Documentos de Académico
Documentos de Profesional
Documentos de Cultura
Experimenter
7
- C.Ul t1
-~
~~'L--
IOtt
i
I OMl3 Pl._I
TO 100CIOP1NI
IC Packaging
Integrated circuits can be found in a
variety of packages. At the present time,
there are more than 120 case ty pes made
by some 70 compa nies aro und the world.
Fig . 2. The split supply con tain four series Of these ma ny case styles, three types a re
barreries grounded at the connect point. Where dominant. (In te rms of qu antity of devices
possible 10 use, the single supply offers the on the ma rket , in a given case ty pe, a bout
advantage of simplicity .
90% o f t heir q uantity can be found in
some variation of one of these three ca se
types). As ye t , no definite standardizat ion
Reliability has been set up among th e ma nufa c turers
Many manufac turers arc turning lo !Cs regarding packaging, pin numbers, and lo-
because of their high reliability. Devices cations, so carefully check the basing be-
built under almost clinica l conditions are fore you plug that IC into the socket.
bo und to be better than a circuit bu iJt on a T hree popular case sty les are pictures in
workbench. As an example, conside r b uild- Fig. 1.
ing the clectricaequivale nt to the l! EP 583, Use a low-wat tage solde ring iro n! 25 to
using the 2 1 transistors and 27 resisto rs: It 40 watts is a good range . Excess heat could
would be necessary to make 80 to 90 "kill" the IC.
solder connectio ns. a real source for poten- Keep com ponent lead s short! Excess
tial tro uble. lead le ngth could cause spurious o r para-
In addition to the adva ntages listed ,
sitic oscillatio ns o r no opera tio n at a ll.
re placement is simple. Schematics a rc easy
If you arc using a power suppl y (o ther
to read , especiall y for the beginner. Areas
than a battery), it is a good idea to bypass
yet to be co nquered in the constru ction of
the power leads. Connect a 0.05 or 0.1
ICs arc: !lo w lo built indu ctors, large-va lue
capacitor from the p ower input to ground
capacitors, and high-va lue resista nce o n an
a t or near the input term inal of the IC.
IC chip. It is presently necessa ry lo con-
nect these components exte rnally.
!Cs can be mounted o n perforated
board or printed-circuit board by either
D
OR GATE
D
ANO GATE
C>
AMPLIFIER
D
FLIP FLOP
8
~ NANO
G"l GATE
lmott~torm l
:~'
:~:
SCT ~aOUTPUT
CLOCK T
CLEAR C <l I NVE RSE
OUTPU T
TYPICAL BUFFER WITH GATED INPUT TYPICA L .lK FLIP FLOP
GATES CONN ECTED AS A
BISTA BLE FLIP FLOP
9
Adder - provides the .wmm a nd carry Occasionally the lime arises when the
operations on two inpu t signa ls. hobbyist needs some thing in the way of
Shift Regi s ter s - providesbistable ga tes other than what he has or what is
storage. available. For example, you need a 3-input
Flip-nop - provides division or count. gate and you have a 4 -input gate; simply
One flip-fl o p divides by 2. provides ground one input. Ground two inputs to
o ne o utput change in state for every obtain a 2-input gate. If you have a dual
two input charges. 2-input gate, such as the HEP 580, and you
Gates. This function comes in a wide need a 4-input gate, tie pins 6 and 7
variety of configurations. There are 2, 3, 4, toge ther and this becomes the o utput:
or more inputs and 4 categories as fo l- inputs are then on pins I , 2, 3, and 5.
lows: Ampli fier s. In digital work it is referred
And: When all inputs go to I , output lo as a buffer. Its or iginal use is to in crease
will go to I . "fan-in" or "fan-out" capability; that is,
Na11d: Output will be I except when the num ber of other units that ca n be
all inputs go to I . con nected in parallel to the input (fan-in)
Or: When any input goes to I. output or ou tput (fan-out). By adding proper
will be I . external biasing it is possible to connect
Nor: Out put will be I except when this unit to linear (audio-rf) usage.
any input goes to I . Flip-Flop. There are a number of types
Nund and nor differ from and and or in of flip-flops available. As mentione d pre-
that inversion has take n place. Refer to viously, a flip-flop (multivibrator) can be
Fig. 4. Note the sma ll o at the input or "made up" by cross connecting two ga tes.
out put of some of the examples. This o The R-S flip-flop is one example. The J-K
indicates that inversion has taken place. flip-nop is similar but has the added
Ga tes can be connected to opera te in a function known as "clock in pu t" shown as
wide variety of applications other than "T" on the logic block of Fig. 5.
those for which they were designed. Some Fundamentally, flip-flops divide by two.
applicatio ns are free-running mult ivibra- By proper connection, division by 3. 4, 5,
tors, bistable. one-shot. amp lifiers, and etc. can be obtained using a few !Cs as
audio mix ers. shown in Fig. 6.
OUT
$ 1
~
, -
T T
0 c 0
- -DIVIDE av 4
10
IC Receiver Accessory
Audio frequer.cy integrated circuits pro impedance headphone jack and function as a
vide the opportunity to develop very useful self-contained unit to provide tunable af
circuit functions- in many instances at a far selectivity, age and about 1 watt of power
lower cost than would be possible using for direct lo udspeaker operation. By the
discrete components. On the o ther hand, the addition of a few d iodes a simple noise
inherent nature of the IC housing and the limiter can also be added. For compactness
conditions under which it must operate also and simplicity, only a resistQr-capacitor net-
make the practical realization of a circuit work was used without any inductive
using !C's, in some instances, just as compli- elements to obtain audio selectivity. The
cated as a circuit using discrete components. resultant selectivity is not as sharp as that
This seeming contradiction was experienced provided by a bulky inductor but is quite
by the author in developing the receiver usable on CW, especially with a transceiver
accessory unit described in this article. So, that already contains a steep-skirted crystal
even if one has no immediate need to build or mechanical SSB filter. The fact that the af
the unit described, it may prove interesting selectivity is tunable also adds to its useful
to still read about and appreciate some of ness. The age feature is not absolutely
the considerations involved in the use of necessary but was added since many trans-
audio integrated circuits. ceivers, although their ave cannot be
General disabled as such, do not provide full age on
The accessory unit described was CW when the rf gain control is at some
designed as a compact, solid-state unit that intermediate setting, as would be normal if
could be plugged into a medium to high one were using the rf gain on CW as the
"volume" control while tuning. The audio
output amplifier was included to eliminate
Power amplfflor I C mounted In hoat sink is the need for going back into the receiver in
shown on l eft. FET and anoclated com order to use its audio output stage. Thus, no
ponanu aro at upper right sldo of heat.sink.
Other compo nonts aro conveniently
modification whatever is required to the
arranged around duol 1tago FET. Poton receiver and the accessory unit can simply be
tiomoters shown on right aro mlniaturo type unplugged from the headphone jack when it
mounted directly on por foroted board
stock. is not desired for operation.
Circuit
A block diagram of the accessory unit is
shown in f ig. 1. The basic stage functions
are relatively simple. The incoming audio is
split into two parts, each going !h of the
HEP592 (a hobbyist version of the MCI 535
dual operational amplifier sold as a stereo
preamplifier). One section of the HEP592
serves as a level detector with an adjustable
threshold . When the positive going portion
of the input signal exceeds the threshold
level, an output voltage is produced which
when rectified is coupled to the gate of a
11
HEP 801
GAIN
CONTROL
THRES-1
HOLD
I
I
HEP80 I FET. The drain-source resistance of tunable audio filter which passes all audio
the FET is normally quite low but increases frequenc ies except the one to which it is set.
to several thousand ohms with increasing Thus, all other audio frequencies are fed
negative gate-source potential. This wide back to Jhe input to oppose wall input
resistance variation is used to regulate the frequencies except one frequency. The over
gain of the power amplifier IC. This "for all result is an audio peaking stage at the one
ward" method of obtaining age is somewhat frequency to which the audio filter is set.
different from the usual "backward" The power amplifier stage is necessary to
method where part of the final output af raise the output level to drive a small
voltage is rectified and then used to control loudspeaker and produces about 3/4 watt
the gain of a preceeding stage. However, it output.
works just as well and with some experimen The wiring diagram of the unit is shown
tation of the circuit time constants, it may in Fig. 2. A l OK potentiometer is used with
even appeal to some as providing more the age portion of the HEP 592 IC to set the
responsive age action than the "backward" threshold value at which age action starts. It
system. can either be brought out as a panel control
The other section of the HEP592 serves or left as a trimmer adjustment. The tunable
as a tunable af filter. Advantage is taken in af filter is of the bridgedT type. As shown, a
constructing the filter circuit of the fact that three-unit potentiometer is needed to cover
most "operationa l amplifier-turned the. complete audio range up to a few
hobbyist" IC's provide a differen tial input. thousand cycles. One could possibly make
That is, so-called inverting () and non only two of the resistor legs variable, but the
inverting ( +) inputs. A portion of the input frequency range will be restric ted to a few
signal is fed to the non-inverting input. The hundred cycles over which the network is
other input is coupled to the output via a effective. Fig. 3 A shows an alternative filter
12
network which can be connected between instead of dir ectly wiring it, for protective
terminals 7 and I 0 of the HEP 592 and purposes.
which requires o nly o ne potentiometer. It is The use o f the HEP 593 power amplifier
rated to be effective from 70 to 10,000 appears si mple from Fig. 2 and although its
cycles, but only the components specified connectio ns are not involved, its placement
should be used. in a unit does present some problems. A heat
Fig. 38 shows a simple no ise limiter sink must be used and Motorola specifically
which can be connected either before or advises against the simple slip-on fin-typt
after the HEP 592. Advantage is taken of head sinks available for T0-5 transistor
the fact that the JC units are operated with a cases. They suggest a 2 inch x 2 inch x I /8
dual-polarity power supply to allow biasing inch piece of aluminum with a center hole
of the diodes for symmetrical clipping. drilled to snugly fit over the transistor case.
Not having any material available, the author
used a heat sink found on a surplus IBM
computer board. The index tab was removed
fro m the HEP 593 and a clean contact area
between the flange of the IC and the heat
sink established. The IC was then glued into
place. Overall, the relatively low voltage gain
( 18-35) o f the IC power amplifier, its cost
and the necessity of using a heat sink does
not compare too favorably with using two
larger case audio transistors which do not
require heat sinks at the I watt level.
Two 9 volt transistor batteries were used
to power the unit shown. The battery
potential slightly exceeds the 8 volt oper-
ating voltage recommended for the units but
seems to cause no difficulty.
Adjustment & Operation
a,,
HlP &01
$ 0
There are no real adjustments that should
be necessary if proper layout procedures are
followed. Some experimentation of the feed-
~I( J;I !II( back resistors in the age amplifier circuit
may be necessary to obtain the best action
F lg. 2. Wiring diagram o f accesso ry unit.
and "bread-boarding" the c ircuit initially
Resistors are Y, watt. Triple section poton-
tlomotor used in audio filter is I RC typo will save time later. The lC's did show some
450502MD502M 0202. tendencies toward instability if too much
cou pling were allowed to exist between the
various input and output circuits. This is
Construction understandable when one realizes that
The photograph shows one construction although the lC's are called "audio" type,
layout possible for the unit. Basically, the their actual response extends up to several
circuit components are grouped around each hundred khz. So, an rf feedback loop can
JC and directly soldered together. The exist which will overload the units but yet
potentiometers for the age and af filter are not produce an audible indication.
mounted directly on the perforated board In operation, the receiver af o utput level
stock. The potentio meters shown are actual- is adjusted to produce adequate audio out-
ly of a miniature type but normal, less put, but not to the point of overloading the
expensive \4 inch shaft types are quite unit. The age threshold control is set as
adequate. Also, although not shown, an IC desired to produce the best age action when
socket should be used for the HEP592 unit, going from a weak to a strong signal.
13
(A)
330 K
)T+ll
w.
0 1 I .01 .I
l 0
20 K 10 K
IB)
-
IN270 ~ .. ~r IN270
1H~ ~
1
,----------7
I I
1/ I
5K 5K
)
-9
Fig. 3. Simpl er audio notch filter (A) r equiring only one potentiometer.
Cap acitors are Aerovox type P123ZNG. Symmetrical noise clipper (B) w hich
ca n be used at input to unit or a t input to power amplifier {p in 4) .
14
Build an 8 Transistor Code Oscillator
With Just One IC
15
C2 C3
SPI
SWl Bl R3
~1 11------------~
PARTS LIST:
ICl HEP 570 R3 10 K SWI SPST SPl 45 U
Al 3 .9 K Cl. 2 .05 F KYl MORSE KEY INTERCOM
R2 l OK POT C3 .001 F 01 3 voe 4" SPEAKER
AOOMl ft00M7
Fig. 2. If the distance between operating stations is held down to a reasonable value a CW intercom
setup can be made with two or more code oscillators. If you 've got kids in the family who can' t find
time to practice their code for the Novice exam, this arrangement is sure to turn the trick.
To couple a pair of code oscillators for has already been used to bring two new
two-way use, it is better to para llel the Novices to the bands. (Learning the code is
keys rather than the speakers themselves. fun if there is some incentive to study ; and
This approach keeps each unit from the a room-to-room CW intercom will work
labor of driving more than its fair share of wonders with harmonics who might
the load. Figure 2 illustrates an exce llent otherwise be reluctant to practice.)
room-to-room interconnect method that
16
Integrated Circuit Audio Filter
A vi ew of an encned WM3A flltet end one with part of the casi ng removed to show tho
internal hybrid inregreted circ uitr y. The encased filter measures only 8110" x 7/10" end
is about 3 / 16" t h ick. Photo courtesy Western Microwave. Los Gatos CA 95030.
17
,.,
IN 0--1
--~1,_.. _ _.1_..,00UT w..
,,
l J T
,...
also in feedback arrangements so the filter
c1rcuits are not loaded down. Unfortun-
ately, by the time you combine sufficient
discrete component re networks and tran- _,.
sistor amplifier stages to have the re filter
duplicate the performance of an inductor
....,., ..
_,tl(OUUICTOWJ
...
net work, the re filter can be as large and as
costly as the latter. The advent of
integrated circuits has changed all that,
however. High gain amplifier circuits and
multiple re networks can be incorporated
[OUT
Ti#
...,..
in one physically compact unit.
The block diagram of the integrated
circuit filter is shown in Fig. 2. Three
multi-transistor operational amplifiers and
the necessary re networks in a feedback
arrangement are combined in the hybrid
filter unit. Three external resistors arc used
and can be manipulated to change the
.. ...
- llC.OdlW;Yf*I
...
operational characteristics of the filter. One
can see some of the filter components in
......
ti1
M'VTI'
...
..
Fig. 2. Block diagram of the Western Micro
wave WM3 filter. ALL of the blocks shown
are contained In the slngle filter unit Fig. 3. The WM3 filter can be used alone
shown In the photograph. Each amplifier [(A) and (8)) to function as a low-pa11 or
is, In fact, a separate integrated circuit hugh-pau filter or in conjunction with a
amplifier. The number refer to the postemplifier ((C) and (0)) to produce e
terminal connections. steeper slope at the cutoff frequency.
18
the unencased view of the fil ter in the
photograph. The encased filter measures
0.8 inches x 0.65 inches x .15 inches thick.
The surface area is about that of a 25 cent
piece, and it is hardly any thicker.
Pe rforrnan c~
The filt er really begins to shine when
one investigates its performance possibil- ,_-:
ities. It can be used as a high-pass filter,
low-pass filter , peaking fi lter, notching -~~=-----.;=----"=
,.---..,!,,---o!
- fll( OU(lll(Y IM I
filte r, etc. The center freque ncy can be
adjusted as desired by an ex ternal potentio-
meter as we ll as the Q if desired. The unit
can also be set up so you can switch select
a variety of different filt er effect outputs.
Graphs portray the performance of this
type of fil ter best. Fig. 3 illustrates the
out put versus freq uency characteristic of
the filter in several low-pass and high-pass
circuits. The filter can be used alone for
these functions, or its o utput used to drive
another ic operational amplifier power
stage to furth er increase the slope of the Fig. 4 . The oxtremely sharp peek ing and
notching charac torl1tic1 of thQ f ilter make
frequency response at the cut-off fre- it ideal for use as an cw selectivity d evice.
quencies. The. peak ing and notching freque ncies can
be tuned from 500 to 1500 c y c les.
Fig. 4 shows the fil ter used as a peaking
or notching filter. Note the extreme
sharpness of the response at the nominal
center frequency of 1 khz. Actually, both As might be imagined from looking a t
the center frequency as well as the the arrangement of the external resistors
sharpness of the filter response can be and the output terminals used for each
tuned by making various of the external specific application of the filter, you can
resistors variable as shown in Fig. 5. Thus, devise various switching a rrangements to
the center frequency of a nominal 1khz select differen t outputs, different specific
filter can be tuned from about 500 to center freq uencies, etc. The possibilities in
1500 cycles. The Q can be varied from this direction are pretty well only limited
about 1 to a maximum o f 100. by your imagination . Fig. 6 shows one
... .
...,
( oCft l
-.-.-
..A.fl(.U\.A" lliOOOCL
- l h .. OO o.< ti1~
'
F ig. 5 . Tho abovo graph s indica te how by simpl e O><tel'nal pot ontiom oter
control, th o froQuoncy as woll a s th e a of the f ilter can be var ied.
19
Summary
The type of filter described is available
now from firms such as Western Micro-
wave. The price of such a filter - about $30
depending upon the type of casing used-
will restrict its use in amateur equipment
until greater sales will invariably bring the
price down to that of regular integrated
Fig. 6. The above d iagram illustratas how
the filter can be ul&d to simultaneously circuits. However, even though one may
produce different output characteri1tlc1. Rf not be using such a filter tomorrow, such
and RQ are chosen according to Fig. 5.
filters will be the type of component that
will become common in amateur equip-
simple circuit which provides simultaneous ment as the micro-miniaturization of com-
or switch selected different output possibil- ponents for use if communications circuits
ities. continue.
. .. W2EEY
Integrated Circuit 6 Meter Converter
At the present time, amateur 6-meter con- which is in the middle of the broadcast
verte~ are built around. nuvistor tubes or band, making the converter ideal for
field-effect transistors. The forme r requires mobile applications. If a 7 MHz difference
a b-supply in the o rder of 1OOV and frequency is desired , a 43.5" MHz crystal is
filament power; both types usually require used and the 3900 pf capacitor is the
neutralization, which is tedious ana re- mixer drain tank circuit is changed to I 00
quires additional co mponents. pF
The 6-meter converter described in tl1is
article uses an integrated circuit in the
front-end , eliminating the need for neutral-
ization and yielding far better gain charact-
eristics I han the nuvistor or all-FET
counterparts.
~I
12V input and output tank circuits for maxi
.. mum signal.
Construction Details
'"" The converter was built on a piece of
~.
~-
aluminum, with teflon sockets and termi
nals used for mounting components, al
~ i:.1~ though if built on a punched phenolic
board with standard terminals the circuit
..
INPUT will work as well. The only precautions to
observe are making the connections as
short as possible and to keep the coils well
separated from each other.
The toroidal coils are not visible in the
photographs because they are mounted
Fig. J. Schemaric under the trimmer capacitors.
The converter costs less than S20.00 in
detune the oscillator plate tank circuit slight parts and outperforms by far many of the
ly. Now tune the mixer drain tank circuit for commercial units that are presently avail
maximum signal, and tune the rf amplifier able for almost twice the cost.
22
Digital Counters
23
Disadvantages: Dimmest of all displays.
Small angle of viewability.
~
I 0
Projected Image
This system consists of 12 light bulbs, 0-9
7
"
~ and + and -, which arc projected on the back
of a ground glass screen. Various bulbs may
' IC
24
Fig. 2. Digital decoder # 1. Reset to 0 by RtSCT lO ZC.AO 8'I' Ml:JllilD(l..,._Y
P\.ACHG 3 IN ON ll5CT UC..
momentarily placing +3.6 v on reset line. All
AU. ~ <:F. $NLNll f\HCTk)i
resistors of a similar function are tho same ARC: ntC SAM( \IWL.
value.
25
Top view of digital decoder 112. At the top Top view of dig Ital decoder # 1 . At the top
are the IC and 2 transistor gates. In the are the 5 flip-flops. The 2nd row has the
middle are the other gates. At the bottom 9 00 driver at the left. and the other 5 1c1
are the drivers.. are the 914 1 1. The other transi11or1 are the
drivers. At the left is a d igital display which
con1l1u of 10-1149 light bulbs encapsulated
in plastic.
26
the wiring of the 914 decoding gater which
is the last to be wired and the most prone to
error.
As you can see in the parts list, there are
a few different ways to build this decoder. If
you plan on using the Nixie/Pixie type of
indicator , then you will need to use such
transistors as the Fairchild 2N3568 or others
Two Pixier indicators, mounted on some
keyboa rd.
with a 60V o r more collector-to-emitter
voltage rating. The 60V line and 10
resistors to the driver transistors is only
necessary when Nixie/Pixies are used also.
Fig. 3a shows how to hook up lamps to the
decoder, and 3b shows how to hook u p
Nix ie/Pixies.
Digital Decoder #2
This is the unit which I use in my
present digital counter. It has only 4 !Cs, so
cuts down on the cost of the unit, but uses
22 NPN computer transistors, which, while
cutting do wn on the cost, adds to the
complexity.
Parts used are as follows: 4 - JK flip-flops
(Fairchild 923 IC); 22 - NPN switching tran-
sistors (Surplus computer); 2 - 2.7K \12 watt
resistors; 5- 470-680D \12 watt resistors;
10- 4 . 7K-6.8K \12 watt resistors; I 0-1 K-2K
\12 watt resistors.
An edgel it indicator showing Internal con - Cost: Well under $10.
struction. Wiring time : About 6 hours.
CAI
Digital Decoder #1
This decoder needs 3 1-41 parts: 5- J K CBI
27
Fig. 4. Digital decoder 1 2. All rasinors of a AESCT TO 2DI> l't WCICHTWJ'
Pl..JCNG !.It ON lllDET U:
similar function are the samo value. To reset
AU. JEIS'l'altS ts .......,. n.N:'TOil
to O. place +3.6 v on resat line momentarily. AlllE.TH(SUC:~
~
t he binary counter, it would be a shame to
depend on the 60 Hz line freque ncy for a
timing standard for such an accurate instru-
ment as a digital counter. Therefore, I am
.... now using a 100 kHz crystal as the standard .
Fig. 7 shows that 1 am using one of J im
Fig. 5. A slmple coun t checker. Fisk's little crystal circuits into a 9 14 mono-
stable pulse shaper. This in turn drives 24
matrix to drive the 7 light bars. As it is
rather involved, I am not including it here,
but those interested may obtain the sche-
matic by sending me a SASE. The cost will
approach $ 15 to $25, depending on parts
availability , and the construction time will
be about 10 or more hours.
Count Checker
A simple cou nt checker which can be
used to see that everything is counting Fig. 6. Addition of an emitter follower to
correctly is shown in F ig. S. Before you original input circuit of tho binary counter.
28
frequency divide section which will give you
the desired accuracy, say 100 Hz. This will
require a divideby-1000 section , or 12
nip-nops. Now you can read out the fre
quency of that vfo to, say, 7034.8 Hz. How
about 20. 15 , and 10? Simple too! for 20
meters. add one more flip-flop to the I 000
UV
divider to get 50 Hz. This way, you get
Fig. 7. double the count, or 14069.6 Hz. For 15,
add a divideby-3 to the divide-by-1000 to
get 21,104.4 Hz. For 10 meters, add two
923 nip-Oops hooked up as 6 "divide by l O"
divide-by-2 for a divide-by-4. . Exciting?
frequency dividers. This results in a "divide
Certainly is! Accurate? Just as accurate as
by one million," giving me a final output
when you zero t hat l 00 kHz crystal with
frequency of one hertz in ten seconds. Using
WWV. Of course you don't have to read out
t h is to trigger the count controller and gate,
the whole number, but can read out the last
I take a count of all pulses in l 0 seconds,
3 or 4 o nly, since yo u should know your
thereby giving me an accuracy to one tenth frequency to within a megahertz. Where do
of a hertz. The decimal point is inserted as you do from here? Well , I've given you some
hertz. The d ecimal poin t is inserted as
ideas o n the fundamentals. The rest I leave
shown in the picture. Various other frequen to you.
cies are tapped off the divide by one million . .. WllLMD
set of nip-nops to ach ieve other values of
counts. This is described in my previous
article.
