Pré-visualização parcial do texto
Baixe Elektor 26 e outras Notas de estudo em PDF para Mecatrônica, somente na Docsity!
up-to-date electronics for lab and leisure ELENTEM GÊ June 1977 45P / USA $ 1.50 Iqudspeaker crossover flters-levitator precisjon fimebase o MAO ms 4 En Tt fe sã 6-04 — elektor june 1977 colofon/decoder ELBICPOI NEGONES Editor Deputy editor W. van der Horst P. Holmes J. Barendrecht G.H.K. Dam E. Krempelsauer G.H. Nachbar Technical editors : Fr. Scheel K.S.M. Walraven Art editor C. Sinke Subscriptions Mrs. A. van Meyel U.K. editorial offices, administration and advertising: 6, Stour Street, Canterbury CT1 2XZ, Kent, U.K. Tel.: Canterbury (0227)-54430. Telex: 965504. Bank: 1. Midland Bank Ltd., Canterbury, A/C no. 11014587, Sorting code 40-16-11, Giro no. 3154254. 2. U.S.A, and Canada: c/o World Way Postal Center P.O. Box 80689, Los Angeles, Cal. 90080, AIC no. 12350-04207. Assistant Manager and Advertising Editorial : R.G. Knapp : T. Emmens Elektor is published monthly on the third friday of each month. 1. U.K. and all countries except the U.S.A. and Cananda: Cover price £ 0.45. Number 27/28 (July/ August), is a double issue, “Summer Circuits”, price £ 0.90. Single copies (incl. back issues) are available by post from our Canterbury office, at £ 0.60 (surface mail) or £ 0.95 (air mail). Subscriptions for 1977, January to December incl., £ 6.25 (surface mail) or £ 11.— (air mail). Subscriptions from July/ August to December 1977 £3.05 (surface mail). 2. For the U.S.A. and Canada. Cover price $ 1.50. Number 27/28 (July/ August), is a double issue, 'Summer Circuits', price $ 3 —. Single copies (incl. back issues) $ 1.50 (surface mail) or $ 2.25 (air mail). Subscriptions for 1977, January to December incl. $ 18.— (surface mail) or $ 27.— (air mail). Subscriptions from July/ August to December 1977, $9.00 (surface mail). All prices include post & packing. Change of address. Please allow at least six weeks for change of address. Include your old address, enclosing, if possible, an address label from a recent issue. Letters should be addressed to the department concerned: TQ = Technical Queries, ADV = Advertisements, SUB = Subscriptions; ADM = Administration; ED = Editorial (articles submitted for publication etc.); EPS = Elektor printed circuit board service. For technical queries, please enclose a stamped, addressed envelope. The circuits published are for domestic use only. The submission of desígns or articles to Elektor implies permission to the publishers to alter and translate the text and design, and to use the contents in other Elektor publications and activities. The publishers cannot guarantee to return any material submitted to them. All drawing, photographs, printed circuit boards and articles published in Elektor are copyright and may not be reproduced or imitated in whole or part without prior written permission of the publishers. Patent protection may exi etc. described in this magazine. The publishers do not accept responsibility for failing to identify such patent or other protection. in respect of circuits, devices, components National advertising rates for the English edition of Elektor and/or international advertising rates for advertising at the same time in the English, Dutch and German issues are available on request. Distribution: Spotlight Magazine Distributors Ltd, Spotlight House 1, Bentwell Road, Holloway, London N7 7AX. Copyright O 1977 Elektor publishers Ltd — Canterbury. Printed in the Netherlands. What is a TUN? What is 10 n? What is the EPS service? What is the TQ service? What is a missing link Semiconductor types Very often, a large number of equivalent semiconductors exist with different type numbers. For this reason, 'abbreviated' type numbers are used in Elektor wherever possible: *:741' stand for uA741, LM741, MC641, MIC741, RM741, SN72741, etc. e 'TUP'or 'TUN' (Transistor, Universal, PNP or NPN respect- ively) stand for any low fre- quency silicon transistor that meets the following speci cations: VCEO, max lc, max he, min Prot, max fTmin. Some “TUN's are: BC107, BC108 and BC109 families; 2N3856A, 2N3859, 2N3860, 2N3904, 2N3947, 2N4124. Some 'TUP's are: BC177 and BC178 families; BC179 family with the possible exeption of BC159 and BC179; 2N2412, 2N3251, 2N3906, 2N4126, 2N4291. e 'DUS' or 'DUG! (Diode Univer- sal, Silicon or Germanium respectively) stands for any diode that meets the following specifications: r— DUS DUG 20V 35mA, 100 uA 250mW! 10pF Some 'DUS's are: BA127, BA217 BA218, BA221, BA222, BA317, BA318, BAX13, BAY61, IN914, 1N4148. Some 'DUG's are: OAB5, 0A91, 