'Err 1,rcochannel interference
suppression using stacked antennasa clever solution
for VHF/UHF radio and TV
If you want to receive a dis-tant or even overseas TV sta-tion that happens to use thesame frequency as a strong
local transmitter, the receivedsignals will be weak andalways suffer from ghost
effects, moire and other inter-ference. In some cases,
reception will be totally impos-sible because of the much
greater signal strength of thestation around the corner.
This article describes an inter-ference reduction method that
may be used for UHF TV aswell as VHF radio (including
amateurs).
Design by W. Fischer, DDORQ
Because of the plethora of satellite TVbroadcast stations currently crammedinto the available bandwidth of cable -TV networks, terrestrial (foreign) pro-grammes are often neglected. Unfor-tunately, most of these programmes(like Belgian, Dutch or Irish ones) arenot on satellite, so interested viewersin coastal regions of the ix mainlandare forced to use 'ordinary' TV anten-nas to receive these programs at just -about -acceptable quality.
Whenever these signals are weakand almost blotted out by interference,
you should always remember thatthere is really no reason to complain.After all, the radiated power, antennaheight and radiation pattern of the rel-evant TV transmitter are such that itsdesignated target area is reliably cov-ered. Any signal that spills outside thenormal coverage area, by special prop-agation conditions or otherwise, is sim-ply good fortune. Many viewers of for-eign terrestrial TV stations can live withthat, were it not for the fact that recep-tion often suffers from a national TVtransmitter or relay that uses the same,
14 Elektor Electronics 6/98Elektor Electronics 6/98
Because of the plethora of satellite TV
broadcast stations currently crammedinto the available bandwidth of cable-TV networks, terrestrial (foreign) pro-grammes are often neglected. Unfor-tunately, most of these programmes(like Belgian, Dutch or Irish ones) arenot on satellite, so interested viewersin coastal regions of the UK mainlandare forced to use ‘ordinary’ TV anten-nas to receive these programs at just-about-acceptable quality.
Whenever these signals are weakand almost blotted out by interference,
you should always remember thatthere is really no reason to complain.After all, the radiated power, antennaheight and radiation pattern of the rel-evant TV transmitter are such that itsdesignated target area is reliably cov-ered. Any signal that spills outside thenormal coverage area, by special prop-agation conditions or otherwise, is sim-ply good fortune. Many viewers of for-eign terrestrial TV stations can live withthat, were it not for the fact that recep-tion often suffers from a national TV
transmitter or relay that uses the same,
14
Design by W. Fischer, DD0RQ
co-channel interferencesuppression using stacked antennas
a clever solutionfor VHF/UHF radio and TV
If you want to receive a dis-tant or even overseas TV sta-tion that happens to use thesame frequency as a strong
local transmitter, the receivedsignals will be weak andalways suffer from ghost
effects, moiré and other inter-ference. In some cases,
reception will be totally impos-sible because of the much
greater signal strength of thestation around the corner.
This article describes an inter-ference reduction method that
may be used for UHF TV aswell as VHF radio (including
amateurs).
or almost the same, frequency.Although there is some internationalco-ordination between TV transmitteroperators as regards channel alloca-tions of high-power TV transmitters, ingeneral no responsibility is assumedfor interference inflicted by a nationaltransmitter on the reception of a for-eign one.
Similar problems often occur withVHF radio reception and amateur radio(for example, beacons using the samefrequency).
Interference readily occurs underfavourable propagation conditions.Particularly with ordinary (AM) TV onUHF, even the weakest station on thesame channel can cause very annoy-ing interference patterns in the nor-mally crisp and clear picture from astrong nearby station.
Co-channel interference may besuppressed to a considerable extent byusing a pair of stacked yagi antennasrather than a single one. Althoughthey are directed at the ‘weak’ station,the two antennas are not stacked toincrease their total gain (by 3 dB intheory as a result of increased direc-tivity). Rather, this is done to cancelout the signal picked up from thestrong station that is the source of theinterference.
S T A C K E D A N T E N N A SThe underlying principle is illustratedin Figure 1. Two identical antennas aremounted at a distance ‘d’, and theirsignals are taken to a power combinerby way of two coax cables having thesame length.
As the antenna array is aimed atthe desired (‘weak’) station, thereceived signals will arrive in-phase atthe inputs of the power coupler. If theantennas are spaced by the so-calledoptimum stacking distance, the gain soobtained is 3 dB in theory. However,this distance is not used here, becausewe aim at cancelling interference, andnot increasing the antenna gain orsharpening the antenna pattern!
A signal picked up from an inter-fering station at an angle α will arrivewith different phase angles at each ofthe two antennas (actually, just thedipoles). In fact, the signal arriving atantenna A2 covers an extra distance, s,of
s = d sin α
The phase shift, ϕ, incurred underwayequals
ϕ = 360 s/ λ
where λ is the wavelength of the sig-nal.
For total eradication of the signalreceived from this direction, thereceived signal voltages have to be
equal and exactly 180 degrees out ofphase.
As shown in Figure 2, the stackedantennas have four so-called ‘nulls’ intheir directivity pattern. These nullsoccur at mirror-image positions withrespect to the main axis of reception.
C O M P U T I N G T H ES T A C K I N G D I S T A N C EThe distance d between the two yagiantennas is computed from
In case the calculated distance d is sosmall as to cause the antenna reflectorelements to touch each other, the con-stant n may be made 1, 2, etc. (integervalue required).
Unfortunately, a phase shift of 180degrees can not be obtained from twoantennas stacked as described here,
dn
=+ ⋅λ λ
α
/
sin
2
15Elektor Electronics 6/98
interferingtransmitter
wantedtransmitter
cable 1
cable 2
adder
L
d
S
A1
As
L
αα
980041 - 11
1Figure 1. Principle of co-channel interfer-ence suppression using two stacked anten-nas directed at the same transmitter.
Figure 2. Position of the ‘nulls’ in the directivity diagramof a horizontally polarized pair of yagi antennas.
Figure 3. A practical case based on two transmitterssharing the same frequency in the UHF band. Transmit-ter X is the source of interference, transmitter Y, thewanted station.
α
α
α
α
A1
180° 0°
A2
direction
980041 - 12
α
A1
A2
180°
150°
transmitterX K55
transmitterY
0°
980041 - 13
2
3
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
because in that case the wanted signalwould be cancelled, too.
Using careful construction, how-ever, co-channel interference may besuppressed by at least 25 dB.
E X A M P L EThe author lives in Southern Germany,and wants to receive an Austrian TV
station situated just across the border.The situation is illustrated in Figure 3.Both station X and station Y use chan-nel 55 (picture carrier at 743.25 MHz).In this example, the desired stationuses transmitter Y, whose reception isto be freed from constant interferencecaused by high-power station X.
Well, these transmitters actuallyexist, as well as the receiver location:
X = ARD (German regional TV), loca-tion Cham, transmitter power approx.100 kW.Y = ORF 2 (Austrian national TV), loca-tion Zug peak, transmitter powerapprox. 2 kW.A = receiver location, Vohburg/Donau.
Using the above equations, the dis-tance between the identical UHF yagiantennas is calculated as follows:
Because it is not possible to mount theantennas at a distance of just 40 cm(result of the first calculation), the nexthigher distance is calculated usingn=1. The result is a distance, d, of1.2096 m, which was actually imple-mented by the author to build thestacked antenna array.
P R A C T I C A L R E A L I S A -T I O N
The two antennas are connected upusing a coax coupler device obtainedfrom an antenna installer. The authorused a Reichelt coupler type RW021-DC. This simple and inexpensive solu-tion does have a drawback, however:its insertion loss of about 4 dB nullifiesany extra gain obtained from stackingthe antennas.
d m= 1 2096.
d m=0 403.
dn
=+ ⋅
=+ ⋅
°
λ λ
α
/
sin
. / .
sin
2 0 403 2 0 0 403
150
λ = =300
743 250 403
..
MHzm
λ =300
f
With weak antenna signals, a dif-ferent method of coupling the antennasignals is recommended. The alterna-tive is the quarter-wavelength imped-ance transformer (stub) whose con-struction is illustrated in Figure 4. Forclarity’s sake, the antenna dipoles aresketched only.
The two antennas are electricallyconnected in parallel by two pieces of75-ohm coax cable having the samelength. The impedance at the junctionof the cables is then 75/2=37.5 ohms.A piece of 50-ohm coax cable (forexample, the ubiquitous RG58) is thenused to step up the 37.5 ohms sourceimpedance to 75 ohms as required forthe download cable. The impedance ofthe matching stub is calculated from
The 50-ohm impedance of our ‘real-world’ RG58 cable is sufficiently closeto the theoretical value of 53 ohmsproduced by the equation.
The simple equation you will need tocalculate the length of the matchingstub is given in Figure 4.
The factor v in the equations is theso-called velocity factor, a material con-stant specified for coax cable by itsmanufacturer. For RG58 cable as usedhere, v is stated as 0.66 in thedatasheets. Using this constant, thelength of the 50-ohm matching stubfor the example in Figure 3 is calcu-lated as follows:
f = 743.25 MHzCable = RG58CU; v = 0.66.
Provided you carefully assemble thestacked antenna array, the signal cou-pler and impedance matching stub,the method described in this articleshould enable a previously interfer-ence-ridden TV signal to be receivedmuch better.
(960041-1)
λ
λ
= [ ]
= ⋅ [ ]=
30000
46 66
fcm
v cm
cm
l
l .
Z Z ZL source downl= ⋅
17Elektor Electronics 6/98
downlead cable
980041 - 14
14
Figure 4. Basic con-struction of theimpedance matchingstub.
4
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
electronicaccelerometer
measure acceleration, deceleration,shock, tilt and vibration
Just over a year agowe published an
Application Note cov-ering the type ADXL05
accelerometer chipfrom Analog Devices.Now it's time to actu-ally use this interest-
ing device in a projectfor home construc-
tion.
From an idea by J. Wilkes
From your physics lessons you mayrecall that acceleration is normallymeasured in metres per secondsquared. Acceleration due to theearth's gravity force is, however, usu-ally expressed as a factor of g, where g isthe internationally adopted value9.80665 m/s2. Any forces considerablygreater than 1 g can produce complexsensations in humans beings: jetfighterpilots, and astronauts can tell you allabout them. For your very own experi-ence, do some travelling up and downin a fast elevator system in a tower flat!
Back to electronics, now. For thoseof you who missed the above -men-tioned article, here's a quick rundownon the main component in the presentproject, the ADXL05.
As illustrated in Figure la, the
accelerometer is contained in a 10 -pinTO100 metal case, and consists of a sen-sor, an oscillator, a demodulator, pre-amplifier, voltage reference and abuffer amplifier. The device will mea-sure accelerations with full-scale rangesof ±5g to ±1g or less. Its typical noisefloor is 500 µg/Hz, allowing signalsbelow 5 mg to be measured. TheADXL05 can measure uniform acceler-ation, such as that due to gravity, aswell as variable accelerations, such asvibration.
Since acceleration is a vector quan-tity, the device has three axes: a sensi-tive axis (X) which is defined as illus-trated in Figure ib, a transverse axis (Y),which is perpendicular to the axis ofsensitivity in the plane of the packagecircle, and a transverse axis (Z), which
420 Elektor Electronics 6/98Elektor Electronics 6/98
From your physics lessons you mayrecall that acceleration is normallymeasured in metres per secondsquared. Acceleration due to theearth’s gravity force is, however, usu-ally expressed as a factor of g, where g isthe internationally adopted value9.80665 m/s2. Any forces considerablygreater than 1 g can produce complexsensations in humans beings: jetfighterpilots, and astronauts can tell you allabout them. For your very own experi-ence, do some travelling up and downin a fast elevator system in a tower flat!
Back to electronics, now. For thoseof you who missed the above-men-tioned article, here’s a quick rundownon the main component in the presentproject, the ADXL05.
As illustrated in Figure 1a, the
accelerometer is contained in a 10-pinTO100 metal case, and consists of a sen-sor, an oscillator, a demodulator, pre-amplifier, voltage reference and abuffer amplifier. The device will mea-sure accelerations with full-scale rangesof ±5g to ±1g or less. Its typical noisefloor is 500 µg/Hz, allowing signalsbelow 5 mg to be measured. TheADXL05 can measure uniform acceler-ation, such as that due to gravity, aswell as variable accelerations, such asvibration.
Since acceleration is a vector quan-tity, the device has three axes: a sensi-tive axis (X) which is defined as illus-trated in Figure 1b, a transverse axis (Y),which is perpendicular to the axis ofsensitivity in the plane of the packagecircle, and a transverse axis (Z), which
Just over a year agowe published an
Application Note cov-ering the type ADXL05
accelerometer chipfrom Analog Devices.Now it’s time to actu-ally use this interest-
ing device in a projectfor home construc-
tion.
20
From an idea by J. Wilkes
electronicaccelerometer
measure acceleration, deceleration,shock, tilt and vibration
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
is perpendicular to both X and Y Thetransverse sensitivity is virtually nil.
CIRCUIT DESCRIPTIONAs shown by the circuit diagram in Fig-ure 2, the accelerometer is battery -powered, and equipped with a digitalreadout in the form of an LC (liquidcrystal) display.
At an acceleration of 0 g, the sensoroutput supplies a nominal 1.8 V. Twothings have to be said about this value.First, it is subject to a tolerance of±0.3 V Second, this value allows posi-tive as well as negative g values to bemeasured. To make sure the DVM indi-cates 0 g at a sensor output of 1.8 V, weneed to raise its negative input (-) withrespect to ground, in other words, addan offset of 1.8 V This is done with theaid of the 3.4-V reference voltage sup-plied by the ADXL05, R1, R2 and pre-set P1.
Preset P3 and resistor R6 allow anydeviation from the nominal 1.8-V (0-g)output level supplied by the ADXL05to be corrected.
Although the use of a DVM modulewith a full-scale readout of 200 mVdoes make for a compact and state -ofthe -art circuit, it does cause someapparently contradicting requirements,for which comprises have to be found.To fully exploit the meter's sensitivityand accuracy, it has to be matched tothe typical sensitivity of about200 mV/g of the sensor. That is not toodifficult: simply design for a buffergain of 0.5 and so create a range of 2 gwhich corresponds to a full-scale read-out of 200 mV On theother hand, the fulldrive margin the sensoris then not exploitedbecause the device hasa range of 5 g. Bychoosing a gain of 1/20(0.05), 5 g then pro-duces a readout of just
911
OSCILLATORDECOUPLINGCAPACITOR
2n
SELF -TEST(ST)
AVOOn
+5V
TOP VIEW
COM
50 mV, which, admit-tedly, is not favourablein respect of DVMaccuracy. There youare, however, with twoswitch -selectable ranges of ±2-g and±5-g ranges as trade-offs between dis-play accuracy and sensor accuracy.
Because the sensor sensitivity is alsosubject to a tolerance of ±25 mV/g, thebuffer gain is
Figure 2. Circuit chgram of the accelerom-eter. A low componentcount is achievedthanks to the use of anintegrated sensor anda ready-made voltmetemodule.
IC1
120V
DEMODULATORCAPACITOR
VREFOUTPU1
VOUT
NOTES:AXIS OF SENSITIVITY IS ALONG A LINEBETWEEN PIN 5 AND THE TAB.THE CASE OF THE METAL CANPACKAGE IS CONNECTED TO PIN 5(COMMON).ARROW INDICATES DIRECTION OFPOSITIVE ACCELERATION ALONG AXISOF SENSITIVITY.
Figure 1. Block dia-gram (a) and axes ofsensitivity (b) of theADXL05.
made adjustable withpreset P2.
The accelerometeris powered by a 9 -voltPP3 battery whoseoutput voltage isstepped down to 5 Vby an 78L05 regulatorin position IC1. Cur-rent consumption will
970012 - 11b
be about 15 mA, ofwhich 2 mA goes onaccount of the DVMmodule. Finally, the cir-cuit contains a number
of decoupling capacitors at criticallocations.
CONSTRUCTIONThe accelerometer is easy to build onthe printed circuit board whose art-work (copper track layout and compo-nent mounting plan) is shown in Fig-ure 3. The only point to note here isthat the sensor has to be soldereddirectly on to the board (do not use anIC socket). This is essential because thesensor has to be level when the boardis level. The external components(on/off switch, battery, range selector
C4
22nC2
100n4
C3
72n
° VREF
DC4 VOUT
IC2
ST VIN
ADXL05ODC VPR
DVMDVM
P30
0
R6
a ran ra=QoDo01 0 oDoa
0 DPM951° DVM
S2a 2g (200mV)
R4C7
100n1%
5g (50mV)
R3C6
1
1%
P2 R5
ffirWA20k MT
980047 - 11
Elektor Electronics 6/98 21
is perpendicular to both X and Y. Thetransverse sensitivity is virtually nil.
C I R C U I T D E S C R I P T I O NAs shown by the circuit diagram in Fig-ure 2, the accelerometer is battery-powered, and equipped with a digitalreadout in the form of an LC (liquidcrystal) display.
At an acceleration of 0 g, the sensoroutput supplies a nominal 1.8 V. Twothings have to be said about this value.First, it is subject to a tolerance of±0.3 V. Second, this value allows posi-tive as well as negative g values to bemeasured. To make sure the DVM indi-cates 0 g at a sensor output of 1.8 V, weneed to raise its negative input (–) withrespect to ground, in other words, addan offset of 1.8 V. This is done with theaid of the 3.4-V reference voltage sup-plied by the ADXL05, R1, R2 and pre-set P1.
Preset P3 and resistor R6 allow anydeviation from the nominal 1.8-V (0-g)output level supplied by the ADXL05to be corrected.
Although the use of a DVM modulewith a full-scale readout of 200 mVdoes make for a compact and state-ofthe-art circuit, it does cause someapparently contradicting requirements,for which comprises have to be found.To fully exploit the meter’s sensitivityand accuracy, it has to be matched tothe typical sensitivity of about200 mV/g of the sensor. That is not toodifficult: simply design for a buffergain of 0.5 and so create a range of 2 gwhich corresponds to a full-scale read-out of 200 mV. On theother hand, the fulldrive margin the sensoris then not exploitedbecause the device hasa range of 5 g. Bychoosing a gain of 1/20(0.05), 5 g then pro-duces a readout of just
50 mV, which, admit-tedly, is not favourablein respect of DVMaccuracy. There youare, however, with twoswitch-selectable ranges of ±2-g and±5-g ranges as trade-offs between dis-play accuracy and sensor accuracy.
Because the sensor sensitivity is alsosubject to a tolerance of ±25 mV/g, thebuffer gain is made adjustable with
preset P2.The accelerometer
is powered by a 9-voltPP3 battery whoseoutput voltage isstepped down to 5 Vby an 78L05 regulatorin position IC1. Cur-rent consumption will
be about 15 mA, ofwhich 2 mA goes onaccount of the DVMmodule. Finally, the cir-cuit contains a number
of decoupling capacitors at criticallocations.
C O N S T R U C T I O NThe accelerometer is easy to build onthe printed circuit board whose art-work (copper track layout and compo-nent mounting plan) is shown in Fig-ure 3. The only point to note here isthat the sensor has to be soldereddirectly on to the board (do not use anIC socket). This is essential because thesensor has to be level when the boardis level. The external components(on/off switch, battery, range selector
21Elektor Electronics 6/98
1
A
B
Figure 1. Block dia-gram (a) and axes ofsensitivity (b) of theADXL05.
1N4001
78L05
100µ 100n
BT1
16V
4µ710V
22n
22n
IC1
9V
S1
D1
C1
C5
C2
C3
C4
ADXL05VIN –
IC2
VREF
VOUT
ODC VPR
DC1
DC4
10ST
5
9
6
1
4 8
2
3
7
R1
22
k
R2
22
k
R4
28
k7
1%
R3
3k
16
1%
C7
100n
C6
1µ
R6
27
0k
R5
47k
20k
P2
MT
50k
P3
5k
P1
MT
S2a 2g (200mV)
5g (50mV)
0DVM
DVM+5V1V8
1V8
8V45V
DVM
DPM951
980047 - 11
3V4
2
Figure 2. Circuit dia-gram of the accelerom-eter. A low componentcount is achievedthanks to the use of anintegrated sensor anda ready-made voltmetermodule.
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
and DVM) are wired up as illustratedin Figure 4. Do not yet connect theDVM module, however, because theaccelerometer has to be adjusted first.
A D J U S T M E N T1. Connect a digital multimeter to
the circuit ground and –DVM ter-minal, and adjust P1 for a readingof 1.80 V. Next,connect the multi-meter betweenground and+DVM. Makesure the board is
level (0 g), andadjust P3 for 1.80 Vagain. Now con-nect the DVM module to theaccelerometer board.
2. Change back and forth betweenthe two ranges. The indicatedvalue should remain the same. Ifnot, adjust P3. Next, adjust P1again until the module indicates0 g. Repeat these adjustmentsuntil a reasonable optimum isachieved.
3. Select the ±2-g range, and holdthe board perpendicular withpin 10 of IC1 pointing upwards.Adjust P2 for a reading of 1.000(1 g). Now turn the board so thatpin 10 of IC1 points downwards.The DVM should then indicate–1.000. If not, adjust P2 until thedeviation for +1 g equals that for–1 g. This deviation is caused bythe above-mentioned (small)transverse sensitivity of theADXL05.
That completes the adjustment of theinstrument. Having ‘boxed’ the project,
you should be ready todo some real-world gtesting in a fast elevatoror one of those stomach-churning machines yousee at funfairs.
Finally a few words about the DVMmodule used in this project. This mod-ule has an option for ‘floating’ mea-surements, which allows the circuit tobe powered by a single battery. Theaccelerometer board has a 5-volt con-nection for powering the DVM mod-ule. It should be noted that the nega-tive connection of the DVM modulerepresents a low resistance. Conse-quently it can not be connected to thenegative terminal of the accelerometerboard. So it is connected to the(buffered) positive terminal instead.The only difference it makes is the ori-entation of the sensor (simply turn theaccelerometer board 180 degrees). Byusing a triple changeover switch inposition S2, the decimal point on theDVM module can be switched accord-ing to the range selected.
(980047-1)
Reference:Accelerometer Type ADXL05, Applica-tion Note, Elektor Electronics April 1997.
23Elektor Electronics 6/98
980047-1H2 H
4
C1
C2
C3C4
C5
C6
C7
D1
H1 H3
IC1
IC2
P1
P2
P3
R1
R2
R3
R4R5
R6
980047-1
5V0
DVM
+
+
+
-
-
S2
S1
BT1
980047-1
3
Figure 3. Copper track layout and componentmounting plan (board not available ready-made).
COMPONENT LIST
Resistors:R1,R2 =22kΩR3 =3kΩ16 1% (Philips MRS25
series)R4 =28k7 1% (Philips MRS25 series)R5 =47kΩR6 =270kΩP1 =5kΩ multiturn 10 turn horizontalP2 =20kΩ multiturn 10 turn horizon-
talP3 =50kΩ multiturn 10 turn horizon-
tal
Capacitors:C1 = 100µF 16V radialC2 = 100nF Sibatit (Siemens)C3,C4 = 22nF MKT (Siemens)C5 = 4µF7 10V radialC6 = 1µF MKT (Siemens)C7 = 100nF MKT (Siemens)
Semiconductors:D1 = 1N4001IC1 = 78L05IC2 = ADXL05JH (Analog Devices)
Miscellaneous:BT1 = 9V battery PP3 blockS1 = rocker switch, 1 make contactS2 = rocker switch, 1 c/o contact, or
3x c/o contact for decimal point ondvm
DVM = DPM951 DVM module (Con-rad)
C1
C2
C3C4
C5
C6
C7
D1
IC1
IC2
P1
P2
P3
R1
R2
R3
R4R5
R6
980047-1
5V
0
DVM
+
+
+
+-
-
-
S2 S2aS1
S2b
980047 - 12
S2c
S1
BT1
BT1
9V
INHI1
INLO2
VDD LK4
LK5
LK6
3
VSS4
COMMON5
6
REFLO7
REFHI8
BP9
BP10
11
DP312
DP213
DP114
5g 2g
DVM
DPM951
Figure 4. Suggestedwiring diagram. Thedecimal point on theDVM module isswitched with two sec-tions of S2.
4
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
elects nics n linerPICs on the Internet
Microchip's family of PICdevices comprises a small
galaxy of single -chip microcom-puters, which achieved tremen-dous popularity mainly because
they are so easy to program.We rummaged around on theInternet for a bit and stumbled
on an overwhelming amount ofinformation relating to PICs.Some of the more essential
sites are discussed here.
The number of sites (on the Internet)covering PIC-related subject reflectsthe widespread interest in these con-trollers. The contents of these sitesranges from simple exchanges of infor-mation to commercial pages offeringready -built circuits for sale. Many do-it-yourself (DIY) enthusiasts, includingmany readers of this magazine, will beinterested mainly in construction pro-jects and free PIC software.
