up-to-date electronics for Topamp 60 clean watts really trying! I played programming make all the contents elektor november 1979 — UK 03 november 1979 volume 5 number 11 page 11-14 To analyse the motion of fast-moving objects it is useful to be able to take a well-timed succession of photographs. One method is to leave the camera shutter open and produce a series of light flashes for the successive exposures. The flash sequencer uses only three ICs and a few other components to control five (electronic) flash guns. page 1 1 -24 For one who had never played around with microprocessors, the TV games computer was a fascinating gadget! The title of the article tells the story: I played TV games . . . . . . and it was fun! page 1 1 -34 The simple crystal- controlled short-wave converter is intended for use in combination with a conventional medium- wave-receiver — a car radio, for instance. m # (§) T « » © i o © 0- 2 O 0 : 3 O © 0 4 © © o: 5 0 © quite a competition! selektor topamp Hybrid audio power amplifier modules are not particularly new: over two-and-a-half years ago, in January 1977, we dis- cussed the subject quite extensively. What is new is the rapid advance in technology that has led to very high quality modules. The Philips types OM931 and OM961, for example, that will provide 30 or 60 very 'clean' watts, respectively, into 4 or 8 ohms. flash sequencer electronics the easiest way ik. wiisoni It is well known that children can generate some of the most original ideas on any subject. We can only envy the straight- forward simplicity of their answers to questions of a tech- nical nature, especially when the subject is electronics . . . remote control motor switch home trainer Budding physical fitness fanatics require an effective training program, but they must avoid overstraining their as yet untrained corpus. The circuit described here is a useful aid: it gives an indication of the amount of effort that can safely be exerted. fuel economiser (W.H.M. van Dreumel) While most things are 'going decimal’ it would appear that motoring costs are 'going logarithmic' and any method of saving money on the road must be greeted with enthusiasm. This particular idea is aimed at reducing the cost of acceler- ation (accelerating costs?). I played TV games (2) Last month, we examined the basic principles of the TV games computer, and discussed the more important instruc- tions. In this second article we will take a closer look at the rest of the instruction set, explain some useful programming tricks and list some useful routines available in the existing 'monitor' software. short-wave converter ionosphere There are so many 'whys' associated with shortwave recep- tion that many of us are completely in the dark about what frequency to choose, what time to listen, and what is likely to be heard. This article about the ionosphere is intended to take some of the guesswork out of shortwave listening. 11-01 11-06 11-10 11-14 11-16 11-18 11-20 11-22 11-24 11-34 11-36 low voltage dimmer 11-39 Take a hefty heatsink, a handful of passive com- ponents, a p.c. board and a brand-new audio power module, and what do you get? A very good power amplifier, with very little distortion, very little noise, very little fuss, and absolutely no calibration or adjustment! I see your point! 11-39 servo-controlled motor 11-40 market 11-41 ►l advertisement elektor november 1979 — UK 15 This kit for new subscribers (offer valid until December 31st 1979) A complete kit (including loudspeaker) for the simple sound effects generator will be sent free to all new subscribers. This offer is valid for all applications postmarked up to and including December 3 1 st 1979. As the name suggests, the simple sound effects generator will produce a range of sounds from that of an American police siren to one closely resembling the 'twittering' of birds. You can become a subscriber by filling out the order card included in this issue and including the text 'subscription and free kit'. The kits will be sent out at the end of January. — elektor — up-to-date electronics for lab and leisure Elektor House, 10 Longport, Canterbury, Kent. Tel. (0227) 54430. UK 16 — elektor november 1979 advertisement kunt u dit lezen ? NEEM DAN KONTAKT OP ! misschien bent U diegene die wij zoeken voor onze redaktie in engeland. If you think you can write in our style and are able to translate from Dutch to English fluently, then contact Mr. R.E. Day, Elektor Publishers Ltd., 10, Longport, Canterbury, Kent, CT1 1PE. The professional scopes you’ve always needed* When it comes to oscilloscopes, you'll have to go a long way to equal the reliability and performance of Calscope Calscope set new standards in their products, as you'll discover when you compare specification and price against the competition. The Calscope Super 1 0, dual trace 1 0 MHz has probably the highest standard anywhere for a low cost general purpose oscilloscope. A 3% accuracy is obtained by the use of stabilised power supplies which cope with mains fluctuations. The price £21 9 plus £17.52 VAT. The Super 6 is a portable 6MHz single beam model with easy to use controls and has a time base range of Ips to 1 0Oms/cm with 1 0mV sensitivity. Price £162 plus £12.96 VAT. Prices correct at time of gomg to press CALSCOPE DISTRIBUTED BY Maplin Electronics Supplies Ltd. P.O. Box 3 Rayleigh, Essex. Tel: 0702 715 155 Mail Order Audio Electronics, 301 Edgware Road, London W.2. Tel: 01 724 3564 Access and Barclay card facilities (Personal Shoppers) Watford Electronics, 33-35 Cardiff Road, Watford. Herts. Tel: 0923 40588 CALSCOPE quite a competition elektor november 1979 — 1 1*01 « Al " I In the recent Summer Circuits edition, we presented 106 circuits — selected from over 3000 entries to our inter- national competition. £ 10.000 worth of prizes were available for the best 20 of these, as selected by you, our readers. Well, you certainly voted — with a vengeance! In all 4414 voting cards were returned, most of which listed 10 cir- j cuits. Over 40,000 votes in all! We soon realised that if we tried count- ing them by hand, the results would not be available until the middle of next year. Fortunately, we have access to a computer. Even at that, it took a total of about 40 hours of computer time! Not to prolong the suspense further than necessary: the points scored by the various circuits are listed in Table 1 . It's not so easy to spot the winner in this list, so: print-out number 2 in Table 2 gives the final positions. Congratulations, Mr. 106! Before decoding these anonymous num- bers, it is perhaps interesting to compare the slightly different voting results obtained from the cards sent in from our four different language editions. Tables 3... 6 give the results from England, France, Germany and Holland, respectively. The higher totals in the German list reflect the fact that more voting cards were returned from there. However, the general tendency in all four lists is surprisingly similar — with one or two notable exceptions: what happened to circuit no. 9 in the Dutch list? Have they forgotten what the sun looks like?! For that matter, we had quite a bit of fun watching the intermediate results, as they became available. As the long, wet summer progressed the 'solar tracker' moved slowly up in the list and the 'moisture sensor' moved down . . . hard luck, sir! And now: the prizes! These are the 20 winners, and their prizes: 11-06 — elektor november 1979 selektor The atmosphere of Venus Measurements from the first spacecraft to orbit Venus are proving, upon analysis, to be an important step in understanding the evolution and thermo- dynamics of the planet's deep atmos- phere. Success in achieving that aim would reflect into work on the diffi- cult problems of the circulation that affects our own weather. In May 1978 an Atlas Centaur rocket blasted off from Cape Canaveral in Florida carrying a Pioneer spacecraft which, after a journey lasting over six months, became the first spacecraft to be put into orbit around the planet Venus; the manoeuvre was successfully completed on 4th December 1978. A few days later four probes launched from a second Pioneer entered Venus's atmosphere at different points. The orbiter and the probes carried a variety of experiments to study the structure and composition of the planet's atmos- phere. One of the experiments on the orbiter is a radiometer measuring infra-red radiation emitted by the atmosphere and clouds, similar to radiometers on board satellites orbiting the Earth and observing our own weather. From observations with this instrument, the temperature of different layers of the atmosphere and of the clouds below the spacecraft can be inferred. The instru- ment has been built jointly by the Department of Atmospheric Physics of the University of Oxford and the Jet Propulsion Laboratory, Pasadena, California. It is in fact the first British- built experiment to travel to one of the planets. Venus has, of course, been visited before by spacecraft. Three Mariner spacecraft from the USA have passed by the planet and nine Venera probes from the USSR have entered its atmos- phere. But the Pioneer 12 orbiter of last year is the first spacecraft to orbit the planet with the aim of observing the day-to-day changes in Venus's 'weather'. Already Known Why is Venus such an interesting planet to Earth-bound meteorologists? To show that, first a description of what was known about Venus and its atmosphere before Pioneer 12 and a brief look at some of the results from the latest mission: Venus is the next planet to the Earth and is somewhat nearer to the Sun. It is about the same size as the Earth and rotates much more slowly; a solar day on Venus is 117 Earth days. The Venus 1 Venus at Orbiter launch Orbiter one month probe launch May 1 978 after arrival Figure 1. Trajectories of the Pioneer orbiter and probes. atmosphere is very deep — equivalent to about 100 Earth atmospheres — and the cloud cover appears virtually complete, so that at visible or infra-red wavelengths no part of the surface can be seen from outside. The first indication of a very high sur- face temperature on Venus, of about 450°C, came from ground-based measurements of the brightness of the planet in the microwave part of the spectrum, made about 1960. Confirma- tion was provided by the microwave radiometer carried aboard Mariner 2 in 1962. By contrast, measurements of the brightness temperature in the infra-red part of the spectrum indicate a tempe- ture of about — 40°C, a value represen- tative of the temperature at the top ot the visible clouds, which emit strongly in the infra-red region. Temperature Profile So, from infra-red and microwave measurements, we can begin to construct a profile of the variation of temperature with height for the Venusian atmos- phere, such as that shown in the dia- gram below. Further evidence for the 2 Figure 2. The atmosphere of Venus, indicating the various regions and the temperature profile. selektor elektor november 1979 — 1 1-07 accuracy of this profile has come from four Venera probes which passed close to Venus during the period 1969-72. Although theoretical calculation does not bear it out entirely satisfactorily, there is general agreement that the high temperature at the surface of Venus arises from the so-called 'greenhouse' effect. Venus's atmosphere and cloud cover together behave in a similar way to the glass in a greenhouse, in that they allow a certain amount of solar radiation to pass but are a very effective blanket to infra-red radiation leaving the planet's surface. This blanketing means that only a small amount of solar radiation needs to get through to the surface to cause quite a high tempera- ture. Some of the opacity in the infra-red region is due to absorption by the clouds and some to the fact that the absorption by water vapour and carbon dioxide under the high pressure of the lower Venusian atmosphere is much greater. A crucial test of the greenhouse hypothesis is to measure the proportion of the solar radiation reaching the surface of Venus in relation to the total amount arriving at its outer atmosphere. Such measurements were first made from the Venus 8 probe in 1972; they suggested that perhaps only one-quarter per cent of the total solar flux falling on the planet penetrated to the surface. Rather better measurements from one of the Pioneer 12 probes last year gave the higher figure of about two per cent, which seems good enough to confirm that the greenhouse mechanism is effective. Figure 3. Scan of Venus from the Pioneer orbiter's infra-red radiometer. forward convincing arguments, taking this and other evidence into account, that the cloud particles are solutions of sulphuric acid - a 75 per cent solution at the top of the clouds, at an altitude of 60 km, and a solution of about 98 per cent at the bottom of the clouds. Vital Clue Spectroscopic measurements from groundbased telescopes have shown that the dominant constituent of the planet's atmosphere is carbon dioxide and that there is little oxygen or water vapour. This seeming lack of water vapour led to a great deal of speculation about the composition of the clouds. A variety of sulphur compounds and mercury compounds, some of them quite exotic, were advanced as contenders. The vital clue came from very careful measure- ments of the polarization of reflected sunlight from the planet by two French astronomers, Coffeen and Gehrels, in 1969. These were interpreted by two Americans, Hansen and Arking, as consistent with reflection from a cloud of spherical particles of about 1 pm radius with the rather precise refractive index of 1.45 ±0.02. In 1973. A T. Young from the USA put RADIOMETER SCAN Figure 4. Brightness temperature of cloud tops in the infra-red near 11 pm wavelength, measured across the part of the planet shown in the previous diagram. 11-08 — elektor november 1979 selektor Figure 5. Scan from equator to equator through the North pole, showing the longitudinal dependence of temperature observed in three of the channels of the infra-red radiometer. (The numbers against the traces show the wavelength of each channel and the altitude represented by the information). Young's hypothesis that the main con- stituent of the clouds is sulphuric acid has been substantially confirmed by direct measurements from Pioneer 12 probes, though much larger particles, thought to be sulphur, were found in addition to sulphuric-acid droplets. I > 185 K < 170 K Figuur 7. Temperature at an altitude of about 100 km, measured by Pioneer infra-red radio meter. The day side is significantly warmer (by 10 K or so) than the night side. Rotation Many observers examining photographs of Venus taken from telescopes in the ultraviolet part of the spectrum have noticed features that change with time. Particularly interesting are some which seem to have a marked tendency to recur at intervals of about four days. Evidence of rapid rotation of the upper atmosphere also comes from measure- ments of the difference in Doppler shift in spectral lines between opposite edges of the planet; they show velocities of about 100 ms" 1 , consistent with four-day rotation. Further evidence comes from the very beautiful photographs taken from Mariner 10 which passed by Venus in 1 973 en route to Mercury. Photo A is a similar photograph taken in the ultra- violet range from Pioneer. Cloud struc- ture suggesting intense zonal circulation appears to be present in both hemi- spheres. The rapidly-moving features are visible only in ultra-violet photographs, so it is supposed that they belong to a variable, thin cloud layer at a consider- ably higher level than the main cloud deck, that is, at about 90 km altitude. Because the solid surface of Venus rotates so slowly, as already described, this evidence of rapid rotation of the upper atmosphere came as something of a surprise. In 1969, Schubert and Whitehead from the USA put forward the theory that the motion was caused by a travelling thermal wave induced by the motion of the Sun relative to the atmosphere. To prove their point, they carried out an experiment with a slowly-moving heat source under an annulus of mercury, and showed that LATITUDE Figuur 6. Simple thermally-d riven circula- tion in Venus's atmosphere. The upper levels of cloud near the equator are heated through absorbing solar radiation. Air above the heated area rises, cooling as it does so. A compensatory sinking, with associated warming, occurs in a smaller region near the pole. the mercury in the annulus developed a velocity in the opposite direction to that of the source and of about four times its magnitude. Comparing the dynamical properties of the upper Venusian atmosphere with the mercury in the laboratory annulus led them to argue that the ratio of atmospheric velocity to the apparent velocity of the Sun, relative to Venus's atmosphere, would be much greater than the factor of four found in the laboratory experi- ment. Infra-Red Further clues regarding the circulation are beginning to come in from the infra-red radiometer experiment on the Pioneer 12 orbiter. The radiometer scans across the planet in the way shown in figure 3. Figure 4 shows the effective temperature at the top of the clouds measured along such a scan, as reported by F.W. Taylor and his co- workers at the Jet Propulsion Labora- 1 10 100 1000 WATER-VAPOUR PRESSURE (mbar) Figure 8. Runaway greenhouse effect. selektor elektor november 1979 — 11-09 'runaway greenhouse effect', and may be