0A95, AA116. *'BC107B', 'BC237B', 'BC547B' all refer to the same 'family' of almost identical better-quality silicon transistors. In general, any other member of the same family can be used instead. BC107 (-8, -9) families: BC107 (8, -9), BC147 (-8, -9), BC207 (-8, -9), BC237 (-8, -9), BC317 (-8, -9), BC347 (8, -9), BC547 (8, -9), BC171 (-2,.3), BC182 (3, 4), BC382 (3, -4), BC437 (8, 9), BC414 BC177 (:8, -9) families: BC177 (:8, -9), BC157 (-8, -9), BC204 (-5, -6), BC307 (-8, -9), BC320 (-1, -2), BC350 (-1, -2), BC557 (-8, -9), BC251 (-2,-3), BC212 (3, 4), BC512 (-3, 4), BC261 (-2, -3), BC416. Resistor and capacitor values When giving component values, decimal points and large numbers of zeros are avoided wherever possible. The decimal point is usually replaced by one of the folloi (pico-) (nano-) (micro- (milhic) (kilo-) (mega-) (giga-) A few examples: Resistance value 2k7: 2700 92. Resistance value 470: 470 2. Capacitance value 4p7: 4.7 pF, or 0.000 000 000 004 7 F . Capacitance value 10n: this is the international way of writing 10,000 pF or .01 uF, since 1 nis 10'º farads or 1000 pF. Resistors are % Watt 5% carbon types, uniess otherwise specified. The DC working voltage of capacitors (other than electro- Iytics) is normally assumed to be at least 100 V. As a rule of thumb, a safe value is usually approxi- mately twice the DC supply voltage. ozr3rao Test voltages The DC test voltages shown are measured with a 20 kS2/V instru- ment, unless otherwise speci É: U, not V The international letter symbol “U! for voltage is often used instead of the ambiguous 'V”. *V“is normally reserved for Volts". For instance: Up = 10 V, not Vb=10V. Mains voltages No mains (power line) voltages are listed in Elektor circuits. Itis assumed that our readers know what voltage is standard in their part of the world! Readers in countries that use 60 Hz should note that Elektor circuits are designed for 50 Hz operation. This will not normally be a problem; however, in cases where the mains frequency is used for synchronisation some mo cation may be required. Technical services to readers e EPS service. Many Elektor articles include a lay-out for a printed circuit board. Some — but not all — of these boards are avail- able ready-etched and predrilled. The 'EPS print service list' in the current issue always gives a com- plete list of available boards. * Technical queries. Members of the technical staff are available to answer technical queries (relating to articles published in Elektor) by telephone on Mondays from 14.00 to 16.30. Letters with technical queries should be addressed to: Dept. TO. Please enclose a stamped, self addressed envelope; readers outside U.K, please enclose an IRC instead of stamps. e Missing link. Any important modifications to, additions to, improvements on or corrections in Elektor circuits are generally listed under the heading “Missing Link” at the earliest opportunity. AME =D ão 6-14 — elektor june 1977 formant — the elektor music synthesiser (2) fFormant= the glelktor musie synthesiser (C. Chapman) Part 2 Having discussed the basics of a synthesiser in last month's introductory article, the second part of this series gets down to some of the practical aspects, namely the keyboard and keyboard electronics. Before beginning the discussion, it must be stressed that “Formant” is not a suit- able project for the beginner. The com- plexity of the synthesiser demands a high degree of competence in soldering p.c. boards and interwiring if an unac- ceptably large number of faults are not to arise. Some knowledge of operational amplifier basics is also almost essential. Nor should the project be undertaken by anyone who does not have access to an oscilloscope, a good multimeter, and preferably a digital voltmeter. Top grade components are also a must. Where specified 1% metal film or metal oxide resistors must be used. All other resistors whould be good quality 5% car- bon film types, while capacitors (except where the capacitance demands an elec- trolytic) should be low loss, low leakage types such as polycarbonate, polyester or polystyrene. Ceramic capacitors should not be used. Semiconductors too should be first-grade devices from a reputable source, not “unmarked, un- tested” manufacturer's rejects. The temptation to save money by buying dubious components should be avoided, as unsatisfactory performance will almost certainly be the result. The synthesiser comprises two separate units, the module unit containing VCO's, filters, power supplies etc, and the manual unit containing the key- ta 520 ê, 2922120) Elano) 609%O azkeys