As a matter of course we shouldstart by mentioning the site run by themanufacturer, Microchip:www.microchip.comHere, all elementary information maybe found. Then there's Parallax Inc.,who supply the famous BASIC Stamp,a miniature board containing a PICrunning a BASIC interpreter. Parallaxcan be found atwww.parallaxinc. com/home/htm
There are many more companiessupplying hardware and/or softwareemploying PIC devices. For example,ITU Technologies from Cincinnati:www.itutech.comThe PIP02 software for the Mini PICProgrammer described our June 1997issue is available from Silicon Studio atwww.sistudio.com
PROGRAMMERSDesigns for PIC programmers and theassociated software may usually beobtained from sites run by enthusiastspublishing their 'thinkware' via theInternet. A PIC programmer for thePC parallel port (P16PRO), completewith printed circuit board design and
Windows soft-ware may befound on theMaribor Uni-versity server(in Slovenia) atwww.uni-mb.si/-tw1205e7b/(our thanks aredue to Mr. E.Dekker forsupplying thisURL).
Another programmer for the Cen-tronics port goes by the name ofTOPIC. It may be found atwww.man.ac.uk/-mbhstdj/topic.htmlHowever, if you want to keep thingsas simple as possible, you are welladvised to look for a PIC programmerfor the serial port. One definitely KISS(Keep It Simple Stupid) and yet well -tried design is, for example, COM84.All details on this ultra -simple designare available onwarthog.ecce.maine.edu/segeeprog84.html
Finally, we should mention the Pro-grammer 2 design atwww.gbar.dtu.dk/-c888600/newpic.htm
lcMicrochip Technology Inc. - NetscapeFile Edit View Go Communicator Help
-; 31 -1--)" sBookmarks Location. ihttp...qwww. microchip.com/
Onli
rothictInformation
Crnicro-. DevicesMary DevicesceLerf, Deviceseveloper's
Toolboxpplicatkin Notes
Consultantsatashrmtsevelopinent Tor-
Knowledge Bateelated Sites
°moonyInformation
vet -viewEditorial News
EVeni,mploymentSeles &Distribution
nninaea
Text Version
IZMICROCHIP?.The Embedded Control Solutions Company-
P LLIV LIZ IT!Welcome to Online @ Microchip,the official web -site offering thelatest data -sheet on our deviceso software updates of our populardevelopment tools.
M.11=EMMOrder Online Today!
Ni1C11.1=1111* Select from 4PlCmienr- sessions and 38
MCU cities asS'rt.a7,, Microchip
presents the 1998PICrnicroT. MCU
Seminars. The seminars includespecial pricing fors of our mostpopular development tools.
P1C17C756 Seminar &1Norksho
Join other 8 -bitMCU designers,
applicationengineers, and
consultingprofessionals in
mastering the thigh-performance features of
Microchip's PIC17C756
Quick FAQ'sDownloading Blues?
CD Request
How Much Is It?Sales & Distribution
Where Is It?Search Us
Did Someone Say Help?Contact Microchip
ffIEIMEMEMP LAB v3.40 "BETA"Specification Update. PIC1605.XProgramming Specifications. PIC1GCEEX
More of "NT., New" il,e2 17110r00,11P
PlCmicro Workshop 201 InvitationThe Complete B Bit MCU Solution
Investor Relations & Stock InfoProgramming SpecificationsSoftware Updates & User GuidesJobs 9 MicrochipPackaging & Reference Material
Send web comment to tyebonastergmicrochle.comce. Microchip Technology Inc 'INN] All RIghle Reseved
The above mentioned programmersare just a few of dozens of designs thatmay be found on the Internet.
LINKSBecause of the sheer volume of infor-mation available on PICs via the Inter-net, it's a good idea to have a look atpages maintained by people who havedone research work.
Eric's PIC page is very extensive anoffers, among others, PIC projects, tipsand tricks, FAQs and a plethora oflinks to other PIC sources and suppli-ers. Go towww.brouhaha.com/-eric/pic/
Dave Tait, who is also responsiblefor the TOPIC programmer, runs apage called 'Links to InternetResources', which contains hundredsof links to books, commercial products,application areas, and other peopleexperimenting with PICs:www.man.ac.uk/mbhstdj/piclinks.html
The South African PIC Microcon-troller Zone atwww.ip.co.za/people/kalle/pic/default.htmhas a nice collection of assorted PICinformation, like ftp sites, web pagesand mail lists.
Finally, if you think you need evenmore information, visit the PIClistArchive Website atwww.iversoft.com/piclist/which can be browsed by subject, dateor author.
(985045-1)
A24 Elektor Electronics 6/9824 Elektor Electronics 6/98
The number of sites (on the Internet)covering PIC-related subject reflectsthe widespread interest in these con-trollers. The contents of these sitesranges from simple exchanges of infor-mation to commercial pages offeringready-built circuits for sale. Many do-it-yourself (DIY) enthusiasts, includingmany readers of this magazine, will beinterested mainly in construction pro-jects and free PIC software.
As a matter of course we shouldstart by mentioning the site run by themanufacturer, Microchip:www.microchip.comHere, all elementary information maybe found. Then there’s Parallax Inc.,who supply the famous BASIC Stamp,a miniature board containing a PICrunning a BASIC interpreter. Parallaxcan be found atwww.parallaxinc. com/home/htm
There are many more companiessupplying hardware and/or softwareemploying PIC devices. For example,ITU Technologies from Cincinnati:www.itutech.comThe PIP02 software for the Mini PICProgrammer described our June 1997issue is available from Silicon Studio atwww.sistudio.com
P R O G R A M M E R SDesigns for PIC programmers and theassociated software may usually beobtained from sites run by enthusiastspublishing their ‘thinkware’ via theInternet. A PIC programmer for thePC parallel port (P16PRO), completewith printed circuit board design and
Windows soft-ware may befound on theMaribor Uni-versity server(in Slovenia) atwww.uni-mb.si/~uel205e7b/(our thanks aredue to Mr. E.Dekker forsupplying thisURL).
Another programmer for the Cen-tronics port goes by the name ofTOPIC. It may be found atwww.man.ac.uk/~mbhstdj/topic.htmlHowever, if you want to keep thingsas simple as possible, you are welladvised to look for a PIC programmerfor the serial port. One definitely KISS(Keep It Simple Stupid) and yet well-tried design is, for example, COM84.All details on this ultra-simple designare available onwarthog.eece.maine.edu/segee/prog84.html
Finally, we should mention the Pro-grammer 2 design atwww.gbar.dtu.dk/~c888600/newpic.htm
The above mentioned programmersare just a few of dozens of designs thatmay be found on the Internet.
L I N K SBecause of the sheer volume of infor-mation available on PICs via the Inter-net, it’s a good idea to have a look atpages maintained by people who havedone research work.
Eric’s PIC page is very extensive anoffers, among others, PIC projects, tipsand tricks, FAQs and a plethora oflinks to other PIC sources and suppli-ers. Go towww.brouhaha.com/~eric/pic/
Dave Tait, who is also responsiblefor the TOPIC programmer, runs apage called ‘Links to InternetResources’, which contains hundredsof links to books, commercial products,application areas, and other peopleexperimenting with PICs:www.man.ac.uk/mbhstdj/piclinks.html
The South African PIC Microcon-troller Zone atwww.ip.co.za/people/kalle/pic/default.htmhas a nice collection of assorted PICinformation, like ftp sites, web pagesand mail lists.
Finally, if you think you need evenmore information, visit the PIClistArchive Website atwww.iversoft.com/piclist/which can be browsed by subject, dateor author.
(985045-1)
PICs on the Internetelectronics on-lineelectronics on-line
Microchip’s family of PICdevices comprises a small
galaxy of single-chip microcom-puters, which achieved tremen-dous popularity mainly because
they are so easy to program.We rummaged around on the
Internet for a bit and stumbledon an overwhelming amount of
information relating to PICs.Some of the more essential
sites are discussed here.
ifr#7^PPIC & AVRprogrammer
Hardware and software to programPIC and AVR microcontrollers
The hardware andWindows 95 soft-
ware discussed inthis article
enables you toprogram
Microchip'shighly success-
ful 16C84 and16F84 microcon-
trollers, as well as Atmel's `AVR' typesAT90S1200, AT90S2313, AT90S4414 and
AT90S8515. All of these RISC 'beasts' are in -
circuit programmable, and offer internalFlash/EEPROM program memory, static RAM,
various input/output lines and another EEPROMfor the microcontroller to write to.
/P 1 d i I-
I I I I I
I I I r II I I
Ii II I I I I II I I I I
I 1
,-
'. _
-21Ar.,4
Design by W. Schroeder
It's almost impossible tothink of modern electronics withoutmicrocontroller applications. Micro -controllers have a vast number of appli-cation areas, ranging from the PCmouse to process control units, notonly in industrial systems but also thelatest washing machines. Admittedly,very few of you will be interested indeveloping their own washingmachine. However, applications thatdo come to mind for home develop-ment may include a timer -controlledwindow shutter, an infrared controlledgarage door, a sound -to -light unit, or a
26 Elektor Electronics 6/98Elektor Electronics 6/98
It’s almost impossible tothink of modern electronics withoutmicrocontroller applications. Micro-controllers have a vast number of appli-cation areas, ranging from the PCmouse to process control units, notonly in industrial systems but also thelatest washing machines. Admittedly,very few of you will be interested indeveloping their own washingmachine. However, applications thatdo come to mind for home develop-ment may include a timer-controlledwindow shutter, an infrared controlledgarage door, a sound-to-light unit, or a
The hardware andWindows 95 soft-
ware discussed inthis article
enables you toprogram
Microchip’shighly success-
ful 16C84 and16F84 microcon-
trollers, as well as Atmel’s ‘AVR’ typesAT90S1200, AT90S2313, AT90S4414 and
AT90S8515. All of these RISC ‘beasts’ are in-circuit programmable, and offer internal
Flash/EEPROM program memory, static RAM,various input/output lines and another EEPROM
for the microcontroller to write to.
26
Design by W. Schroeder
PIC & AVRprogrammer
Hardware and software to programPIC and AVR microcontrollers
digitally controlled laboratory powersupply. In these examples, processorcontrols using more or less powerfulmicrocontrollers are a great option.
The PIC controllers type 16C(F)84
belong in the class of less powerful butalso less expensive devices. These RISCcontrollers have an instruction set ofjust over 30 instructions (hence reducedinstruction set). They can be operated ata clock frequency of up to 10 MHz, andthen offer a command cycle time of400 ns, which equals 1/4th the clock fre-quency.
When more computing power isrequired, or more I/O pins, you should
opt for Atmel’s AVR controllers. Thesmallest of these, the AT90S1200, wasalready the subject of a project forhome construction (Ref. 1,2), and itmay be compared, as far as computingpower is concerned, with the PICdevices. The larger AVR controllersoffer additional features like an inte-grated RS232 interface, a serial SPIinterface, pulse width modulation, twotimers, up to 120 instructions, and so
27Elektor Electronics 6/98
XTAL1 XTAL2
RESET
S1200AT90
IC9
PB6
PB5
PB7
PB4
PB3
PB2
PB1
PB0PD0
PD1
PD2
PD3
PD4
PD5
PD6
20
10
19
18
17
16
15
14
13
12
11
5 4
1
2
3
6
7
8
9
K2
1
2
3
4
5
6
7
8
9
C15
C16
MAX232
R1OUT
R2OUT
T1OUT
T2OUT
IC7
T1IN
T2IN
R1IN
R2IN
C1–
C1+
C2+
C2–
11
12
10
13
14
15
16V+
V-
7
89
3
1
4
5
2
6
C17
C13
C14
R8
27
0Ω
R9
10
0Ω
R10
10
k
R11
10
k
X2
10MHz
C18
18p
C19
18p
C20
100n
R2
27
0Ω
X1
4MHz
C11
18p
C12
18p
C25
10n
T2
BS170
T3
BS170
S5
X3
4MHz
C21
18p
C22
18p
A0
A1
A2
A3
A4
A5
A6
B0
B1
B2
B3
B4
B5
B6
B7
C9
100n
B0
B1
B2
B3
B4
B5
B6
B7
A2
A3
A4
7805
IC2
78L12
IC6
K8
C2
100µ25V
C4
10µ16V
C26
10µ25V
C27
10µ16V
C1
100n
C3
100n
C24
100n
C23
100nD1
1N4001
T1
BS250
R4
10
0k
R7
4k
7
R5
10k
R6
1k
IC5c
5
6
1
C6
100n
S4
S3
8x 1k1
2 3 4 5 6 7 8 9
R3
S1
S2
8x 1k1
2 3 4 5 6 7 8 9
R1
A0
A1
A2
A3
A4
A5
A6
B0
B1
B2
B3
B4
B5
B6
B7
74HC541
IC4
1112131415161718
19
EN
2 3 4 7 8 95 6
&
1
D10
D11
D12
D13
D14
D15
D16
74HC541
IC3
1112131415161718
19
EN
2 3 4 7 8 95 6
&
1
D2
8x 2k21
9 8 7 6 5 4 3 2
R14
D3
D4
D5
D6
D7
D8
D9
B0
B1
B2
B3
B4
B5
B6
B7
A0
A1
A2
A3
A4
A5
A6
PIC16C84
CLKOUT
IC1
OSC1
MCLR
RA4
RA1
RA0
RA2
RA3
RB0
RB1
RB2
RB3
RB4
RB5
RB6
RB7
17
18
13
12
11
10
16 15
14
1
3
9
8
7
6
2
4
5
PIC16C84
CLKOUT
IC8
OSC1
MCLR
RA4
RA1
RA0
RA2
RA3
RB0
RB1
RB2
RB3
RB4
RB5
RB6
RB7
17
18
13
12
11
10
16 15
14
1
3
9
8
7
6
2
4
5
A0
A1
PD1
PD0
PD2
PD3
DEBUG
PMISO
PONL
PMOSI
AT RES
MCLR
TXD5
RXD5
CTS5
RTS5
RXD
CTS
TXD
RTS
PD
3
PD2
MCLR
R13
10
k
IC5d
9
8
1
IC5a 1
2
1
IC5b 3
4
1
IC5e 11
10
1
IC5f
13
12
1
PD
1B
6
PD
0B
7
B7
PO
NL
PM
OS
IB
5
B6
PM
ISO
RESET
K5
10
1
2
3
4
5
6
7
8
9
K4
10
1
2
3
4
5
6
7
8
9
A0
A1
A2
A3
A4
A5
A6
B0
B1
B2
B3
B4
B5
B6
B7
DEBUG
K1
10
11
12
13
14
15
16
1
2
3
4
5
6
7
8
9
10k
P1
B2
B3
B1
A0
A1
A2
RS
R/W
EN
K6
K7
K3
A3
B7
B6
B7
B6
B5
B6
B7
MCLR
RESET
MCLR
MCLR
AT
RE
S
C10
100n
IC5
14
7
C5
100nIC5 = 74LS07
C8
100nIC3
20
10
IC4
20
10
5V
12V
5V
5V
5V
5V
5V
5V5V
5V
5V
5V 5V
U
U
5V
5V
5V
5V
5V
5V
5V
5V
-10/P
C7
100n
C13 ...C17= 1µ / 16V
VPP
CLK
DATA
980049 - 118x 2k2R12 1
9 8 7 6 5 4 3 2
1
Figure 1. Circuit dia-gram of the program-mer and programmingsocket adaptor IC9.
Controller Clock RAM Progr.Flash EEPROM I/O TIMER RS232 SPI PWM
PIC 16C(F)84 10 MHz 36 B 1024 W 64 B 13 1 – – –
AVR AT90S1200 16 MHz 0 B* 512 W 64 B 15 1 – – –
AVR AT90S2313 20 MHz 128 B* 1024 W 128 B 15 2 X – X
AVR AT90S4414 20 MHz 256 B* 2048 W 256 B 32 2 X X X
AVR AT90S8515 20 MHz 512 B* 4096 W 512 B 32 2 X X X
* + 32 × 8 Bit Registers
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
on. The fact that thesedevices process instructionsat the full clock rate makesthem four times as fast as aPIC, assuming the sameclock rate is used.
For comparison pur-poses, Table 1 lists the mainfeatures of the individualmicrocontrollers. Currently,the best available microcon-trollers are the PIC16F84,AT90S1200 and AT90S8515(max. 8 MHz).
H A R D W A R ED E S C R I P T I O NA N DC O N S T R U C T I O NThe hardware and softwarepresented in this articleallows you to programMicrochip’s PIC16C84/16F84as well as all currently avail-able (early 1998) Atmel AVRmicrocontrollers (AT90S1200,AT90S2313, AT90S4414,AT90S8515). All RISC micro-controllers supported by thepresent design are in-circuitprogrammable using a‘shell’ called PICAVR32which runs under Win-dows 95. These controllersoffer an internal Flash/EEP-ROM program memory,SRAM, a plethora ofinput/output lines and anEEPROM to which themicrocontroller has directaccess.
Building the hardwareshould not present difficul-ties if you work carefullyand take your time. With thepossible exception of theready-programmed PIC inthe circuit, only commonlyavailable parts are used. Thefirmware-code PIC, by theway, is available ready-pro-grammed, either directlyfrom the Publishers orthrough kit suppliers adver-tising in this magazine.
The hardware isdesigned in such as way thatPICs as well as the DIL-styleAT90S1200 and AT90S2313
28 Elektor Electronics 6/98
980049-1
C1
C2C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
C20
C21 C22
C23C24
C25
C26
C27
D1
D2 D9 D10 D16H1 H2
H3H4
IC1
IC2
IC3 IC4
IC5
IC6
IC7 IC
8
IC9
K1
K2
K3
K4 K5
K6K7
K8
P1
R1
R2
R3
R4R5 R6R7
R8
R9
R10 R11
R12
R13
R14
S1
S2 S3
S4
S5
T1
T2T3
X1X2
X3
980049-1
- - - - - - - - - - - - - - - -+
-
T
(C)
Seg
men
t980049-1
980049-1
2
Figure 2. Every-thing fits on thisdouble-sidedthrough-platedboard (availableready-madethrough the Pub-lishers).
chips may be programmed and testedby inserting them in a DIL socket onthe programmer board. Mind you,these are the chips you want to pro-gram, not the PIC mentioned above(containing the firmware code).
In addition, all microcontrollersmay be programmed in-circuit via con-nectors K6 (AVR) and K3 (PIC), which isparticularly useful for the AT90S4414and AT90S8515 which come in a PLCCcase.
The programmer board is poweredby a mains-adaptor (15 V, 500 mA) con-nected to K8. When only AVR chips areprogrammed, a 9-volt adaptor is suffi-cient because AVR devices do notrequire an additional programmingvoltage. The supply voltage is stabi-lized by a 7805 voltage regulator (IC2)and its satellite decoupling capacitorsC1-C4 for smoothing and decoupling.A 78L12 regulator (IC6) is responsiblefor the extra 12-V PIC programmingvoltage.
The heart of the PICAVR32 pro-grammer is PIC IC8, which containsthe firmware code and arranges allcommunication with two parties: atone side, the PC, and at the other side,the microcontroller to be read or pro-grammed. The link with the PC isestablished using an ordinary 9-way 1-to-1 sub-D extension cable (socket &plug). The well-known level convertertype MAX232 (IC7) on the program-mer board changes RS232 voltage levels(±12 V) into TTL/CMOS compatiblesignals (5-V), and feeds them directlyto the corresponding I/O pins of the
firmware PIC, IC8. This chip respondsby forwarding control commands setup by the PC software either to the PICor the AVR to be programmed (posi-tions IC1 and IC9 respectively). In thepresent programmer both microcon-troller types are programmed seriallyby way of an SP interface. For this theAVR chips employ I/O pins 5, 6 and 7on Port B (MOSI, MISO, SCK)). ThePICs are programmed via I/O pins 6and 7 of Port B (CLK, DATA). To theseprogramming signals should be addedthe supply voltage (GND, VCC) andthe /RESET or /MCLR signals, all ofwhich are also needed during pro-gramming. All relevant signals aretaken into account in the pinning ofthe in-circuit programming connectorsmentioned above. When designingyour own (application) circuits inwhich this type of programming is tobe employed, you should make surethat ‘collisions’ between the program-mer hardware and the target systemhardware are avoided! In particular, besure to provide series resistors on thesignal lines used during programming.
To prevent the risk of your own cir-cuits under test causing damage to theprogrammer hardware, the latter con-tains an open-collector driver type74LS07 (IC5) as well as transistors T2and T3, which buffer all lines of thefirmware PIC, IC8. The use of open-collector drivers does, however, neces-sitate the switching on of the DIPswitches for the pull-up resistors. Forthe AVR chips, the Port-B lines arepulled up (S1.6, S1.7, S1.8), while for
the PIC chips S1.7 and S1.8 areswitched on. The other switches in S1and all switches in S4 should be ‘on’.All switches in S2 and S3 should be setto ‘off ’.
The /RESET line of an AVR chip iscontrolled via transistor T2. The fallingpulse edge at the /RESET pin is essen-tial to initiate a programming sequenceon an AVR chip. Because the /MCLRpin on PIC chips has to be at 12 V dur-ing programming, T1 is required inaddition to T3 which arranges theactual resetting of the PIC. Normally,T1 and T3 are both switched off. Thejunction R5-R7 supplies a voltage ofabout 4 V, which arrives at the /MCLRinput of the PIC to be programmed,via resistor R6. To reset the PIC, the/MCLR pin is pulled to ground via T3.The programming voltage is appliedby switching on the p-channel MOS-FET. When the MOSFET conducts, itshort-circuits R5, so that 12 V is appliedto R6.
PICs to be programmed are clockedat 4 MHz — other clock frequencies arealso possible. For AVR chips a quartzcrystal is available, and, as an alterna-tive, a complete crystal-controlled oscil-lator.
The DIP switches provide assis-tance with the testing of input/outputoperations of the two microcontrollers.Switches S1 and S4 connect 1-kΩ pull-up resistors to all port lines, while S2and S3 allow port input lines to be heldlogic low. Users should take care not touse the input switched on port linesdeclared as outputs! Doing so (by acci-
29Elektor Electronics 6/98
COMPONENTS LIST
Items available from the Publishers:PCB, programmed PIC and disk,
order code 980049-C.Firmware PIC only,
order code 986509-1.Project disk only,
order code 986019-1.
Resistors:R1,R3 = SIL-Array 8 x 1kΩR2,R8 = 270ΩR4 = 100kΩR5,R10,R11,R13 = 10kΩR6 = 1kΩR7 = 4kΩ7R9 = 100ΩR12,R14 = SIL array 8 x 2kΩ2P1 = 10kΩ
Capacitors:C1,C3,C5-C10,C20,C23,C24 = 100nFC2 = 100µF 25V radialC4,C27 = 10µF 16V radialC11,C12,C18,C19,C21,C22 = 18pFC13-C17 = 1µF 16V radialC25 = 10nFC26 = 10µF 25V radial
Semiconductors:D1 = 1N4001D2-D16 = LED, low currentT1 = BS250T2,T3 = BS170IC1 = PIC16C84/F84 to be pro-
grammedIC2 = 7805IC3,IC4 = 74HC541IC5 = 74LS07IC6 = 78L12
IC7 = MAX232IC8 = PIC16C84-10/P (ready-pro-
grammed, order code 986509-1)IC9 = AT90S1200 to be programmed
Miscellaneous:S1-S4 = 8-way DIP switch blockS5 = 4-way DIP switch blockX1,X3 = 4MHz quartz crystal X2 = 10MHz quartz crystal K1 = 16-way boxheaderK2 = 9-way sub-D socket (female),
angled, PCB mountK3,K6,K7 = 6-way SIL pinheaderK4,K5 = 10-way boxheaderK8 = mains adaptor socket,
PCB mountDiskette containing PICAVR32 pro-
gram, order code 986019-1.
