39
Broadband RF preamplifier The preamplifier described is a push- pull type based on a pair of inexpensive JFETs. With a few changes to the circuit, coverage may be extended downwards to VLF and LF, or upwards to 30 MHz and even VHF (150 MHz). While many designs of broadband amplifiers cover -3 dB frequency response limits of 3 to 30 MHz, or 1 to 30 MHz, the VLF, LF or even the AM broadcast band (BCB, 540 to 1700 kHz) is rarely covered. The pre- amplifier discussed here does, offer- ing a relatively high dynamic range and third -order intercept point, fea- tures which AM BCB Dxers will value highly because of the bone -crunching signal levels put on the air by local AM BCB transmitters. One of the problems of AM BCB is that those sought-after DX stations tend to be buried under multi -kilowatt local sta- tions on adjacent channels. The present amplifier achieves a decade (10:1) response (250 kHz to 2,500 kHz). It retains the 50-Q input and output impedances that are stan- dard in RF systems and is is easily modifiable to cover other frequency ranges up to 30 MHz. As you will soon discover, a good directional antanna is a must for seri- ous LW and MW DX-ing. Some sggested antenna designs are given reference (1). For LW and MW DX-ing A PUSH-PULL AMPLIFIER The basic concept of a push-pull ampli- fier is illustrated in Figure 1. Two iden- tical amplifiers, Al and A2, each amplify one half -cycle of the input sinewave signal. At the input, phase splitting is achieved with the aid of a transformer whose secondary winding has a grounded centre tap (CT). At the output, a similar, but reverse -con- nected, transformer is used to recom- bine the signal components. Here, the centre tap is on the primary winding, and used to feed in the supply voltage. The push-pull amplifier being bal- anced, it has a very interesting prop- erty: even -order harmonics are can- celled in the output, so the output sig- nal will be cleaner than for a single -ended (unbalanced or asym- metrical) amplifier using the same active amplifier devices. PRACTICAL CIRCUIT The actual RF circuit is shown in Fig- ure 2; it is derived from a similar circuit found in Doug DeMaw's excellent book W1FB's QRP Notebook (2). The active amplifier devices, T1 and T2, are type BF256B JFETs that are intended for service up to VHF The choice of JFET is not particularly critical - the author originally used the NTE-451, but the general-purpose MPF102, or a rarer bird like the 2N4416, should also work. The JFETs sit between a pair of similar, back-to-back connected transformers, Trl and Tr2. The source bias resistor for the JFETs, R1, and its associated decou- pling capacitor, Cl, are connected to the centre tap on the secondary winding, B- C, of input transformer Trl. Similarly, the positive supply voltage (approx. 9 V) is applied through a limiting resistor, R2, to the centre tap on the primary of the output transformer, Tr2. High and low frequency supply decoupling is ensured by C3 and C2 respectively. None the less, the amplifier should be powered by a reasonably stable and filtered 9-V sup- ply. Current consumption will be of the order of a few tens of mA. POPULATING THE BOARD The PCB shown in Figure 3 is unfortu- nately not available ready-made. Pop - Design by Joseph J. Carr 14 Elektor Electronics 5/98 Elektor Electronics 5/98 A PUSH - PULL AMPLIFIER The basic concept of a push-pull ampli- fier is illustrated in Figure 1. Two iden- tical amplifiers, A1 and A2, each amplify one half-cycle of the input sinewave signal. At the input, phase splitting is achieved with the aid of a transformer whose secondary winding has a grounded centre tap (CT). At the output, a similar, but reverse-con- nected, transformer is used to recom- bine the signal components. Here, the centre tap is on the primary winding, and used to feed in the supply voltage. The push-pull amplifier being bal- anced, it has a very interesting prop- erty: even-order harmonics are can- celled in the output, so the output sig- nal will be cleaner than for a single-ended (unbalanced or asym- metrical) amplifier using the same active amplifier devices. P RACTICAL CIRCUIT The actual RF circuit is shown in Fig- ure 2; it is derived from a similar circuit found in Doug DeMaw’s excellent book W1FB’s QRP Notebook (2) . The active amplifier devices, T1 and T2, are type BF256B JFETs that are intended for service up to VHF. The choice of JFET is not particularly critical — the author originally used the NTE-451, but the general-purpose MPF102, or a rarer bird like the 2N4416, should also work. The JFETs sit between a pair of similar, back-to-back connected transformers, Tr1 and Tr2. The source bias resistor for the JFETs, R1, and its associated decou- pling capacitor, C1, are connected to the centre tap on the secondary winding, B- C, of input transformer Tr1. Similarly, the positive supply voltage (approx. 9 V) is applied through a limiting resistor, R2, to the centre tap on the primary of the output transformer, Tr2. High and low frequency supply decoupling is ensured by C3 and C2 respectively. None the less, the amplifier should be powered by a reasonably stable and filtered 9-V sup- ply. Current consumption will be of the order of a few tens of mA. P OPULATING THE BOARD The PCB shown in Figure 3 is unfortu- nately not available ready-made. Pop- The preamplifier described is a push- pull type based on a pair of inexpensive JFETs. With a few changes to the circuit, coverage may be extended downwards to VLF and LF, or upwards to 30 MHz and even VHF (150 MHz). 14 Design by Joseph J. Carr Broadband RF preamplifier For LW and MW DX-ing This article is based on circuit diagram no. 7-12 in the book Secrets of RF Circuit Design by Joseph J. Carr, published by TAB Books (McGraw-Hill), ISBN 0-07-011673-3 (pbk), 0-07-011672-5 (hc). While many designs of broadband amplifiers cover –3 dB frequency response limits of 3 to 30 MHz, or 1 to 30 MHz, the VLF, LF or even the AM broadcast band (BCB, 540 to 1700 kHz) is rarely covered. The pre- amplifier discussed here does, offer- ing a relatively high dynamic range and third-order intercept point, fea- tures which AM BCB Dxers will value highly because of the bone-crunching signal levels put on the air by local AM BCB transmitters. One of the problems of AM BCB is that those sought-after DX stations tend to be buried under multi-kilowatt local sta- tions on adjacent channels. The present amplifier achieves a decade (10:1) response (250 kHz to 2,500 kHz). It retains the 50-Ω input and output impedances that are stan- dard in RF systems and is is easily modifiable to cover other frequency ranges up to 30 MHz. As you will soon discover, a good directional antanna is a must for seri- ous LW and MW DX-ing. Some sggested antenna designs are given reference (1) .

Broadband RF preamplifier - WorldRadioHistory.Com · Elektor Electronics 5/98Elektor Electronics 5/98 A PUSH-PULL AMPLIFIER The basic concept of a push-pull ampli-fier is illustrated

  • Upload
    others

  • View
    16

  • Download
    4

Embed Size (px)

Citation preview

  • BroadbandRF preamplifier

    The preamplifierdescribed is a push-pull type based on a

    pair of inexpensiveJFETs. With a few

    changes to the circuit,coverage may be

    extended downwardsto VLF and LF, or

    upwards to 30 MHzand even VHF

    (150 MHz).

    While many designs of broadbandamplifiers cover -3 dB frequencyresponse limits of 3 to 30 MHz, or 1 to30 MHz, the VLF, LF or even the AMbroadcast band (BCB, 540 to1700 kHz) is rarely covered. The pre-amplifier discussed here does, offer-ing a relatively high dynamic rangeand third -order intercept point, fea-tures which AM BCB Dxers will valuehighly because of the bone -crunchingsignal levels put on the air by localAM BCB transmitters. One of theproblems of AM BCB is that thosesought-after DX stations tend to beburied under multi -kilowatt local sta-tions on adjacent channels.

    The present amplifier achieves adecade (10:1) response (250 kHz to2,500 kHz). It retains the 50-Q inputand output impedances that are stan-dard in RF systems and is is easilymodifiable to cover other frequencyranges up to 30 MHz.

    As you will soon discover, a gooddirectional antanna is a must for seri-ous LW and MW DX-ing. Somesggested antenna designs are givenreference (1).

    For LW and MW DX-ing

    A PUSH-PULLAMPLIFIERThe basic concept of a push-pull ampli-fier is illustrated in Figure 1. Two iden-tical amplifiers, Al and A2, eachamplify one half -cycle of the inputsinewave signal. At the input, phasesplitting is achieved with the aid of atransformer whose secondary windinghas a grounded centre tap (CT). At theoutput, a similar, but reverse -con-nected, transformer is used to recom-bine the signal components. Here, thecentre tap is on the primary winding,and used to feed in the supply voltage.

    The push-pull amplifier being bal-anced, it has a very interesting prop-erty: even -order harmonics are can-celled in the output, so the output sig-nal will be cleaner than for asingle -ended (unbalanced or asym-metrical) amplifier using the sameactive amplifier devices.

    PRACTICAL CIRCUITThe actual RF circuit is shown in Fig-ure 2; it is derived from a similar circuitfound in Doug DeMaw's excellentbook W1FB's QRP Notebook (2). The

    active amplifier devices, T1 and T2, aretype BF256B JFETs that are intended forservice up to VHF The choice of JFETis not particularly critical - the authororiginally used the NTE-451, but thegeneral-purpose MPF102, or a rarerbird like the 2N4416, should also work.

    The JFETs sit between a pair of similar,back-to-back connected transformers,Trl and Tr2. The source bias resistor forthe JFETs, R1, and its associated decou-pling capacitor, Cl, are connected to thecentre tap on the secondary winding, B-C, of input transformer Trl. Similarly,the positive supply voltage (approx. 9 V)is applied through a limiting resistor, R2,to the centre tap on the primary of theoutput transformer, Tr2. High and lowfrequency supply decoupling is ensuredby C3 and C2 respectively. None theless, the amplifier should be powered bya reasonably stable and filtered 9-V sup-ply. Current consumption will be of theorder of a few tens of mA.

    POPULATING THEBOARDThe PCB shown in Figure 3 is unfortu-nately not available ready-made. Pop -

    Design by Joseph J. Carr

    14

    This article is based on circuit diagram no. 7-12 in the book Secrets of RF Circuit Design by JosephJ. Carr, published by TAB Books (McGraw-Hill), ISBN 0-07-011673-3 (pbk), 0-07-011672-5 (hc).

    Elektor Electronics 5/98Elektor Electronics 5/98

    A P U S H - P U L LA M P L I F I E RThe basic concept of a push-pull ampli-fier is illustrated in Figure 1. Two iden-tical amplifiers, A1 and A2, eachamplify one half-cycle of the inputsinewave signal. At the input, phasesplitting is achieved with the aid of atransformer whose secondary windinghas a grounded centre tap (CT). At theoutput, a similar, but reverse-con-nected, transformer is used to recom-bine the signal components. Here, thecentre tap is on the primary winding,and used to feed in the supply voltage.

    The push-pull amplifier being bal-anced, it has a very interesting prop-erty: even-order harmonics are can-celled in the output, so the output sig-nal will be cleaner than for asingle-ended (unbalanced or asym-metrical) amplifier using the sameactive amplifier devices.

    P R A C T I C A L C I R C U I TThe actual RF circuit is shown in Fig-ure 2; it is derived from a similar circuitfound in Doug DeMaw’s excellentbook W1FB’s QRP Notebook (2). The

    active amplifier devices, T1 and T2, aretype BF256B JFETs that are intended forservice up to VHF. The choice of JFETis not particularly critical — the authororiginally used the NTE-451, but thegeneral-purpose MPF102, or a rarerbird like the 2N4416, should also work.

