up-to-date electronics for lab and leisure Controlled fluorescent Compact SSB receiver. slektorjune 1982 - 6-03 the principles behind an SSB receiver This article provides the uninitiated with a crash course in transceiver technology in general. compact shortwave SSB receiver 6-16 A single side band (SSB) is normally associated with high cost and com- plexity. This does not always have to be the case! The use of 'direct conversion' techniques results in a compact low cost SSB receiver with excellent performance. electronic starter for fluorescent lights Speed up the ignition of fluorescent light tubes by installing an electronic starter! The tube strikes almost immediately, without any flickering. measuring AC waveform (K. Fitttta) . An insight into measuring techniques needed for fault finding in AC circuits. mobile aerials • Although some CB enthusiasts no longer have to be on the run, many still prefer to use a mobile rig. The article describes how a single aerial can serve both car radio and transceiver purposes. electronic dog whistle In order to keep the dog lovers of the community happy, here is our first dog circuit! Using the UAA 1 003 the 6502 housekeeper published in last month's issue will well and truly tell the time. 6-30 6-32 the 'Poly bus' Constructors who intend to make a complete polyphonic synthesiser, will be confronted with a complex wiring problem. A bus board has been designed, helping to keep the amount of wiring to the bare minimum. fluorescent light dimmer Usually fluorescent lightsare not compatible with existing dimmer circuits. This article explains how to modify fluorescent lights, so that they can be dimmed, and describes a circuit able to fully control them. An active insurance policy which is also a deterrent to would-be thieves! 6-42 6-47 from the 6502 to the 6809 The 6809 CPU manufactured by Motorola supersedes 6502. The beauty of this chip is that it can be implanted into existing 6502 systems. introducing DMOS power FETs • • New power FETs seem to be christened almost every day. Despite all their different names, they all have a great deal in common. This article takes a look at power FETs in general, paying special attention to the fast-switching DMOS branch of the family. solid state relay •••• Electronic relays have quite a lot of advantages over their conventional electro-mechanical counterparts. The solid state relay won't spark or wear out as quickly and perhaps more importantly, it can be built easily at home! missing link market . . . 6-62 6-62 EDITOR: P. Holmes UK EDITORIAL STAFF T. Day E. Rogans TECHNICAL EDITORIAL STAFF | J. Barendrecht G.H.K. Dam E. Krempelsauer G. Nachbar A. Nachtmann K. S.M. Wal raven elektor" 6-12 - elektor june 1982 Look, no hands! For the first time in its 88-year history London's Tower Bridge was raised by remote control today — using an ordi- nary telephone some seven miles away. It was done using a unique interface unit and microcomputer which together form a device called an Information Transfer Module (ITM), a new techno- logy developed by the Brighton-based company ITT Business Systems. The ITM integrates and extends the func- tions of business communications pro- ducts such as telephones, telexes, VDUs and word processors. The telephone used was connected through the public telephone network to an ITM at Tower Bridge. The ITM converted the signal and passed it via an actuator to move a lever which raised the bridge. The lever is usually moved by hand. (785 S) Direct broadcasting by satellite Following the Home Secretary's an- nouncement in the House of Commons on the go-ahead for direct broadcasting by satellite, British Aerospace, Marconi and British Telecom declared their intention of forming a joint company. United Satellites, to provide Britain's first national broadcasting and tele- communications satellite system. The project could be operating by the end of 1985. As Halley's comet is due to reappear the same time, the satellite system may well be called Halley 1. Two satellites will go into orbit: one for broadcasting purposes and the other in case of temporaly system failure. A third satellite on the ground would replace a satellite that failed altogether. Two TV channels will be transmitted, one of which will be paid for by viewers' subscriptions. The signals will be broad- cast in coded form and only those who buy a decoder will be able to receive the transmissions. In addition, the satellite will serve to improve telecommuni- cations. United Satellites have already inves- tigated potential markets, and also the technical and operational means to meet broadcasting and telecommunications requirements on both a short and a long term basis. The company's preliminary work has already involved liaison with Government departments and with the broadcasting industry. Further to this, British Telecom have advised on the development of national and inter- national satellite telecommunications services from the mid-1980s. The next phase will call for further discussions with the broadcasting organ- isations to define technical requirements and the terms on which satellite ca- pacity will be able to be offered for direct broadcasting television services. The requirements for satellite tele- communications services will also be specified in agreement with British Telecom. The Halley 1 project will not only be the first British national system for direct broadcasts by satellite, but it will also promote British satellite systems and services on an extensive scale throughout the world. 1767 SI Four-year old computer relegated to Science Museum Computer developments are progressing at such an alarming rate these days that they are virtually impossible to keep up with. It is disconcerting, to say the least, when brand new systems, each in itself a remarkable feat of engineering and technology, are superseded by better, faster machines almost before they are given a chance to prove their worth. The amount of money involved is mind- bogling. Just recently an IBM System 370/148 computer, which originally cost nearly a quarter of a million pounds, was rel- egated to the London Science Museum! Its former owner. Gulf Oil Corporation, acquired the computer in 1978 (a century ago in computer terms) for its Copenhagen Accounting Centre, where it was used to administrate the Gulf credit card scheme for the whole of Scandinavia. It has now been replaced by a computer which operates at three times the speed and has a four times larger memory capacity. The Science Museum is planning to add the computer to their collection of historic computer hardware and is hoping to eventually have it in running order . . . Where will it all end?? (784 S) 3 * SSB * SSB * SSB * SSB * SSB * SSB * SSB « SSB * SSB « SSB « SSB « SSB « SSB « SSB « SSB » SSB « SSB elektor june 1982 - 6-13 Modulation In principle, a transmitter could merely consist of an oscillator producing a fairly high frequency signal. The signal is then transmitted 'on the air' by way of an aerial. As figure la shows, however, most transmitters are a little more complex than that and contain several com- ponents in addition to the oscillator. Let's look at the block diagram in figure la. An oscillator signal with a frequency of. say. 4 MHz enters an die principles behind an SSB reeeiwr in which to transfer information. The result can be described as a series of RF smoke signals. The switch in figure la may be re- garded as the encoder of a CW (Continu- ous Wave) transmitter. As a matter of fact, the wave is not continuous at all, but is chopped up into little pieces by the encoder. This form of modulation is sometimes referred to as pulse modu- lation. Other forms of modulation also exist, one of the better known being illus- trated in figure 1b. Here the switch has been substituted for a voltage control circuit, which varies the output voltage of the power amplifier in proportion to a microphone signal. In the block diagram in figure 1b a 1kHz signal has been selected as the modulation frequency and the amplitude (or envel- ope) of the output signal can be clearly seen to assume the waveform of the 1 kHz signal. As many will have guessed, this is known as amplitude modulation (AM). As the signal is modulated sym- metrically, a symmetrical output signal is obtained with a peak value that is twice that of the unmodulated carrier a crash course in transceiver technology It is all very well saying that SSB stands for 'single side band', but what does it really mean? This article not only explains terms like 'side band' and 'carrier wave', but also provides the uninitiated with an introduction to transceiver technology in general. Elsewhere in this issue readers are invited to construct their own SSB receiver, but before they get up to their ears in solder and components they might like to take a look 'behind the scenes' and see what they are about to build! amplifier where it is boosted from a couple of mW to 1 00 W, for instance. It then passes through a filter which 'cleans it up' by removing any un- desirable constituents (interference etc). The filter also makes sure that the impedance of the amplifier and the resonance of the aerial are well matched. The signal that is effectively trans- mitted is known as a carrier wave. Even though an adequate receiver is able to pick this up, the carrier wave alone is unintelligible. To allow information to be transferred from a transmitter to a receiver, relevant data will somehow or other have to be added to the carrier wave, it is, in fact, modulated! As its name suggests, a carrier wave serves to carry information. The easiest way in which to modulate the carrier wave is to use the switch shown in figure la. This enables the transmitted carrier wave to be inter- rupted at regular intervals and, provided both the transmitter and the receiver stick to some code (such as the morse code!) this is an effective method Another well-known type of modu- lation is frequency modulation or FM. There is no need to go into details here, but the basic principle is shown in figure 1c. This time the frequency of the carrier wave is modulated, instead of the amplitude. The microphone signal is converted into a control voltage which serves to shift the frequency of the oscillator slightly up and down. The amplitude of the output signal can be seen to remain quite constant. Of course, there are other types of modulation systems apart from the ones shown in figure 1. FM related systems include narrow-band FM and phase modulation (PM), whereas DSSC and SSB, for instance, belong to the AM family. It is the latter two that we’re really interested in. Side-bands DSSC and SSB modulation systems have been around for some time. The basic principles behind them were discovered quite a while ago and are as follows: If an AM transmitter like the one in figure 1b modulates at an audio fre- quency of 1 kHz, a carrier wave of 4 MHz (= 4000 kHz) and two side bands are produced (harmonics), one at 3999 kHz and the other at 4001 kHz. Figure 2a shows what such signals look like on the screen of a spectrum analyser. The two side bands are the mirror image of each other and contain exactly the same information. The carrier wave itself does not provide any information but, as indicated in figure 2a, it does absorb most of the trans- mission energy. In the early days of radio someone had the bright idea to suppress the carrier wave altogether and to channel the transmission energy into the signal carrying side bands. This B » SSB * SSB » SSB » SSB » SSB « SSB « SSB » SSB « SSB « SSB * SSB « SSB « SSB * SSB * SSB * SSB * SSB 6-14 - elektor june 1982 method is known as DSSC, which stands for Double Side band Suppressed Carrier. The result illustrated in figure 2b is that the effective (information carrier) output power is double the amount produced in AM. One step further in this direction leads us to SSB (single side band). Since both side bands are identical one can be suppressed without causing any infor- mation to be lost. Figure 2c shows how in SSB the effective output is again double that produced in double side band systems. When figures 2c and 2a are compared, it is quite obvious that the transmission power is handled a lot more efficiently in SSB than in AMI SSB: the pros and cons Not surprisingly, SSB is the most frequently used modulation system on short wave. Hams operating within this frequency range rarely use anything else. SSB not only gives a better performance and provides the transmitter with much more power, but it also has the ad- vantage that the bandwidth need only be half that required for AM purposes. At a maximum audio frequency of, say, 3000 Hz (sufficient for speech trans- mission) the side bands will extend beyond (below and above) the carrier wave frequency up to 3000 Hz, re- sulting in a bandwidth of 6 kHz. The single side band of an SSB signal only occupies 3 kHz of the transmission range. This means that twice as many transmitters can be squeezed into a certain waveband. In practice, the number is even higher, as no carrier interference can be produced between two neighbouring stations now that the carrier wave has been suppressed. Unfortunately, there are also a couple of disadvantages associated with SSB. For one thing an SSB transmitter is much more complicated and expensive to build than an AM set. But the worst drawback is encountered at the re- ceiving end. As the receiver has to tune into a single side band, its frequency stability has to meet far higher stan- dards than an AM receiver. In short, anyone wishing to try a hand at building the SSB set described elsewhere in this issue should read the instructions very carefully. The receiver The receiver converts information broadcast by the transmitter into a form that listeners can understand. To be able to do this it has to meet two require- ments: First of all, it must be able to select the desired station from a huge quantity of other signals 'on the air'. Next, it must glean the relevant infor- mation from the signal and convert it into an acoustic signal. AM listeners can make do with a crystal receiver. This comprises a tunable LC Figure 1 . Various types of moduletion ere used in transmitters. Some of the best known types are shown here: pulse modulation or c.w. Hal. amplitude moduletion (1b) and frequency modulation (1c). circuit to select the required signal, a diode to recuperate the audio frequency information from the radio frequency signal and finally, headphones to make the modulation audible. If a certain amount of selectivity and sensitivity are required, however, the receiver will have to include a number of selection circuits and the signal will have to be RF amplified. That is why a straightforward AM receiver usually looks like the one in figure 3, a super- heterodyne set. The input signal is mixed with that of an oscillator. The oscillator is adjusted to a slightly higher frequency than the input signal and is tuned together with the input circuit. As a result, the difference between the input and the oscillator signals remains constant (455 kHz) over the entire tuning range of the receiver and the differential signal (the intermediate frequency or IF signal) will be available at the output of the mixer. Now the signal can be extensively filtered to provide the required selec- tivity, because, contrary to the input circuit, the IF signal is at a constant frequency, so that the filter circuits no longer have to be tuned for each particular station. After the necessary filtering and amplification, the IF signal - i * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB » SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB » Is detected and AF amplified. The modulation is then audible in the loudspeaker. So much for AM receivers. In principle, an SSB set closely resembles its AM counterpart, but as the signals involved here are extremely narrow-banded, far better selectivity is required. The straightforward circuit in figure 3 is hardly likely to give satisfactory results. Nine times out of ten, the block diagram of an SSB receiver will look like the one in figure 4. As the circuit has two mixers and two different IF fre- quencies, it is known as a 'double superhet'. This is where we come across an essen- tial difference between SSB and AM systems. A carrier wave is needed to detect the IF signal but, as the article pointed out earlier, this is not present in an SSB signal. So somehow or other the carrier wave will have to be gener- ated in the receiver and added to the signal. As a rule, the wave is added just before the signal is detected, with the aid of a BFO (Beat Frequency Oscil- lator). By tuning the BFO very carefully to exactly the same frequency as that of the (imaginary) carrier wave, the orig- inal modulation frequency (1 kHz) can be recuperated by the detector. This procedure calls for great frequency stability throughout the receiver and especially in the BFO, as the slightest fluctuation leads to a frequency shift in the AF signal. Tuning the BFO requires considerable care and precision. By mistuning the BFO the pitch of any voice can be altered to sound like Donald Duck at one end of the scale and Ivan Rebroff at the other - with a vast vocal repertoire between the two extremes! All in all, SSB receivers are quite difficult to operate and demand a lot of patience and experience. But a radio enthusiast's greatest asset is a steady hand! M QQR * q«;r * RSR . RSR * SSR * SSB * SSB * SSB * SSB * SSB * S 6-16 — elektc 1982 a direct conversion circuit achieving 'superhet' quality A single side band (SSB) is normally associated with high cost and complexity. This does not always have to be the case! With direct conversion, the RF signal is converted directly into AF without producing an intermediate frequency (IF). The use of such a technique results in a compact low cost SSB receiver with excellent performance. The cost of construction bears no relation to its quality! While not being exactly simple, we are sure that this project will bring the world of 20 