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
3PICAVR32- CADELPHI324PICAVR324Weitergabe4Testarr.asm
File Edit Controller
AT90 S 1200
LED Test program f or 41',28.10.97 Stefan Jansen
include "1200definc"device AT90 S 1200
def Temp =r1def Tempt =r17def Temp3 =r18def A_Wert =r1def Delay =r2def Delay2 =r21def Delay3 =r22
Options Help
Controller
COM port
Editor
Assembler options
Auto download
PIC cliff. download
Slow PC
Debug -monitor
Show errors Ctrl+E
PIC: 16C84
4,0 AVR: ATSOS1 200
AVR: AT90S2313
AVR: AT9OS441 4
AVR: ATSCIS8515
PIC1 6F84.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Interrupt vectors : fixed at addresses [Ito 3
Adr $000 = ResetAdr $001 = External InterruptMr $002 = TimerO InterruptMr $003 = Analog comparator Interrupt
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
rjmp Resetrjmp Ext_Intorjmp Tirnfl_cmf
rjmp Ana_ Comp
OW Main
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Interrupt routine External Interrupt
Ld_Into:retr
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Interrupt routine Timer°xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Status: New processor type : AT90S1 200
igEl Start I A My Computer j (C:) I Delphi32
PICAVR32 - CADELPH1324PICAVR324Weitergabe4Testpic.asm
File Edit Controller
PIC16 C84
Test program for PIC 8e AAll outputs are tested: 8 re
4 yellow 1
Attention, the order on poi
list p=16C84
tinclode "PI C.11"tinctocle"TE STPIC.H'
org Ox0000
ESET GOTO START
Options Help
Controller
COM port
IEditor
Assembler options
Auto download
PIC cliff. download
Slow PC
Debug-mon
Show errors Ctrl+E
Font
Background
TARTCLRF PORTACLRF PORTEBSF STATUS.RPOMOVLW V00000000'MOVWF TRISAMOVLW V00000000'MOVWF TRISBBCF STATUS.RPO
BSF STATUS, CARRYBCF INTCON.REIF
ESTPICCLRF TEMPBSF STATUS, CARRYMOVLW Ox09MOVWF COUNTCLRF TEMP
OTE RLF TEMP.!CALL WAITMOVF TEMP.!)MOVWF PORTEDE CFSZ COUNT,1
To program start
Clear Port AClear Port B
Switchto Bank 1Switch PO RTA to
output
Switch PortE tooutput
Switchto Bank 0Set Carry to 0
Set Port B Change interrupt to 0
Set Carry to 0
Load loop variable into WLoad Win COUNT
Rotate PORT B left until into CARRY FlagWait
Status: Running assembler...
If zero then two onwards
StartI A My Com...I GJ (C:( 1 Delphi32 Picavr32 Weiterg..
dent) may cause damage to the rele-vant port of the PIC or AVR chip to beprogrammed.
Low -current LEDs D2 through D12indicate the port -line states. They arecontrolled via IC3 and IC4.
The hardware when finished istested with the aid of a couple of spe-cially written test utilities. First, how-ever, let's have a look at the controlsoftware running on the PC.
WINDOWS 95CONTROL SOFTWAREThe software developed for the PIC &AVR Programmer also enables assem-bly code for PICs and AVRs to bedeveloped, programmed and read. Forthis purpose, the program as well asthe EEPROM memory may be readand written. You can get started withthe programmer once it is connectedup to the PC's serial port (via a 9 -waysub -D cable with no crossed wires),and connected to its power supply. Tobegin with, you should burn a PICwith the file TESTPIC.ASM, or an AVRchip with the file TESTAVR.ASM. Inaddition to a controller test, all LEDoutputs are checked with the aid of arunning light. Follow this procedure.
Launch PICAVR32The PICAVR32 shell first looks for theprogrammer hardware. When thehardware does not report back, a mes-sage reading No COM Port Selected isindicated near the bottom of thescreen.
Select RS232 portSelect a free RS232 port using themenu OPTIONSI COM PORT. Whenthe programmer hardware is found,the status line reads PICAVR32 (PICVersion x.x.) ready. The menu optionCONTROLLER is then accessible.
Load assembly -code fileUnder Filel Open, load the above men-tioned assembler source code file.Alternatively, use the FILEI LOADoption to fetch a finished Hex(adeci-mal) file.
Assemble fileUse FILEI START ASSEMBLER tolaunch the assembly -code to object -code translation.
Download object codeNext, the menu option CON-TRO LLERI DOWNLOAD PRO GRAMis used to actually program your PIC orAVR chip. Provided no errors occurduring programming, the relevant
r Figure 3. Somescreenshots showingPICAVR32 in action.
A30 Elektor Electronics 6/98
dent) may cause damage to the rele-vant port of the PIC or AVR chip to beprogrammed.
Low-current LEDs D2 through D12indicate the port-line states. They arecontrolled via IC3 and IC4.
The hardware when finished istested with the aid of a couple of spe-cially written test utilities. First, how-ever, let’s have a look at the controlsoftware running on the PC.
W I N D O W S 9 5C O N T R O L S O F T W A R EThe software developed for the PIC &AVR Programmer also enables assem-bly code for PICs and AVRs to bedeveloped, programmed and read. Forthis purpose, the program as well asthe EEPROM memory may be readand written. You can get started withthe programmer once it is connectedup to the PC’s serial port (via a 9-waysub-D cable with no crossed wires),and connected to its power supply. Tobegin with, you should burn a PICwith the file TESTPIC.ASM, or an AVRchip with the file TESTAVR.ASM. Inaddition to a controller test, all LEDoutputs are checked with the aid of arunning light. Follow this procedure.
Launch PICAVR32The PICAVR32 shell first looks for theprogrammer hardware. When thehardware does not report back, a mes-sage reading No COM Port Selected isindicated near the bottom of thescreen.
Select RS232 portSelect a free RS232 port using themenu OPTIONS|COM PORT. Whenthe programmer hardware is found,the status line reads PICAVR32 (PICVersion x.x.) ready. The menu optionCONTROLLER is then accessible.
Load assembly-code fileUnder File|Open, load the above men-tioned assembler source code file.Alternatively, use the FILE|LOADoption to fetch a finished Hex(adeci-mal) file.
Assemble fileUse FILE|START ASSEMBLER tolaunch the assembly-code to object-code translation.
Download object codeNext, the menu option CON-TROLLER|DOWNLOAD PROGRAMis used to actually program your PIC orAVR chip. Provided no errors occurduring programming, the relevant
30 Elektor Electronics 6/98
3
Figure 3. Somescreenshots showingPICAVR32 in action.
controller is started, and the LEDs onthe programmer board act as a (slightlymixed-up) running lights.
O P T I O N SThe software offers various optionalfunctions in addition to the straight-forward ‘load-file-and-program-away’sequence outlined above. Help withthe individual menu options can besummoned up by pressing the F1 key.The menus FILE, EDIT and HELPshould be self-explanatory, so they arenot discussed here.Under OPTIONS|CONTROLLERyou can select the controller to be pro-grammed. If the controller type isincluded in the first line of the assem-bly-code file, behind the semicolon,then it is automatically selected whenthe assembly-code file is loaded.The OPTIONS|EDITOR menu entryallows you to select editor cosmeticslike font, colour and type, as well asbackground colour.OPTIONS|ASSEMBLER OPTIONSserves to ‘adjust’ the assembler and theparameter required for calling theassembler. For PIC devices you shoulduse Microchip Assembler Release 1.4. ForAVR devices, use AVRASM Version 1.1.Be sure to obtain the DOS versions ofthese programs, which are notincluded on diskette 986019-1. Theyare, however, available free of chargefrom the Microchip and Atmel Internetsites.OPTIONS|AUTODOWNLOAD ,when enabled, automatically launchesa program download operation, fol-lowing successful source code assem-bly.OPTIONS|PIC DIFF DOWNLOADenables a difference-file download forPICs. It is particularly useful if only cer-tain parameters of a program are to bemodified, or just a couple of bytes. Thisoption is only available for PICsbecause AVR chips have to be com-pletely erased before any program-ming of the program memory.OPTIONS|SLOW PC should beswitched on when using a sluggish PCnot having a serial FIFO for the RS232interface. Use this option if errors occurfrequently when downloading pro-grams.OPTIONS|DEBUG MONITORlaunches a Debugging monitor thatalows hex values to be produced usinga test-output routine. The Debugging
procedure and the relevant routinesare discussed in greater detail in thePICAVR32 Help file.By selecting OPTIONS|SHOWERRORS you enable the program todisplay the assembler ’s error reportfile. The Error-Viewer is also openedautomatically whenever the assemblerwrites an error entry in the Error file.
All functions found under the menuoption CONTROLLER cover thedownloading/uploading of programs,EEPROM and the internal configura-tions of the individual microcon-trollers. The individual functions are,of course, dependent on the selectedmicrocontroller type.CONTROLLER|DOWNLOAD writesa currently loaded hex file or the objectcode (resulting from a previously com-piled assembly-code file) into the con-troller. PICAVR32 has an internal bufferfor the object code to be written intothe microcontroller. A hexadecimal fileloaded using the FILE|LOAD INTEL-HEXFILE32 option is also transferredto this buffer, which has no graphicsdisplay in PICAVR32.
FILE|SAVE HEX enables a controllerto be read, and its contents to be writ-ten into a hex file.
CONTROLLER/UPLOAD reads theprogram memory in the microcon-troller, and provides a window dis-playing the contents word by word inhexadecimal notation. If a microcon-troller is read-protected an ‘empty’program memory is displayed (FFFFhfor AVRs, and 3FFFh for PICs). Outputis in the form of words, because eachof the two controller families has itsown memory organisation. With PICs,the word length is 14 bits, with AVRs,16 bits.CONTROLLER|WRITE TO EEP-ROM enables you to modify individ-ual bytes in the EEPROM range of themicrocontroller. The EEPROMs in thePIC and the AVR are organised byte-wise.CONTROLLER|UPLOAD EEPROMenables the complete EEPROM to beread, and the contents to be displayedbytewise in hexadecimal notation.Under CONTROLLER|CONFIGURECONTROLLER you can define thebasic configuration data of the relevantmicrocontroller. With the AVR chip, the
settings are very simple, because theyonly cover the Flash program protec-tion option. With PICs, the settings notonly control the program memory andEEPROM protection, but also the con-figuration of the oscillator and thepower-up and watchdog timers.CONTROLLER|ERASE CON-TROLLER enables microcontrollers tobe cleared (erased). This option is usefulif you want to reprogram a read-pro-tected PIC.CONTROLLER|START CON-TROLLER performs a hardware resetof the microcontroller.
P R O G R A M M I N G T H EM I C R O C O N T R O L L E R SA discussion of the programming pro-tocols for the two rival microcontrollerfamilies is, unfortunately, beyond thescope of this article. Those of you inter-ested in these matters are referred tothe following publications:- Atmel AVR Enhanced RISC Micro-controller Data Book- Microchip Datasheets DS30430A andDS30189D.
These publications not only containextensive descriptions of the program-ming algorithms and procedures rec-ommended by the respective chipmanufacturers, but also provide dis-cussions on the controller structuresand explanations of all assembly-codeinstructions.
Finally, excellent support for theAtmel AVR series of microcontrollers isavailable from our advertiser EquinoxTechnologies.
(980049-1)
References:1. Electronic Handyman,
Elektor Electronics December 1997.2. Programmer for
Handyman/AT90S1200,Elektor Electronics December 1997.
31Elektor Electronics 6/98
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
A 11flat -panel displays
current & new technologies
Flat -panelplays have been
in use in televi-i- sion receiversand computermonitors since
the early 1980s.Strictly speaking,all electronic dis-
play devices,except the cath-
ode ray tube(CRT) can be
classified as flatpanel types. But a 71tk Ir446.
Im Sony introduceda flat CRT television (Trinitron-rn as early as 1982. What the world is waitingfor now is the thin flat -panel display, which can be hung on a wall like a pic-
ture.ago,
lowering the cost of materials for the monolithic types and, more par-ticularly, the assembly costs of hybrid devices have posed far greater prob-
lems in practical applications than originally expected.
However, in spite of early optimism after its introduction some years
Elektor ElectronicsBy
our editorial staff
6/98Elektor Electronics 6/98
Flat-panel dis-plays have been
in use in televi-sion receiversand computer
monitors sincethe early 1980s.
Strictly speaking,all electronic dis-
play devices,except the cath-
ode ray tube(CRT) can be
classified as flatpanel types. ButSony introduceda flat CRT television (Trinitron™) as early as 1982. What the world is waiting
for now is the thin flat-panel display, which can be hung on a wall like a pic-ture. However, in spite of early optimism after its introduction some years
ago, lowering the cost of materials for the monolithic types and, more par-ticularly, the assembly costs of hybrid devices have posed far greater prob-
lems in practical applications than originally expected.
32
By our editorial staff
flat-panel displayscurrent & new technologies
Electronic display devices can bedivided into emissive and non-emis-sive types. The former include:
CRT (cathode ray tube);PDP (plasma display panel);ELD (electroluminescent display);VFD (vacuum fluorescent display);LED (light-emitting diode);
and the latter,
LCD (liquid crystal display);ECD (electrochemical display);EPID (electrophoretic image display);SPD (suspended particle display);TBD (twisting ball display);PLZT (transparent ceramics display).
The CRT was invented in 1897 in Ger-many (Braun); LCD watches were puton the market in 1972 in the USA; anECD clock was introduced in 1982 inJapan (Seiko); an experimental PDPTV was shown in 1978 in Japan(NHK); the light-emitting diode (laser)was invented in 1962 in the USA(Nathan) and the first trial LED TVwas shown in 1979 in Japan (Sanyo).
Undoubtedly, the bulk of the devel-opment work on flat-panel displayshas been, and still is, carried out in theUSA.
T H E M A R K E TThe CRT display, which has thelongest history (101years), is still the lead-ing type both economi-cally and for displayquality: it has almost 85per cent of the worldmarket. In 1997, hun-dreds of factories allover the world pro-duced 68 million CRTs,and this is expected togrow to 85 million bythe year 2001.
Number two is the LCD, which hasabout 10 per cent of the world market.In 1997, this type of display was man-ufactured in about 30 factories world-wide. The size of aperture (diagonal)of most of these displays varies from10 in to 12 in (9 in to 11 in viewable).Larger sizes are possible, but they areonly made in small numbers becausethese displays require new machinery,which is costly. Coupled with the factthat, owing to the explosion in thenotebook market, the demand far out-strips the production capacity, manu-facturers are loath to spend vast sumsof money on new machinery whilethis situation continues. In 1997, themarket demand for these displays was6.9 million, whereas production capac-ity was only 4.3 million.
Nevertheless, some manufacturersare busy modifying existing factoriesor building new ones to start produc-
tion of 14 in (13 inviewable) displayswithin the next year orso. It is, however,unlikely that these
developments will bring down prices,particularly since many manufacturersof LCDs also produce CRT displays.
L C D V S C R TFrom the point of view of the user, theLCD technology has advanced to thepoint where an LCD can replace a tra-ditional CRT, at least as far as thesmaller apertures are concerned. Tech-nical developments have made possi-ble better resolution, (relatively) largerapertures and higher frequency (cur-rently up to 240 MHz). However,LCDs still suffer from a limited view-ing angle, a limited contrast range, anda higher price than comparable CRTs.
P R I N C I P L E O F L C DIn a liquid crystal display, the applica-tion of a voltage changes the molecu-lar orientation of the liquid crystal, andthe resulting change in optical charac-
teristics, such as double refraction,optical rotation, dichroism, or opticalscattering caused by the reorientation,is converted into a visible change.
The most widely used types ofLCD are the thin-film transistor (TFT)and the super twisted nematic (STN).Basically, each of these is a passive dis-play that uses the modulation of lightwithin a liquid crystal cell. This cellconsists of a layer of liquid crystalabout 10 µm thick sandwichedbetween two glass substrates on whichare formed transparent electrodes.Grooves are etched in the surface ofthe electrode to impart a fixed orien-tation to the liquid crystal molecules.The grooves on each of the electrodesall point in the same direction and areused to direct the molecules mechani-cally, that is, the long molecules followthe grooves. See Figures 1a and 1b.
Owing to inter-molecular forcesbetween the crystals, there is no, or
33Elektor Electronics 6/98
transparentelectrode
liquid-crystalmolecules
supplyvoltage
off
supplyvoltage
on
transparentelectrode
light light 980020 - 11a
transparentelectrode polarizerpolarizer
light
light
off
on980020 - 11b
1
Figure 1. Basic con-struction of a liquidcrystal display (LCD).Liquid crystal is sand-wiched between twoglass substrates andthe construction iscompleted by an air-tight sealing materialaround the periphery.
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
* Nematic = having molecules or atoms ori-ented in parallel lines.
hardly any, randommovement of the mol-ecules which all adoptthe same orientation.
The direction of thegrooves on one elec-trode is, normally, atright angles to that ofthe grooves on theother electrode. In theSTN type, the molecular axis of the liq-uid crystal rotates continuouslythrough 90° between the two sub-strates. Since the pitch of the twist issufficiently large compared with thewavelength of visible light, the direc-tion of polarization of linearly polar-ized light incident vertically on one ofthe electrode substrates is rotatedthrough 90° by the twist of the liquidcrystal molecules as it passes throughthe cell. Therefore, the nematic cellshuts off light when placed betweentwo parallel polarizers and transmitslight when placed between two
orthogonal polarizers.When a voltage is
applied to a twistednematic cell, from a cer-tain threshold voltage,Uth, the molecule axesbegin to align them-selves with the electricfiled. When the appliedvoltage is about 2Uth,
the majority of molecules have theiraxes realigned in the direction of theelectric field, so that the 90° opticalrotary power is eliminated. In thiscase, in completely the opposite fash-ion to the case when there is no volt-age applied, between parallel polariz-ers, light is transmitted and betweenorthogonal polarizers, light will beshut off.
The basic operation as justdescribed is illustrated in Figure 1b. Itis based on the electrooptic effectwhen a twisted nematic cell is placedbetween two orthogonal polarizers. In
this case, light is transmitted whenthere is no voltage applied, and lightis blocked when a voltage is applied.Between parallel polarizers, however,the relationship between transmissionand blocking of light is reversed. So,twisted nematic LCDs provide a blackdisplay on a white background or awhite display on a black background.
A colour LCD contains in additiona colour filter for each pixel. Each ofthese pixels consists of three tiny dots(red, green and blue). This means thata colour display has three times asmany pixels as might be expected onthe basis of the graphic resolution.
The background of an LCD is a dif-fuse source of light, usually fluores-cent, placed behind or at the side ofthe display. A diffusor ensures that thelight falls uniformly over the display.
A serious drawback of an LCD isthat relatively little light is transmitted:in practice only 3–5 per cent. This poorefficiency is caused partly by the polar-izers (50%), and partly by the othertransparent layers, particularly theblack matrix (30%).
Manufacturers are constantly seek-ing to improve the efficiency, which isparticularly important for users ofnotebook computers. In such comput-ers, the life of the battery is already amatter of concern so that any improve-ment in efficiency will be more thanwelcome.
The company 3M has developed aspecial foil on which millions of smallprisms are deposited. These prismsensure that as much of the light aspossible is incident at right angles.Because of the bundling effect of theprisms, the power of the light sourcecan be halved.
A C T I V E O R P A S S I V ELiquid crystal displays come in twomain types: the inexpensive passivetypes (STN = super twisted nematic),and the more expensive active type(TFT = thin film transistor).
In the STN device, the liquid crys-tal is controlled by a two-dimensionalmatrix of conductors. These indium tinoxide conductors are deposited on tothe glass substrate. A potential appliedto the crossing of two conductorschanges the orientation of the crystal,which results in a pixel. Since thematrix is driven by the scan principle,this method is relatively slow. More-over, variations in the electric field dur-ing fast image changes (as when text isbeing scrolled or a cursor is reposi-tioned) cause annoying shadows. Thecontrast and the number of colours tobe reproduced are limited.
In a TFT device, each pixel has itsown transistor cell (whence its name).Transistor are deposited on one of thetwo glass substrates and can be con-trolled accurately and rapidly. This
34 Elektor Electronics 6/98
Black matrix Counter substrate
Color filter
Analyzer
Electricfield
Liquid-crystalmoleculePixel electrode
(metal)
Thin-film transistor
Thin-film-transistorsubstrate
Counter electrode(metal)
Polarizer
Super TFT LCD (IPS)
Black matrix Counter substrate
Color filter
Analyzer
Electricfield
Liquid-crystalmoleculeThin-film transistor
Counter electrode(transparent: ITO)
Thin-film-transistorsubstrate
Pixel electrode(transparent: ITO)
Polarizer
Conventionalthin-film-transistor LCD
Source: Hitachi Ltd. 980020 - 12
2
Figure 2. In an In-Plane Switching-mode(IPS) or Super ThinFilm Transistor (TFT)LCD, the twisting ofthe crystal moleculesis arranged with twoelectrodes positionedat the same side of thedevice.
type of display is therefore very suit-able for use with a large colour spec-trum and rapidly changing images.Owing to the amplification of the tran-sistor, the strength of the electric fieldacross the cell is greater than attainablewith passive displays. This, in turn,results in a better range of contrasts.
V I E W I N G A N G L EOne of the more serious drawbacks ofLCDs is their limited viewing angle of15–40°. The larger the angle the displayis viewed at, the more restricted thecontrast. This is particularly disturbingin the case of a colour display. To solvethis problem, a number of manufac-turers have introduceda new technique, calledvariously In PlaneSwitching-mode (IPS)or Super TFT. Thisenlarges the viewingangle to almost 140°(depending on themanufacturer).
The new technique has the addi-tional benefit of simplifying the pro-duction process. The two electrodesthat switch a pixel are deposited on tothe glass substrate together with thetransistor cell. This means that there isno longer a potential across the wholeLCD, but only across the pixel. Thearrangement is shown in Figure 2.
In the quiescentstate, when there is novoltage across the pixel,the molecules are inparallel with thegrooves in the elec-trodes. Twisting of themolecules within theliquid crystal, as in Fig-ure 1, does thereforenot occur. The polariz-ers, which are at right angles to oneanother, ensure that light is blockedover a large viewing angle, that is, thedisplay is black.
When a voltage is applied, the mol-ecules direct themselves according tothe electric field, which is at rightangles with the orientation of the mol-ecules in the quiescent state. Thestronger the electric field, the greaterthe rotation in the crystal, and themore light is transmitted.
P D P & F E DTwo other important flat-panel dis-plays are the plasma display panel,PDP, and the field emission display(FED).
The plasma display panel was firstintroduced in a paper in the USA in1954 (Skellet). The first trial colour PDPTV was exhibited in 1978 by NHK inJapan. More recently, Fujitsu of Japanand Philips of the Netherlands havecollaborated in the development of a
41 in wide TV screen. At present, theprice of PDP displays is much too highfor commercial exploitation, but it isexpected that it will come downappreciably within the next few years.At the same time, there are still somefundamental problems to be resolved
as well.The basic layout of a PDP is shown
in Figure 3. In current models, apotential of some 100 V is applied tothe electrodes. It is expected that innear-future models this voltage maybe lowered to about 60 V and in latermodels to about 10 V.
The life of a PDP is reckoned to besome 10,000 hours, comparable to thelife a typical television receiver.Although the picture on a current PDPis good, it is not good enough. Annoy-ing grey ghosting appears frequentlywith moving images.
Field emission displays combine thetechnology of the CRT with that of anLCD. This results in the excellent pic-ture obtainable from a CRT repro-duced on a flat panel display. The basiclayout of an FED is shown in Figure 4.
35Elektor Electronics 6/98
Front glassubstrate
Sustainelectrode
Buselectrodes
Separator
Address electrodesPhosphors (R) (G) (B)
Magnesiumoxide layer
Dielectriclayer
Rear glassubstrate
Plasma display
980020 - 13
3
Figure 3. A plasma display panel (PDP)is a good alternative to the LCD, but itwill take some time before this kind ofdisplay will become commerciallyaffordable.
Resistivelayer
Microtips
Cathode conductor
Glass
Column line
Cathode
Anode
One pixel
Glass faceplate
Gate row line
ITO layer
Bluephosphor
Greenphosphor
Greensubpixel
Bluesubpixel
Redsubpixel
Redphosphor
980020 - 14
Field-emission display4
Figure 4. In field emission displays, highelectric fields strip electrons from thetips of the cathode (whereas in a cath-ode ray tube thermionic emission isused). The field-emitted electronsexcite phosphors that may be identicalto those used in CRTs, thus producing aCRT-like image in a flat, thin package.
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
In principle, use is made of phosphorsthat are ignited by an electron beam.
The cathode consists of a strip ofconducting material on to which tinycones (some 10,000 per pixel) aredeposited. The cathode in this type ofdisplay fulfils the same function as theelectron gun in a CRT. A potential of200–800 V applied between the cath-ode and anode ensures that the phos-phor is ignited to generate light.
The FED technology is new. Pre-production models are available todesigners with aperture diagonals of5–6 in.
T H E C R T F I G H T S B A C KSo as to ensure their ruggedness(remember that there is a high vacuuminside), cathode ray tubes are curvedat the front. This curvature causessome distortion in the picture andresults in increased sensitivity to reflec-tions in the glass.
Flat panel type CRTs have beenaround for some time, but only withsmall apertures. However, Sony’sTrinitron™, RCA’s beam-guide CRT,and Mitsubishi’s Diamondtron™ thatare now available in aperture diago-nals of up to 50 in (full colour) are toall intents and purposes large size, flatpanel CRTs. Moreover, the recentlyintroduced Wega FD Trinitron™ fromSony is a true flat panel CRT. Becauseof the frontal flatness of these CRTs,the picture can be seen at large view-ing angles without loss of contrast orcolour resolution. Moreover, the CRTsdo not suffer from annoying reflec-tions They have been available for
some time for use in good-quality tele-vision sets and computer monitors.
The development of the flat panelCRT over the past twenty-odd yearshas been one of overcoming a greatmany technical difficulties. One ofthese was the design of a rugged,mechanical construction able to with-stand the large forces resulting fromthe high vacuum inside the tube. Sucha large, heavy tube is difficult to man-ufacture, but is now produced fromglass that is also used for automobilewindscreens.