:::e.:~-.::~::-:::~.::.
literature.
::-.::o::o.:::
t/616E (General Ca1alog & Ap
plications.
... ~
Write: Burroughs Corp., Electronic Components
Div., Plainfield NJ 0706 l.
High Vacuum/Projected Image readouts: Request
information from :
Industrial Electronic Engineers, Inc., 7720
Lemona Ave., Yan Nuys CA 91405.
Top view of the 24 923'1 in the divide by
Fairchild IC's: Applications brief 36 on 9960
one million" section.
series
Circuit No tes RTL I through 5.
Flip.flops App 120/2, App-1 18/ 2 &
SL2l8
A number of people have inquired about Write: Fairchild Semiconductor, 313 Fairchild
other types of ICs which might be used. Dr., Mountain View CA.
Suppose that you wanted to count up higher Fairchild parts and information : Request "' Design
ing wi th Integrated Circuit Components"
in freq uency. To do this, you wou ld have to Write: Hyer Electronics Co., Denver Technolog
change the time constant of the input shaper ical Center, PO Box 22227, Denver CO 80222.
by pulling in a smaller condenser, and t hen General Digital information:
using h igher frequency ICs Say you wan ted Write: Interstate Electronics Corp. 707 E Yer
to read the o ut put of a 7 MHz vfo. Simple. mo nt Ave., Anaheim CA- Request Sl39A and
Updates. Digilal Equipt Corp., Technical Publica
Use Fairchild 926s (freq. to 8 MHz as t ions Dept., 146 Main St., Maynard MA
opposed to 2 Mil z of the 92 3). Make a 0 1754- Request " Logic Handbook.'"
29
Unique Digital TTY Accessories
J. A. Murphy, KSCBA
30
with a period of 22 ms. Nand gate U4 picks The first sample pulse also loads the
out every other clock pulse and drives start element into output flip-flop U I pin 5
inverter US input pin 3. The signal at U4 and causes the counter outputs, Ul pin 9 ,
pin 4 is the string of positive going sam- U2 pin S, and U2 pin 9 to step from all
pling pulses. The first of these pulses causes (Ill) , the reset condition, to ( Hit) . The
JK flip-flop U3 output pin 6, which has second sample pulse loads the first signal
been high until now, to go low. As long as element into the output flip-flop and steps
the K input, pin 3, of this flip-flop remains the counter to LHL. This process continues
low the following sample pulses will have for pulses three through six. At this time
no effect on its output. The low output at the last signal element has been lo(lded into
pin 6 goes to nor gate input U4 pin 2, the output flip-flop and the counter out-
causing the timing cycle to continue re- puts are LHH. This makes both U4 pin 9
gardless of the signal at the input. Note, and U4 pin IO high, so U4 pin 8 and US
however, that this happens only after 11 pin 9 go low. Inverter output US pin 8 JK
ms of continuous spacing signal. This flip-flop U3's K input, pin 3, go high. The
means that a spacing condition must exist qext sample pulse, which loads the stop
at the input for at least 11 ms to initiate a element into the output flip-flop , also
timing cycle. This provides protection causes JK flip-flop output U3 pin 6 to go
against noise on the signal line. high. This, together with the condition on
PROPER\.Y TIMED 5
DISTORTED 5
INTO REPEATER L_J lJ
O<ITl'\IT Of'----.
REPEATER
31
Ill Piii.
TO TO
uePtN
TO
U41'11111
TO
uel'Ul11
TO ue ll'IHI TO U8 l'IN I
TO ua PIN TO Ill l'IN 9
TOUIHll TOUZl'l'l8
TO ue Piii I TO UI l'IN I
OUTl'UT
up with marks so that 13.5 ms after the input and output leads of tl]e converter
last signal ele ment is sent to the printer the and switch the clock frequencies from 91
output goes lo a marking condition and Hz and 74.2 Hz to t 48.4 Hz and 45 .5 Hz.
stays there until the next character arrives. Of course, th e keyboard may only be
To convert the I 00 wpm o ut put of th e operatf:!d at typing speeds up to 60 wpm,
keyboard to a 60 wpm input for your but that shouldn 't cramp the ty ping style
keyer it is o nly necessary to reverse the of too many of us!
34
Understanding and Using
Integrated Circuits
Rbt. A. Hirschfield, W6DNS
35
tra nsistors are made today. Starling with
one N-lype collector region , many bases
and e mitters arc diffused , with the resul-
tant thousa nds of transistors, being cul
apart a fterwards and pul in ind ividual
packages.
Suppose we want to build o ther
A components - for example, a resistor -
using the same processir.g steps employed
in making transistors. Why no l use the
built-in resistivity of the chemically treated
silicon as a resistor? And why no t c ul apart
the tran sistors in inte rconnected groups,
rather than into individual pieces? One
problem exists: isolating one eleme nt fro m
another. Remember, in a slab , or "wafer ,"
of NPN transistors, all collectors arc in one
region, and are thus shorted together . The
B solution is a four th, P-type regio n, separ-
ating each transistor or resis tor region ( Fig.
3). If we bias this " isolating" region so that
it always forms a reverse-biased P-N junc-
tion with every thing it surrounds, ea ch
"component" in the IC will be effectively
isolated by a nonconduc ting, open diode.
We can nuw form a ny combination of
c transistors, resistors, and diodes we Like,
simply by choosing which "windows" we
open at each step in the fabrication pro-
..".. diffusion
cess .
What Are the Ad vantages o f ! Cs?
Economy - Since we must go through
the photochemical process to make tran-
sistors a nyway, il costs very little mo re to
make several of them; putting ma ny tran-
sistors in one package is more economical
than se parating them. In the long run, an
D IC gives a lower "cost per transistor " than
a discre te transistor circuit.
Ease of Use - Most of the signal pro-
cessing and de biasing is done within the
microcircuit , so less wiring is needed to
bring signals in a nd out of individual
transisto r stages, or to interconnec t bias
voltages.
Performance As we've seen, th ere 's
no thing magic abo u t the ind ivid ual transis-
tors used in an IC; if a nything, they .may
c.olltctor not be quite as good at high frequen cies as
E the ir discre te counte rpar ts, because of the
capacitance lo the .. isolat ion,. regio n . Their
Fig. 2. Construction of a silicon NPN rransisror real advan tage in performance is that we
(crossseclion view). are no lo nger limite d to design ing circuits
36
This photomicrograph of an IC rf amplifier shows NPN transistor elements (T). diodes (D ), resistors (R)
and lhe wire bonding pads (W).
to use a minimum number of expensive ponent match ing is guaranteed. For this
transistors; we can design a better circuit reason , integrated circuits are available
that uses four or five times as many which simply are impractical using single,
transistors, at less total cost than the hand-matched transistors and resistors.
dis c r ete circuit. Moreove r , interstage What Are the IC's Limitations?
coupling and capaci tan ce is much less of a La;ge Component Tolerances -
problem than with " discretes," because of manufactu rers, desp ite their most careful
the microscopic lead lengths between com- co ntrols, canno t prevent a wider variation
ponents. Finally, because all parts of the IC in resistance, transistor beta, e tc., than
are made simultaneously, by accurate would be allowed in conventional circuits.
photographic techniques, excellent com- A I kil resistor might well be 7 50 k.Q on
Fig. 3. Cross-section view of an IC, showing two NPN 1ransis1ors and a resis tor.
37
Similarly, inductance on a chip 1s
impractical, so that ICs designed fo r tuned
amplification generally require external
tuned circ uits.
While PNP transistors arc used, they are
not cap'1b lc o f nearly as good performance
as NPNs, since the chemical processing is
not oplimiz<!d for the m. Two ty pes of PNP
arc available: the /a1era/ PNP (Fig. 4) and
the subslrale PNP (Fig. 5). The lateral PNP
Fig. 4. Laieral PNP (cross sec!ion).
38
Medium-power NP N transistors arc now possibly des tructive currents. Finally, this
practical; they take up a great d ea l of chip seldom-mentioned junction can produce
area because o f the need to s pread th eir undesired su bstrate PNP action , o r even ac t
dissipation over a la rge r region . Their" col- as an SC R.
lector sa turation resistance is not as good Reca lling that o ne definitio n of a tran-
as very inexpe nsive " d iscre te" po wer tran- sistor is that it co n tains a fo rward-biased
sistors, because IC collector conta cts are and a reverse-biased ju nction , it can be seen
necessa rily made fro m above , while ordi- t hat if the collector-base j unctio n of the
na ry p ower transistors, which don't h ave NPN is inadvertently forward-biased , it will
isolat ion regions, have b o tto m collector for m the base - e mitter junction o f a sub-
conta cts, with a shorte r c u rr ent path, a nd strate PNP whether we like it o r not.
he nce lower resistan ce. Fo r most practical Moreover, this cond ition produces exactly
designs, an IC can handle power o utputs to the same trigger conditio n in t he four-layer
a watt or so. but must drive a n ex tern al NPNP IC struc t ure as is used lo turn on a
power transis to r to o btain power levels. four-layer SCR. Under some c irc ums ta nces,
T h is also keeps IC power dissipation with in this p roduces " latch up," w h ich ho ld s un til
the limitati o ns of its small p ackage. the su p ply is turned off, or draws en ough
Parasitics - Here is whe re the ac tual current to damage the IC. In properl y
characte ristics of an IC depart in unex- designed ICs, used according to manufac-
pected ways from those of the sa me circuit turers' recomme ndat ions, parasitic prob-
in discre te form. The c ulprit is t he P-type lems will not arise. Experimenters, using
isolation region , w h ich separates the nonstandard ex ternal hookups, ho wever ,
va rious IC e lements; it is a reve rse-biased may find "parasitics" t he explanatio n for
PN j u n ction (Fig .. 6), whic h is assu med to t he !C's u nex pec ted behavio r.
be an ope n circu it. Bu t suc h a ju nction
p ro duces a parasitic shunt capacitance Co nclusio n
from NPN collector to ground, wh ich can We've peered inside the marvelous,
de teriorate high-frequ ency performa nce of widely acclaimed IC, and found it to be a
the IC. While IC schematics rarely show logical exte nsion of known circuit design s
this PN junction, it must be reme mbered and tra nsistor fabricatio n techniques. It has
by the user who applies su pply voltage a few peculiarities, but offers conve nie n ce,
other tha n those recommended; since the performance, a nd economy in building all
subst ra te is co m mon ly connected to t he types of ha m equip me nt. A basic u nde r-
most negative point in th e circu it , normal sta nding o f w hat go es in to an IC, its
opera tio n d oes indeed keep it reverse- advan tages a nd inheren t limitatio ns, w ill
biased , but an attemp t to operate an N PN prepare us to look at available ICs, an d
collector more nega t ive than this voltage learn ho w to put them to work .
will fo rward-bias the junction, drawing . . . W6DNS
39
IC's for Amateur Use
arc good for computers on ly. 1/8 in. wide, and very thin. The particu lar
Integrated circuits are simp ly very small one I have in mind contains some 32
transist0rs, diodes, and resistors, con- transistors and is the great-grandchild of a
structed on tiny dice, sometimes called two-transisor flip -flop multivibrator, which
chips, of substrate material. can "stand on one leg o r the other but not
Capacitors and inductors arc no t gener- an both. " To make the pesky little thing
ally included in thr ;e devices; therefore, in fooiproof, they included a variety of "con-
communications systems the ex ternal com- stant-current transistors, lockup prevention
ponents are often many times larger than transistors, phase reverse rs, (which reverse
the IC itself. The benefits of ICs can be the "truth table"), and heave n knows what
considerable, however, as in an example else.
such as the HEP 590, which has h igh gain, I did detail all the fun ctions in there at
low internal feedback, and absence of one time, a few years ago, and believe it or
detuning effect over the entire ave range. no t the specs on this thing do include a
Comp uter !Cs truth table. They do work, even tho ugh
This is where the whole thing started , there is not one capacitor in there. Every-
and it was a fascinating story, fo r the thing is direct-coupled. With SV at several
science of electronics at least. In a com- hundred amps to tal on a computer power
puter there is a tremendous need for supply they're happy with I V in the off
hundreds, thousands, and millions of state and 4.SV in the on. They work fast ,
memor y cells, switches, gates, adders, shift too, like being clocked (pu lse driven) by
regis ter s, "scratch" pads, inverters, 2, 10 , or 20 MHz oscillators, o r even higher
counter- dividers, delay lines, digital-to- as each year goes by. Yes, this docs mean
analog converters, an alog-to-digital con- exactly what you're wondering about 2, 10 ,
verters, flip-flops, buffers, comparators, or 20 million pulses, or "clicks of the
parity generators, current drivers, differ- clock" per second.
ential and operational amplifiers, binary So the computer can wo rk fast - so
decoders, and a h c~t of other d evice fast even that they're now worrying about
types - all of which ar~ geared not fo r the the time it takes fo r a signal to travel over
ham but for the computer specialist. printed ribbo n con nectors fro m one tray to
In spite of the cumberso me logic that ano.ther. Sound famil iar? Li ke short leads
early Scientific America n articles declared at VHF and UHF?
40
MSI and LSI Shielding, or great care, or both , should be
Just another word to Jet you know that used at these points. They are;
the biis manufacturers are not content with I. Rf amplifier input -output.
just 30 transistors in an IC. No siree! 2. 1-f input-output, in particular from
Medium-scale integration, or MSI, which the diode region back to the front end -
is just a way station, puts a Jot more particularly touchy on certain i-1 narmonics.
than 30 devices - really invisible to the 3. Af input - output.
human eye now - in that little can. 4 . Overall feedback from the af output
Large-sca le integration (LS I) really gets to the front end, even speaker to antenna,
to be high-density. l have one with I 56 this being often just a loopstick and close by.
leads and 774 announced functions on a In the following rf-i-f IC example you
single chip. will see that certain things can be done o n
There are manufacturers who make !Cs those tiny chips which are very inter-
that amateurs ca n use, as well as some esting - not only for i-f work but for rf
transistors for TV front ends that look also.
great for VHF and UHF. Mo toro la 's HEP 590
Adding ano ther transistor to the design Mo torola has a very interesting JC
of a chip is a matter of around 3 or 4, device for high frequency amplification ,
maybe even less today. You draw up a set the HEP 590. Two outstanding advan tages
of masks and reduce them do wn pho to- can be noted: low internal feedba ck, even
graphically to where each indivicual tran- when using the maximum gain of over 30
sistor-to-be cannot be see n at all except dB , and large ave action without detuning
with a good microscope. of the circuits
Then these masks are used one after the It is packaged in a I 0-pin can some
other to diffuse various materials going 5/16 in. in diameter. The leacjs can be
onto the chip , such as properly doped soldered or inserted in a socket. The
silicon , aluminum, gold, et c. - and event- present suggested net price is $3.99, which
ually you have a wafer with fifty or a is quite low considering what it accom-
hundred, or some other larger number of plishes.
ICs on it. One transistor more o r less is How the HEP 590 Works
thu s on ly a matter of dividing the time Figure 2 shows the internal schematic.
involved in drawing it once - plus the When ave voltage is applied to the base of
engineering time of thinking about how t o Q2, Q3 is turned off and the ac gain will be
do it, and how to test it afterwards. This is at a minimum. This action takes place
one of the main reasons for the seemingly with o111 noticeable change in the operating
large numbers of active devices in some point of QI , whose input impedance re-
!Cs. If there's any possible advantage, in mains constant, with very little detuning
they go! Why not? effect on the input tuned circuit even with
Communication ICs maximum ave voltage on Q2 .
Here is a different st ory right away. The configuration of QI and Q3 re-
Practically every "radio set" for co mmuni- duces the internal feedback to a low figure ,
ca tion work that l 've ever seen or heard of which is generally immeasurable up to
has coils in it (or at least resonators if you several hundred megahertz. With the 30 dB
go to Uli F and microwaves). Now in a of gain obtainable at 60 MHz, this is a great
little flat pack o nly a few mils high how are advantage.
you going to put in any tuned circuits? The Diode DI is for de biasing of Q I , under
answer is, o f co;.irse, you don't. They go conditions of varying temperature. Being
outside. So now where are you? You're up laid o ut on the same silicon die as QI, their
against a conflicting set of requirements. currents will be closely similar, even with
There are some "natural" divisions of severe changes in temperature, with conse-
rf, af, and i-f circuitry where attention quent de stability.
must be paid to the proximity of the The noise figure of a two-transistor pair
components, as illustrated in Fig. I . is lowest when the in put device uses the
41
Fig. 1. Block diagram of typical receiver showing
"trouble spots."
common-emitt er connection. The second and the 590 input lead (pin I ) tapped at 5
device will then have little effect on the turns. The I 00 pF capacitors (C I and C2)
noise figure. are too large for in-band operating units
The actual measured gain a t 60 MHz is but are all righ t for experime n tal use.
over 30 dB in suitable circuits, wh ich Manual co ntrol of the ga in was accom-
consist of no more than tuned and im- plished with the 5 kr2 pot , ave tests to be
pedance-matc hed ind uc tors. done lat e r in i-f service.
By the use of two of these devices in The output of the device (pin 6) was
cascade, bandwid ths of over I0 MHz may connected to the top of L2, and 6 turns of
be obta- at 60 MHz, which shows small wire wound arou nd the cold end.
considerable possibility for amateur micro- I started out with loose-coupled an-
wave amplifier service. tenna, base input, and output lin k, but
soon discovered this is 1101 the way to go
with the 590. It likes lots of coupling, on
both input and output. When this was
, 00 0 10
done, the really large gain of the device
,. ~ ~ 1 C~~S~~~TIOtl.
'
o oo
J
l BOTTOM Vl[WI became evident. Various signals from 80 to
below 40 meters sounded like reception
r------------------,
I
with a superhel receiver (except for the
: :SI( selectivity).
10<>-'-:-1---~ Whe n 1 used the device as an r f amp li-
3K
fier in front of my lab receiver, I had to
APPLY AGC
YOLlAGE MERE
3K DATTERY
------1
7
42
reduce the gain of that receiver by a large The Amperex TAA300
amount. That HEP 590 really has a lot of The home of Amperex being not too
sock. far away in Rhode Island, I visited
down there for half a day and was well
Putting the HEP 590 on 6 Meters rewarded with some new VHF transistors
After removing the 40 meter coil and that look swell, and also in meeting some
putting in the required values for 6 meters, topnotch engineers in the lab.
signals came in right away . Plenty of The T AA300 is a comp lete af amplifier
statio ns were heard, and in particular, using in one of those little I 0-pin cans, and puts
a I 00 ft wire, Q5 signals were heard that out a watt of audio when required -
could not even be found on the dial enough to modulate a couple of watts of rf
without the 590 preamp. It works! The 6m on 6, 2, or 432 - which , along with the
values fo r the inductors are as fo llows: exciter section, will take just about all of
LI - 4 turns at 8 turns per inch, 9 /16 in. the de power of two lantern batteries,
diameter; L2 - 5 turns at 4 per inch, 9/16 rated al 12V and 0.5A.
in. diameter L3 4 turns, wound on L2. And it's good for the receiver also. Plug
Figure 4 shows the 6 meter version of the i-f diode outpu"t into it and there's your
the HEP 590 amplifier. loudspeaker banging out a watt. Let's lake
So the Motorola HEP 590 JC Linear ampli- a quick look at one of these.
fier works as claimed. It is a relatively easy With !Cs, allow for lots of time. Some
IC to practice on, with only th ree active of this time will be spent in puzzling out
devices inside, and can be built up into a three things : the internal circuit, the ex-
circuit and tested in a day's time. ternal circu it, and how to draw one schem-
From working with it so far, I can atic t hat has both.
verify that excellent reduced-size, high- I am counting on the T AA-300 to do a
-frequency i-f a mplifiers for amateurs ca n lot for use as a I W af a mplifier; but once
be made with it. again, remember that these little gems are
c--7RF OUT
LJ,
+
Fig. 3. 40m amplifier built around a HEP 590.
43
IO
not primarily made for experimenters - With printed circuitry it has always
they're made to be wavesoldered into small been my fee ling that why sh ould anyone
radios and TVs by the hundred thousan<ls. build just one? To learn the printing
An engineer in a large company can process invo lved , yes. Fo r building one
afford to sit down at his desk and spend a circuit , no. Now some of these little I 0-pin
week or two figuring how to best put this jobs are good for experimen ters, and you
device into a set because that week can be may also be interested in learning about
spread cost-wise over (hopefully) a large them for business reasons, too - or fo r
number of sets if he does a good job. Can size, altho ugh to really cut down in size
you do this just for one item? calls for some pretty expensive ex tern al
Does the circuit of Fig. 5 look like an components. Look at Fig. 6, the external
audio amplifier? Where 's the input and the circu it. There is a .64 F capacito r on pin
output? What are those five dio des doing in 6, 400 F fro m pin 2 to pin 5, a 25 F
there? Why I I transistors just for an af capacitor on pin 8, and 4 7 F from pin 2
amplifier when the "All-American Five" to ground. With "ordinary" size electro-
design will give you mixer, oscillator, i-f, af lytics three o f these values are each fou r
driver, and af gower stage? times the size of the device itself, but I
Don't think I'm running this thing expect with voice frequen cies some of
down, because I intend to use it. I just these can be cut down a bit.
want you to be prepar.ed for a little "new Then there arc a few application notes
thinking. " to think about, like the stability questio n,
44
and a few others, but they are no t bad. have to add a few things on the outside, of
All wor th-while things take time, and course, as shown in Fig. 8.
this is one of them. Be ad vised, and allo w There are so me good ideas incorpora ted
yourself time enough to st udy it o ut before into this baby - like using two transistors
you pick up that little soldering iro n . for the mixer. So are three fo r o ne i-f stage.
Amperex TAD I OO I don't particularly go for the fo ur transis-
Ambitious, knowledgeable, hardwork- tors in the af s tage, witho ut even the power
ing people have looked at the several- out put unit y et , but it does work .
million-per-year ma rket for j ust plain o ld What doesn't go for the amateur experi-
radios (new ones, of course) and though t menter is the fa ntastic proximity of the
about making ICs for them. And they input, the mixer, the oscillator, the i-f and
made some. I have wor ked with one of the af - all in there together o n that one
them, the Amperex T ADI OO , and here is tiny chip. Just too, too close for me. When
the story as I found it. every thing is running right with all the
There are a dozen or so tran sistors in precaution s advised and an exact copy of
the litt le plastic box only 3/ 4 in. long, as the original printed board is made up and
you can see in Fig. 7, the internal circuitry all the components are in exactly proper
of this one. These transistors are pretty places, it does work as a BC set.
close together in there - much too close One of the precautions listed: "The
for an amateur experimental unit, as you oscillator must be limi ted to 100 mV ;
will see. otherwise it will get into the i-f and the
The manufacturer, ambitious as he was, af." It did! I worked 10 days o n this
did produce large numbers of excellent and particular IC and to make a long story
very tiny receivers with this IC. It has a short, the mixer and the i-f ran o kay at
mixer, oscillator, i-f, detector, and an af times (and at times not), but when the
amplifier - but no power stages. Almost internal local oscillator was u sed , I kept
everyth ing you migh t need , except yo u running into trouble.
Clll
4 ,71(
45
AF
AGC COtnAOI.
""
.....
40HMS
O ~~W OU T Pu f
0 101<2'%
,_..30"1.
72ulm
osc cou..
Bear in mind, such an IC does work fine The people handling this unit are excellent
for mass production of BC sets. It's just engineers and have some other very good
that for an amateur experimenter as a devices; and they are still working with
single unit, there is too much feedback newer and better dt:vices for a complete set
involved. Also, the oscillator is strictly (including an FM job!), so we will almost
limited to the BC band service, which certainly see them again soon.
precludes using it as an amateur converter.