metres into your home, without the expense associated with commercially available equipment. Communication receivers are invariably extremely complex and expensive, but in some cases these facts alone, sur- prisingly, do not guarantee good per- formance. When compared to a number of commercial receivers, the Elektor SSB performed rather well, and actually beat some of them 'hands down'! The article 'The principles behind an SSB receiver', elsewhere in this issue, may suggest that building an SSB requires quite a lot of skill and knowl- edge. The simplified block diagram published in the theoretical article de- picted an average receiver which is in fact very difficult to build. However the circuit diagram of commercial communications equipment will prob- ably put you off completely. The use of complex superhet type circuits is not the only way to design an SSB. A more straightforward approach is to use 'direct conversion'. This principle allows much simpler receivers to be built that still achieve a high performance. The main difference between a direct conversion receiver and a superhet is the fact that the first type does not produce an intermediate frequency (IF). Like the superhet, its input and oscil- lator are mixed, but inasmuch as the oscillator frequency is equal to the input signal their sum and differential products supplied by the mixer are restricted to audio frequency infor- I mation. The audio frequency (AF) part of the receiver (section LPF - low pass filter) filters the signal in order to obtain good selectivity. The oscillator also functions as a Beat Frequency Oscillator (BFO). It has the same frequency as the input signal. From the constructional point of view, the oscillator is one of the difficult parts of the circuit, since a high standard in stability is essential. The main advantage of a direct conver- sion receiver over a superhet design are: • Straightforward and compact con- struction. • Easy to align and control. • Because the oscillator and input signal frequency are identical, pro- blems relating to image frequencies are eliminated. Only the harmonics and sub- harmonics of the oscillator frequency could cause some trouble, but the superhet has the same problem anywayl • Low cost, due to its straightforward approach to construction and design. The filtering necessary for good selec- tivity is applied in the AF stage saving on cost alround. An RF filter for the same bandwidth as the AF one used in this design (-8dB at 3 kHz . . . -60 dB at 5.5 kHz) would cost at least forty pounds! The direct conversion receiver does naturally suffer from some disadvan- tages: • It is susceptible to audio image frequency interference, thereby re- ceiving both side bands instead of one. • The operational range is slightly less than that of a superhet because the mixer stage could work as an AM detector, if the specified input signal strength is exceeded. Versatility The receiver described is suitable for the 20 metre amateur band ranging from 14.00 to 14.35 MHz. This fre- quency range was chosen because it is used frequently and therefore the most interesting bandwidth to listen to. For quite some time now, we have been under the influence of sunspot activity which makes it possible for the 20 metre band to be in use 24 hours a day. So starting off with the 20 metre band get With the addition of converters, the receiver is an ideal starting point from which to build a multiwave band communications receiver. This is partly thanks to its tuning (approximately 0.5 MHz). All the amateur bands with the exception of the 28 . . . 29.7 MHz band can be received easily by using a single converter for each band. The circuit A block diagram of the receiver in its final form is shown in figure 2. I SSB * SSB * SSB * SSB * SSB * > 4 SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB compact shortnme SSB receiver Although this is not as simple as the one shown in figure 1 , it does contain everything necessary. As a matter of interest, the block diagram of an average superhet SSB would probably take up five to six pages. Don't panic, it really is not as bad as you think! The aerial signal first reaches a band- pass filter (BPF1) which determines the tuning range. This signal than has to pass an RF amplification stage and a second filter before it reaches the mixer. The signal from the tuning oscillator is also fed to the mixer via a buffer stage. The low frequency output signal of the mixer is thoroughly filtered by means of two low-pass filters (LPF1 and 2). An AF amplifier is situated between both of these filters and connected to a noise limiter. LPF2 is followed by yet another AF amplifier. This is an auto- matic gain control, used to limit the input level to the mixer and therefore protect the input stages of the receiver from excessive input voltages. A straightforward audio output amplifier completes the diagram. Figure 3 shows the complete circuit, as opposed to block, diagram of the Elektor SSB. Take care to place figures 2 and 3 side by side as both are useful in explaining the workings of the circuit. BPF1 is the input stage made up of LI, Cl , and C52. The tuning range achieved because of this filter network is approxi- mately 500 kHz (from 14to 14.5 MHz). That is sufficient to cover the 20 metre band, without overlapping into the 19 metre band. The dual gate MOSFET (T1) wears a coat of many colours: a pre-amp for the input; a buffer between the oscillator and aerial (to eliminate feedback); an active part of the AGC. Even then it still is not overworked! BPF2 is formed by the network con- sisting of L4 . . . L7 and C6...C13. This is a rather complex filter having a width of approximately 3 MHz with a flat response within the 20 metre band. This all helps to achieve good 'mechanical' stability (sensitivity to mechanical vibrations). The next part is the mixer (T2). The principle of this single passive mixer is shown in figure 4. This ensures that nothing at all is fed to the output when there is no RF signal. It also makes sure that only the input signal, and not the oscillator one, is fed to the output. Transistor T2 is also a very versatile dual gate MOSFET. A high voltage level from the oscillator is required for the mixer stage to switch on and off. To ensure a high standard of frequency stability for the oscillator again a good quality dual gate MOSFET BF900 (T3) was used. This stage is a version of a 'Clapp' oscillator, which has in the past proved itself to be very stable. Tuning is carried out with a varicap diode (D4). These diodes need a control voltage, which in this case is supplied by the regulator IC1. The control Figure 1. A direct conversion receiver is much more straightforward than a 'super fet'. The input signal and oscillator frequency are identical therefore no intermediate frequency is produced. The oscillator also serves a BFO. Figure 2. Complete block diagram of the Elektor SSB receiver. SSR . SSR * SSR * SKR * SSR * SSR * SSR * SSR * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSf SSB * SSB » SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB * SSB elektor june 1982 - 6-19 voltage level is determined by PI. This is a 10 turn potentiometer, eliminating the need for gearing in order to achieve a 'slow tuning dial'. Between the oscillator and mixer there is a buffer stage (T4). Now for the AF part of the receiver. A fairly straightforward low-pass filter LPF1 is positioned directly after the mixer, consisting of L8, Cl 5, C29 and C30. It has a high cutoff frequency (about 10 kHz), because otherwise the noise limiter would not be effective. The noise limiter is basically an ordinary 'diode cutter' (D2 and D3), which also forms part of the AF amplifier (T5and T6). The signal is amplified and filtered once more by a second low-pass filter (LPF2) made up of L9, L10 and C33 . . . C37. This removes any com- ponents of the signal which are above 3 kHz (approximately 66dBs per oc- tave). The automatic gain control (AGC) consists of T7 and its surrounding components. A single transistor detector (T7) rectifies part of the AF output signal of IC2 converting it in to DC. The level of DC is proportional to the strength of the AF signal. This DC voltage is then fed back to the second gate (G2) of the RF stage (T1). If the base/emitter threshold of T7 is exceeded (with strong input signals), the gate 2 source voltage of T1 will automatically drop, thus decreasing its gain. The attack is fast and the decay is rather slow, in order to avoid the annoying 'breathing effect' (pulsations) that can occur with some AGCs. Finally there is IC3. This is an audio amplifier able to directly drive a loud- speaker, requiring only a minimum of external components. The volume is regulated by potentiometer P2. Construction The printed circuit board of the SSB receiver can be cut into two parts if required. The RF and AF sections are separated. Figure 5 shows the RF section, which is also indicated as a circuit diagram in figure 3a. The AF part of the board illustrated in figure 6 corresponds to the circuit diagram in figure 3b. With the exception of the transformer, all the power supply com- ponents can be mounted onto the AF board. The choice of whether the printed circuit board is left in one piece or sawn in two is left to the constructor. In order to achieve a reasonably compact final product, the Elektor prototype, separate the boards and mount them on top of each other. Should you separate them, the boards clearly in- dicate the corresponding interconnec- tions, such as AF signal, AGC voltage, supply and so on. The connection points leading to the off-board com- ponents are also easily recognisable. Remember to connect choke LI 2 when linking the supply voltage between the AF and RF sections. No provision was made for mounting LI 2 onto any of the boards! Both the AF and RF parts are double- sided printed circuit boards. The com- ponent overlay side is really one large copper surface that functions as an earth and screen. Consequently all com- ponents that need to be earthed must be soldered on this side. The holes for the other components have insulation rings. The connections for the FET BF900 and the double varicap KV1236Z are shown in the circuit diagram of figure 3. Take extra care when mounting these. The trimming capacitors C52 and C53 are equipped with three legs, of which only two are used, so be careful to connect them the correct way round otherwise the complete circuit could be shorted out. Now for the coils. Constructors who are not particularly fond of winding these things themselves are probably now going to start worrying! Luckily most of them are standard ready-made chokes. However, not only is it essential to buy the correct coils but also to mount them in the right place. Careless- ness here will defeat the entire object of the exercise. LI and L2 cannot be purchased as ready-made coils and therefore must be wound from scratch. The coils are wound onto ring cores of the type T50-6. LI requires 21 windings of 0.4 mm enamelled copper wire, with a tap exactly on the last but one winding away from the earth connection. L2 needs 12 windings of 0.6 mm enamelled copper wire with a tap 2 windings from the earth connection. Try to ensure the windings cover the complete surface of the ring core. Once the receiver is com- pleted and aligned, it is advisable to glue both coils onto the printed circuit The amplifier, mixer and oscillator sections of the RF section must be screened from each other by mounting tin or copper partitions. The printed circuit board and circuit diagram clearly indicate where they have to be mounted, (see photo 2). We also suggest you try to screen the top of each compartment by a tin cover, to be absolutely sure that the RF amplifier, mixer, and oscillator will be restricted to 'minding their own business'. All possibilities of feedback from the oscillator to the aerial have to be avoided, because this can cause hum and microphony. CCR * CCR * CQR * SSR * RSR * SSB * SSB * SSB * SSB * SSB * SSB *_SS i * SSB * SSB * SSB » SSB « SSB » SSB » SSB * SSB « SSB » SSB « SSB * SSB « SSB ♦ SSB « SSB * SSB « SSB « 3 6-20 - elektor june 1982 A metal case for the housing is best. A cut costs here, as a bad speaker not only screwdriver, until the oscillator fre- plastic case will also do, but then the reduces the intelligibility of the output quency is 14.36 MHz. compartments of the RF section have signals, but can in some cases eliminate The aerial now needs to be connected, to be separated altogether, each one them altogether. Turn PI to its mid position, in other being airtight. The best results are words to the middle of the 20 metre obtained by placing pieces of foam Alignment band, and adjust C52 very slowly to rubber between the boards and the sides Aligning the receiver does not require give a maximum AGC voltage level, and bottom of the case. Leave the any special skill; the procedure is really Bear in mind that it takes the AGC interconnection between the RF, AF quite straightforward. Asa good starting some time to reach its nominal value. If sections and the other components till point, first set the trimming capacitor you are in doubt about the accuracy of last. C52 to its mid position (approximately your setting, then just turn C52 back to Remember when choosing a suitable 10 pF) and C53 to maximum. Now to its minimum setting and start again, case that no provision has been made to be absolutely correct, you should Constructors without a frequency coun- mount components, such as the mains position a multimeter between the ter can, as an interim measure, carry out transformer, aerial socket and so on, wiper connection of PI and ground, the following procedure: turn C53 until onto the printed circuit boards. then adjust PI until +8V is read, a 'Donald Duck-like' voice is heard, The loudspeaker should be inserted into Fortunately the voltage supplied by IC1 after having first connected the aerial, a separate housing, again to prevent is +8 V so all that is actually required is Then continue to turn it still further any undesirable feedback. The loud- for PI to be set to maximum. Connect a until morse signals are heard. PI is set speaker should be of reasonable quality, frequency counter to gate 1 of T2 by to its mid position, and C52 is adjusted with a frequency response range of means of a high impedance probe. Turn as previously described, in order to 200 Hz . . . 3000 Hz. It is not wise to C53, with the aid of a plastic trimming achieve the maximum AGC voltage. "MK-T; f!f?U : Wt* -. fi-J : WK'i : I : «!»^ : : ITx'-. : nr* r SSB * SSB * SSB * SSB « SSB » SSB « SSB » SSB » SSB « SSB » SSB * SSB * SSB » SSB « SSB » SSB » SSB « SSB elektor june 1982 - 6-21 Resistors: R1,R4 = 1 M R2.R9 = 100 fl R3.R23 = 1 k R5.R22.R24 = 2k2 R6.R12,R29 ■ 100 k R7 = 4k 7 R8.R27 = 220 k R10.R11 - 120k R13,R32,R33 = 470 SI R 1 4 = 220 SI R 1 5 = 27 k R16- 150k R17.R30 * 6k8 R18 = 3k 9 R19 = 33 k R20 ■ 1k8 R21 * 82 k R25.R26 = 39 k R28 = 56 n R31 = ion PI = 1 k 10 turns P2 = 2k2 log. Capacitors: Cl - 39 p C2.C3.C 1 6,C 1 9.C20.C26 = 22 n ceramic C4 = 47 p C5 = 220 p C6.C7 = 4p7 C8 = 1 0 P C9 " 1 n MKM C10.C14 = 100 p C11 =68p C12.C13 - 22 p C15.C55 = 10 n ceramic C17.C21 = IOOm/IOV C18 = 100 p/4 V C22 = 1 m/10 V tantalum C23 = 22 p/16 V tantalum C24 = 68 p temperature coefficient 0 C25 - 4p7 temperature coefficient 0 C27 * 560 n MKM C28 M 330 n MKM C29 = 33 n MKM C30 = 680 n M KM C31.C48 . . . C51 "47 n MKM C32 = 2m 2/40 V C33.C37 = 1 20 n MKM C34.C36 = 10 n MKM C35 = 180 n MKM C38.C40 = 1 m/IOV C39.C43.C46 “ 10m/16 V tantalum C41 = 1 m/16 V C42 * 47 m/10V C44 - 100 n MKM C45 = 470 M/10V C47 = 1000 m/35 V C52 ■ 20 p trimming capacitor C53 - 60 p trimming capacitor C54- 150 n MKM attention: C32.C40.C41 and C42 must be soldered vertically! Coils: LI * 21 turns of 0.4 mm enamelled copper L2 = 14 turns of 0.6 mm enamelled copper wire, tap at 2 turns from ground connection L3 = 3.3 mH L4 « 1 mH L5.L7 = 4.7 mH TOKO L6 = 2.2 mH L8.L9.L10 = 47 mH L1 1 = ferrite bead with 4 turns of 0.3 mm enamelled copper wire L12 ■ 4.7 mH (Toko) Semiconductors: D1.D2.D3 = 1N4148 D4 * KV 1236Z (Toko) T1.T2.T3 • BF 900 T4 - BF 451 T5 « BF 494 T6.T7 = BC 550C IC1 = 78L08 IC2 = LF 356 IC3 * LM 386 IC4 = 7812 B1 = B40C500 (round version) Trl = mains transformer LS * loudspeaker 8 ... 