Another difficulty was the use bymost manufacturers of a shadow maskor aperture grill that ensures correctdivision of the electron beam for thethree primary colours.
Sony’s Trinitron™ uses a singleelectron gun with three cathodesaligned horizontally, an aperture grille,and vertically striped phosphors. Thecathodes are tilted towards the centreso that the electron beams intersecttwice, once within the electron lensfocusing system and once at the aper-ture grille. This type of tube is there-fore much lighter and cheaper to pro-duce than three-gun tubes.
The accuracy with which the elec-tron gun can focus the electron beamis of vital importance for the quality ofthe picture. New technologies havemade it possible to modify the electrongun to such an extent that focusing isimproved greatly. This means that thetube need not be made deeper. More-over, the deflectors have been madelarger, which further increases theaccuracy of the beam. Other, smaller
modifications ensure that the electrongun gives a sharp picture at the edgesan in the corners of the tube, and min-imizes any distortion resulting fromspreading of the electrons in the cor-ners.
Other firms, such as Hitachi, Pana-sonic, LG, Mitsubishi and Samsunghave also developed flat panel CRTswhich will shortly be used in 17 in and19 in computer monitors (and, ofcourse, in smaller TV sets).
[980020]
37Elektor Electronics 6/98
12,1" 12,1" 12,1"
12,1" 12,1" 12,1"
980020 - 15
15,1" 15,1"
19" 19"
15,1" 15,1"
5
Figure 5. In the manu-facture of LCDs, use ismade on the productionline of material of fixeddimensions. The use ofother dimensionsresults in lower effi-ciency (profit!) and itmay, therefore, besome time before largeraperture displays willbecome available.
Any size you want,as long as it is standardIn 1997, there were in the whole world only three factories produc-
ing the latest generation of glass substrates for liquid crystal dis-
plays (LCDs) in sizes 550 × 650 mm. As shown by Figure 5, this
size substrate is ideal for the production of LCDs with an aperture
of 12 in. The production of different (more particularly, larger) size
substrates is not economically viable.
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
40 Elektor Electronics 6/98
E N V E L O P ED E M O D U L A T I O NAmplitude modulated (AM) signals aremost conveniently demodulated whentheir envelope is constructed, which isdone by rectification of the signal andits subsequent filtering in a low-passsection—see Figure 45. For the demod-ulation of the AM signal introduced inPart 5, we use experiment file XDE-MOD3.SPP. The result is shown in Fig-ure 46. The low-pass filtering is neces-sary to suppress any unwanted spec-tral components of the signal.
B B C R E V I S I T E DThe envelope detector may be used todemodulate the 198 kHz BBC signal(file BBC188. WAV) with a local oscillator(LO) frequency of 188 kHz, whichresults in an intermediate frequency(IF) of 10 kHz. This is done with fileXDEMOD3.SPP, and the result is heardwith the use of file TMP4.WAV. When westudy the spectrum of file BBC188.WAV
(experiment XDEMOD2B), we will note apeak at 10 kHz, which is the convertedcarrier (IF) of the signal, and a less pro-nounced peak at 5 kHz.
The weak spurious IF of 5 kHzresults from the mixing of the LO sig-nal with the carrier of Europa 1 (trans-mitting at 183 kHz). And, indeed,when file BBC188.WAV is demodulated
with a LO frequency that results in anIF of 5 kHz, Europa 1 will be heard (inFrench). A LO frequency that results inan IF of 19 kHz will also result in thereception of a broadcast transmitter at207 kHz.
S Y N C H R O N O U SD E M O D U L A T I O NThere are other means of demodulat-ing an AM signal. In the discussionabout amplitude modulation, it wasseen that when a signal is multipliedwith the sinusoidal output of a localoscillator (LO), the spectrum is shiftedby the frequency of the LO. It is, there-fore, possible to restore the originalspectrum by a second multiplication.The corresponding experiment, XDE-MOD5.SPP is shown in Figure 47, andthe result in Figure 48. If the LO signalis exactly the same as that used tomodulate the original signal, that sig-nal can be regained by multiplicationand low-pass filtering (Figure 3 andFigure 4 top).
Should the phase of the LO signalbe 90° out of phase with the carrier ofthe received signal, there would be nosignal left after the low-pass filtering. Itis, therefore, essential that the LO is inphase with the carrier of the AM signal.This is called synchronous or coherentdemodulation. When the two frequen-
This sixth and finalinstalment deals
with the subject ofmodulation and
demodulation.
By Dr. Ing. M. Ohsmann
introduction todigital signal processing
Part 6 – Modulationand demodulation
AM
signal
rectifier low-pass scope
980015 - 6 - 11
H (f)
f
Figure 45. Principle ofenvelope demodulation.
45
Before using theprograms on the ESPRESSO disk, copy
the entire folder from the CD-ROM to thehard disk: the programs can then be runfrom the hard disk. This procedure isalso explained in the README file on theCD-ROM.When copying a file or folder under Win-dows, its read-only setting is alsocopied. With many ESPRESSO programs,the read-only attribute causes an errorreport to appear, or graphics to disap-pear. This problem is solved by using theExplorer, selecting all files in theESPRESSO folder on the hard disk, andclicking on file ‘Properties’. Remove thecheck in the read-only box by clicking onit. Close by pressing OK, whereuponeverything should function normally.
O2@@@@@@@?O2@@@@@@@@@@@@@L
?O2@@@@@@@@@@@@@@@@@@1O2@@@@@@@@@@@@@@@@@@@@@@@@@@L?
?O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1??O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?
?O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?
O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@??O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L
O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?O@?2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?
?O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1??O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?
?O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@??O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?
?O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?O2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L
?W2@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1W&@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@7@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@??@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@LN@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@??@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?J@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?hg?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?hg@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?hg@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?hg@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@Lhg3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1hgN@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@hg?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@hg?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@hg?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@hg?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?hf@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?hf@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@5?hf@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@hg@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?hf3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?hfN@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?hf?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@Lhf?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1hf?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@hf?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@hf?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@hf@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?he@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?he@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?he@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?he@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?he3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?heN@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@Lhe?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1he?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@he?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@he@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@he
?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@he?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@he@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?h@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?h@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?h3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?hN@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?h?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@Lh?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1h?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@h?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@h?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@h?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?g?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?g@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?g@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?g3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?gN@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?g?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@Lg?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1g?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@g?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@g?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@g?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@L?f@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@1?f@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?f@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@?f@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@5?f3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@(Y?fN@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0Yg?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0Mhe?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M?hg?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M?
?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M?N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M?3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0MN@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M??@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M?@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M??@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M??@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M??3@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M??N@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0?@M@@@@@@@@@@@@@@@@@@@@@@@@@@@@@0M?3@@@@@@@@@@@@@@@@@@@@@@0M?N@@@@@@@@@@@@@@@0?@M?3@@@@@@@@@@0M?V4@@0?@M?
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor uk
46
t mp. way 1..600 AUTO
tmp1. uav 1..600
cies are shifted in phasewith respect to oneanother, the 'in -phase' and '900 phaseshift' states alternate in the rhythm ofthe frequency difference, resulting in afluctuation of the volume of the audiosignal. For that reason, when a coher-ent detector is used, the receiverrequires a circuit, for instance, a phase -locked loop (PLL), to ensure that thefrequency and phase of the LO signalare correct at all times. Although syn-chronous detection of amplitude mod-ulated signals is complicated, it pro-vides optimum results.
.1
am. way 1.. 600 AUTO
AUTO t mp2. wav
Figure 46. Amplitude -modulated and-demodulated signals.
HIGH -RESOLUTIONSPECTRUM ANALYSERWe have already seen that with the aidof an AM signal generator or by meansof multiplication with a sinusoidal fre-quency the spectrum of a signal maybe shifted. This principle is used inmodern receivers to convert the inputsignal into an intermediate frequency(IF) signal. In a similar manner, it ispossible to enhance the resolution ofour spectrum analysersPEcl. When the origi-nal signal is scanned ata rate of 22050 samplesper second, using aDFT of 4096 samples,the result is an intervalof 5.3833 Hz betweenadjacent samples. To be
programme
able to view a spectrumrange more precisely, it
is possible to shift it into the low -fre-quency region and subject it to low-pass filtering and d own samp ling.
In practice, this will be be done inthe following manner. To resolve signalwElE.wAv at about 5 kHz, that is, toobtain a resolution of about 0.5 Hzwith a DFT of 4096 samples, it is neces-sary to reduce the sampling rate of22050 by a factor 10. At 2205 samples/s,it becomes possible to display signals atfrequencies of up to 1.1 kHz. The5 kHz signal from wElE.wAv can be
1111111111F. Synchro-nous or coherentdemodulation with theoriginal carrier andwith a phase -shiftedcarrier.
AM
1.. 600 AUTO980015 - 6 - 12
shifted into this range when it isreduced to 300 Hz by mixing it with afrequency of 5.3 kHz.
If we start with 40960 samples, 4096remain, and these resolve file XSPEC.SPP:the result is in Figure 49. The carrierand the two signals at its left and right(originally 57 kHz) are easily recog-nized. The signal at a distance of125 Hz is the ARI signal which isswitched on when there is travelnews. The other signal, at 34.93 Hzfrom the carrier is the station identifi-cation signal that indicates in whichregion the transmitter whose signal has
LO
local oscillator
scope
scope
980015 - 6 - 13
Elektor Electronics 6/98 41
cies are shifted in phasewith respect to oneanother, the ‘in-phase’ and ‘90° phaseshift’ states alternate in the rhythm ofthe frequency difference, resulting in afluctuation of the volume of the audiosignal. For that reason, when a coher-ent detector is used, the receiverrequires a circuit, for instance, a phase-locked loop (PLL), to ensure that thefrequency and phase of the LO signalare correct at all times. Although syn-chronous detection of amplitude mod-ulated signals is complicated, it pro-vides optimum results.
H I G H - R E S O L U T I O NS P E C T R U M A N A L Y S E RWe have already seen that with the aidof an AM signal generator or by meansof multiplication with a sinusoidal fre-quency the spectrum of a signal maybe shifted. This principle is used inmodern receivers to convert the inputsignal into an intermediate frequency(IF) signal. In a similar manner, it ispossible to enhance the resolution ofour spectrum analyser programmeSPEC1. When the origi-nal signal is scanned ata rate of 22050 samplesper second, using aDFT of 4096 samples,the result is an intervalof 5.3833 Hz betweenadjacent samples. To be
able to view a spectrumrange more precisely, it
is possible to shift it into the low-fre-quency region and subject it to low-pass filtering and downsampling.
In practice, this will be be done inthe following manner. To resolve signalWD1L.WAV at about 5 kHz, that is, toobtain a resolution of about 0.5 Hzwith a DFT of 4096 samples, it is neces-sary to reduce the sampling rate of22050 by a factor 10. At 2205 samples/s,it becomes possible to display signals atfrequencies of up to 1.1 kHz. The5 kHz signal from WD1L.WAV can be
shifted into this range when it isreduced to 300 Hz by mixing it with afrequency of 5.3 kHz.
If we start with 40960 samples, 4096remain, and these resolve file XSPEC.SPP:the result is in Figure 49. The carrierand the two signals at its left and right(originally 57 kHz) are easily recog-nized. The signal at a distance of125 Hz is the ARI signal which isswitched on when there is travelnews. The other signal, at 34.93 Hzfrom the carrier is the station identifi-cation signal that indicates in whichregion the transmitter whose signal has
41Elektor Electronics 6/98
Figure 46. Amplitude-modulated and-demodulated signals.
H (f)
H (f)
f
f
AM
LO
scope
scope
sine
cosine
signal
tmp1
tmp3
tmp2
tmp4
980015 - 6 - 13
local oscillator
Figure 47. Synchro-nous or coherentdemodulation with theoriginal carrier andwith a phase-shiftedcarrier.
46
47
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
tmpl. way 1.. 600
t mp3. way 1.. 600
Figure 48. Top: coher-ent demodulation; bot-tom: synchronousdemodulation with thelocal oscillator fre-quency shifted by 90°.
Figure 49. High defini-tion spectrum of the57 kHz subcarrier forthe ARI signal.
tmp2. uav 1..600
t mp4. way
just been received is situated. By theway, when the first portion of the fileWD1L.WAV is viewed, that is, when themessage is not yet coming through, the125 Hz peak will not be found, sincethe message recognition modulation isthen not yet actuated.
QUADRATUREMODULATION ANDDEMODULATIONThe discussion on coherent demodu-lation showed that the phase of thelocal oscillator signal must be equal tothat of the carrier frequency, and thatif there is a phase shift between the
tap4. war fr.,/ ya = 0.0 yb = 100.0fa = 0.0 fb = 1102.5 arc -1.0
1..600980015 -6 - 14
two, there is no output. This phenom-enon may be made use of when anAM signal with sinusoidal carrier anda second carrier of the same frequencybut cosine shaped are used. It is thenpossible to convey both signals via thesame channel (Figure 50) and todemodulate them with suitable LO fre-quencies into their original shape. Thisis simulated in experiment xoEmoD4.SPP. In this, the first signal is a triangu-lar one and the second, a sinusoidalone. Modulating processes that usesinusoidal and cosinusoidal carrierssimultaneously are called quadratureprocesses. The main carrier and its con-stituents are normally identified by theletter I (In -phase) and the other withthe letter Q (Quadrature).
RDS DETECTIONA practical application of quadraturedemodulation is found in VHF andUHF radio. A car radio, for instance, isswitched on or off when travel news isbroadcast on a 57 kHz subcarrier. Thisis amplitude -modulated with an audiofrequency of, say, 125 Hz when travelnews is being broadcast. This is the Iconstituent.
Today, the same channel is used forthe radio data system (RDS). To pre-vent the ARI and RDS signals affectingone another, the subcarrier for the RDSis 90° out of phase with the main car-rier. This means that the RDS data areconveyed in the Q constituent. Experi-
A42 Elektor Electronics 6/98
just been received is situated. By theway, when the first portion of the fileWD1L.WAV is viewed, that is, when themessage is not yet coming through, the125 Hz peak will not be found, sincethe message recognition modulation isthen not yet actuated.
Q U A D R A T U R EM O D U L A T I O N A N DD E M O D U L A T I O NThe discussion on coherent demodu-lation showed that the phase of thelocal oscillator signal must be equal tothat of the carrier frequency, and thatif there is a phase shift between the
two, there is no output. This phenom-enon may be made use of when anAM signal with sinusoidal carrier anda second carrier of the same frequencybut cosine shaped are used. It is thenpossible to convey both signals via thesame channel (Figure 50) and todemodulate them with suitable LO fre-quencies into their original shape. Thisis simulated in experiment XDEMOD4.SPP. In this, the first signal is a triangu-lar one and the second, a sinusoidalone. Modulating processes that usesinusoidal and cosinusoidal carrierssimultaneously are called quadratureprocesses. The main carrier and its con-stituents are normally identified by theletter I (In-phase) and the other withthe letter Q (Quadrature).
R D S D E T E C T I O NA practical application of quadraturedemodulation is found in VHF andUHF radio. A car radio, for instance, isswitched on or off when travel news isbroadcast on a 57 kHz subcarrier. Thisis amplitude-modulated with an audiofrequency of, say, 125 Hz when travelnews is being broadcast. This is the Iconstituent.
Today, the same channel is used forthe radio data system (RDS). To pre-vent the ARI and RDS signals affectingone another, the subcarrier for the RDSis 90° out of phase with the main car-rier. This means that the RDS data areconveyed in the Q constituent. Experi-
42 Elektor Electronics 6/98
Figure 48. Top: coher-ent demodulation; bot-tom: synchronousdemodulation with thelocal oscillator fre-quency shifted by 90°.
Figure 49. High defini-tion spectrum of the57 kHz subcarrier forthe ARI signal.
48
49
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor uk
ment xops2.sPP simulates the ARI andRDS demodulator in Figure 51. Thephase of the LO signal is normally con-trolled by a phase -locked loop (PLL) toensure that it is correct with referenceto the carrier. The demodulated signalsmay be seen in Figure 52. The uppersignal represents the sinusoidal stationidentification signal (34.93 Hz, whilethe lower one shows the RDS signal.
FREQUENCYMODULATIONIn amplitude modulation, the modu-lating signal is shifted in the frequencyband by the carrier. In frequency mod-ulation, however, the modulating sig-nal, s(t), modifies the frequency of thecarrier. The transmitted signal is
y(t) = cos[2.7cfct + ps(t)].
Strictly speaking, this treats frequencymodulation as a variation of phasemodulation. There is then no longer asimple connection for calculating thespectrum of the transmitted signal.Even with a sinusoidal signal, the spec-trum soon becomes fairly complex:
y(t) = cos[2.7cfct + ucos(27cfmt)].
Figure 53 shows the spectra when thecarrier frequency, fc = 2000 Hz, themodulating frequ ency,fm = 100 Hz or200 Hz, and u = 2.4048 or 4.0. It will benoted that single spectral frequenciesoccur at a distance f, from the carrier.The carrier itself must not be present inthe signal. The magnitude of the singlefrequencies may be calculated withBesse' functions. Note that the FMspectrum of a transmitted signal iswider than that of the modulating fre-quency. Although the audio band-width is only 15 kHz, the FM spectrumextends up to 150 kHz from the carrier.It is this aspect that makes an FM sig-nal less vulnerable to interference,since the narrow -band interferenceaffects only a small part of the FMspectrum. The requisite Carson band-width is calculated from the band-width and the energy of the modulat-ing signal.
WARBLE SIGNALThe CD-ROM accompanying thecourse contains an example of a simpleFM signal with sinusoidal modulation.In Germany, ARD (Arbeitsgemein-sch aft der often tlich -rechtligen Ru nd -to nkan stalten der BundesrepublikDeutschland or, simply, Germany I)transmits a special signal to indicatethe onset of travel news. This consistsof a 2350 Hz carrier that is frequencymodulated at 123 Hz with a modula-tion index of 1. See Figure 54.
[This is, of course, of no interest toreaders in the UK, where the BBC andIndependent Radio Stations provide a
51
signal 1
carrieroscillator
I -part
tmp3
sine
channel
tmp5
signal 2
j 0
cosine
low-pass
sine
cosine
mixer
signal 1
tmp7
f
LO
local oscillator
low-pass
0 -part
Figure 50. Compositionof a system for quadra-ture modulation.
Figure 52. Top: demodulation ofthe ARI signal; bottom: demodu-lation of the RDS signal.
52
signal 2
tmp9
980015- 6-16
Figure 51. This is howthe ARI and RDS sig-nals are demodulated.
f = 7500 Hz980015 -6-17
tf6p4.(Aav 191.170
trap& tfav 1..801 ano080.5-6-18
Elektor Electronics 6/98 43
ment XRDS2.SPP simulates the ARI andRDS demodulator in Figure 51. Thephase of the LO signal is normally con-trolled by a phase-locked loop (PLL) toensure that it is correct with referenceto the carrier. The demodulated signalsmay be seen in Figure 52. The uppersignal represents the sinusoidal stationidentification signal (34.93 Hz, whilethe lower one shows the RDS signal.
F R E Q U E N C YM O D U L A T I O NIn amplitude modulation, the modu-lating signal is shifted in the frequencyband by the carrier. In frequency mod-ulation, however, the modulating sig-nal, s(t), modifies the frequency of thecarrier. The transmitted signal is
y(t) = cos[2πfCt + µs(t)].
Strictly speaking, this treats frequencymodulation as a variation of phasemodulation. There is then no longer asimple connection for calculating thespectrum of the transmitted signal.Even with a sinusoidal signal, the spec-trum soon becomes fairly complex:
y(t) = cos[2πfCt + µcos(2πfMt)].
Figure 53 shows the spectra when thecarrier frequency, fC = 2000 Hz, themodulating frequency, fM = 100 Hz or200 Hz, and µ = 2.4048 or 4.0. It will benoted that single spectral frequenciesoccur at a distance fM from the carrier.The carrier itself must not be present inthe signal. The magnitude of the singlefrequencies may be calculated withBessel functions. Note that the FMspectrum of a transmitted signal iswider than that of the modulating fre-quency. Although the audio band-width is only 15 kHz, the FM spectrumextends up to 150 kHz from the carrier.It is this aspect that makes an FM sig-nal less vulnerable to interference,since the narrow-band interferenceaffects only a small part of the FMspectrum. The requisite Carson band-width is calculated from the band-width and the energy of the modulat-ing signal.
W A R B L E S I G N A LThe CD-ROM accompanying thecourse contains an example of a simpleFM signal with sinusoidal modulation.In Germany, ARD (Arbeitsgemein-schaft der öffentlich-rechtligen Rund-funkanstalten der BundesrepublikDeutschland or, simply, Germany I)transmits a special signal to indicatethe onset of travel news. This consistsof a 2350 Hz carrier that is frequencymodulated at 123 Hz with a modula-tion index of 1. See Figure 54.
[This is, of course, of no interest toreaders in the UK, where the BBC andIndependent Radio Stations provide a
43Elektor Electronics 6/98
H (f)
H (f)
f
f
LO
signal 1
signal 2
signal 1
signal 2
sine
cosine
local oscillator
mixer
channel
sine
cosine
low-pass
low-pass
carrieroscillator
tmp3
tmp5
tmp7
tmp9
980015 - 6 - 16
I-part
Q-part
H (f)
H (f)
f
f
ARI + RDS
ARI
RDS
980015 - 6 - 17
fC = 1500 Hz
fC = 7500 Hz
Figure 50. Compositionof a system for quadra-ture modulation.
Figure 51. This is howthe ARI and RDS sig-nals are demodulated.
50
51
52
Figure 52. Top: demodulation ofthe ARI signal; bottom: demodu-lation of the RDS signal.
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
53
t mp2. way =fa = 0.0 fb =
4.0 yb =4000.4 cur=
100.0 tmp4. way Cu) ya = 0.0 yb = 100.1-1.0 fa = 0.0 fb = 4000.0 cur= -1.1
t mp6. uav <L,I) ya = 0.0 yb =la = 0.0 fb = 4000.0 cur=
travel service underTP/TA and EON.When the traffic modeis selected on a receiver, iover to the local stationNews is broadcast.Cassette or CD listen-ing will also be inter-rupted, and typicallythe voluem of the set
Figure 53. FM spec-trum when the modu-lating signal is sinu-soidal.
At will switchwhen Travel
Figure 54.trum of the special(warble) signal trans-mitted in Germanyahead of travel news.
...
100.0 t ;Ape. uav ya = 0.0 yb =-1.0 fa = 0.0 fb = 4000.0 cur=
will increase a little todraw the driver's atten-tion to the
bulletin.When used in conjunctionwith EON, the TA/TP mode allows lis-
tening to BBC Nationalstations to be inter-rupted with travelreports for the area fromBBC Local stations.
t mp. way <u) ya = 4.0 yb =
tad100.-1.
980015 -6 -15
EON - Enhanced Other Networks -provides a cross reference to other sta-tions for the Travel Service, and otherfeatures.
Editor]
RADIO TELETYPEThe radio teletype service invariablyuses frequency shift keying (FSK) toconvey teletype signals. Such a signal
100.0 t mp2. way <u) ya = 0.0 yb = 100.fa = 1735.0 fb = 2965.0 cur= -1.0 f a = 1735.0 fb = 2965.0 cur= -1.
44
980015 - 6 - 20
Elektor Electronics 6/98
travel service underTP/TA and EON.When the traffic modeis selected on a receiver, it will switchover to the local station when TravelNews is broadcast.Cassette or CD listen-ing will also be inter-rupted, and typicallythe voluem of the set
will increase a little todraw the driver’s atten-tion to the
bulletin.When used in conjunctionwith EON, the TA/TP mode allows lis-
tening to BBC Nationalstations to be inter-rupted with travelreports for the area fromBBC Local stations.
EON – Enhanced Other Networks –provides a cross reference to other sta-tions for the Travel Service, and otherfeatures.
Editor]
R A D I O T E L E T Y P EThe radio teletype service invariablyuses frequency shift keying (FSK) toconvey teletype signals. Such a signal
44 Elektor Electronics 6/98
Figure 53. FM spec-trum when the modu-lating signal is sinu-soidal.
Figure 54. FM spec-trum of the special(warble) signal trans-mitted in Germanyahead of travel news.
53
54
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor uk
(see Figure 55) has a quiescent state,marked SPACE, which pertains whenthe teletype service is not active. A sin-gle character begins with the START BIT,at which the signal transfers to theMARK state. The start bit lasts, in case ofa transfer rate of 50 symbols, 1/50 =20 ms. This is followed by the data bits.The old Baudot code (which is stillused here and there) uses five bits torepresent 64 alphanumeric characters.Each bit has the same length as thestart bit. These are followed by a STOPBIT that lasts 1.42 data bits whereuponthe system is in state SPACE. There thenfollows a fairly long pause (asynchro-nous transfer), after which the data bitstream is continued
In FSK, the MARK state is given a fre-quency of 850 Hz, and SPACE one of1350 Hz. The signals are transmitted inline with the prevailing state by meansof, for instance, a voltage -controlledoscillator (VCO). The distance betweenthe two frequencies is called the SHIFT.File RTTY1.WAV contains such a signalwhich has a rate of 50 bit/s, MARK =850 Hz and SPACE= 1350 Hz. The spec-trum of the teletype signal may be gen-erated with experiment xurvl.SPP. Thisshows two conspicuous peaks at theSPACE and MARK frequencies. Sincethese signals are well within the range300-3400 Hz, they are readily con-veyed via a radio telephony channel.