46
Repeater Audio Mixer
2N4360
2N4360
2N4360
47
The gain of the opamp is determined by
the ratio of the feedback resistance to the 7090 <
input series resistance. In this case , the gain 8 - I~
icon Pl.., CO~NECT IONS
on all the inputs is unity. If more gain is TOP Yl[W
..--...--~
desired on a particular channel, the input
OUTPU T
resistance can be lowered to change the TO TONE
DECODER
ratio (and therefore the gain). For ex- (SEE TEXTl
The frequency response of the amplifier OUTPUT 6M 6M .34 .34 SPARE SPARE
6~c1g~4iR AUDIO ENAB LE AUDIO ENABLE AUDIO ENABLE
is essentially flat from de up to about 8
SOUHCHEO INPUTS
kHz. Beyond that point, it rolls off. The 22 Notes: ( ANY NU ~ SE R >
48
The SST-1, Solid State
Receiver for 40 Meters
49
transformer secondary has one side tied SWR Bridge and Meter Circuit
directly to ground. Various ground arrange- For tuneup while operating portable it
ments were tried to eliminate one of the was considered desirable to include some
batteries, but the two-battery setup was means of monitoring final collector current
finally decided upon . This does split up the as well as some way of indicating when a
load on the batteries somewhat and in- match to the antenna was obtained, since
creases their life. The rf portion draws random-length antennas are convenient. A
about 15 mA and the audio section any- small pilot light could be used to indicate
where from I 0 to 50 mA on strong audio relative current in the final, but the addi-
peaks. tion of the 0 - 1 mA meter and the swr
bridge has proved its worth in the field.
CW Monitor The meter is set by a three-position
A unijunction audio oscillator is in- swit ch to read either final-stage emitter
current or forward or renected power. For
cluded on the audio ou tput module for
emitter-current readings, the voltage across
monitoring while transmitting. The moni-
tor output is fed into the aud io output
a I on resistor in the emitter lead gives a
fu ll-scale reading of app rox imately I 00 mA
amplifier after th e gain cont rol, and the
with the meter specified . A 2 0 k.Q poten-
monitor level is set for suitable volume
tiometer is connected in series with the
which is independent of the receiver vol-
meter for swr measurements and must be
ume control setting. Voltage for the moni-
set to zero resista nce for final current
tor is obtained from the transmitter supply
indications.
and it is keyed along with the transmitter.
The 100 k.Q re~istor and the 0.01 F The swr bridge is standard in design
capacitor in the emitter lead of the uni- except for the two transistors added as de
junction provide an audio tone of about amplifiers. They were found necessary to
get indications with the low power output
600 Hz; this can be raised by lowering the
value of the emitter resistor if you prefer a of the transmitter. The gain mismatch
monitor pitch of higher frequency . between transistors does not really allow
the swr to be measured accurately, but a
Transmitter good indication of a proper match to a 50
The transmitter, crystal-con trolled for or 72.Q coax line can be obtained by
simplicity, used three 2N697 transis- adjusting for minimum renected power and
tors - two in the rf section and the third as maximum forward power. A two-section
a switch for keying. The keying switch was switching arrangement could be used to
added to reduce the current through the eliminate one of the transistors, thus pro-
key contacts, although it could easily be viding more accurate readings. The bridge
eliminated and the final connected to the condu ctor is a piece of V.. in. copper tubing ;
negative supply continuously, since it does the pickup wires are of 14 AWG solid wire.
not draw any current without drive from The assembly is built in a small channel
the oscillator. The keying arrangement bent up out of aluminum sheet met<1I as
shown was a result of using the transmitter shown in Fig. 2. In order to get readings
which had already been built when the with the bridge it must be mounted off the
transceiver idea came up. main chassis with an insulating spacer and
The oscillator is connected in a Pierce the chan nel connected to ground via the
circuit with the crystal between collector coax shield on the input and output only.
and base. The I k.Q potentiometer in the Construction
emitter controls the drive to the final. The The main housing for the transceiver is
oscillator collector uses a slug-tuned coil a Bud SC-3030, which is 6 x I 0 x 7 in. This
with a sma ll link wound over the cold end volume allows construction without crowd-
to the couple into the final amplifier. ing yet keeps the unit small enough to be
The final, operated without any bias, easily portable. The transmitter is built on
runs class C. It uses a pi netwo rk in the a separate minibox 2V.. x 2V.. x 5 in. and ca n
output circuit. be put together as a separate unit with a
50
couple of leads provided for a crystal aluminum chassis. The front is cut off to
socket on the front panel of the main clear the controls on the front panel and
enclosure. The transmitter is mounted to the rest notched and cu t where necessary
the front panel by the key jack which has to clear the main enclosure. A vertical
one side tied directly to ground. In addi- shield was positioned across the transceiver
tion, a barrier terminal strip was provided between the receiver and transmitter, al-
on the rear panel for connection of an ac though this is probably not absolutely
power supply for fixed station work. The essential.
drive control potentiometer was mounted Th e subchassis provides adequate
inside the transmitter since it is not ad- mounting space at the center (between
justed in normal operation. The final am- receiver and transmitter) for the transmit-
plifier transistor was mounted on a piece of ter batteries and the two 9V balleries arl!
.040 aluminum about 2 in. square which mounted to the rear enclosure wall with a
serves as a heatsink. Since the collector is homemade bracket of sheet aluminum. The
connected to the transistor case, the heat- I. SY penlight cell is mounted on the swr
sink must be.insulated from ground. bridge channel as show n in Fig. 2.
A twice-size PC board layout for the The swr bridge is mounted near the rear
receiver rf-firs t audio section is shown in o f the transmitter and the bridge channel is
Fig. 3. The integrated circuit is soldered insulated from the main enclosure as men-
directly into the circuit board , although a tioned above. Connect ion to the bridge
socket can be used. Don't be loo con- should be made with sma ll coaxial cab le
cerned about damaging the integrated cir- going to the transmitter pi network and the
cuit. The one here had to be removed from transmit - receive switch. Similar coax is
the board once by cutting the leads above used to connect the receiver and antenna
the board and remounted by soldering connector to the switch.
extension leads to it indicating that these The photographs show the general con-
little devices are really quite rugged . With- struct ion and panel layout used, but other
ou t a proper unsoldering tool, it is difficult builders may find other arrangements more
to remove the IC once it has been soldered desirable. The general layout is suggested as
in so a socket might be a good idea even a logical one.
though the board is not laid out for one.
To facilitate parts placement and circuit Adjustment and Tuneup
identification, the composite layout/ When the receiver is operating, the
schematic of Fig. 4 is included. Th is should oscillator freq uency is adjusted to cover 40
help to speed your final assembly process. meters by alternate ly padding the tuned
The audio output amplifier from the circuit with fixed capacitance and varying
discard ed broadcast set was mounted on a the trimmer in series with the main tu ning
sma ll piece of Vector board along with the capacitor to achieve the desired band-
compo nents fo r the CW mo nitor. For- spread. By se tting the trimmer, the en tire
tunately, the radio I d ismantled had been band can be tuned or the bandsprea d
built in two sect ions, with the audio stages limited to just the CW portion. Other
on a separate small circuit board . Other arrangements might give different band-
radios might have to be operated on to get spread, but the transceiver described
just the audio section to use in the trans- covered the first I 00 k Hz of 40m over
ceiver. The input to the amplifier can be about half the dial and the other 200 kHz
located by tracing down the leads going to over the second half, which was felt to be
the volume con trol in the original radio. about right and allowed for monitoring
The wiper arm on the control is connected sideband as well as CHU just above the top
to the input. band edge for time checks as an added
The receiver modules are mounted to a bonus.
small subchassis which was extensively Some trouble with the local oscillator
worked on wit h tin shears. It is made from was experienced at first: Spurious re-
a standard open-end 5 x 7 x l 'h in. sponses were obtained, with 7 MH7, ap-
51
VI
"->
-=-v ~v
}
AUDIO
SECTION l
OUTPUT
rr-<~:5.ZOI
ml
PHOHl.S
VE AKER l' J_
l"OI
l'
I
:
01
DI 0 .1 5 .SV
RECEIVER L4
SECTION
10
...
LOCAL OSC
I 2K
TRANSMITTER .01 ~ 2K
I MEG ] ~220K ~IOOK 22
SECTION IHH5
4 7K
590
CW WOHIT~
+
J l.~V 2N233A
TO
METER SWITCH
0 - 1 mA
53
Fig. 3. Douole-size PC board layout for th e rf/af
section of the receiver.
54
all. For CW work, a more selective audio
bandpass filter can be used either in place
of the low-pass filler inside the t ransceiver
or in th e headphone line. A couple of 88
mH toroids and some 0.5 F capacitors
will increase the selectivity markedly when
connected as an audio filter. For general
operation the filter is not necessary and it
does prevent good copy on sideband sig-
nals.
With the addition of a balanced modu-
lator. this setup could provide the basis for
Photo showing shielding and construction of 40m a simple little sideband rig using the audio
transmirrer and receiver in main chassis housing. section for the speech amplifier, generating
a double sideband signal at 7 MHz.
Receiver performance has been excep- Using currently available transistors,
tional and sure beats using a regenerative this little transceiver can be used to drive a
set with its limited audio output and class C stage to I 5 or 20W for a little more
instability. Drift from a cold start with th is signal, but at these power levels a battery
receiver is practically nil and it can be used of rather monstrous proportions is re-
to copy sideband signals with no trouble al quired; the 2W power level is just about
AUDfO OUT
Fig. 4. This composite sket ch shows posi tioning of components on the etch ed board.
SS
right for the mercury batteries used. And IC1 - RCA CA3028A
20\V isn't really QRP anyway. L 1 - 5 turns 28 AWG spaced over L2
L2, L3, L4 - 36 turns 32 AWG wound on .380
diameter toroid core (A rnold A4 -380-
T abl e of Pa rts
125-SF)
Cl - 140 pF peri section, dual va riable capacitor LS - 12 turn s 28 AWG close-wound on 3/8 in.
C2 - 9 - 35 pF ceramic trimmer dinmoter slug-tuned form
C3 - Main tuning 50 pF variable capacitor LG - 3 turns 22 AWG over B+ end of LS
C4 - 365 pF broadcast va r iable L7 - B&W M lnlductor Y, in. 16 TPI, 1 in. l ong
CR 1 - 5.6V zener, 1N708 or oqu lvn l ont Ml - 0 -1 mA meter (Emico Modol 13)
T 1 - I ntor stage audio transfor mer 1.2- 20 kD Main tuning dial - Millen Type 10039 m i dget
GT -12 PC (Allied Radio) panol dial
BA 1, BA2 - 9V batter ies
BAJ - SV cell
BA4 - Two 12V batteries in series, Mallory
TR289 or equivalent ... W9ZTK
56
Practical IC Regulator Circuits for Hams
57
determine the exact voltage in a circuit
Table I. IC Parametric Limitations at 77F.
because the collectors are freque n tly at the
Parame ter CA30 18 CA301 8A most positive voltage.
Many !Cs have lo w brea kdown voltage
Maximum powe r dissipation:
on y si n gle transisto r 3 0 0mW 3 00 mW ratings, which can be a serious d isad van-
t~:~:-;c~:~/oC
)>85 c
for TA 450 mW 4 5 0 mW tage. In some cases, you may be ab le to
design around the problem. Another prob-
Muclmum collector-to-emitter lem at the experimen ter's level is the fact
voltage 15V 15V
Max:lmum collector-to-base that if one transistor is dest royed, the
vo ltage 20 V 30V whole IC has to be replaced . Th is disadvan-
Maximum collec tor-to-substrate
vo ltage 20V 4 0V tage is somewhat offset by the low S 1.62
Mxlmum emitter-to-base
volt age sv sv cost of the CA30 18, but it is hard to
Maximum collec tor current SOmA SOmA rationalize soldering 24 leads for each
mishap. A discrete transistor could be used
in a crisis wit h the CA30 18. Be ti er yet, use
a socke t fo r breadboarding.
limit for the sum o f th e power dissipatio n Applications
of the individual transistors. For example, The CA30 18 is a natural choice for a
if one t ransistor operates at the 300 mW power supply regulator amplifier. Such an
level, the average power of the remaining application, I feel, is also interesting for
three may not exceed SO mW each (300 + analysis. In Fig. 2, I have drawn a very
3 x SO = 4 SO). The second new rating basic regulated power supply using th ree of
refers to the breakdown voltage of the the transistors. Let's examine its character-
collector-to -substrate diodes. These diodes istics from the ratings o f the IC.
arc always reverse-biased because t he sub- Nearly aJI the current to a load will flow
strate has been con nected to the most through Q4 (pins 12 and I ); therefore, we
negative part of the circu it. It is easy to are limited to SO mA, which is the
1:
Exterior view of the 500 mA pow er supply . The case usod is a Bud CMA 1930.
58
Placement of parts is only
critical with respect to
squeezing every thing into
the box. The 2N5295
transistor is outside th e
rear apron and insulated
from tile main chassis. A
piece of copper 2-112 x
2 x J /32 in. forms a heat
sink for the power tran-
sistor.
maximum collector current o f any single power supply could service almost any
transistor. If you think about it, this is all circuit now operated on small 9 V batteries.
the current required by ma ny circuits. The An FM portable receiver draws peaks of
TRIAD MOTOROLA
F25X MOA920-I
RCA
CA3 018
IBV
117V: Jll l 2 .6 V
4700
100 F +
IK
25V
1200
59
~av
002
[~O""A
<SI
IK
2W
>OOO
(2,,V FlA.l SCALC>
R201\ ~ mA
IOI\ 100 MA
21\ ~ mA
~00 mA 7 TO 18V
Fig. 3. "U1ili1y" power supply wirh 7-18V ourpul. Resisror R decerm ined by currenl requirements:
20S1-SO mA; JOS1-JOO mA; 2S1..,.SOO mA.
less than 20 mA ; AM radios draw even less. variable voltage output, and current limit-
The second rating to consider is the ing by use of the fourth transistor on the
collector-to-emitter voltage. The greatest IC. Incidentally, you might just want to
permissible value is JO.SY, which is well build a power supply like this if you are
under the I SY rating of the transistor. The newly acquainted with semiconductors.
collector- base breakdown voltage is not The 2N5295 will handle 500 mA and
exceeded and the emitter-base junction is higher voltage with ease. If we assume the
never reversed-biased, so the design is clean beta of that transistor is 50 at full load, the
in those respects. Don't forget the collector- base drive required would be 500/ 50, or I 0
to-substrate rating. In this case , the maxi- mA. That is no problem for Q4 of the
mum voltage seen is I 8V: okay for this CA3018. On the other hand , it soon
circuit, but not for the next one. becomes apparent that we need higher than
When considering the power rating, it is 20V breakkdown. Some relief is obtained
safe to assu me we will no t exceed 85C ( no by using the CA30 I 8A, but not q uite
derating necessary), and the power dissipa- enough to hand le the collector - emitter
t io n of Q4 will be much greater than the voltage drop across the Darlington pair. By
total power of QI and Q3. At 50 mA of
current, the power (P=IE) would be .050
( 18 - 9)=0.450\V, which is high for a single
transistor. A reduction in output current or
lowering the input voltage on pin 12 is
necessary to operate within ratings. In any
case, it would be wise to heatsink the IC
for this circuit. You may expect it to get
hot above 30 mA .
The shortcomings of the CA30 18
become apparent as we attempt to design a
more versatile supply such as the one in
Fig. 3. A 2N5295 power transistor was
added to the basic circuitrv in order to
operate to supply to I 8V and 500 mA .
Other changes include the use of a The compleced .soo m A regula ced low-volcage
tern perature-compensated zener diode, power supply.
60
using a zener diode between the high- Wh ile the CA30 18 is not truly a circuit ,
voltage point and pi ns 11 a nd 12 of the IC, it provides an in teresting introduction to
the collector-to-emitte r voltage require- the technology and characteristics of
ments are reduced, as is the collector-to- integrated circuitry. Its versa tility invites
substrate voltage. experimentation. For those who would like
It is not necessary to use a temperature- to duplicate the regulated power supply
compensated zener here. Collecto r-to- sho wn in the pho tos, I include a fu ll-size
emitte r voltages are the only ones above
ratings now. Three CA30 l 8s tested had DC 10 15 V 0 LTS
collector-to-emitter breakdown voltage 5 \\\\\11 \lll lj ll fl/ ///// // 20
~::?'\.\
ratin gs in excess of 23V; however, th is is a 1
0 200 300
weak point in the design . ~\\ 100 400 //'?25
1/1111;,
61
Low Cost Fundion Generator
for the Experimenter
RANGE T RIMMERS
OP AMPS ARE : 22K
Fo1rchlld A704,
Not10nol Sttn4. LM709,
or e-qu1v. 47K X2
IOOK
+18
WAVE
SHAPE R6
IK SIG
-18 R4 OUT
4 70
K
3.9K ;;,OO T
62
quency, enabling the distortion to be con- charging until the limit imposed by th e
stant over the entire range. Due to diffi- opamp and the power supply voltage was
culties in obtaining precise symmetry in reached. However, A I is connected to the
FET characteristics, plus nonideal charac- frequency control pot, and supplies the
teristics of the transfer fun ction itself, it is voltage that is converted into the charging
difficult to get the distortion lower than rate by the various resistors. A 1 is con-
about 1.2%. This is adequate for almost all nected in a positive feedback loop with high
applications. hysteresis. The opamp is sensitive to the
Integrated-circuit operational aml_llifiers voltage between the two input terminals. If
(opamps) are used to generate the tri- the difference is only a millivolt or so, the
angular wave. Unit A2 is connected in a amplifier goes to either its fully positive or
circuit arrangement known as an integra- fully negative limit, depending on the input
tor. Time-determining capacitor Ct is polarity.
charged at a rate directly proportional to Feedback to the input is through R2
the current fed into pin 2 (the inverting and R4. Assume the output of Al is fully
input) of the opamp. Pin 2 is kept at negative. This voltage is clamped to about
ground potential by feedback through the 4V by 02 and Z2 (DI and Z l when
capacitor, so the current equals the voltage positive). This 4V is connected back to the
at point Vc divided by the total series input through R2 (and the 3.9 kQ pro-
resistance, consisting of the resistor on the tective resistor). As the voltage at the
octave switch and the range trimmers. output of A2 gets increasingly more pos-
Thus, it is possible to control the charging itive, a point is reached where the current
rate with RI, and the high current capacity through R4 becomes opposite to and
vs low leakage current of the opa mp allows slightly greater than that through R2.
the charging rate, and hence the frequency, When this happens, the amplifier instantly
to be varied over a 500-to-1 range, com- (in a few microseconds) changes state,
pared to the normal I O-to-1 range for aided by the positive feedback through R2 .
normal capacitance-controlled oscillators. Of course, when the opamp changes state,
If a current were just fed into the the integrator starts to charge in the
opamp, the capacitor would continue opposite direction , thus creating the tri-
63
Closeup view of circuit board.
66
The IC-mitter
Stephen Goldstein
Fig. l. See rexr for Cl , C2, LJ, TI. For CW you can also switch ou r rhe transistor if you want.
Instead of keying the volrage, you could key the ourpu r. If you don 'r want age, simply ground
pin 2 of rhe first IC.
67
An IC Marker Generator
D. A. Poole K4BBC
.)ufl>\jt
0< ~ ..[C T~ 00 '
f---o ()\ f P t
,
'0 .
~"'-o---<) 10 P"; " ON ALL 1,,
l.
, l.
1.9 ;;. T CPO"'"' ' c;"'I ~~ ( 1
1..92)
Fig. 1. IC in terconnection and base diagram for the bandedge marker gen er ator.
68
built on perforated board from individu-
ally pur chased components.
The circuitry consists of a 2 00 kHz
oscilla tor and three divide-by-two stages.
In the oscillator, a Fairchild L 9 14
integ>ated circu it is connected as a
crystal-controlled multivibrator. A 200
kHz crystal is used because it is c heaper
and more stable than the I 00 kH z crys-
tals usually used in marker generators.
With the values shown the frequency is The finished marker genera1or. Nore Iha! pen
within a few cycles of zero-beat with light cells mount handily in cover section of
wwv. mini box.
The output from the 200 kHz oscil- switch , battery holder, and output jack
lator is successively fed to three L 923 are from Lafayette.
integrated circuits, each con nected to The generator has proved to be reliable
divide the frequency by two. A rotary a nd stab le, providing subband identifi-
switch selects the desired output. Three cat ion at a fl ick of the switch . Because
penlight cells supply the 4.SV required l?Y the harmonic output of the divider is
the integrated circuits. quite strong, tight coupling to the receiver
The generator in the photo was built or transceiver is not necessary. Good
in a 4 x 2'!.. x 2'!.. in. aluminum box. The results can usually be obtained by slipping
printed circuit boards with integrated cir- the end of the o u tput wire down into the
cuits are attached to the box with screws shield of the first rf a mplifier tube.
and small space rs. The small ro tary ... K4BBC
69
A Simple Integrated Circuit Q-Generator
A lm os t any int egrated circuit filter. It can be easily switched for broad-
operational amplif ier can be used to band operation so it does not affect no rmal
build this Q-multiplier. Its adva ntages SSB operation. Since it is adjusted for unity
are extreme circuit simplicity and a gain, it does not upset any gain relation-
useful frequency ra11ee that ex tends ship in t he i-f stages. As compared to
from audio frequencies l o almost all vacuum tube and transisto r-type Q-m ulti-
i-f f requencies. Both th e peakinf? f re- pliers, t he circuitry of the unit is extre mely
quency a nd Q can be made variable. simple d ue to the use of an integrated circuit
o perational amp lifier. The unit ca n be suc-
cessfully used on frequencies far higher than
those no rmally used with vacuum tube or
transistor Q-multipliers up to 5 MHz or more,
Bo th vacuum tube and tr ansistor Q-multi-
depending upon the integrated circuit used .
plier circuits find wide application in im-
The circuit can also be used a t audio
proving the selectivity of transceivers and
frequencies, if desired. One can also build an
receivers, parlicularly on CW when t he
audio selectivity unit for outboard use when
Q-multiplier is used to peak the i-f response
it is not desired to make internal mo difica-
following a steep-5kirted crystal or . mech-
tions to a transceiver or receiver.
anical SSB fil ter. The Q-multiplier provides a
very narrow bandpass but if used alone does Circuit Description
not provide steep skirt selectivity. When The Q-multiplier is constructed around an
used in conj unction with an SS B fil ter, h0w- integrated circuit operational amplifier.
ever, the latter provides the necessary skirt Many such amplifiers are available on the
selectivity (see Fig. 1). market at prices starting at a few dollars.
The main requirements for choosing a suit-
able unit are that it have a differe ntial input
(inverting and non inverting in puts). and a
single-ended output and a bandwidth suffi-
cient for the frequency of operation. For
example, Fairchild 709T amplifiers are avail-
Al
['"
The Q-multi plier described in this article Fig. 2. Basic type of integrated circuit operat ional
is meant to be inserted between any i-f stage amp lifier wit h a differen tial input and singl e-ended
in a tran sceiver o r receiver following a n SSB output tho c is needed for the Q-mu lriplicr circuit.
70
aole for about $3 and are usable up to at nant at 455 kHz for use in a 455 kHz i-f
least 500 kHz. I built a unit using a Fairchild chain. Positive feedback is supplied via a I
741 which is usable up to 1-2 MH z. Other kQ potentiometer. As with other Q-multi-
amplifiers such as a Motorola MCI 530 can plier circuits, the Q can be increased by
be used up to 10 MHz. regulating the feedback until a point is
Figure 2 shows the schematic of the type rea"ched when the unit will break into
of operational amplifier which is used and the oscillation. The Q-multiplying effect is most
formula for the output voltage of an ideal effective when components are used for the
amplifier. The Fairchild 741 amplifier which resonant circuit which in themselves have a
I used requires no external frequency com- good Q. The inductor used for the 455 kHz
pensation components. Other amplifiers may Q-multiplier is a molded type which provides
roquire a few external components for this a Q of about 55 at 455 kHz by itself. The
purpose as specified on their data sheet. The circuit can multiply this value by about 50
frequency rolloff components should be times or mo re. Another suitable inductor
c)losen such that the amplifier gain starts to can be obtained by using only one or two
decrease just above the frequency where it is sections frq!11 a regular I mH rf choke. The
used as a Q-multiplier. There is no advantage trimmer capacitor is used to set the fre-
to having the gain "rolloff" at any higher quency of the Q-mulliplier in the middle of
frequency and would just make the amplifier the i-f passband. It can, of coruse, be used as
more susceptible to oscillation due to a stray a variable tuning control for the peaking
feedback path via external components. As frequency of the unit. The input resistor,
noted from the gain formu la, the gain of the although shown with a nominal value of 1 kQ,
amplifier depends upon the ratio of R2 and should be chosen so that the gain of the
RI. If R2 is made equal to RI, the gain is overall circuit is approximately unity.
unity. The Q-multiplier effect is based upon
Construction and Adjustment
replacing R2 with a parallel resonant circuit
which will present a very hlgh impedance at T here is nothing critical about the con-
one frequency and, therefore, maximize the struction of a unit utilizing the circuit
overall gain at that frequency . Positive feed- described as long as the various lead lengths
back is also used to enhance the Q-rnulti- are kept short. The photograph, for instance,
plying effect of the circuit. shows how the various components can be
directly wired together on a sma U piece of
Vector board. The board itself can be
directly mounted near the i-f chain in a
41011
C.. nltOllMU lWOOO,
transceiver, receiver, or together with the
< loll.U:O '1'.t.CMO n ~ ,..~Ol14 1
potentiometer used for feedback control if
the latter is panel mounted.
