Académique Documents
Professionnel Documents
Culture Documents
A B
telephone network
PSTN PSTN
exchange national/international exchange
980018 - 11
Figure 1. Traditionally, part of an Internet link is by way of the plain old telephone system (POTS).
Internet connections make ever-heav- (PSTN), also jokingly referred to as POTS causes a digital connection to be
ier demands on the bandwidth of a (plain old telephone system). Corporate established using this (analogue) tele-
communication channel. Because of (business) users may resort to leased phone line. Using a modem bank and
the increased information density datalines, which are much faster. In a network server, the provider arranges
(these days nearly all web pages have fact, leased lines may supply a band- for the actual connection to the Inter-
graphic elements, and they support width that matches the exact require- net to be established. Depending on
sound and exhibit photographs), a ments of the (corporate) user, so that a the final destination, a Web site being
pretty fast data line has to be available perfect, seamless link may be made visited, or an e-mail message to be
to enable the web pages to be built at with the corporate network. This type of delivered, a second telephone line is
reasonable speed. Moreover, work on connection is far too expensive for indi- employed at the far side. Actually, the
new, heavier services like distributing vidual users, hobbyists and other occa- Internet bypasses a piece of the tele-
digital video via the Internet is under- sional users, who are more or less phone network whenever it allows two
way. Consequently, a 14k4 modem is forced to opt for a modem-PSTN con- end-users to communicate with each
just to slow to surf the Internet, while a nection. Typically, the available band- other. Ultimately, digital data from one
33k6 modem gives reasonable results, width will then depend on the quality of computer is delivered, unmodified, to
and an ISDN connection using a band- the telephone line and the modem another, which may be at the other
width of 64 kbits/s is a good alternative technology used. With analogue con- side of the globe! Once you get this
these days (though overpriced in the nections, that is, connections using a point, it is easily understood why tele-
UK). ISDN even allows you to aggregate standard telephone dial-up connec- phone companies are striving to find a
two channels, enabling a bandwidth of tion, the highest available speed is cur- solution that enables end-users to be
128 kbits/s to be achieved. Looking at rently offered by the asymmetrical connected directly to the Internet. The
the current trend in technological 56 kbits/s standard. telephone exchange then has to cre-
developments, you may safely assume With the increased demand for fast ate the link with the Internet. In a num-
that even that kind of speed will not be datalinks and the rapidly growing ber of cases, the service provider is
sufficient some time in the near future. number of Internet users in mind, new then no longer required. This option is
The speed of a modem is, incidentally, service providers seem to pop up currently under investigation, and there
only one aspect of the casethe inter- almost every day. The coming years are already encouraging signs from a
nal capacity (bandwidth) of the Inter- will no doubt bring us a fierce and manufacturer claiming to have found
net and the providers connection to interesting battle between cable- TV the solution.
the Net must not be bottlenecks either. companies, electricity boards, tele- Internet providers, afraid of being
In more cases than you would care to phone companies and other suppliers abandoned by their millions of users,
think of, the datastream between of information services. Each of these will of course try to counter this devel-
provider and end-user is ultimately will attempt to induce individual Inter- opment by offering services that
much slower than the theoretical net users to go for their offerings of fast reduce the role of the traditional tele-
capacity offered by the modem. Internet services. phone companies.
Lets return to the communication In this context, it is interesting to see that
between the end-user and the service Telephone at the hub the European Commission recently
provider. So far, two systems are in use decided that Internet phone links need
to realise the communication between The drawing in Figure 1 illustrates how not be subject to special regulations.
the Internet and you, the end-user. The todays Internet connections are built This decision may be revised, however,
vast majority of individual users will up. The end-user relies on his/her tele- when we enter the year 2000 and the
faithfully connect to their ISPs by way of phone line to dial the Internet Service EEC will look into the problems again.
the public switched telephone network provider (ISP). In this way, the end-user Until that time, Internet providers are in
HOW
DirecPC
WORKS
DirecPC
Network
CABLE LINE Operations
Center
(NOC)
information is
transferred to
to the
to your ISP WEB SERVER
through your
MODEM your request
is directed
START: Your PC sends 980018 - 12
a request
Figure 2. Using DirecPC the end-user employs a modem and a telephone line to dial into the server of his ISP , sending a
request for information. The requested information is returned over the air by means of a fast satellite link.
Analogue modems
A number of modem types are cur-
rently available for connecting to a
regular telephone line. The theoretical
speeds achieved by these modems
range from 14k4, 28k8 and 33k6, up to
56k. The first three rely on symmetrical
twisted twisted
pair pair
xDSL xDSL xDSL xDSL
IP Router
980018 - 13
Figure 3. With XDSL modem techniques (i.e., ADSL and VHDSL), high-speed data is sent down twisted-pair copper cables,
along with ordinary telephone traffic (PSTN or ISDN ). At the local telephone exchange, the data traffic is sent to an IP router.
Next, the traffic is arranged via the Internet.
In-phase operation
You can find a good theoretical intro-
duction to the subject of mains noise
in the article Mains-noise suppression
in the June 1997 issue of Elektor Elec-
tronics. In the present article, we con-
centrate on common-mode and dif-
ferential-mode noise. The difference
between these can be explained with
reference to Figure 1.
If the noise current flows from the
source of the disturbance through
both mains leads L (live) and N (neu-
tral) to the noise recipient, and then
returns via the protective earth lead,
PE, and the earth, this is called com-
mon-mode or asymmetrical noise. If
on the other hand the noise current
In the last few years, electromagnetic Designing a circuit with no regard to its flows through one mains lead to the
compatibility (EMC) has become an EMC behaviour, and then making it noise recipient and then returns via the
important consideration in the design compliant by packaging it in a metal other mains lead (but not via the
of electronic circuits. The designer has enclosure with a mains filter, will in earth), this is called differential-mode
an obligation to achieve a low level of most cases turn out to be an (expen- or symmetrical noise. As a rule, most
emissions and a high degree of immu- sively) mistaken approach. Adding a differential-mode noise lies under
nity to external noise, not only for cir- mains filter to a circuit which has 500 kHz and most common-mode
cuits that specifically work at high fre- proven to be overly susceptible to noise above 500 kHz.
quencies but also for most computer, external noise, in spite of the use of all
microcontroller and audio circuits. In other countermeasures, can only be
order to meet this objective at an regarded as a desperate solution.
appropriate cost, passive and active Mains filters are thus required even for
Technical data
EMC measures must be taken in all DIY projects, prototypes and laborato- Voltage: 230 V-
stages of the development process. In ry models. With commercial mains fil- Maximum current: 2A
principle, one can say that the later ters, the designers usual problem is Common-mode noise suppression:
that such measures are introduced in first to select a suitable filter from the (-3dB at 18 KHz): 50 dB
the course of development, the more profusion of types available, and then Differential-mode noise suppression:
difficult and expensive they are to to employ it in a manner which more (-3dB at 35 KHz): 50 dB
implement. or less meets the conditions stated on
Figure 2. Equivalent circuit for common-mode noise. Only one lead (L or N) is shown.
100
D
has an attenuation characteristic with due to the saturation of the core are (dB) fDM
three regions, as shown by the dashed undesirable; in other words, the peak 80
line in Figure 3. (Most commercially noise currents should not exceed a
available mains filters have a similar certain value although it is precise-
DM
characteristic.) The low-frequency ly these unpredictable noise peaks 60
passband region is dominated by the which the filter is supposed to render CM
12 dB/octave; the effect of the para- tance of the choke windings. This is for- L Cp
sitic components is negligible in this tunate, since otherwise the leakage C
20
region. In the high-frequency region currents through the capacitors to PE Ls
the situation is exactly the opposite: could become large enough to trip
the parasitic components form a 12- the earth-leakage current breaker 0
dB high-pass filter (fres = 1/2LpCp) (ELCB). In addition, suitable capacitors 10 100 kHz 1
f
10 100 MHz
fCM1 fCM2
which allows very-high-frequency that are approved for use in such filters 982032 - 13
common-mode noise to pass with are only available with relatively small
almost no attenuation. The two points values. Specifically, only so-called
Figure 3. Attenuation characteristics for
of maximum attenuation, between EMI-suppression capacitors meeting
common-mode noise (dashed line) and
which the attenuation is relatively uni- the EN132400 standard can be used;
differential-mode noise (solid line).