MARK = 850 Hz
SPACE = 1300 Hz
START BIT
20 ms
STOP BIT
I I I
Feader bitBIT = 0 = MARK
BAUDOT character "8" = 00110
Figure 55. Typicalradio teletype signal.
850 Hzband-pass
980015 - 6 - 21
Figure 56. Configura-tion of a filter decoderfor radio teletype sig-nals.
low-passSchmitttrigger
BITS
980015 - 6 - 22
Figure 57. Signals ine filter decoder.
trip. tsFav 1..5513 AUTO
tmpl. way 1.. 5513
-H-----41.1.----41114W-441111111-41411111W
41111111. -114=111 -09 -01* --4t mp2. way 1.. 5513
t mp5. way 1.. 5513
tmp6. wav 1..5513
tmp7. way 1.. 5513
Elektor Electronics 6/98 45
980015-6-23
(see Figure 55) has a quiescent state,marked SPACE, which pertains whenthe teletype service is not active. A sin-gle character begins with the START BIT,at which the signal transfers to theMARK state. The start bit lasts, in case ofa transfer rate of 50 symbols, 1/50 =20 ms. This is followed by the data bits.The old Baudot code (which is stillused here and there) uses five bits torepresent 64 alphanumeric characters.Each bit has the same length as thestart bit. These are followed by a STOP
BIT that lasts 1.42 data bits whereuponthe system is in state SPACE. There thenfollows a fairly long pause (asynchro-nous transfer), after which the data bitstream is continued
In FSK, the MARK state is given a fre-quency of 850 Hz, and SPACE one of1350 Hz. The signals are transmitted inline with the prevailing state by meansof, for instance, a voltage-controlledoscillator (VCO). The distance betweenthe two frequencies is called the SHIFT.File RTTY1.WAV contains such a signalwhich has a rate of 50 bit/s, MARK =850 Hz and SPACE = 1350 Hz. The spec-trum of the teletype signal may be gen-erated with experiment XRTTY1.SPP. Thisshows two conspicuous peaks at theSPACE and MARK frequencies. Sincethese signals are well within the range300–3400 Hz, they are readily con-veyed via a radio telephony channel.
45Elektor Electronics 6/98
START BIT STOP BIT
0 1 1 0 0
20 ms
980015 - 6 - 21
leader bitBIT = 0 = MARK
SPACE = 1300 Hz
BAUDOT character "8" = 00110
MARK = 850 Hz
tmp1
tmp2
tmp3
tmp4
tmp5 tmp6
tmp7
tmp.wav
850 Hzband-pass
1300 Hzband-pass
rectification
low-passSchmitttrigger
BITS
980015 - 6 - 22
Figure 55. Typicalradio teletype signal. Figure 56. Configura-
tion of a filter decoderfor radio teletype sig-nals.
Figure 57. Signals inthe filter decoder.
55
56
57
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor uk
810 .0 10 10 .0 10 10 .0 10
- -
R YR YR YR YR 'IRV.0 10 10 .0 .0 .0 .0 10 10 .0 e0 sO
...... .......
RADIO TELETYPEDECODINGAn FSK-modulated signal may bedecoded by a traditional filter decoderas shown in Figure 56; it has beendone so in Experiment xftrTY6.sPP. Theresulting waveforms are shown in Fig-ure 57, while the decoded text is givenin Figure 58. The preferred test text isthe character string RYRYRYR since thiscontains the fastest bit conversion(xurv5.sPP).
FINALLYAlthough this is the end of the six -partcourse, there are a number of aspectsthat, owing to shortage of space, havenot been discussed. Experiments rele-vant to some of these are listed in thebox and contained on the CD-ROMaccompanying the course. The title ofthe disk (see Readers Services towardsthe end of this issue) is ESPRESSO, an
acronym for Elektor Electronics SignalPRocessing Experiments and Simula-tion SOftware. With the knowledgegained by the course, readers will beable to conduct these experimentswithout further discussion. Neverthe-less, a small help programme is con-tained in file ExPs.Doc.
[980015-6]
980015 - 8 - 24
Figure 58. Decoded textof a radio teletype sig-nal.
Additional experiments on the CD-ROMSingle-sideband generationFilter method and phase methodSingle-sideband demodulationPhase modulationAMDSBBCTerrestrial meteorological broadcastsSatellite meteorological broadcastsTime signal broadcastsDTMF (dual tone multiple frequency)
Elektor Electronics 6/98 47L47Elektor Electronics 6/98
R A D I O T E L E T Y P ED E C O D I N GAn FSK-modulated signal may bedecoded by a traditional filter decoderas shown in Figure 56; it has beendone so in Experiment XRTTY6.SPP. Theresulting waveforms are shown in Fig-ure 57, while the decoded text is givenin Figure 58. The preferred test text isthe character string RYRYRYR since thiscontains the fastest bit conversion(XRTTY5.SPP).
F I N A L L YAlthough this is the end of the six-partcourse, there are a number of aspectsthat, owing to shortage of space, havenot been discussed. Experiments rele-vant to some of these are listed in thebox and contained on the CD-ROMaccompanying the course. The title ofthe disk (see Readers Services towardsthe end of this issue) is ESPRESSO, an
acronym for Elektor Electronics SignalPRocessing Experiments and Simula-tion SOftware. With the knowledgegained by the course, readers will beable to conduct these experimentswithout further discussion. Neverthe-less, a small help programme is con-tained in file EXPS.DOC.
[980015-6]
Figure 58. Decoded textof a radio teletype sig-nal.
Additional experiments on the CD-ROMSingle-sideband generationFilter method and phase methodSingle-sideband demodulationPhase modulationAMDSBBCTerrestrial meteorological broadcastsSatellite meteorological broadcastsTime signal broadcastsDTMF (dual tone multiple frequency)
58
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
discharge circuitfor 1.2 V sintered NiCd batteries
Developments in theworld of rechargeablebatteries, and the new
models that haveappeared in recentyears, such as the
metal hydride batteryand the lithium -ion
battery, are mainly ofinterest in the elec-
tronics industry. Manyexperimenters, hob-
byists and other ama-teurs have stuck to
NiCd cells and batter-ies, which have now
been with us sincethe 1950s. The mainreason for this popu-
larity is probably theiruser-friendly behav-
iour (which isexceeded only by sealed lead -acid batteries). Another factor in favour ofthe NiCd battery is its low internal resistance (at least as far as sintered
types are concerned), which enables it to provide fairly large currents(but not as high as lead -acid batteries).
Design by K Walraven
The nickel -cadmium battery ismechanically rugged and long-lived. Ithas excellent low -temperature charac-teristics and can be hermetically sealed.Cost, however, is higher than for thelead -acid or nickel -zinc battery. Inmany applications, the use of a sealedlead -acid battery is to be preferred overthe other two types.
A slight drawback of a sinteredNiCd battery is its so-called memoryeffect, which is fortunately completelyreversible. It should be noted that massplate nickel -cadmium cells and batteriesdo not develop the memory effect in
a.>
any circumstances. The present circuitis, therefore, primarily intended for usewith sintered NiCd 1.2 V cells.
LOW INTERNALRESISTANCEThe ability of NiCd batteries to providefairly large currents (because of theirlow internal resistance - at least as faras sintered types are concerned) is animportant factor for the amateur frater-nity, since many home-made modelunits draw fairly large currents. As acomparison, the d.c. resistance of threetypes of fully charged 1 Ah, 1.2 V
48 Elektor Electronics 6/98Elektor Electronics 6/98
The nickel-cadmium battery ismechanically rugged and long-lived. Ithas excellent low-temperature charac-teristics and can be hermetically sealed.Cost, however, is higher than for thelead-acid or nickel-zinc battery. Inmany applications, the use of a sealedlead-acid battery is to be preferred overthe other two types.
A slight drawback of a sinteredNiCd battery is its so-called memoryeffect, which is fortunately completelyreversible. It should be noted that massplate nickel-cadmium cells and batteriesdo not develop the memory effect in
any circumstances. The present circuitis, therefore, primarily intended for usewith sintered NiCd 1.2 V cells.
L O W I N T E R N A LR E S I S T A N C EThe ability of NiCd batteries to providefairly large currents (because of theirlow internal resistance – at least as faras sintered types are concerned) is animportant factor for the amateur frater-nity, since many home-made modelunits draw fairly large currents. As acomparison, the d.c. resistance of threetypes of fully charged 1 Ah, 1.2 V
Developments in theworld of rechargeablebatteries, and the new
models that haveappeared in recentyears, such as the
metal hydride batteryand the lithium-ion
battery, are mainly ofinterest in the elec-
tronics industry. Manyexperimenters, hob-
byists and other ama-teurs have stuck to
NiCd cells and batter-ies, which have now
been with us sincethe 1950s. The mainreason for this popu-
larity is probably theiruser-friendly behav-
iour (which isexceeded only by sealed lead-acid batteries). Another factor in favour of
the NiCd battery is its low internal resistance (at least as far as sinteredtypes are concerned), which enables it to provide fairly large currents
(but not as high as lead-acid batteries).
48
Design by K Walraven
discharge circuit for 1.2 V sintered NiCd batteries
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
sealed cell is
Standard 110 m52/cellHeavy duty 50 m52/cellSintered 19 m52/cell
ENVIRONMENTALEFFECTSOne of the most serious drawbacks ofNiCd batteries is their effect on theenvironment. This type of battery con-tains cadmium which is toxic. In mostcountries, discarded NiCd batteries aredumped on the rubbish heap wherethey remain toxic for a very long time.It is true, of course, that their life ofsome 500-800 charge/discharge cyclesdoes not cause millions of them to bedisposed of on the rubbish heap. Nev-ertheless, this was a very important fac-tor in the decision of manufacturers ingeneral to discontinue the use of NiCdbatteries in most consumer products.
MEMORY EFFECTAnother disadvantage of sintered (notmass plate) NiCd batteries is, asalready mentioned, their memoryeffect. This manifests itself in the cellretaining the characteristics of previouscycling. That is, after repeated shallow -depth discharges the cell will fail toprovide a satisfactory full -depth dis-charge. Note, however, that Evereadycylindrical nickel -cadmium cells areparticularly noted for their lack ofmemory effect.
The memory effect is a nuisance,because it means that a battery with anominal capacity of, say, 600 mAh, aftera number of charge/discharge cycleshas a useful capacity of only 300 or 400mAh. This has nothing to do with thelife of the battery: even a new batteryif charged as stated will soon lose partof its capacity.
Fortunately, this reduction in capac-ity can be prevented fairly simply.Moreover, batteries that already sufferfrom the memory effect can berestored to their nominal capacity. Thecure is simply to ensure that a batteryis occasionally fully discharged beforeit is recharged. Occasionally meansbefore every third or so recharge. Notethat there are chargers on the marketthat have the discharge facility built in,but this will certainly not be the case inthe less expensive types.
CORRECTDISCHARGINGThere is no need for extensive circuitryto discharge a battery: a simple resis-tor or light bulb will accomplish itreadily. It is, however, necessary tokeep an eye on the discharge time,because otherwise there is the risk thatthe battery is discharged beyond a cer-tain voltage. When this happens, itmay cause polarity reversal in the cells
comprising the battery.Correct discharging can only take
place via a circuit that arranges for thebattery to be discharged to a certainlevel and then disconnects it from thedischarge circuit.
The diagram of such a circuit - seeFigure 1- is pretty straightforward.Nevertheless, such a discharge circuitdoes the job correctly. It causes a bat-tery to be discharged to a level of650 mV This level ensures that the bat-tery is correctly discharged without therisk of polarity reversal. The battery isnot discharged at a constant current,but in short bursts. This allows the bat-tery to 'recover' during the intervals,which, in practice, has been found toextend its useful life.
During the discharge, an LED lightsto show that the process is continuing.Since the diode cannot work from avoltage of 0.65-1.2 V, the voltage has tobe raised. To this end, the astable mul-tivibrator formed by T1 and T2 oscil-lates at a rate of 25 kHz. When T2 is on,current flows through inductor L1, sothat energy is stored in the magneticfield. When T2 is off, the inductor is'discharged' via the LED, whereuponthis lights.
Diode D1 prevents the energystored in the inductor from leakingaway via the base of T1. This mighthappen because the capacitors in thecircuit have fairly high values, whereasthe resistors have low ones. The cho-sen values ensure that the dischargecurrent is sufficiently high. When thebattery voltage is 1.2 V, the dischargecurrent is some 200 mA; at 0.8 V, it hasdropped to about 100 mA, and at0.65 V to around 50 mA. When the bat-tery voltage drops to 0.65 V, the dis-charge process is discontinued.
CONSTRUCTIONThe tiny circuit is best built on theprinted -circuit board shown in Fig-ure 2, but this is not available readymade. However, a small prototypingboard will do very nicely as well.
Inductor L1 is a small choke whichshould be readily available from mostelectronics retailers.
The LED should be a high -effi-ciency type, while, because of thethreshold discharge voltage, D1 mustbe a Schottky type.
USAGEThere is not much that can be saidabout using the discharge unit. It is sim-ply a matter of connecting the 1.2 V bat-tery with correct polarity, checking thatthe LED lights, and disconnecting thebattery when the LED goes (or is) out.
In general, the discharge period willnormally be three to four hours. Asmentioned before, the battery does notneed to be discharged fully before it isrecharged: before every third recharge
1
2
Figure 1. The circuit is basicallyan astable multivibrator oscillat-ing at a rate of 25 kHz.
Figure 2. The discharge circuit isintended for 1.2 V batteries. Ifseveral of these, or a 9 Vrechargeable NiCd battery, haveto be discharged, an appropriatenumber of PCBs are needed.
Parts listResistors:R1, R4 = 4.7 f2R2, R3 = 100 n
Capacitors:CI = 0.22 µFC2 = 0.47 µF
Inductors:L1 = choke, 4.7 mH
Semiconductors:D1 = BAT85D2 = LED, red, high efficiencyT1, T2 = BC63
will be fine.If a battery is suspected of suffering
from the memory effect, discharge andrecharge it two or three times in suc-cession. This action will in almost allcases restore the capacity of the batterycompletely (commensurate with its life,of course).
(980050-I)
Elektor Electronics 6/98 49
sealed cell is
Standard 110 mΩ/cellHeavy duty 50 mΩ/cellSintered 19 mΩ/cell
E N V I R O N M E N T A LE F F E C T SOne of the most serious drawbacks ofNiCd batteries is their effect on theenvironment. This type of battery con-tains cadmium which is toxic. In mostcountries, discarded NiCd batteries aredumped on the rubbish heap wherethey remain toxic for a very long time.It is true, of course, that their life ofsome 500–800 charge/discharge cyclesdoes not cause millions of them to bedisposed of on the rubbish heap. Nev-ertheless, this was a very important fac-tor in the decision of manufacturers ingeneral to discontinue the use of NiCdbatteries in most consumer products.
M E M O R Y E F F E C TAnother disadvantage of sintered (notmass plate) NiCd batteries is, asalready mentioned, their memoryeffect. This manifests itself in the cellretaining the characteristics of previouscycling. That is, after repeated shallow-depth discharges the cell will fail toprovide a satisfactory full-depth dis-charge. Note, however, that Evereadycylindrical nickel-cadmium cells areparticularly noted for their lack ofmemory effect.
The memory effect is a nuisance,because it means that a battery with anominal capacity of, say, 600 mAh, aftera number of charge/discharge cycleshas a useful capacity of only 300 or 400mAh. This has nothing to do with thelife of the battery: even a new batteryif charged as stated will soon lose partof its capacity.
Fortunately, this reduction in capac-ity can be prevented fairly simply.Moreover, batteries that already sufferfrom the memory effect can berestored to their nominal capacity. Thecure is simply to ensure that a batteryis occasionally fully discharged beforeit is recharged. Occasionally meansbefore every third or so recharge. Notethat there are chargers on the marketthat have the discharge facility built in,but this will certainly not be the case inthe less expensive types.
C O R R E C TD I S C H A R G I N GThere is no need for extensive circuitryto discharge a battery: a simple resis-tor or light bulb will accomplish itreadily. It is, however, necessary tokeep an eye on the discharge time,because otherwise there is the risk thatthe battery is discharged beyond a cer-tain voltage. When this happens, itmay cause polarity reversal in the cells
comprising the battery.Correct discharging can only take
place via a circuit that arranges for thebattery to be discharged to a certainlevel and then disconnects it from thedischarge circuit.
The diagram of such a circuit – seeFigure 1 – is pretty straightforward.Nevertheless, such a discharge circuitdoes the job correctly. It causes a bat-tery to be discharged to a level of650 mV. This level ensures that the bat-tery is correctly discharged without therisk of polarity reversal. The battery isnot discharged at a constant current,but in short bursts. This allows the bat-tery to ‘recover’ during the intervals,which, in practice, has been found toextend its useful life.
During the discharge, an LED lightsto show that the process is continuing.Since the diode cannot work from avoltage of 0.65–1.2 V, the voltage has tobe raised. To this end, the astable mul-tivibrator formed by T1 and T2 oscil-lates at a rate of 25 kHz. When T2 is on,current flows through inductor L1, sothat energy is stored in the magneticfield. When T2 is off, the inductor is‘discharged’ via the LED, whereuponthis lights.
Diode D1 prevents the energystored in the inductor from leakingaway via the base of T1. This mighthappen because the capacitors in thecircuit have fairly high values, whereasthe resistors have low ones. The cho-sen values ensure that the dischargecurrent is sufficiently high. When thebattery voltage is 1.2 V, the dischargecurrent is some 200 mA; at 0.8 V, it hasdropped to about 100 mA, and at0.65 V to around 50 mA. When the bat-tery voltage drops to 0.65 V, the dis-charge process is discontinued.
C O N S T R U C T I O NThe tiny circuit is best built on theprinted-circuit board shown in Fig-ure 2, but this is not available readymade. However, a small prototypingboard will do very nicely as well.
Inductor L1 is a small choke whichshould be readily available from mostelectronics retailers.
The LED should be a high-effi-ciency type, while, because of thethreshold discharge voltage, D1 mustbe a Schottky type.
U S A G EThere is not much that can be saidabout using the discharge unit. It is sim-ply a matter of connecting the 1.2 V bat-tery with correct polarity, checking thatthe LED lights, and disconnecting thebattery when the LED goes (or is) out.
In general, the discharge period willnormally be three to four hours. Asmentioned before, the battery does notneed to be discharged fully before it isrecharged: before every third recharge
will be fine.If a battery is suspected of suffering
from the memory effect, discharge andrecharge it two or three times in suc-cession. This action will in almost allcases restore the capacity of the batterycompletely (commensurate with its life,of course).
(980050-I)
49Elektor Electronics 6/98
T1
R1
4Ω
7
R2
10
0Ω
R3
10
0Ω
R4
4Ω
7
C1
220n
C2
470n D1
BAT85
L1
4mH7
D2
T2
BC6392x
980050 - 11
BT
1
Figure 1. The circuit is basicallyan astable multivibrator oscillat-ing at a rate of 25 kHz.
980050-1
(C)
Seg
men
t
980050-1
(C) Segment
C1
C2
D1
D2
L1
R1
R2
R3
R4
T1
T2
+
-
980050-12
Figure 2. The discharge circuit isintended for 1.2 V batteries. Ifseveral of these, or a 9 Vrechargeable NiCd battery, haveto be discharged, an appropriatenumber of PCBs are needed.
Parts listResistors:R1, R4 = 4.7 ΩR2, R3 = 100 Ω
Capacitors:C1 = 0.22 µFC2 = 0.47 µF
Inductors:L1 = choke, 4.7 mH
Semiconductors:D1 = BAT85D2 = LED, red, high efficiencyT1, T2 = BC63
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
The U2514B is anintegrated bipolar
radio circuit suitablefor digital tuning sys-
tems. It contains anFM front end with pre-
amplifier and FMstereo decoder aswell as a complete
AM receiver anddemodulator. Stop -
signal generation isimplemented for FMand AM mode. The
circuit is designed foruse in small radios,
power packs and mul-timedia applications.
A Telefunken SemiconductorsApplication
AM/FMReceiver IC
for digital tuning
HPIOSCE AFNMFMOSC3 FMRF FMDET mpxour
2E1
FMIN 0-qp.2
FMAGC
OSOOIJT8
04
GNDRF CJ
FM front snd
M IN 0 L A M Fr urn end - 0
15
FM TF and
modnini or
A %I'M MI412-
gnal
25
AM IF andumn<nitkinior
7 1.0 12 I 13
0AMOSC FMTHIN
AM] RN
MPXIN
24
Stevendecoderand HCC
Irrly PLL
17 9 22 26.
0METER CFR A A MFM
AMSA DJ
Figure 1. Block schematic of theType U2514B AM/FM Receiver IC.
Features: FM wideband AGC
LO buffer for digital tuning
Integrated stop -signal generation for AM and FM
Adjustable stop -signal sensitivity
Automatic stereo -mono blend
High cut Mute function Pilot canceller Supply voltage range 3-12 V
VRFY
21OO
OUTR
atm
LPF
200 CER
23nb. 0 CTRL&
PIN DESCRIPTIONThe FM preamplifier input, FMIN (pin28) - see the block diagram in Fig-urel- consists of a transistorgrounded -base circuit that providesexcellent noise performance and largesignal behaviour. It is recommended toconnect a source impedance of 100 nto achieve optimal performance. Thedirect current through the amplifyingtransistor is reduced by the internalautomatic gain control (acc). Thismeans that in the case of large input
0 VS
C'ND1'6
980[5&.11
signals the input alternating current isbypassed via the wideband AGC tran-sistor. A capacitor (C13 in Fig-ure 2), isconnected between pins 2 (Fmacc) and4 (GNDRF) to smooth the AGC voltageand to shorten the transistor base toground (pin 4). A tuned r.f. (radio fre-quency) circuit is connected betweenpins 3 (FmRF) and 27 (vs). The amplifiedr.f. signal is fed internally to the mixerinput.
The FM local oscillator consists of atransistor in a grounded -collector con -
50 Elektor Electronics 6/98Elektor Electronics 6/98
P I N D E S C R I P T I O NThe FM preamplifier input, FMIN (pin28) – see the block diagram in Fig-ure 1 – consists of a transistorgrounded-base circuit that providesexcellent noise performance and largesignal behaviour. It is recommended toconnect a source impedance of 100 Ωto achieve optimal performance. Thedirect current through the amplifyingtransistor is reduced by the internalautomatic gain control (AGC). Thismeans that in the case of large input
signals the input alternating current isbypassed via the wideband AGC tran-sistor. A capacitor (C18 in Fig-ure 2), isconnected between pins 2 (FMAGC) and4 (GNDRF) to smooth the AGC voltageand to shorten the transistor base toground (pin 4). A tuned r.f. (radio fre-quency) circuit is connected betweenpins 3 (FMRF) and 27 (VS). The amplifiedr.f. signal is fed internally to the mixerinput.
The FM local oscillator consists of atransistor in a grounded-collector con-
The U2514B is anintegrated bipolar
radio circuit suitablefor digital tuning sys-
tems. It contains anFM front end with pre-
amplifier and FMstereo decoder aswell as a complete
AM receiver anddemodulator. Stop-signal generation is
implemented for FMand AM mode. The
circuit is designed foruse in small radios,
power packs and mul-timedia applications.
50
A Telefunken SemiconductorsApplication
AM/FMReceiver IC
Features:ä FM wideband AGC
ä LO buffer for digital tuning
ä Integrated stop-signal generation for AM and FM
ä Adjustable stop-signal sensitivity
ä Automatic stereo-mono blend
ä High cut
ä Mute function
ä Pilot canceller
ä Supply voltage range 3–12 V
1
Figure 1. Block schematic of theType U2514B AM/FM Receiver IC.
for digital tuning
AP
PLI
CA
TIO
NN
OT
EA
PP
LIC
AT
ION
NO
TE
The
cont
ent o
f thi
s no
te is
bas
ed o
n in
form
atio
n re
ceiv
ed fr
om m
anuf
actu
rers
in th
e el
ectri
cal a
nd e
lect
roni
cs in
dust
ries
or th
eir r
epre
sent
ativ
es a
nd d
oes
not
impl
y pr
actic
al e
xper
ienc
e by
Ele
ktor
Ele
ctro
nics
or i
ts c
onsu
ltant
s.