If the Q-multiplier is not made tunable
the initial adjustment consists of peaking u~
the LC circuit. This can be done with the
unit connected in an i-f strip and using any
test signal centered in the i-f bandpass. The
adjustment is best done with the feedback
contro l set for minimum Q and with a barely
audible CW test signal. The input resistor value
should be then chosen for approximately
o 1wci.tst
unity gain. The adjustment is not difficult
and need not be made. exactly. The output
Fig. 3. Circuil of 1he Q-mul1iplier as consirucred level produced by a test signal without the
for a 455 kHz i-f
Q-multiplier connected is noted. Then when
the Q-multiplicr is used, the input resistor is
l'igure 3 shows the schematic of a practical chosen so that the output level remains
Q-multiplicr circuit with the LC circuit rcso- approximately the same. With most transis-
71
tor i-f stages, the value of the resistor needed bandpass. This will result in some increase in
will be about I-Sn. gain when the Q-multiplier is tuned to the
Bypassing of the Q-multiplicr can be done center of the i-f bandpass but normally ihe
in a number of ways in order to allow for result will not be objectionable. Another
normal SSS reception. If the Q-mulliplier is way to bypass the unit is to replace the LC
made frequency tunable, it can simply be circuit with a simple resis tor equal in value
tuned outside the i-f bandpass. In order not to the input resistor (as in Fig. 2). The
to have the i-f gain decrease too far when switching action can be accomplished by
doing this, the input resistor must be chosen using a l 1..n po tentiometer for the feed-
for unity gain to take place when the back control which also incorporates a
Q-multiplier is just tuned outside the i-f SPDT switch. The switch must he wired such
t hat t he resistor replaces the tuned circuit
when the wiper arm of the potentiometer is
at ground potential.
Summary
The simple Q-multiplier circu it described
can be used for a variety of purposes besides
that of improving i-f selectivity. It is useful
for improving the Q and selectivity of a
variety of tuned circuits as they might be
used in FSK converters, audio filters for
distortion tests. etc. Stagger-tuned circuits
used in series can be formu lated to provide a
variety of bandpass shapes, often replacing
more expensive components where bandpass
shape factor is not important.
The permission of Dick Gcrdee of Optical
The Qmulriplier components can be direclly wired
together using Vector board mounting. In this case, Electronics to present this circuit. which he
a PC board potentiomerer is used ro ser rhe oper o riginally developed, is gratefully acknow-
aring Qrarher rhan making it conrinuously variable. ledged. . .. W2EEY
72
3 Versitile IC Testers
73
Fig. l . Connecrion diagram s for some common !Cs. All are shown from rop view.
outpu t is ident ifi ed , connect a pulse input ing pin 11 to grou nd and pin 4 to +S V is
to the o ther pins and fi nd o ut which o nes appropriate for the SN747 3N fli p-flo p.
give an inverted ou tpu t. This is now suffi- Again , identify the outpu ts as above and
cient to identify t he IC. connect the scope to one of them. Connect
Frequently an IC will be perfec tly good the pu lse generator to the o ther pins. If
except for one inpu t. If it has only a mino r you t ruly have a fli p-flop, you should see
defect (one of four gates bad , a defective the pulse waveform divided down to a
reset on a nip-flop, or such), yo u can break square wave at half its freq uency.
off the pin correspo nding to the d efec t.
This is much more economical than dis- Similar procedures can be fo llowed for
carding the unit. almost any digital IC. Some o f th e newer
Now that we have identified the gates, fun ctio ns are suffi ciently complex to make
let's try some o ther po ssibilities. Connect- a data sheet necessa ry for testing, o r yo u
74
+5V or +3.6V
0PENOING OH
LC. FAMILY
8.2K
8 PIN
Kl PIN
3.9K 270 270 14 PIN
Lc. k~s
1
TO
UCE-H.MllEllED
PINS t<
PARALLEL
may never hit on an appropriate combina- counter can be tested by the above
tion of inputs and outputs. In such a case metho d , it is rather tedious, and the driver
the digital tester of Fig. 2 can still make ca nno t be since it is designed to interface
valid operational tests, although it will be with a high-voltage device. To simplify
of little use in identifica tion. testing of the counter and SN744 l N Nixie
Decad e Counters, Nixie Drivers driver, a special unit was built (Fig. 3)
Counting circuits and readout lube which simu lates decade counter operatio n.
drivers are becoming more popular as their Half of an SN7400N (or any o ther 2-, 3-,
prices decrease. While the SN7 490N decade or 4-input dual, triple, or quad gate) is used
+ov
SLE TEXT
22K
+ 2:00-:SOO'Y()-_,.,VY-~
;~LL_J
r+-g 0 0
D-m
SN7441
SOCKET
75
+-l~V PIN 1 PIN 11
Sl(I , Sl(2
[ill
~
TEST [])
IOOK 302
91K
l.,K
TEST
709 , ,
IOK 709
~IK LMI02
741
LMIOI
2 .4K
PIN 4
SKI
PIN 6 TO SCOPE
SK2
710, 702
PIN 4
SK3
-15V -6V
6.2K
lo produ ce a si ngle pulse each time the and pin connections a re quite nonuniform.
SPOT switch is closed . Thi~ steps the The most useful type of linear JC is the
decade cou nter, and the Nixie lube is operational amplifier, or opamp, and the
observed for proper opera tio n. Obviously, most popular of these is the type 709 .
it is necessary to have a working Nixie They are so popular that the price has
driver in the socket when testing the plumme ted from over $50 to under $2 in
counters, a nd vice versa. reasonable quantities, a nd are available
The 22 kn resistor in series with the from numerous sources. The linear tester
Nixie is for cu rrent limiting. This resistor (Fig. 4) was designed primarily for 709 and
value depends o n the B+ supply and the the 7 10 comparator. It will also lest other
particular tube used. A rule of thumb is to opamps, such as the LM l 0 l , the 702, the
se lect a resistor which insures that all of A 74 1, and the LM I 02 voltage fo llower.
each digit is lit up when the corresponding T he diagram showing the test-circuit equiv-
ca thode is grounded. The value is not alents (Fig. 5) gives the appropriate ou tput
particularly critical. waveforms to look for. The tester d oesn 'l
test for parameters such as de o ffset and
Linear LCs
open-loop gain , but does provide a go/no-
Compared to digital !Cs, linears are a go test.
ho rse from a different stable. If you have The unijunclion and current-source
no idea what type you have, you are quite transistors genera te a linear sawtooth of
likely to destroy it during the testing about I 6 V amplitude. The a pproxima tely
procedure. They are designed for a wide s20n resistor is adjusted so that the
varie ty of positive and nega tive vo ltages, voltage swing is symme trical about ground
76
709, 741 t.tc
702
8V
800 mv
OUT
/WI
/VV1
(lnttmol
ConnKtK>N
when measured at the emitter of the MPS you might try an arrangement o f terminal
6520 buffer. This signal is then appropri- strips and alligator clips, or perhaps a field
ately attenuated and applied to the device of pin jacks. Also, while it is not shown on
input. The 709 output should be a sy m- the diagrams, it is a good idea to bypass the
metrical linear sawtooth of I 6V amplitude. power supply connections to ground as
The 302 output should be similar, but near to the IC sockets as possible , especi-
without attenuation. Testing the 709 with- ally in the linear IC tester.
out attenuation will show its peak-to -peak Unless you're an old pro with ICs,
output swing, without delivering enough testing them is not as boring a job as it inay
current to damage the input. A tester seem. It is impossible to describe a com-
switch is used to control the attenuation. plete test procedure for all digital ICs, since
they fail in as many ways as there are
Building the Testers internal components. It is highly instruc-
The diagrams are more or less self- tive to test the !Cs with both the pin
explanatory. The only special component diagr:tms and the internal schema tics in
used in any of the testers is the patching front of you . It will give you insight into
arrangement for the digital IC tester. I used IC operation and a better knowledge of
some sort of connector block into which logic functi ons than can be obtained from
small conical pins fit. It had been lying the literature. And the money you save
around for so long that I fo rgot where I got from salvaging just a few ICs can equal the
it. If you can't find something like this, cost of the testers.
. .. WA21KL
77
An IC Audio Processor
~-------------------------0+8-12V
5 6
7 4
4 8 3 5
SL630 SL620
200A
3 ge 2 !I
2 10 8 7 50F
50F
IOOF
OUTPU T o-----11-----.-+---t---!J--+---t---~ IOO F
3 F 05F
78
hone lea ds arc automat ic.:ally avoid ed . If th e capac it or fro m pin 6 of the SL620 lo
the usual t ypc of unba lanced input is used. ground .
however , it is connected to pin 5 of the The co nt ro l ra nge of the u nit is illus-
SL630 via a I F c.:apacitor. Pin 6 is then tra ted by F ig. 3. Only a very slight change
left unconnected. The capacitor between in age voltage is necessa ry to co ntrol the
pins 3 a nd 4 of t he SL63 0 provides a output over a 60 dB range. In practice, the
high-frequenc.:y rolloff chara cteristic. The in put can cha nge over a 35 dB range a nd
values shown pro vide a ro lloff st arting a t 3 the output level will remain between 70
k Hz, but this ca n be c hanged. if desired, by and 87 mV.
experimenting with the capacito r value.
The outpu t is taken from pin I of th e Construc tion
S L630 via t he t wo 3 F coup ling capaci- The p hotograp h shows how the au tho r
tors. Part of t his o u tput is coupled to pin I constructed a unit on perforated board
of the SL620 IC via the 0.5 F capacitor. stock. The parts layout is in no way critica l
T he S L620 IC uses th is vo lt age to genera te a nd is just a matter of conven ience. The
an age vo ltage whic h is eve ntually availa ble pho togra ph of the undersid e of the b oard
a t pin 2 of t he S L620. From t here it is is j ust shown to illustrate how easily t he
coupled to pin 8 of the S L630 to co ntrol unit could be adapted for e tc hed PC board
the gain of th e latter IC. What these two co nstruct ion as a club project since o nly
little ICs can accomplish is sho wn nicely b y o ne wire crossover is necessary a nd even
Fig. 2 . The upper curve sho ws a va rying tha t can probably be e liminated by exper-
speech input , while the second c urve shows imen ting with the parts layout.
how the audio output ap pea rs after pro- T he units will o perate eq ually well with
cessing. Notice that when t he speec h input a 6 or I 2 V d e supply and draw up to about
eit her rises ra pd ily or fa lls rapidly the t 5 mA. The operat ing voltage can be
o utput remains essentia lly co nstant. Noise borrowed fr o m a we ll filt er ed point in a
bursts. because of th eir muc.:h sho rt er time transmitter or a battery supply used . In the
dura tion , arc recognized se pa rate ly by the latter case, a usual 9 V batte ry is ideal lo
unit. The noise burst shown occurring use. There are no controls to the unit a nd
during speech , alt hough much higher in so it may be housed easily inside a t rans-
a mplitud e than the speech level, produces mit ter . avoiding on ly any location near
practically no in crease in ou t put. A n auto- high rf fields.
ma tic squelch featu re is also provided.
When there is a pa use in speech . the o ut put Operating Results
is disabled to prevent the ba ckground noise I wmpared o peratio n of the unit lo
buildup commo n to most simple compres- several t ypes o f convent iona l a udio co m-
sors. The pause time before t he o utput is pressors. In ever y case, the unit described
disabled is about I seco nd and ca n be exh ibited a mu ch smoother compressio n
c hange d , if desired , by va rying the valu e o f act ion wi thout pops. clicks, etc. II simply
----
Fig. 2. /Jlustration of the input/output and age
~
AUOlO OUTPUT
EN VELOPE voltage charact eristic of the processor.
11 I - SHORT TIM E CON STANT
1 r---, " -
DE TECTOR OU TPUT
LON G TIM E CONST ANT
A G C VOLTAGE
\ I " DETECTOR OUTPUT
' ~~ !
It I
THE LARGER OUTPUT
CONTROLS THE SYSTEM
79
unit were used as solely a power amp lifier
0 device. Unlike many JCS, however, the
a> 10 I\. necessary bias resistors and capacitors to
~
I\.. suppress parasitic oscillations are all in-
z 20
0 cluded in the unit , thus saving many
~ 30
~ 40
' '\.. external components. Pin 7, marked "mu-
ling," ca n be switched to ground to disa ble
w
.... 50 ' "\
the audio o utput. If manual contro l of the
~ 60 . gain of the unit were desired (instead o f by
the SL620 unit), a potentiometer can be
0 0.4 0.8 1.2 1.6 2 .0 connected between pins 9 and 2 with the
AGC VOLTAGE (V) wiper arm going to pin 8 .
The circuitry o f the SL620 u nit, o n the
Fig. 3. C6n!rol range of !he uni! excees 6? dB.
other hand, is quite differe nt than most !Cs
The graph illustrates the control of the . mp_ut
signal by the SL630 with age volt age supplied oy because of its specia lized fun ctio ns. T 1
the SL620. through T4 are the input af amplifie rs. The
af output is couplea to a d e o utput
amplifier (TI 6-T 19) by means of two
sounded more like the type o f qua lity det ectors (Tl4 and TI S). Tl4 in co njun-
speech processing found o n commercial ction with C l has a short rise and fall time
circuits rather than the usual harsh, noisy co nstant. Tl S in conjun ction with C2 has a
co mp ressor t ype action , which usually has long rise and fall time constant. Thus, any
to be disabled for local contacts because input signal will rapidly initiate age action
the inherent distortio n then becomes so via T 14 (in 20 ms), but after a lo nger t ime
noticeable. The squelch and noise immu - ( 200 ms) Tl Stakes over to co ntro l the age.
nity featur es added also a great deal to the T he effect is ra pid initial age respo nse but
cleanli ness of the speech o utput and should no t fa lse age response to sudden peaks
be particula rly useful in a mobile ty pe after the speech input has sta rted. T6-T8
situation where a great deal of ex traneo us form a trigger circuit wh ich detects sudd en
background noise ca n exist. peak inputs above 4 mV , such as noise
T he IC Internals bursts. When such a burst occurs during a
pause in input it prevents via TIO and Tl 3
The circuitry of the SL630 audio ampli-
the output fro m b eing turned on .
fier is no t too different from that of many
T9 in conjunctio n with C3 forms a sort
IC aud io amplifiers, except t hat it includes
of me mory circu it having a time co nstant
provision for age co ntro l o f its o utp ut over
of abo ut I seco nd. So lo ng as a speech
a wide ra nge. T he input is co upled to a
input is presen t, il does not act but during
different ial amplifier direct ly without the a pause exceeding o ne second , C3 dis-
use o f coupling capacitors for a balanced cha rges to turn o n Tl2 via T ll and turn
input. T he age voltage (pin 8) contro ls th e off the aud io o ut pu t o f the SLG30/SL620
emitter current flow return fo r the differ- combinat io n. The capacitors mentioned
ential amplifier via the transistor whose above for the various time co nstants, Cl ,
base goes to pin 8 and a 3.6 kr2 resistor. C2 and C3 are external to th e IC and their
T he 7 SQQ resistor between ba se and co llec- va l ~e can 'be experimentally changed to
tor acts as a "linearizi ng" element to give suit individua l preferences . W2EEY
the smoot h co ntrol range sho wn in Fig. 3.
The rest of the unit co ntinu es to provid e *If 1111i1s are 1101 available locally, write ro
amplification (40 dB overa ll ),_ ending in a Plessey Miaoelecrro11ics. 170 Fi1111 Street.
series-connected output pa ir . which ca n Farmingdale. l.I.. NY for locatio11 of
provide an outpu t up to 2SO mW if the nearest dis1rib11tor.
80
An Integrated Circuit CWID Generator
P. J. Ferrell, W7PU9
projecl lo develop an automa tic, have appeared in amate ur lit c rature~-n and
A solid-stale CW ID generator was the read e r is re ferred lo th e m fo r b;1ck-
recently initiated by members of the ground 111a1 c rial.
Seattle repeater group. A lth ough there Program Counrer a nd Start /S to p Flip-F lop
have been a number of recent articles This six-s tage ripple count er consis ts o f
concerning such devices, 1 - 2 - 3 -4 ou r start- thn;c dual J K fli p-fl ips. The first five
ing point was the FM Magazine article by cascaded stages arc the program count e r
Woore. 2 The outcome of this project m1:1st and count from 0 to 3 1. The last stage is
be classifie d as an engineering overkill. The e m ployed as 1he sta rt /stop fli p-flop. Euc.:h
resulting CW identification generator stage of a rip ple counter is arra nged to
features a clo cked c haracter generator (for toggle ( cha nge sta te) o n thc out pu t o f t he
flawless CW with variable speed), inexpen- prcc.:cd ing stage. A five-stage progra m
sive RTL integrated circuitry, and a count er has 32 d istin ct s1agcs (2~=32).
co m puter-designed di ode read-only Whe n arranged as shown in Fig. 2. th c
memory matrix. Also included a re "pulse" program count e r adva nces under th e con-
starting, a discrete "hold" voltage available tro l of gal e GI whic h de rives it s input fro m
during ID execution, and a continuously the c haracte r ge ne rato r. Each do t, d ash , o r
adjustable keying speed (from far too slow blank c harac.:ler advances lhe program
lo far too fast). Nol only are these gener- counte r bv o ne count. The last c harac ter
ators ideal for repealer identification , but (numbe r 32 ) rese ts o r clea rs the first fi ve
they may be used lo identify any amateur stages. hut toggles the sta rt/ slop fl ip-flop
station su ch as RTTY, ATV, etc. to the " sto p" o r se t positio n , thereby
The block diagram in Fig. 1 shows the ha lting the c harac ter genera to r.
major divisions of the ID generator. Many A positive pulse into the "preclcar"
excellent articles covering RTL logic design input o f PF6 clears the halt and all o ws the
STOP
START
STOP
START
Fig. 1. CW ID gen erator block diagram.
81
(o) INPU T S TO MEMORY
FROM
G3 (I) INPUTS TO MEMORY
82
TO GATE GI
CO I TO 0 .5 JJF
OUTPUT
CODE
c
33K l STREAM
+3V I MEG
STOP +OOT
(IN)
a positive "+ BLANK" input from the Diode Read-O nly Memory Ma tr ix
diode memory causes the dash generated This is the hard part! Each desired code
by the character generator to be blanked stream requires a distinct and different
read-only memory design. A +DOT voltage
polarity of lh (I) oulpul must be produced by the diode memory
FF5 o o o o o o o o o o o o o o o o for each program counter state that cor-
Ff 4 o o o o o o o o + + + + + + + +
FF3 o o o o + + + + o o o o + + + + responds to a dot in the desired code
FF2 o o + + o o + + o o + + o o + + stream, and a +BLANK is required for each
FFI o + o + o + o + o + o + o + o + blank character. An example is presented
0 I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 in Fig. 4.
+DOT >+ + 0 + 0 0 + 0 0 0 0 0 +++ Suppose that the first letters of the
+BLANK 0 0 0 + 0 + + 0 0 0 + 0 0 0 0 0 desired code stream were "DE (blank,
CHARACTER - : x x x - - x - - blank)W7". The program counter states are
D E W 7 shown corresponding to the required out-
puts from the diode memory matrix. A
Fig. 4. Diode memory inputs and ourpuls. dash is seen to be the "default" condition;
i.e. if neither a +DOT nor a +BLANK
out - which results in the transmission of a occurs, a dash results. Each of the 32
blank character. program counter states must be accounted
Thus, the role of the diode read-only for since they all appear on each ID
memory matrix is to provide just that execute cycle.
sequence of +DOT and +BLANK inputs to Figure 5 shows the method for decoding
the character generator which results in the the program counter. The five terminals
transmission of the desired code stream . marked FF I through FFS are connected to
Gate G4 is the output with a plus repre- either the I or 0 side of the respective
senting "key down and a zero represent- flip-flop. If any of the five connections is
ing "key up. low, then the whole common line is low.
Output
-f---f~f--rf-.t-IJliil!~~!No~
line
+3V~:..__
f
FF I FF2 FF3 FF4 FF5
(I or O) (I or 0) (I or 0) (I or O) (I or 0)
83
FF5 o 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 + + + + + + + + + + + + + + + +
FF4 o 0 0 0 0 0 0 0 + + + + + + + + 0 0 0 0 0 0 0 0 + + + + + + + +
FF3 o 0 + + + + 0 0 0 0 + + + + 0 0 0 0 + + + + 0 0 0 0 + + + +
FF2 o 0 + + 0 0 + + 0 0 + + 0 0 + + 0 0 + + 0 0 + + 0 0 + + 0 0 + +
FFI 0 + 0 + 0 + 0 + 0 + 0 + 0 + 0 + 0 + 0 + 0 + 0 + 0 + 0 + 0 + 0 +
0 I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
l - 0
l
E
l l l
w 7
l
0
l -
B
l
F
0 4 12 8 24 28 20 16 x x x x
I 5 13 9 25 29 21 17
3 7 15 II 27 31 23 19 x x
2 6 14 10 26 30 22 18 x x x
DOT MAP 5-INPUT BLANK MAP
KARNAUGH MAP
The onl y time the common line can be memory. The details of these methods and
high is when all fiv e input connections a re the underly ing theory are beyond the
high . For any given connectio n , th is will scope of this artic le, but for those who arc
occur exactly once during each program fascinated by this stuff, standard tex ts arc
counter cycle . This type of diode decoder available whi ch will q uickly dispel t he a ura
is often called an and gale since it has a of "black magic" that seems to surround
high output o nly when all inputs a re high. this area. 8
Because o f th e diode out put from the For our purpose, a graphical reduc tion
common line, these decod ers may be technique known as a Karnaugh map will
paralleled to obtain the required +DOT and be e mployed . F igure 6 illustrates the
+BLANK functions. This paralleling is process for the code stream DE W7DBF.
ofte n referred lo as an or gate since any T he po larity of the I ou tput levels o f
high input results in a high output. nip-Oops FF I through FF 5 arc shown as
In the example of Fig. 6, the desired the program counte r s te ps from 0 (all FFs
code strea m has 15 do ts and 9 blanks. If clear) to state 31 (all FFs set). The sam ple
we employ a separate diode decoder for code strea m begins with a blank and has
each one, then a tota l of 24 decoders three bla nks separat ing the DE from the W
would be required: 15 would be paralleled A Karnaugh map organization of pro-
to give thr +DOT signal and the remaining gra m counter states is presented , nanked
9 would be paralleled to provide the on the left by the DOT map a nd on the
+BJ..ANK signal. Each decod er requires 6 right by the BLANK map fo r the desired
diodes for a grand total of 144 d iodcs to code strea m. Reduc tion is accomplished by
build a straightforward read-only me mo ry "folding" the map about a ny of th e
using th is te chnique. dividing lines a nd pa iring the marks (dots
Fo rtunately fo r us, the English philo- or blanks) wh ich o verlap. For example,
sopher George Boole p ublished h is /u 11es1i- fo lding the DOT map about the vertical
ga1io11 of rile Laws of Th ough1. in which he centerli ne pairs dot 5 with dot 2 1, and dot
resolve d the ambiguity of the words and 9 with dot 25, and do t 15 with dot 3 1.
and or by means of a kind of a lgeb ra. In Successive pairing, then pairing pairs, etc .