25 0/1 W Listening A few metres of wire strategically placed is sufficient for the aerial. A genuine aerial for the 20 metre band is a vertical rod approximately 5 m in length. Constructors who are new-comers to this particular field of electronics may need some time to get used to the alignment procedures, but, don't worry, test signals to try out the receiver are available in abundance. As stated before, whenever you switch on your receiver there will always be something to listen to. Most European amateurs will not be very active in the early hours of the morning (who is?), making it a good time to tune into South American or Asian stations. Because of the large number of morse transmissions in the 20 metre band, a course in morse will be useful and will obviously increase your listening pleasure. The quality and performance of this direct conversion receiver is really impressive. The sensitivity of the prototype proved to be no less than 0.1 5 pV with a signal to noise ratio of 10 dB. In practice this means that the receiver will stand up to any com- parison made with commercial, ready- made receivers. With this design the quality certainly does not correlate with the low cost (which can often be the case with do-it-yourself circuits circuits). One final point! Although the power consumption is not low, you are not going to notice any appreciable in- crease to your electricity bill. 40 mA requirement for average output volume levels is very high, implying that a portable version of the SSB receiver is possible. The easiest way to achieve this is to connect two 9 V power packs in series, giving a total voltage level, of 18 V. The combined power packs are connected in parallel to C47. Alkali-manganese power packs have a capacity of 500 mAH, giving sufficient power for 1 0 hours use. H rlCQR CCR * CQR » QCR * QCR * CCR * QCR . QQR SRR * SRR * SSR * SSB * SSB * SSB « SSF strike a light! Standard fluorescent light tubes, or to give them their scientific name 'low pressure gas discharge tubes', are not as straightforward as they seem. The normal opaque-looking tubes are in fact made of clear glass with an internal coating of fluorescent powder, and filled with mecury vapour (and a little argon). The vapour is under an ex- tremely low pressure (about 0.00001 absolute atmospheric ata.). By ionising the vapour under the influence of a powerful electrical field, a gas discharge (discharge of electrons) occurs. When an electric discharge passes through the mercury vapour a small quantity of visible light and a large amount of invisible ultra- violet light is generated. As already mentioned, the internal surface of the glass tube is coated in a thin layer of fluorescent powder. This converts the ultra-violet (short wave) radiation into visible (long wave) light with a continuous spectrum. The kind and colour of the light produced from discharge sources depends on the choice of powder. This is one reason why tubes are available in various colours and shades. A detailed look at this aspect can be found in the article on the fluorescent tube dimmer published elsewhere in this issue. A little argon (an inert gas) is added to the mercury vapour as a kind of catalyst, helping in the lighting-up or striking process. The start or strike voltage level required for the tube to ignite depends on the temperature of the gas. The lower the voltage the higher the temperature of the gas has to be. For this reason, filaments are mounted at either end of the tube. They pre-heat (pre-ionisation) the gas to promote ignition. Once the gas has started to discharge (tube lights), the glowing discharge (staying alight) can be main- tained by a relatively low voltage. As electronic starter for fluorescent lights When fluorescent light tubes are switched on they tend to light up in rather hesitant manner and the flickering effect is often irritating. Manufacturers have produced special, 'rapid starting' types, but these are more expensive and this makes them less popular. However even standard fluorescent tubes can speeded up by means of the electronic starter described here. The tube strikes almost immediately - an amazing effect, when you see it for the first time! a matter of interest, exceeding the glow discharge voltage by a large amount will cause a progressive reduction in the internal resistance of the tube. As the resistance decreases, the current flowing through the gas increases (current density), making some form of current limitation necessary. A choke, which dissipates very little power in the form of heat, is used as an induction coil together with a starter to produce a high ignition voltage. The choke also tends to sup- press RF interference caused by the gas discharge inside the tube (avoiding 1 mains borne interference). The starter not only serves to generate 2 an induction voltage, but also switches the current through the filaments. The most common type of starter includes a helium filled lamp containing a bimetal (thermal) electrode. This is shown in figure 2. The switch contacts are open when the device is quiescent. If the heiiun mains switch (SI) is closed, the entire mains voltage will be applied to the starter. This is enough to fire the helium lamp. The design structure of the starter only allows a low current (about 0.1 A) to flow. The heat produced by the gas discharge causes the bimetal switch to close. Asa result, a high current passes through the filaments which pre-heat the gas for a while. When the bimetal switch closes the helium lamp is internally short- circuited and so goes out. After a while, the temperature inside the lamp will drop to such a low level that the bi- metal switch will open. In resisting this c*? abrupt current cut-off, the choke gener- HI ates a momentary high voltage by self- IjE induction and feeds this to both ends of ^ the tube. Consequently, the fluorescent tube will light. When this happens, the voltage across the starter contacts is equal to the glow discharge voltage of the tube. This is too low to restrike the small helium 'lamp' in the starter again. F ' 9ur Since the bimetal switch remains open, J ^ te the starter will be inactive when the the tube is lit. A capacitor is connected in parallel to the starter to suppress any fi uori RF interference in the fluorescent tube. Unfortunately, the tube will not remain lit after the initial strike operation. More often than not, the temperature inside the tube will not be high enough to maintain the gas discharge straight away. The current could also be practi- cally zero when the bimetal switch is opened, in which case insufficient induction voltage is available. The tube will have to be struck several times for it to remain lit. As mechanical starters are rather slow, a visible delay occurs between every start session. This explains why the tube flashes on and off for a while after it is turned 'on'. To . A capacitor is connected in parallel irter to suppress any RF interference by the gas discharge inside the Capacitors: Cl = 15 n (seetex C2 = 100 n/630 V Semiconductors: D1 - ER 900 diac Thl “TIC 106D thyris avoid such flickering, we will have to make sure that the tube is sufficiently pre-heated and that the strike oper- ations take place in quick succession. This is exactly what the electronic starter does. The circuit Figure 3 shows the circuit diagram of the electronic starter. When dealing with its operation, let us assume that switch SI is closed and that the voltage at the anode (with respect to the cath- ode) of the thyristor is positive. As long as the tube is not lit, the voltage across the electronic starter contacts is equal to the mains voltage. When the voltage level in capacitor Cl, charged by way of the divider R1/R2, reaches the break-down level of the diac (about 30 V), the thyristor will conduct and Cl will discharge. A relatively high current will now flow through the filaments, and the choke. As a result, a magnetic field is created inside the choke. When the mains voltage assumes a negative polarity, the positive current will continue to flow through the choke, but only for a while. With the dying magnetic field the current drops to zero. At this point the thyristor will switch off and the maximum voltage (approxi- mately) of the negative half cycle of the mains will appear across C2/R4. Together, LI and C2 form a resonant circuit and this will now cause the capacitor to rapidly charge and dis- charge to approximately double the mains voltage. This high voltage level will readily ignite the tube. When the following positive half cycle of the mains supply arrives, the thyristor will once again turn on and repeat the procedure. The whole process is re- peated at a rate of 50 times per second. After several cycles the tube will be sufficiently hot to remain lit and so the voltage across the starter contacts will drop to the 'glow discharge' level. This is much too low for the diac and there- fore the thyristor to conduct. Conse- quently the electronic starter will re- main in a quiescent state. Figure 3. Only eight components are involved in the electronic enough to fit inside the case of the 'old'. 6-24 - i AC waveforms Building the circuit Although the theoretical aspects behind the electronic starter require a detailed explanation, the construction is simple and straightforward. Constructors will be pleasantly surprised by the format of the circuit and printed circuit board. Only eight components are involved! Figure 4 shows the printed circuit board for the circuit. It is compact enough to fit inside the plastic {not aluminium!) case of a conventional starter, which saves having to modify the fluorescent tube holders in any way. Make sure that the connection wires of the thyristor are not in contact with the metal heat sink. If necessary, glue the thyristor to the board with a touch of epoxy resin. Resistor R4 and capacitor C2 must be mounted on the copper side of the printed circuit board. The heading photograph shows the finished product from two different angles. As mentioned earlier, we do not recommend the use of a metal starter The photograph shows the finished product from two different angles. As mentioned earlier, we do not rec- ommend the use of a metal starter case for safety reasons. Open the starter case with care. Remove the 'helium' lamp and the capacitor from the contact pins. Do not cut the connection wires of the capacitor too short, as they can be used to solder the printed circuit board to the contact pins. Assemble the starter carefully and insert it in a fluorescent tube holder. The circuit is designed for fluorescent lamps in the 20 ... 65 W range. In the event of a 20 W tube not igniting right away, lower the value of Cl to 10nF. The value of this capacitor really depends on the type of fluorescent tube used. Incidentally, the same applies to C2. If fluorescent tube power ratings below 20 W are selected, it may be necessary to try out different capacitor values. M Please note that this circuit is patented by Philips (Mullard) no. 1223733. Testing and measuring electronic equip- ment can lead to a variety of problems. Before endeavouring to come up with any solutions, let us examine a few examples of the type of situation that is likely to arise. Figure 1 shows three different ways in which to evaluate an AC waveform. The peak amplitude U pp corresponds to 100% of the amplitude in either the positive or the negative half of the waveform. The root-mean-square value U rrns is about 71% of the ampli- tude and the average value 0 is only about 64% of the peak value. These percentages may seem rather strange, but that is because electronics is subject to the general laws of physics rather than any arithmetical correlation. measuring AC * waveforms . . . no problem when you know how When testing a circuit it is often very difficult to know exactly what to measure. With a digital multimeter measuring DC voltages is quite straightforward. But what about AC voltages? The constructor then has to decide what to measure: peak, average or rms (root-mean-square) values. This depends on which of the three alternatives gives a reliable indication of whether the circuit is working properly or not. Then again, which one is actually being displayed by the meter? K. Fietta The mathematical relationships between the three possible values for sinewave voltages can be expressed as follows: h .^EP.JLo ,s V2 2.2 U The basic formulae can be varied in form to provide the appropriate values. It is not our intention to delve into the physical features of test and measure- ment equipment here. It is enough to know that a moving coil meter (that is one without a permanent magnet) indicates the arithmetical mean of an AC waveform and that a moving iron meter displays a root-mean-square value. To find out the principles behind this, it is advisable to refer to the subject in a good electronics book. A digital volt- meter may also be used to measure AC voltages, but only if a rectifier is connected in the input circuit. That brings us to the problems men- tioned earlier, for both moving coil meters and DVMs require a rectifier. Although multimeters very often in- corporate an average responsive and a peak responsive rectifier to measure the average and peak values, respect- ively, the scale of the actual meter is calibrated to the rms values of a sine wave. This fact should be remembered when measuring other waveforms or results could become very confusing. One thing is clear: rectifiers play an important part in meters and it is a good idea to find out why before going any It should be noted that although a moving iron meter is suitable for measuring AC waveforms elektor june 1982 - 6-25 measuring the rms value of AC voltages, it is only used in heavy current engin- eering because of its low internal impedance. Rectifiers Figure 2a illustrates a peak response rectifier. Resistor R2 represents the high impedance input of a sensitive moving coil meter, a DC measurement amplifier or a DVM. Provided the R2 R1 par- ameter is met, the voltage levels of figure 2b can be expected to appear across capacitor Cl . The capacitor will be charged to the U pp level on the rising edge of the positive half cycle. When the voltage drops, the capacitor can only be discharged very slowly by way of R2. The leakage is compensated for during the following positive half cycle. As a re- sult, the DC voltage U is produced; this is the actual measurement voluge. Where Urms=10 v - U= 10V-V2 = 14.1 V. As this method is mainly used to measure the rms value of sine wave volt- ages, the meter scale indicates 10 V which corresponds to the rms value when a DC voltage of 14.1 V is applied. The peak response rectifier behaves extremely well in the case of non sine wave signals (waveforms other than sine waves). The peak values are of course also indicated accurately. However, errors will occur in the reading when the sine wave is degraded by spurious signals or other interference. Then the 'true' rms voltage value cannot possibly be deduced from the reading. Discrep- ancies of a more serious nature arise if AC voltages are produced with different half-cycle peak values and which are then mixed with a DC voltage. Peak meters are frequently used in audio to measure signal strength for recording levels. Figure 3 shows an average response rec- tifier. The current flowing through the meter is always in proportion to the actual value of the signal under test. Due to the mechanical characteristics of the meter, the values are integrated and the result displayed is the average. As was already seen in figure 1 , the average value of a sine wave signal is only 10% below the rms value. Again, the meter scale is calibrated for rms values. An average response rectifier responds fairly precisely to a square wave signal. At a duty cycle of 50% the instrument indicates 11% in excess of the true value. Now of course readers will point out the fact that when the duty cycle of a square wave signal is 1:1, the peak, average and rms values should all be the same ... so why does the meter miss the mark by 11%? Answer: because the scale has been calibrated to provide the rms values of sine wave signals. Note that average response meters are universal in that they provide fairly accurate rms readings even when the sine wave signals are distorted (up to 10% harmonics). A good example of this is the VU meter which monitors Figure 2b. The capacitor is charged during the positive had cycla and keeps its charge even when the voltage is no longer applied. signal modulation in tape decks and cassette recorders. Measuring rms values The rms of an AC voltage is defined as: The level of AC voltage required to produce the same amount of heat from a specific resistance as an equivalent DC voltage over a predeter- mined time interval, irrespective of the waveform. The relationship between the rms and peak values of a sine wave signal with respect to power is illustrated in fig- ure 4. All the values in the 'u' curve are squared, so that the values in the new curve, 'u 2 ' are positive. Since power P = U 2 /R, the root of the mean of the square is obtained as follows: U 2 pp/2 = U 2 rms . Figure 4. This graph helps to establish the relationship between U 2 rms and U 2 pp . The result is: U rms - U pp /\£. This confirms the assumption made earlier that U rms = U pp /v2 (for sine wave voltages). The relationships could also be illustrated by means of compli- cated and highly aesthetic integrals, but that would only confuse the issue — and the constructor! The question is, how to obtain the real rms value on a scale, irrespective of the waveform of the input signal? A moving iron instrument is not suitable here, as its internal losses are high. In other words, a comparatively large amount of power has to be fed to it before it will even start to indicate anything. As hobbyists are not in the habit of working with anything in the order of kV, kW, kVA or kA, they can forget about the moving iron meter for a while. The mathematical method for measuring rms values (for which special ICs are 5 Figure 5. A quasi rms response rectifier. The diode/resistor network 'shapes' the behaviour of the correct reading for a corresponding voltage value in a circuit ■toff angle (such as dimmer controls). that involves a change in the phase available nowadays) Is very complicated. First the signal is squared. An RC net- work acts as an integrator, an 'average shaper', so to speak. Finally, a rooter circuit extracts the root of the average value, providing the rms value of the signal at the output. In practice, this method uses a slightly different ap- proach. In order to obtain as wide a dynamic range as possible, the squared signal is divided by the output signal at the input. The 'rooter' at the output is then omitted. The division is logarithmic to allow small signal levels to be de- tected as well. In all honesty, the whole process is rather complicated and takes hours to explain, so let's forget about it for now! An alternative method is based on the physical principle behind the rms definition: a resistor wire is heated and the amount of heat is measured using a thermocouple. Obviously these very low voltages are rather difficult to measure. Diagnosis: this method is totally unsuit- able for amateurs. Therapy: none! Quasi rms measurement A compromise is reached between economy and practicality by measuring the quasi rms. There is no point in over- measuring AC waveforms doing the economy aspect by merely substituting the average response recti- fier in figure 3 for an integrator in the form of an RC network. Instead, it is preferable to set slightly higher par- ameters. Figure 5 shows a quasi rms rectifier, which naturally has nothing to do with the accurate measurement pro- cedures used