siendo, board. These two units are interconnec- ted by cables with plug and socket terminations and can be separated for transportation or storage. The keyboard is a 37-note C to C key- board fitted with Kimber-Allen two- pole normally open contact blocks. The keyboard used in the prototype was an SKA type. This keyboard is rec- ommended for the project and the descriptions given will relate to it, though other types may also be suitable. The keyboard consists of an aluminium chassis with dimensions similar to those given in figure 1, to which the keys and return springs are assembled. The key contacts are depressed by a plastic actuator on the underside of each key, which protrudes through a hole in the chassis (see figure lc). Contact blocks are supplied separate from the key- board, and the first task is to assemble the contacts to the underside of the chassis as shown in figures 2a and 2b. The contact blocks must be spaced away from the chassis so that the actuator (A) just touches the movable contact with the key in the rest pos- ition. A strip of 3 mm thick perspex or other plastic makes a suitable spacer (F). The simplest method of fixing the con- tact blocks in position is to glue them to the plastic spacer using quick setting epoxy adhesive, taking care not to get any adhesive into the “works” of the contact block. For those preferring a more easily serviceable assembly the slot in the underside of the contact will accept a rectangular section metal strip which will clamp the contacts to the chassis using nuts and bolts. Keyboard interface The principle of the synthesiser key- board, which was briefly explained in part 1, is shown again in figure 3. The function of one contact-set on the key- board is to provide a control voltage to the voltage-controlled modules of the Figures 1a, 1b and 1c. Mechanical details of the SKA keyboard. (A) Keys. (B) Contact actuator. (C) Chassis. Figures 2a and 2b. Showing the mounting of the key contact blocks. (A) Key. (B) Actuator. (C) Chassis. (D) Contact block. (E) Divider board. (F) Spacer. | | “= a formant — the elektor music synthesiser (2) elektor june 1977 — 6-15 :|1b 12345 6789 313233343536 37 123 Sid 190 35 46 Si AR mn Fam An CRP AoA CRP Tp = jd Ah 25+3* EESRpea ER formant — the elektor music synthesiser (2) elektor june 1977 — 6-17 inherent in this simple approach. 6a Firstly, since the switch in figure 4 Kov. corresponds to a key contact of the key- board, the leakage resistance of the switch when open is the leakage resist- ance of 37 key contacts connected in parallel, which can be quite low, especially in a humid environment. This could be overcome by increasing the value of the capacitor so that it dis- charges more slowly, but it would then o take much longer to charge from the keyboard divider chain, which would result in unwanted “glissando” effects. The solution is to use a double sample- Sb and-hold circuit, as shown in figure 5. The pre S and H circuit stores the out- put of the keyboard on a small capaci- tor Cl, the output being buffered by a source follower FET Tl. Before the voltage on Cl can decay due to the key contact leakage the voltage at the source of T1 is transferred to a larger capacitor C2 by an electronic switch T2. The “off” resistance of this switch is much higher than that of the keyboard, and T3 has a high input resistance, so C2 can hold its charge for quite a long time. T2 is switched by a gating pulse controlled by the second contact set of the keyboard. The cathode of DI is normally at -14V and T2 is thus cut 7 off. When a key is depressed the gate pulse takes the cathode of DI up to Portamento +14 V and T2 turns on. Kov sra 6 Portamento control When playing up and down a scale the control voltage from the source of T3 would normally consist of a series of discrete steps, as shown in figure 6a. This would give rise to equally discrete changes of pitch, the minimum change in pitch being one semitone interval of the tempered scale, as with any other keyboard instrument. However, many 8 instruments are characterised by the ability to make continuous (smooth) changes of pitch, one example being the trombone with its slide. This style of playing is known as “portamento”. The circuit of a portamento stage is Octaves *º2[7/— Snap Emo given in figure 7. It consists simply of a “eoarse” source follower FET, preceded by an RC network that integrates the stepwise Key” output of T3 to give a much smoother change as shown in figure 6b. Pl con- trols the “smoothness” of the change. Note that, due to the FET tolerances, Ob the source resistors R2, R4 and R6 must Kov