described with the aid of figure 8, which compares the atmospheres of Mars, Earth and Venus. Suppose the atmospheres began to form by gas escaping from the interiors at a time when the surface temperatures were determined by the balance between solar radiation being absorbed and long wave radiation being emitted, at values given by those on the left- hand side of the diagram. Water vapour and carbon dioxide accumulating in the atmosphere, through the blanketing of the greenhouse effect, cause the surface temperature to rise; eventually clouds may form, intensifying the greenhouse effect and, thereby, raising the surface temperature still further, until in the end some balance is reached. For Mars, the atmosphere is so thin that no significant cloud has formed and the blanketing effect of the atmosphere is small. On Earth, in the equilibrium state, most of the water is in liquid form, while for Venus, on these assumptions, the surface temperature has always been above the boiling point of water at the surface pressure, so we would not expect to find any liquid water. If water has been present, in a similar amount to that on Earth, it would have been the main constituent of the early Venusian atmosphere. No other gases would have been present to prevent ultra-violet solar radiation dissociating water vapour at the top of the atmos- phere, so the hydrogen thereby pro- duced would escape and the oxygen would be consumed in various oxi- dation processes at the surface. The large amount of carbon dioxide remaining in the atmosphere, too, instead of in carbonates in the rocks, is consistent with this atmospheric history. Enough has been learned about the atmosphere of Venus to show that its evolutionary composition and physical structure pose very interesting problems. Our aim is to be able to model and to understand how the transport of heat, momentum and minor constituents is organized within Venus's atmosphere and how the atmosphere has evolved to its present state. Analysis of observations from Pioneer is already proving to be a big step towards further understanding. If we can solve these problems about Venus's atmosphere, which is so differ- ent from that of the Earth, one im- portant outcome will be that we shall tackle the difficult problems of the circulation of our own atmosphere a great deal more confidently. (spectrum no. 163) (500 S) tory. The most interesting feature is the very warm part of the cloud tops at about 79 degrees North, which is interpreted as a substantial clearing in the clouds enabling the radiometer to view much deeper and warmer levels of the atmosphere. Average tempera- tures at various latitudes for other levels, viewed by other channels of the infra-red radiometer, are shown in figure 5. They all illustrate the rather interesting fact that the polar regions are warmer than the equator at these levels. T aken together, the measurements temperature contrast between the day side and the night side of the planet; in fact, no difference has yet been identified in the infra-red data for levels below about 90 km. Some indica- tions of variations with longitude are beginning to emerge from the infra-red data, and they can possibly be associ- ated with a rapid zonal circulation. For higher levels, interpretation of the infra-red radiometer measurements by the Oxford group shows significant differences between the day and night side temperatures, as shown in figure 7, indicating that thecirculation in Venus's ionosphere is of a different character than that lower down. The Greenhouse Effect Returning to the lower part of Venus's atmosphere, below the clouds, measure- ments from the pioneer 12 probes have point to a circulation gently rising over a large part of the equatorial and mid- latitude regions and sinking in a smaller region near the pole (figure 6); the sinking air warms as it subsides, and it clears the clouds away. Superimposed on this overturning circu- lation should be the rapid zonal motion, which ensures that there is very little confirmed that the atmosphere is 95 per cent carbon dioxide, the other five per cent being nearly all nitrogen. As already stated, water vapour is notice- ably absent, compared with the amount in the Earth's atmosphere. An interesting explanation of this was put forward in 1969 and 1970 by Ingersoll, Rasool and De Burgh, in the USA. It is called the 11-10 — elektor november 1979 topamp It is surprising, in a way, that the adver- tising boys haven't yet come up with some phrase like 'Compact Power ®' to describe the output stage in amplifiers in the low and medium price brackets. After all, quite a few of these are equipped with a hybrid power amplifier module by now. Usually, these modules belong to the generation described in the article '1C Audio', referred to above. But now, there is something new and better, as we shall see. A better circuit and improved thermal stability have proved possible. The next generation When developing the OM931 and OM961, every effort was made to achieve low distortion and good thermal stability. Each module contains two Darlington output transistors and a ceramic sub- strate on which all other internal com- ponents are mounted. The (internal) circuit diagram is shown in figure 1; when discussing this, it will sometimes be a help to refer to figure 2: a com- plete circuit, including all external components. The input stage is a PNP differential amplifier, T1 and T2, with a current- source (T3) in the 'tail'. Resistor R3 is virtually equal to R1 + R2; this means that the dissipation in T2 is almost identical to that in T1, so that the long- tail pair is in thermal balance. This, in turn, means that the DC offset at the output is kept to a minimum. The output signal from T1 (across R2) is passed through a buffer stage. T4, to the driver (T5). Capacitor Cl provides frequency compensation; however, the value is smaller than usual since a rather uncommon frequency compensation circuit is used (as can be seen in fig- ure 2). Components T6, PI, R11 and R12 set the bias current for the output stage; the latter consists of two Darlingtons, T9+T10 and T11+T12. R12 is included to counteract the effect of supply volt- age variations on the bias setting. In the complete circuit (figure 2), an elec- trolytic is connected between the output Hybrid Hifi Hybrid audio power amplifier modules are not particularly new: over two-and-a-half years ago, in January 1977, we discussed the subject quite extensively. What is new is the rapid advance in technology that has led to very high quality modules. The Philips types OM931 and OM961 , for example, that will provide 30 or 60 very 'clean' watts, respectively, into 4 or 8 ohms. In this article, we will take a closer look at these 'lightweights with a heavy punch'. topamp elektor november 1 979 — 11-11 (btstrj Figure 1. The internal circuit of the power amplifier modules OM931 and OM961. Main specifications of the OM931 and OM961 OM931 OM961 supply voltage i 23 V ± 26 V t 31 V i 35 V quiescent current 80 mA 80 mA 100mA 100 mA output power, 4 ) 30 W — 60 W — output power, 80^^ — 30 W — 60 W clipping level at 1 kHz, 4 ft. d * 0.7% 40 W 40 W 75 W 75 W THD at 1 kHz, 1 W 0.02% 0.02% 0.02% 0,02% input sensitivity 0 7 v RMS tovrms IOVrms 14 V RMS input impedance 10k open-loop gain 80 dB 110,000x1 closed-loop again 24 dB 115.7 x) feedback factor 56 dB 1630 x) frequency response at 10 dB below maximum output power 30 . . . 40.000 Hz -1 dB power bandwidth (—3 dB, d = 1%) 20 . . . 40,000 Hz signal-to-noise ratio at 50 mW output power 75 dB signal-to-noise ratio at maximum output power > 102 dB output DC offset voltage ± 20 mV supply ripple rejection > 65 dB output impedance 50 m absolute maximum supply voltage, OM931 ± 40 V OM961 ± 45 V maximum case temperature 95° C Note (1 ): for THD < 0.2% at all frequencies from 20 Hz to 20 kHz (FTC specification). (pins 3 and 4) and pin 8. This provides 'bootstrapping', with the result that the collector impedance 'seen' by T5 is much greater than R13+R14+R15, so that a high open-loop gain is obtained. Bootstrapping can cause trouble, unless due care is taken in the design. If D2 is omitted, things could go wrong . . . The voltage on pin 8 can easily rise above the positive supply voltage. Without D2, the voltage at the base of T9 cannot rise more than about 0.5 V above posi- tive supply — T9 will be in saturation by then. If we consider the fact that 'clipping' in the output stage is rather nasty (the relatively long recovery time leads to audible distortion) it is obvious that it is a good idea to 'clip' at an earlier point in the circuit — even if this means sacrificing a few hundred milli- volts of output swing under full drive — provided a much shorter recovery time can be achieved in this way. The boots- trap circuit provides an ideal solution: split the series resistor into two parts (R13 and R14) and connect the junc- tion to positive supply through a diode (D2). The values of the two resistors are chosen such that D2 starts to conduct at a signal level just below that required to drive T9+T10 into saturation. With D2 conducting, the bootstrap mechanism is effectively put out of action; the remaining collector impedance for T5 is the relatively low value of R13, and so the open-loop gain collapses. This makes for a vastly improved recovery charac- teristic after the amplifier has been driven into clipping. Those readers who would like to know more about this are referred to the literature listed at the end of this article. What remains, by and large, is protec- tion circuitry: T7, T8, R16...R24, C3, C4, D3 . . . D8. When this circuit starts to operate, T7 or T8 turns on, preventing further drive to T9+T10 or T11+T12, respectively. The base- emitter voltages of T7 and T8 depend on both the output voltage and the out- put current. Diodes D7 and D8 are included to protect the output devices against excessive voltage spikes, such as could occur if the protection circuit came into operation while driving a heavily inductive load. Now for the external components Figure 2 is the circuit for a power ampli- fier using the OM931 or OM961 , as pro- posed by Philips in an application note. A symmetrical power supply is used, so that the loudspeaker can be DC-coupled — no output electrolytic is required. C5 is the bootstrap elco. C7 and R8 provide a well-defined load at high fre- quencies, to maintain unconditional stability. LI and R7 drastically reduce the effect of a capacitive load — this might otherwise result in 'ringing'. Negative feedback from the output to the inverting input is provided by R4, R5, C3 and C4; C4, in combination with R4 and R5, provided so-called 'lead' topamp 11-12 — elektor november 1979 I 2 23- 35V* Figure 2. A complete power amplifier using either the OM931 or OM961 . This circuit is a Philips design. Parts list Resistors: R1.R5 = 10 k R2 = 4k7 R3 = 330 n R4 - 680 tl R6 = 22 n R7 = 2T22/1 W R8 = lOn/’/r W Capacitors: Cl * 1 p/63 V C2 = 270 p C3 - 47 m/10 V C4 = 1 20 P C5 = 100 m/40 V C6 = 470 m/40 V C7 = 100 n C8.C9 = 10 m/63 V Semiconductors: IC1 = OM931 or OM961 Miscellaneous: Heatsink, 0.8°C/W (OM961) or 1.4° C/W (OM931) LI ■ 4 ... 6 mH; 40 turns on R7, CuEm, 0.6 mm p compensation. Another useful pre- caution. At audio frequencies, the closed-loop gain is determined by R4 and R5. To be more precise, the gain is The components R2, R3 and C2 deserve special mention. In combination with R1 (in parallel with the source im- pedance provided by the preamplifier), these components ensure that the open- loop gain rolls off above a certain fre- quency. Something of this kind is necessary to keep any amplifier with feedback stable; by placing these components in front of the amplifier (effectively outside the feedback loop), there is no danger of overload inside the loop. TIM (Transient Intermodulation Distortion) is avoided in this way. The Table summarises the main speci- fications of the two amplifiers built according to the circuit given in figure 2, and using the OM931 or OM961. The figures given speak for themselves . . . Let's get cracking A printed circuit board design is given in figure 3. This board is suitable for a single (i.e. mono) power amplifier; for stereo, two p.c. boards are required. The mechanical details of the amplifier modules themselves are given in fig- ure 4. When mounting the OM931 or OM961 module on the board and on the heat- sink, the module should be mounted about half an inch off the board; the edge of the board will be almost flush against the heatsink. For a stereo ver- sion, the two modules can be mounted on a common heatsink, provided the latter has a sufficiently low thermal ] resistance. The symmetrical supply voltages can be read off from the table. Note that, when using an unstabilised supply of the type shown in figure 5, the supply voltages given should be available at full drive. 3 Figure 3. Printed circuit board for a single (mono) power amplifier according to the circuit given in figure 2. topamp elektor november 1979 — 1 1-13 4 L 8 31 f j t 5.75 - - - 0.5 80023 4 Figure 4. Mechanical dimensions of the OM931 and OM961. 5 Figure 5. A symmetrical, unstabilised supply for the power amplifier(s). The current rating of the transformer and rectifier diodes depends on the number of modules to be connected, the maximum output power and the loudspeaker impedance. Details are given in the text. Under no-drive conditions, higher volt- ages will be found; however, the maxi- mum ratings (+/— 40 V for the OM931 and +/— 45 V for the OM961) should never be exceeded — and a safety mar- gin should be allowed for the mains voltage rising to 10% above its nominal value. The current rating for the transformer and rectifier in figure 5 will depend on the output power required, the load impedance, and the number of modules running on the same supply. Per mod- ule, the current consumption is as follows: OM931 , 30 W into 4 12: 1 .25 A OM931 , 30 W into 8 El: 0.9 A OM961 , 60 W into 4 El: 1.75 A OM961, 60 W into 8 El: 1.25 A For a stereo amplifier, obviously, the total current consumption will be twice that given above. Some care should be taken when wiring up the amplifier(s). Bad wiring can ruin the performance of even the best ampli- fier design; it can even lead to a con- siderably higher distortion percentage! This is not so surprising when one con- siders that the heavy current flowing in the posititve supply lead during full drive only flows during the positive half of the output signal swing — effectively, it is half-wave rectified. The same applies in the negative supply lead. This means that there are an awful lot of higher harmonics floating around! And it only takes a little bit of stray capaci- tance or inductance for them to find their way back to the input of the amplifier . . . Keep the supply wiring short and direct, therefore, and as far as possible away from the input wiring. Heavy gauge wire is also a good investment — it keeps the resistance down. The return lead from the loudspeaker should be connected direct to the supply electrolytics, not to supply common on the p.c. board. In a stereo amplifier, don't give in to the temptation to use the same lead for two jobs: separate wires should be used for all supply lines, loudspeaker returns, etc. Screened cable should be used for the input wiring. If the case is to be con- nected to supply common, this should be done at the input, not at the supply. All this may seem rather overdone. But it would be a pity to buy good amplifier modules and then ruin their perform- ance by skimping in the final construc- tion! Finally, all due care should be taken with LI, R7 and the connections to these components. Virtually the whole of the output signal current runs through LI, and a bad joint would ruin the out- put damping factor. M Literature: 1. 