R1
R3 Siemens B82723-A2202-N1
L1
C1 C2 C3
N X2 X2 Y2 N Miscellaneous:
470n 470n 4n7
220k
R2 B82723-A2202
-N1 275V
F1 = fuse, 2A slow, with PCB mount
2x 5mH6/2A holder.
K1,K2 = 3-way PCB terminal block,
C4 pitch 7.5mm.
Y2
4n7 Case: Bopla SE432DE.
982032 - 11
PCB, order code 982032-1, see
Readers Services page.
Figure 5. The complete schematic diagram of the mains filter with its nine components.
H3
H2
F1 2AT
982032-1
L1
P K1 C2
C1 R1
C4
N
R3
R2
C3
K2
H1
H
N P
dent resistor) across the mains lines. the mains voltage (and for the varistor
Such a component has a very high characteristics) a nominal varistor volt-
resistance for voltages lower than its age of 275 volts appears to be appro-
rated voltage, and a very low resis- priate. For the varistor to do its job
tance for voltages above this value. properly, a series resistance is neces-
Considering the allowed tolerance for sary to absorb the excess voltage.
The source impedance of the mains
network, which rises with
increasing frequency,
signal wires
comes into play here. In the
frequency range in question
good (10 MHz and above), the
source resistance Zg is
greater than 150 , which is
circuit more than enough for prop-
er operation of the varistor.
Data Valid
Byte mode
982041 - 13b
1284 adapters connected via the ISA are mirrored back into this region. For Connectors
bus and also support ECP. It addition- example, selecting addresses 0x378H The IEEE 1284 standard does not limit
ally specifies a number of modes that and 0x778H with newer types of ISA itself to defining the structures of the
the adapter must support. These are cards gives access to two different ports and the electrical signals to be
summarized in Table 4. registers located in two different used. The mechanical specifications
The ECP register model is comparable banks, while with older-model cards of the connectors to be used are also
to that of a standard parallel port, but these two addresses will both access included in the standard.
it also makes use of an important fea- the same register. Three different types of connectors
ture of the ISA interface. In the stan- The advantage of this technique (types A, B and C) are proposed in the
dard IBM architecture only the first (which is called aliasing) is that new standard. Type A is a 25-pin sub-D
1024 register or I/O addresses are cards can have invisible registers and connector, which is long since a famil-
used. This is the address space starting thus can increase the maximum num- iar feature of the PC. Type B is the well-
at 0x000H and ending at 03FFH. Ten ber of available registers. With this known 36-position Centronics connec-
address bits (AD0:9) are required to approach, other I/O registers located tor, which can be found on many print-
address this region. on other cards will never be disturbed. ers. Type C, a miniature 36-pin con-
In order to keep costs down, the hard- In this way the new cards remain com- nector with retaining clips, is especial-
ware of older-model ISA cards uses patible with the older cards, which ly intended for use in new designs. This
signals from the ISA bus to help in can only decode 10 address bits. last type of connector takes up less
decoding the address. This automati- The ECP register model makes use of space on the circuit board, and it is
cally limits the number of addressable the aliasing technique. It defines six easily attached and removed thanks
registers to 1024. PCs that are more registers that use only three I/O to the retaining clips. An additional
modern decode more address bits addresses (see Table 3). Two registers important advantage of this connec-
and thus have a larger I/O address located at addresses 0x000H and tor is that with it two additional signals
space available. A consequence of 0x400H, respectively, have somewhat are added to the interface: Peripheral
this is that locations in the first 1K block different names and functions Logic High and Host Logic High. These
of the address space can be doubly depending on the ECP mode in which signals can be used to ascertain
addressed, since the limited address- they are used, as can be seen from whether the equipment at the other
decoding logic of older-model ISA the table. The fourth and fifth columns end of the cable is powered on. In this
cards generates mirror addresses in of Table 3 list the various ECP modes way a certain degree of power man-
this block from higher-level addresses. and identify the functions these regis- agement is made possible by the IEEE
Software can address the new register ters have for each mode. The actual 1284 standard. (982041)
banks by adding 1024 (0x400H) to the mode must always be selected in
base address which is used. The new advance by writing one of the mode
addresses thus lie outside of the origi- codes to register 0x402H.
nal 1K address region, but in fact they
The receiver
described here is of
the Direct Conversion
type and is designed
for the 20-metre band
or, in new money,
14 MHz. The compo-
nents used are read-
ily available from the
junk box or can be
modified from stan- The 20-metre amateur radio band The RF part is concerned with the
(14 MHZ to 14.350 MHz in IARU Region 1) receiving of the signal. It consists of
dard components. is great for daytime DXing. During two sub-circuits: RF amplifier and fil-
The whole idea of periods of increased sunspot activity,
the band is open all the time, and the
ter, T1-L3-C2, and a VFO (variable-fre-
quency oscillator) T3-L1-C11. The
building simple radio 20-m band is buzzing with activity. band-pass filter around IC1 serves to
When there is little or no solar-flare clean up the received signal and make
projects is that they activity, the band is still usable for it much better to listen to. The fourth
can be constructed medium-distance communication dur- essential building block is the LM386
ing daytime hours and at sunrise and audio amplifier, which raises the
from the bits and sunset. Because the so-called MUF cleaned signal to loudspeaker level.
pieces found in the (maximum usable frequency) only Inductors L1 and L3 are home
exceeds 14 MHz around noon during made and the two tuning capacitors in
shack (a shack is a the winter months, a fairly large skip the receiver are preferably made from
zone has to be taken into account. tuning capacitors having larger capac-
radio amateurs Unfortunately, discussions on the intri- itance values more about this fur-
shed). cacies of radio wave propagation and ther on.
the impact of the solar cycle on radio The received signal from the aerial
communication are beyond the scope is fed into the source of a FET-based RF
of this article. The good news, how- amplifier, T1, configured in grounded-
ever, is that there are excellent books gate mode to present a low impedance
available on these subjects. to the aerial. The amplified signal is
tuned to the desired frequency by
CIRCUIT DESCRIPTION means of inductor L3 and variable
The circuit diagram shown in Figure 1 capacitor C2. This selected output is
may appear to be rather complicated at taken to a pair of diodes wired as a bal-
By Eric Edwards GW8LJJ first blush. There are four sub-circuits. anced mixer. The other signal used for
12V
* see text A 12V E 5V7 I 3V2
1 voir texte
* siehe Text
R19
B 1V4 F 2V3 J 6V
100
* zie tekst C C 10V6 G 5V2 K 1V4
*
C29 C4 R3 D 4V9 H 3V8 L 11V9
12V
1k5
M 5V8
10 63V 100n
T2
D1
E
L1
*
12...13V6
L3
* 2H2
12V 31mA
P1 T3 K1
C3 C5 BC 0V
a b
2x 550C C8 D3
1N4148
100n 1n D
C10
1k D2 100p 1N4001
F BF256B
15p
A R2 R4 R5
C6 C7 C9 8 C27 C28
T1
470
680
15k
C11 IC1
BF256B 150p 50p 4 100n 1000
220p 470p 16V
C1 B
C2
R1
10p
1k
100p
12V 12V
0V
R12 R17
1k
10
12V C17
L2 * R7 C20 K K
L
4n7 C23
10k
C22
J R11
4n7
G 150k J R16 47 1000 16V
1mH 0V
C15 R9 C16 75k IC2 16V
R6 2 0V 8
10k J C19 6
R14 3
C26
1M
0V 1 6 1 M 0V
470n 4n7 IC1a 10k C21 5
C13 T4 3
4n7 IC1b
7 LM386
2
J 5 100
10 4
J 16V
J 63V 7
10 LS1
63V H
10
BC550C J R18
I C24
R8 R10 R13 R15
C12 C14 C18 P2
M
4k7
2k7
1k
33k
10k 22 C25
IC1 = TL072 log. 16V 8
10n 22 100 16V
16V 100n
980036 - 11
H1
tnemgeS )C(
T
R8
R6
R7
C3
C18,C26 = 100F 16V radial
L3 T4
C1 L2 R12 C18 C22,C28 = 1000F 16V radial
R13
C13
R9
P1 C19 C16
Inductors:
R19 C20 R11
C29 L1 = 15 turns, 36 SWG (0.2 mm dia)
R14 C17 enamelled copper wire on Neosid
C5 10K1 coil assembly (purple core).