The
cont
ent o
f thi
s no
te is
bas
ed o
n in
form
atio
n re
ceiv
ed fr
om m
anuf
actu
rers
in th
e el
ectri
cal a
nd e
lect
roni
cs in
dust
ries
or th
eir r
epre
sent
ativ
es a
nd d
oes
not
impl
y pr
actic
al e
xper
ienc
e by
Ele
ktor
Ele
ctro
nics
or i
ts c
onsu
ltant
s.
F.lfitennu
0
Figure 2. Typicalapplication of theU2514B circuit inconjunction with aType U4285B phase -locked loop.
Parts listResistors:R1 = not usedR2, R17 = 100 kn.R3, R4 = not usedR5 = 330 ki-2R6, R7 = 5.1 k52R8 = 150 ki-2R9 = not usedR10 = 33 knR11 = 6.8 kc-2.R12 = not usedR13, R14, R61 = 56 k52R15, R16, R19 = 10 kn.
R18 = 47052R20 = not usedR21 = 1.8 ki-2R22 -R59 = not usedR60 = 22052
R62 = 5.6 1<-0.R63 = 22 52R64, R65 = 12 k -Q.RP1 = 20 kf2 preset potentiometerRP2 = not usedRP3 = 100 ki-2
Capacitors:C1 = not used
v,to Micmeal ti l Le r
A1.4
C2 = 0.015 I.LF
( lo.H
C3, C21, C23 = 10 pFC4 = not used
CIRL4 0t -n_ (Ark MEFER
0 0 0 0CH' C11 C1 C13 C1-1
0 0 0 0SW01 SW02 SWai SW04
C5, C18, C28, C29, C30, C65, C70, C71,C73 -0.1 µF
C6, C14 = 10µFC7 = 0.47µFC8, C16, C26, C69 = 100 pFC10 = 0.0068µFC11 = 0.068 tLFC12, C13, C67 = 0.01µFC17 = not usedC19, C20, C68 = 0.0022µFC22 - 22 pFC24 = 390 pFC25, C61, C64, C72 = 100µFC27 = 180 pFC31 = 15 pFC32 = 3.9 pFC33 -C60 = not usedC62, C63 = depends on crystalC66 = 270 pFCT1, CT2 = 15 pF trimmer
Semiconductors:CD1, CD5 = KV1591A-2CD2 = not usedCD3, CD4 = BB814
Integrated circuits:IC1 = U2514B
Inductors & filters:L1 = not usedL2 = Toko 7PH-Type A119ACS-19000ZL3, L4 = Toko 7KL Type 291ENS2054IBL5 = Toko 7P Type A7BRS-12938XL6 = Toko 7PH Type A119ACS-18999NL7 = Toko 7P Type 7MC-312162N0CFI = Murata CSB456F10CF2 = 10.7 MHz ceramic filter, BW
180 kHz or 150 kHzCF3 = 455 kHz ceramic filterAM capacitance diodes = Toko
KV1591A-2AM bar antenna:
LW1 = 220µ,H (59 turns, 10 x 0.04 flexwire)Rpw2 = 25 ki-2 (23 turns) if w1 circuit isin resonance at 1 MHz)Ferrite material: diameter 10 mm, length80 mm
Miscellaneous:X = crystal, 4 MHz
Elektor Electronics 6/98 5151Elektor Electronics 6/98
Parts listResistors:R1 = not usedR2, R17 = 100 kΩR3, R4 = not usedR5 = 330 kΩR6, R7 = 5.1 kΩR8 = 150 kΩR9 = not usedR10 = 33 kΩR11 = 6.8 kΩR12 = not usedR13, R14, R61 = 56 kΩR15, R16, R19 = 10 kΩR18 = 470 ΩR20 = not usedR21 = 1.8 kΩR22–R59 = not usedR60 = 220 ΩR62 = 5.6 kΩR63 = 22 ΩR64, R65 = 12 kΩRP1 = 20 kΩ preset potentiometerRP2 = not usedRP3 = 100 kΩ
Capacitors:C1 = not used
C2 = 0.015 µFC3, C21, C23 = 10 pFC4 = not usedC5, C18, C28, C29, C30, C65, C70, C71,
C73 =0.1 µFC6, C14 = 10 µFC7 = 0.47 µFC8, C16, C26, C69 = 100 pFC10 = 0.0068 µFC11 = 0.068 µFC12, C13, C67 = 0.01 µFC17 = not usedC19, C20, C68 = 0.0022 µFC22 = 22 pFC24 = 390 pFC25, C61, C64, C72 = 100 µFC27 = 180 pFC31 = 15 pFC32 = 3.9 pFC33–C60 = not usedC62, C63 = depends on crystalC66 = 270 pFCT1, CT2 = 15 pF trimmer
Semiconductors:CD1, CD5 = KV1591A-2CD2 = not usedCD3, CD4 = BB814
Integrated circuits:IC1 = U2514B
Inductors & filters:L1 = not usedL2 = Toko 7PH-Type A119ACS-19000ZL3, L4 = Toko 7KL Type 291ENS2054IBL5 = Toko 7P Type A7BRS-12938XL6 = Toko 7PH Type A119ACS-18999NL7 = Toko 7P Type 7MC-312162NOCF1 = Murata CSB456F10CF2 = 10.7 MHz ceramic filter, BW
180 kHz or 150 kHzCF3 = 455 kHz ceramic filterAM capacitance diodes = Toko
KV1591A-2AM bar antenna:
LW1 = 220 µH (59 turns, 10×0.04 flexwire)Rpw2 = 25 kΩ (23 turns) if w1 circuit isin resonance at 1 MHz)Ferrite material: diameter 10 mm, length80 mm
Miscellaneous:X = crystal, 4 MHz
Figure 2. Typicalapplication of theU2514B circuit inconjunction with aType U4285B phase-locked loop.
2
3
U25148AMFM CrtrLA
CIRLIEL
26 23 22
Aht0S1' FMOSL'
0 output (open drain)ucasum,0 I/O
SCA SDAI/O inpuI1output (open drain
Microcontrofler
figuration. The nega-tive resistance at thebase of the transistor isgenerated by an external capacitor (C21in Fig-ure 2) between pins 5 (FMoscE)and 4 (GNDRF). Another external capac-itor (C22 in Figure 2) is connectedbetween pins 6 (Fmoscs) and 5(FMoscE) to increase the resistance atthe emitter and to lead to a higheroscillator swing. The negative resis-tance at pin 6 is about 250 n. Conse-quently, the resonant Lo (local oscilla-tor) tank resistance of about 5 kn,which depends on the Q(uality) factorof the coil, is transformed to this mag-nitude via a capacitor.
A resistor (R8 in Fig. 2) is connectedbetween pins 8 (oscouT) and 14 (VREF)which determines the amplitude of theoscillator voltage that is fed to the PLL(phase -locked loop) circuit. The Tele-funken U428xBM family of PLL circuitsis recommended as members of thisfamily offer a high signal-to-noise ratioand low current drain.
The AM antenna coil is connectedbetween pins 1 (amix) and 14 (vREF). Toensure that the ACC operates correctly, acoil impedance of about 25 kn isrequired. The AM oscillator must beloaded by an external tank referred toVREF.
The IF output ((Four - pin 10) ofboth the am and the FM mixer has to beloaded into external IF tank circuitsreferred to VREF. The Q -factor of the IFcoils must be not lower than 50 n.
The ceramic resonator of the stereodecoder PLL circuit is used as a stop -sig-nal detector for AM signals. For this pur-pose, the parallel resonance frequencyof the resonator, which, unloaded, isabout 456 kHz, is reduced by an inter-nal load capacitor to 455 kHz. It isimportant that the AM IF is 455 kHz.The internal loading capacitor isdefined by the current through pin 9
igure 3. Block dia-gram showing theU2514B interfaced wia microcontroller.
R30058.13
(AMSADJ) to ground.An external resistor
(R19-RP3 in Fig. 2)between pins 9 (AMSADJ) and 14 VREF)permits the alignment of the stop -sig-nal centre frequency. The width of thestop window is typically 1.2 kHz. If AMsearch mode is not actuated, the pin ispulled to ground internally.
The input impedance of the AM IFamplifier is 3.3 1(52 in accordance withto the required impedance of mostceramic filter. The input refers to VREF(pin 14).
The input impedance of the FMIF amplifier is 330 n in line with therequired impedance of most FMceramic filters. The input refers to GND(pin 16).
A tank circuit (L2 -R11 -C16 in Fig-ure 2) is connected between pins 15(FMDET) and 14 (VREF). The discrimina-tor coil must be adjusted so that thevoltage at pin 11 (AFSM) is 1.2 V at10.7 MHz to ensure that the FM stopsignal is generated correctly.
A capacitor (C28 in Figure 2) is con-nected between pins 11 (AFSM) andGND for smoothing the FM audio out-put. As the deviation of the FM signal(max. 75 kHz) might be greater thanthe stop -signal window (42 kHz), FMaudio output smoothing is necessary togenerate a modulation -independentstop signal. In AM search mode, C28smoothes the FM demodulated AM IFsignal.
The METER pin (17) is driven by acurrent that corresponds to the level ofthe FM IF or AM IF signal. The requiredexternal load consists of parallel net-work R8 -C14 (see Figure 2). The voltageat this pin controls the high cut andmono/stereo blend function in thereception mode. These functions maybe modified by the designer by choos-ing different values of the load resistor.
The reception mode, search mode,
mute function, and search sensitivity,are selected by applying a control volt-age to pin 22 (=La). If this voltage ishigher than 800 mV, the receiver circuitis in the reception mode, otherwise inthe search mode. When the voltage isbetween 800 mV and 1.3 V, the audiofrequency (AF) level at pins 18 and 19(ouTR and OUTL respectively) isreduced (mute function). In the searchmode (0-700 mV), the voltage leveldetermines the degree of the searchsensitivity.
The signal at pin 23 (CTRLB) indicatesstereo or mono reception: stereo if thevoltage is higher than 1.2 V, otherwisemono. Furthermore, it is possible toforce the receiver circuit to mono byapplying an external voltage <800 mVto pin 23. If the search mode is selected,the low active stop signal appears atpin 23. If output CTRLB is applieddirectly to the I/O port of a microcon-troller, its high level must be adaptedby linking pin 23 to ground via a suit-able resistor.
An internal voltage regulator gen-erates a stable reference voltage of2.4 V, which is needed for all functionblocks of the ic. An external capacitor(C29 in Figure 2) must link pin 14 toground in order to achieve stabilityand noise suppression.
When pin 26 (AMFM) is linked toground, the receiver circuit is switchedto the AM mode; if the pin is open, theFM mode is selected. Switchingbetween these two states may be per-formed by a microcontroller withopen -drain I/O ports. A capacitor (C30in Figure 2) must be connectedbetween AMFM and ground for FMmode operation. It serves for smooth-ing the control voltage of the pilot can-celler.
A ceramic resonator of 456 kHz par-allel resonance frequency (at 30 pF on -chip load capacitance) is connectedbetween pin 20 (cERE5) and ground(necessary for the pilot regeneration).It should be mounted very close topin 20 to avoid spurious radiation. InAM <1.1 V) search(VLA< 800 mV) mode, the resonatoris used for stop -signal generation. Theparallel resonance frequency is thenreduced to 455 kHz by adjusting thecurrent into pin 9.
A phase -locked loop (PLL) circuit isused for the pilot regeneration of thestereo decoder. Therefore, a loop filterformed by network R5 -C10 -C11 is con-nected between pin 21 (LPF) andground.
The multiplexed signal is applied topin 24 (mPxix) and thence fed to thestereo decoder. The input resistance atthis pin is about 10 kn. It is recom-mended to align the channel separa-tion by an RC network between MPXINand MPXOUT (pin 25) owing to toler-ances in the group delay of the IF filter.
A52 Elektor 5/9852 Elektor 5/98
figuration. The nega-tive resistance at thebase of the transistor isgenerated by an external capacitor (C21in Fig-ure 2) between pins 5 (FMOSCE)and 4 (GNDRF). Another external capac-itor (C22 in Figure 2) is connectedbetween pins 6 (FMOSCB) and 5(FMOSCE) to increase the resistance atthe emitter and to lead to a higheroscillator swing. The negative resis-tance at pin 6 is about 250 Ω. Conse-quently, the resonant LO (local oscilla-tor) tank resistance of about 5 kΩ,which depends on the Q(uality) factorof the coil, is transformed to this mag-nitude via a capacitor.
A resistor (R8 in Fig. 2) is connectedbetween pins 8 (OSCOUT) and 14 (VREF)which determines the amplitude of theoscillator voltage that is fed to the PLL
(phase-locked loop) circuit. The Tele-funken U428xBM family of PLL circuitsis recommended as members of thisfamily offer a high signal-to-noise ratioand low current drain.
The AM antenna coil is connectedbetween pins 1 (AMIN) and 14 (VREF). Toensure that the AGC operates correctly, acoil impedance of about 25 kΩ isrequired. The AM oscillator must beloaded by an external tank referred toVREF.
The IF output (IFOUT – pin 10) ofboth the AM and the FM mixer has to beloaded into external IF tank circuitsreferred to VREF. The Q-factor of the IFcoils must be not lower than 50 Ω.
The ceramic resonator of the stereodecoder PLL circuit is used as a stop-sig-nal detector for AM signals. For this pur-pose, the parallel resonance frequencyof the resonator, which, unloaded, isabout 456 kHz, is reduced by an inter-nal load capacitor to 455 kHz. It isimportant that the AM IF is 455 kHz.The internal loading capacitor isdefined by the current through pin 9
(AMSADJ) to ground.An external resistor
(R19-RP3 in Fig. 2)between pins 9 (AMSADJ) and 14 VREF)permits the alignment of the stop-sig-nal centre frequency. The width of thestop window is typically 1.2 kHz. If AM
search mode is not actuated, the pin ispulled to ground internally.
The input impedance of the AM IF
amplifier is 3.3 kΩ in accordance withto the required impedance of mostceramic filter. The input refers to VREF
(pin 14).The input impedance of the FM
IF amplifier is 330 Ω in line with therequired impedance of most FM
ceramic filters. The input refers to GND
(pin 16).A tank circuit (L2-R11-C16 in Fig-
ure 2) is connected between pins 15(FMDET) and 14 (VREF). The discrimina-tor coil must be adjusted so that thevoltage at pin 11 (AFSM) is 1.2 V at10.7 MHz to ensure that the FM stopsignal is generated correctly.
A capacitor (C28 in Figure 2) is con-nected between pins 11 (AFSM) andGND for smoothing the FM audio out-put. As the deviation of the FM signal(max. 75 kHz) might be greater thanthe stop-signal window (42 kHz), FMaudio output smoothing is necessary togenerate a modulation-independentstop signal. In AM search mode, C28smoothes the FM demodulated AM IF
signal.The METER pin (17) is driven by a
current that corresponds to the level ofthe FM IF or AM IF signal. The requiredexternal load consists of parallel net-work R8-C14 (see Figure 2). The voltageat this pin controls the high cut andmono/stereo blend function in thereception mode. These functions maybe modified by the designer by choos-ing different values of the load resistor.
The reception mode, search mode,
mute function, and search sensitivity,are selected by applying a control volt-age to pin 22 (CTRLA). If this voltage ishigher than 800 mV, the receiver circuitis in the reception mode, otherwise inthe search mode. When the voltage isbetween 800 mV and 1.3 V, the audiofrequency (AF) level at pins 18 and 19(OUTR and OUTL respectively) isreduced (mute function). In the searchmode (0–700 mV), the voltage leveldetermines the degree of the searchsensitivity.
The signal at pin 23 (CTRLB) indicatesstereo or mono reception: stereo if thevoltage is higher than 1.2 V, otherwisemono. Furthermore, it is possible toforce the receiver circuit to mono byapplying an external voltage <800 mVto pin 23. If the search mode is selected,the low active stop signal appears atpin 23. If output CTRLB is applieddirectly to the I/O port of a microcon-troller, its high level must be adaptedby linking pin 23 to ground via a suit-able resistor.
An internal voltage regulator gen-erates a stable reference voltage of2.4 V, which is needed for all functionblocks of the IC. An external capacitor(C29 in Figure 2) must link pin 14 toground in order to achieve stabilityand noise suppression.
When pin 26 (AMFM) is linked toground, the receiver circuit is switchedto the AM mode; if the pin is open, theFM mode is selected. Switchingbetween these two states may be per-formed by a microcontroller withopen-drain I/O ports. A capacitor (C30in Figure 2) must be connectedbetween AMFM and ground for FM
mode operation. It serves for smooth-ing the control voltage of the pilot can-celler.
A ceramic resonator of 456 kHz par-allel resonance frequency (at 30 pF on-chip load capacitance) is connectedbetween pin 20 (CERES) and ground(necessary for the pilot regeneration).It should be mounted very close topin 20 to avoid spurious radiation. InAM (VAMFM <1.1 V) search(VCTRLA<800 mV) mode, the resonatoris used for stop-signal generation. Theparallel resonance frequency is thenreduced to 455 kHz by adjusting thecurrent into pin 9.
A phase-locked loop (PLL) circuit isused for the pilot regeneration of thestereo decoder. Therefore, a loop filterformed by network R5-C10-C11 is con-nected between pin 21 (LPF) andground.
The multiplexed signal is applied topin 24 (MPXIN) and thence fed to thestereo decoder. The input resistance atthis pin is about 10 kΩ. It is recom-mended to align the channel separa-tion by an RC network between MPXIN
and MPXOUT (pin 25) owing to toler-ances in the group delay of the IF filter.
3
Figure 3. Block dia-gram showing theU2514B interfaced witha microcontroller.
To drive both the compensationnetwork to pin 24 and an optional RDS
(radio data service), the output imped-ance at pin 25 (MPXOUT) is low. Thedirect voltage is 1.2 V in FM mode(depending on the discriminator coilalignment) and 0.8–1.2 V in the AM
mode (depending on the signal level).The open-collector output at pins 18
and 19 (OUTR and OUTL respectively)requires and external resistor toground (R7 and R6 in Figure 2 respec-tively). The deemphasis may beachieved by an additional parallelcapacitor (C13 and C12 respectively).
F U N C T I O N A LD E S C R I P T I O NIn the FM mode, the antenna signal isfed via a tuned r.f. circuit to the inte-grated pre-stage, which consists of atransistor grounded-base circuit. Toprotect the pre-stage against overload,an automatic gain control (AGC) isincluded on the chip.
A tuned r.f. circuit on the collector isnecessary for amplifying and filteringthe FM signal, which is fed internally tothe mixer. It consists of a double-bal-anced Gilbert Cell.
The local oscillator (LO) signal isgenerated by an integrated oscillator.The buffered LO signal is used to drivea PLL. THE IF signal (10.7 MHz) is cou-
pled out at the mixer’s output and fedvia a ceramic filter to the demodulator.The demodulated audio signal is avail-able at pin 25 (MPXOUT)
In the AM mode, the signal is feddirectly to the mixer. The antennaimpedance must be higher than 25 kΩto ensure correct operation of the levelcontrol in case of large signals. The LO
signal is generated by an integratedoscillator. The buffered LO signal isused to drive the PLL. If the AM searchmode is required, the IF must be455 kHz. The IF output is fed via aceramic filter to the demodulator. Thedemodulated audio signal is availableat pin 25 (MPXOUT).
When a control voltage is applied topin 22 (CTRLA), the mode of the receivermay be selected.
The search mode is selected whenthe control voltage is <800 mV.
The search sensitivity may beselected by varying the control voltagein the range 100–800 mV. When thecontrol voltage is 100 mV, the highestsensitivity is achieved.
In the reception mode, muting ispossible by varying the control voltagein the range 0.8–1.4 V. When the con-trol voltage is 0.8 V, the highest mutedepth is achieved.
The output at pin 23 (CTRLB) indi-cates whether the receiver is operating
in the stereo or mono mode. When thecontrol voltage applied to pin 23 <1.1V, the receiver is forced to mono.
In the search mode (VCTRLA<0.8 V),the internally generated stop signal isavailable at pin 23 as a low active sig-nal.
If both conditions
for AM: VMETER>VCTRLA
for FM: VMETER (90/R8) >VCTRLA
and
1.1 V<VAFSM<1.3 V
for AM: current adjust into pin 9 for fpin 20 = 455 kHz
for FM: detector coil adjust to Vpin 11 = 1.2 V for 10.7 MHz
are fulfilled, a stop signal is generated.[980058]
53Elektor Electronics 6/98
Pin Symbol Function AM FM
1 AMIN AM antenna input VRef High impedance
2 FMAGC FM AGC time constant VRef VRef – 80 mV
3 FMRF FM RF tank High impedance VS / 0 to 1 mA
4 GNDRF Ground RF GND GND
5 FMOSCE FM oscillator emitter VRef – 2×VBE / 0 A 0.95 V
6 FMOSCB FM oscillator base VRef – VBE / 0 A 1.7 V
7 AMOSC AM oscillator VRef /0.3 mA High impedance
8 OSCOUT Buffered AM/FM oscillator output VRef /0.3 mA VRef /0.7 mA
9 AMSADJ Current input for AM stop-signal adjustment AMsearch=VBE, AM=0.1 V 0.1 V
10 IFOUT AM/FM IF output VS /in 50 µA VS /in 0.4 mA
11 AFSM AF smoothing 0.8 to 1.2 V 1.2 V
12 FMIFIN FM IF amplifier input VBE to GND/ 0 A VBE to GND
13 AMIFIN AM IF amplifier input 3.3 kΩ to VRef 3.3 kΩ to VRef
14 VREF Reference voltage input VRef = 2.4 V VRef = 2.4 V
15 FMDET FM discriminator VRef /1 µA VRef /0 A
16 GND Ground GND GND
17 METER Field-strength output 0 to 2,3 V 0 to 2 V
18 OUTR AF output right 0 to 2.3 V/0.15 mA 0 to 2.3 V/0.15 mA
19 OUTL AF output left 0 to 2.3 V/0.15 mA 0 to 2.3 V/0.15 mA
20 CERES Ceramic resonator 456 kHz for AM search and for pilot-PLL in FM mode 0,1 to 2.3 V 0.1 to 2.3 V
21 LPFLow-pass filter for pilot-PLL AMLow-pass filter for pilot-PLL AM search and FM
0.2 V0.5 to 2 V
-0.5 to 2 V
22 CTRLA Control input for mute, search mode and search sensitivity 0 to VRef 0 to VRef
23 CTRLBControl input for forced mono, Control output for stop function,Mono/stereo information
0.1 V to VS ,30 kΩ 0.1 V to VS, 30 kΩ
24 MPXIN Stereo decoder MPX input 0.8 V 0.8 V
25 MPXOUT AM/FM MPX output 0.8 to 1,2 V 1.2 V
26 AMFM AM/FM switch and pilot canceller time constant GND 1.54 V
27 VS Supply-voltage input 3 to 12 V/in 5 mA 3 to 12 V/in 9 mA
28 FMIN FM antenna input VRef – VBE /0 A 1.5 V
The conductancetester goes one step
further than the usualcontinuity tester found
in so many work-shops. It contains abuzzer to indicate avery low resistancebetween two points
along a conductor orcircuit and an LED
display to show theorder of resistance
between these pointswhen the buzzer does
not sound.
Design by L. Koch
conductancetester
with buzzer or LEDindication
Arguably, one of the most useful gad-gets in a small workshop is a continu-ity tester which enables the rapidchecking of whether a conductor or cir-cuit is open -circuited or short-circuited.It normally uses a buzzer to indicate ashort-circuit (that is, a continuity in theconductor or circuit). Such a device isimproved considerably if it is given ameans of showing the order of resis-tance between the two probes whenthe buzzer remains quiet. This quicklysolves the question of 'is it a bad con -tad?' or 'is it a break in a cable?', and soon. The simple indication provided inthe present tester shows at a glance therelative magnitude of the resistancebetween the two probes.
INDICATORFor a rough indication of a measuredvalue of resistance (or its reciprocal,conductance) a liquid -crystal display(LCD) or other fairly expensive indica-tor is not necessary, and in the presenttester the indicator is formed by anumber of light -emitting diodes
(LEDs). These are driven by the well-known display driver IC Type LM3915.This circuit is designed specifically todisplay the value of an analogue volt-age via a row of LEDs.
The LM3915 contains a referencevoltage source and an accurate decadicpotential divider. The voltages at thetaps of this divider are applied to aseries of comparators. These compara-tors are driven sequentially in line witha rising input to the IC. The compara-tor outputs can drive an LED directly.The row of LEDs may be set to the dotor bar mode. The brightness of thediodes can also be adjusted in accor-dance with individual needs.
One of the attractions of theLM3915 is that it requires relatively fewexternal components (other than theLEDs). The high -impedance input cir-cuit of the IC accepts signals at levelsfrom 0 V to 1.5 V below the supplyvoltage. Provided that the input signalsdo not exceed ±35 V, there is no needfor an external protection circuit. Theinput voltage is indicated in 3 dB steps.