    The JFETs sit between a pair of similar,back-to-back connected transformers,Tr1 and Tr2. The source bias resistor forthe JFETs, R1, and its associated decou-pling capacitor, C1, are connected to thecentre tap on the secondary winding, B-C, of input transformer Tr1. Similarly,the positive supply voltage (approx. 9 V)is applied through a limiting resistor, R2,to the centre tap on the primary of theoutput transformer, Tr2. High and lowfrequency supply decoupling is ensuredby C3 and C2 respectively. None theless, the amplifier should be powered bya reasonably stable and filtered 9-V sup-ply. Current consumption will be of theorder of a few tens of mA.

    P O P U L A T I N G T H EB O A R DThe PCB shown in Figure 3 is unfortu-nately not available ready-made. Pop-

    The preamplifierdescribed is a push-pull type based on a

    pair of inexpensiveJFETs. With a few

    changes to the circuit,coverage may be

    extended downwardsto VLF and LF, or

    upwards to 30 MHzand even VHF

    (150 MHz).

    14

    Design by Joseph J. Carr

    BroadbandRF preamplifier

    For LW and MW DX-ing

    This article is based on circuit diagram no. 7-12 in the book Secrets of RF Circuit Design by JosephJ. Carr, published by TAB Books (McGraw-Hill), ISBN 0-07-011673-3 (pbk), 0-07-011672-5 (hc).

    While many designs of broadbandamplifiers cover –3 dB frequencyresponse limits of 3 to 30 MHz, or 1 to30 MHz, the VLF, LF or even the AMbroadcast band (BCB, 540 to1700 kHz) is rarely covered. The pre-amplifier discussed here does, offer-ing a relatively high dynamic rangeand third-order intercept point, fea-tures which AM BCB Dxers will valuehighly because of the bone-crunchingsignal levels put on the air by localAM BCB transmitters. One of theproblems of AM BCB is that thosesought-after DX stations tend to beburied under multi-kilowatt local sta-tions on adjacent channels.

    The present amplifier achieves adecade (10:1) response (250 kHz to2,500 kHz). It retains the 50-Ω inputand output impedances that are stan-dard in RF systems and is is easilymodifiable to cover other frequencyranges up to 30 MHz.

    As you will soon discover, a gooddirectional antanna is a must for seri-ous LW and MW DX-ing. Somesggested antenna designs are givenreference (1).

  • ulating the board should not presentproblems as only a handful of normal-sized components are involved. Theconstruction of the transformers is dis-cussed below. It is advisable to housethe preamp in a metal case. In somecases, particularly at HF and VHF (seefurther on) a screen may be requiredbetween the input and the output sothat they can not ‘see’ each other. Inthis way, you prevent oscillation orother instability.

    B U I L D I N G T H ET R A N S F O R M E R SIn the case of this project, the key tosuccess is a pair of carefully woundtransformers. By comparison, the con-struction of the rest of the amplifier isfar less critical, as long as all connec-tions are kept as short as possible.

    The cores used are binocular (‘two-hole’ or ‘pig’s snout’) ferrite beads typeBN-43-202 from Amidon Associates (3),or beads made from 4C6 material(Philips Components).

    Each transformer contains threewindings. In each case, the ‘B’ and ‘C’windings are 12 turns of 0.3-mm(SWG30; AWG30) enamelled copperwire wound in a bifilar manner. Thecoupling link on each transformer iswinding ‘A’. On the input transformer,Tr1, ‘A’ consists of 4 turns of 0.15-mm(SWG38 or AWG36) lacquered copperwire (CuL), while on Tr2 it consists oftwo turns of the same wire. The pri/secratios of the transformers (1:9 for Tr1and 36:1 for Tr2) match the source anddrain impedances to 50 Ω input andoutput impedance. For clarity’s sake,the winding and connection referenceletters and numbers at the output ofthe amplifier are shown in Figure 4a.The other drawing, Figure 4b, is forthose of you who are interested inhigher frequencies, say, up to 30 MHzor perhaps even VHF, when a ferrite(or iron-powder) ring core is used.More about this further on.

    First, let’s agree on what a ‘turn’ is.Look at Figure 5, which shows whatwe mean by one turn on a ferrite binoc-ular core (left-hand drawing) and twoturns (right-hand drawing).

    Start with the coupling winding, A,on the input transformer, Tr1. Strip,scrape or burn the lacquer for about5 mm from one end, and tin it with sol-der. Pass the wire through one of theholes of the core, across the barrierbetween the two holes, and then backthrough the other hole. This ‘U’shaped turn counts as one turn. Tomake transformer Tr1, pass the wirethrough the holes three more times, tomake four turns. The wire end shouldbe back at the same core side as thestart. Cut the wire to allow a shortlength to connect to the PCB. Removethe lacquer coating from this free end,and tin the exposed part. Now label

    these two ends ‘A1’ and‘A2’ using small piecesof paper. The primaryof Tr1 is now com-pleted.

    The secondarywindings, B-C, arewound together in thebifilar manner, and consist of 12 turnseach of 0.3-mm lacquered wire. First,twist the two wires together using anelectric drill with speed control. Taketwo pieces of wire, each approximately85 cm long, join theends together andchuck them up in anelectric drill. The otherends of the wire arealso joined, andanchored in a benchvise. Back off and pullthe wire taut. Turningthe drill (careful, thewire may break andwhip around the drillchuck), causes the twowires to twist together.

    Keep twisting until youobtain a pitch of about3 to 5 turns per cm.

    To start the sec-ondary winding,remove the insulationof both wires at oneend of the twisted pair,

    and tin the exposed parts. Label onefree wire end ‘B1’, the other, ‘C1’. Passthe wire through the hole of the coreopposite ‘A1’, and wind 12 turns. Cutthe twisted pair wires off to leave

    about 1 cm free. Tin theends as before.

    Connecting the freeends of the twistedwire requires an ohm-meter or continuitytester to see which wiregoes where. Identifythe end that belongswith ‘B1’, and label it‘B2’. The other pair is‘C1-C2’. ‘C1’ is con-nected to ‘B2’ to makethe centre tap.

    15Elektor Electronics 5/98

    Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk

    A1

    A2

    TR1 TR2

    UDC

    A

    B

    979012 - 12

    CT

    1

    Figure 1. Basic opera-tion of a push-pullamplifier. Thanks tothe balanced opera-tion, even-order har-monics are cancelledout at the output.

    T1

    BF256B

    T2

    BF256B

    R1

    27Ω

    C1

    K1D1

    1N4001

    C4

    100n

    C5

    10µ 63V

    R2

    27Ω

    C3

    100n

    C2

    10µ 63V

    979012 - 11

    TR1, TR2 = BN-43-202 Amidon;4C6 Philips

    TR1 TR2

    A

    A

    1

    2

    1

    2

    1

    B

    B2

    C

    C

    A

    A

    B

    B

    C

    C

    2

    1

    1

    2

    1

    2

    2

    Figure 2. Circuit dia-gram of the broad-band preamplifier.Provided you payattention to the designand construction ofthe transformers(using this month’sDatasheets), the pre-amp can be modifiedfor operation at VLF(down to 10 kHz orless) or VHF (up to150 MHz).

  • 17Elektor Electronics 5/98

    Transformer Tr2 ismade just like Tr1, butwith only two turns onthe coupling winding,rather than four.

    Before removing the labels andmounting the transformers onto thePCB, subject them to a final, thorough,wire-for-wire check using your ohm-meter.

    M O D I F Y I N G F O RH I G H E R A N D L O W E RF R E Q U E N C I E SA variation is to build the preamplifierfor the shortwave bands (up to30 MHz). This can be accomplishedeasily enough. First, reduce C1 to100 nF. Second, build the transformerson a toroid (ring) core rather than abinocular one. In Dough DeMaw’soriginal design (op-cit.) a type FT-37-43ferrite ring core wasused with the same12:12:2 and 12:12:4 turnsratios. Winding lac-quered wire on a ferritering core is not difficult— see Figure 4b. Coresmade from 4C6 material(from Philips) shouldalso be suitable for useabove 50 MHz, but this

    has not been tested inpractice.

    Alternatively, selecta powdered-iron core,such as the Amidon T-

    50-2 (red) or T-50-6 (yellow). About20 turns will be needed for the ‘A’ wind-ing on Tr2, and seven turns for the ‘A’winding on Tr1. You may want toexperiment with other core types andturns counts to optimise for the specificsection of the shortwave spectrum youwish to cover. A tip: use Amidon’s con-densed design information on toroidcores found on this month’s Datasheets.

    The third variation is to make theamplifier cover much lower frequen-cies (e.g., well down into the VLFregion). The principal changes neededare the cores for Tr1 and Tr2, the num-ber of turns of wire needed, and thecapacitors. The type 43 core (from Ami-

    don) will work downto 10 kHz or so, butrequires a lot moreturns to work effi-ciently in that region.The type 73 material,which is found in theBN-73-202 core, willprovide an AL value of8,500, as opposed to‘just’ 2,890 for the BN-

    43-202 device. Doubling the number ofturns in each winding is a good start-ing point for amplifiers below 200 kHz.The type 73 core works down to 1 kHz,so with a reasonable number of turnsshould work in the 20 to 100 Hz rangeas well.

    (979012-1)

    References:1. Small loop antennas for MW AM BCB, LF

    and VHF reception, Elektor ElectronicsJune and July/August 1994.

    2. Published by the ARRL, 225 Main Street,Newington, CT 06111, USA. Internet:www.arrl.org.

    3. Amidon Associates, 2216 East Gladwick,Dominguez Hills, CA 90220, USA. Inter-net: www.amidoncorp.com.

    979012-1

    (C) Segment

    C1

    C2

    C3

    C4

    C5

    D1

    H1

    H2

    H3

    H4

    IN1

    K1

    R1

    R2

    T1

    T2

    TR1

    TR2

    979012-1

    TT

    979012-1

    (C)

    Seg

    men

    t

    3COMPONENTS LIST

    Resistors:R1, R2 = 27Ω

    Capacitors:C1 = 1µF solid MKTC2,C5 = 10µF 63V radialC3, C4 = 100nF

    Semiconductors:D1 = 1N4001T1,T2 = BF256B (see text for alterna-

    tives)

    Miscellaneous:K1 = mains adaptor socket, PCB

    mountTr1,Tr2 = Binocular core, Amidon

    type BN-43-202, or Philips coremade from type 4C6 material (seetext).

    Windings:Tr1A = 4 turns 0.15 mm CuL wire

    (SWG38/AWG36)TR1B = 12 turns 0.3 mm CuL wire

    (SWG/AWG30)TR1C = 12 turns 0.3 mm CuL wire

    (SWG/AWG30)Tr2A = 2 turns 0.15 mm CuL wire

    (SWG38/AWG36)TR2B = 12 turns 0.3 mm CuL wire

    (SWG/AWG30)TR2C = 12 turns 0.3 mm CuL wire

    (SWG/AWG30)

    Figure 3. Printed cir-cuit board artwork(board not availableready-made).

    Figure 4. Naming thewire ends of the pri-mary and secondarywindings (a), andusing a toroid core tomake the transformersif you want to use thepreamp for VHF (b).On a toroid core, too,the bifilar windingtechnique is used.

    A1

    A2

    TR2

    Ub

    979012 - 13

    B1

    C2

    B2

    C1

    C1

    B2

    C2

    B1

    A2A2

    A1

    A1

    4

    979012 - 14

    5

    Figure 5. One turn(left) and two turns(right) on a binocularcore.

    Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk

  • 7?, Monitor/TVrefresh rate meterobviates eye strain

    The fieldfield frequency of a computer monitor or TVscreen is a compromise between the require-ment for as many horizontal lines as possibleand the avoidance of flicker. Such brightness,luminance or colour fluctuations if fewer than25-30 per second may cause eye strain and

    headache. It is often possible to increase the fieldfrequency and so prevent such discomforts. The

    meter described in this article enables the fieldfrequency to be determined.

    The number of complete pictures persecond on a monitor or televisionscreen is the frame frequency, which ishalf the number of rasters per second,that is. half the field frequency. Euro-pean television uses a field frequencyof 50 Hz, and American televison oneof 60 Hz (these frequencies are thesame as those of the mains supply or, inthe US, the household AC supply).Since the consequent frame frequencyof 25 Hz and 30 Hz respectively is low,most broadcast television companies

    * Definition is a measure of the resolution of thesystem which in turn depends on the number oflines per frame.

    The number of complete pictures persecond on a monitor or televisionscreen is the frame frequency, which ishalf the number of rasters per second,that is. half the field frequency. Euro-pean television uses a field frequencyof 50 Hz, and American televison oneof 60 Hz (these frequencies are thesame as those of the mains supply or, inthe US, the household AC supply).Since the consequent frame frequencyof 25 Hz and 30 Hz respectively is low,most broadcast television companies

    The field frequency of a computer monitor or TVscreen is a compromise between the require-ment for as many horizontal lines as possibleand the avoidance of flicker. Such brightness,luminance or colour fluctuations if fewer than25–30 per second may cause eye strain and

    headache. It is often possible to increase the fieldfrequency and so prevent such discomforts. The

    meter described in this article enables the fieldfrequency to be determined.

    Monitor/TVrefresh rate meter obviates eye strain

    * Definition is a measure of the resolution of thesystem which in turn depends on the number oflines per frame.

  • Visit our Web site at http://ourworld.compuserve.com/homepages/elektor uk

    1

    K2

    5_0 01510

    4_0 14 YSYNC

    3-0 0113HSYNC

    2_0 0127_0 rt.1_0 o-6

    K1

    +

    9V13mS

    14

    3

    1N4148

    02

    H

    CTRD11/101DEC

    CT=0

    IC5

    0

    3

    4

    4017 5

    9

    CT -5

    2

    L10

    5

    L11

    14

    13 IN

    5

    CTRDIV101DEC

    CTPO

    IC42

    3

    4

    4017 5

    7

    9

    CT ->5

    3

    4

    7

    10

    6

    5VC6

    :100n

    25k

    14

    4

    2

    5

    6

    7

    MDX

    IC3

    8x °7

    4051

    G8

    20

    6

    8

    10 9

    T3

    119

    !RFD9120

    1N4148U- C)

    use a system of inter-laced scanning. In thissystem, the lines of suc-cessive rasters are notsuperimposed on eachother, but are inter-laced.

    In the past, considering the avail-able bandwidth and the then currenttechnology, a frame frequency of 25 Hz(or 30 Hz) was a reasonable value.However, many people can discernfluctuations of 25 per second and they,therefore, experienced eye strain orheadache when watching television oran old computer monitor.

    Manufacturers of television setshave solved this problem by increasingthe frame frequency to 100 Hz with theaid of a digital buffer memory. Theresult is a stable picture. The solutionhad to be incorporated in the televisionreceiver because the transmission stan-dard cannot be altered.

    The situation was even worse incomputer monitors, since users spendmany hours a day in close proximity ofthe screen. Moreover, the definition* ofa monitor is much better than that of atelevision receiver. This means that theframe frequency (sometimes calledrefresh rate) of a monitor needs to bemuch higher than in a televisionreceivert. Whereas in early monitors,

    IC6

    *zie tekst* voir texte

    *see text*siehe Text

    Figure 1. Circuit dia-gram of the refreshrate meter. Measure-ments may be taken

    ivia an electronic eyeor via an electricallink.

    5V IC2e

    10

    22p

    11

    PDO

    PD1

    PD2

    PD3

    PD4

    PDS

    R136

    0'Cl

    AT90S1200

    16PC

    XT1 XT 2

    PBO

    PB1

    Pa2

    PB3

    PB4

    PB5

    PB6

    PB7

    RESET

    RS

    RIW

    5

    2

    13

    14

    15

    16

    17

    K3

    75`000000

    -0-0-00

    16-0-0

    X1

    32-758 kHz IC 2.21C2

    Er-2

    12

    IRFD9120

    modern displays use aframe frequency well inexcess of 100 Hz (at thetime of writing - early1998 - a frequentlyencountered value is132 Hz). Modern video

    cards can produce such high -valueframe frequencies with a bandwidth of250 MHz. These values ensure anabsolutely stable display.

    4

    CIRCUIT DESCRIPTIONThe meter, whose circuit diagram isshown in Figure 1, is highly suitable forquickly monitoring the frame fre-quency of a computer monitor. All thatneeds to be is to hold the unit in front

    5V

    + 4IC2f C2b

    13 IC2a 3

    C11

    E,1 061iov

    980040 - 11

    of the display and it will indicate theframe frequency used. It may also beconnected to the video output (or VGAoutput) when it will indicate the linefrequency and the field frequency.

    The circuit is based on ICi, anenhanced RISC processor from Atmel'sAVR family (see data sheets in our Jan-uary 1998 issue). The device makesavailable 15 I/0 lines, 1 Kbyte flashmemory, and 64 bytes EEPROM. It hasa number of other facilities as well,such as an 8 -bit timer and a watchdog.Its internal structure is shown in Fig-ure 2.

    This article is not the place to give adetailed analysis of the operation of the

    On the CD-ROMThis project is one of the many contained onthe CD-ROM `bilD-µC Hard&Software 97-98'.This CD-ROM contains more than 100 pro-jects resulting from the design competition inour July/August 1997 issue. On the disk canbe found all kinds of background information,including source codes of several projects,including the present one. If you wish to pro-gram the processor yourself, this CD-ROM isa must. See the Readers Services section

    1..V.r.2 mn r2NN4. ed F. PA r Jeie way red..1.

    :1 IA I. lukr:uk ckil...2": FAYNI

    use a system of inter-laced scanning. In thissystem, the lines of suc-cessive rasters are notsuperimposed on eachother, but are inter-laced.

    In the past, considering the avail-able bandwidth and the then currenttechnology, a frame frequency of 25 Hz(or 30 Hz) was a reasonable value.However, many people can discernfluctuations of 25 per second and they,therefore, experienced eye strain orheadache when watching television oran old computer monitor.

    Manufacturers of television setshave solved this problem by increasingthe frame frequency to 100 Hz with theaid of a digital buffer memory. Theresult is a stable picture. The solutionhad to be incorporated in the televisionreceiver because the transmission stan-dard cannot be altered.

    The situation was even worse incomputer monitors, since users spendmany hours a day in close proximity ofthe screen. Moreover, the definition* ofa monitor is much better than that of atelevision receiver. This means that theframe frequency (sometimes calledrefresh rate) of a monitor needs to bemuch higher than in a televisionreceiver†. Whereas in early monitors,

    modern displays use aframe frequency well inexcess of 100 Hz (at thetime of writing – early1998 – a frequentlyencountered value is132 Hz). Modern video

    cards can produce such high-valueframe frequencies with a bandwidth of250 MHz. These values ensure anabsolutely stable display.

    C I R C U I T D E S C R I P T I O NThe meter, whose circuit diagram isshown in Figure 1, is highly suitable forquickly monitoring the frame fre-quency of a computer monitor. All thatneeds to be is to hold the unit in front

    of the display and it will indicate theframe frequency used. It may also beconnected to the video output (or VGAoutput) when it will indicate the linefrequency and the field frequency.

    The circuit is based on IC1, anenhanced RISC processor from Atmel’sAVR family (see data sheets in our Jan-uary 1998 issue). The device makesavailable 15 I/O lines, 1 Kbyte flashmemory, and 64 bytes EEPROM. It hasa number of other facilities as well,such as an 8-bit timer and a watchdog.Its internal structure is shown in Fig-ure 2.

    This article is not the place to give adetailed analysis of the operation of the

    1

    Figure 1. Circuit dia-gram of the refreshrate meter. Measure-ments may be takenvia an electronic eyeor via an electricallink.

    On the CD-ROMThis project is one of the many contained onthe CD-ROM ‘µP-µC Hard&Software 97-98’.This CD-ROM contains more than 100 pro-jects resulting from the design competition inour July/August 1997 issue. On the disk canbe found all kinds of background information,including source codes of several projects,including the present one. If you wish to pro-gram the processor yourself, this CD-ROM isa must. See the Readers Services section

    Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk

  • -L-GND

    -III-XTAL1 XTAL2

    8 -BIT DATA BUS

    PROGRAMCOUNTER

    PROGRAMFLASH

    INSTRUCTIONREGISTER

    STACKPOINTER

    HARDWARESTACK

    INSTRUCTIONDECODER

    CONTROLLINES

    GENERALPURPOSE

    REGISTERS

    z

    STATUSREGISTER

    WATCHDOGTIMER

    MCU CONTROLREGISTER

    TIMING ANDCONTROL

    A

    TIMER/COUNTER

    INTERRUPTUNIT

    EEPROM

    44

    H PROGRAMMINGLOGIC

    4 SPI

    DATA REGISTERPORTB

    DATA DIR.REG. PORTS

    PORTB DRIVERS

    DATA DIR.REG. PORTD

    1111111PORTD DRIVERS

    PBS - PB7

    processor, so thedescription is restrictedto the functions it ful-fils in the presentmeter.

    When the processoris clocked at 12 MHz, it achieves anoperating speed of 12 MIPS (millioninstructions per second), which is farmore than required for the meter. Con-sequently, a clock frequency is usedthat makes simple timing possible.Crystal X1 is a readily available compo-nent with a fundamental frequency of32,768 kHz. It is used extensively in thewatch and clock industry, since its fre-quency is a multiple of 2 (215). This sim-plifies the derivation of a signal of1 Hz.

    An additional benefit of a low -fre-quency crystal is its low current drain.

    1111161,

    figure 2. Block dia-gram of the enhancedRISC processor, whichis housed in a DIL20ase.

    RESET

    980040 -12

    The oscillator isdesigned around twoseparate inverters toensure stable operationof the crystal. The clocksignal is applied to pin 5

    of IC1 via buffer IC2f.The B ports of IC1 (pins 12-19),

    which contain the signals for drivingthe liquid -crystal display (LCD), arelinked to connector K3.

    The control signals for the LC dis-play, RS, RiW and E, are available at theports PD0, PD1 and PD3.

    The external interrupt input, PD2, isused for monitoring external signals.Which signal depends on multiplexerIC3 it is the output of the electronic eye,T2, the horizontal synchronization sig-nal (reduced to 1/100th via IC4 andIC5), or the vertical synchronization

    I

    ExperimentationThe software for the processor was developed with a program that can bedownloaded from the Internet: http://www.atmel.comAt this site will be found an assembler, a debugger and software for the devel-opment board. The processor is programmed with the aid of a serial pro-grammable interface (SPI). In short, all requisite software is available free viathe Internet. The processor may be programmed with the 'Handyman' pub -

    signal. The multiplexer is controlled asappropriate via ports PD4 and PD5.

    Network Rj-C1-IC2a-IC26 providesa reset pulse when the supply isswitched on.

    The electronic eye consists of T2 anddigital buffer IC2c.

    The power supply is slightly moreelaborate than usual since it is pro-vided with an electronic on/off switch.This automatically switches off thesupply when the meter has notreceived any input signal for 30 sec-onds.