R15 R10
T2
R3
L1
C23
turns, 36 SWG (0.2 mm dia.) enam-
C7
elled copper wire on Neosid 10K1
R5
C26
C9 C8 C10 P2 C28 and 10T1.
C25
H2
H3
T T LS
C11
Semiconductors:
D1,D2 = 1N4148
D3 = 1N4001
T1,T3 = BF256B
T2,T4 = BC550C
IC1 = TL072CP
IC2 = LM386N-1
Miscellaneous:
K1 = mains adaptor socket, PCB
mount
Printed circuit board, order code
980036-1, see Readers Services
page.
Loudspeaker, 8 , 1 watt
36 SWG (0.2 or 0.25 mm dia.) enamelled rewind of the VFO coil may be neces-
copper wire wound on Neosid type sary to solve the problem. In general,
10K1, 10F1 or 10T1 formers. The RF however, you should not have trouble
input tuning coil, L3, consists of 15 in finding the centre of the band. Right
turns (L3a, primary winding) and you now have the signal from the sig-
3 turns (secondary winding, L3b). Both nal generator, which should be modu-
are close-wound and connected to the lated to help resolve it. Reduce the size
pins of the 10K1 base as illustrated in of the signal until it is just audible,
Figure 2. An idea may be to secure the adjust L3 for the strongest signal. You
main winding, L3a, with candle wax are now ready to try the receiver on
or super-glue. When set, wind L3b in air. Remove the signal generator and
the same direction on top of and in the switch it off. Also switch off your com-
centre of L3a. Secure this winding and puter because those things generate so
remove insulation for soldering to the much noise you will think your
base pins of the plastic former. Check receiver is hyperactive. Hopefully by
the continuity of the coils at the base now you would have picked up a sig-
pins, and make sure that the screening nal off air. Adjust C2 and C11 for the
cans may be fitted later without caus- most comfortable reception together
ing short-circuits with excess solder with adjustment of the volume con-
blobs at the pins. trol. If AM (Amplitude Modulation) is
The RF choke, L2, is a ready-made breaking through, carefully adjust pre-
miniature type. It should be noted, set P1 (across the diodes ) for maxi-
though, that it probably gives inferior mum attenuation. It works!
performance when compared with a Throughout the preparation of this
real RF choke removed from some old article I have been constantly listening
equipment, probably because of the to many DX (long-distant) amateur sta-
relatively low Q (quality) factors of tions in SSB and CW, and have received
those modern miniature chokes. Radio some very good SSTV pictures from dis-
rallies and amateur radio suppliers will tant stations.
have these older (physically larger) If a signal generator is not available
types. for calibration, a general coverage or
The audio amplifier is of a well- ham radio receiver could be used.
proven design with a chip that is eco- Place the aerial lead of the commercial
nomic to buy. receiver near the oscillator circuit of
The printed circuit board designed your home-brew receiver, but not too
for the receiver has large copper areas close as to pull the oscillator of your
at the solder side to reduce stray radi- home-brew. This is not as accurate as
ation, and to enable all component the signal generator but will give you
leads to be kept as short as possible, an idea where you are. Once you are
which is essential to keep the circuit receiving off-air stations, these can be
stable. Use a slow motion dial for the compared with those on the commer-
main tuning, and ensure all wiring is cial receiver providing that receiver is
rigid, with the receiver built on a solid good enough! The main tuning control
chassis or framework. Do not use a does exactly what it states and the RF
front panel that bends as you tune or tune control peaks the signal to maxi-
you will never be able to track anyone mum. You may have to back this off
down. Good rigid construction with for those stronger (S9+) signals
neat short wiring is the secret to this
good performer. START LISTENING!
Enjoy building and using this little
IN USE receiver, I think you will be surprised
Set the tuning capacitors about half with the results.
mesh (mid-way) and the coil cores half One thing to bear in mind whether
way down the former, adjusting with you are listening on a home-brew or
a non-metallic tool, otherwise you will commercial short wave receiver is that
be left with iron dust but no core. With the reception is very much affected by
the aerial and the all-important earth the time of day/night and also the time
connected, turn the volume to about of year. Weather and atmospheric pres-
half way. sure will also affect the reception.
If a signal generator is available, set Radio amateurs use different fre-
it to 14.20 MHz. This is centre of the 20- quency bands at different times of the
metre band. The output of the gener- year, day and night.
ator is connected to the aerial socket of Finally, a directional aerial is much
the receiver, and is adjusted so that the better than a long untuned wire,
signal is heard but not too strong. If no although you will receive many DX sta-
signal is heard and providing all is tions with a long wire aerial.
okay with your construction, tune the 73s and good DX.
main tuning capacitor to bring in the (980036-1)
signal. If still nothing is heard, you
may be too much off frequency; this is
easily detected by swinging the signal
generator until a signal is heard. A
output
Figure 2. Timing diagram IC1e 10
of the parking sonar.
t 980030 - 13
3 IC1a
10p
IC1b
12V12V D8 L1
10H
7806 6V
R3 1N4001
1 2 3 4
1 1
470
D9
C14 C15 C16
IC2a IC1c
1 100 10
P1 3 5 6 100n
& 1 15V 25V 10V
4k7 C2 2 0 1W3
150k
R1
5k6
R2
1n 6V
f0
40kHz
14 C17 14 C18
R4 IC1 IC2 6V
1k 7 100n 7 100n
D1
R10
R5
120k
390k 1N delay
4148 30cm - 3m 6V
IC2c
4 P2 R6 9 C5 R8 IC2d IC1d
6 8 12 D3 R9
5 & 1k2 10 & 1k 11 9 8
13 & 2k2 1
47k 10n R11 R14
IC2b D2 1N4148
BZ2
1k8
1k8
window R7 R22 D7
C3 C4 20cm C6
6V 1N4148
82k
56k
detector D4 D5 D6
470n 470n >5 33n
R16
12
1N4148
120k
1N4001
IC1f
1 R12
4V
1k
13
P3 5 IC1e T1
3 7 6V 11 10
100k C13 R20 IC3b 1
1 6 R13
R15 IC3a 1k2 C7
2M2
2 R19
3k9 3n3
BC557B
100
C8
R18 R21 470n
1n 330k 100k
2V5
X1 R17
C11 C12 8
150k
C9 C10
IC1 = 74HC04 IC3
100n 220 4
10 100n IC2 = 74HC132 10V
10V
IC3 = OP279G
980030 - 11
0
Figure 3. Complete circuit
C15 C14
P3 C10 C9 diagram of the parking sonar.
4 D9
The transducers are X1 and
H4
H1
12V C8
L1
R19
R17
R16
D3
R2 R9
R5 C5
C6 R4
IC2 R6
IC1 D1 C4
R12 D2
C1
P1 D4
R13 C3 P2
R3 T1
R11
D7 Figure 4. Printed-circuit
D5 C7 R14
H3
H2
Parts list
Resistors:
R1, R17 = 150 k
R2 = 5.6 k
R3 = 470
R4, R8, R12 = 1 k
R5 = 390 k
R6, R20 = 1.2 k
(C) Segment
980030-1
R7 = 82 k
R9 = 2.2 k
R10, R16 = 120 k
R11, R14 = 1.8 k
R13 = 2.2 M
R15 = 3.9 k
THE DESIGN
The block diagram of the tester is 1 8 UCE
shown in Figure 1. The Centronics DAC
interface in the diagram is linked to 11-bit
the printer port of the computer. Six counter
2 C E
data lines of this port are used for con- IB
3
trolling the counter and the ADC, DAC B B
while a handshake line (input) is used Centronics
interface 4
to read back the serial data from the NPN E
PNP C
(ECP).