1C Audio, Elektor January 1977 2. Negative feedback - how thick to lay it on, Elektor March 1977 3. Equip (1), Elektor April 1976 11*14 — elektor november 1 979 flash sequencer r as l 1 as i fas i r as i f] lash sequencer sequencer sequencer sequencer sequencer To analyse the motion of fast moving objects it is useful to be able to take a well-timed succession of photographs of the object under investigation. Ideally these photographs should be superimposed on the same frame (multi-exposure technique) so that the relationship between the various positions of the object can be examined in detail. However, the cost of the camera needed for this type of photography is out of the reach of most amateurs (and many professionals). An alternative method is to leave the camera shutter open and produce a series of light flashes for the successive exposures, thereby producing similar results. This article describes an electronic flash sequencer which generates a series of five flashes. It is intended for photographic enthusiasts who require something extra from their existing equipment. Admittedly, professional flash sequencers are commercially available but at fairly high prices and financially, therefore, not the first choice of many amateurs. □ as i A as i as i ^ft“ as i IIS 1 ft ft* IIS 1 ■ ft Sift* as i rf# IIS 1 2 ft as i ft ^fft' as i il ash sequencer sequencer sequencer sequencer sequencer sequencer sequencer sequencer sequencer sequencer sequencer This design uses five flash units. These are fired in succession with intervals adjustable between 10 ms and several seconds. The shortest time is dependent on the flash duration and this in turn determines the resolution of the move- ment being analysed. In the interests of economy, especially with regard to the total number of exposures expected, it may be considered impractical to use other than elec- tronic flash guns (unless, of course, you grow your own bulbs). Circuit description. The circuit of the sequencer consists basically of a four stage ripple counter as shown in figure 1 . The camera contact, via inverter II, fires the first thyristor Thl and, at the same time, triggers MMV1. At the end of its pulse duration, the negative going edge at the output of MMV1 triggers MMV2 and fires thyristor Th2. And so on until Th5 has fired. It will be obvious that it is possible to continue the chain for any number of stages — and there- fore flashes. The intervals between flashes are set by potentiometers R6 . . . R9. Each thyristor is automatically turned off after firing when the capacitor in the flash unit becomes discharged, causing the thyristor hold current to collapse to below its critical value. To test the firing sequence without a camera, switch SI is placed in the 'test' position and switch S2 is used to simulate the camera contacts. Any contact bounce in the camera or S2 is eliminated by the circuit itself: MMVI will not retrigger and the flash guns require a longer time to reset. The sequencer can also be put through its paces without flash units being connected if desired. Light emitting diodes can be used as shown in figure 2a. The thyristor rapidly charges the capacitor through the LED causing it to flash. Once the capacitor is charged, the thyristor turns off and the capacitor then discharges through the resistor across it. A 12 volt 100 milliamp power supply can be used and a circuit for this is shown in figure 2b. As an alternative the sequencer can be powered by eight size AA or C dry cells. Construction The construction of the sequencer should not present any problems, all components being readily available. For the thyristors any 5 amp 400 volt type will perform satisfactorily. Plugs and sockets for connecting the flash units are available at photographic shops. The controls for exposure intervals can be realised in different ways, 500 k or 1 Mohm potentio- meters can be used but switched resis- tors offer some marked advantages . . . A discrete step control permits repeatedly exact settings once the ~ost effective interval has been estab- >~ed Individual resistance values can be determined by rule of thumb: ' * for every incremental 5 ms interval. With the 500 k or 1 Mohm suggested for R6 . . . R9 this amounts to a maximum interval of 2.5 or 5.0 seconds. It should be noted that the above can only be an approximation since the combined component tolerances can result in an error of up to 50% either way. If greater accuracy is required, one of the following methods can be used. If ordinary potentiometers have been used these could be calibrated by hand. With switched resistor banks, each step can be trimmed with the aid of a variable resistor to be sub- stituted by a fixed resistor once a value has been arrived at. Figure 3 shows a possible front panel layout giving an indication of the size (which can be an important parameter for the photographer). M 1 1-16 — elektor november 1979 electronics the easiest way Some of the most delightful obser- vations about electronic communi- cations have been boldy put to paper by primary school miniprofessors. Take these historical explanations for example. Question: "When was the radio inven- ted?' Answer: 'On page 24.' 'The radio was invented in the pre-me times.' The Romans did not have radios. They used smoke signals in both the A.C. and D.C. times.' Children have a knack for discarding everything but what they consider to be the most essential information. One boy brusquely wrapped up all of man's yearn- ings, struggles and triumphs in this eight word package: 'Progress was from electricity to radios to now.' Here's a remark as charming as child- hood itself: 'I was thinking the radio electronics die easiest «ji> It is well known that children can generate some of the most original ideas on any subject. We can only envy the straightforward simplicity of their answers to questions of a technical nature, especially when the subject is electronics . . . was invented before the telegraph. When I learned different, all the thoughts I was going to say went in a swallow down my throat.' Another tiny historian concluded: The Dark Ages lasted until the invention of electricity.' Through the years, the youngest generations' fund of knowledge has proved to be a glittering gold mine of wit and unconscious wisdom, often conveniently unhampered by hard facts. Each new subject seems to be a fertile new field for off-centred interpretation and lopsided logic. Digging into facts about Marconi produced such notable nuggets as these: 'Marconi was born in 1874,supposably on his birthday.' 'It took much hard work for Marconi to think out how to invent the radio. He had to keep thinking around the clock, twelve days a week.' 'In just a few short years he became a sensation overnight.' 'He expired in 1937 and later died from this.' Recently a bright-eyed little radio enthusiast came up with this endorse- ment: 'Every time I think how the radio gives us so much fun, I have joy feels all over.' A skeptical classmate of hers absorbed all the statistics regarding the number of ham radio operators, but got his skepticism across in one crushing state- ment: The total amount of ham oper- ators today is more for saying than believing.' It must run in the familiy. Two years later his younger sister reported: The number of ham operators we have today is an adsurbly large fact of a number.' The subject of hams has stumped many eager young scholars. Here are three more futile but imaginative explan- ations: 'Ham operators look something like people.' 'They are one of the chief by-products of electricity.' 'The meaning of them has a very short memory in my mind.' The elementary school youngster's mind seems to be a vast storehouse of miscellaneous misinformation — half true, half false and wholly delightful. His fund of knowledge about electricity includes such fascinating items as these: 'Electricity has been with us forever and maybe even longer.' 'Would the average person be able to keep up with the news if it was not for electricity? The chances are 999 out of a hundred.' 'In electricity, opposites attract and vice versa.' 'If you see lightning, no you don't. You see electricity.' 'From now on, I will put both gladness and wonder in my same thought about electricity.' Here's one I’ve been trying to figure out for five years: "You should always capitalize the word electricity unless it is not the first word in the sentence.' This next little girl seemed to be giving it all she had when she wrote: 'Correct my being wrung, but tell me true or false. Do negative charges go through electrons or through protons? I wrecked my brain trying to think which. ' But I'm afraid others are more non- chalant in their pursuit of knowledge: 'Protons are bigger than electrons in case I ever want to know.' Psychologists tell us that half learning a fact incorrectly is often the first step to learning it right. So let's be philo- sophical as we buzz through these fractured facts about electrons and protons: '100 electrons equal 1 radio program.' 'When the switch is on, electrons are constantly bumping into each other inside the wire. There is really quite an overpopulation of electrons.' 'Once I saw in an educational cartoon about how electrons move. Electrons are very interesting folks. All their ways are hurry ways.' 'Electrons carry the negative charge while protons take the affirmative.' 'Electrons are the same as protons only just the opposite.' 'I think I admire the electron more than anything else about electricity because it weighs only about one over 2000th as much as a proton but can still hold its own.' When questioned, children offer the ever present possibility that however far from right their answers may be, the next wrong answer could be more witty and thought-provoking than the correct one. Sometimes they don't know and electronics the easiest way elektor november 1979 — 11-17 they know they don't know, but that doesn't keep their answers from being charming: 'Ideas about how radios work have advanced to the point where they are no longer understandable.' 'Did I pass the test about how to get a ham radio operator's license and why not?' 'I have found radios to be easier to listen to than to tell how they work.' Take three small boys, mix them up thoroughly with several pounds of strange facts, then shake up with an examination and you have the perfect formula for instant confusion. 'The way vacuum tubes work, as I understand it, is not very well under- stood.' 'Many questions have been aroused in my mind about vacuum tubes. As a mattery fact, the main trouble with vacuum tubes is that they give more questions than answers.' 'In electricity, positives are attracted by negatives for the reason of search me.' Often a grownup can only envy the simplicity of a child's way of expression, as is the case of the lass who remarked: 'When I learned we were going to see a movie about ham operators all over the world, I told my feet to quiet down but they felt too Saturday to listen.' In their world of uncertainty, once they know a fact for certain, they hang on to it tenaciously, e.g.: 'Another name for the radio is radiotelephony, but I think I will just stick with the first name and learn it good.' Children, like mountain climbers, must always make sure that their grasp on a fact is firm, even though they want to leap far beyond. Otherwise, they may find themselves trapped on a mental ledge. There is usually at least an element of truth in the most absurd answer. Sometimes they aren't wrong at all. It's just the way they put it that's so funny: 'Radio has a plural known as mass communication.' 'Water scientists have figured out how to change river currents into electric currents.' 'The best thing live wires are good for is running away from.' 'Quite a bit of the world's supply of electricity goes into the making of ham radios.' 'Many things about electronic com- munication that were once thought to be science fiction now actually are.' Members of the primary school set certainly have their own opinions, and few are hesitant to express them: 'All the stuff inside a ham radio is so twisted and complicated it is really not good for anything but being the stuff inside a ham radio.' 'Electronics is the study of how to get electricity without lightning.' How about this unforgetable remark: 'Last month I found out how a radio works by taking it apart. I both found out and got in trouble'. And you can’t argue with the young fellow who reported: 'When currents at 200 to 240 volts go through them radios start making sounds. So would anybody.' Just what is a vacuum? Here are five answers, fresh from the minds of nine- year-olds: 'Vacuums are made up mostly of nothings.' 'A vacuum is an empty place with nothing in it.' 'Vacuums are not anythings. We only mention them to let them know we know they're there.' 'There is no air in vacuums. That means there is nothing. Try to think of it. It is easier to think of anything than nothing.' 'A vacuum tube contains nothing. All of its parts are outside of itself.' Another lad wrote of this frustrating experience: 'I figured out how a vacuum tube works twice but I forgot it three times.' One of his classmates reported: 'When I learned how empty vacuum tubes are, I would have fainted if I knew how.' If you're at all hazy about other parts in a radio, hang on. These next thoughts will leave you only slightly worse off than before: 'An electron tube can be heated two different ways. Either Fahrenheit or Centipede.' 'When you turn a radio on, the tubes get hot. The hotter anything gets, the faster the molecules in it move. Like if a person sits on something hot, his molecules tell him to get up quick.' 'In finding out that radio tubes get hot, the fun is not in the fingers.' 'Transistors are what cause many radios to play. Transistors are a small but important occupation.' 'We now have radios that can run on either standard or daylight time.' One student had many tussles with his spelling book. When he finished writing one particular sentence, the battleground looked like this: 'ter- manuls do not agree with themselves spelingly and pruncingly.' With apologies to Mr. Webster, I would like to present a pocket-size dictionary of pint-size definitions, compiled from school children's reports. Should any of them prompt Webster to turn over in his grave, he would have to do so with a smile: 'Axually, a choke coil is not as danger- ous as its name sounds.' 'Electromagnets are what you get from mixing electricity and magnets together.' 'Think of a volt. Then yippee, because now you have had the same thought as Voltaire, after who this thought was named.' Another lad had the right information, but the wrong answer: There are some things about electricity we are still not sure of. These things are called whats.' If the kids don't know all the answers, they can always do what their parents once did — try to slide by on a guess or two: 'A radio telescope is a thing you can hear programs by looking through it.' ‘Current electricity is electricity that is currently in use.' Children are so full of questions, they can't possibly wait for someone to tell them all the answers. That's why they plunge recklessly ahead on their own, like so: 'Sound travels better in water than in air because in water the molecules are much closer apart.' 'I have noticed that if a portable radio is turned in different directions, the station talks loudest behind its back.' 'Although air is hollow it is not just for looking through. It is also for having radio waves running through it and trying to answer questions about.' 'Radio waves would not be all that important to study if it were not for ears.' 'Someone in here said that FM has shorter waves than shortwave radios. Is this so? I think it is because I think I was the one that said it.' (If you can't believe yourself these days, who can you believe?) An obviously more confident young man proclaimed' 'Much has been said about how radio waves travel. Radio waves are both hearable and talkable.' The last word must go to this moppet who was doing well — until the last word: 'I believe the radio is one of the most important inventions of all time. Of course my father works at a radio station, so I may be a little pregnant.' That's one young writer who would have done fine if she had just stopped while she was ahead (which is good advice for grownup writers, too). By kind permission of 73's magazine. M 11-18 — elektor november 1979 remote control motor switch Remote control systems for models use various ways of coding the control signals. One way is to use pulse-width modulation: pulses are sent with a repetition rate of 20 ms and a pulse length of 1.0... 2.0 ms, where the pulse length defines the command. The circuit described here belongs in this category. The position of a three- way switch is determined by the length of the pulse received. If the switch is used to control a motor, it can be arranged so that the motor turns one way if the pulse length is 1 .0 ... 