1938, eighty-fou r years later , Prof. D. E. allows a reduction from the o riginal 144
Shan n o n ( th c l 11fo r111ario11 Theory diodes to a total of 48 diodes arnnged as
S/u11111011) put Boole's algebra o r Boolean shown in Fig. 7. Note tha t in Fig. 7, the
Algebra to use in the Symbolic Analysis of 0 /1 nip-flop li nes arc reversed for FF:! and
Relay and Swirching Circ11i1s. Th is c lassic FF4. Th is reversa l ma terially simp lifies
pape r has revolutionized sw itching design , printed cin:ui t const ru ction .
and has led lo the develop me nt of minimi- As an examp le o f pairing. consider 1hc
zation techniques which can d ramatically blanks in position' 0. 4. 8. and 1~ of the
reduce the diode count of our ready-only BLANK map o f Fig. 6. Ir the map is fo lded
84
NO GERMANIUM SILICON
CONNECTION DIODE DIODE
+ +
R 4.7K
~
R
~
-...r
R
~
-...r
R ~
Jl"'T
R
-
-11""1
~ +DOT
R
-
.,
-Jl"'T
R
v ~
TO
CHARACTER
GENERATOR
R
~
R
. ~
....-1
~ + BLANK
~ ~,
....
. R -.....,
'> '~ '> I\ 'l ~ I\ 'l I\ I\
+3V 0 I
'I 0 0 I I 0 0 I
FFI FF2 FF3 FF4 FF5
about the secon d verllrnl line (the one again they differ only in fF4 . In pan C. a
separat ing positions 4 and 12). 1hc11 hl;1 n k com parison of t he 0 /8 with 1he 4 / 1:::>pa irs
4 pairs with blank 12. and b lank 0 pairs s hows tha t these differ o nly in FF3. T hus.
with blank 8. If we fold again. then <ill four if a 1hree-inpu 1 d iod e dcc.:o dcr of the type
blanks coincide with each ot her. This presen ted in Fig. 5 were co nnected to f'F I
is illustrated in Fig. 8. In pan A. the CO). FF:> (OJ. and l F S (0 ). ii would give a
program counter Sl <iles for positions 0 and +Bl.A\'K for all four dcsi rcd program
8 arc compared. They differ in cxac lly o ne coun ter states ( 0. 4. 8. and 1 2 ). Withou1
1 1 posi1ion (F F4) as all pairs mus1. /\ this r,duc1ion. 24 diotks !six fo r each
single diod<' decoder of the type sh own in blank) would haw been required ra 1hcr
fig. 5 rnuld ge l bo1h blanks si mply hy tlwn the four aclually required. T his d iode
neglect ing to con nect to F F4 . I l even saves decoder may h e fo u nd in Fig. 7 as the firs t
o ne a11tl diode. line in lhc +B LANK gro up .
Pa rl B of Fig. 8 presen ls the sa me The foregoing example illustra t es the
cornparbo n for blank 4 and bh1n k 12. a nd use of a Karnaugh map. The actual reauc-
lion is performed as in F ig. 8 . It should be
no t ed that so me stat es will not pair, such as
FF5 00 0 00 0 00 0 blank 2 7 in Fig. 6 . To pick up this blank, a
FF4 o+ o+ ft.II five-input diode decoder is required.
FF3 00 0 ++ + 0 + From Fig. 6, we see t hat for blank 27, t he
FF2 00 0 00 0 00 0
program count e r is in st a t e "++O++" and
FFI 00 0 00 0 0 0 0
the resulting diode decod e r can be fo und in
08 4 12 Pairin g Fig. 7 as the bott o m diode line.
A and B
The rule w he n pairing s tates is tha t the
(A) (Bl (Cl
two program counte r states can differ in
Fig. 8. Actual reduclion procedure. exac tl y one flip-flop position. All o ther
85
EXECUTION BEGINS
D B 0 I 2 3 4 5 6 7 8 9 +
s G G R
s G G G G R
s G G G G R
s G G G G R
s G G G G R
s G G G G R
s G G G R
s G G G G R Fig. 9. Computer printout.
s G G G R
s G G G G G R
EXECUTION COMPLETE
positions must agree, including any compared. This greatly increases the
omitted connections as in Fig. 8, part C. If already great tedium of such calculations.
there is disagreement in more than one FF Computer Optimization
posi tion , then these two states do not pair. In order to minimize the pain of diode
After completing the reduction process, memory design , the task was subcontracted
check to make sure that every necessary to a digital computer. The Seattle repeater
state has been covered a t least once by one group is extremely fortunate in having
of the final decoders: otherwise , you may remote access to the University of Washing
be surprised at the resulting code stream. ton Computer Center's Burroughs BSSOO
This type of ca lculation, once understood, computer, one of the nicest hardware/soft-
is not particularly difficult , but it certainly ware systems ever put together. The resulting
is tedious and has lots of room for errors. program in extended ALGOL accepts the
Slight changes in the desired code stream desired code stream (in dots, dashes and
(even position) can have a huge impact on blanks) as an input and performs a com-
the diode count. plete Boolean reduction for both +DOT
For example, the diode count for the and +BLANK diode decoders.
code stream in Fig. 6 is 48. If just the DE is If the specified code stream is less than
slid one count to the right, a new code 32 characters (more than 32 characters are
stream is formed which starts with two not allowed), the n the computer assigns
blanks, and has two blanks between the DE the necessary trailing b lanks and performs
and Ill. The diode count for this new the required reductions. and repeats 1hc
stream is SS . If this new stream is shifted process for eal"11 shifl of 1hc ..:ode stream
two positions to the left , so that the two until all the trailinl! blanks haw become
lead ing blanks become trailing blanks, then leading hlanks. Thl" nHk stream version
the diode count becomes 85 . These effects having the smalkst d1mk L'Ollnt is printl"d
are unpredictable, and for complete out alonl! with diode aml n:sistor counts
opt imization each code-stream version and an :11.:tual map of the cnt1rt diod<'
must be reduced separately, Jnd the results rtad-on ly mcmory matn'( . l: xampks of 1hc
86
d igita l compu ter prin tout art: shown in Fig. ll EP 5701. Virtually any PNI' silicon t ra11-
9. sisto rs will functio n in t h.: clock ci1cuit. A
HE P 57 is a good choice. Bo t h germa n ium
Const ruc tion and sil icon d iodes a re used in the dio de
The fou r dual J K flip-flops art! Motoro la read-o nly memory. Ge rmaniu m d io d es arc
MC79 0 P (or HE P 57~ ) and the q uad e mployed for I he and fu n ctio n , si nce the ir
:!-in p ut gat e is a Mo to ro la MC7 :!4 P (o r lo w j un ct ion capaci ta nce will no t load the
0
Cl :ll (/)
:iimm
a a a a
rfr~
a
0..., )>
a
2 2 a>
C m -!
-
e -s..
..., )> -!
-! r
mm
a D a D I G
:ll
a I I II I I I I I I I
...
1~
II II II II I I I N
I I D D I I I I I I
D II D II I I I I I
i'
0 .l>
~
a .,. I D I I D I I vi
::J
~~
II II II II 2 II
I . .. II I I D I I I
I I D I DD I D I D D I co
~ ,;~ '"' I D I I I I I I I I I I
0
t. D D I I I I I I
."'...,,..
Circuit board (borrom).
0 (/)
+ t
0
~
..... -"
0 + <o ;o <o
m ..
..fl... OQ 0
3 z )>
~ "
-< .,. ~ m r 0 ...
UI )> l -< UI
-!> -! 0 .
Cl -
m0 c
)>
\>I )J )J
"'
2 m -! Cl l\J < I\)
m 2o )> .....
-!
:! s: ~ ")(
:ll m
)> .,, )> (/)
- i n -i
0 )> - ()
...{\.. s Q
:ll :! (")
G> CJ)
0
z 0 I:)
~
0
87
co
co
Here's what the board looks lrkc w hen the flatpacks and diode matrix are soldered in. The vert ical
placement of the diod es helps to keep the size down to this 3 x 5 in. circuit board.
JK flip-flop s in the program count~r. ca n he found and read 1hcm over care lully.
Either s il icon o r germanium diodes may be T he additional background material will
used in the or function. w11h s1 li<.:on signal amply repay the effort involved.
dio des prefe rred sin<.:c they result in a T he Seattle repeat er group can supply a
higher noise margin fo r the memory. Cheap m o derate number o f t inned epoxy glass
diodes arc availablt: from various solid-s tale c in.:uit boards for this ID generator. Th e
supply houses. Poly Paks feature s SO boa rd is not drilkd. but assembly instruc-
sili<.:on o r germanium diodes for S I. tions and a com put e r op1imized diode map
for the circuit boa rd is included. Be c.:crtain
Bo th sid.:s of 3 x S in. d o uble-s ided PC to specify the desired code st rea m, keeping
board :ire shown in fig. I 0 . A one-sided in mind the a bso lute limit of 32 c haracters
board was used fo r th.: firs! few models. (dots. da shes. and blank s ). Unit cost is $ I 0.
wit h ;1 second 3 x 3 in. hoard u sed to and they ma y he obtai ne d from th: Seattle
complete the matrix connections. This repealer group. 1 8~35 46th PI. S.. Seattle
""co rdwood"" const ruc t ion is a pleasure to WA 98188.
loo k al. bul a nightma re l o wire up. I f a
Ac k11 owle d gemen 1. The author expresses
diod.: go.:s ""west'" o n a w rdwood s1yk
apprcc.:ia1ion t o I-; 7F VO fo r art photog-
gcnaator, it is bes1 l o throw il away, since
raphy. and t o K?.\I WC for th~ snapsh ot.
un soldering about 80 dio d es and rcsis1o rs
and the n ge t ting t hing ~ back toge t her is Re fe re nces
t. I\ Digiral ldenrifica1ion Generaror, llarry
eve n w o rse than the initial cons truction T odd K6ZCI. 73 ~t .iga11nc . ~lard1 1967.
e ffort . /\ d o uble -sided e poxy glass PC 2. I C Repearer lden 1ifier, I om ll"oorc \\'l.16111 ~t.
board is recommended fo r the ID gener- I \I ~l;iga1111c. Januan t 969.
3. Au1oma11c CW l denrlfier John Connor'
ator. An operational gcncra1or s ho uld be II 6t\ YZ. llam K.1dio ,\lag:i1111c. Apnl 196\1.
endosed in a me tal box wi t h all leads .J . A D1qiral Morse Code Generaror Jerry ll:ill
K ll' LP.QS'l.J unc 1969. '
bypassed for rf. Even Vil F / U II F fie lds 5. Digiral Logic Devices, OST . July 1968.
have the ability to drive RT L logic <.:ir- 6. MicroLogic for Non -logic Users Hank 01,on
cui1ry absolutely crazy. Be fore undertaking W6GXN. 73 ~h1guinc. June 1967.
7. The M1cro-Ulrimaric, 1 0111 Pkkcring \\l l Cl W.
the cons tructio n of this ID gen.:ra1or. the 73 ~lagJLine. June 1966.
builder should obtain as many of the 8. Log1ca"1 Design of Digiral Compu1ers, Mon1-
rden:nccd I D generator anicles 1- 2 - 3 - as
4 !!0l11Cr} Phi-ier. John ll"il c~ & SoO\. 1958.
89
IC Square-Wave Generator
90
a reliable astable multivibrator circuit which
produces rectangularly shaped waves.
As stated earlier, the method used for
varying the frequency destroys the output
wave's symmetry. When R 1 is varied, the
output may change from a square wave to
a rectangular wave or pulse, for instance.
or course, this change in wave shape has
no effect upon the frequency as it is varied.
In order tq correct the wave shape, the
output of the multivibrator is fed into a
Fairchild L 923 J-K flip-flop. The aclion
of this flip-flop is shown in Fig. 3. The
L 923's output changes on ly when the input Fig. J. Operotion of the L 923 flip-flop. Note tho!
signal goes negative. Notice that the output rogardleu of the shaps of tho input waveform, the
is always a perfectly symmetrical square output is always. a perfocHy symmetrica l squaro wavo.
wave, regardless of the shape of the input
waveform. The input can be spikes, pulses,
rectangular waves or any other waveform stance, by feeding a 100 kHz signal from
which has a fast negative going portion. It your receiver's crystal calibrator, you can
can also be seen from the diagram that lock the generator at 100 kHz, 50 kHz,
33 kHz, 25 kHz, etc. Of course, this will
result in excellent frequency stability and
accuracy. Be careful not to feed too much
signal into the sync tcnninals, as you might
cause erratic operation.
The output level is controlled by a 5-k
pot at the output of the flip-flop. The actual
t .u
value isn't too important as long as it isn't
too low. Otherwise you might load the flip-
,....
""'"
LOO
.....
fl op too much. Don't go below l k. The u se
of a log-taper pot will permit adjustment
down into the millivolt region for low-level
.... audio work. The output voltage is about
two volts into a high-impedance load.
The supply voltage for the unit is taken
from two 1.5 volt D cells in series. Current
drain is less than 40 mA. Remember that
pin 8 of both !C's is connected to the +3 V
and pin 4 of poth is connected to ground
or minus. A colored line or flat portion on
the edge of the ! C's body identifies pin 8.
Diodes D 1 and D 2 can be almost any signal
Fig. 2. Complete square wa..,o 9enorator. Bendswitch-
diode. Parts values should be followed rather
in9 capacitors are IOj/0 or boHor tolerance. Resistors.
oro I/ wott.
closely to insure adequate band coverage.
the output frequency is one-half the input
frequency. This means that the multivibra- Construction
tor is actually operating at twice the fre- The printed circuit layout is given for
quency indicated on the front panel dial. those who want to make their own PC
The generator puts out a beautiful square boards. You can get an idea of the front
wave to I MHz and beyond. A slight amount panel arrangement from the photo. Actually
of overshoot on the rising portion of the the re is nothing critical about layout or con-
squa re wave is normal at high frequencies. struction so you can arrange things inside
The S (set) and C (clear) inputs are both to your liking. I used a 4 x 5 x 6 minibox
grounded, and the P (preset) input and for my unit, which is just right if you use
the Q output arc disregarded. a Millen 10039 vernier dial as 1 did. This is
Synchronization pulses are fed into gate a compact unit, and using a larger d in!
C 1 for locking the generator's frequency to will mean using a larger cabinet. The Millen
some external source or oscillator. For in- dial is rather expensive and maybe you'll
91
;.
l_L1_Lr~ 8
1401
l_Ll_Ll e
TT-rTT l TTTTT
o.T ~
2.2k
SYNC
Fig. 4. Full sito layout of PC board. This is a bottom view with components mounted on top.
want to use one of the imports and make pacitors you'll only have to calibrate the
your own scale. Since I'm on the subject of lowest frequency band. On each succeeding
cost, I might as well say that the whole band, the frequency is ten times the fre-
project will come to about $20.00 with all quency at the same point on the previous
new parts, including the ~lillen dial. With band. The simplest method of c:ilibration
an imported dial, you can probably knock is with a 60-Hz sine wave, which can be
$5.00 off that figure. supplied internally on most scopes.
To calibrate the lowest band (band A)
Cali bration the following procedure can be used. First,
You might have noticed by now that the allow tJic scope to warm up for a few minutes
scale on my dial is not linear. This is be- until it becomes stable. Turn off the in-
cause I used a linear taper pot for R1o since ternal sync of the scope. Apply the 60-Hz
it was available. fm not particular alJul sine wave to the vertical input of the scope
such things but if you prefer a more linear and adjust the sweep frequency until you
scale, I would suggest trying a log- or scmi- obtain six full crcles on the screen. Since
log taper pot. Keep in mind the fact that you are not using the internal sync, you'll
most vernier dials turn only 180' as opposed have to adjust it very carefully to stabilize
to the normal 270 turn of a pot. You might the pattern. With six full cycles on the
have to adjust the position of the pot in screen, the sweep frequency is now set al
the dial to insure proper bandspread. 10 Hz. '.'/ext, feed the square wave from
Calibration can be achieved only by the the generator to the scope and tune the
use of a scope or frequency counter. If generator's frequency until you obtain one
you don't own one maybe you can gain full cycle of a square wave. De careful not
access to one for about fifteen minutes or to move the sweep frequency of the scope.
so. By using 10%, or better, tolerance ca- The square wave generator is now set at
I 0 Hz and can be marked on the dial.
Tune the generator again until two full
cvclcs arc visible on the screen. The scale
c:m now be marked at 20 Hz. This process
can be continued on up to 100 Hz. After-
wards, go back and repeat it all over again
to make sure you have the proper calibra-
tion. Once you have made certain that there
arc no errors, you can mark the rest of
the bands as outlined previously. On band
E, you can listen to the signal Oil a broad-
cast receiver to sec if it checks out. The
bands Oil m)' unit did not exactly come out
in multiples of ten because 1 used 20% tol-
(o) (b)
erances capacitors from my junk box. Even
Fig. 5. Bosic diagram for tho IC's. so, they came uul very close.
92
o o
0
0 ......
0 ..-...0: l(j o o o o o IJ
I
0
'f-:-
Fig . 5. Full site printe d-ci rc uit board used in the IC squa re-wa ve ge ne rato r.
93
An Audio Sinusoid Generator
Howard Phillips, W6FOO
Do you need an audio ge nerator which which can be connected in series for LI = 88
delivers a sine wave over the' frequency range mh o r o nly one of the windings can be used
of 500 hz to 5 khz? The osci llator described for L2 = 22 mh.
here will do the job while maintaining a The o pera tion of the oscillator is better
nearly co nstant output voltage over t he understood by considering the functional
range of I khz to 5 khz. diagram illustraterl by Fig. 2. The current
An aud io oscillator which generates a which flows through L2 is in phase with the
low-harmonic-content sinusoid has many voltage developed across the tuned circuit.
uses. Examples include two-tone generators The output voltage from the X 1 comple-
fo r use in amplifier linearity tests and AFSK mentary amplifier is divided between L2 and
oscillators for encoding Teletype signals for LIB. T he d rive voltage to L 1B does not vary
radio transmission. The oscillator described with frequency between 1 khz and 5 khz.
here is well suited for the latter application
because swit ching different capacitors into
the circuit allows generating two equal
amplitude non-simultaneous tones using the
same oscilla tor.
OSCU. .tif04
94
Fig. 6 shows a graph which can be used to
select a value of C for any desired oscillator
frequen cy. The error bars above and below
the plotted data points reflect the 20%
tolerance of the capacitors used to obtain
the data.
The choice of semiconductors is not
critical except that silicon transistors and
silicon diodes should be used. The bias
conditions arc designed to make use of the
forward voltage drop (0. 7 volt) across a
silicon diode. The power supply voltage
values are not critical. If the semiconductor
I
-
distortion of 4% at 3.5 khz and a 1.4% \ I
waveform distortion at 500 hz. Filters can
\
be used to decrease the waveform d istortion
to less than 0.1 % if required.
The o utput voltage is plotted as a
function of frequency in Fig. S. The de-
- -r
I
-
crease in output voltage at lower frequencies L-.
I ~
is cau sed by : (a) The Q of the resonant II
circuit decreases at lower frequ encies; (b) I \
the inductive reactances of L2 and LIB
decrease at lower frequ encies, and the ..
&: '-
0 '- ~:
~;
.
e- " -
voltage drops across the I 00 emitter
resistors become appreciable. The oscillator
can be made to operate at lo wer frequencies
. ) L- !
I
.. -
-
I
I I
I
I
I
I)
~
by increasing the values of LI and L2.
' 1 s ~ a.1as10 1 1 ~
95
A Six Meter IC Converter
G. W. Reynolds, K2ZEL
TIHS convcr1 er 1s onl' m odule o f a solid with 5/8" nylon (plast ic) standoff in sulat o rs.
'late mobile recl'Jver :J JHI transm 111er de- The met all ic s hi eld s were made o f sheet cop-
sig ned . construc ted :Jnd test ed by thl author. pe r and soldered to s uppo rt intt vecior b oa rd
The advent of solid stale devices make ii pos- clips . All rC$iStors used arc Y.i watt dissipa-
sible to desi[!n 111:;:hl y stahlc receivers an d tion and th e capacitors arc eit her disc cera m-
tr:Jnsmilla\ 1d11ch 11ill operat e directly from ic or mylar printed c ircuit type. The Jacks
the car b:J tt er)'. The 1iiot1ul:Jr system was sel- for wuplin g t he convert er t o the an t enna and
ect ed 10 insure adcquJ lc' 'h1cld i11:;: between to the tuna hh: 1f module (o r receiver ) arc
sec tio ns o f the receiver and the transmitter stand:mt ON(' j acks. Coil forms arc 1/.1" ny-
a nd to facilitate ch.111 j!es 111 the ,y,1e111 :Jt a lon rod. Co mpo nent layou t is s h ow n in
future d.11e 1f desired . F.ig . I.
The convener is con, tructcd of 111 0 .. !Cs" Align ment Pro cc dur~
.111d a">ociatcd componen1'. The one 1s Jn rf After comple ting constructio n use a g rid
ampli fier o t JO dh (!.tin a nd the s~co1H.l i~ a dip meter or tunne l dipper to adjust the fre-
crystJI o'cillator. 1111xer. am plifie r s tJge of quency of the LC combinati o n s to the ap-
.J O dh i;;orn . 1-valuJllnt! lll>crlion los,es. the propriate frequencies per chart.
u,aall gJin ot the n111verter is 55 d h. The
T uned Circu it Frequency
factors .con,11kred 1d1en selecting the parti-
LI CI 50 mhz
cu1<1r " IC" U\Cd Wl' rc cost , availability and
L4 C2 50 mh z
su 11 ahili1 y . 1 he l:11rd11ld 70 3 opcrn tes as a
LSC3 50 mhz
differc n t1Jl Jmplif1 lr ,,f 20 to 30 db ga in at
L6C4 Xtal Freque ncy
frequenc ies up to 200 mhz and 1s inexpen- L7C5 I F o ut.
sive and readily ava1lahlc . The RC A CA-3028
"'more ex pe nsive and can be used as a n rf dif- The if' o utput will b e the di ffer en ce be-
krcntial amplifier wi t h nea rly the same ga in twee n the c rystal frequency and the 50 mliz.
fi11ure. hut its advantai;c in this converte r is Connect the co nve rter to a receiver set
I hat it can be used a> :1 mi xer . local I crys tal and apply p ower to t he con verter. Using the
or tuned ci rc uit controlled) oscillato r and grid dip o scillat or in the wavcmetcr condition
amplifie r of 40 db i;ain. rhc c ircuit employs adju st C4 for maximum rf ~ignal when the
a thi1d overto ne crystal in the series reso nant di p per is near L6.
frt'd hack p osition o f th e oscillato r circ uit . Next apply a 50 mhz mo dulated Si!!nal
The shield p!Jccd .1cross IC is necessary
t o prevent 111duct1vc feedback and sclf-oscil-
l.11Jo n of t l11s .. IC .. The tuned circuit LI C l
is provid ed to mat ch the ant e nna inrut t o the
con vcrtc1 .111d link co upled th rough a sh ield
10 rcducc c ross-talk .ind birdies bein g fed
throu j!h the co n verter to the tunable if mo d-
ule (or receiver).
-- 4 ~1- - - - -- -
Con st rue I ion
Con st rucl ion was ma d e on a vector board Fig. 1. Componen t l ayout .
u sintt vector hoard di r s as mounting termin- fro m a signal gcncrafor t o pin 2 of IC2 and
als. Thi' form o f con~ truction permits sub- adjust the receiver for the signal at the if o ut-
sti1uio n of clcc t ricnll y similar com p o n e nts put. Adjust CS for maximu m output. Dis-
w1 tho u1 detrime nt t o the circuit co nfigura connect the signal generat o r and apply signal
t ion. The vector board was cu t t o fit the to Ji and udj u st first C3, then C2, and then
stand ard 5"x2 1/.i"x2 Y.i " 111inibox a n d mo unted C l for maximum signal ou t put. Repeat ad-
96
'"
L7
f'Mf---':,........_ ___
.... IZV SI
.___...__.~ ___,
Fig. 2.