be selected on test, and this will be “Ofiser” described in part 3. em E : Ô Overall tuning, frequency - E [ox modulation and offset balance FETs connected as source followers differ in two essential respects from ideal voltage followers. Firstly, between gate and source there is always the gate- source voltage of the FET, which means that the source is always at a higher voltage than the gate. This appears as an undesirable positive offset voltage at the source of the FET, and since, in this cir- | cuit, three FETs are connected in cas- Octaves 109% “fine” 6-18 — elektor june 1977 formant — the elektor music synthesiser (2) 9 cade these offset voltages are additive. Secondly, the gain of a source follower is slightly less than unity, which means that a | V change on the gate does not produce a 1 V change at the source. The offset voltage is compensated for in the circuit of figure 8. This comprises two IC op amps. IC3 is connected as an inverting summing amplifier, while IC4 is connected as a unity gain inverter to restore the correct sense of the control voltage. A negative voltage controlled by P4 may be summed with the control voltage input (KBV”) to cancel out the positive offset voltage. The gain losses in the sample and hold and portamento stages are compensated by increasing the current through the keyboard divider chain by means of P6 in figure 3 until a control characteristic of 1 octave/volt is obtained at the out- put of IC4. The circuit of figure 8 performs two additional functions. By adding a vari- able DC voltage to the control voltage the entire tuning range of the syn- thesiser may be shifted. P2 provides an adjustment of about 5 octaves, while PS provides a fine adjustment of about * one semitone, so that the synthesiser can easily be tuned to match other instruments. | A further input is provided for fre- quency modulation, for example to DR +see text ] son [Sto Tt... T4=BEZ4S A, BF244 À ICT... ICO = 4A7AI C, MC1741 CPI (MINI DIP) D1,D2 = 1N4148] 6-20 — elektor june 1977 formant — the elektor music synthesiser (2) provide vibrato. The modulation level can be adjusted by means of P3, and with P3 fully clockwise the “sensitivity? of this input is about | octave per 500 mV. Gate circuit Because of the action of the sample and hold circuit, once a key has beem depressed the control voltage remains at the KOV output until another key is depressed. This would cause a note, once pressed, to sound indefinitely were it not for the envelope circuits that control the attack, sustain and decay of the notes. Gate pulses to control the sample-and- hold circuit and to control the envelope | shapers are derived from the second set of keyboard contacts. As shown in fig- ure 9, these are all connected in parallel and fed with 4.7 V DC from IC1. Whem a key contact closes the output of ICS immediately goes to +4.7 V. C6 charges! via P7 until its voltage exceeds the volt age on the inverting input of IC6 (IC6 functions as a comparator) when the output of IC6 will swing positive. Whem the key is released the output of ICS will become zero and C6 will discharge rapidly through D2 so that the output of IC6 will swing negative. The RC network P7/C6 provides an adjustable delay that compensates for a difference in closing time between the two sets of contacts. For example, should the gating contact close before the control voltage contact then the synthesiser would first sound a note determined by the residual voltage on) C1 in figure 5. Then when the control voltage contact closed the correct note would sound. The delay network ensures that the gate pulse is delayed until after the new control voltage has, been applied to Cl. However, since C6 discharges rapidly through D2 the gate pulse ends immediately the key is released. CS and R25 at the input tol IC5 help to suppress noise due to con- tact bounce Construction Figure 10 shows the complete circuit of the keyboard interface, while the printed circuit board and componente layout are given in figure 11. The p.c. board for the divider chain will be given next month. In the third part of the article the description of the module unit will also begin, together with wiring details for the keyboard interface. Photo 1. A Kimber-Allen contact block as used in the synthesiser. Photo 2. Closeup of the contact blocl mounted on the underside of the keyboard. Photo 3. View of the completed keyboai with the divider chain p.c. board in position. : E a ne | digibell envelope shaper elektor june 1977 — 6-21 u ground. CI will now begin to charge II & enve & through R2 and R3, so the base voltage