    The direct voltage output of themains adaptor is applied to capacitorC3 via transistor T3. Whether this tran-sistor is on or not depends on thepotential at its gate. When S1 ispressed, the gate becomes negativewith respect to the source and the tran-sistor comes on. The meter is then actu-ated. Shortly afterwards, IC1 enables T1via port PD6, so that, even when S1 isreleased, the supply remains on. Whenport PD6 is made low, the meterswitches itself off again.

    CONSTRUCTIONThe meter is conveniently built on thesingle -sided printed -circuit board inFigure 3.

    The on/off switch, Sj, and the elec-tronic eye, T2 must be fitted on theenclosure of the meter. All other com-ponents are soldered on to the PCB.Mind the polarity of T3.

    The inductor, L1, is easily made. Itconsists of two turns of 0.3 mm enam-elled copper wire around a small fer-rite bead.

    Connector K2 should be fastened tothe mother board with screws, nutsand washers before any of its pins aresoldered.

    There are various ways of linkingthe motherboard to the LCD board.Ribbon cable is best if a permanent linkis wanted. On the prototype, a single -row IC socket with wire wrap pins isused. The LCD has a single row of pinsthat protrude from the board at twosides. Provided everything is measuredaccurately, this type of link is easilyuncoupled.

    The display board must be fitted tothe mother board on four 10 mm spac-ers. The prototype construction isshown in Figure 4.

    When all construction work is com-pleted, connect a 9 V battery to themeter and press S. The display willthen read 'searching ...' with under-neath it 'Hz'. When a measurementvalue has been established, the displayreads 'Refresh Rate'. If the displayremains grey, it is possible that the con-trast has not been adjusted correctly.Turn P1 until some text appears on the

    processor, so thedescription is restrictedto the functions it ful-fils in the presentmeter.

    When the processoris clocked at 12 MHz, it achieves anoperating speed of 12 MIPS (millioninstructions per second), which is farmore than required for the meter. Con-sequently, a clock frequency is usedthat makes simple timing possible.Crystal X1 is a readily available compo-nent with a fundamental frequency of32,768 kHz. It is used extensively in thewatch and clock industry, since its fre-quency is a multiple of 2 (215). This sim-plifies the derivation of a signal of1 Hz.

    An additional benefit of a low-fre-quency crystal is its low current drain.

    The oscillator isdesigned around twoseparate inverters toensure stable operationof the crystal. The clocksignal is applied to pin 5

    of IC1 via buffer IC2f.The B ports of IC1 (pins 12–19),

    which contain the signals for drivingthe liquid-crystal display (LCD), arelinked to connector K3.

    The control signals for the LC dis-play, RS, R/W and E, are available at theports PD0, PD1 and PD3.

    The external interrupt input, PD2, isused for monitoring external signals.Which signal depends on multiplexerIC3 it is the output of the electronic eye,T2, the horizontal synchronization sig-nal (reduced to 1/100th via IC4 andIC5), or the vertical synchronization

    signal. The multiplexer is controlled asappropriate via ports PD4 and PD5.

    Network R1-C1-IC2a-IC2b providesa reset pulse when the supply isswitched on.

    The electronic eye consists of T2 anddigital buffer IC2c.

    The power supply is slightly moreelaborate than usual since it is pro-vided with an electronic on/off switch.This automatically switches off thesupply when the meter has notreceived any input signal for 30 sec-onds.

    The direct voltage output of themains adaptor is applied to capacitorC3 via transistor T3. Whether this tran-sistor is on or not depends on thepotential at its gate. When S1 ispressed, the gate becomes negativewith respect to the source and the tran-sistor comes on. The meter is then actu-ated. Shortly afterwards, IC1 enables T1via port PD6, so that, even when S1 isreleased, the supply remains on. Whenport PD6 is made low, the meterswitches itself off again.

    C O N S T R U C T I O NThe meter is conveniently built on thesingle-sided printed-circuit board inFigure 3.

    The on/off switch, S1, and the elec-tronic eye, T2 must be fitted on theenclosure of the meter. All other com-ponents are soldered on to the PCB.Mind the polarity of T3.

    The inductor, L1, is easily made. Itconsists of two turns of 0.3 mm enam-elled copper wire around a small fer-rite bead.

    Connector K2 should be fastened tothe mother board with screws, nutsand washers before any of its pins aresoldered.

    There are various ways of linkingthe motherboard to the LCD board.Ribbon cable is best if a permanent linkis wanted. On the prototype, a single-row IC socket with wire wrap pins isused. The LCD has a single row of pinsthat protrude from the board at twosides. Provided everything is measuredaccurately, this type of link is easilyuncoupled.

    The display board must be fitted tothe mother board on four 10 mm spac-ers. The prototype construction isshown in Figure 4.

    When all construction work is com-pleted, connect a 9 V battery to themeter and press S1. The display willthen read ‘searching …’ with under-neath it ‘Hz’. When a measurementvalue has been established, the displayreads ‘Refresh Rate’. If the displayremains grey, it is possible that the con-trast has not been adjusted correctly.Turn P1 until some text appears on the

    2

    Figure 2. Block dia-gram of the enhancedRISC processor, whichis housed in a DIL20case.

    ExperimentationThe software for the processor was developed with a program that can bedownloaded from the Internet: http://www.atmel.comAt this site will be found an assembler, a debugger and software for the devel-opment board. The processor is programmed with the aid of a serial pro-grammable interface (SPI). In short, all requisite software is available free viathe Internet. The processor may be programmed with the ‘Handyman’ pub-

  • Visit our Web site at http://ourworld.compuserve.com/homepages/elektor uk

    Parts listResistors:

    = 10 kf2.R2= 1MQ.R3 = 100 kf2.R4, R5 = 47 kf2.R6, R7 = 4.7 kf2.R8 = 3.9 kf2.R9 = 15 kf2.Pi = 25 kf2. (27 kf2.) preset poten-tiometer

    Capacitors:Ci = 10 [IF, 16 V, radial02 = 22 pFC3, 04 = 4.7 [IF, 16 V, radialC5-09 = 0.1 [IF

    Inductors:Li = 2 turns of 0.3 mm enamelled

    copper wire on a ferrite bead

    Semiconductors:Di, D2= 1N4148Tl = BC547T2 = SFH309-4 (do not use -F ver-

    sion)T3 = IRFD9120

    Integrated circuits:IC1 = AT90S1200 (Order no

    986510-1)IC2 = 74HC04IC3 = 4051104, IC5 = 4017IC6 = 78L05

    Miscellaneous:Ki = clip for 9 V batteryK2 = 15 -pole HD -sub -D connector,

    female, right-angled, for boardmounting (for link to VGA connec-tor on computer)

    K3 = 14 -way SIL header= push-button switch, 1 make

    contactX1 = crystal 32,768 kHz9 V batteryLC module, 2x16 charactersPCB Order no.980040 (see Readers

    Services towards the end of thisissue)

    Programmed controller Order no.986510-1 (see Readers Servicestowards the end of this issue)

    with the optical sensor is not the onlyway: measuring direct at the output ofthe video card is also possible. This iseffected by linking the meter via a suit-able cable to the output and waitinguntil the measurement results appearon the display.softwareThe software is contained in theprocessor. Immediately after the meterhas been switched on, a routine isstarted by which the two inputsof the meter are scanned for250 ms. If the optical sensordetects a signal, or the verticalsynchronization appears on therelevant connector, the appropriate

    3

    on the display. The measurement iscontinued as long as the signal is pre-sent at the relevant input. When it isno longer there, both inputs arescanned again for 250 ms. If no inputsignal is detected for 30 seconds, themeter switches itself off automatically.Make sure that the switching off occursonly when no light is incident on themeter. It is advisable to cover the meterwhen it is not in use.

    The source code may be found onthe CD-ROM 'OTC Hard &Software97-98' in directory NL/01.

    [980040-1]

    Figure 3. Printed -cir-cuit board for therefresh rate meter.

    Figure 4. Photograph ofthe completed proto-type. Note how the LCDboard is fitted to themother board on four

    with the optical sensor is not the onlyway: measuring direct at the output ofthe video card is also possible. This iseffected by linking the meter via a suit-able cable to the output and waitinguntil the measurement results appearon the display.softwareThe software is contained in theprocessor. Immediately after the meterhas been switched on, a routine isstarted by which the two inputsof the meter are scanned for250 ms. If the optical sensordetects a signal, or the verticalsynchronization appears on therelevant connector, the appropriate

    on the display. The measurement iscontinued as long as the signal is pre-sent at the relevant input. When it isno longer there, both inputs arescanned again for 250 ms. If no inputsignal is detected for 30 seconds, themeter switches itself off automatically.Make sure that the switching off occursonly when no light is incident on themeter. It is advisable to cover the meterwhen it is not in use.

    The source code may be found onthe CD-ROM ‘µP-µC Hard&Software97-98’ in directory NL/01.

    [980040-1]

    (C) Segment

    980040-1

    C1

    C2

    C3

    C4

    C5

    C6 C7

    C8 C9

    D1 D

    2

    H1 H2

    H3

    H4

    H5H6

    IC1 IC2

    IC3 IC4 IC5

    IC6

    K1

    K2

    K3

    L1

    P1

    R1

    R2

    R3

    R4

    R5

    R6

    R7

    R8

    R9

    S1

    T1

    T2

    T3

    X1

    -+

    980040-1

    C

    E

    (C)

    Seg

    men

    t980040-1

    3

    Figure 3. Printed-cir-cuit board for therefresh rate meter.

    Figure 4. Photograph ofthe completed proto-type. Note how the LCDboard is fitted to themother board on four

    Parts listResistors:R1 = 10 kΩR2 = 1MΩR3 = 100 kΩR4, R5 = 47 kΩR6, R7 = 4.7 kΩR8 = 3.9 kΩR9 = 15 kΩP1 = 25 kΩ (27 kΩ) preset poten-

    tiometer

    Capacitors:C1 = 10 µF, 16 V, radialC2 = 22 pFC3, C4 = 4.7 µF, 16 V, radialC5–C9 = 0.1 µF

    Inductors:L1 = 2 turns of 0.3 mm enamelled

    copper wire on a ferrite bead

    Semiconductors:D1, D2 = 1N4148T1 = BC547T2 = SFH309-4 (do not use -F ver-

    sion)T3 = IRFD9120

    Integrated circuits:IC1 = AT90S1200 (Order no

    986510-1)IC2 = 74HC04IC3 = 4051IC4, IC5 = 4017IC6 = 78L05

    Miscellaneous:K1 = clip for 9 V batteryK2 = 15-pole HD-sub-D connector,

    female, right-angled, for boardmounting (for link to VGA connec-tor on computer)

    K3 = 14-way SIL headerS1 = push-button switch, 1 make

    contactX1 = crystal 32,768 kHz9 V batteryLC module, 2×16 charactersPCB Order no.980040 (see Readers

    Services towards the end of thisissue)

    Programmed controller Order no.986510-1 (see Readers Servicestowards the end of this issue)

    Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk

  • DCF-controlledLED clock

    Replica ofRhine Tower

    Clock

    Visitors to Germany mayhave seen the world's

    largest digital clock which ishoused in a 234 metres high

    transmitter tower on thebanks of the Rhine in Dus-

    seldorf. The time is indicatedby vertically arranged lampsthat shine their light through

    `portholes' (glazed roundwindows). This articledescribes a miniature

    replica of this Rheinturmuhr(Rhine Tower Clock). It is

    controlled by an Atmelprocessor and synchro-nized by a DCF module.