454Elektor Electronics 6/98Elektor Electronics 6/98
Arguably, one of the most useful gad-gets in a small workshop is a continu-ity tester which enables the rapidchecking of whether a conductor or cir-cuit is open-circuited or short-circuited.It normally uses a buzzer to indicate ashort-circuit (that is, a continuity in theconductor or circuit). Such a device isimproved considerably if it is given ameans of showing the order of resis-tance between the two probes whenthe buzzer remains quiet. This quicklysolves the question of ‘is it a bad con-tact?’ or ‘is it a break in a cable?’, and soon. The simple indication provided inthe present tester shows at a glance therelative magnitude of the resistancebetween the two probes.
I N D I C A T O RFor a rough indication of a measuredvalue of resistance (or its reciprocal,conductance) a liquid-crystal display(LCD) or other fairly expensive indica-tor is not necessary, and in the presenttester the indicator is formed by anumber of light-emitting diodes
(LEDs). These are driven by the well-known display driver IC Type LM3915.This circuit is designed specifically todisplay the value of an analogue volt-age via a row of LEDs.
The LM3915 contains a referencevoltage source and an accurate decadicpotential divider. The voltages at thetaps of this divider are applied to aseries of comparators. These compara-tors are driven sequentially in line witha rising input to the IC. The compara-tor outputs can drive an LED directly.The row of LEDs may be set to the dotor bar mode. The brightness of thediodes can also be adjusted in accor-dance with individual needs.
One of the attractions of theLM3915 is that it requires relatively fewexternal components (other than theLEDs). The high-impedance input cir-cuit of the IC accepts signals at levelsfrom 0 V to 1.5 V below the supplyvoltage. Provided that the input signalsdo not exceed ±35 V, there is no needfor an external protection circuit. Theinput voltage is indicated in 3 dB steps.
The conductancetester goes one step
further than the usualcontinuity tester found
in so many work-shops. It contains abuzzer to indicate avery low resistancebetween two points
along a conductor orcircuit and an LED
display to show theorder of resistance
between these pointswhen the buzzer does
not sound.
54
Design by L. Koch
conductancetester
with buzzer or LEDindication
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
RESISTANCEMEASUREMENTSince the LM3915 is designed for indi-cating voltages, and the present circuitis intended for measuring conductance(or its reciprocal, resistance), a meanshas to be devised to convert voltageinto ohms or siemens. In the presentcircuit, this is done by placing an addi-tional potential divider at the input ofthe LM3915 and ensuring that the divi-sion ratio is influenced by the magni-tude of the resistance between the testprobes.
In the circuit diagram of the testerin Figure 1, the external potentialdivider at the input of ICI is formed byD1, R2 and R3. The resistance, if any,between the test probes is connected inparallel with R3 via resistor R1 and pre-set P1. This means that this resistanceaffects the division ratio of the divider,and thus the signal applied to pin 5 of'Cl.
The design arranges for resistancevalues of 10 n to 7.5 1(52 to be indicatedin seven 3 dB steps by the sequentiallighting of D2-D9. The first diode, D1,lights when the resistance is lower than10 52; this level may be preset to 0 n(that is, full conduction) with P1.
Transistors T1 and T2 in parallelwith D1 ensure that in case of very lowresistance between the test probes onlyD1 lights and that the buzzer, Bz1, isenergized.
The diagram in Figure 1 shows IC1configured for the dot mode, whichkeeps the current drain low. This is
D11
.L1
:100n
3
MODE ° L10
L9
RHI IC1 L8
L7
LM3915 Ls
L5
L4
L3
L2
L1
SIG
REFOUT
RE FADJ
RLO
effected by leaving pin 9 (MODE) of theIC open. If the bar mode is wanted,pin 9 must be linked to pin 3.
CONSTRUCTIONThe tester must, of course, be as com-pact as possible so that it can be car-ried about in one's pocket.Consequently, the printed -circuitboard for it (see Figure 2) is small.Building the tester on this board issimplicity itself, as is wiring it up.There are only three connections tobe made: two for the probes
Parts listResistors:Ri = 75 52R2 = 2.2 kf2R3 = 17.8 k52R4 = 1.2 k52R5, R6 = 10 k52R7 = 100 52P1 = 47 k52 preset potentiometer
Capacitors:C1 = 0.1 µF
Semiconductors:D1-D9 = LED, 3 mmD10 = zener diode 2.7 V, 400 mWD11 = 1N4148T1, T2 = BC557C
Integrated circuits:Bz1 = buzzer, 12 VBt1 = 9 V battery with connecting clip2 off test probes
Figure 2. The printed -cir-cuit board for the conduc-tance tester is not avail-able ready made.
D9
D8
10
11
12
3
14
15
D72x
- BC557CD6
-111.
D5
6
17
18
D414 -N.
D314 6-6.D2
(arrows),two for thebuzzer(Bz1), andtwo for thepower supply.
The tester is best powered by a 9 Vbattery. Since the current drain of thecircuit is at most 30 mA (with buzzersounding), an alkaline -manganesebattery should last about a year innormal use.
The tester, complete with battery, isbest housed in a small plastic case.
[980045]
R5 R6 R7
BZ1+0
980045 - 11
Figure 1. The circuit dia-gram of the conductancetester is an example ofsimplicity.
Figure 3. Photograph ofthe completed prototypetester.
Elektor Electronics 6/98 55
R E S I S T A N C EM E A S U R E M E N TSince the LM3915 is designed for indi-cating voltages, and the present circuitis intended for measuring conductance(or its reciprocal, resistance), a meanshas to be devised to convert voltageinto ohms or siemens. In the presentcircuit, this is done by placing an addi-tional potential divider at the input ofthe LM3915 and ensuring that the divi-sion ratio is influenced by the magni-tude of the resistance between the testprobes.
In the circuit diagram of the testerin Figure 1, the external potentialdivider at the input of IC1 is formed byD1, R2 and R3. The resistance, if any,between the test probes is connected inparallel with R3 via resistor R1 and pre-set P1. This means that this resistanceaffects the division ratio of the divider,and thus the signal applied to pin 5 ofIC1.
The design arranges for resistancevalues of 10 Ω to 7.5 kΩ to be indicatedin seven 3 dB steps by the sequentiallighting of D2–D9. The first diode, D1,lights when the resistance is lower than10 Ω; this level may be preset to 0 Ω(that is, full conduction) with P1.
Transistors T1 and T2 in parallelwith D1 ensure that in case of very lowresistance between the test probes onlyD1 lights and that the buzzer, Bz1, isenergized.
The diagram in Figure 1 shows IC1configured for the dot mode, whichkeeps the current drain low. This is
effected by leaving pin 9 (MODE) of theIC open. If the bar mode is wanted,pin 9 must be linked to pin 3.
C O N S T R U C T I O NThe tester must, of course, be as com-pact as possible so that it can be car-ried about in one’s pocket.Consequently, the printed-circuitboard for it (see Figure 2) is small.Building the tester on this board issimplicity itself, as is wiring it up.There are only three connections tobe made: two for the probes
( a r r o w s ) ,two for theb u z z e r(Bz1), andtwo for thepower supply.
The tester is best powered by a 9 Vbattery. Since the current drain of thecircuit is at most 30 mA (with buzzersounding), an alkaline-manganesebattery should last about a year innormal use.
The tester, complete with battery, isbest housed in a small plastic case.
[980045]
Elektor Electronics 6/98
D5
D4
D7
D6
D9
D8
D3
D2
D1
T1BC557C
T2
R5
10
k
R6
10
k
R7
10
0Ω
R4
1k
2
R2
2k
2
R3
17
k8
D10
2V7400mW
REFOUT
REFADJ
LM3915
IC1
MODE
SIG
RHI
RLO
L10
17
16
15
14
13
12
11
10
L9
L8
L7
L6
L5
L4
L3
L1
18L2
9
5
8
4
6
7
3
2
1
50Ω
P1
R1
75
Ω
BZ1
D11
1N4148C1
100n
2x
BT1
9V
980045 - 11
1
Figure 1. The circuit dia-gram of the conductancetester is an example ofsimplicity.
(C) Segment
980045-1
BZ1
C1
D1
D2
D3
D4
D5
D6
D7
D8
D9
D10
D11
H2H3
H4
IC1
P1
R1
R2
R3
R4
R5
R6
R7
T1T2
- 980045-1
+
+
T
(C) Segment
980045-1
2Parts listResistors:R1 = 75 ΩR2 = 2.2 kΩR3 = 17.8 kΩR4 = 1.2 kΩR5, R6 = 10 kΩR7 = 100 ΩP1 = 47 kΩ preset potentiometer
Capacitors:C1 = 0.1 µF
Semiconductors:D1–D9 = LED, 3 mmD10 = zener diode 2.7 V, 400 mWD11 = 1N4148T1, T2 = BC557C
Integrated circuits:Bz1 = buzzer, 12 VBt1 = 9 V battery with connecting clip2 off test probes
Figure 2. The printed-cir-cuit board for the conduc-tance tester is not avail-able ready made.
Figure 3. Photograph ofthe completed prototypetester.
55
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
Field ProgrammableAnalogue Array
MPAA020 from Motorola
The MPAA series ofField ProgrammableAnalog Arrays from
Motorola is a newfamily of products that
have programmableanalogue buildingblocks that can be
configured to createcircuit functions that
solve real -world signalprocessing problems.When used with sup-
porting CAD tools andmacro library func-
tions, these productsallow the user to
address analogue cir-cuit design problems
with low risk and mini-mum analogue exper-
tise.
By M Kupfner
Table 1. Performance specifications of the MPAA020
Specification Typical Value
System Master Clock Frequency (clock) Internal Sampling Clock Rate TBD 1 MHz (max.)
Maximum Signal Frequency Recommende
Maximum Signal Frequency Nyquist
200 kHz
500 kHz
Input Signal Range 0.5 V to (Vdd - 0.5V)
Analog Output Drive100 pF (max.)
1 IQ (min.)
DC Offset < 10 m V
Harmonic Distortion 1 kHz
Harmonic Distortion 200 kHz< 0.1 %< 0.5 %
Differential Non -Linearity < 0.15 LSB
Integral Non -Linearity < 0.24 LSB
Slew Rate 10 V/ms
Signal to Noise Ratio (SNR) > 60 dB
Power Supply Rejection Ration (PSRR) TBD
Power Dissipation (max.)
Each cell individually selectable200 mW(10 m W/cell)
Operating Temperature Range -40 to + 85 °C
Topology: (1) study of geometrical properties and spatial relations unaffected by contin-uous deformation, such as twisting or stretching. Mathematical approaches employingtopology are of great importance in modern theories of the four fundamental interac-tions (gravitational, electromagnetic, weak and strong); (2) in electronics: generic circuitstructure or collection of working structures.
A56 Elektor Electronics 6/9856 Elektor Electronics 6/98
Topology: (1) study of geometrical properties and spatial relations unaffected by contin-uous deformation, such as twisting or stretching. Mathematical approaches employingtopology are of great importance in modern theories of the four fundamental interac-tions (gravitational, electromagnetic, weak and strong); (2) in electronics: generic circuitstructure or collection of working structures.
The MPAA series ofField ProgrammableAnalog Arrays from
Motorola is a newfamily of products that
have programmableanalogue buildingblocks that can be
configured to createcircuit functions that
solve real-world signalprocessing problems.When used with sup-
porting CAD tools andmacro library func-
tions, these productsallow the user to
address analogue cir-cuit design problems
with low risk and mini-mum analogue exper-
tise.
By M Kupfner
Field ProgrammableAnalogue Array
MPAA020 from Motorola
Specification Typical ValueSystem Master Clock Frequency (clock) Internal Sampling Clock Rate TBD 1 MHz (max.)
Maximum Signal Frequency RecommendeMaximum Signal Frequency Nyquist
200 kHz500 kHz
Input Signal Range 0.5 V to (Vdd – 0.5V)
Analog Output Drive100 pF (max.)1 kΩ (min.)
DC Offset < 10 mV
Harmonic Distortion 1 kHzHarmonic Distortion 200 kHz
< 0.1 %< 0.5 %
Differential Non-Linearity < 0.15 LSB
Integral Non-Linearity < 0.24 LSB
Slew Rate 10 V/ms
Signal to Noise Ratio (SNR) > 60 dB
Power Supply Rejection Ration (PSRR) TBD
Power Dissipation (max.)Each cell individually selectable
200 mW(10 mW/cell)
Operating Temperature Range -40 to +85 °C
Table 1. Performance specifications of the MPAA020
The first available Field ProgrammableAnalog Array, the MPAA020, is an ‘elec-tronic breadboard’ that provides anideal medium for quickly designing,debugging, and implementing a widearray of analogue circuits, therebyreducing development cycle times.
Additionally, the analogue arrays,when used with Motorola’s MPA seriesof field programmable gate arrays(FPGA), allow the user to implementfield programmable mixed-signaldesigns, using the products and soft-ware from each of the individual FPGAtechnologies. This mixed-signal designcapability extends the user’s flexibilityeven further by providing the capabil-ity to simultaneously design, debug,and implement, both the analogue anddigital aspects of a system topology.
A P P L I C A T I O N SMotorola’s field programmable arrayscan be used in a wide variety of appli-cations. The first analogue array prod-
ucts are aimed at industrial motion,process, and power control applica-tions. Additional applications to whichthe analogue arrays are suited includecommunications (low to medium fre-quency applications), process control(temperature, heating and cooling sys-tems, pressure control, and so on),automotive, and medical instrumenta-tion, and measurement systems.
A N A L O G U ET E C H N O L O G YThe technology for Motorola’s fieldprogrammable analogue array is basedon switched capacitor (SC) technology.The MPAA020’s switched capacitor cir-cuit topology (see box) is designed tobe insensitive to parasitic capacitances;consequently, arbitrary signal routeingis possible with minimal loss in signalintegrity caused by these parasitics..Also, since the capacitors are integratedon silicon, the capacitance ratios aretightly matched, allowing precise ana-
logue signal processing without theneed for calibration or feedback. Ana-logue resources in the MPAA020 arecontained in configurable analogueblocks (CABs) that incorporate aswitched-capacitor CMOS op amp,comparator, capacitor arrays, CMOSswitches, and SRAM (static RAM). Datastored in SRAM control the switchesthat program various capacitance val-ues (static and dynamic) in the inputand feedback signal paths of theop amp. Analogue functions such asprogrammable gain stages, adders,subtractors, rectifiers, sample&hold cir-cuits, and first-order filters can beimplemented in a single cell (CAB).Higher level functions, such as biquadfilters, PLLs (phase-locked loops), leveldetectors, and so on, can be imple-mented using two or more cells.
S C T E C H N O L O G Y I NT H E M P A A 0 2 0Switched-capacitor cells are invariablyimplemented in quantity in an inte-grated circuit. The layout of a singlecell is shown in Figure 1 (which, by theway, is a macro contained in theEasyAnalog™ software).
The clock (switching) phases of thecell are designated Φ 1 and Φ 2, and thecapacitors C1 and C2. The amplifica-tion, α, of the cell is, as in a standardconfiguration, α = C1/C2.
In accordance with Shannon’s the-orem, the clock frequency, fclk, must beat least twice the highest frequencyoccurring in the signal, fsig. If this con-dition is not met, aliasing effects willoccur. In the MPAA020, the clock fre-quency is 1 MHz max, which meansthat the signal frequency must notexceed 500 kHz – in practice, 200 kHz.
A R C H I T E C T U R EThe MPAA020 contains 41 op amps,100 programmable capacitors, and 6864switches. The switches control circuitconnectivity, capacitor values, andother selectable features. The array is
57Elektor Electronics 6/98
switched-capacitor basicsIn switched-capacitor (SC) technology, a resistance isreplaced by a switched capacitance. The current througha resistance, and thus through a switched capacitor, isdirectly proportional to the voltage, U, applied across theresistance.
If, in the diagram, the switch is as shown, the capacitor,Cs, is charged to a charge Q = UCs. When the switch is inthe other position, the capacitor is discharged. Theprocess of the capacitor being charged and dischargedcontinuously results in an average current I = UCs/Ts,where Ts is one charge/discharge cycle, that is, 1/fs (fs is theswitching frequency). Substituting fs for 1/Ts givesI = UCs fs. In analogy to Ohm’s law, I = U/R, it follows thatR = 1/Cs fs. In other words, the ‘resistance’ is inversely pro-portional to the switching frequency and the capacitance
of the capacitor.Because of its easy programmability, switched-capaci-
tor devices find application in adaptive filtering, anti-alias-ing, phase-locked loops, and digital signal processing(DSP).
R
CS
980059 - 15
C1
C2
Uin
Uout
980059 - 11
φ1
φ2
φ1
φ1
φ2
Figure 1. Layout of asingle switched-capaci-tor cell.
1
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
SINik Pierwm
Eli
E E
Colon Nnibirri
I=1 El El I=1
structured in a grid that contains 20CABs arranged in a 4 x 5 matrix (seeFigure 2).
The programmable CABs rely onthe configuration logic in the upperportion of the chip to control the con-nectivity within the array and func-tionality in each CAB. Two buses movedata from the shift register to the CABs;the data control bus retrieves andmoves the data from the shift registerand the transfer control bus latches thedata into local SRAM.
Custom functions can be added tothe chip to meet customer specificapplications. An 8 -bit programmablebandgap voltage reference is availableto each CAB.
Op amps are provided on the chipperiphery that can be configured forunity gain buffering or filtering-for
lIchn
41rdiaa - I 17
instance, anti-aliasing or smoothing fil-ters (external resistors and capacitorsare then required).
Configuring an analogue designwithin the array is performed bydownloading 6 kbits of data via RS232communications from a PC orEPROM. The data stream containsinformation to configure the individ-ual cell, the cell interconnections, inter-nal bandgap voltage reference, andU0s. During this configuration down-load process, all cells are placed in apower down mode.
Figure 2. Block dia-gram of the TypeMPAA020 field pro-grammable analogarray.
SINGLE CABThe diagram of a single configurableanalogue block is shown in Figure 3.Each capacitor array consists of 255 sta-tically switched capacitors, whichmeans that the array can be pro-grammed in 255 stepped values.
The static switches, which can beset only once during the program-ming, are also used to control therouteing resources.
The 'resistance' (switched -capacitoreffect) is obtained by dynamicswitches, which also affect the phase ofthe input signal.
Additional switches enable thenumber and kind of limited local con-nections between two adjacent CABs.There are, however, also global lines toenable non -adjacent CABs to be inter-connected.
2 0 -CELL VERSIONThe MPAA020 is a field programmableanalogue array based on a general pur-pose analogue cell that may be config-ured, either alone or in combinations,as any of a wide range of analoguefunctions from simple comparators tocomplex filters. These cells arearranged in a 4x5 array with support-ing circuitry to provide input/output
Figure 3. Diagram of one of the Config-urable Analogue Blocks on which theMPAA020 is based.
A58 Elektor Electronics 6/98
structured in a grid that contains 20CABs arranged in a 4 × 5 matrix (seeFigure 2).
The programmable CABs rely onthe configuration logic in the upperportion of the chip to control the con-nectivity within the array and func-tionality in each CAB. Two buses movedata from the shift register to the CABs;the data control bus retrieves andmoves the data from the shift registerand the transfer control bus latches thedata into local SRAM.
Custom functions can be added tothe chip to meet customer specificapplications. An 8-bit programmablebandgap voltage reference is availableto each CAB.
Op amps are provided on the chipperiphery that can be configured forunity gain buffering or filtering—for
instance, anti-aliasing or smoothing fil-ters (external resistors and capacitorsare then required).
Configuring an analogue designwithin the array is performed bydownloading 6 kbits of data via RS232communications from a PC orEPROM. The data stream containsinformation to configure the individ-ual cell, the cell interconnections, inter-nal bandgap voltage reference, andI/Os. During this configuration down-load process, all cells are placed in apower down mode.
S I N G L E C A BThe diagram of a single configurableanalogue block is shown in Figure 3.Each capacitor array consists of 255 sta-tically switched capacitors, whichmeans that the array can be pro-grammed in 255 stepped values.
The static switches, which can beset only once during the program-ming, are also used to control therouteing resources.
The ‘resistance’ (switched-capacitoreffect) is obtained by dynamicswitches, which also affect the phase ofthe input signal.
Additional switches enable thenumber and kind of limited local con-nections between two adjacent CABs.There are, however, also global lines toenable non-adjacent CABs to be inter-connected.
2 0 - C E L L V E R S I O NThe MPAA020 is a field programmableanalogue array based on a general pur-pose analogue cell that may be config-ured, either alone or in combinations,as any of a wide range of analoguefunctions from simple comparators tocomplex filters. These cells arearranged in a 4×5 array with support-ing circuitry to provide input/output
58 Elektor Electronics 6/98
2Figure 2. Block dia-gram of the TypeMPAA020 field pro-grammable analogarray.
Figure 3. Diagram of one of the Config-urable Analogue Blocks on which theMPAA020 is based.
3
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor uk
signal buffering, programmable refer-ence voltage, cell to cell interconnec-tions, and so on.
Each cell's function may be pro-grammed to connect with any of theother cells in the array. Unused cellsare powered down individually tominimize power dissipation.
Digital interface circuitry is pro-vided to write the analogue circuit con-figurations data to on -chip SRAM inthe same manner as Motorola's digitalFPGAs (serial and parallel PROM, andmicroprocessor mode).
Analogue circuit design is simplifiedwith EasyAnalogTM design software,which handles bit level circuit configu-ration details, allowing the user to doanalogue design using functionalmacros in an easy to use point -and -click graphical environment.
E AS YANAL 0 G TM
PROGRAMThe EasyAnalogTM program may becompared with a circuit editor. Afterthe desired function has been selected,it is placed on one or more free CABs,and the input and outputs connectedas needed. Freeware EasyAnalogTMruns under Windows 95 or NT andmay be downloaded from the Internethttp://sps.motorola.com/fpaa
The available functions are listed inTable 2.
PERFORMANCEA summary of the MPAA020's perfor-mance features is listed in Table 1. TheMPAA020's circuit topology providesadvantages relative to chip area, pincount, and rail -to -rail voltage swingwith improvements being made to thesignal-to-noise ratio and signal band-width in arrays that will be introducedlater this year. One of these, theMPAA132, has a signal-to-noise ratio
p 7.1idia bi-' ...Awl*
LI'
0_42 V.
El ODE=I= LECI DO El El DEIP.I+bli
_._
r
"N.
-IliaP LHetP3 r
-3_
- -k-7-ig --._W1
1
O.-.-._-_17 I
:
0M
EEr
LI
LI
Vi -i-4) MI 11Q a-4ma
Figure 4. Example of aconfiguration possiblewith EasyAnalogTMsoftware.
(> 80 dB), a bandwidth of 1 MHz, inter-nal feedback facilities, for instance, forautomatic gain control (AGC), and amultiplier. A higher voltage family isalso planned which will operate inapplications with voltages greater than10 Y.
FINALLYUse of the array may initially be prob-lematic since it is housed in a 160 -pinQuadratic Flat Pack (QPF) with a pinpitch of 0.25 mm. This makes manualsoldering a little difficult.
[980059]
Table 2. Partial macro library - analogue functions
Single cell macro functions
Gain stages (max= 20' min= 1/20)Sum & diff. amps (3 weight inputs)Sample&holdTrack&holdIntegratorN -path integratorDifferentiatorHalf -wave rectifierFull -wave rectifierLPF rectifierCosine filterDecimatorBi-linear filter
Multi -cell macro functions
Low Q biquad filter (high- or low-pass); 2 cellsLow Q biquad filter (notch, band-pass); 3 cellsHigh Q biquad filter (high- or low-pass); 2 cellsHigh Q biquad filter (notch, band-pass); 3 cellsMaximum corner frequency = T/10Minimum corner frequency = T/100LimiterInterpolatorSchmidt triggerVoltage -controlled oscillatorSine -wave oscillatorSquare -wave oscillatorTriangle -wave oscillator
Elektor Electronics 6/98 59
signal buffering, programmable refer-ence voltage, cell to cell interconnec-tions, and so on.
Each cell’s function may be pro-grammed to connect with any of theother cells in the array. Unused cellsare powered down individually tominimize power dissipation.
Digital interface circuitry is pro-vided to write the analogue circuit con-figurations data to on-chip SRAM inthe same manner as Motorola’s digitalFPGAs (serial and parallel PROM, andmicroprocessor mode).
Analogue circuit design is simplifiedwith EasyAnalog™ design software,which handles bit level circuit configu-ration details, allowing the user to doanalogue design using functionalmacros in an easy to use point-and-click graphical environment.
E A S Y A N A L O G ™P R O G R A MThe EasyAnalog™ program may becompared with a circuit editor. Afterthe desired function has been selected,it is placed on one or more free CABs,and the input and outputs connectedas needed. Freeware EasyAnalog™runs under Windows 95 or NT andmay be downloaded from the Internethttp://sps.motorola.com/fpaa
The available functions are listed inTable 2.