in maths and physics. The behaviour of a 'real' rectifier (its curve) is imitated so effectively by the diode/ resistor network D1/D2/R3 . . . R6 that the deviation of the reading remains within the tolerance range permitted for rms measurement equipment. This type of circuit is particularly suitable for measuring distortion and for calculating power levels. Rms meters can also be used for other purposes as well, as we Measuring voltages in static converter circuits Fortunately, it is possible to carry out rms measurements without using the 'quasi 'method. For, as it was mentioned at the beginning of this article, the reading may be multiplied by a correc- tion factor to obtain the correct rms value. For this the type of rectifier used in the meter must be known. In most cases, an average response rectifier will be involved and the scale will already be calibrated to read rms, in other words, it was multiplied previously by a factor of 1.11. This figure only holds good for The relationships between peak, average and rms values have already been dealt with. However, where voltages in static converter circuits are concerned, the change in waveform will cause consider- able errors to occur. This is because the phase cutoff angle is not taken into account. This is how the rms value is related to the phase angle: Urms = Upp Vj; (*-¥>+ 1/2 sin 2 yj)' The formula for calculating the average value looks a little more straightforward. By multiplying it by the correction fac- tor mentioned above, the actual reading 2s/2 The two formulae may be plotted in relation to the phase angle as curves in a graph. The graph helps the constructor calculate the reading required for a certain rms value. If for instance, the voltage in a dimmer circuit is to be measured and its value should be 170 V rrns , the phase angle in the U rms curve will be 81°. At this value the vertical axis intersects the U curve at 1 26 V. If this is what the meter displays, the true rms will be 170 V. H 6. Zapf, The behaviour of measurement devices when measuring non sine wave voltages, Grundig Tl. L 1982 - 6-27 One of the biggest problems faced by breakers is which aerial system to choose, so that their transceiver will operate efficiently. This is a universal problem encountered by the whole spectrum of radio communication system users (HAMs, etc.). It is a fact of life that an aerial is probably the most vital part of any system. No matter what the quality or power rating, a transceiver will be made impotent by a badly designed or maladjusted aerial. Further restrictions are imposed on the designers when an aerial has to be mounted upon a vehicle. Unfortunately, because of practical and safety consider- ations the normal highly efficient static mobile aerials coil to 'make it long again'. Three different types of loaded coil mobile antennas are possible. Figure la shows the BLC (Base Loaded Coil) and figure 1b the CLC (Centre Loaded Coil) type. Both these designs are compact and reasonably short. Even though the rod is approximately 1 metre in length, the induction of the coil enables the entire unit to have a resonance of 27 MHz. The current distribution along each aerial is shown on the right-hand side of figure 1. These diagrams give a general picture of the way the aerials behave. As a general rule, the longer the aerial and the greater the current passing along it. the more radiation it produces. As a matter of interest, the CLC type has a better performance than the BLC, because the length of rod carrying a maximum level of current is greater than the BLC. By far the simplest to build is the BLC (figure la). This type is also easily and cheaply available professionally built. Readers are reminded that the BLC is the only 27 MHz CB mobile aerial that can be used legally on British roads. The use of any other type, as described in this article, should be confined to drive ways, private roads, and when on holiday abroad (check each country's Although 'some' CB enthusiasts no longer have to be on the run many will still prefer to use a mobile installation in the car. This article describes how a single aerial can serve both a car radio and a CB transceiver operating within the legal 27 MHz 'FM' band. It also discusses the various merits of different 'possible' aerial designs. systems are totally unsuitable. A mobile aerial has to be compact and fairly short if only to comply with the existing laws. Many readers are probably wondering why the majority of VHF/UHF aerials are vertical rods of various descriptions. The main reasons for using vertical as opposed to horizontal polarisation are as follows: • They are simple and unobtrusive and easily mounted onto vehicles; • single element antennas give all-round coverage (omnidirectional) irrespect- ive of the direction which the car is • it is an accepted standard for mobiles working within the UK. Readers should not worry about the term 'ground-plane' aerial. Basically any rod (or whip) aerial becomes one of these if it is mounted on the metal roof Before delving too deeply into the problems surrounding the use of ordi- nary telescopic car aerials, it is a good idea to look at the 'possible' types usable for 27 MHz. The simplest and most commonly used mobile aerial is the %X. For a standard 54 X aerial to have a resonance frequency of 27 MHz, it would have to be approximately 2.7 metres long. Stick that on a car and see what happens to the driver when confronted by the local bobby! The only alternative is to physically shorten the rod, and 'electrically' lengthen it in order to retain the 27 MHz resonance. This is achieved by addinga 'loaded coil’ (to the shortened rod). In other words: cut it down to a size (length) that can be mounted on a car, and then add a Figure 1. Three different ways in which to electrically lengthen a rod aerial with the aid of a loaded coil. The graphs plot the level of current passing along the aerial in each example. Note that only the BLC is legal in the UK. 6-28 - elektor june 1982 regulations first). The BLC principle can also be utilised when modifying a standard car aerial for CB and the modification circuit is described later on in the article. A CLC is rather impractical from a constructional point of view. A normal rod aerial has to be cut into two equal lengths, the coil being fitted between the two halves. The result would prob- ably be rather unstable. Figure 1c shows a TLC (Top Loaded Coil) aerial. This type can be easily built and has the best overall performance. In order to maintain equal resonance, a capacitive load has been included in the form of a 'capacitive hat'. The 'hat' may be either a metal lid or a couple of metal spokes. The TLC has two advan- tages over the BLC and CLC types: the length carrying the maximum current is greater (see the graph in figure 1c) and due to the careful construction of the 'hat' the induction of the coil is reduced considerably. This results in more radiation (yield) and less 'mismatch' losses, leading to a better performance. There are various ways in which to build a TLC 'hat'. Figure 2 shows one method. A coil is wound around a piece of PVC 'conduit', one end of which is connec- ted to one vertically and four horizon- tally mounted 'spokes'. The other end is obviously attached to the aerial. The 'spokes' may be knitting needles (the old-fashioned metal type!) or bicycle spokes that are cut to size. The coil has a total diameter of 19 mm and consists of 24 turns of 1 mm tp enamelled copper wire. The wire must be wound very tightly, without leaving spaces. The other end of the coil is linked to the top of the rod aerial with the aid of a terminal block. The coil can be made waterproof by means of a plastic coating spray or an epoxy resin. This is highly recommended, as most car aerials have to withstand all kinds of weather. In any case, the coil will be considerably damaged, if any water manages to trickle in. Note that the TLC mount causes no interference to the FM wave band. Therefore there is no need to remove it when using the car radio. Before explaining the modification circuitry a short note on the use of shortened car aerials. These normally have a length of % X for the FM wave band (about 70 cm). Although this is far too short, the addition of a loaded coil together with a modification circuit as shown in figure 3 will make it resonate at 27 MHz. One aerial, two radios . . . Whether the mobile aerial is a home-built or a bought 27 MHz type, problems are bound to arise once the aerial is used for both the car radio and the CB trans- ceiver. It would be dangerous, to say the least, to simply connect the transmitter output of the transceiver to the input of the car radio . . . and hope for the best. Few car radios will appreciate, or even survive, this kind of treatment. In order to avoid damaging the car radio, a filter system has to be installed. The simplest solution would be to connect an effective high-pass filter (which would eliminate any signals below 80 MHz) in series with the car radio aerial input. The FM wave band (87 . . . 108 MHz) can then be received without any interference on the car radio, while simultaneously transmitting on CB. Unfortunately, this kind of filter also ‘cuts out' any long and short wave signals that the aerial picks up. For this reason a different approach was looked for. Figure 3a shows the complete filter circuit as it would be mounted on a printed circuit board. The filter is made up of two separate sections, the lower section of which (L6 . . . L8, C4 . . . C6) contains an aerial modifi- cation network for the 27 MHz trans- ceiver. This enables the transceiver to be used at full power (4 W) despite the Blektor june 1982 - 6-29 shortened aerial modification. Using the trimmer capacitors C5 and C6, the set may be adjusted to a minimum VSWR. Readers should note that the circuit shown in figure 3a is designed for a BLC type using a normal car aerial, in which L8 acts as the loading coil. If either a CLC or a TLC is used, L8 (and C4) may be omitted. The modification network will then resemble the circuit in figure 3b. The filter designed to protect the car radio against high-risk 27 MHz transmit- ter signals is shown at the top of figure 3a. As can be seen, it isn't a high-pass but a highly selective filter. It consists of a band-stop filter (L3 . . . L5, Cl . . . C3) for the 24 ... 30 MHz fre- quencies and a by-pass filter for the FM wave-band. The resonant circuits LI, L2/C1, C2 are included in the by-pass filter and are tuned to approxi- mately 95 MHz. The filter circuit is quite effective. Frequencies within the band-stop range are suppressed by around 60 dB. There- fore, using the authorised CB trans- mission power of 4 W, not more than 0.5 pW interference reaches the aerial input of the car radio. A very satisfac- tory state of affairs. Construction Some of the coils used are not available ready-made, so readers will have to 4 PVC Figure 4. Coils L6 and L7 can be wound, one beside the other, around a piece of PVC conduit. make them themselves. However they are not difficult to wind, as there aren't any taps or secondary windings. Three of the eight coils required (L3 . . . L5) are in fact easily obtainable chokes. Details concerning the construction of the other five are provided in figure 3. L6 and L7 can best be wound around a piece of PVC conduit in the manner indicated in figure 4. For ease of construction, a printed circuit board has been designed for the circuit and is shown in figure 5. Once the coils are made, the filter can be built in a matter of minutes. Although this cannot be seen in figure 5, the board is in fact double-sided. There is a wire link, as opposed to a copper track, connecting the lower side of L7 to earth. This allows the modification circuit needed for a CLC or TLC aerial to be con- structed (as shown in figure 3b) without the need for any drastic changes to the printed circuit board. C6 is soldered in the position of the wire link (becoming C7) and therefore no longer acts as a trimmer capacitor for L7. By shorting out C4 and L8 with wire links, the circuit will resemble the one in figure 3b. An important point to note is that the dotted lines as shown in figure 3 have a specific purpose. The optimum oper- ation of the circuit is only guaranteed when the 'radiating' section of the aerial modification network is screened from the band-stop filter. This is done by mounting a metal partition on the board, in the position denoted by the dotted Finally, the link between the printed circuit board and the aerial should be as short as possible to prevent unnecessary dissipation. If possible, the printed circuit board should be mounted just below the car aerial. M 6-30 - ele 1982 electronic dog 1 Some time ago a particular type of tweeter came onto the market ac- companied by an enormous amount of publicity such as 'over 300 W' and 'without a crossover network', etc. The 'Hallelujah Chorus' of the advertising fraternity was convinced that it would take the world by storm. The piezo tweeter did not in fact receive the universal acclaim expected and as result they are still relatively cheap and easily available. The article is certainly not going to argue the pros and cons of this tweeter, let's just say, that for certain appli- cations it is ideal. electronic dog whistle high quality ultra-sonic dog call Most if not all the circuits published in electronic magazines have always catered for other hobbies. During the last few years Elektor has designed numerous circuits for photographers, musicians, moviemakers, model railway enthusiasts and so on. But, 'where oh where' are the circuits for the dog owners of this country? After all there are millions of people who are proud of 'man's best friend'. In order to keep this section of the community happy, we hereby publish our first dog biscu . . . sorry, circuit, and we assure everyone that electronics is not going to the dogs. The Piezo tweeter horn The main difference between normal dynamic horns and the piezo is its construction. The latter has a membrane driven by a small plate of piezo ceramic material. The result is a horn with a very small dynamic mass. Incidentally the same principles are employed in certain ceramic cartridges and most commonly in cigarette lighters. The impedance of a piezo tweeter resembles that of a capacitor (see figure 1 ), rather than that of a resistor (normal dynamic type). Consequently this type of tweeter has a very high efficiency, in other words a good input to output sound pressure level (dBs) relationship. Therefore it can be driven by a battery powered circuit and made to reproduce very high frequencies. Just right for the dog circuit! Doggy ears Have you ever wondered why your dog pricks up its ears from time to time when no apparent audible sound is present? As most readers will know dogs are able to perceive audio frequencies outside the human hearing spectrum. This is for both ends of the scale. Con- sidering a frequency of 20 kHz, the average person will not hear it at all (there are exceptions) irrespective of the volume level. On the other hand, animals and in particular dogs, are sensitive to these tones and will react instantly; unless they are asleep or just lazy. Anyway whistles producing such frequencies are useful, allowing dogs to be called from great distances without waking up the whole neighbourhood. Mind you, even using one will not guarantee that fact because dogs are not the only ones able to hear it! Canaries, young children and some adults are likely to hear it as well! There is also the probability that all the dogs in the neighbourhood will respond and land on your doorstep. The circuit The high frequency tone required can be derived by driving the piezo tweeter with the circuit as illustrated in figure 2. A square wave instead of a sine wave is applied in order to keep battery consumption as low as possible. The tone is produced by means of N1 . . . N3, R1 and C2, which constitute an astable multivibrator. Due to the fact that the Piezo horn forms a capaci- tive load, the wave forms of the signal will have high peaks. That is why the Schmitt trigger inverters N1 . . . N3 and N4 . . . N6 (all 6 inverters are present in the 40106 1C) have been connected in parallel and supplied with an output stage, consisting of T1/T2 and T3/T4 respectively. N4 . . . N6 invert the signal coming from N1 . . . N3. In this way a 'power oscillator' is constructed. When fed by a 9 V battery, this 'power oscil- lator' supplies an a.c. voltage having an amplitude of 15Vp P and a frequency of approximately 21 kHz. Could not be better for our needs! Sound pressure Figure 3 shows the frequency response of the Piezo tweeter. In this case we are mainly interested in the 20 kHz range and fortunately enough the horn reaches its maximum efficiency at this frequency. This curve was recorded with a controlled voltage of 4 V r ms and a microphone held at a distance of 457 mm from the horn. The Elektor dog whistle supplies a voltage of 1 5 Vpp. The rms value of this voltage is approxi- mately 6.5 V, because we are dealing with a square wave voltage having a slightly unsymmetrical duty cycle. Using 1982-6-31 Figure 1. The impedar af the Piezo horn. this value (6.5 V rms ) and extending the microphone distance to 1 metre, will result in a sound pressure of 101 dBI! Quite a lot for 20 kHz! Be Warned Care should be taken when using the whistle. Even though the user may not be able to hear it, remember 101 dBs are being produced which is going to give somebody or other a headache. 