    Design by D. de Mulder

    If you have never visited Diisseldorf(about 35 miles northwest of Bonn andonly 27 miles from the Dutch border)you can see photos of the world'slargest digital clock on the Internetand download a display version fromit (http://www.diisseldorfde), which isreproduced in Figure 1.

    Thirty-nine of the sixty-two win-dows along the height of the tower

    form the 'clock'. They indicate, fromtop to bottom, tens of hours, singlehours, tens of minutes, single minutes,tens of seconds, and single seconds-see Figure 2.

    CIRCUIT DESCRIPTIONIn the electronic replica of the tower inFigure 3, diodes D2-D40 show the time;

    A26 Elektor Electronics 5/98Elektor Electronics 5/98

    If you have never visited Düsseldorf(about 35 miles northwest of Bonn andonly 27 miles from the Dutch border)you can see photos of the world’slargest digital clock on the Internetand download a display version fromit (http://www.düsseldorf.de), which isreproduced in Figure 1.

    Thirty-nine of the sixty-two win-dows along the height of the tower

    form the ‘clock’. They indicate, fromtop to bottom, tens of hours, singlehours, tens of minutes, single minutes,tens of seconds, and single seconds—see Figure 2.

    C I R C U I T D E S C R I P T I O NIn the electronic replica of the tower inFigure 3, diodes D2–D40 show the time;

    Replica ofRhine Tower

    Clock

    Visitors to Germany mayhave seen the world’s

    largest digital clock which ishoused in a 234 metres high

    transmitter tower on thebanks of the Rhine in Düs-

    seldorf. The time is indicatedby vertically arranged lampsthat shine their light through

    ‘portholes’ (glazed roundwindows). This articledescribes a miniature

    replica of this Rheinturmuhr(Rhine Tower Clock). It is

    controlled by an Atmelprocessor and synchro-

    nized by a DCF module.

    26

    Design by D. de Mülder

    DCF-controlledLED clock

    f

  • Visit our Web site at http://ourworld.compuserve.com/homepages/elektor uk

    1

    Figure 1. The RhineTower Clock on whichthe present replica is

    !Lased.

    diodes D46-D48 portray the lighting ofthe revolving restaurant, and diodesD41-D45 represent aircraft warninglights.

    The arrangement of the light -emit-ting diodes on the display boardshown in Figure 4 coincides with thediagrammatic representation in Figure2, so that a faithful small-scale replicais obtained. This arrangement is, ofcourse, not necessary: the LEDs mayjust as well be arranged as a tradi-tional, horizontal bar.

    The use of a programmed Atmelcontroller saves a lot of components.Apart from the controller, IC1, andvoltage regulator IC3, only one moreIC is needed: IC2.

    Display driver IC2 is a very usefulcircuit element, which contains all thatis needed to drive an LED display inmulti -mode operation under the con-trol of a microprocessor or microcon-troller.

    The device is linked to the micro -controller via two control lines, WRITE

    and Mode, which cause either 4 -bitcontrol information or an 8 -bit (onebyte) data word to be displayed. Thesequential data words are automati-cally stored in an 8 -byte buffer at eachpositive WRITE pulse. They may be dis-played directly or in decoded form(hexadecimal or binary code -to -7 seg-ment). In the present application thedecoder is, of course, not used. Theindividual LEDs of the clock can beenabled by the microcontrollerthrough direct control of the eight seg-ment drivers.

    Normally, IC2 can control up toeight 7 -segment displays with com-mon anode in multiplex. Since the seg-

    Figure 2. The arrange-ment of the LEDs onthe display board isidentical to that of theoriginal clock.

    ment and digit lines are all linked toconnector K1, it would be possible touse a standard 7 -segment LED displayfor indicating the time (with suitablymodified software in IC1), but theresult would no longer be a replica ofthe Rhine Tower Clock.

    When the tower display is used,K1 is linked via a length of flatcableto its counter part, K2, on the displayboard. On this board, all segmentand digit lines are used to drive theLEDs. In the circuit diagram, thegroups of LEDs (with interconnectedanodes) are connected (from left toright) to the relevant digit line of thedisplay driver as shown in the table.

    The horizontal segment lines run

    0

    000000000

    0 000000

    000000000

    0 0 000000

    000000000

    } restaurant lighting

    hours: tens

    hours

    } tower lights (red)

    Iminutes: tens

    minutes

    }tower lights (red)

    Iseconds: tens

    seconds

    980035-11

    Elektor Electronics 5/98 27

    diodes D46–D48 portray the lighting ofthe revolving restaurant, and diodesD41–D45 represent aircraft warninglights.

    The arrangement of the light-emit-ting diodes on the display boardshown in Figure 4 coincides with thediagrammatic representation in Figure2, so that a faithful small-scale replicais obtained. This arrangement is, ofcourse, not necessary: the LEDs mayjust as well be arranged as a tradi-tional, horizontal bar.

    The use of a programmed Atmelcontroller saves a lot of components.Apart from the controller, IC1, andvoltage regulator IC3, only one moreIC is needed: IC2.

    Display driver IC2 is a very usefulcircuit element, which contains all thatis needed to drive an LED display inmulti-mode operation under the con-trol of a microprocessor or microcon-troller.

    The device is linked to the micro-controller via two control lines, WRITE

    and Mode, which cause either 4-bitcontrol information or an 8-bit (onebyte) data word to be displayed. Thesequential data words are automati-cally stored in an 8-byte buffer at eachpositive WRITE pulse. They may be dis-played directly or in decoded form(hexadecimal or binary code-to-7 seg-ment). In the present application thedecoder is, of course, not used. Theindividual LEDs of the clock can beenabled by the microcontrollerthrough direct control of the eight seg-ment drivers.

    Normally, IC2 can control up toeight 7-segment displays with com-mon anode in multiplex. Since the seg-

    ment and digit lines are all linked toconnector K1, it would be possible touse a standard 7-segment LED displayfor indicating the time (with suitablymodified software in IC1), but theresult would no longer be a replica ofthe Rhine Tower Clock.

    When the tower display is used,K1 is linked via a length of flatcableto its counter part, K2, on the displayboard. On this board, all segmentand digit lines are used to drive theLEDs. In the circuit diagram, thegroups of LEDs (with interconnectedanodes) are connected (from left toright) to the relevant digit line of thedisplay driver as shown in the table.

    The horizontal segment lines run

    27Elektor Electronics 5/98

    1

    Figure 1. The RhineTower Clock on whichthe present replica isbased.

    restaurant lighting

    tower lights (red)

    hours: tens

    minutes: tens

    hours

    minutes

    seconds: tens

    seconds

    tower lights (red)

    980035 - 11

    2

    Figure 2. The arrange-ment of the LEDs onthe display board isidentical to that of theoriginal clock.

    Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk

  • from top to bottom in sequence A, B,C, …, DP. This is a total of eight lines,which is, in principle, one short in thecase of the single (hours, minutes, sec-onds) LEDs. Therefore, specialarrangements have been made for thethree bottom diodes in the relevantrow, D10, D24 and D38. The anodes ofthese diodes are commoned and then

    linked to the digit 8 line (pin 18 of K1,K2 and IC2).

    On the display board, the sequenceis exactly the opposite. There, the threeLEDs occupy the top position of therelevant rows.

    Diodes D41–D48, whose anodes arelinked to the digit-7 line, are driven viathe segment lines in such a manner

    that diodes D41–D45(aircraft warning) flasheach second, while theother diodes, D46–D48(restaurant lighting)

    light continuously.Resistors R4 and R5 may serve as

    series resistors for the LEDs in case adifferent display or other kind ofexperimental circuit is connected to K1.For the same reason, pin 1 of K1 isstrapped to ground, and pin 2 carriesthe 5 V supply line. Similarly, resistorR7 is intended for experimental pur-poses: it may be used as a series resis-tor for an LED (in place of JP1) whichshows whether the supply is on ornot. If the circuit is used as a replica ofthe tower clock, the three resistors maysimply be omitted.

    28 Elektor Electronics 5/98

    7218A

    IC2

    MODE

    SEGA

    SEGB

    SEGC

    SEGD

    SEGE

    SEGF

    SEGG

    DIG1

    DIG2

    DIG3

    DIG4

    DIG5

    DIG6

    DIG7

    DIG8

    ID0

    ID1

    ID2

    ID3

    ID4

    ID5

    ID6

    ID7

    28

    19

    12

    11

    13

    14

    10

    WR

    27

    25

    24

    26

    DP15

    16

    23

    20

    17

    22

    21

    18

    6

    5

    7

    8

    9

    3

    1

    2

    4

    K1

    10

    11

    12

    13

    14

    15

    16

    17

    18

    19

    20

    1

    2

    3

    4

    5

    6

    7

    8

    9

    R1

    47

    k

    R2

    47

    k

    R6

    10

    k

    R5 R4

    X1

    12MHz

    R3

    22

    0Ω C1

    4µ763V

    D1

    C3

    47p

    C4

    47p

    C2

    4...22p

    C5

    100n

    C6

    100n

    K2

    10

    11

    12

    13

    14

    15

    16

    17

    18

    19

    20

    1

    2

    3

    4

    5

    6

    7

    8

    9

    D41

    D42

    D43

    D44

    D45

    D46

    D47

    D48

    D30

    D31

    D32

    D33

    D34

    D35

    D36

    D37

    D38

    D39

    D40

    D25

    D26

    D27

    D28

    D29

    D16

    D17

    D18

    D19

    D20

    D21

    D22

    D23

    D24

    D11

    D12

    D13

    D14

    D15

    D2

    D3

    D4

    D5

    D6

    D7

    D8

    D9

    D10

    D491N4001

    R7

    JP1 C8

    1000µ25V

    C9

    10µ63V

    C7

    100n

    7805

    IC3K5

    980035 - 12

    * **

    5V

    5V

    5V5V

    2

    1

    3

    zie tekst*see text*voir texte*siehe Text*

    INT0/P32

    INT1/P33

    89C1051/

    P10/IN+

    P11/IN–

    RXD/P30

    TXD/P31

    T0/P34

    T1/P35

    IC1

    P12

    P13

    P14

    P15

    P16

    P17

    RST

    P37

    X1X0

    20

    10

    12

    13

    14

    15

    16

    17

    18

    1911

    54

    2

    3

    1

    6

    7

    8

    9

    89C2051DCF77

    3

    Figure 3. Circuit dia-gram of the DCFclock. A DCF moduleis linked to the DCF77input.

    D41–D48 (aircraft warning lights and restaurant lighting): digit line 7

    D39–D40 (tens of hours) digit line 6

    D30–D37 (single hours) digit line 5

    D25–D29 (tens of minutes) digit line 4

    D16–D23 (single minutes) digit line 3

    D11–D15 (tens of seconds) digit line 2

    D2–D9 (single seconds) digit line 1

  • 29Elektor Electronics 5/98

    980035-1(C) Segment

    C1

    C2

    C3

    C4

    C5

    C6

    C7

    C8

    C9

    D1

    D2

    D3

    D4

    D5

    D6

    D7

    D8

    D9

    D10

    D11

    D12

    D13

    D14

    D15

    D16

    D17

    D18

    D19

    D20

    D21

    D22

    D23

    D24

    D25

    D26

    D27

    D28

    D29

    D30

    D31

    D32

    D33

    D34

    D35

    D36

    D37

    D38

    D39

    D40

    D41 D42D43

    D44D45

    D46D47 D48D

    49

    H1

    H2 H3

    H4

    IC1

    IC2

    IC3

    JP1

    K1

    K2

    K5

    R1

    R2

    R3

    R4R5

    R6

    R7

    X1

    2

    1

    3

    +

    -980035-1

    980035-1(C) Segment

    4

    Figure 4. Printed-circuit boards for the clock. The control and display boards are single-sided, notthrough-plated. The three boards for the display must be cut apart before any work is done.

