate components especially the pins of the ICs (test at the pins of ICs and not the soldered joints!). ■ Check the currents (generally they are stated on the circuit diagram or in the text). Don’t be too quick to suspect the ICs of overheating. a If possible check the operation of the circuit in separate stages. As a general rule, follow the course of the signal. ■ Check the contents of any PROMs or EPROMs fitted. ■ While checking voltages, currents, fre- quencies or testing the circuit with an oscilloscope, work systematically and take ■ It is always a good idea to do any fault finding as a combined operation with a friend, two heads are better . . . ■ Be wary of ’red herrings’ when fault tracing. Do the simple checks first. ■ Finally, remember our constant com- panion Murphy is looking over your shoulder. If that part of the circuit cannot possibly be wrong and you haven't checked it - that's where to start looking. ■ . . . And don’t forget to switch the power on and check the fuses! 4-60 elokto 4-61 TTL ICs Important features of all ERX models are rec- tangular current limiting (for driving non-linear loads); +10%. -20% voltage adjustment; over- voltage protection; a holding time of 30 ms typical, 20 ms minimum to enable orderly shutdown at power failure; remote error voltage sensing; and selectable 115230 VAC in- put. Cost of the new switcher is $179.00, single unit price. Kepco Inc 131-38 Sanford Avenue Flushing. NY. 11352 USA. Telephone 010 1 212 461-7000 TWX 710-582-2631 (3417:1 :F) 240-Watt switched power supply with green baize under- trim and non-slip feet. R A Kent Engineers 243 Carr Lane Tarieton, Preston Lancashire PR4 6YB Telephone: (077 473) 4998 (3417:9) tips specially made as re- placement in Weller TCP and ECP temperature con- trolled irons. The range has recently further been improved by an additional treatment of the areas wet- ted by the solder. Various tip designs are available as shown in the photo- graph. Cobonic Limited 32 Ludlow Road Guillford Surrey GU25NW Telephone: (0483) 505260 Telex: 28604 (3382:18:F) New in their ERX series of single-output L-chassis switching power supplies, Kepco Inc. and TDK jointly introduce a 240 Watt unit, available in 5-, 12-. 15-, and 24 Volt models. These are dual-FET forward con- verters. operating at a fre- quency of 100 kHz with an efficiency of 80%, achieved through the use of a recently developed TDK ferrite called H7C4. As with the other members of the series, the noise densi- ty of the new ERX unit is low enough to comply with the VDE 0871/6.78 re- quirements from 450 kHz to 30 MHz. Driver iCs feature Bimos // technology Sprague has introduced the UCN5800A and UCN5801A latched drivers, high voltage, high current integrated circuits which, fabricated using bipolar/ MOS technology, provide a very low power latch with maximum inter- face flexibility. Type dependent, the devices contain four or eight CMOS data latches, a bipolar Darlington driver stage for each latch, and CMOS, PMOS, and NMOS compatible inputs for latch control circuitry. Both units have open-collector output and integral diodes for inductive load transient suppression. The output transistors are capable of sinking 500 mA and will sustain at least 50 V in the Off state. Appli- cations include use with relays, solenoids, stepping motors. LED and incan- descent displays, and other high-power loads. Sprague Electric UK Ltd. Salbrook Road Saifords Surrey RH1 5DZ. (3417:5) The 3S-TIP is a range of long-lasting soldering iron Kits for making morse-code keys A low-cost kit, available from R A Kent Engineers, contains all the necessary components for making morse-code keys. Developed for radio amateurs, the key can be assembled in less than Made of solid brass, the key is pivoted on ball race bearings and has solid- silver contacts to ensure accurate and reliable performance. It has fine- pitch threaded screws with instrument knurled heads to allow for precise ad- justments to be made. The kit is supplied com- plete with detailed as- sembly instructions, and the manufacturers can also supply a French- polished hardwood base I HCMOS TTL j compatible oscillators A complete range of HCMOS quartz crystal clock oscillators Is now I available from M-Tron i through UK stockist and j distributor MCP Electronics. Frequencies between 3.0 and 25 MHz are available. The waveform rise and fall time under ten TTL loads exceeds TTL requirements. designed as a pin-tor-pin Xj' 1 1 j j j INJ compatible replacement for the industry standard 1 1 j I Wjy W MP7533 and otters the ad- m INflfl ditional advantages of ^\!l | tHIH lower output capacitance I and very high linearity If and accuracy. The low ,n 9 symbols incorporated \ output capacitance results on ,he LCD Three different i in higher speed, taster set- versions are available with | fling, and easier interfac- ing with output amplifiers. The linearity and accu- racy are ol the order ex- ®SSSE« peeled ol a 12 -bit device. Micro Power Systems (UK) ttk .. 9H I ... I QJ Limited .jS cJM I J Orion House IbmJ 49 High Street Addtestone | Surrey KT15 ITU |J; *_4 Telephone: (0932) 57315 | Telex. 