1 ,25 ms; it is stopped for pulses between 1.25 and 1 .75 ms; finally, a pulse width from that the input pulse width is only 1.1 ms, it will have returned to logic '0' before FF1 is clocked. The Q output of FF1 will therefore become '0' and the 0 output will become '1' (C and D, respectively, in figure 2). T1 is turned on, the relay pulls in, and the motor will run, say, anti-clockwise. While all this is going on, MMV2 and FF2 are also doing their stuff — with only one or two minor differences. The period time of MMV2 is set (by R2 and C2) to 1 .75 ms — E in figure 2 — so it takes that much longer before FF2 is clocked and its Q output becomes logic 0. Note that if it was already at remote control motor snitch 1.75 to 2.0 ms causes it to turn in the opposite direction. The circuit is given in figure 1 , and figure 2 illustrates the pulses at various points. The incoming pulses are fed to the trigger inputs of two monostable multivibrators (or 'one-shots'), MM VI and MMV2, and to the 'data' inputs of two flip-flops (FF1 and FF2). Let us assume that a 1.1 ms pulse is received (A in figure 2). MMV1 is triggered, so that it produces a 1 .25 ms output pulse (as determined by R1 and Cl). The Q output from this one-shot (B in figure 2) is used to clock FF1. With this type of 'data flip-flop', the logic level at the 'data' input is trans- ferred to the Q output on the positive- going edge of the clock signal. As can be seen from figure 2, this corresponds to the end of the 1.25 ms period deter- mined by MMV1 . Since we are assuming logic 0, it stays that way! Since T2 is controlled by F2's Q output, this tran- sistor will be turned off. Relay 2 drops out (or stays out), which is exactly what we want. If the incoming pulses are made longer than 1 .25 ms, the data input of FF1 will be at logic 1 when it is clocked. Its Q output will become 'O', turning off T1 so that Rel drops out. Both poles of the motor are connected to positive supply. No power, so no rotation . . . Finally, if the incoming pulses are made still longer — more than 1 .75 ms — the Q output of FF2 will become logic 1. Now, at last, T2 is turned on; Re2 pulls in, and the motor starts to run in the opposite direction. So much for the main circuit. Only two points remain to be mentioned. The cross-connections between the Q output of FF1 and the R input of MMV2, and between the Q output of FF2 and the R input of MMV1, ensure that only one of the two relays can be pulled in at any time. Strictly speaking, this is an unnecessary refinement, but it only costs two bits of wire. Three Schmitt-trigger NAND gates, N1 . . . N3, perform a double function. When the low-voltage (4.8 V) supply for the electronics is first connected, C5 is not charged and so the output of N3 is at logic 1 . This sets FF1 and resets FF2, so that both transistors are turned off and the motor is stationary. Further- more, if the supply to the motor (8.4 V) drops below a level determined by PI, the flip-flop consisting of N1 and N2 changes state, again turning off the motor. This is done to prevent damage to the accumulator by excessive discharge. Since quad NAND gate packages contain four NANDs, there's one left. remote control motor switch elektor november 1 979 — 11-19 This can prove useful if the circuit is to be triggered by negative control pulses: N4 can be wired in series with the input, to invert them! It was mentioned in passing that a low- voltage supply is used for the electronics proper. Since only a few milliamps are required, this supply can be derived from the battery that powers the receiver. LI and C7 can be added to smooth the supply — although in oractice LI will often prove unnecessary, "he supply for the motor will normally be provided by a separate accumulator, "he voltage will, of course, depend on the motor used; the relays must also pull in reliably on the same voltage (and ~3ve sufficiently heavy-duty contacts!). ; a different supply voltage is used, the value of R5 (in series with PI) will have to be altered accordingly. The voltage at the wiper of PI should be set to approxi- mately 2,2 V with fully-charged bat- ter es. The motor will then be switched off when the voltage drops by about 10%. Another good way to adjust PI is to set it so that the motor is turned off automatically when it is held stationary under power. Note that this adjustment should be done very slowly, since C6 and R9 provide a considerable delay! Once the protection circuit has cut in, the only way to reset it is to disconnect the 4.8 V supply for a few seconds. The same applies when setting up the model: if the battery that powers the motor is installed last, the protection circuit will already have detected a 'low' battery. As before, the 4.8 V supply will have to be disconnected for a few seconds. If this is felt to be a nuisance, a reset push- button can be included in parallel with R8 and C5. One final note. If the motor is found to run the wrong way, the connections to the motor should be reversed — not those to the battery! Otherwise the protection circuit won't work ... H 11-20 — elektor november 1 979 home trainer Budding physical fitness fanatics require an effective training program, but they must avoid overstraining their as yet untrained corpus. The circuit described here is a useful aid: it gives an indication of the amount of effort that can safely be exerted in the course of the training course. It is only a coarse indication, of course, but adequate for normal use. All specialists agree on one point: regular training is the key, and only a limited amount of well-chosen exercises home are required. The home trainer described irsiiiKT cise and another half-minute break; and so on. At the outset, five one-minute work-outs are enough for one day. After about four weeks, an extra minute can be added; from then on, a further minute is added every two weeks until finally, after 12 weeks, a total of 10 minutes hard work (with five minutes relaxation) is permitted. It is sufficient — and therefore advisable — to go through this routine every other day, or three times a week. If only general fitness is desired, there is no need to extend the five minutes a day. Going up to ten minutes is only worth while for real enthusiasts. All exercises can be used that bring more than one-sixth of the main muscles into play: for example, push- ups, knee-bends, touching your toes, running, high jumps and so on. Obvi- ously, special training gear (home Even the ancient Romans knew that physical fitness is something to be desired: 'mens sana in corpore sano', as Juvenal declared. Nowadays, the number of training programs being put forward would seem to indicate that a large amount of mental effort is being dedicated to working out how other people should 'work out'. Maybe this isn't quite what Juvenal had in mind, but there is something to be said for 'sensible gymnastics' — if nothing else, at least a good training program can help to improve physical fitness with a minimum of effort. Apparently, in our highly efficient society, this is a desirable goal. timer for systematic PT here is based on a system evolved at Leeds University: so-called Circuit Training. This system has the advantage of combining two desirable goals: im- proving stamina and toning up the most important muscles. Several variations of the same basic system exist, and one of the most popular ones is the basis for this circuit. The idea is to work hard for one minute and then take a 30-second breather; then another minute of strenuous exer- trainers of one kind or another) can also be used. Equally obviously, it is a good idea to use various different types of exercise — one minute of each, say. During the one-minute exercise periods, you are supposed to really exert your- self. Keeping one eye on the clock is not easy under these circumstances. And this is where the 'Home Trainer' comes in. At the end of the first minute it gives a (welcome?) indication that it is time for a breather; after a further indication home trainer elektor november 1979 — 11-21 it recalls you to your duty, and so on. Two different frequencies are used, so that there is little danger of getting out of step. The tones last for about two seconds. In case of doubt, two LEDs clearly indicate what you are supposed to do: Green for Go and Red for Stop. Like traffic lights, only without the amber. The circuit From the description given above it is to be expected that thecircuit will be fairly simple. It is. A single 555 timer and a few standard TTL ICs do the whole job. The 555 timer gives the basic clock pulses, at one-second intervals. A counter, consisting of two 7490s, derives the 60-second and 30-second intervals from these clock pulses. One minute after the circuit is first switched on. the output of NAND gate N7 goes to logic 0. This triggers monoflop MMV1. During the two-second output period of this monoflop, a multivibrator (N5 and associated components) produces a 750 Hz 'take-a-breather' indication. At the same time, the logic 0 from N7 causes a flip-flop (N3 and N4) to flip — or should it be flop? — so that the green LED is turned off and the red LED is turned on. As mentioned earlier, red means Stop . . . After a further 30 seconds, the counter (IC2 and IC3) resets. The output of IM8 now becomes logic 0. This triggers monoflop MMV2 (IC5), turning on the 1500 Hz 'get moving' signal for two seconds and resetting the N3/N4 flip- flop (if it flipped before, it will now flop, or vice versa) so that the green LED lights. Two clear and unambiguous indications that it is time to get back on the job. The only calibration point in thecircuit is the 100 k preset potentiometer in the basic clock generator circuit. The cali- bration procedure is as easy as it is Table first through fourth week fifth and sixth week seventh and eighth week ninth and tenth week eleventh and twelfth week from thirteenth week 5 x 1 minute 6x1 minute 7 x 1 minute 8x1 minute 9x1 minute 10x1 minute Note: for enthusiasts only! The rest of us keep to 5 x 1 minute every other day. obvious: PI is adjusted until the one- minute exercise interval lasts for one minute. A few seconds over or under are unlikely to affect the effectiveness of the training program. A straightforward 5 V power supply, as shown in the circuit, is sufficient. The total current consumption is less than 150 mA, so a bell transformer will be more than adequate. Using the circuit is even easier than calibrating it. After switching on, SI is set to position 1 - 'Reset' - so that the counters are reset to zero. Having dressed suitably and moved the furniture out of the way, SI is set to position 2. The first strenuous minute starts: operation 'physical fitness by the clock' is under way. The merciless mechanical mentor will let you know when to let up and when to get going again. The only mental exercise required of the ardent (perspiring) student is to keep track of the number of rounds. As mentioned earlier, five one-minute sessions are enough for most people. Real enthusiasts can derive their training program from the accompanying table. One final word of advice: those who are in any doubt at all about their bodily health must consult a doctor before embarking on any strenuous exercise. M 1 1-22 - elektor fuel economiser celerometer. It is a fact that a 'smooth' driver rarely accelerates at more than 1 m/s 2 . How does your figure stand — are you a 5 m/s : driver? The accelerometer How do we measure acceleration in a practical sense? There is a very simple method right before the eyes of far too many motorists: those little mascots dangling on a string. When the car is stationary or moving at a constant speed, the mascot hangs straight down (disregarding any possible complications due to the relativity theory). If the speed is increased the mascot will swing back on the string; the greater the acceleration the farther it will swing back (see figure 1). The accelerometer in the fuel economiser . is based on this principle. As shown in figure 2, the heart of the device is ‘a i weight on a plate on a rod in a box' . . . 1 home made of course. As the car ac- celerates the weighted strip will swing on its spindle, and in doing so, it varies the frequency of an audio oscillator. At low acceleration rates, the output fre- quency will be so low that it is virtually inaudible. Increasing acceleration will produce a low buzz. Really taking off will be rewarded by a distinct tone. A sort of Swinging Strip Controlled Oscil- lator (SSCO), really. The mechanical details can be seen in figure 2. The strip can be made from a piece of copper laminated board with a collar soldered to the upper end. A bolt is passed through this and fitted to a base plate allowing the strip to rotate Not a novel method of mounting loudspeakers but an audible aid to smoother (and therefore cheaper) motoring. While most things are 'going decimal' it would appear that motoring costs are 'going logarithmic' and any method of saving money on the road must be greeted with enthusiasm. This particular idea is aimed at the cost of acceleration (accelerating costs?). In other words, if you resisted the impulse to put your foot down quite so hard all of next week, how much petrol will you save? How do you try it? W.H.M. van Dreumel It is, of course, possible to calculate fairly precisely how much energy it takes to accelerate your particular car from, say, 20 to 40 miles per hour (no, please do not phone the A. A. or R.A.C.) Briefly the figures go like this: if your car is initially traveling at x metres per second and t seconds later your speed has increased to y metres per second, the acceleration (a) was y — x metres per second in t seconds, or a = ^(m/s 2 >. Why (m/s 2 )? Simple, because speed is measured in metres per second and not miles per second (we've been doing it all wrong folks). How does all this help us? In short, it is possible to determine the rate at which the speed of a car is increasing by an acceleration-measuring device: an ac- fuel economiser elektor november 1979 — 11-23 S e ) r r d I, r ; i II is :h in c- :h' an id a >le DO )ts is nt vn ns he ng he ng ser f in 'a ac- ing ies At rre- illy vill off . A icil- in n a h a lOlt □ a freely. The side view in figure 2c will clarify this. At the lower end of the strip, a heavy nut can be used as a weight — how heavy can best be found by experiment. As can be seen in figures 2b and 2c, an LED and an LDR are mounted on either side of the box so that the LDR cannot 'see' the LED when the strip is hanging straight down. The LED should only start to illuminate the LDR after the strip has swung back through a small angle. Different indication character- istics can be achieved by tailoring the shape of the cut-out in the strip: an almost square shape as shown will give a fairly abrupt changeover from a low to a high frequency while a wedge shape will give a more gradual increase. The box for the prototype was made by soldering pieces of copper laminated board together and the author actually filled the bottom of this box with heavy engine oil to damp the movement of the weighted strip. However, that is up to our more wealthy readers to experiment with. We have yet to try porridge, as a cheaper alternative. The complete unit can be mounted at a suitable point in the car. The strip must be free to swing back, of course, and it must hang straight down — two restric- tions that limit the choice of suitable positions in the car somewhat. If the unit proves too sensitive — beeping at even quite modest acceleration — there are two solutions. One is to use a heavier weight, but this may involve dismantling half the unit. The alternative may therefore be preferable: mount the box at an angle, in such a way that the strip rests against the side of the box when the car is stationary. A certain minimum acceleration is then required before it even starts to move. The circuit The astable multivibrator circuit that provides the warning tone is shown in figure 3. It is a standard '555' con- figuration, that has been described in various guises more than once . . . The output frequency depends on the value of the LDR, R1 and Cl . As more or less light falls on the LDR, the frequency will be higher or lower. The range of frequencies produced can be modified by selecting different values for Cl . A high impedance loudspeaker should be used with 60 fZ being an absolute minimum. If only a lower impedance loudspeaker is available then a series resistor must be used to make the total resistance over 60 SZ. We are not suggesting that this article will prevent the next oil crisis, but it may help to make a small reduction on the motoring costs of our readers. 'Since using the Elektor accelerometer Fuel Economiser, I'm now a 1 m/s 3 driver'. M Figure 2. Mechanical details of a more sophisticated version of the weight on a string principle Figure 3. The circuit of the audio oscillator. 11-24 — elektor november 1979 I played TV games . . . The load. Store, Branch, Compare, 'Miscellaneous' and 'Program Status' instructions were all dealt with last month. As illustrated in Tables A . . . E in that article, these instructions are sufficient for quite interesting little programs. However, as the extended version of the same program on the new ESS record illustrates, programs can be made rather more sophisticated by the use of the remaining instructions: Arithmetic, Logical and Rotate. (The Input/Output instructions cannot be used in the basic version of the TV games computer). Arithmetic Even though the computer will not normally be required to do sums, the so-called arithmetical instructions are add or subtract operation, provided the 'with carry' bit (bit 3 in the PSL) is set. If the WC bit is not set. Carry or Borrow information is ignored - in practice, this has proved even more useful! — The Inter-Digit Carry bit (IDC): this gives the Carry or Borrow information that applies between the lower four and the upper four bits in the register affected. This information can be ignored when binary arithmetic is performed, but it may be essential when doing decimal calculations. — The Overflow bit (OVF): since large numbers (greater than 7F) can be interpreted as negative numbers, things can go wrong in an addition. For instance, 70 + 28 will give the result 98 — but this is equivalent to a I pkfyinl TV games ....and it was fun! Last month, we examined the basic principles of the TV games computer, and discussed the more important instructions. In this second article we will take a closer look at the rest of the instruction set, explain some useful program- ming tricks and list some useful routines available in the existing 'monitor' software. With this information, and a little practice, it should be possible to develop quite interesting programs. At this moment, we've got half-a-dozen ideas, and we hope to get them on ESS records in the not-too-distant future! quite useful. As shown in Table 8, a complete set of add and subtract instructions are available; the only other instruction under this heading is 'decimal adjust register'. Both addition and subtraction are straightforward : 03 + 05=8; 19 -02 = 17; 28+ 13 = 3B; and so on. The calculations are per- formed in 8-bit true binary and negative numbers are two's complement, so that the hexadecimal calculations are valid. As a result of these calculations, three bits in the Program Status Lower will be set or reset: — The Carry/Borrow bit (C): to be precise, this is set to 1 by an addition that generates a carry, and to Oby a subtraction that generates a borrow. However, in most cases it is sufficient to know that this