of T2 will fall and the note will gradu- ally decay as T2 slowly turns off. DI is reverse biassed during this time and prevents Tl from influencing this part of the circuit. It may be necessary to experiment with (W. Kunz) the value of Cl to ensure that the note has decayed completely before it is abruptly cut off by the sequencing of the Digibell circuit. K This circuit is intended to improve the sound of the “Digibel”, which was [= published in Flektor No 14 (June 1976). The tones produced by the original Digibell circuit, which is reproduced in figure 1, terminate abruptly at the end of each note in the sequence. The simple envelope shaper described here causes the notes to die away exponen- =+0Ob tially, which is much more natural and acceptable to the ear. The circuit, which is given in figure 2, functions as follows: in the interval Pinti 1C4 between notes TI is tumed off so the base of T2 is biassed to about 1.5 V by do a RI, DI, R2and R3. Immediately a note PRC Es appears at pin 12 of IC7 it will be ampli- fied by T2. However, pin 11 of IC4 will go high for the duration of the note, tuning on TI which shorts RI to 1 sv — O 14[ 16) nf ale] gana EO es o e a Ic5 3 ap Bono 76168 orou oe 1€7 Mera E de bo holy 0123456789). Ic3 ofé 7442 E e =1C1=7400 ASR O N345 = C4 =7400 EE ES E RS -.. D8= DUG ls dy E E me 1/27473 : EE voou somo a EE 7490 A FEZ ceara EM n DU c7 &º saga | precision timebase for frequency counter elektor june 1977 — 6:23 sy O) A caes o fos [o so E Raio! Power supply The supply circuit uses a voltage regu- lator to keep the number of necessary components to a minimum. This IC should be fitted with a heatsink of about 19ºC/W maximum. Capacitors CI and C2 cut down switching spikes from the diodes. C4, in parallel with C3, is a decoupling capacitor which stops high frequency signals appearing on the supply line. Construction The receiver circuit and the divider circuit are mounted together on one circuit board. Figure6 shows the printed circuit board and component Figure 1. Block diagram of the precision iebase, consisting of a Droitwich receiver with PLL and a frequency divider. Figure 2a. Circuit diagram of the receiver, which consists of a selective amplifier feeding into a PLL, Figure 2b. If the signal begins to clip in the second stage, two limiting diodes may be 2b pe cl ca) laced across the tuned circuit. 2x R m [330p. AAnN9 Figure 3. Circuit diagram of the frequency RT divider, which provides the timebase fre- quencies, x see text Parts list for figures 2,3 and 6 Inductors: L1,L2=470uH C11,C13=10n C12,C18=47 u/6V P1=4k7 preset Resistors: Capacitors: R1,R5=15k C1=2n2 R2,R6=33k C2,C7 = 1 n5% ceramic R3,R7,R10=1k C3,C8 = 330 p 5% ceramic R4,R8 = 680 2 C4=164u/6V R9,R17 = 220 2 5,09 = 2u2/6 V R11,R12R14=4k7 C6=470p R13,R18=2k2 C10=47n L3 = see text and figure 5 Semiconductors: T1 = BC559C,BC179C or equ T2,T4,T5,T6 = BC547,BC107 or equ. T3 = BC557,8C177 orequ. D1=LED 1C1=567 IC2,1C3,1C4,1C5,1C6,1C7 = 7490 Miscellaneous: earphone, e.g. crystal 2 k im- pedance or greater 3 ld Er a & AREA ár E É à E SEN SE é da di RP ger A SE ar] [o] [aa] [ama] [Roe ra] [ria ro BO" opta, ota, toa onto ro a ofiita, dão Ooh 6-24 — elektor june 1977 precision timebase for frequency counter D1..D4= 4x1N4004 c a] cá) ee feio 1c1 uA7805 Aa sagas layout. A low-loss type capacitor, such as ceramic capacitors, should be chosen for CI, C2, C3, C7 and C8. The circuit board should be mounted in a metal box. Coil L3, the aerial coil proper, is wound on a ferrite rod 10 mm in diameter and 20 em long using enam- elled copper wire 0.2 mm in diameter (see figure 5). L3 and C19 should be mounted firmly in position to prevent any change in the frequency from occurring, as this circuit is highly selective. Tuning The circuit can be tuned with the aid of a portable radio receiver. A place in the house should be chosen where there is good reception from Droitwich, and the input circuit L3/C19 should be set up here. The aerial coil and capacitor may be connected to the Droitwich receiver by means of screened micro- phone cable, which should not be longer than 20 metres. Having decided upon the best position for the aerial coil and upon the length of the cable, the receiver can now be. tuned. An ecarphone (preferably with a screened cable) should be connected to the receiver, the trimmer capacitor C19 can then be adjusted until the Droitwich transmitter is heard on both receivers. Pre-set potentiometer Pl is then slowly adjusted until LED DI lights up. By quickly switching the unit on and off several times, it is possible, to check whether the receiver is