20 kHz at high volume should not be aimed at any human or animal in close proximity. It's similar to sitting in front of the speakers of 1000W disco system for a few hours. Keep in mind that the long term side effects of all this are not known, but to be on the safe side (like smoking) it's better to accept the possibility that it could 'damage your health'. Finally, to play it safe we suggest you equip your dog and yourself with ear protectors and then try it. Have fun! H 6-32 -elaktor june 1982 In a very short time every electrical appliance will be talking to you: the washing machine, vacuum cleaner, cooker and probably, the kitchen sink. This 'desirable feature' (?) is already evident in the new generation of digital clocks that are fast beginning to appear. A clock that actually tells the time is not such a bad idea after all, especially for the visually handicapped. The UAA1003 from ITT has been designed specifically to form the basis of a talking clock. It incorporates a complete speech generator designed specifically to 'tell the time'. Further- more, it can be connected directly to talking dock give the 6502 housekeeper the gift of the gab! More and more 'chattering chips' are appearing on the market. In December 1981 Elektor introduced the Talking Board with its extensive vocabulary. But, as this article points out, computers are not the only ones to talk. Even digital clocks can now be 'conversed with' thanks to the UAA 1003 from ITT, a single chip speech generator. Once the 1C and a few other components have been added to the 6502 housekeeper described last month, the clock will well and truly be able to 'tell' the time! the seven segment outputs of any (existing) digital clock. Last month, Elektor published a versa- tile clock of its own, the 6502 house- keeper, and so it seemed a good idea to draw up a circuit for it using the UAA 1003. After leaving the 'operating table', the clock will be able to express the time both in digits and words. As mentioned earlier, the speech cir- cuit can be connected to 'ordinary' digital clocks, with the proviso, that their displays are of the CC (common cathode) type. The speech generator The UAA 1003 is a speech generator 1C in a 40-pin package. The 1C is shown in the form of a block diagram in figure 1. Digital techniques are used to store and process the required phonemes. By using data and redundancy reduction methods, it was possible to store a vocabulary of about 20 words and inte- grate the necessary control, decoder and D/A converter circuits, all on a single chip. Every word generated by the speech 1C contains a number of step-shaped pulses, each one having a fixed pulse duration of 10 ms. Every pulse is made of up to 1 28 different amplitudes which can each assume 16 values. This corre- sponds to 4-bit amplitude modulation. Different word segments are linked up according to the digital control signals that are applied. The 1C is currently available in two languages, English and German. Let's examine the 'insides’ of the 1C as shown in the block diagram in figure 1. When the speech generator is 'switched on' via either of the two start inputs, the intermediate input data is read in. The decoder ROM and the control cir- cuit establish the word order according to the data entered and then address the corresponding word parameters, after which the address logic fetches the speech segments from the speech ROM. The output digital code is processed inside a data regenerator before being sent to a D/A converter which delivers the actual speech signal. The speech generator 1C has a special feature in that it receives its time data from the clock's seven segment connec- tions. However, the data inputs of the 1C will only function provided the circuit is connected to a digital clock with common cathode displays that are not multiplexed. Not all the segment connections are needed to decode the time. Segment connections c and d serve to decode the hour tens, a, b, e, f and g the hours, d, e and f the minute tens and finally. Figure 1. Block diagram of the UAA 1003. Phonemes are stored and processed in a digital clock 1982-6-33 a, b, e, f, and g the minutes. The data inputs of the 1C have an internal pull- down resistor, enabling them to be connected directly to the segment out- puts of the clock. The pin assignment is as follows. There are two start inputs, pins 14 and 15. When the 1C generates a positive pulse at pin 14 the time is announced in the manner described above. If this is pro- duced at pin 15, however, the time is preceded by an alarm signal that lasts about one second. The 'busy' output (pin 12) is a kind of open collector out- put and has a low impedance while the time is being output. It may be used to control any external devices that are hooked up to the clock. A DC voltage is applied at pin 18 so as to calibrate the oscillator frequency of the 1C. The set frequency is available at pin 16 (a kind of open collector out- put too) for measurement purposes. An external reference current must be applied to pin 34. The amount of current determines the level of the output signal. The speech output (pin 33) again produces an output current, as a result of which a resistor will also have to be connected to it in order to provide an output voltage. Pins 17 and 19 constitute the stand-by power supply connections. They allow the 1C to be connected to a stand-by supply whenever it is not used to indicate the time. This comes in handy if the circuit is battery-powered, for instance, but there is no need to go into that here. Pins 20, 1 and 35 are the 'normal' power supply connections and the remaining 1C pins are all data inputs. Adapting the circuit to the 6502 housekeeper As readers will remember, the 6502 housekeeper is more than just a clock. It can be used for timing all sorts of processes in the home, darkroom, workshop, etc. In short, a device well worth endowing with the power of speech! One minor problem has to be dealt with first: the displays on the housekeeper are multiplexed and, re- member, that is precisely what the UAA1003 does not cater for. Don't worry, this can be remedied by adding a couple of ICs, by way of an interface, to the circuit. Figure 2 shows the various signals that control the displays in the 6502 house- keeper, The display segments are driven by PAO . . . PA6 and lines PB3 . . . PB6 make sure that the four required dis- plays are multiplexed. Using a set of D flipflops, the segment data belonging to the various displays must now be stored to allow all the signal infor- mation to be applied to the speech 1C simultaneously. To ensure that the right information enters the right flipflops, the PB signals are used to read in the data on the PA lines. This means that the flipflops corresponding to the seg- ments in display 6 must receive a clock pulse from line PB6, and so on. If we take a closer look at the waveform on PB6, as shown in figure 3, the rising edge of the signal can be seen to appear virtually at the same time as the data on PAO . . . PA6 (for LD6). The rising edge on PB6 must be slightly delayed, initially to make absolutely sure that the correct signals are read into the flipflops. This is taken care of by the R 1/Cl delay network included in the circuit diagram in figure 4. A similar delay technique is also employed on the other PB lines. The flipflops (IC2. . . IC6) are situated to the left of figure 4. The seven seg- ment data required by the UAA 1003 is permanently available at the outputs of the flipflops (as if the clock were a non- multiplexed type, after all). Theoreti- cally, therefore, the flipflop outputs could be linked directly to the data inputs of the speech 1C, were it not for another slight snag . . . The data on the PA lines is inverted with respect to the segment information. Fortunately, this can easily be remedied by connecting the 0 outputs of the flipflops to the data inputs instead of the Q outputs. That just about covers all there is to say about the circuit diagram. We've already dealt with the UAA, so that only leaves the output amplifier, an LM 386 in this case. A bandpass filter consisting of R10, C5, R11,C6,C7and P2 is included between IC1 and IC10. Potentiometer P2 acts as the volume control. Finally, the stabilised 5 V voltage is provided by a 7805 chip, IC11. The whole circuit consumes about 150 mA current. PI effects the only calibration needed for the circuit. This adjusts the | internal clock frequency of the speech 1C. The adjustment may either be carried out by ear (until the voice sounds human!) or by measuring the frequency at pin 16 of the 1C. This should be about 25.6 kHz. Connecting up the circuit The circuit shown in figure 4 can be con- nected to the 6502 housekeeper with- out any difficulty. Lines PAO . . . PA6 and PB3 . . . PB6 belonging to the talking clock board are simply linked to the corresponding connections on the main board of the 6502 housekeeper. The power supply may be connected up right after the bridge rectifier on the housekeeper power supply board. The ALARM input may be linked to one of the TO. . .T3 switch outputs. Whenever the corresponding output goes high, a short alarm signal will be emitted, after which the time is an- nounced. Usually, of course, push- button SI is depressed to make the clock 'speak', but then the time indi- cation will not be preceded by an alarm signal. What about other digital clocks? Other digital clock can be made to talk too, but this does call for a little more time, effort and components. The simplest solution is to connect the circuit to a non-multiplexed clock with common cathode displays, as this, after all, is what the UAA1003 was de- signed for. In that case, components IC2 . . . IC9, R 1 . . . R4 and Cl . . C4 may be omitted. The input of IC1 (points A, B . . . P) are connected directly to the corresponding display segments in the clock. Segment c per- taining to the hour tens display is there- fore linked to point P, segment d to point N, and so on. The logic levels of the digital clock pins from which the required signals are derived must meet the following par- 0 V < Ui < 0.3 V (segment 'off') 1.5 V < Uh < 5 V (segment 'on) The 'low' level is usually correct due to the pull-up resistors at the inputs of the UAA1003. The 'high' level should not be a problem either, as the oper- ating voltage of a display segment is at least 1.6 V. Making clocks with multiplexed displays talk is a different matter. Since in this case all the components must be mounted on the board (to store inter- mediate multiplexed data), the seg- ment connections must be linked to inputs PAO . . . PA6 and PB3 . . . PB6 in the normal manner. Note that the inputs respond to TTL levels here <0V fuse (see text) fuse holder for pcb quiescent voltage level before checking the delay time. A very high value (more than 1000 ffF) may lead to an excessive leakage current and cause problems. Practical points Construction of the printed circuit board will present no problem. How- ever, if a socket is used for I Cl it is important to ensure that C4 is dis- charged each time before fitting the 1C. In practice the unit can be mounted virtually anywhere that is convenient, including an existing switch box. In the latter case it must be noted that the dimmer control will not be compatible with normal house wiring and extra cable runs will have to be fitted. A total of 4 wires plus earth must run between the switch box and the light fitting. Further to this, a mains supply is also required. An alternative is to fit the entire electronics into the light fitting, if at all possible. This method requires only three wires to the switch control It is not possible to advise exactly what modifications are required, as 'standard electrical wiring practices' may well not prevail, especially if yours is an older property. Enough to say that if you are at all unfamiliar with the electrical 'arrangements' in your house, it may well be advisable to invite your friendly electrician in for an evening and gently steer him towards the subject. M elektor june 1982 - 6-47 Electronics are finding their way into the car more and more these days. This is not confined to the up-market models either. The application of the majority of electronic circuits in the car are related to energy and cost saving. This normally takes the form of electronic ignition and timing systems of varying complexity. Another obvious appli- cation of electronics is the protection against theft of the vehicle. As well as protecting the car, the alarm system described here also provides protection for accessories such as radio, cassette deck and CB rig. In many cases it is not the car itself that is stolen, but its contents! ear alarm an active insurance policy Although cars are insured against theft, most motorists will agree that it is preferable not to make use of the policy. The major advantages of the circuit described in this article are automatic resetting and protection against false alarms, which is not only a good thing for the owner, but also for the neighbourhood. Alarm systems Alarm systems will always be a matter for discussion. This is especially true when deciding which type of system to apply and how extensive the coverage needs to be, since as far as electronics is concerned, the complexity could be infinite. W. Schuster Commercially available systems usually come in one of three guises. The basis of a very popular alarm system is a type of 'tilt switch' which is used to activate the alarm. Effectively, this consists of one or more switches which are sensitive to any slight movement of the vehicle. This makes it almost impossible for a would- be thief to touch the car without acti- vating the alarm. However, the major disadvantage of this system is that the alarm cannot differentiate between various kinds of vibration. They tend to be triggered by passing vehicles, strong wind and pedestrians who inad- vertently touch the car. Far more sophisticated alarm systems are based on ultrasonic or infra-red prin- ciples. These do not react to the move- ment of the vehicle, but they certainly provide excellent protection for the interior of the vehicle. However, instal- lation and setting up require a fair amount of time and effort. The system must be designed to cater for fluctu- ations in temperature (which can be large inside a vehicle) and prevent false triggering by the movement of insects inside the vehicle. The latter holds par- ticularly true for ultrasonic based systems. The third and simplest type of alarm is triggered by courtesy light door switches. This is a good compromise between cost and efficiency. With the help of some electronic circuitry the construction of a reliable alarm instal- lation should not prove to be too diffi- cult. The following circuit is based on this principle. Operation of the system The simpler the circuit, the more reliable it is likely to be, and so this type of circuit is the basis for the vast majority of car alarm systems. How does it work? When leaving the car the system will be energised, either auto- matically or by a switch that is hidden somewhere inside the car (underneath the dashboard, for instance). A lamp on the dashboard (which can be either a LED or a commercially available 12 V indicator) will light for approximately 1 minute showing that the alarm is activated. During this time period the occupants of the car must leave it and close the doors. The alarm will remain silent while the car doors are being opened and closed. The alarm will be primed 6 seconds after the light goes out. If a door is now opened, the alarm will sound after a 6 second delay. It will continue to sound for a period of 1 minute, by which time your average thief will be attempting to fade dis- cretely into the background. A useful advantage of this circuit is its reset facility. This is fully automatic ensuring that any further attempts will have the On returning to the vehicle, the rightful owner would simply turn off the alarm by means of the hidden switch during the 6 second delay. (This should be practised as any fumbling would cause a certain amount of embarrassment . . . ) CMOS ICs in the car There are a number of reasons why CMOS ICs are suitable for use in the car. The most important is their wide supply voltage range (between 3 and 15 volts), eliminating the need for voltage regu- lators. With a supply voltage of 12 V, a noise immunity margin of better than 5 V can be reached - a figure that is far superior to any other logic family. Another advantage, of course, is their extremely low current consumption. The quiescent current of CMOS devices can be considerably less than the normal self-discharge rate of the car battery. The only real disadvantage of using CMOS ICs is the problem associated with handling. This however, ceases to exist once the 1C is mounted on a printed circuit board. The circuit Figure 1 shows the complete circuit diagram of the car alarm. The system is activated by means of the hidden switch S2 which, when closed, supplies power to the circuit via diode D1. Initially, the flipflop consisting of gates N1 and N2, will be reset. This is ensured by the time constant of capacitor C4 and resistor R5 which holds the pin 8 input of N2 low for a period of time. The initial state of the outputs of the flipflop will therefore be low and high for the Q and 0 outputs respectively. The Q output is used to control the N3/N4 oscillator which will be switched off with a logic 'O' at pin 1 of N3. The 'high' output of Q is fed to the clear input (pin 21 of IC3. The contents of this seven stage ripple counter will now be cleared and ready for action. For the C4/R5 time period, the output of N6 will be high, switching on the lamp Lai via T2. This gives a visual indication that the alarm is primed. During this time period, opening the door will have no influence on the cir- cuit because the trigger input of the flipflop is 'latched' high by the output of N6 via T1. The circuit will remain in this condition until C4 charges via R5. With the values shown in the circuit diagram this will be about one minute, by which time the trigger threshold of N6 will be reached. Its output going to logic 'O' will have two results: Transistor T2 will switch the indicator lamp off and C5 will begin to charge via R7. After about 6 seconds (the time con- stant of C5/R7) T1 will release the set input at pin 13 of N1. The flipflop will not alter its state yet, it will require the operation of the door switch to do this. The alarm circuit is now fully 