L
1
ST Y." Oia. closewound L7 JOT Y." dia. Y 1 Xtal 43 .5 mhz for 6.5 ic .5 mhz jf
L 2T Y. " dia. " Le 3T X" dia.
2
L 2T 14" d ia .
3
L ST 14" dlo. c SO pf A Y. watts
4
Ls STY. .. diu. c 21 SO pf C 50vdc
Tap 2T c 3 SO pf C1 4 745 pf trimmers
L GT Y. " dia. C 4 SO p f
6
Tap 2T Cs 100 p f
justments CS, C3, C2, C l wit h th e generator signal generato r and connect the anten na.
at antenna in put (J 1) because of slight inter- Happy listening to six meters.
action between adjustments. Remove th1; ... K 2ZE L
97
An IC Audio Notch Filter
A bridged-T audio notch fucer combined achieve reasonable narrowness and high at-
with a n IC amplifier produces a highly tenuation at the notch freQuency, expensive
versatile, wide-ra nge audio rejection filter capacitors were necessary and the filter
with both variable frequency a nd variable could on ly be used in very high impedance
"Q" contro ls. circuits. The use of an integrated circuit
Audio filters are certainly nothing new. amplifi er with a notch filter in a feedback
They have long been used to in:iprove the arrangement, howeve r, produces a notch
selectivity of a receiver or transceiver when filter of very high Q using inexpensive
it was not desired to "dig" into the i/ components. It is neither critical as to circuit
circuitry and improve the selectivity a t the rf impedance nor does il introduce any overall
level. The diladvantage of such a method of circuit insertion loss.
selectivity iaiproverrit:nt is that the selec-
Basic Circuit
tivity takes place late in the receiver pro-
cessing chain. Therefore, when one is lis- The circuit of the IC notch filter is shown
tening to a weak station, a strong station in Fig. 1. The actual notch filter consists of
near in frequency can control the ave or the bridged-T network - the ganged 50 K
overload the receiver stages. potentiometer and the two .05 mf capac-
Nonetheless, audio selectivity is easy to itors. The circuit presents a very h.igh attenu-
apply and can take the form of either an ation at one frequency wltich is related to
aud io frequency peaking or notching type the time constants. of the circuit. The fre
fu nction. Audio peaking can easily be pro- quency of maximum attenuation can be
vided by a number of fixed frequency filt.:r changed by either changing the value of the
designs and numerous inexpensive units are resistive or capacitive legs. As was men-
available from surplus outle ts. The disadvan- tioned, however, unless special precau tio ns
tage of the peak ing approach is that most are taken, t he bridged-T network alone will
filte rs which are of any real use produce a tend to prod uce a very broad rejection no tch ,
"ringing" effect. The sound is unnatural and which is particularly unsuited to CW work.
definitely very tiring if the filter is con- The integrated circuit, however, corrects this
stantly left in {he circuit without any pro- situation in the following manner: the input
vision for disabling it. Stations can also be signal passes through the bridged-T net work
lost when 5canning a band unless tuning is to one input of a differential operatibnal
done very slowly when using the filter on amplifier (a Motorola MC 1533 in this case).
cw. Feedback through R4 is coupled from the
The no tching type filter, on the other amplifier output back to the signal input
hand , is usable on both CW and phone. It point. The other amplifier input (-) then
does not cause any ringing effect and does receives a combination of the original input
not mask any signals when quickly scanning and feedback signa ls. Since the nature of a
a band. A single notch filter can only differential amplifier is such that when the
eliminate the o ne freque ncy to which it is (+) and (-) inputs receive equa l level signals,
set, but when a receiver already has a good the output is zero, R 2 and R3 are chosen
phone filter - such as the multiple crystal or such that this condition exists and inftoite
mechanical types found in SSB trans- attenuation takes place al the notch
ceivers - one notch frequency possibility frequency.
seems to suffice in most qrm situations. There may be situations where a some-
Audio notc h fil lers built around passive what broader rejection notch is desired with
components o nly are fairly old, but their use correspondingly less max imum attenuation
has disappointed many operators because to at the notch freq uen cy. This adjustment is
98
provided by making RI variable, as shown in Fig. I shows the use of a Motoro la
Fig. I. As the wiper arm on RI moves from MC I S33 operational amplifier, bu t almost
right to left , the feed back voltage aro und the any similar un it will suffice. A number of
operational amplifier decreases and the ef- inex pensive "surplus" IC o perat ional ampli-
fective Q of the bridged-T netwo rk is re- fiers arc available from such suppliers as Po ly
duced. T hus, if desired, RI can be .made Pak s, Lyn nfield, Massachusetts 0 1940. Other
variable and fu nct ions as a "Q" control. un its may differ in their voltage req uire-
Othe rwise, RI can be a fixeQ value resistor ments and rolloff compensation needs (the
and R2 is connected to the junction of RI RC network between pins 9 and IO on the
and R4 to produce a single frequency, deep ~C l S33}, but t his information is usually
no tc h audio filter. In this case, the only supplied with the unit. It should be noted
variable contro l would be the dual SOK o hm that a simple integrated circu it audio ampli-
pote ntiometers which vary the time constant fier cannot be used ; such units do not have
of the bridl(ed-T network a nd hence the differe ntial inn11 t circuits.
notch freque nc y. The dual potentiometer is
capable of varying the notch frequency over Construction
about a I 0 : I frequency range - 300 to
3,000 cycles approx imately. This range T here are very few precau tions to be
should cert ainly suffice for most applica- observed in constructing the unit because of
tions but, if desired, the range can be the nature of its operation. One possible
changed by using differe nt (but equal) values method of construction which the author
of capacita nce for C l and C2. explored is shown in Fig. 2. All of the circuit
The differential o perational amplifier components arc mounted on a piece of
used may seem ~ n unusual component to vcctorboard which in turn is mounted on the
rear pote ntiometer of the dual SOK ohm
IOOK
po tentiometer. A piece of foa m plastic
Cl
00
material is glued between the underside of
.
Q CONTlllOL
the vcctorboard and the rear potentiometer
to achieve the mounting.
, The parts layout shown for the vector-
A.UDK> IOIC
INPUT ......,,..,..._...._.._.WY--+-WY.._-< board in Fig. 2 need not be followed
1- 101(
OHMS AUDIO exactl y, a lt hough it was the simplest which
OU Tl'\IT
IKOHW t he author could devise. If an in tegr.ited
circuit packaged in a dual-inline containe r is
used {rectangular with S-7 connections o n
* O[Tfl'IMIN[
tt ,_, VAL.U BY
,,
l'llllST USIN Cl IOK POT
99
The resistors used in the notch filter can Mounting and Operation
either be Y. of !12 watt sizes. The capacitors The mounting or placement of the no tch
need be rated no higher than the maximum filter in a receiver or transceiver is fa irly
value of the supply voltage used for a nexible. T he unit can be inserted be tween
particu lar IC. The capacitors used in the almost a ny two audio stages. DC blocking
bridged-T network sho uld be of good quality capacitors must, of course, be used to
to achieve the sharpest notc h selectivity . prevent other than audio frequencies passing
Disc ceramic types are acceptable, although, through the filter. Alternatively, the unit can
if possible, low-loss types such as the be used completely ex ternal to a receiver or
Aerovox Pl 23ZN series are preferred . The transceiver. The audio output fro m the
current demand from the power supply is in receiver or transceiver can be taken from a
headphone jack or fro m the lo udspeaker
terminais through a transformer (4- 10 ohms
to I K o hm or more). The gain of the
operational amplifier, in the latter installa-
tion, is more than sufficient to drive any pair
TUtMINAl$
NUMH,_(0 of medium to high-impedance headphones.
AS SUN
ntOM Operatio n of the no tch filt er is ex tremely
fO,S IDl
simple. When not in use, the filter is adjusted
fo r maximum low-frequency attenua tion. As
qrm develops, the filter is used (in con-
j unction with tuning of the receiver
bandpass) to eliminate the most severe inter-
Fig. 2. No partic ular ci rcuit layou t Is fering beat (on SSB o r CW). The result is an
nocossarv but the components eon so com almost complete attenuation of the inter-
pactly be grouped o n vocto rboard thot the fering signal while still retaining the full
ontlro circuit mounts on tho back of tho fidelity of the desired signal. The difference
frequ ency contro l potontlomotor .
between this method of qrm e liminatio n a nd
that which depends u pon a severe reduction
the o rder of a fe w milliamperes, but the
in bandpass to a ccomodate only the desired
o perating voltages should be obtained from a
signal is q uite startling in terms of fid elity
well-filtered source to avoid a ny possible
and ease of tuning.
hum problems.
The dual SOK o hm potentiometer used in Summary
the bridged-T circuit is a standard linear Many thanks a re due to Herman Gelback,
taper type. Although one can purchase such W7JPU, of the Boeing Company, who com-
potentiometers from various supply houses, pletely developed the original circuitry of
particular attentio n should be paid to being the IC notch filt er and who a llowed the
"cost-conscious" abo ut this ite m. The autho r to present this description of its
a uthor has fou nd suc h potentiometers avail- operatio n. Herman has also suggested that
able for as low a s 35 as compared to prices besides the MC I 53 3 IC, another suitable
of SJ for similar units at regular catalog off-the-shelf IC would be the General Elec-
prices. tric PA-230. which sells for just over $3
100
An Integrated Circuit Electronic Counter
Front viow of tho int~gratedcircuit froquoncy counter. The noon counting docados are on tho left count
control' aro on tho right. '
101
RESET OUT
+200
OUT
+200 '""
"'"'
"'"'
000 ....
I HI,
,,....
OR
l'IJ\.S[S 3V
I MHl OSC 0
0 ~ POW[A
AHO 2 ~1 TO ALL S(CT IONS
SUPPLY
OIVIOUt 0
+2ZUUO) UV JO HU WAJUCU!:S
60 ... TO ltCVR
Fig. I. Blocl diagram of tho completo inlegralad circ uil fr oquoncy counter. Any number of decades moy be
usod, but for proper display, tho units decade should bo lo tho right, the Ions decade lo ifs loft, olc.
changes. Therefore, it combines the con- units decade is allowed to count, one is
venience of an analog display with the ac- passed on to the tens decade, like,vise for
curac) of a digital display. The last digit each ten pulses the ten decades receives it
in this case usually vacillates between two passes one on to the hundreds decade, etc.
adjacent numbers because of the one hertz The decade counters, after the units de-
per gating period error inherent in a digital cade, are not gated since they only receive
count. pulses if the units decade is supplying them.
The counter consists of three main sec- Although the decades count by binary flip-
tions. First, a frequency divider divides the fl ops, suitable feedback circuits make them
signal from a l :\1Hz standard down to 10 count in decimal instead of binary. A decod-
kHz, 1000 Hz, 100 Hz, 10 Hz, or 1 Hz, as ing network and ten transistors allow one of
required. A time base derived from the 60 ten neon lamps on the decade to be turned
Hz line could have been used but this would on to display one digit of the measured fre-
have limited the accuracy to 0.1% and quency. Each decade can also be reset to
would only have permitted the 10 Hz and 1 zero by n reset pulse from the control sec-
Hz ranges. This section also applies 10 kHz tion.
markers to the remainder of the frequency
measuring setup. The 10 kHz pulses are rec-
tangular in shape and have strong harmonics Digital integrated circuits
above 30 MHz. Therefore, they might as The rounter uses RTL integrated circuits
well be used as markers. because of their low cost. These have been
Second, a control section takes the desired described in 73 magazine both in articles
time base frequency and turns on the units and integrated circuits, and in two excellent
counter for the correct length of time. It articles about TC electronic keyers; therefore,
also shapes the input signal, so thnt the they will only be described briefly here. The
units counter will accept it, turns on the render who is not familiar with RTL circuits
high voltage supply for the neon lights dur- should review these references before try-
ing the display period, and supplies a reset ing to understand the counter in detail. He
pulse to all counting decades at the end of might also find it advisable to build the
the display period. "Kindly Keyer", before he builds the rount-
Third, the counter proper consists of as er, as the writer did. Although the rounter
many counting decades as the builder de- could probably be built and made to work
sires, one for each digit to be displayed. The by just following the diagrams, a previous
units decade is gated by the control section knowledge of RTL circuits, gained by build-
and only counts pulses when the control sec- ing a simpler device, will help in trouble
tion wants it to. For each ten pulses the shooting.
102
Oscillator and frequ ency dividers
A 1-MHz crystal oscillator is used as the
main frequency standard at WlPLj. One
MHz is used instead of the usual 100 kHz
because a 1-MHz crystal gives better stability
than a 100-kHz crystal if one wants to pay
a reasonable price for the crystal. This is
probably because the 1-MHz crystal can be
AT cut. The oscillator and a divider to 500
kHz are mounted in a separate box so that Fig. 2. Tho oscillotor ond 2: I fre qu ency divide r used
with the decimal counter. This unit was built into a
the oscillator can be kept on a ll the time
se parate box ond may be used for ob taining markers
for better sta bility. Also, 500 kHz can be os doscribod in tho t ext. Although the FET is o Mo-
used for other purposes including future torol J N 126, n MPF-105 is loss oxponsive nd
p lans to use it lo synchronize a phase-locked would probobly work. IC I is o Firchild 923 or ono.
oscillator for the first conversion of the re- ho ll Motorolo MC-790-P; IC2 is o Foirc hild 900
ceiver. If the builde r already has a frequency or one holl of Motorolo MC. 799-P.
standard, it is not necessary to build an-
other crystal oscillator for the counter- the
existing one can be worked in easily. Con- The first three 5: 1 dividers are identical
versely, if the builder is interested in fre- except for time-constant values. The output
quency measurement but does not yet want of the 10-kHz divider is fed through a buffer
to build the counter, he can build the oscilla- to the station receiver and frequency mea-
tor and the dividers down to 10 kHz and at suring equipment. The marke rs are very
least have markers for his receiver. The strong through 30 MHz, the limit of the au-
oscillator and first divider arc shown in thor's receiver. If the receiver calibration
Fig. 2. . cannot be trusted to 10 kHz, the 50-kHz
The remainder of the frequency divider test button, shown in dotted lines in Fig. 2
section, Fig. 3, consists of 2: l a nd 5: 1 di- and not used by the aulhor, can be provided.
viders. The 2: 1 dividers are simply J-K Pushing this button makes the 50-kHz mark-
flip-fl ops; the 5:1 dividers are J-K flip-flops ers louder and the other 10-kHz markers
with an RC network and inverter on the set turn into 25-kHz markers. The counter prop-
input which only allows every fifth input e r does not read correctly while this is being
pulse to produce nn output pulse. done, but this doesn't mntter since identify-
ing the markers is done separately from
The 5: 1 divider can be best understood making the fina l count.
from the diagrnm and waveforms of Fig. 4. The divider form 500 Hz to 100 Hz uses
Without an input signal, the inverter input a discreet high-beta transistor instead of a
is held high by th~ connection to positive gate, so that a higher resistor value and,
voltage thru R 1 The inverter output is, therefore, a smaller capacitor value can be
therefore, low so that low appears on the set used. The dividers to 10 Hz and 1 Hz use
input of the flip-flop. If the 0 output of the decade dividers, with four J-K B.ip-Rops
nip-flop is initially high, the first negative in order to avoid even larger capacitors.
going transition on the toggle input will make This type of circuit could have been used
it go low. This change will be passed on to for all the dividers nnd would have elimi-
the inverter through C1 and this will make 11ated the need to ndjust the divjders. The
the set input go high so that the 0 output circuit of these dividers will be described
cannot go low again when more input pulses in the section on the counting decades which
come in. C, will charge through R1 and, a fter use the same circuit.
a delay. the inverter output and the set The switch, S., selects the divider freq uen-
input will go low again so that the flip-flop cy whose period is equal to the desired gate
can respond to a n input pulse. lf the djvider time and is calibrated in factors, by which
is adjusted correctly, it will pass every fifth the counter reading must be multiplied,
input pulse. Other division ratios can be rather than in gate time. The X60 position
obtained, and maybe it would work with a takes the time base from the ac line instead
division ratio of ten, but the ratio of five of the dividers, and is useful in adjusting
makes the division rntio very stable. In fact, the dividers. For example, to adjust the di-
it does not go out of adjustment for a change vider whose outp ut frequency is 50 kHz the
in the supply voltage from 3.0 to 4.0 volts. input switch is set to 50 kHz, the multiplie r
103
!:!Q!l - Al.L r\.IP'1.0f'S utC r.t.lltCHllD )IAU.) 011
lfl WOTOltOL.4 MCTtoP
....
""
'"'
n
h--,~-~-,
10 HI --0
0 IOO Ht-<>
~~ MK Off : ~:: ::: S~T~l"'- l)H1
Fig. 3. Tho frequency divide rs used in tho IC counter. Integrated circuih IC I through JCS are one-half of
Fairchild 914's or part of Motorola MC-789-P or MC-724.P; IC6 is a Fairchild 900 or one.half e Motorola
MC-799-P.
switch to 60 and the counter should r ead one that is conducting turns the other one
833. This reading will jump around a bit, off. The positive half cycle of the input sig-
due to instability in the ac line frequency, nal will make IC8 conduct and once it is
but the reading for the 10-kHz divider will turned on, the high output from IC9 will sup-
only vacillate between 166 and 167. ply holding current through R. to keep it on.
The negative ha lf cycle will then over-
come this holding current and turn off IC8
Control section whereupon the holding current will be re-
The input selector switch, 5 2 (Fig. 5), moved and IC 8 will continue not to conduct.
selects the desired input which can be either A small amount of noise riding on the input
a signal input for measurement, or one of signal will not be able to overcome the hold-
the divider outputs for self checking. IC8 ing current and will, therefore, not make
and IC9 can be regarded either as an a mpli- the circuit change state. The resulting rec-
fier with positive feedback, or as a flip-flop. tangular wave is fed to the units decade at
They make the signal into a rectangular all times and the necessary gating is done
wave with sh arp edges and reject noise in the first J-K flip-flop of the units decade.
which may appear on the input signal. At Provision for gating already exists in the
any instant of time, either IC8 or IC 9 will J-K and it is simpler to use it than to do the
conduct, but not both at once, because the gating in the control section.
104
The remainder of the control section can IC7 and, reset can only occur if all three
exist in either of two states, count or dis- inputs to IC 4 are low. The trailing edge of
play. We will discuss these quiescent states the pulse still toggles IC 7 , however, and we
before . we examine how it gets from one to are in the display mode; displaying the
the other. In the count state the l output of number of input pulses that occurred be-
IC7 is high and the 0 output is low. If S 1 is tween two timing pulses.
not in the X l position, the high l out-put of The switch, S3 , is used if you want to
IC 7 turns on Q3 and turns off Q4, thereby count, or display, for a multiple of the basic
turning off the neon lamps. The low 0 out timing period. The switch itself does not
put of IC7 goes to the gate input of the units switch the counter to display or count, since
decade and allows it to count. It also turns only the timing pulses can be a llowed to do
off Q1 so thnt the "gate on" light will be this; rather, it prevents the counter from
illuminated and it per its the "gate pulse" going into the other state. The "display" po-
light to be turned on if a gate pulse is pres- sition of this switch is useful if you have
ent. The opposite conditions exist in the dis- just made a critical count and want to hold
play state. Power is applied to the neon it a few seconds to make sure of writing
lamps through Q., and a high output is sup it down correctly. It is also useful if the cir-
plied to the gate so that further counting cuit for blanking the neon lamps isn't work-
cannot occur, and both Q1 and Q 2 are turned ing or isn't yet built and you want to make
on so that the two gate lamps are shorted a reading on the higher ranges. In this case
and not illuminated. it is difficult to read the display on the auto-
To understand how we change state, as matic position because you will see both the
sume we are on display and S3 is in the auto- counting and the display, but placing S3 on
matic position. IC 1 and IC 2 form a mono "display" will hold the last count and allow
stable multivibrator which supplies the reset you to read it. The switch can be thrown to
pulse and the trigger for IC 7 The positive- "display" either during count or during dis-
going edge of the rectangular wave from play. In either case, a timing pulse 'viii still
the frequency dividers turns on IC 1 momen- switch IC7 from count to display at the right
tarily and this makes the output of IC2 go time, but the next timing pulse will not put
high. Furthermore, this holds IC 1 on until it back on count due to the high level on
R2 charges up ~ again, whereupon the out- the clear input. Also, the counter will not
put of IC 2 goes low again. The result is a be reset due to the high input of IC4 which
short pulse which occurs once every timing will hold its output low.
period. Since we are on display and auto- The count position of S3 is normally used
matic, this pulse will be passed by IC3 , IC4 , only on the XI position of S., and is used
IC 5, and IC6 , inverted each time, a nd appears when you wnnt a gate time of several sec
as a high pulse to reset the counters. The onds for an error of less than one Hz. This
trailing edge of the pulse from IC2 will tog- is useful in the ARRL Frequency Measur
gle IC 7 , putting us in the count mode. The ing Tests where it is desirable to use a 10-
next pulse from IC 2 will not reset the coun- second gate time in order to obtain an ac-
ters because IC 4 has a high input from curacy of O.lHz. With this arrangement,
Fig. '4. Tho buic 5: I frequency divider using e J-K flip-flop, nd RC circuit end an inverter, olong with the
waveforms.
105
,._OM J I, 9<>-------o~~s
OAIJTO 1-+-.~~0<S~P\.~~~---------<> ~T ~S
D<tAO<
TO SI,
+200
+3.)
'''"
.......
2.U
Z.2'
Fig. 5. The control section of tho digital freque ncy counter. ICI is a one-half a Fairchild 914, one-fourth a
Motorola MC-724-P or one-third a Motorola MC-792.P. IC4 is one-third a Motorola MC-792-P. IC2, IC3,
ICS, and IC9 aro onosixlh of Motorola MC-789-P, onofourth of Motorola MC-724-P, or one-half of Fair-
child 914. IC6 is a Fairchild 800 or one-half a Motor ola MC-799-P. IC7 is Fairchild 923 or onoholf a
Motorola MC-790-P. QI and Q2 are 2N3877's or Poly Pak 2N1893's.
if you start a ten second run and the signal will find that running a multiple second
starts to fad e, you can stop the test at the count is much easier than reading about it.
next timing pulse by throwing the switch Jn wiring the counter it should b e remem-
to display and still obtain a meaningful erc<l that the supply to the neon bulbs is
reading. To make a ten second run, you a 200-volt square wave because of the lamp
start with S,. on display, and throw it to blanking circuit, and also, the collectors of
count when everything is ready. The next Q, and Qz (Fig. 5) have 60-volt pulses on
timing pulse will put you in the count mode, thtm since they turn on neon lamps. Both of
hut the next one will not put you back on these must be kept away from the inputs to
display. the fC"s; otherwise, erratic operation will re-
Each timing pulse will flash the "gate sult. In particular, the 200-volt lead to the
pulse"' lamp once, and after it has flashed counte rs must not be cabled with the signal
ten times. you put 53 back on display. The and gate inputs to the counters and the leads
next pulse will put the counter on display to I1 and l e must be kept at least an inch
.md you will be able to read the frequency :l\\,ty from the leads of 5 3 . If the counter
in tenths of hertz. With a little practice, you shows any erratic operation which cannot
106
be easily explained, the blnnking circuit It is necessary to use discrete transistors
should be disabled by grounding the base of to drive the neon lamps at the present state
Q3 so that the lamps are on continuously, of the art, but these arc not expensive, ex-
and 11 and 12 should be shorted to ground. pecially if Poly Paks 0 2Nl893s are used.