    Apart from decoupling capacitorC6, the display drivers needs no fur-ther external components, but IC2does. Capacitors C5 and C7 (DCF mod-

    ule) are decoupling elements on thesupply line. Capacitor C1 provides apower-on reset; the resistor in serieswith it may be omitted when an

    Atmel controller is used.Diode D1 at the output (pin 9) of IC1

    lights at every correctly received DCFpulse and is, therefore, a useful aid in

    Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk

  • 5

    Figure 5. The DCF moduleboard is populated mostly withSMT components. It is linked tothe control board via a 3 -core(screened) cable.

    pointing the ferrite rod antenna of theDCF module in the right direction.

    The pulse output of the module isapplied to input pin 11 (P3.7) of ICi,whereupon it is decoded (see box).

    The DCF module is linked to thecontrol board via terminals earth (1),+ 5 V (2) and DCF pulse (3). Resistor R6is a pull-up element for the open -col-lector output of the DCF module.

    Owing to the external 12 MHzquartz crystal, the internal oscillator ofIC2 is so accurate that even when theDCF synchronization fails, the clockcontinues without discernible error. Ifnecessary, however, the accuracy maybe enhanced with trimmer C2. Notethat it is not possible to use the clockwithout the DCF module since this isneeded when the circuit is first takeninto use to set the clock accurately.

    Power is provided by a 9-12 Vmains adaptor rated at 150 mA. Theoutput of this is connected to K5. Thecurrent drawn by the control board,incl. the DCF module, is small: about18 mA. In operation, the display boarddraws about 120 mA.

    CONSTRUCTIONBefore any work on the boards isstarted, that for the display should be

    Parts list

    Resistors:Ri, R2 = 47 kf2

    = 220 f2R4, R5, R7 = see textR6= 10 kf2

    Capacitors:Ci = 4.7 [IF, 63 V, radialC2 = 4-22 pF trimmer03, C4 = 47 pFC5-07 = 0.1 [IFC8= 1000 [IF, 25 VC9 = 10 [IF, 63 V, radial

    Semiconductors:D1, D46-D48 = LED, high efficiency,green

    D2-D40 = LED, high efficiency, yellowD4i-D45 = LED, high efficiency, redD49 = 1N4001

    Integrated circuits:ICi = 89C1051 or 89C2051 (Atmel), pro-grammed Order No. 986505

    102 = 7218A (Intersil, Harris, Plessey)IC3 = 7805

    Miscellaneous:

    PC1-PC3 = soldering pinX1 = 12 MHz quartz crystalK1, K2 = 20 -way box header with pro-tective shell for board mounting

    K5 = socket for connecting mainsadaptor for board mounting

    JP1 = 2.54 mm pin strip and pin jumper20 -way flatcable terminated into mating

    connectors for box headerHeat sink for IC3 (e.g., Fischer CK35SAfrom Dau)

    DCF module (see Stippler advert in thisissue)

    PCB with programmed controller,Order No 980035C

    Decoding processThe DCF clock is a compact design: it uses only 829bytes of machine code. The software is relatively straight-forward and consists of two dissimilar blocks. One ofthese enables the reading of serial data (via P3.7) andstoring the information in various registers. Figure 6shows the make-up of the code word transmitted by theDCF transmitter. When a valid code is read for at least twoconsecutive minutes, the data is used to set the clock.

    The controller maintains its own clock function. Aftera reset, several registers are erased and then used tostore the information as to number of hours, minutes andseconds.

    Also after a reset, timer 0 is set to the 8 -bit auto reloadmode, whereupon it generates an interrupt every 250 ps.

    At an interrupt, the content of the accumulator and theprogram status word (PSW) are stored. Two registers are

    maintained by this routine: one of them counts from 0 to100 (250 ps x 100 = 25 ms), and the other from 0 to 40(40 x 25 ms = 1 s). The resultant 1 s pulse incrementsthe registers, whereupon the clock becomes available.

    When valid DCF information has been received, thecontents of the hour register and minute register are syn-chronized.

    Reproducing the information on a display requiresmore arithmetic. First, a byte is split into two digits, afterwhich the hexadecimal code is converted into a position.For in -stance, decimal 7 has the hexadecimal code 0111and causes the seventh LED to light. The conversion iscarried out by successive comparison of the with the dig-its 0-9. When the position code has been found, it isadded to the display code via an instant instruction.

    A30 Elektor Electronics 5/98

    pointing the ferrite rod antenna of theDCF module in the right direction.

    The pulse output of the module isapplied to input pin 11 (P3.7) of IC1,whereupon it is decoded (see box).

    The DCF module is linked to thecontrol board via terminals earth (1),+5 V (2) and DCF pulse (3). Resistor R6is a pull-up element for the open-col-lector output of the DCF module.

    Owing to the external 12 MHzquartz crystal, the internal oscillator ofIC2 is so accurate that even when theDCF synchronization fails, the clockcontinues without discernible error. Ifnecessary, however, the accuracy maybe enhanced with trimmer C2. Notethat it is not possible to use the clockwithout the DCF module since this isneeded when the circuit is first takeninto use to set the clock accurately.

    Power is provided by a 9–12 Vmains adaptor rated at 150 mA. Theoutput of this is connected to K5. Thecurrent drawn by the control board,incl. the DCF module, is small: about18 mA. In operation, the display boarddraws about 120 mA.

    C O N S T R U C T I O NBefore any work on the boards isstarted, that for the display should be

    30 Elektor Electronics 5/98

    5

    Figure 5. The DCF moduleboard is populated mostly withSMT components. It is linked tothe control board via a 3-core(screened) cable.

    Decoding processThe DCF clock is a compact design: it uses only 829bytes of machine code. The software is relatively straight-forward and consists of two dissimilar blocks. One ofthese enables the reading of serial data (via P3.7) andstoring the information in various registers. Figure 6shows the make-up of the code word transmitted by theDCF transmitter. When a valid code is read for at least twoconsecutive minutes, the data is used to set the clock.

    The controller maintains its own clock function. Aftera reset, several registers are erased and then used tostore the information as to number of hours, minutes andseconds.

    Also after a reset, timer 0 is set to the 8-bit auto reloadmode, whereupon it generates an interrupt every 250 µs.

    At an interrupt, the content of the accumulator and theprogram status word (PSW) are stored. Two registers are

    maintained by this routine: one of them counts from 0 to100 (250 µs × 100 = 25 ms), and the other from 0 to 40(40 × 25 ms = 1 s). The resultant 1 s pulse incrementsthe registers, whereupon the clock becomes available.

    When valid DCF information has been received, thecontents of the hour register and minute register are syn-chronized.

    Reproducing the information on a display requiresmore arithmetic. First, a byte is split into two digits, afterwhich the hexadecimal code is converted into a position.For in-stance, decimal 7 has the hexadecimal code 0111and causes the seventh LED to light. The conversion iscarried out by successive comparison of the with the dig-its 0–9. When the position code has been found, it isadded to the display code via an instant instruction.

    Parts list

    Resistors:R1, R2 = 47 kΩR3 = 220 ΩR4, R5, R7 = see textR6 = 10 kΩ

    Capacitors:C1 = 4.7 µF, 63 V, radialC2 = 4–22 pF trimmerC3, C4 = 47 pFC5–C7 = 0.1 µFC8 = 1000 µF, 25 VC9 = 10 µF, 63 V, radial

    Semiconductors:D1, D46–D48 = LED, high efficiency,

    greenD2–D40 = LED, high efficiency, yellowD41–D45 = LED, high efficiency, redD49 = 1N4001

    Integrated circuits:IC1 = 89C1051 or 89C2051 (Atmel), pro-

    grammed Order No. 986505IC2 = 7218A (Intersil, Harris, Plessey)IC3 = 7805

    Miscellaneous:

    PC1–PC3 = soldering pinX1 = 12 MHz quartz crystalK1, K2 = 20-way box header with pro-

    tective shell for board mountingK5 = socket for connecting mains

    adaptor for board mountingJP1 = 2.54 mm pin strip and pin jumper20-way flatcable terminated into mating

    connectors for box headerHeat sink for IC3 (e.g., Fischer CK35SA

    from Dau)DCF module (see Stippler advert in this

    issue)PCB with programmed controller,

    Order No 980035C

  • visit our Web site athttp://ourworld.compuserve.com/homepages/elektor uk

    cut into separate boards as indicatedby the relevant lines.

    Populating the boards should notpresent undue difficulties. Do notoverlook the two wire bridges on thecontrol board and the many on thedisplay board.

    Owing to the closeness of thetracks, soldering must be carried withgreat care.

    Make sure that the correct polarityof all diodes and electrolytic capacitorsis observed.

    The broad tracks at the edges of thedisplay boards serve to interconnectthese boards. They are best linked bylaying them flat on the workbenchwith the track layout upward. Theconnecting wires should be as short asfeasible. The boards may be stiffenedby soldering stout wire or metal stripsat their edges. It is also possible to usesmall, metal clips available from sta-tionery suppliers.

    The DCF module is connected asshown in Figures 5 and 6. When thefour screw terminals are viewed fromabove (Figure 6), the terminals fromleft to right are: DCF output (inverted);DCF output; supply line (+ 1.2-15 V);and earth.

    Figure 5 shows the SMD-populatedtrack side of the DCF board and theconnecting leads to the control board(from left to right: earth; + 5 V, DCFoutput). If the connecting leads arelong(ish), screened 3 -core cable shouldbe used.

    The outputs of the module areopen -collector outputs of n -p -n tran-sistors that can handle voltages up to30 V and are able to switch currents ofup to 1 mA to ground.

    Do not yet insert IC1 and IC2 intotheir respective sockets, and check thecompleted boards carefully. Thenswitch on the power and check thatthe supply lines are as specified. Thepotential at the junction of IC3 and C9,

    year(8 bits)

    1

    2

    10

    month 8(5 bits) 4

    1

    day of week 2,(3 bits)

    day of month(6 bits)

    Elektor Electronics

    P3

    4 2

    hour(6 +1 bits)

    5/98

    pin 20 of the socket for IC1, pin 19 ofthe socket for IC2, and pin 2 of K1must be + 5 V ± 5%.

    Switch off the supply, insert IC1and IC2 into their sockets, and connectthe DCF module and the displayboards.

    Switch on the supply, whereuponthe restaurant lighting should come onand the aircraft warnings lights shouldstart to flash at 1 -second intervals. Atthe same time, the display shows00:00:00. The clock will set itself to thecorrect time when the DCF signal hasbeen received flawlessly for at leasttwo minutes. Faultless reception isindicated by the regular lighting ofdiode D1 at each DCF pulse.

    Owing to the multiplexing of thedisplay the reception in the immediateproximity of the control and displayboards may suffer from interference. Itmay, therefore, be necessary when tak-ing the clock into use for the first timeto place the DCF module away fromthe remainder of the unit. Alterna-tively, connect the module to theantenna via a longer, screened cable.Reception difficulties may also beencountered when the clock is close(that is, less than one metre) to a tele-vision set or similar equipment.