tuned| correctly. If properly tuned the LED) should light up as soon as the circuit) is switched on. If this is not the case) then the potentiometer should bel readjusted. With strong reception the sound becomes poorer since limiting ce in the amplifier. However this will not, affect the timebase frequencies. If the field strength of Droitwich is so great that the signal begins to clip in T2, then! Parts list for figures 4 and 7 Resistors: R1=1509 Capacitors: c1,C2=100n C3 = 2200 u/16 V c4=470n C5=10u/6V Semiconductors: D1,D2,D3,D4 = 1N4004 D5 = LED eg. TIL209 IC1 = uA7805 or LM129 Miscellaneous: Tr = mains transtormer, 9 V, 0.5 secondary 13 6-26 — elektor june 1977 slotless model car track (3) Having described the multiplex encoder, infra-red transmitter and receiver in the previous two articles, the next important component in the control chain is the servo amplifier, which converts the variable width control pulses into a corresponding position of the steering servo. Servomechanisms of the type used in model control consist of a high grade micromotor driving two control horns through a 200: | reduction gear and a rack and pinion mechanism. A poten- tiometer mounted on the output of the gear train provides positional feedback information. Figures la and 1b show the general appearance and internal construction of a servo. In figure lb the motor and potentiometer can clearly be seen in the lower half of the case and the gear train in the upper half. The connecting cable terminating in a five-pin plug contains two leads for the motor supply and three leads for the potentiometer con- nections. Figure 2 shows a schematic representation of a servo. The function of the servo amplifier is to compare the desired position of the servo (which information is contained in the control pulse width) with the actual position of the servo as indicated by feedback from the potentiometer. The servo then applies power to the servomotor to move the servo in one direction or the other until the required and actual servo positions are the same. To conserve power current is supplied to the motor only when the servo position is to be changed (unlike some designs that require alternate current pulses of | opposite polarity to maintain a stable servo position). Block diagram The principle of operation of the servo system is shown in the block diagram of figure 3. Control pulses from the multi plex decoder are fed to a high input impedance non-inverting buffer stage at the input of the servo amplifier. The output of the buffer stage is connected to the trigger input of a monostable. When this is triggered by the positive- going edge of the control pulse its out- put goes to logic “0”. It remains in this state for a time which varies according to the position of the servo feedback potentiometer. As this potentiometer is driven by the servo gear train the monostable thus provides a reference pulse whose width is proportional to the servo position and can vary between 1 and 2 ms. The required position of the servo is thus contained in the control pulse train in the form of positive going pulses whose width can vary from 1-2 ms depending on the position of the control joystick, while the actual pos- ition of the servo is contained in a train of simultaneously negative-going pulses whose width is proportional to the actual servo position. The control pulses from the output of the buffer stage are combined with the reference pulses from the monostable in a summing circuit, with three results: 1. If the control pulse is longer than the reference pulse the output of the summing circuit will be a positive pulse whose width is equal to the difference in width between the two pulses. 2. If the control pulse is shorter than the reference pulse the output will be a negative pulse whose width is equal to the difference between the two pulses. 3. If the pulses are equal (i.e. the servo is in the correct position) the output will be zero. From here it is a fairly simple step to provide the required drive to the servo- motor, The positive pulses can be ampli- fied and used to drive the servo in the direction that increases the pulse length from the reference monostable, while the negative pulses are amplified and used to drive the servo in a direction which will reduce the pulse width of the reference monostable. The servo will thus always move so as to make the con- trol pulses and reference pulses of equal width. When the servo reaches the correct position the control and refer- ence pulses cancel and no current is fed to the servomotor. One small trick has to be employed to make the system function satisfactorily. slotless model car track (3) elektor june 1977 — 6:27 O a a da Figure 1. External and internal views of a servo. Figure 2. Schematic representation of a'servo. The feedback potentiometer is driven through a 200 : 1 reduction gear. Figure 3. Block diagram of a servo system. 