IC
The tester is based
on an 11-bit counter,
IC1. This chip, which IB
9V
CIRCUIT DESCRIPTION
The circuit in Figure 3 is largely iden-
tical to that of the 1993 tester: the only
real change is the substitution of an
AD557 for a ZN425 in the IC2 position.
The UCE for the transistor on test is
generated by digital-to-analogue con-
verter IC2, amplifier IC3a and driver T1.
The DACs generating the various
base currents are discrete designs. The
base currents for the n-p-n transistor
on test are produced in a simple man-
ner: resistors R8R11 convert the logic
level (5 V) of the three MSBs output by
10
11
16
1k
1k
K1 9 1 16
0 D0 Vout C12 8
36 18 SELECT CTR12 7 2 15 3 T1
1 D1
IC2 A IC3
35 17 11 6 3 14 1 100n
IC3a
ERROR
CT=0 2 D2 B 4
34 16 5 4 2
3 D3
BD139
27k
33 15 3 5 R6
IC1 4 D4
AD557
32 14 2 6
5 D5 JN
31 13 CT 4 7 9
74 6 D6 CE IC3 = LM358
30 12 13 8 10 R3 R4 R5
HCT 7 D7 CS
27k
27k
33k
29 11 BUSY 4040 12
8
28 10 10 14
+ 9 12 13
27 9 15
10
26 8 D6 1
11
25 7 R8 D1
24 6 8 150k
NPN C E PNP
D4
23 5 D3 R9 1N4148
T5 T6
22 4 D2 150k D2
B B
21 3 D1 R10
20 2 D0 150k 1N4148
19 1 R11 D3
37k4 E C
Centronics 1N4148
T2
R12
5V
100k
T3
C13 R13
100k
100n T4
R14
100k
14 20 5
R19 R20 R17 R18 R15 R16
REF + 1 7
A0 IC3b
8k66
17k4
34k8
27k
27k
27k
2 6
IC4 A1
47k
16 3 R24
DATA A2
4
18 A3
IO CLK 5 R23 R21
3x BC547
A4
17 6
270k
ADDR A5 R25
1k
A6
7 1V82
19
SYS CLK 8
A7
15 TLC 9
CS A8
1541 11
A9
R22
12
A10
REF 270k
13 10
16V3
IC5 IC7
TR1
LM317 7805 5V
D4 D5
K2
150
2k2
4x R27 R26
1N4001
D6 D7 C2 C5 C9 D10
C7 C8
1k8
R28
100 100 100n 100n 100
63V 35V 25V
C4 C6
150
R29
2x 9V
3VA3 C3 100 100
35V 25V
100
IC6
332
35V R30
D8 D9
Figure 3. Circuit dia-
1V82
gram of the BJT980022 - 11
2x 1N4001 LM337
tester. All compo-
nents are readily
available types.
value is low. This has the drawback, stable, is that UCE is increased from 0 V between 0 V and 9 V. This value is
however, that the drop across it, that to 9 V in 256 steps during the test brought to the requisite level by atten-
is, the test voltage, is very small. It is period. This shows the effect of UCE on uator R21-R22. The resulting signal is
therefore raised 48 by IC3b, before it the current gain. applied to input A3 of the ADC.
is applied to digital-to-analogue con- The test value is limited because the The power supply is straightfor-
verter IC4. measurement range of the ADC is ward based on a traditional design.
The resulting signal to input A0 of only 5 V. Given a voltage amplification Only the auxiliary negative line is a lit-
the ADC is shown in Figure 2. This of 48 and a test resistance of 1 , it is tle out of the ordinary. An alternating
indicates that with seven base currents clear that IC (or IE) cannot exceed current flows through C3: one half
a voltage proportional to IC (in of an 104.2 mA. Since the base current is cycle via D8 and the other half via D9.
n-p-n device, IE) is measured. The rea- 175 A maximum, the current gain Each of the half cycles charges C4; the
son that the current rises in the course cannot be higher than 595. resulting potential across this capacitor
of time, although the base current is In practice, the measured UCE lies is held at a level of 1.8 V by IC6d.
Parts list
4
H2
R10 R7
C10 R9
Resistors: K1
R13 D2
R1, R2, R25 = 1 k R11
R3, R4, R6, R15, R17, R19 = 27 k R14
R1
IC2
R5 = 33 k C1 D3
IC1
C
R7 = 10 , 5 W 36 18
D1
R8R10 = 150 k R8 R22 B PNP
R12
C11
R11 = 37.4 k, 1%
R2
1
R21
R12R14 = 100 k R6 E
R16 = 34.8 k, 1% C13 R25
R5
R18 = 17.4 k, 1% T1 C
R24
R4
IC3
R20 = 8.66 k, 1% B NPN
R3
R21, R22 = 270 k 1
T2 T3 T4
R23 = 1.00 , 1% E
C12
IC4
R24 = 47 k
R23
R15
R16
R18
R17
R20
R19
R26 = 2.2 k 980022-1
R27, R29 = 150 1-220089
tnemgeS )C(
R28 = 1.8 k 1
R30 = 332 , 1% 19 1
Tr1
D8
C6
Capacitors: C3
D6
D4
D9
C1 = 220 F, 25 V, radial
C2 = 100 F, 63 V, radial C4 ~ K2
C3C5 = 100 F, 35 V, radial D10
C2
C6, C9 = 100 F, 25 V, radial
C7, C8, C10C13 = 0.1 F
D7
D5
C5
Semiconductors:
R26
R30
R29
R28
R27
D1D3 = 1N4148 C8 ~
IC6
IC5
D4D9 = 1N4001 C7
D10 = LED, high efficiency
T1 = BD139 IC7
C9
T2T4 = BC547B
T5 = n-p-n transistor on test
T6 = p-n-p transistor on test
Integrated Circuits:
IC1 = 74HCT4040
IC2 = AD557JN
IC3 = LM358
IC4 = TLC1541IN
IC5 = LM317 (SOT220)
IC6 = LM337 (SOT220)
IC7 = 7805
Miscellaneous:
K1 = 36-way Centronics socket,
right-angled, for board mounting
K2 = 2-way terminal block for board (C) Segment
mounting, pitch 7.5 mm 980022-1
Mains socket with integral fuse
holder and fuse (200 mA slow
blow)
2 off 2 mm banana socket, blue (c)
2 off 2 mm banana socket, black (b)
2 off 2 mm banana socket, red (e)
DPDT mains on/off switch
Mains transformer 29 V, 3.3 VA,
e.g., Monacor/Monarch VTR3209
Enclosure 11550135 mm
(4.5325.32), e.g., ESM
EB11/05)
PCB Order no 980022*
Software Order No. 986005-1* (The
DOS version of the software is still
available under Order no. 920144)
* These items are available as a sin-
gle item under Order no. 980022-C
Figure 4. The printed- CONSTRUCTION Start by laying the wire bridges and
circuit board is Building the tester is straightforward, fitting socket K1. Follow this by fitting
designed to house all particularly if it is done on the printed- first the passive and then the active
components, includ- circuit board shown in Figure 4. The components. Mind the polarity of the
ing the mains trans- board is compact and designed to electrolytic capacitors, transistors and
former, but excepting house all components except the integrated circuits.
the on/off switch. on/off switch, but including the mains No special sockets are used for the
transformer. test terminals, which instead consist of
Figure 5. Photograph
of the completed pro-
totype. The tester is
linked to the PC via a
standard printer cable.
* TA = traffic announcement
TP = traffic programme **CD-quality audio broadcasts will in themselves be insufficient incentive for the majority
PTY = programme type of listeners to splash out on a DAB receiver.
TDC = transparent data channel Moreover, it is a well-established fact that although most of us spend a lot of money on
RT = radio text audio hi-fi equipment, this is usually installed in the living room, while most of our listen-
GPS = Global Positioning System ing is done on a simple (often portable) receiver in the kitchen, bathroom or bedroom.