A test can then be made to see i the trouble The transistors arc used as shunts scross the
still exists. Except for these precautions, no lamps. This makes gating simpler a nd also
other difficulties should be encountered with limits the voltage across each transistor. For
the unit. a given count one lamp must be on and the
Counting decades other nine off. The driver for the desired
Fig. 6 shows the circuit on one counting lamp must have low level on all its inputs
decade, including neon lamp drivers. The so that the transistor will not conduct, a llow-
gate input on the units decade is connected ing the lamp to light. The other nine drivers
to the gate output of the control section, must have high level applied to at least one
but the gate inputs on the other decades input; this will be sufficient to extinguish
must be grounded since each must accept the lamp, regardless of what appears on the
any pulses put out by the prececding decade. other inputs. The gating of the lamps could
The actual counting is done by four J-K hnvc been done entirely with IC's but this
flip-flops and, with the help of the table method was found to be simpler and cheap-
shown, the reader can follow the count as er, at least at the present sta te of the art.
an interesting exercise. The input pulse fol- A B c D E F G H
lowing the ninth count makes the decade Count B A D c F E H G
go back to zero and passes a negative tran-
0 H L H L H L H L
sition on to the next decade making it count
L H L H H L H L
once. H L L H l H H L
IC 5 through IC 12 are needed to amplify l H L H l H H l
the voltage output of the J-K flip-flops. The 4 H L l H H L L H
J-K's give only one-volt output with light 5 l H H l H L l H
external loading due to the fa ct tha t they 6 H L H l H L l H
internally load their own outputs. This was L H L H H L L H
not found sufficient to drive the resistor 8 H L l H L H L H
gates used for the neon lamp drivers. An 9 l H H l L H l H
inverter, however, gives almost full supply Tobie I. Truth toblo showing th e proper lovol on
voltage when lightly loaded and drove the each of the logic line' of the decade counter in
resistor matrix satisfactorily. Fig. 7.
107
no
220 lk IUSIST"S '-"E" l/2W
fll[.SISTO..S NOT MAflK(O
Afll( IO!k r.14W
o~5
~'
~'
o~a
Fig. 6. A typical counting decade. In thi cir~uit integrated circuits IC I !hrough IC4 are one-hall Mo-
torola MC-790-P's or Fairchild 923's. ICS !hrou gh IC9 are one-idh Motorola MC-789-P's, one-lourlh Mo-
lorola MC-724-P's or one-hall Fairchild 914's. All lransito" are 2N3877' or Poly Pals 2N I 893's. All neon
lamp are NE-2's.
In testing the decades, + 200 volts must lated and the ! C's will work on any voltage
not be applied unless all transistors, whichfrom 3.0 to 4.5 volts, although 3.6 10%
are in place, have neon lamps across them. is recommended by the manufacturer. The
Otherwise, if a transistor is not conducting,
power supply used by the author is shown
the collector voltage rating will be exceeded
in F ig. 7. An 8-amp transformer was used
since there is no neon lamp limiting the because it didn't cost much more than a
voltage. Also, if +200 volts is applied to a2-amp one in the same series. The 2-amp
decade but +3.5 is not, all lamps should unit would probably work and would save
light since the logic circuitry only acts tospace and weight. For the 200-volt supply,
short out the undesired lamps. No harm is anything from 150 volts on up would work,
done by this and it is a quick way to check although with anything much over 200 volts,
the lamps and driver transistors. If a lamp the 220K collector resistors must be in-
does not light under this condition, its creased or a dropping resistor must be pro-
driver transistor should be suspected first.vided. If this voltage is taken from a supply
powering other equipment, it must be re-
Power supply membered that the current drawn will be
The counter, as shown in Fig. 6, requires a 40 mA peak square wave at 5, 50, 500,
about one ampere at 3.5 volts and 40 mA at or 5000 Hz which may cause a buzz to be
200 volts. Neither supply needs to be regu- heard on the other equipment.
108
I
Construction References
1. Skeen, "Low Con P rtd1io n Fr rquency ?i.leaau re
The individual counting decades are built ment' ', QST, J anua ry, 1965, p. 32,
on See-Zak MM-492 boards and the remain- 2. Bran ine, "An Electronic Co unt er for Amateur U1e,"
13, D ecembe r, 1966. J>. 20.
der of the unit on a See-Zak MM-512 board 3. Pickering, "The Ml croU l1imn t lc", 73, June, 1906,
mounted on See-Zak R-25 and R-212 rails. p. 6.
See-Zak M-25 terminals are used .for the 4. Daughters, "The Ki nd ly Kcyer" , 73, July. 1066,
p . 46.
larger components, including the Fairchild 5. Lancaster, "U1inR New Low CoKt l ntegrftled Ci r
!C's. The hole spacing on these boards is cuit&", Electro nic World , lluch, 1966, p, 50.
0.2" whereas the Motorola !C's require 0.1" 6. Olson, " MicroLogic for ~on Loglca l U1ers", 73 ,
June, 1967, p, 58.
spacing; therefore, seven extra ~e inch holes 7. Sud in g. "'A n InttgratedCircuit F'requen cy Counter ,"
must be drilled for each Motorola IC inbe- 73 , ~o,ember. 1967. page 9.
109
100 KHZ Thin-line Pulse Generator
Digital integrated circuits arc an entirely whose harmonics a re usable to 432 MHz or
new kind of electronic component. These higher. And it can be built without benefit
fin ished, ready-to-go devices contain com- of special instruments and knowledge.
plex transistor circuits in tiny, convenient
packages. Until recently they weie too ex- The thin line pulse
pensive fo r one-off projects, but a burgeon- One rather surprising result of higher
ing marke t and competition between manu- mathematics is that a ll repetitive signals arc
facturers have brought some prices to the composed of hnnnonically related sine and
dollar apiece level. In fact, very good digital cosine waves. For example, the familiar
IC's are now available on the surplus market. square wave is composed of a funda mental
Perhaps because they are so new, it is freque ncy, which sets its basic repetition
hard to see applications for digital IC's out- rate, and of odd harmonics only of its funda-
side the computer nnd industrial control mental, which contribute to its square cor-
scene. It takes a little while to adjust mental ners. H the harmonics' amplitude or phase
perspective, too, before their input-output relationship is upset, the square wave is dis-
clrnracteristics begin to seem natural. Yet, torted. This feature makes the square wave
there arc applications for them which arc very useful for amplifier testing, but its har-
not very difficult. For instance, how about a monic content is not very good for frequency
frequency standard? standard applications.
Ordinary 100-kHz frequency standards are Now suppose that we start adding up sig-
usually audible up to a few tens of mega- nals of F, 2F, 3F, and so on, phased in so
hertz. A good one m i~h t be usable at 50 that they a ll reinforce each other once per
MHz. The circuit described here uses a dual cycle. Let's say thay are all the same amp li-
NANO gate to generate a 100 kHz signal tude. What would we get? See Fig. l A.
@ ~
100 KCS THIN LINE. GENERATOR
~
IN1: toNl:
ASHE ClttTRONIU LAB REF. t:nU
OUTl'llT
>-
*,.. ~
110
The five equal amplitude sine waves peak thin line pulse appears as a damped oscilla-
sim ultaneously at the beginning of the fun- tion. But these problems do not interfere
damental's cycle. Everywhere e lse, until near with constructing the generator, because
the end of the cycle, they are more or less the very simple NAND gate circuitry con-
out of phase. Trying to see what will happen, tains no critical elements or adjustments.
we try adding the first two frequencies.
Fig. Ill, the result, might suggest something How it works
to a mathematician. There are four circuit sections, shown in
Fig. 3. A 100-kHz crystal-stabilized oscilla-
tor sets the basic frequency, and a dual
NAND gate circuit converts the oscillator
output to a thin line pulse. A 1-Hz astable
generates the output marking signal. A 6
volt de power source is provided by n volt-
age doubler Zener-regulated supply.
Multivibrator oscillators are not ordinarily
very stable frequency sources. But if the os-
cillator is designed to run slightly below
required frequency, and an appropriate crys-
tal is connected between transistor base
terminals, oscillations are stabilized at the
CA) crystal frequency.
The crystal docs not change the multi-
vibrator's style of operation. It synchronizes
'II the astable to its own frequency, by trigger-
I
I ing the OFF transistor into conduction short-
I
I
ly before normal RC turn-on. The output
I
I
is a squarish wave with good fall time, but a
I long rise time as shown in Fig. 4A.
I
In passing through the first NAND gate
I,, I
the pulse is squared up and becomes slightly
unsymmetrical. See Fig. 4B. A differentiat-
(8) ing network, C7 and Rll,converts the square
Fig. I. Five sine wave' (A) end the wa veform as a wave into the pulses shown in Fig. 4C. These
result of point.by.point addition ( 8 J. pulses, applied to the second NAND gate,
As the number of frequencies is increased, reappear as the thin line pulses shown in
their amplitudes tend to average to zero F ig.4D.
everywhere except at the beginning of the Since one CW signal sounds just like an-
cycle. Herc, they a ll add up to a short, sharp other and there may be several in the vicinity
pulse. It follows that a short, repetitive, one- of a check point, a marker feature is re -
sidcd pulse should contain odd and even quired. This is provided b y the 1-Hz astable,
multiples of the fundamental frequency. which paralyzes the second NAND gate part
An ideal thin line pulse has infinite fre- of the time. Its base bias resistors are un-
quency conten t. 0 No real signal could meet equal, giving a distinctive duty cycle to the
this spec, but a fast digital IC can produce output signal. A switch disables the astable
a very workable approximation. Fig. 2 if a continuous signal is required. Fig. 5
shows n Tektronix 545A view of the gen- shows the output when the second astable
erator output and tests with other scopes is operating: the output is locked in the up
indicate the real pulse has better rise time condition during half of each 1-Hz astable
and sharper corners than shown here. cycle.
If this pulse is viewed on a low-perform- Sometimes an astable oscillator will refuse
ance service variety scope, its appearance to start oscillating when it is turned on. I t
will be greatly changed. There will be an does not start because both transistors a re
apparent loss in amplitude, since the pulse in saturation. This reduces loop gain so
occurs and terminates before the slow cir- that available noise cannot be amplified
cuitry can properly respond. The apparent around the loop. It would ne ver start with-
duration is increased, also because of the out some strong, outside inte rference.
slower viewing circuitry. And the fast pulse ~smith, 4flJJliul J"1th ,mat k 1 /or R ndio and Commu ni
may excite circuit resonances, so that the calion.i Envintu, Oo,cr Publica tions, 1961.
111
through the large coupling capacitors C5
and C6. The reverse B-E breakdown voltage
of these transistors is not known, so the
diodes are provided to prevent the turnoff
voltage exceeding 2 volts or so.
IOO 200 DC power for the Generator circuitry
NAN()!(~
comes from a voltage doubler supply based
on a low-current filament transformer. I ts
Fi g. 2. Re al ci rcuit output as seen by a Tektro nix
545A oscilloscope . A fastor scope shows shortor riso-
design is conventional, but a large capacitor,
timo e nd sharper corners. Cl2, is provided across its output to mini-
mize noise on the supply line. The supply
A pair of diodes, Dl and D2, provide a could be replaced with some batteries,
reliable remedy. The diodes are a rranged shunted by a 50 F or larger capacitor to
so that base bias must come from whichever absorb transients. The original breadboard
collector is at the higher voltage. If both ran very well, powered by four flashlight
transistors arc in saturation, their collectors batteries.
are nt perhaps l volt, which cannot provide
enough base current to keep the transistors
in saturation. This contradictory situation Construction
does not arise in the real circuit, which The generator is built in a Premier #PMC
starts reliably. 1008 3x5x7 inch heavy a luminum box. Its
Additional diodes, DS through DB, appear top cover was refinished in light green enam-
in the base circuit of the 1-Hz astnble. el, and fo ur ~ inch grommets in the bottom
These are protective diodes. T he collector piece serve as protective fe et.
swing at turnoff of about 5 volts is con- Inside the box, the 6.3-volt transformer
veyed powerfully to the opposite base
+I +I ..and cheater cord connector are mounted on
+I
Cl
1()1()
OUTPUT
.
CT
.,,
""
NOTll
LID,~ COUUI: TUNNG ltSISTOltS
no 9( SCLCCT0 Ol.#nNO TIM[-Ul'I
Ol,DtO Sll...ICOH, lS PN MJN,
Cl,C2 6 "4,M MATCHIHG COMflONfHTS
'"'
ASTAI\.(
+I
rf
111 \IAJ:.
I
I
0
uo--~--,__~
Fig. 3. Schemat ic of tho 100 Hz thin-lino g onorator. IC ACUO<
112
6-32 threaded shaft to one end of each sim-
(l
IO 20
plifying assembly.
Component assembly on the boards is
largely a matter of plugging in Vector T9.4
'
~ .. 1 D I
on one side of t11e board nnd practically
all components on the other side. This ap-
proach seems a little inflexible but is slrnight-
forward and looks good.
IO
"' Possible board orien tation problems may
~
be overcome by working out n handling and
.. (8)
wiring procedure that doesn't require con-
stant reference to actual components. A good
~I .. ~ L_ approach assumes that the board is only
turned over an imaginary hinge at its
bottom edge, so that top down when one
10 20
side is up becomes bottom up when the
MICLllOSCCONOS
other side is down. This preserves left-right
(C)
rel:itionships. Another useful convention is
that all supply wiring goes to left-hand end
.. of components.
Wiring is carried out one network (plus
~1 :
which occupy most of the box. Tw~ banana
jack output connectors are placed on the
right-hand side, just below the panel.
The circuit boards are cut to 4~ x 5
.___I__.__---L-[
cu 1.0
inches, from Vector %2 inch pattern A stock Bli2!l?
and mounted paraJlel to the panel. The up- Fig. 5. Tho second NANO gato loch in its up posi-
per board is spaced an inch from the panel, tion part of the tima to produce an intermitte nt
and carries both astable oscillator circuits. output.
The other board is mounted one half inch
below, and carries the digital IC and the mounting holes in the board from the com-
power supply circuitry. Assembled, the two ponent side, and their lends brought to the
boards make a sandwich with wiring sides T9.4 lugs.
together. Then the other components arc mow1ted
Both boards are mounted on the same on the board s.
four centers. These are through the second product. Diode and electrolytic capacitor
hole diagonaJly inward from each corner. mounting polarity shou ld be double checked.
The l inch 6-32 internaJly threaded spacers The T9.4 lugs mar need a little bending
are modified by adding a short length of before they will take a good grip on the
113
cies not available from the supply lines, very
careful bypassing is indicated.
High-frequency bypassing consists of C9,
a .O l 11F disc ceramic capacitor at'ross the
IC supply terminals on the wiring side of
the b oard, and ClO, a 100 picofarad capaci-
tor soldered directly between supply termin-
als on the IC. The capacitor leads are pro-
vided with spaghetti insulation and placed for
minimum open space betw.een the capacitor
leads and the !C's supply leads.
Testing before fina l assembly is very easy,
because the odd appearing board layouts
go together giving a structure that opens out
like a book. The hinge is the four leads
between boards. Leave transfonner leads long,
so thnt the circuit may be tested well free
Inside iha assombl od ih in-lin o 9 onorator show ing the
of its cabinet.
component side of the power supply a nd IC board.
The upper half of the Premier box is pre-
pared by a powerful cleaner which removes
its origina l paint. After thorough removal
componi:nts, but no component soldering is of the cleaner, the metal is roughened with
done until everything is insta lled. wet sandpaper, rinsed in vinegar solution
Trimmer capacitor C3 is mounted on its and then clear water, leaving a very good
tabs just under the top panel. Then a small surface that does not require priming for
screwdriver access hole is drilled over it in excellent paint adhesion. Watch out for
the panel, before painting, for vernier fre- greasy fi ngerprints.
quency adjustment after fina l assembly. Rustoleum #868 Green applied from a
Certain components are matched before convenient spray can gives a fine fini sh.
.....
-.........
installation. An ohmmeter and a capacitor
checker will do a satisfactory job of select-
. . .. .... .
.., ~
ing Cl and C2, and R4 and R6, for equal
values. These components arc chosen alike
for best symmetry of the I 00-kHz oscillator ' .....
operation. It might be good planning to
leave these components unsoldered until tun-
ing is completed, but everything else can be
soldered to the board at this point. Note
. 'lllllU ~.
~~,;\) .
;IWll) 4l.IDIJ.
thnt the R3 and RS sites do not get resistors ;i911t~.
until later. ~~~
T wo optional capacitor sites are included.
These are for C4, an additional and prob-
ably unnecessary padder across the crystal;
and C7 A, which can be added to increase
:. ...
: ~-
the width of the thin line pulse.
Apparently, the digital IC comes in a
specially designed package for testing before
use. To mount the IC, solder a % inch
piece of #22 wire in each of the T9.4 View of th o c o mp o nent side of th o asta ble oseilla
lugs carrying supply and signal voltages to tors board.
the IC. Place the IC be tween the two rows
of lugs, bend the wires against the proper Follow instructions on the can. After drying,
terminals, and solder. Ko other mounting is the fresh, clean enamel will take water-
required. proof India ink, applied with a Leroy draft-
The originnl breadboard showed a lot of ing pen. When the ink is thoroughly dry, a
transient noise in its supply circuit. This final coat of Rustoleum #717 Clear fini shes
originated from the IC, which was trying the job. The enamel is soft a t first, but
to get big chunks of current to manufacture hardens into a coat durable in normal lnb
pulses. Since the IC cannot deliver frequen- use.
114
quency near enough to WWV for final zero-
ing with the capacitor. Verify tuning rnnge
on both sides of WWV.
Correct values for R3 and RS may be
approximated very quickly if a good trig-
gered scope is available. Try selecting re-
sistors for a period of 11.4 microseconds with
......
THRU.OCD STIA>
the crystal removed.
~HT SI)[ OF SUPfl\.'T
I IC 90UllO
115
Using the First Ham Integrated Circuit
For several years, 73 has been a leader te rmines an ! C's cost, so that the LM270,
in publishing articles on ham applications which is about the size of a single medium
for cmrently available digital and analog power transistor, is already cost competitive
IC's. Until now, however, no !C's had been with the less complex discrete-component
produced specifically for two-way-radio use. circu its it replaces. As volume commercial
At least one IC manufacturer, recognizing use of the circuit increases, the circuit is
the need for such circuits, in the potentially likely lo be available at even more attrac-
large commercial, military, and amateur tive prices.
market, is now aiming a major development
effort toward communication "subsystems on Inside the can
a chip", and it is expected that other man-
ufacturers will follow. Besides the direct The LM270 consists of several separate
benefits of improved performance, decreased functions, designed to work together in a
size, and lower component costs, which will self-contained system, lo produce control
reach amateur radio through commercially- voltages for external use, or to respond to
built rigs, the new specialized chips will applied control signals. H eart of the cir-
enable even the casua l homcbrcw artist cuit is a balanced series-shunt variable at-
lo construct sophisticated, complex equip- tenuator, formed by the four transistors in
ment he might previously have considered Fig. 1, which allows a large gain control
beyond his reach. range, with low distortion (for inputs less
The first Communication IC now available than 100 mV p-p), and which can be di-
is the nlional Semiconductor Uvl270 Audio rectly coupled to other parts of the system,
ACC/Squelch Amplifier. It is basically an e liminating the transformer or capacitor
operational amplifier, whose gain is con- coupling necessary with all other variable
trolled by a de voltage, plus a built-in sen- arrangements. From a twelve volt supply,
sitive squelch threshold detector. The ten the gain vs. control voltage relationship is a
pin circuit replaces entire sections of to- smooth curve, as in Fig. 2, which gives a
day's transmitters, receive rs, or transceiv- constan t gain of +40 db for control voltages
ers, and makes speech compression, VOX, be tween zero and +2 volts, and is effectively
receiver squelch, and other functions prac- "shut olT" a bove + 2.6 volts.
tical in even the simplest homebrew rigs.
While the chip contains 36 junction devices
(tra nsistors and diodes), and 20 resistors,
it is size, rather than complexity which de- 40
-,.....lo...
JD
2D
r\
m
+1D ' \
~ D
"" - 1D
\
1
- 20
\
- JD
\.. ......
- 40
1.9 2.0 2.1 2.2 2.J 2.4 u 2.6
Vac CONTRO L. VOLTS
116
... volts, so that symmetrical output clipping
occurs.
Two identical gain control inputs, pins
3 and 4, are provided, which allows con-
trol by two independent sources at the same
time, such us simultaneous AGC and squelch.
By bypassing pin 2, the gain control inputs
become emitte r-follower positive peak de-
tectors. The control inputs are protected by
6.5 volt zeners (Q33 and Q34). U the con-
trol input is expected to rise above +6.5
volts, a lOK series resistor at that input
should be used to prevent excessive dissipa-
tion in the zeners.
Fig. 3. Differential input circuit a/Ill squelch de
teclor.
Remote gain-controlled audio amplifier
A simple application is a preamplifier,
A separate subsystem within the LM270 Fig. 5, whose gain is manually controlled,
is the squelch detector, F ig. 3. Using the noiselessly, by a de voltage from a remote
same input differential amplifier as the var- location, rather than running long, capaci-
iable gain circuit, the high gain peak de- tive coax signal lines to and from that loca-
tector formed by Q20, Q36 and Q21 re- tion. Pin 4 is bypassed b y an external capaci-
sponds to very small inputs (as little as tor, to eliminate noise pickup. Since the
a millivolt, depending on setting or the ex- gain-control curve, Fig. 2, is approximately
ternal threshold pot), by rapidly discharg- logarithmic, n linear pot w ill give a desirable
ing an external capacitor. In the absence of logarithmic audio attenuation characteristic.
input signal, C(ext) ch arges above +2.6 V, For illustration, the second control input
which, when tied to the gain control input, is shown connected to an IC logic gate, of
keeps the output amplifier "off". A mom- the DTL RTL or TTL varieties now avail-
entary input peak above the threshold causes able at l~w cost. This gate, operating from
Q21 to rapidly discharge C(ext) below a five volt supply, can be p art of a logic
+2 volts, turning the amplifier full y "on". arrangement to overrfde the remote con-
This arrangement gives a fast attnck, slow trol, and shut off the amplifier under pres-
release squelch, which catches first speech ent conditions. The resistors and capacitors
syllabics, and waits long enough to avoid shown biasing the single-ended input are
"culling out" between words. used to illustrate one way of operating the
The complete circuit appears in Fig. 4. inp uts at a fi xed de voltage; subsequent
A de tailed explanation of each part, too examples will show simpler schemes.
lengthy for inclusion here, may be found
in the references.
JY~----J
required is tJmt both inputs be at equal de
potential, somewhere between + 4.5 volts
and the positive supply. Like an "op amp", ~lOMCLIHE
the LM270's de output voltage stays at ap- :-;,---i
proximately half of the positive supply volt-
age, for all supplies between +4.5 and +24 Fi11.; 5. R emote or digital control amplifier.
117
CUI.CUI w1t1u AIU,.. COllltlCTIO'll
118
Yee
,..
r---- ,..
.........vv...........-1 ~
OUT
....
CHAflCilNG
RESISTOR
SOUUCH llK
THRUHOLO
119
Twin-tee constant amplitud e audio
oscillator with remote level control
Oscillation occurs in a twin-tee, op-amp
type circuit, when total feedback gain equals
unity (including filter losses). Conventional
methods of regulating oscillator amplitude
usually rely on nonlinear loading of the
gain stage. With the LM270, however, gain
may be set by detecting the output, and
using this to force the gain to exactly the
minimum value required to sustain low dis-
lorl ion oscillation. The "ACC Oscillator"
circuit, Fig. 13, automatically compensates
for changes in oscillator load impedance.
The exact amplitude at which this action
Fig. JO. Fast a/lack, slow release squelch action.
occurs is set by an external pot, and may
"
I .ll"''
-= 11V
~~
-s
... ... . . ..
u1
"'"''""
HlH
uoe
120
feedback network in the preceding exam-
ple, to make a regulated-output AM if align-
ment generator, F ig. 14. If the AGC thresh-
old voltage, which determines the ampli-
tude or stabilized output, is varied at a slow
(audio) rate, the output rf amplitude will
be forced, by the AGC feedback, to track
the audio modulation. ... ......