1.2ms " g803-3 Since the control pulses occur only once every 25 ms it follows that the pulses from the summing circuit do also. If such short pulses were applied direct to the motor then motor and gear train friction and inertia would prevent the servo from responding. Pulses from the summing circuit are thus “stored” on two capacitors, one for the positive half of the servo amplifier and one for the negative half. This voltage can be ampli- fied and used to drive the servo even in the interval between control pulses. Damping The control system so far described is a purely “on-off” feedback control system. Unfortunately this system, as it stands, will not work satisfactorily in practice. When the servo reaches the correct pos- ition power to the motor will be cut off. However, due to its mechanical inertia the motor will not stop dead but may overrun by as much as 15 revolutions. The control loop will immediately see a positional error and will reverse the motor. When the correct position is again reached power to the motor will be cut and the motor will again overrun, and so ad infinitum. The effect of this is that the servo oscillates or “hunts” about the correct position which, while it may not be disastrous from a steering point of view unless it is extremely pro- nounced, nevertheless increases power consumption since the motor is turning almost all the time. The motor inertia can be electronically compensated by introducing a lag term into the feedback loop. This is shown as block 3 in figure 3. An output is taken from the servo amplifier and fed back to the reference monostable. When the reference pulse is longer than the con- trol pulse the feedback makes the refer- ence pulse shorter than it should be, so that power to the motor is cut off just before the servo reaches its correct pos- ition. When the reference pulse is shorter than the control pulse the feed- back lengthens the reference pulse, and again the power is cut off just before the servo reaches the correct position. In both cases the inertia of the motor will make it overrun to the correct position. Dead zone Since the servo amplifier has only a finite loop gain it is not possible for the servo to assume exactly the correct position. There is always bound to be a small but finite difference between the control pulse and the reference pulse, Once the servo has moved to within a certain tolerance from the correct position the output pulses from the summing stage become so narrow that insufficient output is available to over- come motor friction. However, these pulses can cause the motor to hum and the output transistors to overheat. The solution is to introduce a “dead zone” into the output amplifier. This means that the output transistors will conduct only when the driving signal exceeds a certain amplitude. Once the servo has moved sufficiently close to the required position power to the motor is completely cut off. The dead zone in the present design amounts to about one percent of the total servo travel. This means that, once the servo is within one percent of the required position power to the motor will be cut, or to put it another way the control joystick must be moved by at least one percent of its travel before the servo will respond. This accuracy is more than adequate for model control appli- cations. Complete circuit Figure 5 shows the complete circuit of the servo amplifier. TI is the input buffer stage. The differentiating net- work C1/R6 converts the leading edges of the control pulses into short spikes to trigger the reference monostable which slotless model car track (3) elektor june 1977 — 6-29 R+ Iz B+ tz shorter than the control pulses. The waveform at the junction of R4 and R5 exhibits positive-going spikes. The volt- age at the junction of C6 and C7 falls below 2.4 V, turning on the lower half of the output stage. Figures la to 12b show what happens when the control and reference pulses are almost equal. Figure la shows a reference pulse (lower trace) that is only 40us longer than the control pulse. The lower trace in figure 1lb shows that the voltage at the junction of C6 and C7 does not have time to reach its correct value since T4 and TS are turned on for only a short time. The input voltage to the output stage thus