bit will be in- terpreted correctly in any following negative number (-68). This type of ambiguous result is indicated by the setting of the overflow bit: if two positive numbers are added or subtracted and the result is 'negative' the OVF bit is set. Similarly, if a positive result is obtained from a calculation on two negative numbers. So much for addition and subtraction. In practice, it is often sufficient to know that clearing the 'WC' bit results in a straightforward calculation, without any unexpected 'carries' or 'borrows'. Decimal Adjust Register This instruction allows BCD sign magni- tude arithmetic to be performed on packed digits. Full details are given in the instruction manual. So far, we have got by quite well without it; the only time it might have been useful (for a Table 8 Arithmetic description example comments Add to register Zero (ADDZ) 81 R0: = R1 + R0 Add Immediate (ADDI) 84xx xx = data Add Relative (ADDR) 88yy yy = displacement Add Absolute (ADDA) 8Czzzz zzzz = address Subtract from register Zero (SUBZI A1 R0: = R0 - R1 Subtract Immediate (SUBI ) A4xx xx = data Subtract Relative (SUBR) A8yy yy = displacement Subtract Absolute (SUBA) ACzzzz zzzz = address Decimal Adjust Register (DARI 94 played TV games . . . elektor november 1 979 — 1 1-25 decrementing time display on the screen) it seemed simpler to subtract six at each '0 — F crossing', as follows: F707 TMI, R7 r- 9802 BCFR A706 SUBI, R7 U-etc. Logic The instruction set includes AND, Inclusive Or (IOR) and Exclusive Or (EOR) instructions, as summarised in Table 9. The corresponding logic oper- ations are given in Table 10; for most practical applications, it is easier to describe the effects in words: AND An AND instruction causes two groups of 8 bits to be compared; in the result, only those bits will be logic 1 that were 1 in both of the original groups. This instruction can therefore be used as a 'data mask'. As an example, assume that some type of delay routine or 'clock' is counting in R3, and that the three least significant bits are used to determine the screen colour. This can be achieved as follows: 03 LODZ, R3 4407 ANDI, R0 8406 ADDI, R0 CC1FC6 STRA, R0 After 'screening out' the five higher bits by means of the AND instruction, the 'Background enable' bit is added, and the result stored in the PVI. Inclusive Or Once again, two groups of eight bits are compared; in this case, however, all bits that are logic 1 in either of the two groups will be 1 in the result. Another way of looking at this is to say that only those bits will be logic 0 in the result that were 0 in both of the original groups. A complementary data mask, in other words! Both AND and IOR instructions can also be used to set or reset one or more bits in a group of eight, without affecting the others. In the example given above, for instance, if the contents of R3 are to be used for both screen and background colours: 03 LODZ, R3 6408 IORI, R0 CC1FC6 STRA, R0 The Inclusive Or instruction is added to ensure that the Background enable bit is always set. Exclusive Or Quite apart from its 'logical' function, this instruction can be used as a 'selective inverter'. If we take one group of 8 bits as the original data and exor it to a second group, the result will be that some of the bits in the first group will be inverted, as apecified in the second group. Complicated? Not really. Each bit in one group specif ies what happens to its partner in the other: if it's logic 1, the partner is inverted; if it's logic 0, the corresponding bit in the other group is PHILIPS PROGRAMMABLE VIDEO INTERFACE (PVI) Ltectiour. i components ‘ , SHAPE 3 0A34 25 25 data-address 0A36 25 26 data-address 0A38 25 27 data-address 0A3A 27 28 data-address 0A3C 00 29 data-address 0A3E 00 40 data-address 0A40 00 41 data-address 0A42 00 42 data-address 0A44 70 43 data-address 0A46 50 44 data-address * SHAPE 4 0A48 60 45 data-address 0A4A 50 46 data-address 0A4C 50 47 data-address 0A4E 50 48 data-address 0A50 00 49 data-address Start address: 09C7. Return to monitor by operating PC key. routine is given in Table 12. After the presets and a 'wait for VRLE' routine, the first scan is requested: '3F0183 BSTA, UN'. After the scan, the two highest bits in R1 indicate the 'scan status'. If bit 6 is at logic 1, this was the first scan and so a further scan is required; the program branches back to the 'wait for VRLE' routine. After the second scan, bit 6 is at logic 0 and bit 7 indicates whether one key was depressed during the two scans: it is one if this is the case, and zero if no key or two or more keys were operated. Note that 'key operated' (bit 7 is logic 1) corresponds to a negative number, so the condition code will be set to 10. A further possibility, not used in this routine, is to reset only bit 7 at address 089F. Bit 5 in R1 will then indicate if a key is (still) depressed. To get back to the routine given in Table 12, after the second scan (when reaching address 0FE6, in other words) the lower five bits in R 1 give the number of the operated key. The corresponding hexadecimal numbers are listed in figure 3a; the indications at the top left- hand corner correspond to the key indications suggested for the monitor routine. It should be noted that these numbers are only valid if bit 7 in R 1 is logic 1, as mentioned above; otherwise, '00' will appear if the data at address 089F was cleared completely, or else the previous key code if only bit 7 was reset. These key codes can be ideal for many applications. It is particularly useful that the lower four bits are identical for both keyboards, and the fifth bit indicates which keyboard was used. However, in some cases an alternative code is more suitable, and this is obtained by the second part of the routine (from address 0FE6 to 0FF5). The translated key codes shown in figure 3b will be transferred into Register 0. This code has several advantages. For the sixteen 'number keys', the data simply corresponds to the key number. All other keys are distinguished by the fact that bit 7 is logic 1 ; furthermore, bit 6 is logic 1 for the '+' and ' keys only. Similarly, bit 5 uniquely identifies the RCAS and WCAS keys. The only disadvantage is that the upper control (UC), lower control (LC) and reset keys (the latter only if the key is wired as part of the keyboard) are all translated as 80, since they are not used in the monitor routines. Finally, an additional subroutine using the keyboard scan routine is included from address 0FF6 on: 'Wait for key release'. This routine simply repeats the keyboard scan until the indication '30', for 'no key', is obtained. Some little routines After the extensive discussion of the keyboard scan routines, it should come as a welcome relief to take a look at some little subroutines. I played TV games . . . elektor november 1979 — 11-29 dear duplicates The instruction '3F009E' (BSTA, UN, 009E) causes 'FE' to be loaded into the four 'vertical offset duplicate' addresses: 1 F0D, 1F1D, 1 F2D and 1F4D. The result is that only the basic objects will appear on the screen, without any duplicates. Alternatively, any other desired vertical offset can be loaded by first storing it in R0 and then starting the subroutine at address 00A0. Only register zero is used in this routine. dear objects All object shape data can be cleared by storing 00 at all addresses from 1F00to 1F4F. This is accomplished by a subroutine starting at address 01 6E. Any other data present in R0 (FF, say) can be loaded into all these addresses by starting the subroutine at address 016F. Registers used: R0 and R2. split register The 8 bits in a register can be written as two hexadecimal characters. Sometimes it is useful to actually separate these two characters. A subroutine, starting at address 035E, splits the data in R1. If the original data in this register was 'XY', the subroutine will leave '0Y' in R1 and load '0X' into R0. Text display routines There are, of course, several other small subroutines available in the monitor software, Flowever, most of these are closely related to the text display routines, and so it is easier to treat them as a separate group. initiate P VI This subroutine (starting at address 0161) presets the PVI for text display. It has the following effects: - objects size 2 ('AA' in 1FC0); — correct colour (yellow objects, blue screen), — '00' in 1FC3 (form/pos); - sound off; Table 12 0FD0 20 EORZ, R0 0FD1 CC089F STRA, R0 0FD4 7712 PPSL, RS, COM 0FD6 7509 CPSL, WC, C 0FD8 ► ■0CIFCB LODA, R0 0FDB F440 TMI, R0 0FDD -9879 BCFR 0FDF 3F0183 BSTA, UN 0FE2 F540 TMI, R1 0FE4 — •1872 BCTR 0FE6 01 LODZ, R1 0FE7 — -1'A05 BCTR 0FE9 0430 LODI, R0 0FEB 7510 CPSL, RS 0FED 17 RETC, UN 0FEE + -451 F ANDI, R1 0FF0 0D6122 LODA, I/R1 0FF3 7510 CPSL, RS 0FF5 17 RETC, UN 0FF6 r> ►3B58 BSTR, UN 0FF8 F430 TMI, R0 0FFA L_ -987 A BCFR 0FFC 17 RETC, UN Registers used: R0, R1\ R2\ R3‘; Subroutine levels used: 2 for 'keyboard scan'. 3 for 'wait for key release'. presets for key scan wait for VRLE Gosub 'Keyscan repeat if 1 s* scan load ’30' if no key else load translated code and return (keyscan (no key) return l keyboard scan f and decode wait for key release, then return — disable score ('AA' in 1FC8 and 1 FC9); — clear objects ('00' in 1F00 . . . 1F4F). Note that all object position data is set to 00 by this routine! Furthermore, the background data is not cleared; the background is merely made 'invisible' by giving it the same colour as the screen. Registers used: R0, R1.R2. message data When writing a text on the screen, a lot of complicated data must obviously be loaded into the 'object shape' area in the PVI. Fortunately, several characters are pre-programmed in the monitor software, as listed in Table 13. The first 28 (up to and including the 'x' sign) are deliberately programmed; the rest are 'accidental'. A complete scan of all characters and other shapes that can be obtained in this way is included as one of the routines in File 2 on the ESS 006 record. To obtain one line of text on the screen, the codes derived from Table 13 must be loaded into addresses 0890 . . . 0897: eight characters in all for each line. If spaces are required, the code '17' must be stored in the corresponding addresses. In some cases, it may be useful to first store 8 spaces and then store the one or two characters required. There is a subroutine for this, starting at address 02D9; it uses R0 and R2. A program example may serve to clarify the points discussed so far. The routine given in Table 14 (derived from Table 7 in last month's article) will produce a complete display of the most useful characters. After the usual 'interrupt inhibit' instruction, the first step is to initiate 3 0 System Left-hand Right-hand keys keyboard keyboard uc 0F RCAS 03 WCAS 07 c 0B H r- UJ ■ STRT BP REG 8 9 A B 0E 92 06 0A 12 16 1 A LC PC MEM 4 5 6 0D 01 05 09 11 15 19 RESET - ♦ 0 1 2 0C- 00 04 08 10 14 18 * Note that this code is only obtained if this key is wired as part of the normal keyboard — not if it is wired direct to the reset input, as in the suggested keyboard layout. b System Left-hand Right-hand keys keyboard keyboard UC 80 RCAS 90 WCAS 93 c 0C HI H ■ STRT BP REG 8 9 A B 8A 84 87 08 09 0A 0B LC PC MEM 4 5 6 OH 80 8D 81 04 05 06 wi RESET - ♦ 0 1 2 3 80' _c L E0 00 01 02 03 30 * no key operated *see note under figure 3a. 1 1 -30 — elektor november 1979 I played TV games . . . the PVI, as described above: '3F0161'. Then R3 and R1 are preset, for the total number of characters (42 = 2A) and the number of characters per line (07) respectively; the desired character codes are stored from address 0930 on. The 'load 8 spaces' routine is included as the next step ('3F02D9'). Not that it is strictly necessary in this case (we're already loading seven characters in each line, and one more space could easily be added), but it serves to illustrate the principle. The following small loop (from 090C to 0912) transfers the first line of code numbers (from address 0953 on) to the 'message line scratch' (from address 0890 on). We now come to the next monitor subroutine: load Mime This monitor subroutine (at address 020E) translates the codes stored in the message line scratch to the corresponding shape data for the four objects, and stores the results in a 'display scratch' (from address 0800 to 088F, for all six lines!). Since this routine uses all four active registers (R0 . . . R3), it would alter the character count data in R3. One solution would be to use the Load Immediate instruction at address 0907 as scratch, as described earlier. In this program, an alternative solution is used: the upper register bank is selected before branching to the subroutine. The next step is to check whether all characters, for all six lines, have been loaded. As long as this is not the case the program branches to address 0927, bringing us to the next subroutine: Table 13 character code character code character code character code 0 00 A 0A P 14 7 5F 1 01 b 0B r 15 8A 2 02 C 0C = 16 nil) AA 3 03 d 0D space 17 1 BB 4 04 E 0E + 18 T BC 5 05 F OF - 19 1 DF 6 06 G 10 1 A : (2) E6 7 07 L 11 X IB . F7 8 08 1 12 1 (3) A2 9 09 n 13 Notes: (1 ) this n is slightly larger than the 'official' version (code 13), and looks better between capitals. (2) similarly, this colon is larger than that obtained by code 1 A, which can be useful. (3) the exclamation mark is too small, actually, but no better version exists . . . (4) the 0 (code 00) can be used as the letter O; similarly, a 5 makes a good S and a 2 will pass for a Z. Table 14 0900 0902 0905 0907 0909 090C 090 F 0912 0914 0916 0919 091 B 091 D 0920 0922 0924 0927 0929 092C 092 E 0930 0937 093 E 0945 094C 0953 7620 PPSU, II 3F0161 BSTA, UN (clear/initiate PVI) 072A LODI, R3 ■0507 LODI, R1 3F02D9 BSTA, UN (load 8 spaces) •0F4930 LODA, I-R3 (messline data) CD4890 STRA, I-R1 5978 BRNR, R1 7710 PPSL, RS 3F020E BSTA, UN (load Mline) 7510 CPSL, RS 5 BOA BRNR, R3 •0C1E89 LODA, R0 F410 TMI, R0 wait for '+' key release 1879 BCTR 1 F0038 BCTA, UN return to monitor 7710 PPSL, RS 3F02CF BSTA, UN (scroll) 7510 CPSL. RS 1 B57 BCTR. UN 5F A2 1 7 8 A 1 7 E6 F7 sixth line 02 16 17 18 19 1 A IB fifth line AA 13 00 14 15 05 BC fourth line 0E 0F 10 12 DF 1 1 BB third line 07 08 09 0A 0B 0C 0D second line 00 01 02 03 04 05 06 first line DATA Start address: 0900 scroll To be more precise, this subroutine (from address 02CF) should be listed as 'scroll and load 8 spaces in Mline'. It has the following effects: — all object display data in the display scratch is moved up one line, from sixth to fifth, from fifth to fourth, and so on; the data for the first line is lost; — the code for 'space' (17) is loaded in the eight message line scratch pos- itions. Since this routine uses registers 0, 1 and 2, it is again padded by register-bank- select instructions. Unnecessary, in this case, since the only register data that must be preserved is that in R3 — but once again included to illustrate the principle. After this routine, the program branches back to address 0907, to load the next line. Once all six lines have been loaded, the branch instruction at address 091 B will not be executed: the data in R3 are now zero. An uncommon program ending follows: — wait for '+' key release — the program is started by operating this key, and the microprocessor is so fast that it will have finished the program before you have time to release the key! — return to monitor at address 0038. This transfers control back to the monitor program in such a way that it takes care of putting the text on the screen, without first writing any message of its own! In most cases, however, this easy way out will not be possible. A further played TV games . . elektor november 1 979 — 1 1*31 monitor subroutine is then required to get the message on the screen: Table 15 display six lines The six lines on the screen each consist of all four objects; lines 2 ... 