Figure 5. Frequency
spectrum of a DAB eral services are divided into packets Although Eureka 147 is a combined
transmission in of fixed length. Each packet has as European development, great interest
Band III. (Courtesy: address the identification of the service in it has come from beyond Europe,
Eureka) to which it belongs. The capacity of culminating in the setting up of the
each service may be increased if cer- WorldDAB Forum early last year. The
tain service addresses are transmitted forum was set up to take DAB from the
then released is used for data services; more frequently than others. development laboratory into the indus-
this is termed Fast Information Data try and thence to the consumer.
Channel. The FIDC may be used for TODAY It is hoped by manufacturers and retail-
paging or alarm systems. Ten years have elapsed from the orig- ers alike that DAB will take off in
The MSC is divided into subchan- inal concept of Eureka 147 in 1987 to earnest this year. Unfortunately, the
nels (SC), whose capacity must be a the introduction of the first DAB con- signs for this to happen are none too
multiple of 8 kbit/sec. An arbitrary part sumer products (by a number of Euro- encouraging. There are two problems:
of the MSC may be used for data ser- pean as well as Japanese manufactur- one is the dearth of DAB receivers (since
vices. ers) at the International Consumer no one really knows what the consumer
In the so-called Stream Mode, a Electronics Fair in Berlin last year. wants** and is prepared to pay); the
subchannel will cater for only one ser- Design of suitable chips went hand second is that, in general, the public has
vice at a fixed allocated data rate. In in hand with the development of the not (yet) been converted to DAB.
certain circumstances, this leads to an equipment. This design was under- Nevertheless, the WorldDAB
uneconomical use of the capacity, but taken by a group of research estab- Forum reckons that this year about 100
it has the advantage that the receiver lishments led by Philips. The chips will million people in Europe will be in
can evaluate the data in a straightfor- be manufactured by Philips, Bosch, reach of DAB services: about 40 million
ward manner. Temic and SGS Thomson (at the time in Germany, 25 million in the UK, and
In the Packet Mode, the data of sev- of writing: early 1998). 35 million in Sweden, Belgium, Den-
DAB in the United Kingdom and plans for initial DAB services in late 1996. A group of
Toronto-based commercial broadcasters have established
some Commonwealth countries full-time transmitters on Torontos CN Tower to carry the
programming of 15 radio stations. At the same time, CBC
In the United Kingdom, the Government set the frame- announced some months ago that it would bring DAB to
work for the development of DAB in the 1996 Broadcast- 75% of the population over the next five years.
ing Act. The BBC began an operational DAB service in
September 1995 in the London area and is now building India is looking to the Eureka 147 system as the future
up its transmitter network so that 60% of the British pop- of radio there. The public broadcaster, All India Radio,
ulation will be covered by the middle of this year. The reg- started preliminary studies and experiments and set up a
ulatory body for commercial radio, the Radio Authority, has test transmission system in Delhi. Over the past year or
announced a fast burn approach to licensing private so, work has been orientated towards DAB planning, satel-
radio multiplexes at both national and local level. A num- lite distribution of a DAB ensemble and its relay in major
ber of private stations are already operating a pilot ser- cities.
vice.
In Singapore, DAB was demonstrated at the Asia Telecom
Australia looks to be moving ahead towards adopting the 97 by the Singapore Broadcasting Authority in association
Eureka 147 system. in 1996, the Government-appointed with Deutsche Telecom. Further trials, using both L-band
Digital Radio Advisory Committee expressed support for and VHF, took place last year.
Eureka 147 as the appropriate delivery system for Aus-
tralia. Pilot services are operating in three major Australian In South Africa, a seminar in late 1996 on DAB organized
cities. Tests have also been carried out of satellite-deliv- by the South African Broadcasting Corporation and Sen-
ered DAB. tech, the transmission provider, led to the formation of a
South African DAB Association. Sentech has set up a test
Canada has remained at the forefront of Eureka 147 devel- facility in Johannesburg to serve as a pilot and develop-
opments after commercial radio and CBC announced their ment system.
7
980017 - 58
Brief specification
Measurement range 10100 V m1
Frequency range 10 Hz 2 kHz (1 dB)
5 Hz 5 kHz (3 dB)
Measurement error <10%
Sensor capacitance 80 pF
Input capacitance 5 pF
CMRR >60 dB
R17
U
10k
C11 U
R6
D1
1M8 3
2x 3V R16 47p
1 C1
BAS45A
R1
3
1
100
1
IC2a
2
R18
C15
6k8
500V 2
100V/m
10
10V D13
C4 R2
C2 C3
90V/m
3
22M
D12
R12 9 10 80V/m
1n 47p 10p MODE L10
R3 15k 11
C8 6 L9 D11
1%
2M2
RHI 12
E1 R9 1% 7 L8 70V/m
6 REFOUT 13
3V 22 10k R14 L7 D10
7 5 IC3 14
E0 IC2b 100k SIG L6 60V/m
R7 5 R15 15
10k L5 D9
15k
3k9
P1 16 50V/m
E2 R10 1% LM3914 L4
P2
4M7 17
4k7 L3 D8
C7 R5 8 18
C6 R11 R13 REFADJ L2 40V/m
22M
C9 4 1
U RLO L1
15k
4k7
D7
1n 100p 30V/m
R4 22 2
D3 1% 10V D6
22M
6 20V/m
7
C5 IC1b D5
5 S1 D15 U 10V/m
10p D4
500V 2x R8 BT1 1N4001
D5 MPR (10V/m)
BAS45A 1M8
8 8 D6, D7 MPR (25V/m)
C12 C13 C14
9V
IC1 IC2
100 4 100n 4 100n
16V IC1 = TLC272
IC2 = TLC272
980039 - 11
2 4 4 4 4
10
10
150
150
front plate rear plate
30 x 30 cm 30 x 30 cm
6
3 3 3
300
300
20 20
10
10
4 4 4 4
10 150 10 10 10
150
(dimensions in mm) 980039 - 12
300
C12
C11
Note that for emissions complying 3 D3 H3
C10
R18
R17
R16
H2
-
mendations (which is the case with
R7
E2
BT1
most modern VDUs) only the three R8
C8
IC2
C6
R11
+
IC1
C14
D15
After it has been buffered and
E0
P1
C4 C7
S1
amplified, the reference potential of
980039-1
C13
R6
R10
R12
R9
R3
C15
D1
D2
P2
C2
tnemgeS )C(
1-930089
mentation amplifier.
C9
IC3
R13
R2
a single 9-V (dry or rechargeable) bat-
C1
R1
D13
D12
D11
D10
D8
D7
D6
D5
H4
CONSTRUCTION
The most demanding part of the cir-
cuit is the sensor, a DIY air capacitor.
This is produced from two 3030 cm
(C) Segment
980039-1
Figure 3. Component
layout of the single-
sided printed-circuit
board for the E-meter.
Miscellaneous:
BT1 = 9-V battery with terminal clips
S1 = single-pole on/off switch
Enclosure = Bopla E430 (from
Phoenix Mecano Ltd, phone 01296
398 855) or similar
Aluminium sheet, 1 mm thick, 2 off
300300 mm
4 off M4 nylon screw, 15 mm long
4 off M4 nylon nut
1 off M3 metal screw, 15 mm long
2 off M3 metal screw, 10 mm long
4 off M3 metal screw, 20 mm long
(for board mounting)
4 off spacers, 15 mm long (for board
mounting)
5 off M3 nut, nylon or metal
5 off nylon spacer, 10 mm long (for
sensor)
2 off solder tag, 3 mm
Insulated circuit wire as required.
PCB Order no. 980039 (see Read-
ers services towards the end of
this issue)
Testing, don't
To be able to apply the information remove IC !
D6 lights
presented in this instalment, you have
to have a working IC tester available,
together with all the files found on the
OKAY not OKAY
diskette with order number 986014-1 or
CD-ROM 986001-1 (see further on).