decays quickly and the output stage is turned on for only part of the interval Parts list to figures 5 and 13. Resistors: R1,R9,R11=10k R2,R6=47k R3=3k9 R4,R5,R7 = 4k7 R8,R16 = 5k6 R10 = 100 k R12=220 0 R13=27k R21=270 9 R22 = 390 k Capacitors: O n ceramic 7 n MKM 7 4/6 V tantalum C4,C5 = 100 n ceramic C6,C7,C8 = 242/6 V tantalum Semiconductors: Ti,T4,77=BC557B T2,73,75,76=BC547B C 557 B (AC 188) C 547 B (AC 187) D1,D2=1N4148 Miscellaneous: M1,P1 = control servo with built-in potentiometer between control pulses. Figures 12a and 12b illustrate the same situation, but in this case the reference pulses are slightly shorter than the con- trol pulses. These waveforms should serve to illus- trate the operation of the servo ampli- fier, and will act as a guide to any faultfinding (which hopefully should not be necessary). Construction As the servo amplifier has to fit inside a small car (or other model) the con- struction is necessarily extremely minia- ture, and the p.c. board and component layout are shown in figure 13. A minia- ture soldering iron and 22 SWG solder should be used. Resistors should prefer- I2 + e ably be 1/8 W. Construction should be commenced by mounting all the semi- conductors, pushing the transistors into the holes as far as possible. Next the capacitors may be mounted, and finally the resistors, which are stood on end, To obtain the neatest results it is advisable to mount the components one at a time and solder and trim the leads before mounting the next component. The numbered connections to the servo potentiometer and motor refer to servoes manufactured by Graupner. Other types of servoes may be num- bered differently. Most servoes use a 5k feedback potentiometer, but there are some which use a 1 k pot. For use with these servoes the values of R8 and R16 should be altered to 1k2. H desired, transistors T8 and T9 can be germanium types AC 188 and AC 187, since these have a lower saturation volt- age than silicon types. wu 6-30 — elektor june 1977 levitator leviltator In view of the current interest in magnetic levitation for applications such as hovertrains (see Elektor 22 p. 2-08) it was felt that a magnetic levitator would make an interesting laboratory demonstration for students. The practical difficulties of cryogenic engineering and the fate of the brass monkey led to the abandonment of the superconducting approach at an early stage of development, and the system finally decided upon was controlled ferromagnetic attraction. The system provides an amusing demonstration and an interesting introduction to feedback control systems. In order to levitate an object it is necessary to balance gravitational attrac- tion with an equal and opposite force. Anyone who has ever played with permanent magnets will know that it is possible to lift a ferromagnetic object with a magnet. Controlled levitation, however, is a different matter. For the object to float in a stable position it is necessary for the gravitational force to be balanced precisely by the attractive force of the magnet. Since the force between the magnet and the object varies as the inverse square .of the distance between them it is fairly easy to see what will happen. If the object is initially placed in such a position that the magnetic attraction is weaker than the gravi- tational attraction then the object will tend to fall away from the magnet. As it does so the magnetic force will become weaker, so the object will fall even further, and the magnetic force will become weaker still. If, on the other hand, the object is placed so that the magnetic force is stronger than the gravitational force the object will be attracted upwards. The magnetic force will become even stronger, so the object will be attracted to the magnet even more, until it finally contacts the magnet. Only if the magnetic and gravi- tational forces are exactly balanced will the object remain in position, and even then any slight disturbance will send it either up or down. In other words, the system is unstable. To make the system stable a positional feedback control system is required, which adjusts the strength of the mag- net according to the position of the object. If the object moves close to the magnet then the field strength of the o ) Paiao na 1 jo ; current ) voltage- Rs PID reflector — ams | | ] controller IC1 ...IC3 ai es at Es ! , T2..14) ) 1 À | ER : ea [es Pal | PS 75 Re ) az ferrous R13 3 object ] + e current Ê E z dependent ss E E as z E E A z É E feedback SE MARE RO $ ZE nem ts E 8 e GE 5 ———————O) uv I—O 12v O) a mi supply earth IC7, T5 T6, T7, IC6 A —S 12V di 