'active'. An entrance to the car by an uninvited guest will result in the set input of the flipflop being taken low. Things really start to happen now. The high appearing at the Q output starts the N3/N4 clock oscillator running at the same time as the 'clear' is removed from IC1 by Q. The counter outputs at pins 9 and 6 are 'summed' together with the clock signal. The resultant outputs of gates N7 and N8 will operate the relay (via T3) 12 times in 6 seconds. After a short interval the cycle is repeated, three times in total. The indicator lamp on the dashboard will also light in sympathy. This method of sounding the horn is for two reasons. Firstly it is quite 'energy conscious' and secondly, the horn will sound different from normal, and there- fore, hopefully, easily recognisable by the car owner. At the 64th clock pulse at pin 1 of IC3, about the same time that the would-be thief is attempting to merge with the nearest crowd, the Q1 and Q7 outputs will coincide with a logic 1 output. Gate N5 will now provide a reset pulse for the flipflop. This will stop the horn from sounding but it will not disable the alarm circuit. It will simply wait with infinite patience for the next customer. Additional protection The shaded areas in the circuit are 'op- tional extras', that is, the circuit will also elektor June 1982 - 6-49 function correctly if they are not in- cluded. The components around S3 and T4 form an anti-sabotage circuit. The experienced car thief will attempt to open the bonnet of the car first in an effort to disable any electronic protec- tion circuit fitted. With the circuit here things do not get off to a good start for him. Switch S3 is operated by the bonnet which, when opened, makes the connection between terminals 9 and 7. The charge on C7 will now switch T4 on and sound the horn immediately for about 20 seconds (until C7 discharges). Our unwelcome friend will be wise if he drops the bonnet and moves on. This will make S3 bridge the contacts 8 and 9 to allow C7 to recharge via R4. In a few seconds the alarm will again be fully active. The second option is a connection to the ignition switch, shown in the circuit diagram at point 6 (top left-hand corner). This ensures that the alarm is always disabled when the ignition is switched on. 5 < H H1 ° h o o-w-o o-w2§ i W Te o^Hl^ ' a x t . t o{m lo us 'O' O — ►fro OK l 0-II-OOC3 Mens lo t o o jp-m | o Figure 2. A suggested track pattern and component overlay for readers who wish to make a printed circuit board. C4,C5- 22 p/16 V tant. C6 - 33 n MKS C7- 100 p/16 V Resistors: R1,R7,R8 = 1 M R2 “ 15 k R3,R4 = 22 k R5 = 2M2 R6 = 47 k R9= 10M R10.R1 1,R1 5.R1 7 = 10k R1 2,R13 = 1 k R14 = 220 n R16 = 1M2 Cl = 100 n MKS C2 = 4p7/16V C3 = 1 n MKM D1 - 1N4004 D2 . . . D10= 1N4148 T1.T4- BC547B T2 - BC 140 T3 = BD 136 IC1.IC2 = 4093 IC3 - 4024 Miscellaneous: S3 = 2-way double pole switch Rel = 12 volt relay Lai = 12V/50. . . 100 mA light bulb or LED with 1 k resistor in series Construction and installation The circuit can be constructed on a piece of Veroboard and fitted in a small plastic box. Small is the operative word here because the completed circuit must be hidden and this will be easier if its size is kept to a minimum. The relay for the horn should be a standard car head- lamp or horn relay. It will also be less apparent that it is an addition under the bonnet. The object of the exercise is to make the whole installation as unob- trusive as possible in order to escape the attention of the more experienced thief. For instance, use black cable for all wiring under the bonnet and keep it out of sight as far as possible. Do not fit the relay near the horn. It is strongly advis- able to cover the horn connections with a few layers of tape so that a disconnec- tion here is as difficult as possible. Remember, the greatest enemy of the car thief is time and the longer we can delay him the better chance there is of him giving up and moving on to an easier victim. M 6-50 - elektor june 1982 from Following the latest trend in high-speed systems. Motorola has developed a microprocessor that has an internal 16- bit structure. One of the reasons why the 6809 is known as a Super 6502' is that its registers have the same names as those in the 6502. The features of the two systems are in fact very similar, except that the Motorola chip is much faster and more powerful. The differ- ences in structure are shown in figure 1. die 6502 to the 6809 a new 'super' 6502! The 6809. As always in the ever advancing world of electronics a popular and worthwhile microprocessor, has been superseded once again by a chip with a greatly improved performance: the 6809 CPU, manufactured by Motorola. The beauty of the 6809 is that it can be implanted into existing 6502 systems without any difficulty, thereby creating a new 'super' 6502. With just a few minor hardware modifications, constructors will then have at their disposal a much faster, more powerful computer with new fascinating programming facilities. Register 6502 X-Register 16 Bit Y-Register 16 Bit Stack Pointer 16 Bit Accu A 8 Bit Direct Page Reg. variable Status Register 8 Bit Program Counter 16 Bit As can be seen, the 6809 contains an additional 8-bit accumulator and a variable 'direct page register'. The 6502 CPU, on the other hand only had a single page zero. The 6809 also makes 256 direct pages available. The 6809 has a further advantage in that its two accumulators, A and B, may be com- bined into a 16-bit D accumulator. The' instruction set will look familiar to 6502 operators. Very little has in fact been altered in the mnemonics and addressing modes. The branch commands are particularly effective. The processor can branch within the -16 . . . +15,-128 . . . +127, or —32768 . . . +32767 address ranges. New instructions, such as BRA (branch always) and BSR (branch to subroutine). from the 6502 to the 6809 allow programs to be stored in any area of memory, without having to rely on absolute addresses and without having to alter a single byte. Such programs are known as 'relocatable' routines. The system introduces a new addressing mode, the 'program counter relative' mode. This is extremely powerful, and enables any memory location to be addressed (at a certain moment) that corresponds to the contents of the program counter. As the saying goes, "What you gain on the swings, you lose on the roundabout" and the same applies here, for 6502 fans will have to give up one of their favour- ite addressing modes, the indirect indexed mode (as in LDA-(POINT),Y, for instance). Unfortunately, indirect addressing modes cannot be indexed on the 6809. However, as we have already seen, plenty of other valuable facilities are available instead. The indexed addressing takes a slightly different form. The opcode consists of a single byte and is followed by a 'post byte', which may contain a 5-bit displacement. The next byte or byte pair either represents an 8-bit or a 1 6-bit displacement in two's complement. The effective address is calculated by adding up the index and the displace- ment: index (contents of X, Y, S, U, A, B or C registers) + displacement = effec- tive address. If a displacement is made within the —16 ... +15 range an instruction in the index addressing mode will only contain two bytes: the opcode and the post byte. Although there is no actual indirect indexed addressing mode, memory may also be accessed indirectly in the in- dexed addressing mode. What happens is that the pointer (the sum of the index and the displacement) indicates the memory location in which the ADH of the effective address is stored. The ADL is stored in the following memory location: In the 6809 CPU, the ADH and ADL are always located in that order, after the operation word. But, as readers will remember, this was the other way around in the 6502 (ADL, ADH). An indirect facility is extremely useful, as it enables arrays and symbol tables to be drawn up in high-level pro- gramming languages. The accumulators may also be used as index registers. This means not only can they be incremented and decremented, but they can also be employed during operations in arithmetic or binary (Boolean algebra). In other words, the index can be calculated. This is known as the accumulator indexed mode. The 6809 CPU contains two stack pointers, S and U, and is therefore already one up on the 6502. Sisal 6-bit stack pointer with the same function as that of the 6502. Return addresses from sub- routines and from machine registers are automatically stored on the S stack. It is also used to execute interrupts. As its name suggests, the user stack I the 6502 to the 6809 □ O'-'C, 033 Figure 1. A comparison of memory organisation in the 6809 and the 6502. pointer (U pointer) is purely at the disposal of the programmer. It is also 16 bits wide and mainly used as an input buffer and loop pointer during text editing. When the 6809 and 6502 systems were compared at the beginning of the article, both were seen to have a fairly similar programming structure. Even the ad- dressing modes are almost identical, the only difference being that the 6809 CPU provides a more 'powerful' instruc- tion set and is faster than the 6502. All in all, it really is worthwhile to update the 6502 system and convert it into a 6809. What makes it even more tempt- ing is the fact that: - only the hardware needs to be slightly modified; - more software is available for the 6809 CPU than for the 6502; - BASIC, FORTRAN, PASCAL and a cross assembler (for all commercial processor types) are all provided on diskette for the 6809 system. Cross assembly may be 'bi-directional' such as, say, from the 6809 to the Z80, or - and there is one standard floppy disc control format for all 6809 systems, whereas various formats exist for the 6502. But it is time to answer the question of how can a 6502 system be converted into a 6809 computer? First of all, mount the 6809 CPU together with a 4 MHz quartz crystal and two capacitors on a piece of Veroboard and mount the unit on a 40-pin DIL connector. Now simply substitute the 6502 for the 6809. The pin assignment is shown in figure 2. This piggy-back construction is illustrated in the photograph. Figure 2. Pin assignment of the 6809 CPU. The pin numbers in brackets correspond to the 6502. The conversion procedure: - Remove the 6502 CPU from its socket. - Insert the 6809 piggy-back board in the now empty socket. - Replace the 6502 operating system (stored in ROMs or EPROMs) by the 6809 version. Use may be made of the ASSIST 09 monitor program, for instance, published in the Programming Manual mentioned below. A text editor, a linker/loader and a disc operating system (DOS) are also avail- able for the 6809, which means that the Junior Computer (in combination with a floppy disc system, of course) can now be 'taught' to run in FORTRAN and PASCAL. In the end, the machine will be completely polyglott! Background literature: MC 6809-MC 6809E; 8-bit Micro- processor Programming Manual; M6809 PM (AD); 1.3. 1981; Motorola (including ASSIST 09) Macro Assemblers Reference Manual; 6800, 6801, 6805, 6809; M68 MASR (D); Motorola. M introducing DMOS | sr FETs The term VFET will sound familiar to most readers, although few are likely to have actually seen one 'in the flesh'. Not that they are much to look at, but it does go to show that VFETs have as yet failed to attract the amount of popu- larity they deserve. Way back in 1976 (see the Elektor April issue of that year) VFETs were billed to be the (almost) ideal output transistors for (audio) amplifiers. Due to their high price and poor availability, however, they never quite made it into the limelight. But then, this is just one of those vicious circles, for components don't drop in price and become easy to obtain until they are already popular . . . About a year ago, a new branch was welcomed to the VFET family: the DMOS series. Basically, they are very similar in operation to VFETs, but their structure is slightly different and their switching times are much faster. DMOS FETs are in fact mainly promoted as fast switches. They are predicted to take over a large share of the power transistor market and can be used in converters, switching power supplies and in relay control and motor speed control sys- tems. In addition, some types are de- for RF purposes. DMOS family has same fundamental structure, the construction of the gate may vary from one manufacture to another. Generally speaking, VMOS FETs are better suited as RF ampli- fiers than their DMOS suc- cessors. The latter, on the other hand, are more verti- cal in structure (as will be seen later on) and are therefore capable of handling higher voltage levels. Before we go any further, let's take a look at the main characteristics of the VFET family as a whole and disregard their individual traits for the moment. First of all, we need to find out how FETs differ from their well-known bipolar counter- parts. (Anyone with a special interest in this field might like to read the data books referred to at the end of this article.) To put it in a nutshell, FETs cost less than bipolar types, switch faster (in a few nanoseconds), afford higher input impedances with low drive parameters and have widely extended the range of circuit possibilities. At the time of going to press, the new DMOS transistors were still very difficult to get hold of in the retail trade and those that were to be had were far from cheap. Nevertheless, we have every reason to believe that this situation will change within the not too distant New power FETs seem to be christened almost every day: VFETs, HEXFETs, DMOS, TMOS and SIPMOS, to mention but a few. Despite their different names, they all have a great deal in common, as far as their characteristics, structure and applications are concerned. This article takes a look at power FETs in general, paying special attention to the fast -switching DMOS branch of the family. introducing DMOS power FETs FETs Even 'ordinary' MOSFETs roducing DMOS power FETs 1982-6-53 all that often, so it might be a good idea to recap on some of their features. Normally, MOSFETs have a high input impedance and a fairly average mediocre gain. They are suitable for use at high frequencies (up into the gigahertz range), but can only handle low power. Consequently, they are mainly used in receivers. Their basic operation is shown in the form of a block diagram in fig- ure 1. The source and the drain are both bonded with an n zone within a p sub- strate. Thus, as in ordinary transistors, a npn structure is involved. This may be represented as two diodes connected back-to-back, as a result of which no current is allowed to flow from drain to source. When the gate is made positive, elec- trons collect in the p material bordering the gate (electrons are negatively charged particles and are drawn by the positive gate). The p material around the gate now contains an excess number of electrons and has therefore become an n region. A channel is thus formed between source and drain consisting entirely of n doped material. Further- more, since conduction can take place, current can now flow. The higher the voltage across the gate, the wider the channel and the lower the resistance between source and drain. Figure 2 shows a VFET in cross-section. Again, a p region separates the source and drain, both of which are bonded with n regions. The principle is the same as in figure 1 : when the gate is made positive, a conductive channel is formed in the p region, allowing a current to flow between drain and source. That covers the basic operation of a VFET. The 'V', by the way, stands for vertical (the direction in which the cur- rent passes through the substrate) and has nothing to do with the V-shaped groove in the substrate. The reason why a VFET can handle high power better than an ordinary FET is purely due to its format and not to any great technological achievement. The cost of semiconductors is largely determined by the size of the chip. An ordinary, planar power FET would have to be relatively large in order to cope with the same amount of power. The area occupied by the drain con- nection has been economised on in the VFET and the drain is now situated underneath the chip. Furthermore, the channels are formed by means of dif- fusion, enabling the VFET to operate at much lower tolerance levels. The result is a much smaller chip incorporating a few thousand FETs in parallel, (as can be seen in Photo 1). Thus, it is not a question of a single VFET being able to take on an army of amps, but a whole host of them hold the fort! DMOS FETs will seem quite straight- forward in comparison. Here the gate is completely surrounded by an insulating layer of silicon dioxide (Si0 2 ) and the 6-54 - elektor June 1982 source occupies the whole upper surface. As opposed to the VFET, where the gate is embedded, the gate in the DFET juts out slightly forming a little 'bump'. In photo 1 the gate is in the shape of a square, but other patterns, such as hexagonal (HEXFETs, etc) are also possible, according to the preferences of each particular manufacturer. So much for the structure of DFETs. It should be noted that some types specifi- cally designed for audio or RF appli- cations do not follow this rule. The DMOS structure just described has a disadvantage in that the gate combines a certain amount of internal resistance with a rather large capacitance (several nano- farads). When driven with a signal in the MHz range, the gate may well get so hot under the collar that the whole FET will go up in smoke! This is where VFETs are at an advantage, for their gate can be made of aluminium, which considerably reduces the internal resist- ance. This is also the reason why DFETs are advertised as switches rather than RF components. But what you lose on the roundabout, you gain on the swings, and DFETs are able to deal with relatively high voltages. Great field intensity is produced at the bottom of the V shaped groove in VFETs and the various etching and dif- fusion processes down there are very difficult to