6 are 3e actually the duplicates, of course. To F get the desired text on the screen the A object shape data for each line must be A retrieved from the display scratch at the ® correct moment, and stored in theobject F shape areas in the PVI. 6 The monitor subroutine that does this 7 starts at address 0055; it uses registers 2 R0, R1 and R2. To obtain a correct dis- play; the 'COM' bit in the PSL must be set (instruction: 7702 = PPSL, COM). Furthermore, control must be trans- ferred to this routine at the end of each frame; the return from subroutine will not occur before the sixth line has been displayed. This means that all further program checks or other routines can only be executed just before or during the 'frame end'. — As an illustration, the program given in Table 14 can be modified according to Table 15. All text display routines are now incorporated in the program. However, the disadvantage will be obvious when the PC key is operated: initially, the monitor will scroll, reload the data from message line scratch to display scratch, scroll again, and then add the line 'PC=\ All this doesn't improve the display . . . Interrupt facility Last month, our advice regarding the interrupt facility could be summed up in three words: Don't use it. However, we didn’t follow our own advice: witness the 'space shoot-out' program on the new ESS record! Not that we consider ourselves expert in this field, but at least we now have some experience to pass on. Two or three tricks, in particular. m id it re 8 . ie at in «y iy er Selecting interrupts The PVI generates interrupt requests each time an object (or duplicate) is completed, and at the end of each frame. As long as the Interrupt Inhibit bit in the Program Status Upper is not set, all of these interrupt requests will be acknowledged. No matter what caused the interrupt (object 1 complete? duplicate 3 complete? end of frame? or whatever . . . ), the results will be the same: the interrupt inhibit bit is set by the processor, the running program is interrupted, and the program section starting at address 0903 is run as a subroutine. If we assume that only the end-of-frame interrupt is of interest in a program, all others must be ignored. This is not too difficult: the 'sense' bit in the PSU is ogic 1 at the end of the frame, so the interrupt subroutine at address 0903 can be started as follows: 0903 B480 TPSU, sense 0905 36 RETE — change the instruction at address 0924 to '1 F095A' (instead of 1 F0038}; add the following section of program: 095A -►[♦•0C1FCB LODA, R0 0950 F440 TMI, R0 095 F 1—9879 BCFR 0961 0C1E88 LODA, R0 0964 F420 TMI, R0 0966 1C0000 BCTA 0969 7702 PPSL, COM 096 B 3F0055 BCTA. UN 096 E 1 B6A BCTR, UN wait for VRLE return to monitor if 'PC' display 6 lines Table 16 0900 0903 0905 0906 0908 090A 090C 090 D 0910 0912 0914 0917 0919 091 A 091C 091 F 0921 0923 0925 0928 092 A 092 D 092F 0930 0931 0932 0934 0935 0936 0939 093C 093 E 0941 0944 0947 094A 094 D 094 F 0951 0954 0956 0958 095A 095 D 095 F 1 F0958 BCTA, UN B480 TPSU, sense 16 RETC B440 TPSU .flag — -1808 BCTR 7640 PPSU, flag 20 EORZ, R0 CC089F STRA, R0 r- 1B02 BCTR, UN 1— U-7440 CPSU, flag U-3F0181 BSTA, UN — 9A38 BCFR 01 LODZ, R1 451 F ANDI, R1 0D6122 LODA, I/R1 E4E0 COMI, R0 — 182E BCTR F480 TMI, R0 1C0000 BCTA C804 STRR, R0 3F02CF BCTA, UN 0400 LODI, R0 D0 RRL, R0 D0 RRL, R0 D0 RRL, R0 0608 LODI, R2 82 ADDZ, R2 Cl STRZ, R1 *-004961 LODA, I-R1 CE4890 STRA, I-R2 -5A78 BRNR, R2 3F020E BSTA, UN ►0C1E8A LODA, R0 6C1E8C IORA, R0 6C1E8D IORA. R0 6C1E8E IORA, R0 44F0 ANDI, R0 1—9870 BCFR BSTA, UN r*- *-7420 CPSU, II L-IBTC BCTR, UN 7620 PPSU, II 3F0161 BSTA, UN 7702 PPSL, COM 1 B73 BCTR, UN vertical interrupts only set/reset flag on alternate frames; keyboard scan routine (no key I translate key code branch if '+' key return to monitor if control key save data in R0 and scroll R0 x 8 > load Mline wait for key release display 6 lines wait for interrupts clear/initiate PVI and set COM bit 0961 05 BC 0A 15 BC 17 17 1 7 data 0 0969 0 B 0E 10 12 AA 17 17 17 data 1 0971 0A AA0F 0A AA 10 17 17 data 2 0979 0D 0E 0B 56 BC 1 7 1 7 1 7 data 3 0981 0E AA0D 17 17 17 17 17 data 4 0989 0E 12 AA0D0E 17 17 17 data 5 0991 0E AA 0D 0E 1 7 1 7 1 7 1 7 data 6 0999 0F 12 AA 17 17 17 17 17 data 7 09 A 1 OF 56 AA 1 7 17 17 17 17 data 8 09A9 11 00 11 17 17 17 17 17 data 9 09B1 05 14 0A 05 05 17 17 17 data A 09B9 15 12 10 00 11 0A 0D 0E data B 09C1 AA 12 0C 0E 17 17 17 17 data C 09C9 0A 0A 15 0D 12 10 17 17 data D 09D1 AA0E BC BC 1 7 1717 17 data E 09D9 10 0E AA BC 12 11 0E 17 data F 11-32 — elektor november 1 979 I played TV games . . . I If the sense bit is not set, the TPSU instruction will result in the condition Table 17 code 10. The 'return and enable inter- rupt' instruction (RETE) is then ex- ecuted, terminating the interrupt subroutine! Only if the sense bit proves to be logic 1 , at the end of the frame, will the following interrupt routine be executed. Usually, that is, because there is one minor problem — but we'll come to that in a minute. A more extensive interrupt select procedure is also possible. In the 'space shoot-out' program mentioned above, the program actually starts as follows: 3900 1FQ90B BCTA, UN (to main program) 3903 B480 TPSU, sense 0905 1C0A1O BCTA (to vertical interrupt routine) 3908 1 F09D5 BCTA, UN (to object interrupt routine) 390B 7620 PPSU, II (main program starts here) In this case, if the sense bit is set the conditional branch at address 0905 will be executed, starting the end-of-frame interrupt routine. Otherwise, this branch instruction will be ignored and the following (unconditional) branch will start the object-complete interrupt routine. The latter starts with a further check routine: 99D5 0C1FCA LODA, R0 object 3 0908 F402 TMI, R0 complete? 09DA 36 RETE return if not The final result is that only two basic interrupt requests will be acknowledged: frame-end and object 3 (or duplicate 3) complete. All other object or duplicate complete interrupts will be ignored. When testing this program, one problem was found: Sometimes, the frame-end routine was missed. This error was traced to the fact that an 'object 3 complete' interrupt just before the frame end initiates the corresponding routine - and the latter 'over-runs' the frame end, so that no vertical interrupt was found! The solution, in this case, was simple: make sure that no 'object 3 complete' interrupts can occur just before the end of the frame, by selecting a suitable sequence of 'vertical offset duplicate' values. 09 (30 0903 0905 0906 0908 09OA 090C 090 E 0910 0913 0916 0918 091 B 091 E 0920 0922 0924 0926 0928 092A 092C 092 F 0931 0934 0936 0939 093C 093 F 0942 0945 0948 094 A 094 D 094 F 0952 0954 0957 0958 095B 095 D 095 F 0961 1F0990 BCTA, UN B480 TPSU, sense 16 RETC B440 TPSU, flag — 1804 BCTR 7640 PPSU, flag --1B02 BCTR, UN l*-7440 CPSU, flag -►0D1FCC LODA, R1 0E1FCD LODA, R2 C90B STRR, R1 CE095C STRA, R1 3F0055 BSTA, UN 0702 LODI, R3 0602 LODI, R2 0500 LODI, R1 B440 TPSU, flag | — 1802 BCTR 0604 LODI, R2 ♦-0418 LODI, R0 CC096D STRA, R0 04E0 LODI, R0 CC0984 STRA, R0 04CD LODI, R0 CC0985 STRA, R0 0E4963 LODA, I-R2 CC0987 STRA, R0 CC098A STRA, R0 3F035E BSTA, UN 3F0967 BSTA, UN 0498 LODI, R0 CC096D STRA, R0 040 E LODI, R0 CC0984 STRA, R0 046 D LODI, R0 CC0985 STRA, R0 01 LODZ, R1 3F0967 BSTA. UN 0500 LODI, R1 - FB4B BDRR, R3 r*- 7420 CPSU, II L 1 B7C BCTR, UN vertical interrupts only flag on alternate frames save joystick data display 6 lines joystick data! (IFCCI presets for subroutine split register presets for subroutine joystick data! (IFCD) wait for interrupts 0963 89 71 41 29 address data (continued on next page! -+■ Note: at addresses 096D. 0983 and 0985 either of the alternatives given can be entered. The program modifies these instructions as requiredl Start address: 0900. Interrupt enable A closer look at the program section given above (addresses 0900 to 090B) will lead to a surprise: the main program starts (at address 090B) by setting the interrupt inhibit bit! This means that no interrupt requests will be acknowledged — so what's the point of including interrupt routines? Obviously, at some point in the program the interrupt inhibit bit must be reset. It is, after storing all kinds of initial data in the PVI and presetting a whole series of 'scratch' bytes in the program. Then, at address 09D1 to be precise, the following two instructions are inserted: 0901 i*7420 CPSU.II "1 wait for 0903 ^-1670 BCTR, UN J interrupts The processor will go round and round this loop, until an interrupt occurs. The interrupt routine will then be executed (again setting the interrupt inhibit bit, automatically); at the end of the inter- rupt routine, a 'return' instruction will cause the processor to jump back into this 'wait' loop. Note that the interrupt inhibit bit is reset in the loop, so that it is unimportant whether a 'normal' return instruction (17, say) or a return- and-enable-interrupt instruction is used. As an illustration of the use of inter- rupts, a program is given in Table 16. Not that the same results couldn't have been obtained without using this facility! The data given from address 0961 on corresponds to a series of sixteen words, one for each of the 'number' keys. If other words are required, the data can be derived from table 13. Note that each word must consist of 8 letters or less; if less than 8 letters are used, the remaining pos- itions on each line must be filled with spaces (code 17). Joysticks Saved to the last, because we have very little experience with them . . . The basic principle is fairly straightforward, however. Two addresses in the PVI, 1FCC and IFCD, correspond to the left-hand and right-hand joysticks, respectively. When the flag is set, the vertical direction of each joystick is scanned and the results are stored at the corresponding address; if the flag is not set, the horizontal setting is scanned. The data in the two PVI addresses is only valid at the end of the frame — when the sense bit is at logic 1, in other words. A low data value in address 1FCC or 1 FCD corresponds to 'up' or 'right', depending on the setting of the flag during the previous frame (when the actual A-D conversion took place). The actual range of values obtained varies from one joystick to another. Unfortunately! This means that it is not easy to write a piogram that is suitable in all cases. In fact, the 'space shoot-out' program on the new ESS recording does contain a joystick -scan routine . . . but elektor november 1979 — 11-33 I played TV games . . . 0967 7710 PPSL, RS 0969 0700 LODI, R3 096B F401 TMI, R0 0960 - 1802/9802 BCTR/BCFR 096 F 0701 LODI, R3 0971 “►440E ANDI, R0 0973 C2 STRZ, R2 0974 D2 RRL, R2 0975 82 ADDZ, R2 0976 0506 LODI, R1 0978 81 ADDZ, R1 0979 C2 STRZ, R2 097A — -►0E4278 LODA, I-R2 097D r 5B04 BRNR, R3 097F 00 RRL, R0 0980 00 RRL, R0 0981 DO RRL, R0 0982 00 RRL, R0 0983 L*.44E0/440E ANDI, R0 0985 CD6829/ STRA/IORA, 6D6829 0988 CD6829 STRA, I/R1 0988 — - F96D BDRR, R1 098D 7510 CPSL, R5 098 F 17 RETC, UN 0990 7620 PPSU, II 0992 3F0161 BSTA, UN 0995 04CC LODI, R0 0997 C80F STRR, R0 0999 0702 LODI, R3 099B -►0610 LODI, R2 099 D 0508 LODI, R1 099 F ►7710 PPSL, RS 09A1 3F02CF BSTA, UN 09A4 7510 CPSL, RS 09A6 [-►0E49CC LODA, I-R2 09A9 CD4890 STRA, I-R1 09AC 1—5978 BRNR, R1 09AE 04C4 LODI, R0 0980 C876 STRR, R0 09B2 7710 PPSL, RS 09 B4 3F020E BSTA, UN 09B7 7510 CPSL, RS 09B9 0504 LODI, R1 09BB 1 — 5A62 BRNR. R2 09BD 1 — FB5C BDRR, R3 09BF 7702 PPSL, COM 09C1 1F095F BCTA, UN 09C4 01 0F 0C 00 09C8 01 0F 0C 0C 09CC 0F 11 0A 10 17 00 0F 0F 09 D4 OF 11 0A 10 17 00 AA 17 preset R3 3 x R0 set R1 , R2 (clear/initiate PVI) address preset scrall Messline data address preset load Mline basic message data it's blocked! The text included with the record explains how to re activate it. Obviously, this is a very unsatisfactory state of affairs. However, we have a solution. The program given in Table 17 can be used to test and 'calibrate' joystick controls. It reads the data in the two PVI addresses, with the flag both 'on' and 'off', and displays the results on the screen as follows: FLAGON (= horizontal) 1FCC 75 (= left) 1FCDAD {= right) FLAG OFF (= vertical) 1 FCC 11 (= left) 1FCD83 (= right) The data found at the two addresses is updated on the screen as required. The values given above (75, AD, 11, 83) are just examples, without any special meaning. If the joysticks are wired as shown in the original article, address 1 FCC should correspond to the left-hand joystick; 'Flag on' should correspond to vertical movement; and low data values should be obtained at the extreme 'up' and 'right' positions. Now, a request. If those readers who have a set of joysticks could let us know the results obtained (both with joysticks centered and in the various extreme positions) we can get some idea of the tolerances involved. It would also be interesting to know what value is obtained when no joysticks are connected — our prototypes read '0D' in that case. With this information, it should be possible to work out some kind of 'universal' joystick routine. Then we can start developing suitable programs! In conclusion 'And that' quoth he 'is that.'. Prac- tically all our experience, up to the minute of going to print, is included in these two articles, If we find any more tricks, you'll be the first to know. Meanwhile, we hope that you can start developing interesting programs! M Misprint The information provided with the first ESS record for the TV games computer states that the speed of the 'surround' game can be modified by altering the data at address 0D02. Wrong! it should be address 0D20. 11-34 — elektor november 1 979 short-wave converter The circuit is simplicity itself. With SI in the position shown, the aerial is con- nected to the input bandpass filter. This consists of two LC resonant circuits (LI , Cl, C2 and L2, C3, C4), tightly coupled by C5. The input filter is followed by a self- oscillating mixer stage, built around a dual-gate MOSFET (T1) and a crystal. The desired output frequencies are fed through a further bandpass filter con- sisting of three LC networks (L3/C9, L4/C10 and L5/C11) and a coupling capacitor (Cl 2) to the aerial input of the medium-wave receiver. This receiver is used to tune in to the desired short- wave station. sImhIwim* converter The converter is preset to a particular short-wave band. Table 1 gives the values for LI , L2, C5 and the crystal for the various short-wave bands. If several different bands are to be available, these components would have to be switched; a simpler and more reliable solution is to build several converters. In some cases, the short-wave band may not convert exactly to the medium-wave tuning range. If necessary, a slightly different crystal frequency can be used. The alignment procedure is straight- forward: — Tune in to a short-wave broadcast that is converted to approximately 1400 kHz, and adjust C12 for maxi- j mum signal strength. — Tune in to a short-wave station that appears near 1 500 kHz in the medium-wave band, and adjust C4 for maximum signal strength. - Finally, adjust C2 for maximum signal strength at a station that appears near 1300 kHz. — The adjustments of C4 and C2 are repeated until no further improve- ment can be obtained. It will be apparent from the circuit that the other position of SI connects the aerial direct to the medium-wave receiver and turns off the converter. M Table 1 Band (metres) LI, L2 (pHI C5 (pF) X-tal (kHz) 75 8.2 10 2300 60 4.7 10 3600 49 3.9 10 4600 41 2.2 8.2 5800 31 1.2 8.2 8300 25 0.82 6.8 10500 19 0.56 5.6 13900 16 0.39 4.7 16400 13 0.27 2.7 20100 11 0.22 2.2 24400 Table 1. The input bandpass filter and the crystal frequency must be chosen for the desired short-wave band. ter short-wave converter elektor november 1979 — 1 1-35 \wr~v* • n ®1 □Sfbssit °.°®J R1 = 100 SI R2.R6.R7 = 47 Cl R3 = 22 k R4 = 100 k R5 « 3k3 Capacitors: Cl = 82 p C2,C4 « 7 ... 80 p (trimmer) C3 = 100 p C5 « see table 1 C6 = 100 n C7.C9.C10.C1 1 = 68 p C8 = 10 p/16 V, Tantalum Cl 2 = 10. . . 40 p (trimmer) L1.L2 = see table 1 L3.L4.L5 = 270 pH Semiconductors: T1 = 3N211 Miscellaneous: X-tal = see table 1 SI = three-pole two-way switch Figure 2. Printed circuit board design and component layout. ionosphere Why is long distance shortwave reception possible? Why is MW only good over short distances during the day? There are so many « 'whys' associated with shortwave reception that many of us are completely in the dark about what frequency to choose, what time to listen, and what is likely to be heard. This article about the ionosphere is intended to take some of the guesswork out of shortwave listening. Table 1 HF broadcast bands Frequency (kHz) Band (m) 2300 2945 120 3200 3400 90 3900 4000 75 4750 5060 60 5950 6200 49 7100 7300 41 9500 9775 31 11700 11975 25 15100 15450 19 11700 17900 16 21450 21750 13 25600 26100 11 Table 2 Amateur bands Frequency IMHz) Band (m) 1.8-2 160 3.5 - 4 80 7 - 7.3 40 14 14.35 20 21 - 21 45 15 27 29.6 10 Long distance radio communication is only possible because of the iono- sphere — a region of the earth's atmos- phere which is between about 90 and 320 km high (60 to 200 miles). Ionis- ation of the ionosphere is attributed to ultraviolet radiation from the sun. The ionised part of the ionosphere is not a single region, but is made up from several different layers. The E layer At about 100 km (70 miles) above the surface of the earth is the lowest useful region of the ionosphere, the E layer. The E layer is so low in the atmosphere that free ions have little distance to travel before they recombine with an electron, this forms a neutral particle which will not reflect radio waves. For this reason the E layer is only useful during the daylight hours and is usually much stronger around noon. It almost fades away after sundown. A phenomenon worth mentioning is 'sporadic E' which is generally of little interest to the shortwave broadcast listener. Sporadic E's are made up of irregular patches of relatively dense ionisation floating in the E layer. These patches are usually found in equatorial regions, but also form in temperate climates in the summer months. How- ever, they can appear at almost any time. The why’s and wherefores are not completely understood, making E pre- dictions virtually impossible. Communication distance via a single E 'hop' is most common between 650 km and 2000 km (400 and 1200 miles) — see figure 1. Signals are gener- ally very strong but may vary over wide ranges. Sporadic E is what usually causes television signals to be received over long distances. TV DXing is a very interesting hobby in itself, but is out- side the scope of this article. The F layer The area or region of the atmosphere which is the real workhorse of long distance communication is the F layer. It is about 280 km (175 miles) above the earth. During the day however, it splits into two separate areas, the FI and F2 layers. They are located about 225 km and 320 km (140 and 200 miles) high respectively on days when the ionisation level is high. A good day! After sunset they combine back into the single F region. The maximum single hop distance of the F layer is about 4000 km (2400 miles) - see figure 2, which also shows the relative heights of the various layers. The F region is at such a high altitude that recombination of ions and electrons into neutral particles takes place at a very slow rate. The level of ionisation starts to decrease after sundown, and becomes progressively weaker until reaching its lowest level just before sunrise. This progressive decrease in the ionisation level can be noticed by the early disappearance of stations that were operating on frequencies close to the highest useful frequency of the day. The D layer Below the E layer is a region of the ionosphere which doesn't help com- munications at all, but rather hinders it! This region is called the D layer. Radio transmissions on frequencies ionosphere elektor november 1979 — 1 1-37 lower than about 4 ... 