The tester is very likely to function
ICCL : x mA Test failed:
properly if the message
ICCH : x mA Ent: rep Esc: exit
IC Tester ent = repeat test
1:Test esc
any key
PINORDER pin_order
Library of IC found is shown 980029 - 2 - 12
Links the individual IC pins to the
columns that supply the test vectors.
PINORDER may only follow PINS, and
the pins are identified using their pin This defines the function of each indi- separated by spaces (not obligatory,
numbers. Individual pin numbers vidual pin. The following functions are though), and all pins must be included
should be separated by a space char- available: in pin_definition. Pins which should
acter. All IC pins have to be identified O output not be tested are defined as outputs
in pin_order , including pins which I input (O), and not tested in the VECT line
are not tested. G ground pin (X). PINDEF should precede the first
V Vcc (+supply) pin VECT. After a VECT line, the pin func-
PINDEF pin_definition The individual pin functions may be tions may be redefined using PINDEF.
ICCL
When this instruction is encountered,
test OKAY test not OKAY the regulated power supply measures
the IC supply current. After the test,
the supply current is indicated as
Test ok Test failed! ICCL. ICCL may only be used once for
Ent: rep Esc: exit Ent: rep Esc: exit any one IC.
ent = repeat test ent = repeat test
esc esc ICCH
980029 - 2 - 13
When this instruction is encountered,
the regulated power supply measures
VECT test_vector Z test if pin is at high impedance. the IC supply current. After the test,
A test vector may consist of the fol- X do not test this pin. the supply current is indicated as
lowing elements: ICCH. ICCH may only be used once for
1 output pin: check if pin is at 1; The individual elements may be sep- any one IC.
input pin: apply 1 to pin. arated by spaces (not obligatory). If a
0 output pin: check if pin is at 0; pin is defined as GND, it should have The following points should be noted
input pin: apply 0 to pin. a 0 at the relevant position in the test when writing your own test vectors
for ICs not included in the default
library.
IC Tester Normal keyboard functions:
4 4:Trace
w All IC inputs have to be made logic 0
up/dn = scroll up/down and logic 1 at least once, in a manner
ent = accept next menu
esc = return to main menu that ensures that this change can be
detected on at least one IC output.
Select Library up/dn
ent
w All IC outputs have to go 0 and 1 at
74 x x x esc least once during the test, and also Z
(high-impedance or tri-state) when an
output can assume this state.
Select IC: up/dn w With ICs having a sequential internal
ent
74 : 00 esc circuit, IC inputs driving the clock
input of a register flip-flop may not
change state in unison with inputs of
Insert IC:
these flip-flops, if this transition equals
Press Enter
ent the active edge of the clock input. For
esc
example, the clock input of a 74:74
may not change from 0 to 1 when the
pin 1 = lower left-hand position level at the data input changes at the
pin 14 = upper left-hand position ent, up/dn = next test vector
V = Vcc pin V11 0011 00 esc same time. This proviso also applies to
G = GND pin 01 1111 00G dn = skip repeat loop (if used)
clock and enable inputs of synchro-
0 = IC input at 0
1 = IC input at 1 nous counters.
Z = IC output at high-Z (3-state)
0
0
0 = IC output at 0, should be 0 D6 lights w If an enable input of a latching flip-
1 = IC output at 1, should be 0
0
Z = IC output at Z, should be 0
flop toggles, the level at
1
1
0 = IC output at 0, should be 1
Figure 4. Structure of the data input of the flip-
1 = IC output at 1, should be 1 measure ICCH, ICCL,
1
Z = IC output at Z, should be 1 end of test Menu option 4: Trace. flop is not allowed to tog-
ERR = error in output state
gle at the same time. This
applies to rising as well as
Ent: Restart ent = repeat test falling pulse edges.
Esc: Exit esc
980029 - 2 - 14 To close off this section, Figure 9 shows
Waiting for Host esc Any errors occurring during the com-
Esc: Cancel pilation process are recorded in
Figure 8. Structure of ERR.OUT only. They do not appear on
Menu option 8: the PC screen.
sync detected on no Remote mode.
RS232 i/f ?
TVCHK.EXE
This is a kind of shell program that
yes launches ICTVC.EXE, and enables test
Synchronized to host vector scripts to be debugged. For this
program to operate you have to con-
* * Connected * * esc nect the IC Tester to your PC via the 3-
Esc: Exit wire serial link. TVCHK should be
980029 - 2 - 18 launched with an appended parame-
ter which is either the COM port num-
ber (1-4), or the COM port address (in
name 4040
hex) followed by the associated inter-
9 ; 12-bit CMOS ripple carry counter
pins 16 rupt line (1-7). Example: TVCHK 2.
pinorder 10 11 1 15 14 12 13 4 2 3 5 6 7 9 8 16 Obviously, we are talking of the COM
; /cp mr q11 q10 q9 q8 q7 q6 q5 q4 q3 q2 q1 q0 gnd vcc port to which the IC Tester is con-
pindef I I O O O O O O O O O O O O G V
vect 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 nected! A screendump illustrating
vect 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 some of the options offered by TVCHK
vect 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 is shown in Figure 9.
vect 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1
vect 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1
EDT.BAT
vect 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1
vect 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 This extremely small batch file is used
vect 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 to launch the word processor you will
vect 1 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 be using to load, modify and save test
vect 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1
repeat 4 vector files. If you do not want to
vect 0 0 0 0 0 0 0 0 0 0 0 X X X 0 1 employ EDIT.COM, change EDT.BAT as
vect 1 0 0 0 0 0 0 0 0 0 0 X X X 0 1 required to make it point to your
endr favourite DOS text editor. Type the
vect 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1
repeat 8 name of the file you want to process
vect 0 0 0 0 0 0 0 0 0 0 X X X X 0 1 after EDT. In case VECT.TVC is too
vect 1 0 0 0 0 0 0 0 0 0 X X X X 0 1 large for your wordprocessor, consider
endr creating the part to be appended as a
vect 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1
repeat 16
separate file. Once the new test-vector
vect 0 0 0 0 0 0 0 0 0 X X X X X 0 1
vect 1 0 0 0 0 0 0 0 0 X X X X X 0 1
endr
vect 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1
Figure 9. Example
repeat 32 illustrating the opera-
vect 0 0 0 0 0 0 0 0 X X X X X X 0 1 tion of a test vector
vect 1 0 0 0 0 0 0 0 X X X X X X 0 1 script. The diagram
endr and the script are all
vect 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1
repeat 64
you need to put the
vect 0 0 0 0 0 0 0 X X X X X X X 0 1 4040 12-bit counter
vect 1 0 0 0 0 0 0 X X X X X X X 0 1 through its paces.
endr
vect 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1
repeat 128
vect 0 0 0 0 0 0 X X X X X X X X 0 1
vect 1 0 0 0 0 0 X X X X X X X X 0 1 CD4040
endr
vect 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 Q11 1 V 16
DD
repeat 256
Q11
vect 0 0 0 0 0 X X X X X X X X X 0 1 Q5 2 Q5 Q10 15 Q10
vect 1 0 0 0 0 X X X X X X X X X 0 1
endr Q4 3 Q4 Q9 14 Q9
vect 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1
repeat 512 Q6 4 Q6 Q7 13 Q7
vect 0 0 0 0 X X X X X X X X X X 0 1
vect 1 0 0 0 X X X X X X X X X X 0 1 Q3 5 Q3 Q8 12 Q8
endr
vect 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 Q2 6 Q2 R 11 RESET
repeat 1024
vect 0 0 0 X X X X X X X X X X X 0 1 Q1 7 Q1 CLK 10 CLOCK
Q0
vect 1 0 0 X X X X X X X X X X X 0 1
endr 8 GND 9 Q0
vect 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1
vect 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1
vect 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 980029 - 2 - 19
iccl ; measure IC current
icch ; measure IC current
1
or its consultants.
received from
information received
Electronics or
experience by Elektor Electronics
electronics industries or their
on information
not imply
A National Semiconductor
Application
The
not
in
for easy decoding to address 80h nal that indicates to the LM78 that it ISA is active will not be a problem,
and 84h, the normal addresses for may drive data on to the ISA data bus. since even a single bit of Serial Bus
outputting of POST codes. Inter- IOWR# is the standard ISA com- communication requires 10 microsec-
rupt will be set only when writing mand to the LM78 that it may latch onds, in comparison to less than a
to port x0h or x4h is taking place. data from the ISA Bus. microsecond for an entire ISA com-
The register can be read via ports SYSCLK is the standard ISA system munication.