control. Fortunately, these snags do not exist in planar DMOS FETs and the latter also have a higher break- down threshold. DFETS: do they come up to scratch? For one thing, DFETs dissipate about the same amount of power as a transis- tor in a similar package. Then there are types that can withstand up to 1000 V and others that can switch up to 25 A. As in bipolar transistors, the maximum current level may even be higher than that - for brief periods! introducing DMOS power FETs Photo 1 . A DFET consists of a large number of FETs that are connected in parallel. The square in the top I eft hand corner represents the gate, whereas the rest of the upper surface is taken up by the source (the whole top surface is plated through). Constructors are recommended to go by the Rds(on) ( = maximum on-resistance) rather than rely on the current ratings provided by the manufacturer. The lower the Rds(on) the more current the FET can handle. Be sure not to exceed the maximum dissipation rate! The gain of a FET is expressed in terms of its slope and is a couple of amps per volt, the threshold voltage being one or two volts. An example of the current voltage ratio is given in figure 4. Since a MOSFET is involved, no power is required to drive the gate, as there is no current flow. Thus, the power gain of DFETs is ideal: it is infinite! Unfor- tunately, this feature does not have any practical advantages. A fair amount of power is certainly needed during the switching process, as the gate capaci- tance of several nanofarads has to be transferred. If the capacitance transfer takes too long, in other words, if the gate is fed a slowly changing voltage, the FET will be unable to switch as fast as usual. Although the whole FET family is noted for its remarkably rapid switching capabilities (they switch cur- 4 Figure 4. These graphs show the characteristics of a FET. Similar curves apply to other members of the power transistor family. cing DMOS | FETs elektor ji 1982 - 6-55 rent in about twenty nanoseconds), this speed can only be reached provided the gate voltage is a perfect square wave. In practice, the gate voltage looks far from symmetrical, as can be seen from the (slightly exaggerated) example given in the second photograph). The top trace shows a symmetrical square wave driving a CMOS 4049 inverter. The out- put of the 4049 is connected directly to the gate of a DMOSFET (in this case a BUZ 10). The signal edges leave a lot to be desired and tend to form 'kinks' half-way down the curve. The bottom trace represents current passing through the FET. Clearly, it takes the CMOS inverter quite a while to alter the gate voltage, for the gate capacitance can only be transferred with a couple of milliamps. As the 4049 is designed as a TTL buffer, it enables more current to flow to ground than to the positive connection. Not surprisingly, the falling edge is much steeper than the rising edge. But why is the strange kink formed in both edges and why is it more pro- nounced in the slower, rising edge? Well, the gate/drain capacitance is mainly responsible for this. Figure 5 shows a simplified equivalent circuit diagram which valve lovers will immediately recognise as the 'Miller' effect. The rising voltage across the gate causes the drain voltage to drop. The signal alteration is passed on to the gate by way of the gate/drain capacitance and, as a result, the gate voltage will only be able to rise very slowly. This situation continues until the drain voltage cannot drop any further. The effect is clearly visible in Photo 1, where the gate volt- age is relatively constant while the drain voltage alters. In addition, there is almost always a certain amount of inductance in the source connection and this enhances the effect by making the source slightly negative. At a higher supply voltage, the gate/drain capaci- Photo 2. When a power FET is driven from CMOS, considerable delays are caused because the driver cannot produce enough From top to bottom: The CMOS buffer control, the waveform at the gate of the FET and the current passing through the FET. voltage. This will produce fairly even edges in the drain current. A higher voltage allows the power, required to drive the gate, rise quickly (with the square of the voltage) and does not cut down the time. same thing applies to the driving power requirements, which is why some manufacturers provide graphs showing the gate voltage waveform at different drain tance transfer will obviously take longer. In short, the actual switching time is mainly determined by the circuit driving the gate. The time achieved depends on the drain source voltage (the higher this is, the longer the process takes), on the gate capacitances (which in turn depend on the FET used) and on the driver cir- cuit (regulated by the user). Photo 3 shows a FET driven from TTL, which is a lot faster. High speed switching does, however, entail one or two difficulties. If a current of a couple of amps is flowing through the FET and is interrupted in a matter of nano- seconds, appallingly little self induction is needed in the drain network to cause a considerable peak voltage ('spikes'). The peak voltage must be added to the supply voltage and should the sum exceed the drain source voltage rating of the FET, the transistor will 'kick the bucket' at once. The solution is to con- struct the circuit carefully and connect a freewheeling diode to the power supply. Alternatively, a zener diode may be connected in parallel to the FET. It is not really advisable to use an RC net- work, as a slowly decaying oscillation can rarely be avoided and, in the event of an ill-chosen RC time, it could make matters far worse! 'Spikes' in the drain voltage also affect the gate voltage by way of the drain/ gate capacitance. If the gate is driven at a high on-resistance, the maximum gate/ source voltage may easily be exceeded - and the constructor will end up having to buy a new FET. Either drive the gate with a low on-resistance and/or connect a zener diode between the gate and source. Readers will have gathered from the above that this type of power FET does not incorporate an internal protective diode (zener diode). This is not necess- ary, because of the relatively high gate capacitance, as a result of which 'spikes' can only be caused by an inordinate amount of static charge. The lack of diodes has the advantage that the con- structor can drive the gate without any compunction. Negative voltages in par- ticular will no longer present any problems (provided they are not too large). All in all, due care must be taken with regard to static charges when handling DMOSFETs! Paralleling DFETs Normally speaking, DFETs can quite easily be connected in parallel, because the semiconductor material provides greater resistance at rising temperatures. The Rds(on) will then increase. This ensures that the hottest transistor will automatically consume less current and therefore dissipate less heat. Figure 4a shows what effect this has on the graph: the maximum current is lower at a high temperature. But the opposite is true of current levels below 2A. 6-56 - elektor June 1982 introducing DMOS power FETs So far, so good. Should FETs with mismatched VqS characteristics be con- nected in parallel, the FET with the minimum gate voltage will be driven 'on' first and will temporarily have to do all the work. A second problem may involve oscillation at extremely high frequencies (above 100 MHz). The con- structor should keep this in mind and try to match the Vqs levels of the FETs to within about 5% of each other. To be on the safe side, include a couple of low value resistors in each gate connection. Two birds are killed with one stone: the oscillation is suppressed and the drive potential is better distributed. Cooling DFETs are available in the same packages as bipolar transistors. They are easy to mount on a heatsink (whether they are insulated or not). Cooling is absolutely vital where FETs are involved. When we discussed how to connect two DFETs in parallel, we mentioned the fact that the Rds(on) has a positive temperature coefficient and that this was an advantage in that particular instance. Unfortunately, this behaviour certainly does not benefit dissipation, for the hotter the FET and the greater its resistance, the higher the dissipation. The result is a vicious circle: the temperature rises even further! This may lead to regenerative feedback and inevitable death of the expensive DFET. Such detrimental effects are avoided by keeping the temperature as low as possible. By cooling the transistor, the saturation voltage risk is kept to a minimum and any overheating is pre- vented. The best rule-of-thumb is simply to use a 50% larger heatsink than normal. Figure 6. Provided the switching speed parameters are not set too high, DMOS FETs manner. In figure 6a the DFET is driven directly from a CMOS gate with a supply voltage of about 10 V. In figure 6b the DFET is driven from TTL with an open collector output. In most cases, the pull-up voltage than the 5 V TTL supply. Background literature The 'HEXFET Data Book' from International Rectifier makes an excellent read. The Siliconix ’VMOS Power FETs Design Catalogue' also provides plenty of information. Then there's ITT's book on 'VMOS transistors, their features and applications'. Other titles include: 'Hitachi Power MOSFETs' by Hitachi and 'SIPMOS Power Transistor’ by Siemens. 1982-6-57 solid slate relays the modern method of switching mains Electronic relays have quite a lot of advantages over their conventional electro/mechanical counterparts; they don't spark, wear out as quickly, cause less or interference, and they require a control current of only a few milliamps. Even so, compact solid state relays are extremely rful. The ability to handle up to 8 A is not at all Solid state relays perform in exactly the same manner as conventional mechanical relays, but, as their titel would suggest, contain no moving parts. However their design is a little more critical if long term reliability is to be achieved. The solid state relay (SSR) to be described here can be used in complete safety as the control circuit is totally isolated from the load. More- over the control voltage can be varied over a wide range which is more than can be said of its mechanical counter- part. The pros and cons It can be considered that the conven- tional relay provides a near perfect solution to its job, after all, it has been with us for a long time. So why do we need to employ solid state devices? In principle both types have more in common than just the term relay. Both require relatively low control cur- rent, which need bear no comparison to the switching load. Both also 'elec- trically' isolate the control current from the load. This aspect is clearly illustrated Here the similarity ends, for the conven- tional type uses mechanical switch contacts to switch the load current. The contacts are mechanically activated by an electromagnet controlled by a low current source. The electronic relay, on the other hand uses a triac or thyristor to switch the load. In this case isolation is achieved by the use of an opto coupler. The use of electronic relays certainly removes many of the main drawbacks associated with the conventional type: Arcing, contact bounce, and wear are the downfall of the mechanical relay (MR) and cause no end of problems to designers. Unfortunately the SSR does create new ones! It cannot stand the same degree of overload that a MR can. We also have internal losses to the load voltage to contend with in critical conditions. A drop of 1 or 2 volts to the load voltage is possible, when the switch is 'closed', but this is generally not too inconvenient. However, the inability to handle even small overloads is a very important factor, which must be kept in mind at all times. This is due to the fact that the triacs or thyristors, used in the SSR, will not withstand an excessively high voltage across it. Further to this an excessively rapid increase in the load voltage will also cause the semiconduc- tor to break down. Another consider- ation is that triacs cease to conduct if the load current falls below a specific value, the 'holding current'. Zero-crossing points Now we come to real and unquestion- able advantages of the SSR over the MR. Where mains voltages are concerned it is kinder for motors, light bulbs and other equipment to be switched on at a time when the AC waveform is actually at zero. This is termed (logically enough!) the zero-crossing point. Readers will be aware, for example, that the filament resistance of an ordinary light bulb is low when cold (or switched off) and rapidly increases when the lamp is switched on. If this occurs when the mains waveform is at a peak (maxi- mum voltage) it follows that a surge current results across the lamp filament. If this happens consistently, as it often can, the life of the filament will be significantly shortened. It will now be apparent why switching on at the zero-crossing is so important. This is totally impossible with our old friend the mechanical relay. One minor disadvantage with the SSR described in this article is that the supply is never totally isolated from the equipment. This is because a semicon- ductor is used instead of an actual mechanical switch. A small leakage current through the thyristor/triac and surrounding circuits will always occur. It is so small however, that it can be discounted in most applications. A com- parison between the SSR and the MR relays is given in table 1, but it must be emphasised that this is very generalised and does not take into account, particu- lar uses where one type of relay may be far superior for a specific purpose. Isolation An inherent characteristic of the mech- anical relay is the complete isolation between the control voltage and the load voltage. The same degree of iso- 6-58 - ele 1982 lation with a SSR needs much thought, if it is to be reliable and still able to cope with a wide control voltage range. The smaller drawing in figure 1 illus- trates diagrammatically how isolation is achieved in the MR. No electrical con- tacts exists between its coil and con- In the SSR, an opto coupler provides the seperation between control and load voltages. The Elektor SSR Working from left to right we first have the input and control circuit D5, T2 and the transmiting side of the opto coupler (IC1). Next is the 'receiver' part of IC1, the zero-crossing delay switch (T1) and what can be termed the 'ignition' circuit made up of thyristor Thl and the diode bridge D1 . . . D4. Finally the brawn; triac Tri 1 , switching the load on and off. To drive the control circuit a DC voltage of 3 ... 32 V is applied to the input. The FET (field effective transistor) T2 serves as a current source for the LED within the opto coupler. A typical source current is about 5 mA, which of course will remain constant irrespec- tive of the input voltage. The value and therefore tolerance of the FET will determine the source current. Anything between 3 and 7 mA is suf- ficient. Diode D5 protects the opto coupler by ensuring the correct polarity of the control voltage. When current flows through the LED, the photo transistor (receiver of IC1) conducts, thus cutting off T1. This in turn triggers the gate of thyristor Thl by way of R5. When Thl conducts, it applies a gate current, via the diode bridge, to the triac to enable it to switch on. Now only the forward voltage of the triac (about 2 V) is present in the relay circuit. The relay is pulled in! The other important condition to be met in order for the triac to remain 'switched on', is that the load current should not be less than the hold current (approximately 60 mA). So far, it may seem that the triac switches on immediately the relay is Table 1 Comparison between mechanical (MR) and solid state relay (SSR). vibration and shock stability temperature stability logic compatibility multiway contact change-over switching isolation overload capacity (switching current) quietness of operation switching stability leakage current when off bistable types (NC/NO) drop load voltage driving capacity protection against overloaded negligible negligible Source : Siemens components report 1 5 (1977) book 5 MR poor yes yes excellent L 1982-6-59 triggered. The zero-crossing detection is in fact rather subtle and is all to do with the voltage divider R4/R2. Their values and therefore, relationship to each other ensures the opto coupler cuts off T1 when the AC voltage, rectified by the diode bridge, is below 30 V and not before! 30 V is pretty close to the zero-crossing of the AC voltage, and remember the triac can only switch on the load when T1 is cut-off. Above 30 V, even with a conducting photo transistor, the base/ emitter voltage of T1 will exceed 0.6 V because of R4 and R2. T1 therefore continues to conduct, preventing both Thl and Tril from being activated or In order to switch off the relay, obvi- ously the control current to the opto coupler (LED) has to be terminated, allowing T1 to conduct continuously. The triac will, however, continue to remain 'on' even without a gate current, as long as the load current is high enough (above 60 mA). But upon the next AC zero-crossing the load current will drop below this level switching itself off automatically, and remaining off until the next time the relay is trig- The other components ensure the safety and stability of the circuit. Resistor R3 ensures the photo transistor does not conduct until the LED is illuminated. Capacitor C2 connected to the gate of Tril prevents the triac from switching on as a result of mains borne inter- ference. The RC network R1 and Cl acts as a transient protection, also for the triac. As already mentioned an excessively rapid increase in the load voltage is enough to destroy the triac. This mani- fests itself as noise and 'spikes' in the AC waveform. Cl serves to smooth out these 'spikes' and so that Cl in itself 3a Figure 3a. Relationship between the tolerated case temperature and the load current of the triac. The load capacity reduces considerably when the temperature of 85°C is exceeded. does not become a danger to the triac, R1 limits its charging capacity. Cooling and capacity Most domestic solid state devices, such as light dimmers, contain 400 V rated components. The thyristors, triacs and diodes are often TIC106D, TIC226D and 1 N4004 types. Although for normal applications these will suffice the safety margin, is rather low, especially con- sidering that peak voltages of 320 V may have to be handled from time to time. Professional and small industrial types tend to have heavier duty com- ponents and use 600 V rated items. Obviously the choice is up to you, but, as the difference in price is only mar- ginal it is better to use the higher rated components if you can. As shown quite 3b 5 ♦ B i Irms IA| Figure 3b. The power dissipation of the triac related to the load current. Essential for choosing the correct heat sink. explicitly in the circuit diagram we strongly recommend the use of the 600 V types TIC106M, TIC226M and 1N4005. Using the values indicated for R1 and Cl, the relay will cope with a switching load of up to 1 kW. If a higher load is envisaged, then Cl should be changed for a capacitor of between 22 pF . . . 