8 Mhz can be almost completely absorbed (not re- flected) by the D layer. Of course, the highest frequency absorbed and the amount of absorption is a function of ionisation, which is directly related to the height of the sun. The D layer is strongest during the noon hours in mid- summer. In the winter it is much less intense. Only high angle radiation can manage to pass through the D layer and be reflected back to earth. Since low angle radiation is used for long distance communications it can be seen why only short distance communication is possible on low frequencies when the D layer is ionised. Recap From the above discussion it is apparent that the relative reflectivity of the different layers of the ionosphere is greatly influenced by the sun. The F layer being the highest and most useful layer for long distance com- munication. It is useful around the clock, but becomes progressively weaker as the night draws on. The E layer is is useful for much shorter communi- cation distances, with the lower fre- quencies being reflected better. How- ever, when the D region becomes ionised it begins to absorb those lower frequencies. This limits their use to short distance communication during the day. This effect can best be heard at sunrise in the summer, by listening to the medium wave band. Before dawn many long distance stations should be heard, but as the sun starts to rise (first light) these stations will begin to fade away. Sometimes this takes only a few minutes. At dusk the long distance stations begin to be heard again and become increasingly stronger as dark- ness progresses. Sunspots and other effects There are of course many things which effect the ionosphere and its ability to reflect radio signals. Sunspots Sunspots have, on average, an 1 1 year cycle between the minimum and maxi- mum number of spots, however the cycle may vary between 9 and 13 years. The high and low number of spots vary greatly from cycle to cycle but usually the high count has sharper changes than the low. Sunspot cycles should not be thought of as being sinusoidal. There are times when the number of sun- spots increase to a relatively high level during a period when the norm would be quite low. These isolated highs do not usually last for more than a few months. During the low part of the cycle the ionosphere is relatively weak and high frequency reception conditions are at 11-38 — elektor november 1979 ionosphere their poorest. When the sun has a large number of spots the ionosphere is strong and communication is good up to the higher limits of the HF band (30 MHz... 50 MHz). SIDs and SWFs Sudden increases in solar activity such as solar flares trigger very fast changes in the various layers of the ionosphere. When these conditions occur the varia- tion in the absorption of the D layer is particularly sudden and may last from only a few minutes to a few hours. This suddenness has led to the term SID 'sudden ionospheric disturbance'. SIDs and SWFs (shortwave fade-outs) vary widely in intensity and duration, however the effects tend to be greater in times of high solar activity. Solar radiation There are two principle kinds of solar radiation, ultraviolet light and charged particles. The light travels the distance to earth in about 8 minutes and the effects on the ionosphere are fairly rapid. The particles on the other hand, are moving at a much slower speed and may take up to 40 hours to have any effect on communications. These effects are usually high absorption by the D layer and the production of an aurora, and they sometimes reccur every 27 days — the rotation time of the sun. This reccurrance can continue for as many as 4 or 5 rotations of the sun dependent on the strength of the orig- inal phenomenon. Multi-hop It is possible for a signal to 'hop' more than once, see figure 3. Even though ground reflections and ionospheric absorption take a toll on the signal strength, communications more than half way around the world are possible using multi-hop paths. The signal levels are usually somewhat lower and suffer higher distortion and more fading than do single-hop signals. Fading Fading is sometimes caused when the signal takes two or more paths before arriving at the receiver site with phase d if Terences. If one or more of the paths are unstable, then the changing phase can completely obliterate the signal. Other things like weather fronts and moving air masses also tend to cause unstable radio conditions. The term fading covers an almost infinite variety of phenomena. Angle of radiation and 'muf' The angle at which the transmitted signal strikes the ionosphere has much to do with the 'skip distance’. The distance between the closest and farthest points that communication can be carried out on a given frequency is called the skip zone. In figure 4, point | B is the shortest skip and point A is the longest skip distance for 21 MHz, the distance between these two points is the skip zone. For 14 MHz the skip zone is between points A and C. By studying figure 4 it can be seen that low angle radiation, (the radiation leaving the antenna parallel with the earth's surface) has a longer skip distance than does the radiation going up at a greater angle, i.e. high angle radiation. It should be noted that the bending effect is not only dependent upon the angle at which the waves hit the ionosphere, but also on their frequency. The 'maximum usable frequency' (muf) is the highest frequency that is usable for communications at a given time. The muf also has an effect on the skip distance, as can be seen in figure 4. With a muf of about 28 MHz only the very low angle radiation is being reflected back to earth. As the frequency is lowered the ionosphere appears more intense, therefore reflecting radiation that has higher angles of incidence (see 21 and 14 MHz). This effect can also be heard by listening in on fre- quencies close to the muf at a time when the ionosphere is getting weaker — the skip distance seems to getting longer when in fact the closer stations, which require high angle reflections, are fading away leaving the more distant stations which are being reflected at lower angles. It is apparent from the above discussion that for good long distance communica- tion it is important that the antenna concentrates most of the transmitter power into low angle radiation. The receiver antenna should also be construc- ed so that most of its 'gain' is for low angle radiation. If shorter range com- munication is desired then a lower frequency should be used, together with a higher radiation angle to produce stronger signals. Predictions Making predictions about reception and ionospheric conditions is indeed a tricky business because there are so many variables. However, by taking into account as many known factors as possible, and relating them to past experience, it is possible to make general statements about band condi- tions at a given time for a given fre- quency. Where and when to listen The 90 m and 75 m bands are seldom usable beyond 300 km (180 miles) during the day, but longer distances are usual at night. Static and other atmospheric noise makes use of these bands in the summer months somewhat of a problem. The 60 m, 49 m and 41 m bands have characteristics similar to the two lower bands except the daytime distance is much greater. These three bands also tend to stay open more often at night than do the higher frequency bands. The 31 m, 25 m, and 19 m bands are the real DX bands. During high sunspot years they are open almost continuously. They are especially good in the dawn and dusk periods when the solar activity is low. The 16 m and 13 m bands have very variable propagation which depends on the level of solar activity. During high solar activity the bands are good for very long distance listening, however, they become almost useless during periods of low solar activity. Conclusion The sun is the main factor that domi- nates all radio communications beyond the local level. Radio conditions vary with such obvious cycles as the time of day and season of the year. Since these parameters change with latitude and longitude it is possible to have an almost infinite number of unique communication variations. There are less obvious changes in the ionosphere which are also controlled by the sun, sunspots and other solar radiation. These and many other factors must be taken into account when selecting a frequency which will yield the desired communication path. The optimum results may not always be realised however, the familiarity gained from this article should help reduce the margin of failure and add greatly to one's enjoyment of shortwave listening. M low voltage dimmer low voltage dimmer As is well known, the NE556IC contains two identical universal timers. The device is thus ideally suitable as the basis of a compact, low-loss dimmer circuit for low voltage lamps. One timer is used as a clock generator, whilst the other functions as a monostable multi- vibrator with variable pulse width. As can be seen from the circuit diagram, only a few ancillary components are needed to complete the dimmer. The first timer of the NE556 is connected as an astable multivibrator and provides the required clock signal. The clock frequency is determined by the values of R1, R2 and Cl, and is in the region of 1 kHz. The pulse width or duration of the clock pulses is thus approximately 10ps. The clock signal is fed to the trigger input (pin 8) of the second timer, which is connected as a monostable. The output of the monostable controls a power transistor (T1), which in turn switches the load (i.e. the lamp) on and off. Thus by varying the duty-cycle of the monostable (by means of PI), the lamp is turned on for a greater or smaller length of time, thereby varying its intensity. With the component values shown in the circuit diagram, the duration of the output pulses (pin 9) from the mono- stable can be varied by a factor of 10. The maximum pulse duration (dis- counting the effect of PI) can be calcu- lated from T = 1 .1 x R4 x C2, which in the case of the circuit shown equals roughly 0.4 ms. Thus with a clock fre- quency of 1 kHz, the duty-cycle can be continuously varied between 60 and 96%, which in practice represents quite a suitable range. These values are obtained around the mid-position setting of PI . If the wiper of PI is set to one of the end stops, the circuit will fail to function properly. For this reason it may be worth experimenting with various value resistors in series with PI to make the adjustment range less sensitive. The supply voltage of the circuit can lie anywhere between 5 and 15 V. H 5... 15 V 80001 elektor november 1979— 11-39 Whether a model railway is micro- processor-controlled or hand-operated, a visual display of the 'system status' is always worth while. If nothing else, it makes for an impressive control panel. For some functions, it is even essential to have a clear overview — unless, of course, your main aim is to realistically imitate crashes and derailments. The points, in particular, are extremely important. As many model railway enthusiasts will have discovered, it is not at all easy to see what position the points are in from a distance. Even mechanical ‘point position indicators' are not always particularly clear. The indicator described here provides an unambiguous display on the main control panel. Different coloured LEDs can be used to provide a clear indication at a single glance. The circuit could hardly be simpler. Electro-mechanical points with built-in end switches are used. One of these switches is open and the other is closed when the points are set. The closed switch turns on the corresponding transistor, lighting one set of LEDs. The pushbuttons, electronics and one LED out of each pair can be mounted in the control panel; the other LED in each pair can be mounted alongside the tracks near the corresponding set of points, to give an on-the-spot indication. M 11-40 — elektor november 1979 servo speed control for model boats As shown in the circuit diagram, two 6 V accumulators are used to power the circuit. The upper battery supplies the power when the boat is moving forwards; the lower one is only used for reversing, so it can be much smaller. Potentiometer P2 is controlled by the servo. In the middle of its range, the voltage between the slider and supply common is zero. When the servo alters the setting of this potentiometer, a positive or negative voltage (depending adjustment: the coupling between the servo axle and the potentiometer spindle is tightened when the relative position is correct. A small offset of the poten- tiometer can be compensated by ad- justing PI: a voltmeter is connected between the slider of P2 and supply common, and PI is adjusted so that the meter reads 0 V. The next step is to set P3 to maximum. The servo is moved to one of its extreme positions — 'Full speed ahead', for Sinned CHMilrol for model boats The speed of a model boat can be controlled by varying the supply voltage to the main motor, via remote control. Normally, this control is fully electronic. In the circuit described here, however, a mechanical link is included; the remote control receiver drives a servo-motor and this, in turn, drives a potentiometer that controls the speed of the main motor. (U. Passern) on the direction in which it is rotated) is applied to the non-inverting input of IC1. The output of IC1 will therefore swing either positive (turning on T1 and T3) or negative (turning on T2 and T4). The main motor should be connected so that the boat moves forwards when T3 is turned on. Zenerdiodes D1 and D2 and capacitors Cl and C2 take care of the stabilisation and smoothing of the reference voltages, so that power supply fluctuations have little effect on the motor control. Even so, it is advisable to include interference suppression on the main motor. The first step when setting up the unit is to make sure that the mid position of P2 corresponds to the neutral position of the servo. This is a purely mechanical instance — and P3 is slowly turned down until the maximum permissable voltage across the main motor is ob- tained. Not more than 6 V are available, obviously, but this adjustment makes it possible to use lower voltage motors without danger of burning them out. The transistors need adequate cooling. A heatsink with a thermal resistance not greater than 2.8°C/W should be used, and the transistors must be mounted using mica insulating washers. H 80014 market elektor november 1979 — 11-41 Micro keyswitches Modern Electronic equipment has long been in het fore front with regard to miniaturi- sation, conserving both space and energy. LvNvT MICRO This in turn has made control of access to circuitry even more essential in view of the high portability of equipment. Whereas miniature switches are available for such applications, lock cylinders have tended to remain of much larger dimensions causing difficulty in mounting (space problem) and having low security value. A considerable uplift in quality and security is now available in the new MICRO KABA Locking Cylinder. The internationally well proven advantages of the Kaba design are packed into a tiny 12 mm diameter cylinder operated by a key that can be inserted either way up. Eight pairs of tumblers offer over 10,000 key combinations. High grade brass and nickel silver precision engineering and the well tested security of the Kaba design over several decades, give improved functioning and long life. The universal cross-shaped pro- file of the MICRO KABA cylinder makes it possible to achieve secure assembly into switch housings. The range includes versions with one or two key withdrawal positions. Micro Kaba is not only suitable for electrical key switch applications but also for general use in original equipment where small size is essential. This opens up new possibilities with the use of a tiny lock having big security features. Kaba Locks Limited, Woodward Road, Howden Industrial Estate, Tiverton, Devon EX 16 5HW, Tel.: Tiverton (08842) 56464, Telex: 42564. (1305 M) Single board microcomputer Fairchild have recently launched their 'Spark-16' microcomputer boards in the UK. The heart of this very powerful microcom- puter is Fairchild's recently introduced 9440 Microflame' CPU, a 16-bit, 10-12 MHz bipolar microprocessor. Assembled on a board measuring eight inches by ten inches it is suitable for applications requiring input/ output capability or for use as a basis for more complex systems. The main features of the 'Spark-16' microcomputer are 8K bytes of dynamic RAM, 4 K bytes of PROM, memory control with direct memory access capability. All input and output lines are TTL-compatible. The serial port features a switch for selecting either RS232C or 20 mA current loop operation. A total of thirteen data rates, between 50 and 9600 baud, are also switch selectable. Memory and I/O expansion can be achieved via an SI 00 size edge connector. The 4 K byte on-board PROM can be supplied with 'Firebug', as a resident program. This is an interactive assembler, debugger, editor and monitor designed for program generation in assembler language and evaluation of the 9440 'Microflame' system. 