    When reception has been satisfac-tory for more than two minutes, thedisplay will show the correct time. Ifreception fails during the two minutes,a new period of two minutes is started.Once the clock runs, however, it needsto be synchronized with the DCF sig-nal only once or twice during everytwenty-four period.

    Figure 6. Make-up ofthe 35 -bit data wordtransmitted by DCF.

    10

    miscellaneousapplications

    2

    8

    minute(7 +1 bits)

    see text

    980035 - 13

    [980035]

    31

    cut into separate boards as indicatedby the relevant lines.

    Populating the boards should notpresent undue difficulties. Do notoverlook the two wire bridges on thecontrol board and the many on thedisplay board.

    Owing to the closeness of thetracks, soldering must be carried withgreat care.

    Make sure that the correct polarityof all diodes and electrolytic capacitorsis observed.

    The broad tracks at the edges of thedisplay boards serve to interconnectthese boards. They are best linked bylaying them flat on the workbenchwith the track layout upward. Theconnecting wires should be as short asfeasible. The boards may be stiffenedby soldering stout wire or metal stripsat their edges. It is also possible to usesmall, metal clips available from sta-tionery suppliers.

    The DCF module is connected asshown in Figures 5 and 6. When thefour screw terminals are viewed fromabove (Figure 6), the terminals fromleft to right are: DCF output (inverted);DCF output; supply line (+1.2–15 V);and earth.

    Figure 5 shows the SMD-populatedtrack side of the DCF board and theconnecting leads to the control board(from left to right: earth; +5 V, DCFoutput). If the connecting leads arelong(ish), screened 3-core cable shouldbe used.

    The outputs of the module areopen-collector outputs of n-p-n tran-sistors that can handle voltages up to30 V and are able to switch currents ofup to 1 mA to ground.

    Do not yet insert IC1 and IC2 intotheir respective sockets, and check thecompleted boards carefully. Thenswitch on the power and check thatthe supply lines are as specified. Thepotential at the junction of IC3 and C9,

    pin 20 of the socket for IC1, pin 19 ofthe socket for IC2, and pin 2 of K1must be +5 V ±5%.

    Switch off the supply, insert IC1and IC2 into their sockets, and connectthe DCF module and the displayboards.

    Switch on the supply, whereuponthe restaurant lighting should come onand the aircraft warnings lights shouldstart to flash at 1-second intervals. Atthe same time, the display shows00:00:00. The clock will set itself to thecorrect time when the DCF signal hasbeen received flawlessly for at leasttwo minutes. Faultless reception isindicated by the regular lighting ofdiode D1 at each DCF pulse.

    Owing to the multiplexing of thedisplay the reception in the immediateproximity of the control and displayboards may suffer from interference. Itmay, therefore, be necessary when tak-ing the clock into use for the first timeto place the DCF module away fromthe remainder of the unit. Alterna-tively, connect the module to theantenna via a longer, screened cable.Reception difficulties may also beencountered when the clock is close(that is, less than one metre) to a tele-vision set or similar equipment.

    When reception has been satisfac-tory for more than two minutes, thedisplay will show the correct time. Ifreception fails during the two minutes,a new period of two minutes is started.Once the clock runs, however, it needsto be synchronized with the DCF sig-nal only once or twice during everytwenty-four period.

    [980035]

    31Elektor Electronics 5/98

    6

    Figure 6. Make-up ofthe 35-bit data wordtransmitted by DCF.

    Visit our Web site athttp://ourworld.compuserve.com/homepages/elektor_uk

  • 418/433 MHzshort-range communicationlicence -exempt data transmission

    and remote control systems

    In the UK, two smallsections of the 70 -cm

    band, around418 MHz and

    433 MHz, have beenavailable for some

    time for licence -exempt wireless com-

    munication usingtype -approved low -

    power devices(LPDs), also called

    short-range devices(SRDs). In this articlewe look at the use of

    these frequencybands, and some

    general design con-siderations for SRD

    modules.

    The history of licence -exemptuse of radio transmittersstarted a few decades ago withthe private use of low -powertransmitters for radio -con-trolled models. Much later, CBradio in the 27 -MHz band waslegalised and type -approvedradios for this band werefreely available. However,the use of type -approvedradio -control transmitters aswell as that of CB rigs was tied toa licence fee and some administration.And then, the mere disappearance ofthe dreaded paperwork and, possibly,a test, to get your licence was hailed asa great liberalisation. The result of thisde -regulation on part of the govern-ment authorities was a boom in thesales of CB radios. A few years later, theovercrowded CB band was extendedwith quite a few channels, highertransmit power was allowed, as well asFM (frequency modulation).

    Although small sections of the40 MHz and 173 MHz bands (the latterexclusively in the UK) have long beenavailable for low -power devices like'stage microphones' and remote con-trols (for example, for garage doors),nothing was allowed without paying alicence -fee and proof that the equip-ment was type -approved.

    Many of the regulations, but not thetype -approvals, have been relaxed orchanged radically over the past fewyears. Initially, licence -exempt remotecontrol systems appeared for the CB27 MHz (11-m) band. The real activitydid not start however until two smallsections of the 70 -cm band, 418 MHzand 433 MHz, were 'released' for low -power type -approved devices (LPDs),also called short-range devices (SRDs).

    O

    In Europe, the national radio regula-tion authorities (in the UK : Radiocom-munications Agency, RA) have theirown say about the use of the SRDbands. In the UK, the 418 MHz sectionmay be used for telemetry transmittersand receivers, telecommand and in -building security equipment, while the433 -MHz section is only available forin -vehicle equipment including radiokeys. This is in contrast with manyother European countries, where thesection around 433 MHz is much widerand also available for all of the afore-mentioned applications, and evenvoice communications using 10-mWFM handhelds.

    In the UK, the specification withnumber MPT1340 is applicable to allLPDs using the 418 MHz and433 MHz sections of the 70 -cm band.The Radiocommunications Agency(RA) is an Executive Agency of the DTI(Department of Trade and Industry)responsible for the allocation, mainte-nance and supervision of the UK radiospectrum. The RA can be contacted atthe following address: Radiocommu-nications Agency, New King's BeamHouse, 22 Upper Ground, LondonSE1 9SA. Tel. (0171) 211 0211, fax 211

    434 Elektor Electronics 5/98Elektor Electronics 5/98

    The history of licence-exemptuse of radio transmittersstarted a few decades ago withthe private use of low-powertransmitters for radio-con-trolled models. Much later, CBradio in the 27-MHz band waslegalised and type-approvedradios for this band werefreely available. However,the use of type-approvedradio-control transmitters aswell as that of CB rigs was tied toa licence fee and some administration.And then, the mere disappearance ofthe dreaded paperwork and, possibly,a test, to get your licence was hailed asa great liberalisation. The result of thisde-regulation on part of the govern-ment authorities was a boom in thesales of CB radios. A few years later, theovercrowded CB band was extendedwith quite a few channels, highertransmit power was allowed, as well asFM (frequency modulation).

    Although small sections of the40 MHz and 173 MHz bands (the latterexclusively in the UK) have long beenavailable for low-power devices like‘stage microphones’ and remote con-trols (for example, for garage doors),nothing was allowed without paying alicence-fee and proof that the equip-ment was type-approved.

    Many of the regulations, but not thetype-approvals, have been relaxed orchanged radically over the past fewyears. Initially, licence-exempt remotecontrol systems appeared for the CB27 MHz (11-m) band. The real activitydid not start however until two smallsections of the 70-cm band, 418 MHzand 433 MHz, were ‘released’ for low-power type-approved devices (LPDs),also called short-range devices (SRDs).

    In Europe, the national radio regula-tion authorities (in the UK : Radiocom-munications Agency, RA) have theirown say about the use of the SRDbands. In the UK, the 418 MHz sectionmay be used for telemetry transmittersand receivers, telecommand and in-building security equipment, while the433-MHz section is only available forin-vehicle equipment including radiokeys. This is in contrast with manyother European countries, where thesection around 433 MHz is much widerand also available for all of the afore-mentioned applications, and evenvoice communications using 10-mWFM handhelds.

    In the UK, the specification withnumber MPT1340 is applicable to allLPDs using the 418 MHz and433 MHz sections of the 70-cm band.The Radiocommunications Agency(RA) is an Executive Agency of the DTI(Department of Trade and Industry)responsible for the allocation, mainte-nance and supervision of the UK radiospectrum. The RA can be contacted atthe following address: Radiocommu-nications Agency, New King’s BeamHouse, 22 Upper Ground, LondonSE1 9SA. Tel. (0171) 211 0211, fax 211

    In the UK, two smallsections of the 70-cm

    band, around418 MHz and

    433 MHz, have beenavailable for some

    time for licence-exempt wireless com-

    munication usingtype-approved low-

    power devices(LPDs), also called

    short-range devices(SRDs). In this articlewe look at the use of

    these frequencybands, and some

    general design con-siderations for SRD

    modules.

    34

    418/433 MHzshort-range communicationlicence-exempt data transmission

    and remote control systems

  • 0507. Internet: www.open.gov.uk/ radiocom.

    Document number I-ETS 300 220describes the type-approval require-ments for 418/433-MHz SRDs. Accord-ing to the RA, new equipment can onlybe type approved to this standard pro-vided parameter limits stated inMPT 1340 are met.

    E X - I S M F R E Q U E N C I E SA long time ago, the current SRD bandsection at 433 MHz was part of aslightly larger section reserved for ISM(industrial, scientific and medical)equipment producing RF radiation.Mainly as a result of pressure fromlicensed radio amateurs who use thispart of the band on a shared and/orsecondary basis, the use of ISM equip-ment has been phased out, and theband section is no longer identified assuch, at least not in the UK. Severalother ISM frequency bands are definedin the UK, including 167 MHz, 83 MHzand 40 MHz, all subject to strict regu-lations, the most essential of whichbeing very low ERP (effective radiatedpower) levels.

    The exact frequency allocation ofthe 418 MHz and 433 MHz SRD bandsis shown in Figures 1a and 1b. Itshould be noted that the channel divi-sion and channel widths have beenadopted by SRD manufacturers, therebeing no strict RA regulation in thisrespect.

    It is expected from radio amateursusing the 70-cm band to accept theactivity of low-power SRD users in thispart of the band and not cause inter-ference. Likewise, SRD users have tolive with interference caused by radioamateurs, or prevent interference byusing low transmission rates, surecodes, high redundancy and selectivereceivers. All of this is, of course, in thehands of the manufacturers of SRDs,because the users are not allowed tomake changes to type-approved equip-ment.

    A N E W S R D B A N DMeanwhile, because they are so small,the 418 MHz and 433 MHz SRD bandshave become quite overcrowded. Anew band, around 886 MHz, is ‘identi-fied’ by the relevant authorities for useby SRDs (Figure 2), with reference toCEPT Recommendation T/R 70-03. Inthis band, it is planned to reserve sev-eral channels exclusively for securityapplications. Some channels in the pro-posed frequency range are, however,still in use for analogue cordless tele-phone sets of the CT2 generation.

    For all SRD bands, the intention hasalways been to arrive at unified regu-lations. In the UK, however, the Radio-communications Agency “has notadopted CEPT recommendationTR 01-4 which allows general low-

    35Elektor Electronics 5/98

    1 2 3 4 1314

    15

    16

    433.

    720

    MH

    z

    434.

    120

    MH

    z

    LPD/SRD Band

    Range:Width:Channels:Raster:

    433.72 MHz to 434.12 MHz400 kHz1625 kHz

    Specification: MPT 1340For Vehicle Radio Keys only.Licence-exempt, max. 10 mW ERP

    980038 - 11a

    1 2 3 4 5 6 7 843

    3.72

    0 M

    Hz

    434.