85h and 86h. clock, typically 8.33 MHz. This clock is
Value RAM containing the monitoring used only for timing of the ISA inter- TYPICAL APPLICATION
results: temperature, voltages, fan face. All other clock functions within The setup of a typical application is
counts, and WATCHDOG limits. the LM78, such as the ADC and fan shown in Figure 4. It monitors tem-
This register consists of a total of 64 counters, are done with a separate perature, the speed of three fans, and
bytes. The first 11 bytes are all of asynchronous internal clock. seven power supply voltages. The
the results, the next 19 bytes are the A typical application designed to chassis intrusion detector is based on
WATCHDOG limits, and are use the POST RAM would decode the a photodiode and discrete bistable.
located at 20h3Fh, including two LM78 to the address space starting at When the chassis has been opened
unused bytes in the upper loca- 80h, which is where POST codes are
tions. The next 32 bytes, located at output to. Otherwise, the LM78 can be Figure 4. In this application, the
60h7Fh, mirror the first 32 bytes decoded into a different desired LM78 monitors temperature, fan
with identical contents. The only address space. speed of three fans, and seven
difference in the upper bytes is that To communicate with an LM78 reg- power supply voltages. It also mon-
they auto-increment the LM78 ister, first write the address of that reg- itors the O.S. output of up to eight
Internal Address Register when ister to Port x5h. Read or write data LM75 digital temperature sensors
read from, or written to, via the ISA as well as an optical chassis intru-
Bus (auto-increment is not available sion detector.
for Serial Bus communications).
4
When the LM78 is started, it cycles
through each measurement in
sequence, and it continuously loops
through the sequence about once
every second. Each measured value is
compared to values stored in WATCH-
DOG or Limit registers.
When the measured value violates
the programmed limit, the LM78 will
set a corresponding interrupt in the
Interrupt Status. Two hardware inter-
rupt lines, SMI# (System Management
Interrupt) and NMI/IRQ# (Non-mask-
able interrupt/interrupt request), are
fully programmable with separate
masking of each interrupt source, and
masking of each output. In addition,
the Configuration Register has control
bits to enable or disable the hardware
interrupts.
INTERFACE
The LM78 decodes only the three low- from or to that register via Port x6h. A (intruded), the relevant data are stored
est address bits on the ISA Bus, result- write will take IOWR# low, while a thanks to a battery backup. The
ing in the following base addresses. read will take IORD# low. bistable may be reset via the LM78.
If the Serial Bus interface and ISA A link with the System Manage-
Port x0h: POST codes from the ISA Bus interface are used simultaneously, ment Serial Bus of the PC may be
Bus. there is the possibility of collision. To established via the I2C Bus.
Port x4h: POST codes from the ISA prevent this from occurring in appli- A software-controlled 5 V power
Bus. cations where both interfaces are used, supply is provided with the aid of a
Port x5h: The LM78 Internal Address read Port x5h and if the Most Signifi- MOSFET. This supplies the entire cir-
Register. cant Bit (MSB), D7, is high, ISA com- cuit, which draws a current of only
Port x6h: Data Register. munication is limited to reading Port about 1 mA during normal operation,
x5h only until this bit is low. A Serial and about 10 A in the standby mode.
IORD# is the standard ISA Bus sig- Bus communication occurring while [980009]
F I R F I LT E R S *
The program FIRFIL1 is capable of
much more than the generation of
echoes. As an experiment, generate a Figure 24. Impulse response of a
sweep signal and pass it through the 24 band-pass filter.
filter defined in ECHO3.SPP. This results
in 256 echoes with rather unusual val-
ues as may be seen on the oscillogram
in Figure 23. This clearly indicates that
the sweep signal is passed over a cer-
tain frequency range only: that is, the
filter is a band-pass section.
Unfortunately, it is not easy to aver-
age the sound intensity of 256 echoes
in a manner that results in a band-pass
function. Before the actual technique 980015 - 4 - 14
is discussed, we will take a look at how
finite impulse response (FIR) filters*
function. y0 = a0x0+a1x1+a2x2+
A FIR filter that generates echoes yk = a0xk+a1xk1+a2xk2+
with time delays 0, 1, 2, and sound Only the first term remains, so that
intensities a0, a1, a2, is described by where xk and yk are the input and out- y0 = a0.
the equation put values respectively.
To find the impulse response of the When k = 1,
filter, that is, the value of yk when x0=1
* Also known as non-recursive filters or and all other values xj=0, we must first y1 = a0x1+a1x0+a2x1+
transversal filters. calculate the value of
F I LT E R D E S I G N
10
Since the frequency response of a fil-
ter may be determined by a DFT of its 0
impulse response, the question arises 0 1990 2000 5000 5200 8000 8200
whether the impulse response can be f (Hz) 980015 - 19
derived from the frequency response
by an Inverse Discrete Fourier Trans-
form (IDFT), to which the answer is:
yes. Moreover, when the impulse 30 Figure 30. Screen
response is known, the filter may be images of the filter
simulated with the aid of program FIR- design program.
FIL1.EXE. Filters in digital signal proces-
sors are frequently designed in this
way.
The design of FIR filters is simpli-
fied by program SPECFIL1.EXE. The oper-
ation is shown diagrammatically in the
lower row of Figure 27. From the fre-
quency response of a filter, its impulse
response is determined by an IDFT,
which is modified by a window func-
tion (see box). This enables the prop-
erties of the filter to be fine-tuned. The
impulse response of the filter is then
written into a file, which processes the
FIR program as a data input. A file
(XFILDES.1.SPP) to compute a simple fil-
ter is shown in Figure 28.
The filter specification, the sam-
pling rate, the number of samples in
the impulse response, the parameters
of the window function, and the value 980015 - 4 - 20
NiCd
Dear EditorI am looking for a PCB er, when I saw the output config-
design system that allows me to uration: unbalanced! I had already
first simulate the electrical opera- omitted the unbalanced input from
tion of a circuit and then use turn the A/D conver ter circuit. How 987010 - 11
the approved circuit diagram into about doing a microphone pre-
PCB artwork. Having read all amplifier with a balanced input to
reviews and adver ts in Elektor match the above mentioned con-
14V...17V
Electronics, I am none the wiser. verter?
9V...15V
Therefore, I am curious to know B. Gillisberger (by email) 12V B80
what software is being used in the C1000
Elektor Electronics design labora- Were pretty sure that the micro- 4x 1N4001
tory. phone preamplifier meets your 220
T. Will (by email) demandsit just lacks a bal- 0W25 9V6...11V2
anced output. Enter our
The PCB design process adopt- SSM412-based converter from 1N4148 NiCd
ed by ourselves is by no means the article Balanced/Unbalanced
1k
IC Tester AD557
329-4700, fax (617) 326-8703. tling to 12 LSB for a full-scale step in 800 ns.
name 4001 Internet: www.analog.com. The AD557 is available in two package configura-
pins 14
tions. The AD557JN is packaged in a 16-pin plastic,
pinorder 1 2 3 6 5 4 13 12 11 8 9 10 7 14
pindef I I O I I O I I O I I O G V Features 0.3-wide DIP. For surface mount applications, the
vect 0 0 1 1 1 0 1 1 0 1 1 0 0 1 4001 Complete 8-Bit DAC AD557JP is packaged in a 20-pin JEDEC standard
vect 1 0 0 1 1 0 1 1 0 1 1 0 0 1 1A 1 14 Voltage Output: 0 V to 2.56 V PLCC. Both versions are specified over the operating
vect 0 1 0 1 1 0 1 1 0 1 1 0 0 1 Internal Precision Band-Gap Reference temperature range of 0C to +70C.