'Baby BASIC' is available in PROM as an option. The Spark-16' contains 50 basic instruction types for a total of 2192 different instructions with eight addressing modes. Fairchild Camera and Instrument (UK) Ltd., 230 High Street, Potters Bar, Herts, EN6 5BU, Telephone: Potters Bar (0707) 51111. (1309 M) Based on the case designed for CSC's series of handheld frequency counters, the case measures 3 x 6 x 1 % inches (76x 152 x 38 mm), and comes complete with assembly screws.a screw-in antenna connector, a red transparent plastic front panel, a subminiature jack preconnected to a battery snap connec- tor, and a battery compartment cover. The front panel provides sufficient space for key- boards, speakers, microphones or controls. 0 0 // Customer specified colours can be provided for orders of 1 000 units or more. Continental Specialties Corporation, Shire Hill Industrial Estate, Saffron Walden, Essex CB1 1 3AQ, Telephone: Saffron Walden (0799) 21682. Plastic case for handheld electronic products New from Continental Specialties Corpor- ation is a grey plastic case specifically de- signed for small, portable electronic products such as handheld calculators, counters, remote-control units, communication devices, portable meters, benchtop projects and telephone accessories. (1308 M) 11-42 — elektor november 1979 market Soldering on A new soldering station is now being pro- duced by Antex (Electronics) Limited. The TCSU2 has a temperature range of 270 c C -430°C with a visual indication of the soldering iron tip temperature. Four square LEDs, as shown in the photograph, will light showing tip temperatures of 270°, 300 s , 330° or 360° C. The new station will be supplied with the XTC - 50 watt or the CTC • 40 watt miniature soldering iron, both irons being fitted with a thermocouple sensor and operating on the fully earthed 24 volts supply from the soldering station. The irons are supplied complete with 3 long life iron-coated bits with tip sizes of 0.5 mm, 1 mm, 2.3 mm for the model CTC and 2.4 mm, 3.2 mm and 4.7 mm for the model XTC. Burn-proof silicone covered 5-core cable connects the thermocouple sensor in the tip of the iron with the electronic circuit of the soldering station. Zero voltage switching ensures the absence of transient spikes. Current leakage is negligable and the accuracy of temperature settings is about 2%. The mains, switch, light and fuseholder are all easily accessible at the front of the unit. An on off light shows when the iron in use has reached the required temperature. The circuit also incorporates a 'fail-safe' system to prevent excessively high temperatures. Antex ( electronics ) Limited, Mayflower House, Plymouth, Devon, Telephone 0752 • 67377. (1302 M) New digital multimeters added to the TM500 range The latest entries to the Tektronix TM500 series of modular instruments are two 3’/2 digit Multimeters, the DM505 and the DM502A. The DM505 intended for applications where low capital cost is important and provides the five basic measurements of DC voltage and current, AC voltage and current, and resist- ance in two ranges, (high and low). With the high/low resistance feature, the low setting is used for in-circuit measurements where it is important not to forward bias diode junctions. The maximum imposed voltage is 0.2 V in the low resistance range, and 2.0 V in the high range, the latter being useful where actual measurement on diode junctions is needed. Extra features on the DM502A are dBV and dBM measurements, a fast-response tem- perature range of — 55° C to +200° C, true RMS readings, and autoranging for volts, ohms and dB measurements. The DM502A's combination of autoranging and dB measurements make it an excellent choice for communications applications. In addition to the convenience of autoranging, the DM502A provides direct readout on the display of the total dB reading. There is no need for the mental addition of a scale setting to the display readout. This saves time and eliminates a potential source of error. Pushbutton selection of all functions and ranges plus easy-to-read half-inch LED dis- play digits make the DM505 and DM502A fast and easy to use. A choice of front panel or rear connector inputs is selctable by push- button, a feature which allows easy inter- connection with other TM500 instruments while retaining the ability to revert to exter- nal measurements when needed. Tektronix U.K. Ltd., Beaverton House, P. O. Box 69 Harpenden, Hertfordshire, Tel.: Harpenden 63141. (1284 M) New silicone encapsulants for electronics A new range of Kommerling 2-component silicone compounds, primarily intended for encapsulation and sealing in the electrical and electronics industries is now available in the U.K. through L.B. Chemicals Ltd. The materials are available in soft and medium grades, the soft grades being used where protection from vibration is required while the medium grade is a general purpose coating and encapsulating product. Advantages of the products include widely variable processing times by simple alteration of catalyst ratios, low shrinkage, easy peal-off for repair work and excellent electrical and moisture protection. The materials cure at ambient temperatures without evolution of heat and are thus suit- able for treatment of delicate assemblies which would be damaged by elevated tem- po' itures. L. B. Chemicals L td. , 216 Moss Lane. Bramhall, Cheshire, Tel: 061 440-9559. (1285 M) market elektor november 1 979 — 1 1-43 Between sensor and processor With microprocessors in mind, Siemens has designed a new mos device which converts analogue sensor signals into digital pulses. Designated SAB 3060, this analogue-to-digital converter has a standard 8-bit word length. One of the principal features of this new device is an integrated capacitor network to achieve a very high conversion linearity. The SAB 3060 compares each incoming analogue signal eight times with a sub-divided reference voltage. In each case, it is determined whether the measured value is larger or smaller than the particular reference value. First, half the reference voltage (V re f /2) is offered, followed t>V V re f/4, V re f/8 and so on until the eighth value is reached (V re f/256). By means of this successive approximation, the original ana- logue value is directly converted into a digital 8-bit word. Originally, resistor networks were used for the approximation process. Capacitive cells, how- ever, are more suitable for mos technology. Parasitic capacitances capable of falsifying the result can be suppressed by judicious arrange- ment of capacitors. Additional driver ampli- fiers are not required, as the voltage sources for the measured and reference values are only capacitively loaded. The SAB 3060 is a 18-pin d.i.p., the supply voltages are +5 V and +12 V. The measuring range extends from 0 to +8 V, the reference range from 1 V to 8 V. Special care was taken to achieve a linearity of ± 2 l.s.b. (least signifi- cant bit), in other words ± 0,08% of the range final value. The precision is ± 1 l.s.b. The SAB 3060 has as its core a charge equal- isation converter, which is to be seen in the circuit layout (see photo) as a central capaci- tance field with a total of 256 m.o.s. capaci- tors. Measured and reference values are com- pared in the comparator, from where the digital 8-bit serial information is passed to the converter register. By way of result and out- put registers and an output driver, the digital value is then presented in parallel form. Out- put and converter controls are also integrated. Around 1000 transistors and other elements are on 7.5 mm 2 of silicon. The SAB 3060 is intended as a link between sensors and microprocessors, e.g. when sensors acting as the 'five senses' need to supply direct information on a variety of status such as speed, temperature spacing, length or quantity. When supplied with such values in digital form, microprocessors or microcomputers can subsequently issue instructions for analogue processes by way of actuators. These actuators close the loop between automatic detection and a specific response. Siemens Limited , Siemens House, Windmill Road, SUN BUR Y-on-THAMES, Middlesex TW16 7HS. Tel: (09327) 85691. (1281 M) Liquid crystal displays A new range of liquid crystal displays from Industrial Electronic Engineers, and desig- nated IEE-POLARIS. are now available in 2 models: high performance for use in rela- tively severe environments, and economy for use in mild environments. Both models are available with either reflec- tive or translucent polarizers and come equipped with DIL strip connectors for ease in mounting to PC boards or standard sockets. The user can mount the display with connec- tors in such a fashion as to allow replacement of the display without removing the two con- nector strips from their fixed position. These LCDs feature: 314 to 8 digits, .350" to .700" character height, low 25 pW typical average power consumption, choice of 3 to 9 or 4.5 to 13.5 voltages, and temperatures of -10° to +55° C. I EE's LCDs, which are direct sunlight-read- able, can be displayed continuously for up to two years without battery change, and are compatible with available low power, low voltage, CMOS drive circuitry. The crystal material is environmentally tested for stab- ility and the package hermetically-sealed to assure a long life of greater than five years. Custom models are available using the cus- tomer's font or numeric style together with symbols, decimals, etc. IEE, 7740 Lemona Ave., Van Nuys, CA 91405, U.S.A., Tel.: (213) 787 0311. (1301 M) Floppy disc controller from GECS GEC Semiconductors have announced a single-board Universal Floppy Disc Controller called the iSBC-204. This is fully compatible with the new Intel iSBC-80or iSBC-86 single- board computers and with most single- density, soft-sectored standard and mini floppy disc drives. The iSBC-204 controls two drive surfaces. However, with the addition of a second Intel 8271 floppy disc controller, up to four drives can be supported. It has a direct memory access channel allowing single-board computers to process in parallel with disc transfer operations, and programmable track-to-track access, head settling and head-load times. The wide drive compatibility range of the iSBC-204 is achieved without compromising performance by program control specifying the operating characteristics. The controller can read, write, verify and search either single or multiple sectors and has on-board data separation logic performing standard FM encoding and decoding. The iSBC-204 can be mounted ina one-slot Intel iSBC system chassis or iSBC-604/614 cara cage and interface with the drive(s) on either low-cost flat ribbon cable or twisted- pair conductors with individually wired con- nectors. GEC Semiconductors Limited, East Lane Wembley, Middlesex HA9 7PP, Tel.: 01-9049303. 11-44 — elektor november 1979 market Component tester MTL Microtesting Limited have recently announced their appointment as sole UK distributor for Huntron Instruments, the manufacturers of the Huntron Tracker which is the first of a new generation of portable test equipment incorporating a new technique for detecting and isolating faulty components either 'in' or 'out' of circuit. The Tracker utilizes a scope display, two non-polar leads and three impedance ranges to test a broad range of solid state components such as integrated circuits, bipolar transistors, field effect transistors, diodes, LEDs, unijunctions, gate control switches, capacitors etc. Simple easy to understand 'scope images visualise the condition of a device under test, indicating 'shorts', 'open circuits' and leaks'. MTL Microtesting Limited, 115 The Butts Road, Alton, Hampshire, Telephone: Alton (0420) 88022. (1307 M) vanadium blades. Wide, narrow and Philips types are available. Plastic handles are an unusual feature in the hammer range and a totally secure patented connection between metal head and the handle is used. Cushion hand grips are fitted which, it is claimed, absorb virtually all shock. OK Machine & Tool (UK) Ltd., 48a The Avenue, Southampton, Hants SOI 2SY, Telephone: 0703 38966/7 (1303 M> Digital multimeter Recently announced by Telonic Berkeley UK is the Data Tech 3%-digit, six function digital multimeter produced by a division of the American Penril Corp. The Model 30LC has a basic DC accuracy of 0.1%. A large 0.5" high Liquid Crystal Display (LCD) is used for low power drain from four off-the-shelf, dispos- able, size D flashlight batteries. Either alkaline or zinc-carbon batteries may be used. When using alkaline batteries, up to 2400 hours of battery life from one set of batteries is poss- ible if measuring DC voltages and over 1300 hours with average use of all six functions. The Model 30LC uses a single DVM LSI chip for its analogue to digital conversion. Auto- matic zero and polarity are included. Func- tion and range can be selected by rotary switches. Functions include AC and DC voltage and current, resistance to 0.1 ohm resolution and a diode test feature. A low battery sensing circuit flashes LOW BAT symbol in the LCD display area when approximately 100 hours of operation remains prior to battery replace- ment. Batteries can be changed in less than one minute. When the input exceeds 1999 counts, overrange is indicated by the three least significant digits blanking while the most significant digit '1 ' stays on. The instru- ment is packaged in a high impact plastic case with metal top and bottom. Options include internal 10 amp current range, carrying case, RF probe, high voltage probe and demodulator probe. Telonic Berkeley UK, 2 Castle Hill Terrace, Maidenhead, Berkshire SL6 4 JR, Telephone: 0628 28057. (1304 M) New range of hand tools The new Profil 2000 range of hand tools from OK Machine & Tool (UK) Ltd is the result of extensive technical and ergonomic research. Apart from their appearance, which is uncon- ventional, the tools have other significant differences including sweat absorbing handles and total rustproofing to contribute towards comfort and durability. Initially the range comprises various types of pliers, screwdrivers and hammers suitable for electrical, mechanical and general engineering use. The pliers, made of high alloyed carbon- steel with hardened cutters, have unique sweat absorbing handles and are finished like theother tools in the range in black chromium plate. Several types are available including wire cutters and strippers as well as fine nose strippers. The screwdrivers have red PVC and black Cellidor padded handles with chrome int advertisement elektor november 1979 — UK 25 [TOTAL] Name ^ for your copies of Elektor It it evident that in your profession and/or hobby the design ideas published in Elektor are referred to time and time again. We are therefore now introducing this new cassette style binder to keep your copies of Elektor clean and in order. The chamfered corner of the cassette allows instant recognition of each months isme without the need to thumb through pages of previous months issues. No wires or fastenings are used so copies are easily removed and replaced and each cass ette will hold one year's volume of Elektor. Their smart appearance will look good on any laboratory shelf. "P /sfA A range of 3V2 digit LCD multimeters vl offering high precision and extended w battery life. All feature 0.5" LCD read-out /M with ‘battery low' warning, inputs protected against overloads and tran- nH sients, Auto- polarity, Auto-zero, rugged ABS cases and a full 1 -year warranty. The LMM-200 is a compact handheld multimeter with 0.5% basic accuracy and 15 different ranges. It measures voltage from 0. 1 mV to 500V, current from 0. 1 uA to 2 Amps, and resistance from 0.1 A to 2MA. The LMM-2001 is an identical instrument but with 0. 1 % basic accuracy. The LMM-100 has an adjustable handle, a 2,000 hour battery life and is ideally suited to field or bench use. It measures voltage from 0. 1 mV to 1 KV, current from 0. 1 uA to 2 Amps, and resistance from 0. 1 A to 20M A . 0,1 % basic accuracy. Lascar Electronics Ltd., Unit 1, Thomasin Road. Basildon. Essex Telephone No: Basildon (0268) 727383. To: Lascar Electronics, Unit 1, Thomasin Road. Basildon, Essex. Please send me Data LMM-100 £82.17 LMM-200 £41 34 LMM-2001 £52 84 TEST LEADS £2 53 i enclose cheaue/P O value There's a lot going on at Breadboard! Seventy exhibitors showing and selling everything that the hobby electronics enthusiast could want! Demonstrations of electronic organs — computer kits — audio gear. Radio Station S22 breadcasting throughout the show. See your voiceprint! . Get your own weather details direct from Tiros M! Test your reactions — and your strength. Careers in Electronics — get the advice and information that could start you off on a rewarding and interesting career. It's worth going to Breadboard! Royal Horticultural Halls Elverton Street Westminster London SW1 December 4 -8th 1979 Admission £1 (students 70p)