P E R F O R M A N C EA summary of the MPAA020’s perfor-mance features is listed in Table 1. TheMPAA020’s circuit topology providesadvantages relative to chip area, pincount, and rail-to-rail voltage swingwith improvements being made to thesignal-to-noise ratio and signal band-width in arrays that will be introducedlater this year. One of these, theMPAA132, has a signal-to-noise ratio
(>80 dB), a bandwidth of 1 MHz, inter-nal feedback facilities, for instance, forautomatic gain control (AGC), and amultiplier. A higher voltage family isalso planned which will operate inapplications with voltages greater than10 V.
F I N A L L YUse of the array may initially be prob-lematic since it is housed in a 160-pinQuadratic Flat Pack (QPF) with a pinpitch of 0.25 mm. This makes manualsoldering a little difficult.
[980059]
59Elektor Electronics 6/98
Figure 4. Example of aconfiguration possiblewith EasyAnalog™software.
Single cell macro functions
Gain stages (max=20’ min=1/20)Sum & diff. amps (3 weight inputs)Sample&holdTrack&holdIntegratorN-path integratorDifferentiatorHalf-wave rectifierFull-wave rectifierLPF rectifierCosine filterDecimatorBi-linear filter
Multi-cell macro functions
Low Q biquad filter (high- or low-pass); 2 cellsLow Q biquad filter (notch, band-pass); 3 cellsHigh Q biquad filter (high- or low-pass); 2 cellsHigh Q biquad filter (notch, band-pass); 3 cellsMaximum corner frequency = T/10Minimum corner frequency = T/100LimiterInterpolatorSchmidt triggerVoltage-controlled oscillatorSine-wave oscillatorSquare-wave oscillatorTriangle-wave oscillator
Table 2. Partial macro library – analogue functions
Visit our Web site at http://ourworld.compuserve.com/homepages/elektor_uk
Elektor Electronics 6/98
Measuring liquid levels is a subject which never fails tocause heated debates. Does a technique exist whichmakes it possible to provide not only a ‘tank empty’
alarm, but also a continuous indication of the tank con-tents? The circuit described here operates in one of twomodes: stand-alone or remote (using RS-232 communi-cation). Once you’ve built the circuit, the fuel level in the
tank of a heating system, or the water level in the garden-ing tank, is no longer a secret.
60
Design by M. Vacher
liquid-levelgauge
an original application for the ST62T20
X1
4MHzC1
22p
C2
22p
S1
C3
100n
C11
330n
C12
100n
C10
100µ16V
C6
C7
C4
1µ16V
C5
100µ16V
C8
C9
K1
R2
10
k
R3
10
k
R4
10
0k
R6
47
0Ω
R7
47
0Ω
VPP/TEST
ST62T20
RESET
TIMER
IC1
OUTOSC
PA1
PA2
PA3
PB0
PB2
PB1
PB3
PB4
PB5
PB6
PB7
NMI
PA0
20
IN
18
17
16
15
13
14
12
11
10
19
1
43
6
7
9
8
2
5
S2
1110
1
IC2e
13
121
IC2f
5 61
IC2c
9 81
IC2d
R1
220k
R5
4k7
D4
1N4148
D1
1N4148
SENSOR
D3 D2
C13
100n93C06CB1
IC5DOUT
DIN
CSSK
2 1
4 8
53
67
T1
BC547
RE1
12V
MAX232
R1OUT
R2OUT
T1OUT
T2OUT
IC3
T1IN
T2IN
R1IN
R2IN
C1–
C1+
C2+
C2–
11
12
10
13
14
15
16V+
V-
7
89
3
1
4
5
2
6
K2
1 21
IC2a
3 41
IC2b
K378L05
IC4
IC2
14
7
IC2 = 74HC14
5V
C6 ...C9 = 1µ /25V
12V
TxD
C
NO
NC
12V
5V
970056 - 11
1
2
3
RE
Figure 1. The liquid-level gauge consist of a handful of partsonly; most of the work is done by IC1, a pre-programmedST62T20 microcontroller.
1
Looking at the increasing number ofcircuits based on a microcontroller, onemay start to wonder how it was pos-sible, only a few years ago, to designelectronics worthy of interest. If themicrocontroller has changed the faceof electronics for home construction,we have to recognise, though, that theappearance of this type of componenthas also caused a considerable reduc-tion in comprehensibility of circuitsbased on a microcontroller. On theplus side, the use of a ‘micro’ presentsa number of indisputable advantages,including a much simpler circuit dia-gram, and cost reductions due tofewer components and a smallerboard.
T H E E L E C T R O N I C SA look at the circuit diagram given inFigure 1 allows you to discover thatthe electronics are reduced to a fewelements only. IC1, A microcontrollertype ST62T10 from SGS-Thomson is atthe heart of the circuit. One externalpart the controller can’t do without isan EEPROM, IC5. The third elemen-tary component is the ubiquitousMAX232 which looks after the previ-ously mentioned RS232 compatibilityof the circuit.
The reasons for choosing an ST6controller for this application aremainly that the 62T20 version is afford-able (approx. £5), and that variousentry-level development systems areavailable, for example, the ST6 StarterKit, or the ST6 Programmer describedin Elektor Electronics November 1996.The latter allows you to write yourown programs an also ‘burn’ ST6 con-trollers.
When S2 is pressed, the ST6 is reset.Components R4 and C4 then brieflypull the reset input of the CPU logiclow, and so cause the microcontrollerto be re-initialised.
A number of lines from port B, PB4to PB7, connect the processor to theEEPROM (electrically erasable pro-grammable read-only memory), a type93C06B1. The EEPROM has a serialcommunication channel, and offers acapacity of 256 bits organised as16 words of 16 bits each (default modeused here), or 32 words of 16 bits each.
The on-chip oscillator uses a 4-MHzquartz crystal and the usual pair ofsatellite-capacitors, C1 and C2. Thecrystal is connected between theOSCIN and OSCOUT pins of themicrocontroller.
The 3-way DIP switch block, S1,has a function in the calibration of thecircuit. It will be discussed further on.
The component at the right-handside of the circuit is a MAX232. There’spractically no way to avoid this com-ponent because it provides everythingyou need to convert TTL levels (assupplied by the microcontroller) into
RS232 levels (as used by the PC portwith the same name), and vice versa.Here, the MAX232 sends data suppliedby the processor to the serial interfaceof the PC, which uses them for furtherprocessing.
In the lower left corner of the cir-cuit diagram you can see the electron-ics required for stand-alone operationof the liquid-level gauge: we’re talkingabout the relay which is directly actu-ated by the PA3 line of port A (pin 16).Actually, the port line controls therelay via a switching transistor, T1. Therelay contacts are connected to PCBterminal block K2. Depending on theway an external circuit, pump, valveor actuator can be controlled, the con-tacts have to be normally-open (NO)or normally-closed (NC) types.
The power supply is entirely con-ventional, being based on a three-pinvoltage regulator type 78L05 (IC4).
H O W I T W O R K SSchmitt-trigger gate IC2f acts as an RCoscillator in which R1 is the resistiveelement. The ‘capacitor ’ you wouldnormally expect to see in the oscillatoris formed by the two probes immersedin the liquid. As the capacitance is afunction of the liquid level, so is thefrequency produced by the Schmitt-trigger oscillator. This frequency ismeasured by the microprocessor, andthen translated into a correspondinglevel of the liquid in the tank. The pre-viously mentioned display enables thenecessary values to be stored in non-volatile memory (EEPROM 9306). Thisprinciple applies to different liquids.The author tried it with water andheating fuel (light oil). As illustrated inthe following table, the frequency vari-ation obtained with the immersionprobes depends on the liquid in thetank.
Let’s look at the level calculation ingreater detail. The signal supplied bythe oscillator is applied to the TIMERinput of the microcontroller, whichmeasures the total duration of 32,768cycles of the signal. In this way we geta 16-bit number which is proportionalwith the liquid level. Next, the ST6micro performs the following calcula-tion:
Level (mm) =250 (x−Nlow) / (Nhigh−Nlow)
where x is the measured value, Nlowthe ‘low’ calibration level stored inEEPROM, (Nhigh−Nlow) the differencefrequency stored in EEPROM and rep-
resenting a liquid level of 250 mm. Anyother value is possible provided youmodify the relevant parameter in thesource code file before assembling theprogram. The source code file is avail-able on a disk, the order code is976015-1.
As already mentioned, this circuit canbe used in one of two modes:
stand-alone: the system operates inde-pendently, the relay being actuatedwhen a programmed level is reached(for instance, to switch on a pump).
remote (RS232): the measured level isperiodically sent to a serial link (run-ning at 9,600 baud). The information isthen processed by another system (amicrocontroller or an alarm system).The liquid-level is directly supplied inmillimetres. The syntax of the ASCIImessages is
=00xxx<CR><LF>
where xxx is the measured liquid-levelin millimetres.
C O N S T R U C T I O NThe artwork for the printed circuitboard designed for this project is givenin Figure 2. The circuit board is single-sided and, unfortunately, not availablethrough our Readers Services.Although fitting all the parts isstraightforward soldering work, youshould observe the orientation ofpolarised components like integratedcircuits (including the voltage regula-tor), electrolytic capacitors, LEDs,diodes and transistors.
The microcontroller should, ofcourse, contain the right program. For-tunately, you can obtain it ready-pro-grammed through our Readers Ser-vices (order code 976515-1), or as partof a kit from a kit supplier.
The integrated circuits are mountedin sockets. There is only one wire linkon the board. Since the LEDs are onlyused during the calibration procedure,they may be fitted close to the PCBsurface. Be sure not to fit DIL switchS1 the wrong way around. The contactmarked ‘1’ should be close to resistorsR7 and R6. When the switches are inthe ‘on’ position, the levers should bemoved towards the ST6.
Measuring just 80×62 mm, the cir-cuit board should not be too difficultto mount in a plastic (ABS) case ofyour choice. The two immersionprobes may be made from plastic-cov-ered curtain rods with a diameter of4 mm. Their length will obviouslydepend on the depth of the tank. Theprobes should be fitted at a centre-to-centre distance of about 10 mm. Therods should pass through four holes inthe case (two additional holes at the
61Elektor Electronics 6/98
Liquid min. level(0 mm)
max. level(250 mm) resolution
Water 280 kHz 70 kHz 0.07 mmFuel 280 kHz 230 kHz 1 mm
rear side to keep them in place). Insidethe case, the plastic covering isremoved locally for the connection tothe two oscillator inputs marked SEN-SOR. The four holes are sealed withtwo-component glue or a potting com-pound to make the enclosure water-tight. In spite of this measure, the casemust also be mounted well out ofreach of the liquid in the tank.
The power supply is a bit more rus-tic. The supply voltage is furnished bya mains adaptor with an output capac-ity of 12 V/40 mA. This output voltageneed not be regulated, because it onlypowers the relay coil. The rest of theelectronics draws its supply currentfrom the 78L05 three-pin voltage reg-ulator on the board. The 93C06 (or93C46) EEPROM memory works in 16-bit mode (‘ORG’, pin 6 not connected),which allows circuits of the old gener-ation to be used.
S E T T I N G U POnce all components are in place onthe board, you are ready to proceedwith the calibration of the circuit. Thisprocedure consists of two phases:
1. ‘Low’ level calibration, probes notimmersed.The switches in S1 should be set as fol-lows: SW1-1 on, SW1-2 off, SW1-3 off.After switching on the supply voltage,or after a reset, followed by a stabilisa-tion period of about 2 seconds, the sys-tem launches the frequency measure-ment. The value assigned to Nlow iswritten into EEPROM. The end of theprocess is signalled by one of the twoLEDs coming on (the green one, D3,for ‘okay’, or the red one, D2, for‘error’).
2. ‘High’ level calibration, probesimmersed in liquid with a columnheight of 250 mm.Set the switch contacts as follows: SW1-1 on, SW1-2 off, SW1-3 on. Again, afterswitching on the supply voltage, orafter a reset, followed by a stabilisationperiod of about 2 seconds, the systemlaunches the frequency measurement.The value assigned to Nhigh as well asthe result of (Nhigh−Nlow) are writteninto EEPROM. The end of the processis signalled by one of the two LEDscoming on (the green one, D3, for‘okay’, or the red one, D2, for ‘error’).
Programming the levels at which therelay is actuated.This procedure is similar to the one forthe calibration, only switch SW1-2 isset to the ‘on’ position. The relay isenergised when the liquid tops the‘high’ level, and is switched off whenthe liquid drops below the ‘low’ level(hysteresis).
Note that the three switch contactsin S1 should be returned to the ‘off ’position for normal use of the liquid-level gauge.
O P T I O N SThis circuit may be used with or with-out a serial RS232 link. If the computersystem (or microcontroller) you wantto use for the processing of liquid-leveldata accepts TTL levels, then theMAX232 may be omitted. As it is notrequired in the present application, theRxD line is not implemented in theserial link. If you plan to extend theprogram with additional functions,you may connect pin 15 (PB0) of themicrocontroller to the R1OUT pin.This allows the RxD signal to beapplied to the R1IN pin. It is then alsoworthwhile to replace K1 with a 3-pinheader whose contacts may be con-nected to wires to implement theabove mentioned function.
(970056-1)
Warning. This circuit is not designedor approved for use in tanks contain-ing highly flammable, explosive,aggressive or corrosive liquids.
970056-1
970056-1
C1
C2
C3C4
C5
C6
C7
C8
C9
C10C
11
C12
C13
D1
D2
D3
D4
IC1
IC2
IC3
IC4
IC5
K1
K2
K3
R1R2
R3
R4
R5 R6
R7
RE1
S1
S2T1
V1
V2 V3
V4
X1
970056-1
SENSOR
TxDT+ 0
CNC
NO
COMPONENTS LIST
Resistors:R1 = 220kΩR2,R3 = 10kΩR4 = 100kΩR5 = 4kΩ7R6,R7 = 470Ω
Capacitors:C1,C2 = 22pFC3,C12,C13 = 100nFC4 = 1µF 16V radialC5,C10 = 100µF 16V radialC6-C9 = 1µF 25V radialC11 = 330nF
Semiconductors:D1,D4 = 1N4148D2 = red low-current LEDD3 = green low-current LEDT1 = BC547IC1 = ST62T20 (order code 976515-
1)IC2 = 74HC14IC3 = MAX232NIC4 = 78L05IC5 = 93C06CB1
Miscellaneous:K1,K3 = 2-way PCB terminal block,
raster 5mmK2 = 3-way PCB terminal block,
raster 5mmS2 = presskey, Multimec CTL3S1 = 3-way DIP-switchX1 = 1 MHz quartz crystalRe1 = relay, 12V, PCB-mount (e.g.,
Siemens V23057-B0002-A201)Optional: floppy disk with microcon-
troller source code, order code976015-1 (see Readers Servicespage).
2
Figure 2. Copper tracklayout and componentmounting plan of thesingle-sided printedcircuit boarddesigned for the pro-ject (board not avail-able ready-made).
Elektor Electronics 11/9762
C ngeover
Make 5 -02 3 -o -02
-01 -01
Break 3 L4
104n.
1
SAS max. 01.3.0.1
IIMINE2MEMEMIIIN
fladMOMENIMIlbitE inummunnunam f1/UDl E9.22EEP.TEMMff
Mounting hole ayout View onto the terminals
863006 13
#
DATASHEET 6/98
V23057 Card Relay E
Passive ComponentsRelays
UM8250A
Integrated CircuitsMicroprocessor, Interfacing DATASHEET 6/98
69E
lektor Electronics
6/98
V23057 Card Relay EUpright mounting, immersion cleanable. For PCBmounting, pin arrangement suits 2.5 mm and 0.1-in.grid in accordance with DIN 40801 and DIN 40803,fine.
V23057-B0*** with creep distances and clearances> 4mm*.
1 changeover contact, single or bifurcated, or 1make contact, single.
V23057-D0*** with creep distances and clearances> 8 mm*. 1 make or 1 break contact, single.
* between winding and contact.
ManufacturerSiemens. Siemens plc, Siemens House, WindmillRoad, Sunbury-on-Thames, tel. (0932) 752677, fax(0932) 752671. UK distributor: ElectroValue.
Ordering Code
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15h h h h h h 1 h h h h h 1 h h h h
Basic type numberof card relay EV23127 = suitable for processing on soldering linesV23057 = immersion cleanable
VersionB0 = with creep distances
and clearances > 4mm*.1 changeover contact,single or bifurcated,or 1 make contact, single.
D0 = with creep distancesand clearances > 8 mm*. 1 make or 1 break contact, single.
Coil numberSee Table 1
Contact(s)See Tables 2 and 3
UM8250AAsynchronous Communication Element (ACE)
ManufacturerUnited Microelectronics Corp (UMC), No 3, Li-HsinRoad II, Science-Based Industrial Park, Hsin-chuCity, Taiwan, R.O.C. Fax: (+886) 3-577-4767.Internet: www.umc.com.tw/. UK Distributors: FutureElectronics (01753) 76 3000; Ambar Cascom(01296) 395 635.
General DescriptionThe UM8250A is a programmable AsynchronousCommunication Element (ACE) chip fabricated usingthe S-Gate NMOS process. It performs serial-to-par-allel conversion on data characters received from aperipheral device or a modem, and parallel-to-serialconversion on data characters received from a CPU.
The CPU can read the complete status of the ACE atany time during the functional operation. It alsoincludes a programmable baud rate generator capa-ble of dividing the timing reference clock input bydivisors of 1 to (216–1), and producing a 16x clockfor driving the internal transmitter logic.
ApplicationAnalogue-to-Digital Converter, Elektor ElectronicsJune 1998 SupplementThe following tables of ACE register functions areoffered as a help to program the above converter.
Interrupt Identification Register (IIR)The ACE has an on-chip interrupt capability thatallows for flexibility in interfacing popular micro-processor presently available. In order to provideminimum software overhead during data charactertransfers, the ACE prioritizes interrupts into four lev-els. When addressed during chip-select time the IIRfreezes the highest priority interrupt pending, and noother interrupts are acknowledged until the particularinterrupt is serviced by the CPU.
Block 1 Block 2 Block 3
for V23057-BO... for V23057-DO...
Interrupt Identification Register Interrupt Set and Reset Functions
Bit 2 Bit 1 Bit 0 Priority Level Interrupt Type Interrupt Source Interrupt Reset Control0 0 1 — None None —
1 1 0 Highest Receiver Line Sta-tus
Overrun Erroror
Parity Erroror
Framing Erroror
Break Interrupt
1 -0 0 Second Received DataAvailable
Receiver Data Avail-able
Reading the ReceiverBuffer Register
0 1 0 Third Transmitter HoldingRegister Empty
Transmitter HoldingRegister Empty
Reading the IIR Register(if source of interrupt)
orWriting to the Transmitter
Holding Register
0 0 0 Fourth MODEM Status
Clear to Sendor
Data Set Readyor
Ring Indicatoror
Data Carrier Detect
Reading the MODEM Sta-tus Register
Reading theLine Status Register
#
V23057 Card Relay E
Passive ComponentsRelays
UM8250A
Integrated CircuitsMicroprocessor, Interfacing DATASHEET 6/98DATASHEET 6/98
70E
lekt
or E
lect
roni
cs6/
98
Bit n
o.
Regi
ster
add
ress
0 DL
AB =
00
DLAB
= 0
1 DL
AB =
02
34
56
70
DLAB
= 1
0 DL
AB =
1
Rece
iver
Buffe
r Re
g-is
ter
(Rea
dOn
ly)
Tran
smitt
erHo
ldin
g Re
g-is
ter
(Writ
eOn
ly)
Inte
rrup
t Ena
ble
Regi
ster
)
Inte
rrup
tId
ent.
Regi
s-te
r (R
ead
Only
)
Line
Con
trol
Regi
ster
Mod
em C
on-
trol
Reg
iste
rLi
ne S
tatu
sRe
gist
erM
ODEM
Sta
-tu
s Re
gist
erSc
ratc
hRe
gist
erDi
viso
r La
tch
(LS)
Divi
sor
Latc
h(M
S)
RBR
THR
IER
IIRLC
RM
CRLS
RM
SRSC
RDL
LDL
M
0Da
ta B
it 0*
Data
Bit
0En
able
Rec
eive
dDa
ta A
vaila
ble
Inte
rrup
t (ER
BFI)
‘0’ i
f Int
erru
ptPe
ndin
g
Wor
d Le
ngth
Sele
ct B
it 0
(WLS
0)
Data
Ter
min
alRe
ady
(DTR
)Da
ta R
eady
(DR)
Delta
Cle
ar to
Send
(DCT
S)Bi
t 0Bi
t 0Bi
t 8
1Da
ta B
it 1
Data
Bit
1
Enab
le T
rans
mitt
erHo
ldin
g Re
gist
erEm
pty
Inte
rrup
t(E
TBEI
)
Inte
rrup
t ID
Bit
(0)
Wor
d Le
ngth
Sele
ct B
it 1
(WLS
1)
Requ
est T
oSe
nd (R
TS)
Over
run
Erro
r(O
R)De
lta D
ata
Set
Read
y (D
DSR)
Bit 1
Bit 1
Bit 9
2Da
ta B
it 2
Data
Bit
2En
able
Rec
eive
rLi
ne S
tatu
s In
ter-
rupt
(ELS
I)
Inte
rrup
t ID
Bit
(1)
Num
ber o
fSt
op B
its(S
TB)
Out 1
Parit
y Er
ror
(PE)
Trai
ling
Edge
Ring
Indi
cato
r(T
ERI)
Bit 2
Bit 2
Bit 1
0
3Da
ta B
it 3
Data
Bit
3En
able
MOD
EMSt
atus
Inte
rrup
t(D
ESSI
)0
Parit
y En
able
(PEN
)Ou
t 2Fr
amin
g Er
ror
(FE)
Delta
Dat
a Ca
r-rie
r Det
ect
(DDC
D)Bi
t 3Bi
t 3Bi
t 11
4Da
ta B
it 4
Data
Bit
40
0Ev
en P
arity
Sele
ctLo
opBr
eak
Inte
rrup
t(B
I)Cl
ear t
o Se
nd(C
TS)
Bit 4
Bit 4
Bit 1
2
5Da
ta B
it 5
Data
Bit
50
0St
ick
Parit
y0
Tran
smitt
erHo
ldin
g Re
gis-
ter (
THRE
)
Data
Set
Read
y (D
SR)
Bit 5
Bit 5
Bit 1
3
6Da
ta B
it 6
Data
Bit
60
0Se
t Bre
ak(T
xD)
0Tr
ansm
itter
Empt
y (T
EMT)
Ring
Indi
cato
r(R
I)Bi
t 6Bi
t 6Bi
t 14
7Da
ta B
it 7
Data
Bit
70
0Di
viso
r Lat
chAc
cess
Bit
(DLA
B)0
0Da
ta C
arrie
rDe
tect
(DCD
)Bi
t 7Bi
t 7Bi
t 15
* Bi
t 0 is
the
leas
t sig
nific
ant b
it. It
is th
e fir
st b
it se
rially
tran
smitt
ed o
r rec
eive
d.
Table 1 Coil voltages
Coil numberOrder Code,
Block 2
Nominal voltage Operating voltage range at 20 ºC Resistance at 20 ºC
V DCMinimum voltage UI Maximum voltage UII
V DC V DCΩ
001 6 4.2 10.6 80 ± 8
002 12 8.3 21.5 330 ± 33
006 24 16.8 40 1200 ± 180
013 48 33.6 79 4700 ± 700
023 60 42 98 7200 ± 1080
Table 2 Contact specifications V23127-A0*** and V23057-A0***/-B0*** with single contacts
Ordering Code, block 3 A101 A201 A401 A102 A202 A402
Contact material silver, gold-flashed silver nickel silver-cadmium
oxidesilver, gold-
flashed silver nickel silver-cadmiumoxide
Contacts (see also base terminals)
Max. switching voltage as perVDE 0110 group C
V DCV AC
300250
Max. switching current A 5/151 8/151 5/151 8/151
Max. power rating 2, DC voltage50...330
(voltage dependent)
35...330
(voltage dependent)
up to24 V: 10030 V: 80200 V: 30
250 V: 50
50...330
(voltage dependent)
35...330
(voltage dependent)
AC voltage VA 1250 2000 1250 2000
Max. continuous current A 8
Mechanical life Ops. approx. 2 x 107
1 higher current may flow for a maximum of 4 seconds for up to 10% of on time.
WWW
W
up to24 V: 10030 V: 80
200 V: 30
250 V: 50
double, change-over
Table 3. Contact specifications V23127-A0*** and V23057-A0***/-B0*** with bifurcated contacts
Ordering code, block 3 B101 B601
Contact material silver, gold-flashed gold F
Contacts (see also base terminals)
Max. switching voltage as perVDE 0110 group C
V DCV AC
300250
3630
Max. switching current A 4/101 0.2
Max. power ratingDC voltageAC voltage
WVA
voltage-dependent500
5—
Max. continuous current A 6 2
Mechanical life Ops. approx. 2 x 107
1 higher current may flow for a maximum of 4 seconds for up to 10% of on time.
double, change-over
single, make