1 pF (depending on the load), with a 250 V AC or 600 V DC voltage rating stability. Switching domestic fluorescent light tubes requires something out of the ordinary, due to the self-inductance of the choke used in the starter. In this case R1 needs to be 10k, in order to increase the transient damping. The actual load capacity of the SSR is also dependent on the cooling of the triac. With good cooling (not exceeding a temperature of 85°C), the maximum current can be as high as 8 A, achieving a power handling of 1.8 kW. Without the use of any heat sink whatsoever, current is 1 A, which is still very good as it gives you 225 W to play with. For full power a heat sink with a ther- mal resistance of 4°C/W or less, is re- quired. The triac should be mounted onto it using heat conductive paste. As a matter of interest a 15°C/W type allows a load of 3 A (650 W). Constructors should not find any diffi- culty in working out the exact heat sink requirements for any particular load to be applied, for figure 3a indicates the maximum tolerated case temperature of the triac for the corresponding load currents. First subtract the highest possible environmental temperature (say 30°C or 86°F) from the maximum temperature show in the graph for the load current required. Then divide the result by the dissipation value corre- sponding to the maximum load as found from figure 3b. In order that you get the maths right here is an example. With a maximum load of 1 kW, and a nominal mains voltage the current is 4.4 A. This results in a T c maximum of 95°C (see figure 3a), and a dissipation of 7 W (see figure 3b). Allowing for an environmental tempera- ture of 30°C, the thermal resistance needed for the heat sink is calculated by using the following formula. 95°C - 30° C _ 65°C _ Q Table 2 shows the specifications of the SSR. Attention should be paid to the minimum load and leakage (maximum reversed) current values. 60 mA mini- mum load or holding current, basically means, that equipment consuming less than 1 5 W cannot be controlled accu- rately. The maximum reversed current or leakage of 1 0 mA should not present any problems in most cases, although it is enough to cause a glow in very low rated light bulbs. Construction Figure 4 shows the printed circuit board layout. The size actually allows you to cut it to any shape, within reason, required. By reducing the overall width of the board it will fit quite nicely into mains power supply case type PSC100 or PSC200 as supplied by West Hyde Developments Ltd. of Aylesbury. Care should be taken to isolate the printed circuit board as parts of it are carrying the full mains voltage. Make sure that any test leads and terminals are well insulated. Mount the heat sink somewhere unobtrusive, remember it is also conducting the mains! Just be very very careful! A careless approach may prove fatal. Resistors: R1 =47 0/ R2 = 22 k R3.R4 = 1 I Capacitors: Cl = 1 00 n/600 V (400V.se C2= lOOn Semiconductors: T1 = BC 547B T2 = BF 256A D1 . . . D4= 1 N4005 ( 1 N4004, si D5= 1N4148 IC1 = TIL 1 1 1 Tril = TIC 226M (TIC 226D, see Thl = TIC 106M (TIC106D, see 1982-6-61 and that certainly won't help you or us. Although we are negotiating, there are still distribution problems to be solved before you can receive a 'heavenly' copy of Elektor. The printed circuit board contains 4 connections; two for the control input and two for the load. Use insulated terminals mounted onto the board rather than soldering pins as this will eliviate the possibilities of arcing, short circuits and so on. Keeping the soldered joints as small as possible is also going to help, especially when mounting the opto coupler, otherwise whats the point in isolating the control voltage from the load. A variety of applications TheSSR can obviously be used wherever an MR would be used. There are so many applications that we are certainly not going to itemise them all. Irrespec- tive of the application you will find the following hints useful. If the relay is going to be used as a simple light switch, then the opto coupler becomes superfluous, as a small mains switch or miniature toggle is sufficient. Mind you the switch will have to have a minimum rating of 250 V 0.5 A. In this case IC1, D5, T2 and R3 are not needed. A single pole switch connected to the track connec- tion points for pins 4 and 5 of IC1 is all that is required. This SSR is ideal for the 6502 house- keeper (Elektor May 1982). The digital circuit of the housekeeper can be used to trigger a number of SSRs. The current source is then omitted (T2 and D5),as we are dealing with only one kind of control logic, 5 V. The opto coupler is driven directly via a resistor which is substituted for D5. By means of a wire link the drain and source track points for T2 are also connected. The value of the coupling resistor is pro- portional to the input current (between 3 ... 5 mA). With a 5 V control voltage a 680 £2 resistor is sufficient. Final remarks When dealing with any project associ- ated with the mains supply great care should be taken at all times. Make sure the outer case does not touch any of the components. Should you be using a metal case then the usual pre- cautions such as earthing and so on apply. The load supply line must include Literature: 'Switching mains-powered equipment' Elektor May 1979,p. 5-13 Waiter Briinnier: ‘Eiektronisches Lastrelais ( ELR )' Siemens Components 18 f 1980), Book 2, from p. 69 onwards Horst Schierl: 'Solid-State-Relais, ein voll- elektronisches kontaktloses Relais mit galvanischer Trennung', part 1 and 2, Siemens Bauteile Report 15 (1977), Book 5,p. 163 and book 6, from p. 198 onwards H The output unit for the polyformant Our Bumper Summer Circuits issue Over 100 circuits for rainy afternoons. 250 mA, 7.5 VA max. (non switching rates 240 V a.c. 2 A). Switch body size, excluding pins, is only 20.3 x 8.4 x 9.3 mm max. With a typical contact resistance of only 1 8 MS2 (10 mV/10 mA), this switch is sealed against flux ingress and is washable. Additionally, it has a hinged, transparent, dust cover with a function is completely digital, obviating the need for a capacitor to store the error voltage. Operating over the range * 199.9 mV, the ZN 450 also features an on-chip clock and precision reference voltage and consumes less than 35 mW of power. Apart from the more obvious uses as a DVM or multimeter, the ZN450 can equally well be applied to such devices as digital ther- mometers. pressure gauges and weighing The DVM evaluation kit is available, price Ferranti Electronics Limited Fields New Road, Chadderton, Oldham, Lancashire OL9 8NP, Telephone: 061-624 0S1S (2342 I 6-64 - elektc 1982 Hectaphone power supply Hectaphone power supply is a completely new concept in both design and construction. The outer casing of the supply has been used for housing the power supply and to meet with guides at a pitch which will enable it to ange of power supplies use toroidal have been dramatically reduced. This s the reliability of the supply and of nding equipment. Power MOSFETs are a substantially lower head- able to operate s This reduces losses significantly, especially for higher current rating. Both the AC input and DC output are fused. There is an LED DC Highams Electronic Communications Ltd., 96, Cobham Road, Wimborne. Dorset. Telephone: 0202-693514 (2297 Ml Lightweight 25 MHz bandwidth miniscope The new Ballantine 1024A mini oscilloscope, available from PPM Limited, has been de- signed to suit the needs of the field engineer. and light weight and small size have been achieved without reduction in instrument performance. The 1024A's specification is equal to laboratory bench scopes two or three times larger and heavier; it is shock and weather proof and will operate in harsh environments. The 1024A weighs 2.1 kilos and measures 87 mm x 203 mm x 220 mm. The Ballantine 1024A provides a 25 MHz bandwidth in each of its two vertical input channels. The wide 25 MHz frequency re- sponse extends 1024A use to fast signals, and the instrument has a passive delay line, so that the leading edge of fast rise pulses can be displayed when using internal triggering. The scopes are reliable and run with less than a 9°C hot-spot rise in ambients from 0° to 50°C. The containing cases are dust, splash, and EMI proof. The shock and vibration resistant CRT and solid internal construction of the 1024A make it dependable in de- manding field conditions. The Ballantine Model 1024A mini oscillo- division vertical deflection sensitivity in 9 calibrated range steps in two channels. Fre- quency response is from DC to 25 MHz at the 3 dB point. There is also X-V operation with equal amplifiers. Time base speeds are from 1 microsecond per division to 0.5 seconds per division in a 1 -2-5- 10 sequence, expandable by an X 10 magni- fier to 100 nanoseconds per division. The internal trigger sensitivity is 0.35 division from DC to 5 MHz. increasing to 2 divisions at 25 MHz. Three coupling modes, dc, ac. and ac fast can be selected on both internal and external triggers. The CRT display area is 8x10 divisions, each division equals 0.5 cm, and the 1 KV accelerating voltage gives a bright, high resolution easy -to-read trace. PPM. Hermitage Road, St. Johns. Woking, Surrey GU21 1TZ. Telephone: 04867-601 1 1 (2293 M) Mini enclosure with battery compartment OK's PacTec HP series enclosures are now available with a battery compartment for standard 9 V batteries. Called the HP-BAT- 9 V the enclosure has a removable battery 'hatch' in its back panel, together with the battery clip and lead, and, as with other en- inexpensively 'customised' to individual speci- fications. Measuring 1.12 in (hi x 3.60 In (w) x 5.75 in (d), the case is constructed of ABS pact resistance and an attractive textured appearance, and is ideal for housing all hand- Four standard colours are offered, grey, tan, black and blue, but special custom colours are also available. Other options include belt- clips. shoulder straps, wrist straps, construc- tion of UL-listed flame retarding material and EMI/RFI shielding. OK Machine & Tool t UK I Ltd, Dutton Lane, Eastleigh, Hants S05 4AA. Telephone: 0703-610944 (2287 Ml DIL switches Erg Components is to launch a major new range of dual in-line switches. These are fully sealed, have colour-coded actuators and hinged, transparent, dust covers. The range comprises 24 switches in a variety of switching configurations. All switches in the new range are designed to meet BS9565 and exceed MIL-S-83504. Switching ratings are 30 V 250 mA 7.5 VA max. (non switching 240 V typically 18 mil. Top and base sealed dual in-line switches in the SpectraDIL 023 series will be on the mar- ket soon. The efficient top and base sealing allows flow soldering and solvent cleaning without affecting switch performance. Single throw, ganged and changeover styles are included. Luton Road, Dunstable. Bedfordshire LU54LJ, England. Telephone: 0582-62241 (2288 Ml New digital multimeter A new hand-held digital multimeter, designed for applications in the computer and tele- communications testing and servicing markets, has been announced by SEI. Introduced to meet market demand for a highly portable porates two important new design features. The input terminels are at the top, enabling the operator to 'probe' the circuit under test, whilst holding the instrument in one hand. The 3’/i digit LCD display is at the base, and is sloped for easier reading. Both these design features, combined with ergonomic placing of switches, are intended to make SEI's new everyday usage. The meter is fully protected against short duration transients and will withstand 250 V RMS into any input, on any SEI's new digital multimeter covers a resist- ance range of 0 to 20 MV, with diode test facility, end a voltage range 0 to 1 KV (max) dc and 0 to 750 V RMS (max) ac. Current range is 0 to 2 A, both ac and dc, which is protected by a single 2 A fuse. The meter, which is powered by a PP3 battery, comes complete with carrying case and probes. Salford Electrical Instruments Limited, Eccles, Manchester M30 OHL, Telephone: 061-7895081 (2338 M) Templates for pcb design seen introduced by LINEX of Denmark, and fessional users who are involved in the design The templates are available in scales of 1 : 1 (one template), 2 : 1 (set of 2) and 4 : 1 (set of 4) and they contain the most commonly used figures for printed circuit layouts, circuit views and component views. Component out- lines include potentiometers, diodes, resistors. All component dimensions and terminals are and dimensions are provided with mm and 0.1 '' divisions in the respective scales. All the templates in the series are produced with ink bosses so that they can be used for tracing with technical pens. A comprehensive leaflet illustrating the templates is also available and this leaflet suggests methods and instructions on how best to use the templates. Pell tech Ltd.. Station Lane, Witney. Oxon 0X8 6YS. England. Telephone: Witney 72130/72014 1STD 0993) (2349 M) Video monitors Thandar Electronics have recently announced the introduction of a complete range of professional video monitors. Each monitor is supplied fully operational in chassis format with a choice of black and white or green phospor tubes with the option of standard or non glare screens. The range of monitors are primarily aimed at the OEM test and measurement, computer and video markets although they are ideally Designated the TV2, TV5, TV9 and TV 12 each type is very competitively priced with price breaks for both the single and multiple Thandar Electronics Ltd.. London Road, St. Ives, Huntingdon, Cambs. Telephone: 0480-64646 (2339 M) Cassette recorder for personal computers The ECR81 Enhanced Certified Recorder has been designed specifically as a storage medium for personal computer systems and incorporates a number of features which are lacking in machines designed for the audio such systems. The circuitry includes a signal peak performance. One of the problems with personal computer systems is that of achieving low cost program storage. The difficulty with using ordinary portable recorders is that the level of the output signals from most minicomputers is very low which leads to errors or loss of signal on playing back the tape. Also, tape stretching may occur with ordinary recorders and this can cause computer clock pulses to miss a list of information. The ECR81 is fitted with a long life head matched to TDK's high bias "Super Avilyn" cassette tapes. Output level is preset in the factory thus eliminating the need for 'volume control' adjustment. A 'write protect' micro- switch is fitted to protect accidental tape erasures. Controls include fast forward and re-wind tape search. Monolith Electronics Co. Ltd., 5-7 Church Street, Somerset. Telephone: 0460-74321 (2292 M) elektor— — If you experienced difficulty in obtaining this magazine take this form along to your newsagent and ask him to reserve a copy for you each month. To the newsagent: If you experienced difficulty in fulfilling our customers order, contact our distributors: Seymour Press, 334 Brixton Road, London SW9 7 AG. Surname 1# POSITIVE LIGHT SENSITIVE | UUIIilll AEROSOL LACQUER Enables YOU to produce perfect printed circuits in minutes. Method: Spray cleaned board with lacquer. When dry, place Tracker car computer ■ BRANDLEADING ELECTRONICS iow AVAILABLE IN KIT FORM JV' |, . | ii ! AT-Hifl VOYAGER Car Drive Computer NAME. I ENCLOSE CHEQUE (SI/POSTAL ORDERS FOR £ KIT REE CHEQUE NO 24 hr. Answerphone PHONEYOURORDERWITHACCESS/BARCLAYCARD A SEND ONLY SAE IF BROCHURE IS REQUIRED VOYAGER £59.95 I £119.90| I MAG I DICE 1 £9.95 | £19.90 1 PRICES INC. VAT. POSTAGE & PACKING 1982 sc/mputer w JUNIOR COMPUTER BOOK 1 a personal computer at a very reasonablf Price - UK £ 4.50 for anyone wishing to become familiar 'te opportunity to build and program Overseas £4.75 JUNIOR COMPUTER BOOK 3 single-board Junior Computer into the next, transforming the basic, complete personal computer system. Overseas £ 5.00 300 CIRCUITS for the home e. the basic to the very sophisticated. Price - UK £ 3.75 DIGIBOOK - provides a simple step-by-step introduction to the basic theory and application of digital electronics and gives clear explanations of the fundamentals of digital circuitry, backed up by experiments designed to reinforce this newly acquired knowledge. Supplied with an experimenter s PCB. Price - UK £ 5.00 Overseas £ 5.25 FORMANT - complete constructional details of the Elektor Formant Synthesiser - comes with a FREE cassette of sounds that the Formant is capable of producing together with advice on how to achieve them. Price - UK £ 4.75 Overseas £ 5.00 the piactlcd rfroduction to a pcwetlU system SC/MPUTER (1) - c microprocessor system SC/MPUTER (2) sc/mputer BOOK 75 provides ELEKTOR circuits AND CO-STARRING : Csppy the capable capacitor -P 5 i Cl E 8