8812984 (3382:14:F) ; Standard versions have a frequency tolerance ot ±50 p.p.m. and a stability | ot 100 p.p.m. over the tern- i perature range 0 to 70 °C MCP Electronics Limited 26-32 Rosemont Road Aiperton Middlesex HAO 4QY Telephone: (01) 902 6146 (3382:15:F) I 10-bit CMOS multiplying DAC Miniature DiL DPMs A new range of miniature LCD DPMs (digital panel meter) has been intro- duced by Lascar Elec- tronics. All types utilise surface mount techniques to vastly reduce the overall size. The DIL format is claimed to make the meters particularly easy to use by low or high volume Each meter is also sup- plied with a 'snap-in' bezel for fast fitting. Standard features include auto-zero, auto-polarity, 200 mV fsd, programmable decimal points and 'Low Battery' in- dication, together with a range of useful engineer- 1 character heights of 15mm, 12.5mm, and j 10mm, the latter type | (DPM400) being the world's l smallest off-the-shelf DPM. With a one-off price of £16.95 it is also probably the cheapest. Lascar Electronics Ltd. Module House. Whiteparish Salisbury. Witts SP5 2SJ. Telephone: (07948) 567 Telex: 477876 (3417:8:F) Digital LCR meter New trom Advance House of Instruments is the SOAR Model 5700 digital LCR meter which provides a wide selection of measurements including contact resistance on switches and relays, inter- nal resistance of batteries, and junction capacitance in semiconductors. The model 5700 features auto- ranging and a measure- ment mode which auto- matically selects the optimum range in measur- ing unknown component values. Two VH digit LED displays, each with a maximum reading of 1999, are provided for measure- ment indication. Advance House of Instruments Raynham Road Bishop's Stortford. Herts CM23 5PF Telephone: (0279) 55155 Telex: 81510 (3417:3:F) ^ j 15 3 ■ b 4-65 ELECTRONICS HOBBYISTS How long will you continue to play with simple projects and obsolete technologies? Electronics Hobbyists all over the world are experimenting with Microprocessors and their most exciting applications. Now Dynalog makes it possible even for you to get into the exciting world of Microprocessors. Dynalog has specially designed two LOW COST Microprocessor Trainers based on the popular chips 8085A and Z80A. Advanced Hobbyists, Students and Beginners can use them initially for learning and then for experimenting with Microprocessor Applications. Write today for Dynalog Micro-Systems 14, Hanuman Terrace, Tara Temple Lane, Lamington Read, Bombay 400 007 Tel: 36242 1,353029 Telex: 01 1-75614 SEVK IN Gram: ELMADEVICE Branches and representatives at: Pune. Bangalore. New Delhi, Hyderabad and Chandigarh classified ads. Advertisers Index CONDITIONS OF ACCEPTANCE OF CLASSIFIED ADVERTISEMENTS 1 ) Advertisements are accepted subject to the conditions appearing on our current rate card and on the express understanding ihat the Advertiser warrants that the advertisement does not contravene any trade act inforce in the country. 2) The Publishers reserve the right to refuse or withdraw any advertisement. 3) Although every care is taken, the Publishers shall not be liable for clerical or printer's errors or their consequences. 4) The Advertiser’s full name and address must accompany each advertisement submitted. The prepaid rate for classified advertisement is Rs. 2.00 per word (minimum 24 words). Semi Display panels of 3 cms by 1 column. Rs. 150.00 per panel. All cheques, money orders, etc. to be made payable to Elektor Electronics Pvt. Ltd. Advertisements, together with remittance, should be sent to The Classified Advertisement Manager. For outstation cheques please add Rs. 2.50 ACE COMPONENTS 4 04 AFCO I. & C. LTD 4.06 APEX 4 08 APLAB 4 04 ATRON ELECTRONIC 4 67 COMPONENT TECHNIQUE .4 12 COSMIC 4 76 DEVICE ELECTRONICS 4 11 DYNATRON 4 05 ELCIAR 4 08 ELCOT 4 73 ELTEK BOOKS N KITS 4 67 IEAP 4 14 KAYCEE 4 66 KLAS ENGINEERING . • 4 14 LEADER ELECTRONICS 4 12 LEONICS 4.66 MECO INSTRUMENTS 4 04 MOTWANE 4,07 OSWAL ELECTRONIC CO 4.02 PIONEER ELECTRONICS 4.14 PLA 4.70 PRECIOUS KITS 4.13 RAJASTHAN ELECTRONICS. CG. 4.02 SAINI ELECTRONICS 4.08 SCIENTIFIC 4.10 SUCHA ASSOCIATES 4 12 TESTICA 4.10 TEXONIC 4.10 UNLIMITED ELECTRONICS 4.72 VASAVI 4 67 VISHA 4.75 YABASU 4 09 ZODIAC 4.72 Rs 90/ - 2) 1 2 Melodies horn for vehicles Rs 80 Kits price list Re 1/-. (50% Advance for VPP) PERFECT ELECTRONICS 453. Ganpati Ali. Wai - 41 2 803 (Satara) WANTED EIe Components are normally available with the following companies: INTEGRATED ELECTRONICS INSTRUMENTS 82-174 Red Cross Road Secunderabad 500 003 Phone: 72040 SIUKON ELECTRONICS 315. Lamington Road. Shop No. 4 Kalpana tension. Opp. Police Station Grant Road, Bombay • 400 007 Ph. 350644 VISHA ELECTRONICS 1 7. Kalpana Building, 349, Lamington Road Bombay ■ 400 007 Phone: 362650 DYNALOG MICRO SYSTEMS Tara Temple Lane, Lamington Road Bombay - 400 007. Phone: 353029. 3< ELECTROKITS 20. Narasingapuram Street (First Floor) Mount Road Madras • 600 002 R.N.NtT39HST/83 MH/BYW-228 LIC No 91 Watts News? After 30 years of communicating our finest efforts to you— we still have more news for you. Cosmic is now breaking every sound barrier in maintaining its sophisticated electronic image by touching perfection in the manufacturing of its T ape 'iscks/ recorders. Stereo Systems. Amplifiers, Turntables. Head-Phones. A single dominant factor has encouraged us to keep expanding and that is consumer satisfaction. Stay tuned to us. cdsithc We are Sound!