    120

    MH

    z

    LPD/SRD Band

    Range:Width:Channels:Raster:

    417.9 MHz to 418.1 MHz200 kHz825 kHz

    Specification: MPT 1340Low-power general telemetry, telecommand and alarms.Licence-exempt, max. 0.25 mW ERP

    980038 - 11b

    1 2 4 53 7778

    79

    80

    868.

    000

    MH

    z

    870.

    000

    MH

    z

    Future SRD band

    Range:Width:Channels:Raster:

    868 MHz to 870 MHz2 MHz8025 kHz

    CEPT Recommendation T/R 70-03 refers.

    980038 - 12

    1

    Figure 1. Frequencyallocation and (man-ufacturer-proposed)channel division ofthe two 70-cm SRDbands available inthe UK.

    Figure 2. Plans areafoot to open the886-MHz band forSRD use.

    2

    Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk

  • power devices tooperate in thisband”. Hopefully,the 886 MHz SRDband will be gracedby cross-Europeanstandards, andreceive an ETS(European Telecom-munication Stan-dard).

    L P DM O D U L E SIn this country,RadioMetrix andRadioTech are themain suppliers ofready-made, type-

    approved receivers and transmittersfor short-range communications in the418/433 MHz bands. In this context, weshould also mention the activities ofthe LPRA, the Low-Power Radio Asso-ciation, who publish an interesting andhighly topical newsletter, as well asmaintain a fine Internet web site atwww.lpra.org.uk

    The ready-made, UK type-approved LPD modules fromRadioMetrix and RadioTech come in avariety of frequencies and transmitpowers, depending on your applica-tion and country of use. Modules arealso available for digital communica-tions between, say, a PC and a printer,the radio link effectively acting as avery long RS232 cable. All LPD mod-

    ules we have seen so far contain SMDparts to keep the overall size as smallas possible.

    The simplest versions of SRDs usedto rely on an amplitude-modulatedtransmitter (Figure 3) and an associ-ated regenerative receiver (Figure 4).Note that such systems are probablyno longer allowed under RA specifica-tion MPT 1340. The transmitter consistsof a one-transistor oscillator. Modula-tion is obtained by applying the datasignal to the base of the transistor. Asingle surface-acoustic wave (SAW)resonator is used as the frequency-determining element. A highly similarcircuit for experimental use was pub-lished in Elektor Electronics July/August1993, page 54. Note, however, that thisdesign is based on frequency modula-tion (FM) using two varicap diodes,while the SAW has a fine-tuningadjustment.

    The receiver shown in Figure 4 alsocontains just one transistor. It is biasedto act as a regenerative oscillator, inwhich the received antenna signalcauses the transistor to switch to highamplification, thereby automaticallyarranging the signal detection. Next,the ‘raw’ demodulated signal is ampli-fied and shaped-up by opamps. Theresult is a fairly clean digital signal atthe output of the receiver. The logic-high level is at about 2/3 of the supplyvoltage, i.e., between 3 V and 4.5 V.

    The range of the simple systemshown in Figures 3 and 4 is muchsmaller than that of more expensiveunits, mainly because of the low trans-mit power (approx. 1 mW) and the rel-

    36 Elektor Electronics 5/98

    T1

    R3

    10

    0Ω

    R4

    20

    0Ω

    R2

    6k

    8

    R1

    4k7

    C3

    3p

    C4

    7p

    C1

    470pC5

    470p

    L2

    68nH

    X1

    433,92 MHz

    DATA IN

    ANT.

    980038 - 13

    L1

    27nH

    3

    Figure 3. Circuit dia-gram of an ultra-sim-ple 433-MHz SRDtransmitter usingamplitude modulation(not type-approved byRA).

    IC1bIC1a

    R15

    47k

    R10

    100k

    R11

    100k

    R18

    10Ω

    R7

    12

    k

    R8

    68

    0Ω

    R6

    18

    k

    R5

    20

    k

    R9

    5k

    6

    R14

    3M

    C5

    33p

    C6

    1n

    C8

    1n

    C7

    2p

    T1D2

    D1

    1N

    C4

    2p

    L1

    2µH2

    C9

    C21

    47µ

    R12

    6M8

    D3

    1N4148

    C10

    4µ7

    4148

    2x

    ANT.

    DATA OUT

    *

    C12

    10µ

    980038 - 14

    VCC

    *1,5 Wdg.

    *1,5 wdgn.

    *1.5 turns

    *1,5 spires.

    4

    Figure 4. Circuit of asimple AM receivermodule using a singletransistor in the(regenerative) RF sec-tion (not type-approved by RA). Insome cases, there’san additional pream-plifier stage using asecond transistor.

  • ative insensitivity and wide-bandnature of the receiver. Moreover, ampli-tude-modulated noise is not sup-pressed in any way.

    For more demanding applications,FM (frequency modulation) is the obvi-ous alternative. Block diagrams of anLPD-type 433-MHz FM transmitterand its associated receiver are shownin Figures 5 and 6 respectively.

    The transmitter is automaticallyactuated by means of pulse edge detec-tion, and uses an accurately defined 4-ms time slot to transmit, as soon asdata pulses are detected at the input.When the data signal is removed, thetransmitter automatically returns tostandby mode afterabout 200 ms. As in thelower-spec transmitter,the frequency stability isderived from a SAW res-onator. The main advan-tage of these resonatorsis their low cost. On thedown side, they are subject to rela-tively large production tolerances, andtheir temperature stability is a far cryfrom that of a quartz crystal.

    To keep the bandwidth of the fre-quency modulated transmitter outputsignal within limits, the frequency devi-ation is limited (±2.5 kHz to ±20 kHz,depending on the SAW type and man-ufacturer). The input data rate is alsolimited (low-pass filter). The upshot isthat the highest data rate of the FMmodules is about 10 kBit/s (using a high-est modulation frequency of 5 kHz).

    The antennas used for SRDs are tra-ditionally produced in the form of aprinted-circuit board track, while ⁄-lambda flexible anten-nas are also seen occa-sionally.

    The FM receivermodule shown in theblock diagram (Figure 6)is a superheterodynedesign. Here, too, anSAW resonator is used in the oscillatorto ensure frequency stability. All of theintermediate-frequency (IF) filteringcan be done with low-cost 10.7-MHzceramic filters. Because of the possiblefrequency offset caused by the SAWresonator, a fairly large bandwidth(approx. 280 kHz) is required anyway.Most ready-made SRD receiver mod-ules are compatible with 5-V or 3-V sys-tems.

    Thanks to miniaturisation, SRD

    modules with evenhigher specifications

    are not necessarily larger, but dearerand more complex.

    Higher-spec transmitters achievebetter frequency stability thanks to theuse of a crystal-controlled synthesiser,while the harmonics suppression isalso better as a result of extensive fil-tering at the output.

    Likewise, high-end SRD receivermodules are usually double-conversiontypes using synthesiser tuning and nar-row-band IF filters. The technology

    used to manufacture these modules isthe same as found in handhelds for the70-cm amateur radio band.

    D A T A T R A N S M I S S I O NFor simple data transmission applica-tions, such as a remote control link,you need a suitable encoder at thetransmitter side, and a matchingdecoder at the receiver side. Speciallydesigned integrated circuits are avail-able like the MM57410N fromNational Semiconductor, theMC145026/MC145028 from Motorola,

    37Elektor Electronics 5/98

    data inputswitching threshold

    hysteresisFM

    modulatorRF output

    stage

    pulse edgerecognition

    Integratedantenna

    supply regulationRF carrier enable

    IN Ub = 5 - 12V980038 - 15

    Figure 5. Block dia-gram of a typical FMtransmitter module forone of the 70-cm SRDbands.

    5

    selectionRF input stage

    oscillator

    AF amplifierpulse shaper

    IF amplifierFM demodulator

    mixer

    Ub = 5V OUT980038 - 16

    Figure 6. FM receivermodules may be super-heterodynes or evendouble-conversiondesigns.

    6

    Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk

  • SW -DIPS18

    1 9 1AO DOUT

    170__020-0 10 2 16Al OSC10

    11 3 151M

    3A2 OSC20--Ct

    09 4p

    12 4 14A3 TE0-10 0-0cc 5R.-0 13 5 13A40

    14 6 12ago HT12E--A5 D2A6 D1

    15 7 110-0800 16 8 A7 DO

    5V

    START

    0

    IN OUT

    HT433-1/T

    SW -DIPS

    Figure 7. Applicationcircuit for an addressable digital link usingan SRD and frequencymodulation (FM). A

    VT

    OSC1

    OSC2

    DIN

    A4 D3HT12D

    A5 D2

    A6 D1

    A7 DO

    DATA

    5V

    or the HE8 andHT12 from Hei-land Electronic(D-48351 Ever-swinkel, Germany.Tel. +49 2582-7550,

    fax +49 2582-7887).An example of an addressable sig-

    nal transmission link for the 433 MHzSRD band is shown in Figure 7 (note:this application may not be allowed inthe UK). The encoder IC type HT12Esupplies its data directly to the modu-lation input of the transmitter. Simi-

    DATA

    HT433-1/R

    980038 - 17

    larly, the HT12D decoder IC is founddirectly at the output of the receiver.On the DIP switches in the encoderyou set the same address as in thereceiver to be addressed (multiplereceivers may be addressed by a singletransmitter).

    In addition to the receiver address,four data bits may be applied to theinput. Here, these four bits come frompush-button switches. By applying thetransmit -enable signal (/TE), theencoder is prompted to supply a 12 -bitserial word (consisting of 8 address

    Figure 8. A professional1data modem designedfor use with a 70 -cmSRD (not type -approvedby RA).

    bits, and 4 data bits) to the transmitter.The decoder receives the 12 -bit wordand extracts the first 8 bits as anaddress, and the remaining 4 bits asdata. The four data bits only appear atthe output if the received addressmatches the DIP switch setting in thedecoder. The 4 -bit dataword is firstlatched and then used to control exter-nal devices (here, LEDs are used).

    To make sure the transmissionarrives securely at the decoder, theencoder transmits the 12 -bit serialword four times each time the /TEinput is activated. The decoder with-holds the data until three identical, suc-cessive, copies have been received. TheVT output then flags the availability ofvalid data.

    This process is very well suited toslow data transmission. For higher datarates, a microcontroller is an obviousalternative to special encoder/decoderICs. Note, however, that opting for amicrocontroller (like a PIC) almostalways means that you have to writeyour own software aimed at achievingsecure and reliable transmissions.

    If data is to be exchanged betweenequipment having a serial interface,the first solution that comes to mind isoften one as adopted in the project'Long-distance IrDA link' published inElektor Electronics May 1997.

    For more demanding telemetryapplications, special data modems areemployed in combination with high -end 418 -MHz SRD modules. A trans-mission protocol is then used toimprove the data security. Often, theAX.25 protocol is employed, a spin-offof the X.25 protocol which has been inuse for several years for amateurpacket radio.

    MODULATIONTECHNIQUE -A BOTTLENECKWhile most data transmission modulesapproved for SRD use are usually saidto use 'FM', in practice the actual mod-ulation method is FSK (frequency shiftkeying). Though simple from a designand technology point of view, FSK isburdened by a large bandwidthrequirement which is the chief cause ofthe relatively short distances that can becovered. Assuming a receiver band-width of 25 kHz (at -36 dBm) is beingused for data transmission, then thehighest achievable data rate using FSKwould be a measly 500 bits per second!Consequently, professional applicationsof SRD m