1B 2 13 4A
vect 1 1 0 0 0 1 1 1 0 1 1 0 0 1
vect 1 1 0 0 1 0 1 1 0 1 1 0 0 1 1Y 3 12 4B
Single-Supply Operation: +5 V (10%)
vect 1 1 0 1 0 0 1 1 0 1 1 0 0 1 Full Microprocessor Interface Application Example
2Y 4 11 4Y
vect 1 1 0 1 1 0 0 0 1 1 1 0 0 1 Fast: 1 s Voltage Settling to 12 LSB PC-aided BJT transistor tester revisited, Elektor Elec-
vect 1 1 0 1 1 0 0 1 0 1 1 0 0 1 2A 5 10 3Y
Low Power: 75 mW tronics April 1998.
vect 1 1 0 1 1 0 1 0 0 1 1 0 0 1
vect 1 1 0 1 1 0 1 1 0 0 0 1 0 1
2B 6 9 3A No User Trims Required
vect 1 1 0 1 1 0 1 1 0 0 1 0 0 1 7 8 3B Guaranteed Monotonic Over Temperature
vect 1 1 0 1 1 0 1 1 0 1 0 0 0 1 983004 - 13 All Errors Specified TMIN to TMAX
vect 1 1 0 1 1 0 1 1 0 1 1 0 0 1 Small 16-Pin DIP or 20-Pin PLCC Package
iccl
Low Cost
vect 0 0 1 0 0 1 0 0 1 0 0 1 0 1
icch
General description
The AD557 DACPORT is a complete voltage-out-
put 8-bit digital-to-analogue converter, including out-
name 4002
pins 14 put amplifier, full microprocessor interface and preci-
pinorder 2 3 4 5 1 12 11 10 9 13 7 14 6 8 sion voltage reference on a single monolithic chip.
61
4/98
Integrated circuits Elektor Electronics
Special Applications DATASHEET 4/98 March & April 1998 DATASHEET 4/98
name 4009*
Electrical Specifications (@ TA = +25C, VCC = +5 V unless otherwise noted) pins 16
Model Min Typ Max Units pinorder 3 2 5 4 7 6 9 10 11 12 14 15 8 11 16 13 4009
pindef I O I O I O I O I O I O G I V O 1 VCC VDD 16
RESOLUTION 8 Bits vect 1 0 1 0 1 0 1 0 1 0 1 0 0 1 1 X 1Y 2 15 6Y
vect 0 1 1 0 1 0 1 0 1 0 1 0 0 1 1 X
Elektor Electronics
RELATIVE ACCURACY1 0 to + 70C 1/2 1 LSB 1A 3 14 6A
vect 1 0 0 1 1 0 1 0 1 0 1 0 0 1 1 X 2Y 4 13 NC
OUTPUT Ranges 0 to + 2.56 V vect 1 0 1 0 0 1 1 0 1 0 1 0 0 1 1 X 2A 5 12 5Y
Current Source +5 mA vect 1 0 1 0 1 0 0 1 1 0 1 0 0 1 1 X
3Y 6 11 5A
Sink Internal Passive vect 1 0 1 0 1 0 1 0 0 1 1 0 0 1 1 X
Pull-Down to Ground2 vect 1 0 1 0 1 0 1 0 1 0 0 1 0 1 1 X
3A 7 10 4Y
8 9 4A
OUTPUT SETTLING TIME3 0.8 1.5 s vect 1 0 1 0 1 0 1 0 1 0 1 0 0 1 1 X
iccl 983004 - 15
FULL-SCALE ACCURACY4 @ +25C 1.5 2.5 LSB
vect 0 1 0 1 0 1 0 1 0 1 0 1 0 1 1 X
TMIN to TMAX 2.5 4.0 LSB
icch
ZERO ERROR @ +25C 1 LSB
TMIN to TMAX 3 LSB name 4010*
MONOTONICITY5 TMIN to TMAX Guaranteed pins 16
pinorder 3 2 5 4 7 6 9 10 11 12 14 15 8 1 16 13
DIGITAL INPUTS TMIN to TMAX 4010
pindef I O I O I O I O I O I O G I V O
Input Current 100 A vect 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 X 1 VCC VDD 16
62
1A 1 14
vect 1 1 0 0 0 1 1 1 0 1 1 0 0 1
OPERATING TEMPERATURE RANGE 0 +70 C vect 1 1 0 0 1 1 1 1 0 1 1 0 0 1
1B 2 13 4A
1 1Y 3 12 4B
Relative Accuracy is defined as the deviation of the code transition points from the ideal transfer point on a straight line from the zero vect 1 1 0 1 0 1 1 1 0 1 1 0 0 1
the the full scale of the device. 2Y 4 11 4Y
vect 1 1 0 1 1 0 0 0 1 1 1 0 0 1
2 Passive pull-down resistance is 2 k.
vect 1 1 0 1 1 0 0 1 1 1 1 0 0 1 2A 5 10 3Y
3 Settling time is specified for a positive-going full-scale step to 12 LSB. Negative-going steps to zero are slower, but can be improved
vect 1 1 0 1 1 0 1 0 1 1 1 0 0 1 2B 6 9 3A
with an external pull-down. vect 1 1 0 1 1 0 1 1 0 0 0 1 0 1 7 8 3B
4 The full-scale output voltage is 2.55 V and is guaranteed with a +5 V supply.
vect 1 1 0 1 1 0 1 1 0 0 1 1 0 1 983004 - 17
5 A monotonic converter has a maximum differential linearity error of 1 LSB.
vect 1 1 0 1 1 0 1 1 0 1 0 1 0 1
Specifications subject to change without notice.
vect 1 1 0 1 1 0 1 1 0 1 1 0 0 1
iccl
vect 1 0 1 1 0 1 1 0 1 1 0 1 0 1
icch
#
electronics on-line
Radio stations
Free radio! on the Internet
The Radio Carolines which some
of you may fondly remember have
been towed ashore and super-
seded by radio stations not using
transmitters at all. The Internet has
been discovered as a replacement
for those good old radio waves.
From now on, free radio will reach
you by telephone line.
These days you do not to take for granted that each CD is pre-
need to have a general cov- ceded by a commercial. Likewise you
erage receiver anymore to can pick a book title, sit back and lis-
be able to listen to radio sta- ten to the book being read to you. The
tions from distant shores service called AudioBooks at
and exotic regions. The http://www.audionet.com/#audiobooks
modern equivalent of the has a number of titles available.
short-wave receiver consists
of an Internet browser, an SCANNER FANS
Internet connection (via an The above mentioned web site also
Internet Service Provider) offers some very special services. For
and the so-called RealAudio example, AudioNet allows you to lis-
plug-in module. ten in on certain non-broadcast radio
communication channels. For example,
An interesting web site the site
demonstrating the many http://www.policescanner.com
possibilities of webbed radio allows you to eavesdrop on the mobile
may be found at radio network used by the police in
http://www.audionet.com. Dallas, New York, Los Angeles or
Audionet calls itself The Plano.
Broadcast Network on the Aeroplane spotters are probably
Internet. Its web site pro- more interested in the site
vides access to a wide vari- http://www.audionet.com/simufite
ety of radio programmes which provides a live feed to the Dal-
and radio services. Using las/Fort Worth Traffic Control Tower. In
various menus you may this way, you can listen to communi-
choose from sports reports, cation between Traffic Control and
news, lectures, shows and pilots.
concerts.
Thats not all, however. REALAUDIO
The web site also offers If you want to listen to these webbed
exclusive services to listen- Free Radio stations, you need to
ers, like the spoken book extend the functionality of your
and compact disc (CD) lis- Netscape or Microsoft Internet
tening. Pick a CD title from explorer with the Real Audio plug-in.
the stock list, and listen to it This program, RealPlayer 5, may be
via the Internet. Because of downloaded free of charge from
the American principle http://www.realaudio.com
There Aint No Such Thing after supplying your name and e-mail
As Free Lunch address.
(TANSTAFL), you will have (985029-1)