WIUTA
Conclusion '"""'
11.flil
121
Integrated Circuit TV Sync Generator
122
Fig. J. Scope trace photo of sync and blanking signals and horizontal and vertical rlrive.
in posit ion relative lo the vertical pulse by to set on a negat ive-going trigger edge. (Set is
3 1.75 s. This will cause the ou t put of the interpreted as meaning termi nal Q in the
integrator to be shifted by a small amount drawings will go to +3V a nd terminal Q will
on alternate fields. This shift is enough to go the OV.) An analogous situatio n exists for
cause the lines on alternate fields lo pair the clear input and t he reset condition of the
rather than be spaced equally. The group of flip-fl op. These flip-flops also have a de clear
pulses that remed y this situatio n arc called which resets the flip-flop on a JV pu lse o r
equalizing pulses. level which overrides the inputs on all o t her
Six pulses. each of half the area and twice terminals.
the repetition rate of the ho rizontal pulses, The Clock and Frequency Dividers
arc placed before and after the vertical sync Basic components of any accura tely inter-
pulses. Being the same o n every fie ld , they l a ce d scanni ng system are frequency
serve to buffer the integrator o utput against dividers. One frequency divider divides the
shifts of the horizontal sync pulses which input freq uency of 3 1.5 k Hz by 525 to
now occur farther away in time. obtain a signal at the field rate of 60 Hz.
The complete sync, blanking, and drive Another unit divides by two to obtain 15.75
signals are shown in Fig. 1. This has not been k Hz, t he line rate. In the sync generator
intended as a rigorous treatme nt of the described in this article , the dividers are
television synchron izing system. Fo r the actually electronic binary coun ters requiring
interested reader, the subject is covered in no adjustments. The simples t of all these
many texts. circui ts will d ivide o nly by powers of two.
All gates are negative logic 11a11d or An example is the 4: I counter of Fig. 2 used
positive logic nor. As this unit was designed lo obtain a 3 1.5 k Hz signal from a 126 k Hz
with negative logic in mind , the analysis oscilla tor. II is necessary to apply feedback
proceeds most easily from this viewpoint. lo count by numbers other than a power of
The flip-flops toggle on the negative-going two. The 15: I , 7: I, and 5: I counters that
edge of a trigger (T) pulse provided the set make up the 525: I divider i~t Fig. 2 are
(S) and clear (C) terminals are left floating examples of this type of coun ter. Briefly
or arc grounded. A signal of +3V o n the set slated, the output of the nand gate associ-
and OV on the clear will cause t he fli p-flop a ted with each cou nter resets the en tire
123
Fig. 2. Master osci/Jator frequ~ncy dividers.
counter to zero when the required coun t is change to 3 V. The rate o f change is adjusted
reached . by the potentiometer so the input to the
The 7: I counter has so me nu nor varia- inverter rises to the point necessary to
tions. The last stage is reset by the action of sa turate the transistor in the inverter after
the previous stage resetting. A capacitor is 1.27 s. Thus, 1.27 s after the input transi-
placed across the input of this counter. tion takes place, the output transition will
These two modifications are necessary to take p lace from 3V to OV.
prevent extraneous pulses from resetting or
triggering the counter stages. This is fast
logic and the offending pulses are less than 4 70111"
124
HOfUZONl lAL 12
The master oscillator operates at 126 kHz. SYNC
lilATC I
This frequency was chosen because crystals IN't'CltTEO
for this frequency are less expensive and
more easily o btained t han lower frequency
crystals. The circuit is a Pierce oscillator
constructed from a digital inverter. A second SYNC OUTPU1'
Pulse Generation
Four monostable multivibrators are used
to generate the following pulses: equalizing
INVCRTCO
pulses, 2.54 s; horizontal sync pulses, 5.CB
s; horizontal blanking pulses, I 0.16 s; and Fig. S. Output gating.
vertical sync pulses, 27.3 1 s.
The 27 .3 1 s monostable is a standard
design contructed from two gates. The other
monos tables are constructed from one Generating the Gating Signals
inverter each. They are not true monostables The main counter has 525 unique s tates
in the sense that they d o no t have feedback. of each of its flip-flops. These states can be
d ecoded to provide gating signals.
Gate 3 is the vertical blanking signal. It is
PINl,C1~8
C4
14
04
I
GATE J
generated directly as a combination of the
P1N 14, C2 IO
GATE l
s tates of two fli p-flops o f the counter. Thjs
- - -- IH V111TEO
gate is about 3 horizontal lines longer than
the recommended width. This is virtually
unnoticeable and the add itional logic neces-
sary to correct it hardly seems worth the
additional complication and expense.
Gate I gates off the horizo ntal sync
pulses and gates on the equa lizing pulses.
Gate 2 ga tes off the equalizing pulses and
ga tes on the vertical syn c pulses. Gates I and
2 arc both produced by gating pulses from
the main counter at the begin;1ing and end
of the gate times and using these pulses to
set and clear flip-flops made of t wo cross
connected nand gates. The circuitry is shown
in Fig. 4.
Fig. 4. Gate generacion.
Output Gating
Gates I, 2, and 3 arc used to turn off and
They will wor k in this ap plicat ion because on the basic pulses at the required times.
the half period of the driving waveform is Gate 3 (Fig. 5) gates off the horizontal
longer than the resquired pulse width. The blanking pulses during the vertical blanking
circuitry is shown in Fig. 3. All ou tputs are interval and it itself is inserted as the vertical
invert ed. blanking signal.
125
Possible Changes wiring can be rapidly depicted. This unit
IL is possible to generate "on" and "off" cou ld idea lly be constructed on a single
pulses and use these to set and clear a printed circuit card , although the layout of
flip -flop to generate gate 3 analogous to the such a large printed circuit is not a simple
circuit s for genera ting gates I and 2. In task.
amateur service this additiona l circuitry On all integrated circuit sockets, pin 4 is
hardly seems worth the additional cost and ground and pin 11 is connected to +3.6V .
complication. If the unit is intended for The crystal socket was mounted on a
commercial service, th.is modification will be bracket on the underside of the chassis. All
necessary. Other modifications that would discrete components were mounted on a
normally be required for commercial service directly on the IC sockets. I do not recom-
would be sepa rate width and position flip- mend this procedure as it makes an already
flops and associated gati ng for the vertical crowd c d situation worse. A possible
and horizontal drive signals. Another "must" improveme nt would be to mount all discrete
components on a piece of Vector board
. .,
moun ted on the underside of lhe chassis.
The nrnlliturn potentiometers are a con-
.. ..
lucnSPJ I lwc1n,.I ... ll
j venience but satisfactory operation should
..
LANK
IMC710f'I
C<
J.. cnoPf lwc790f' j
C> C2
lwcnoPI
Cl
D
12
TOP OF
.1
be possible with ordinary carbon potenti-
ometers. Four phone sockets are used for
sync, blanking, and vertical and horizontal
lwcno j lwc790PI jwcnoP I lwc19oP I I. C drive connect io ns.
PINS
DOWN The large empty space on the chassis, the
D D3 Dt DI
extra output connectors, and the two spare
fwCJu' I fwc1etPI lwcn . ,,I lwc125P I terminals on the power block are for pos-
00000
t E3
sible expansion to color. The IC socket is a
E cntf' J lwcnPI 16-pin type that works as well for these
1.27 Z...$4 .$0810112131
JU I'' )H 14-pin packages.
~
}It
~~1
the generator in sections, the first section to
be wired should be the clock and frequency
dividers. Observe the input and output of
[ ____.=_-_ the 15 : I divider on a dual-trace oscilloscope.
There should be 15 input pulses for every
~ outpu t pulse. I f you do not !ave dual-trace
Fig. 6. The complered sync generator. facilities, you can mix the tw.J signals in a
spare gate. Although the output does not
look the same as with a dual-trace scope, the
is the inclusion of alternative methods of number of pulses that should occur dttring
master frequency control, such as a method tile o tttpttt pulse for each cou nter can be
to lock the generator to the power line or determined fro m Fig. 7, which shows scope
some externa l source (such as netwo rk or trace photos of the gating and coun ting
remote site sync generator). circuits. The other dividers may be checked
Cons truction in the same way.
The unit is constructed on a 6 x 12 in. See that the 15.75 kHz and delayed
chassis o f which o nly 6 x 7 in. is used for the 15. 75 kllz signals are in phase. They will be
sync generator (Fig. 6). The primary con- if you have not made a wiring error. Adjusr
sideration to bear in mind is case of wiring. the delay so that there is a 1.27 s between
Each integrated circuit has a possible four- the leading edges of the two signals. Again,
teen or more connections to it and space for without a dual-trace oscilloscope, mix the
126
Equalizing pulsos
2.54 us
Horizontal sync
5.08 us
Sync
Gate 1
Gato 2
Pin 1J, BJ
Pin 9, BJ
Pin 1J, B2
Pin 9, B2 (output)
7: 1 Counter
P in 2, 8 1
Pin 13. 8 1
Pin 9, B l
Pin 9, C l
P in 2, C l
Pin 1J, C)
Pin 9, C2
Pi n l J, C2
127
inverted nondelayed output with the normal accurate indication of the output waveform.
delayed out put. The pulse on the oscillo- Note, however, that 110 adjustments have to
scope from the o ut put o f your mixing gate be made while looking at this waveform so if
has a width equal to the delay time. you trust your wiring and previous adjust-
Wire the monostables and their inverters ment procedure, you can be confident of
next. Look at the output of the inverters having a correct sync signal. If you still insist
and adjust the pulse widths to the correct on seeing the sync waveform, it is possible to
values. Remember the pulses are negative add a single flip-flop as a 2: I counter
going at the o utputs o f t he inverters. Wire connected to the output of t his 2: I divider
gate 3 first , then gate 1, and finally gate 2. to the external sync on your scope and you
The out puts of t he three gates may be should be able , after some adjustment, to see
compared with the scope pictures in Fig. 7 . the sy nc signa l.
Their widths should be: gate I , 57 1.5 s; gate Since the outputs of t his unit are the
2, 19 0.5 s; gate 3, 2857 .5 s. Lastly , wire t he outputs of gates, they will 110 1 drive a low
output gating and con nect il all together. impedance load. All input s to o ther equip-
Sync, blanking, and drive should no w appear ment should be high impedance and the line
al their respective outp ut terminals. Chee)\ should 1101 be terminated with a resistor.
the appearance of the blanking and drive T his situation is ideal in my sta tion as most
with an oscilloscope. The sync waveform equipment is of IC contruction and the
may only be viewed reliably using an oscillo - output of the sync generator simply drives
scope with delaying sweep. The delay has to other gates. If you absolutely require low
be longer than two fields, as the horizontal impedance signals, the outputs of the gates
sy nc pulses alternate their position every may be buffered with simple em itter
ot her field. Without these precautions you followers.
can not be certain that your scope trace is an
128
A Digital Readout for your VFO
R. Factor
129
Type Quan To tal
SN7490 IO $20
SN7475
SN7441 AN 4
5 $ 10
$ 12
...
~'
B57 50
(Nixie) 5 $25
I Mii l Xtal _LL
sn
T his is effective ly for the "worst case ."
i.e .. no ex ternal oscillator, using I MHz
crystal , 5-digil readoul with sior:it!e. lJ y
eliminating the I 0 11 rcsolut 1011. using an
existing 100 KH z calibrator. and tkkring
the most sign ifican t digit. the prkt ht'
co mes.drasticall y lower. TO
COUPHUI
"'""'
GAT [
SN7441AN 3 $ 9
B5750 3 fil_
$38
".
VA(
130
de at about 500 mA, B+ at about 15 mA, T he purpose of the vfo buffer is to
and IV of signal at the vfo frequency. To make sure that the vfo in the rig is not
get the 5V, lap the filament supply (6.3 or, disturbed. An emitter follower is ideal for
preferably, I 2.6V and connect it to the th is application. The component values
bridge rectifier shown in Fig. I . If your shown arc for a supply of 150V, as this is
fil ament supply is 6.3V, .it may be neces- the most common ly used. No trouble
sary to connect another 6.3V, 500 mA should be encountered if the pickoff capa-
transfor mer in series wit h the filament citor ( 12 pF in the schemati c of Fig. 2) is
supply. Nixie tubes like l 70V at about 3 at least an order of magn itude sma ller than
mA across them, bu t since they tend to act the capacitor from the cathode to ground.
as voltage regulators, current limiting is The ac tual capacity will be much sma ller,
necessary. Measure your B+ supply , sub- since the pickoff capacitor is in series with
tract I 90V, and calcula te a resistor that the transistor base, but it doesn't hurt to
wil l give 3 mA per Nixie with this voltage be safe. This completes the necessary mod-
across it. Be su re to make a power calcula- ification o f the rig. No damage to the fron t
tion also, as a 2\V resistor may be required. panel - see?
Small series resistors are con nected to each
Nixie to equalize current; the resistor just Circuit Descrip tion
calculated goes to the junction of the series The theory behind the electronic count-
resistors. er i~ that the nu mber of cycles of the signal
+!!V +!!V
47
47K
220K~
l-3!!ll8
1.0 MHz
~ I
I \
\
I \
'9. 1 1ns~
I I
!
j-J u
nn'--!!
I I
I I
' !
,
' ' ''
'
131
4 IDENTICAL M ODULES
""
OUT OF I.A.NO
INDICATOR
~+12
2'~
in a given period can be counted. If the chain, the first stage of which is the least
period, for example, is I secon d, the significant digit. After I 0 pulses, the first
number cou nted comes out to be cycles stage is reset and a carry pulse genera ted.
per second. Deriving the period precisely is Since the cou nter has five stages and the
done by counting down an accurate refer- counted frequency is in the MHz range,
ence oscillator and then using the counted you can see that the counter will overflow
down signal to gate the signal to be several times during each count interval.
measured. However, we know what the first digit will
The circuit works as fo llows (see Figs. 3 be, so there is no point wasting a counting
and 4). The input signal goes into a divider stage o n it. All the stages arc identical
132
except fo r the most significant digit, which the I MHz output of another oscillator
uses discrete components since special de- can 't be used. Approximately JV (peak-to-
cod ing is required. peak) is required. You can also use a I 00
After the counters have counted, the kHz oscillator and delete one divider stage.
SN744 I AN decoder/ drivers ground the ap- T he MC724P quad 2-input gate shapes the
propriate Nixie cathodes and the corres- signal into a sq uare wave suitable for
ponding numbers light up. Between the triggering the dividers.
counters a nd the decoder/d rivers, one addi- The first two dividers I used in my unit
tional stage is necessary - a buffer storage are Fairchild L958s. I used them because
register. The reason for this is that the gate I had them lying around, and it is okay to
time is 0.1 second , which will cause blur- use SN7490s here. A section of IC8 is used
ring during counting. If you wish to read as a buffer to make the two IC families
out once per second , this isn't too bad. compatible. If you use 7490s, dele te this
However, the dial is more useful when it's gate.
responsive, and you can zero in a frequen- ICs 3, 4, and 5 are the remaining
cy much more quick ly when you don 't dividers. At time zero, le t us suppose that
havo.: to wait a full second to sec where the IC6, the gate nip-flop, is reset (gate at logic
dial turn took you. By using storage 0). The carry pulse from the last divider
registers and a few logic gates, il is possible brings the gate high , starting the count.
to make ten 100 ms measurements in JO.I One carry pulse later (JOO ms), IC6 is again
seconds with no digit blurring. How this is reset, stopping the count. (The abbrevia-
done brings us to the timing circ uitry. tion "ms" means milliseconds; s" means
First , an accurate frequency must be microseconds.
generated. The circuit shown in Fig. 2 is a Here the fun begins. IC6 Q, being high,
stable oscillator designed for 32 pF crys- sets half of the and condition o n the first
tals. A trimmer is included for fine fre- two sections of IC7. IC3 is sti ll counting
quency adjustment. There is no reason why from binary 0 to binary 9 at a rate of I 000
NUU( NIXIE
READOUTS ..fAOOUTS l
~
OR1Vt:RS ORIV(RS
QUAD DlCAOt
LATCHES COUNTERS
,.f.S !T
l lNl:
SlR08E
LINE
Fig. 4. Logic block di,1gram. The lefr block gives rea!1our in l 0 Hz incremcnrs ar l 01 ms interva ls wilh-
our blurring. The righr block reads in I 00 Hz i ncrements; rhe update i s adjusrable and blurring occurs
during l 00 ms coun t.
133
times per second. Its output states are bases in parallel and grounding the emitters
decoded by the remaining inputs to IC7 in of the appropriate group, a cheap and dirty
the following seq uence : At binary 4 and 6, "adder" is constru cted.
ro
pins I 2 and 13 are both high, causing pin Ot~ T
OCtOOCltS
11 to go low, pin 8 lo go high, and thus
strobing data into the buffer register. The
strobe is performed twice because doing so
eliminates the need for an additional de-
code - the same data is strobed each time.
At binary 8, pins 2 and I of IC7 are
both high, causing pin 3 to go low and pin
6 to go high, thus resetting the counter and [11.TlltA SWITCH
WAfCN ON
TlltANSiSTOlllS
lHCIO Olt [QUIV
OANOSWITCH ro NIXI[ c.:.tMOOU
resetting the last two 7490s in the dividing
chain. However, these two dividers are Fig. 5. Using a wafer switch to ground the
appropriate Nixie drive for "adder" function.
reset to 9 instead of to 0. This means that
just 0.1 millisecond later, when IC3 l'arries,
IC4 and IC5 will also, starting the count
again. Note that the total time between the
end of the first coun t and the beginning of
the next is just I ms, allowing almost I 0
measurements per second.
...
,.., .,.,
c111tr )
Backwards or Forwards?
If your rig is of the subtractive mixing
type, you are probably wondering how the
readout frequency corresponds to the out-
put frequency of the rig. It doesn't -
everything is upside down. Of course, it Fig. 6. General decoding scheme for all BCD
will be accurate on one frequency in the states.
middle of the band, but that's little com-
fort. Despair not. Instead, connect the If your vfo starts from some frequency
Nixies to the counter backwards. (Sw itch other than an even megahertz, the decoder
0 for 9, I for 8, 2 for 7, 3 for 6, and 4 for must be designed to ma tch. If the vfo is,
5.) Through the magic of mathematics , for say 3.9 to 4.4 MHz (as is the HX500),
ever; I 0 Hz increment of the vfo frequen- where 3.9 corresponds Lo the top of the
cy, there is a I 0 Hz decrement of the band and 4.4 to the bottom, then the
readout frequency. You ca n now see why I decoding is as fo llows: 4.399.99 to
said that the rig can be additive or subtrac- 4.300.00 is 000.00 to 099 .99 kHz o n the
tive, but not both ... unless you happen to
have a two-position, 50-pole switch a-
round .
134
dial. Thus, 3 must be decoded as O; 2 as 1 ; time base. Since the count interval is so
1 as 2; 0 as 3; and 9 as 4. To do this the short , you can sample once every tenth o f
gale that decodes a 3 binary state should a second, count for a hundredth of a
drive the 0 (and 5) digit line, etc. The second, perform housekeeping functions in
decoding section is fairly simple since only 2.0 ms, and have a display visible for
5 states must be decoded, compared to the almost 90% of the time without using
10 states decoded by the SN7 441 AN. stor age registers. One Nixie stage is elimin-
Figure 6 and Table 1 give the general ated, and JC6 is unnecessary.
decoding logic for all possible binary- To make the stripped-down version ,
coded-decimal states. Considerable simplifi- substitute the circuitry of Fig. 8 for the
bottom half of Fig. 2. It works as follows:
As the dividers count up fro m 0 , the diodes
decode a count of 88 and generate a 1 ms
long rese t pulse. Al tbe transition to 90,
the count gate decodes a binary 9 on the
most significant digit and goes high for I 0
ms, form ing the gate interva l. Thus, the
SN7490s esse ntially perform the timing
fun ctions.
An additional Nixie stage can be elimin-
ated by deleting the most sign ificant digit
fro m the count chain. This also saves the
trouble of figuring out the decoding cir-
cuitry. Since any vfo will be accurate to
the nearest I 00 k Hz, a second of mental
work will co mpute the correct number.
135
istor - there should be a large ac compo-
nent. Pick a value fo r the resistor so that
when the voltage is at an instantaneous
minimum, the current through the resistor
is about 100 mA. If all this is too much
trouble, don't bother. Almost any transis-
tor you choose can handle the power even
without the resistor.
-. Final Thoughts
~
."f""'lt<
I have described a fairly simple and
relatively c.heap way of obtaining accurate
~~~lllJilllliiillliii frequency calibration. There arc some in-
~ 2 herent limitations. The main limitation is
Rear view of counting circuit. that the counter doesn't actually measure
the transmitter out put frequency. Since
the assorted conversion oscillators are in-
around before erratic operation occurs. evitably crystal-control led, there is little
However, it is recommended that the unit cause to worry about significant errors.
be carefully bypassed and shielded; first , When you install the unit, it might be a
because often there is a considerable a- good idea to set the vfo at 000.00 and
mount of rf noating around, and, secon dly, zero-beat the transmitter output with your
because the pulses in the digital circuitry calibrator by adjusting the trimmer capaci-
are very fast and can radia te noise to the tors on the conversion crystals.
receiver. If you can't build the counter described
Mechanically , my unit was constructed because your rig doesn't fit the criteria,
on two boards which were then sand- there is still hope. For instance, if it is
wiched together. This saved about I 0 sq in. additive and subtractive, you might con-
There is no reason why the unit can't be sider using the bidirectional counters that
built on one board. Final dimensions exclu- are now available. Direction of count can
sive of case were 5 by 4- 1/2 by 2-1/4 in.
Something this size could probably be
installed inside the cab inet of the rig. If TABLE I.
you wish, the Nixies and their drivers ca n
Connection Chart for all Binary Combinations
be placed on a separate board whi<!h should
easily fit in the space taken by the typical CONNECT
dial drum or scale assembly.
T he unft requires SV and so me conven- N= B INARY A B c D
ient high voltage. I described the B+ 20 21 22 23
"supply" earlier. The SV supply is wort h a 0 0000
few words. Regulation is not critical, but if 1 1000 20 2i' ~ 23
the supply has lots of hum or noise, filter it
2 0100 2 21 ~ 23
out before it gets to the oscillator. The
oscillator only takes 5 mA or so and an RC 3 1100 20 21 22 23
filler should be sufficient. 4 0010 2 21 22 23
I used some nondescript power transis-
tor for the series regulator. Any thing that 5 1010 20 21 22 23
c;111 handle an amp and SW of dissipation 6 0110 ~ 21 22 23
should be sufficient.
The resistor shown bridging the transis- 7 1110 20 21 22 23
tor bypasses some of the current without, 8 0001 ~ 21 ~ 23
theoretically, greatly affecting the regula-
tion. Measure the voltage across the tra ns- 9 1001 20 2T ~ 23
136
10 KHz
RESET
IC 8 ;J;470
Fig. 8. By substituting this circuit for the botto m Re f. Macleish, Kenneth, "A Frequency Counter
half of Fig. 2, a cheaper but less accurate digital For the Amateur Station," QST, October 1970,
counter results. pp.15.
be controlled by the band switch. These counter that might solve your problem.
units are presently more ex pensive than the If you fee l as I do, that electronic
7490s. equipment should have uses comprehen-
If your rig uses some odd vfo freq uency, sible to the layman, witness an additional
perhaps with a 455 kHz added o r sub- characteristic of the digital dial. Take vfo
tracted, you can try several things. On e is knob firml y in hand and give it a spin. The
to mix the vfo with a signal that will last two digits of th e display will be
subtract the o dd number. Another is to random numbers! Not wishing to become
preset the counter with the reciprocal of an accessory before the fact, 1 won't
the odd number and set to 9 all digits more attempt to suggest what you might use
significant than the o dd number. I don't those random numbers for, except perhaps
know how well these ideas will work, since compiling a random number table. But that
it was not necessary to try them, but they shouldn't stop you from thinking ...
se.:m to be theoretica lly sound. The refer- ... WA21KL
ence gives a design for a more elaborate
137
The Mod 2 Digital ID Unit
138
lines, one at +2V, the other two al OV, tcr to the zero state. All Q' lines become
produces a OV o utput on the o utput high and all Q lines become low. Approxi-
line of the gate. mately 2V and OV are fed into the diode
matrix , which decodes the cou nter number
Symbol:
int o an instruction for the oscillator keying
logic. In the DI U there are four basic
Nam/ gate: l\n inverted and gate; dev ice instructions: (I) send a dit, (2) send a dah,
used to give a low outpu t when all input (3) send neither dit nor dah (blank) , and
lines are high. Example: 3 input lin es. (4) stop.
+2 V on all 3 input lines of ga te,
p roduces a OV output on the ou tput If the diode matrix decodes the first
line of the gate. sequence count ( 0) to be instruction num-
Symbol: D- ber I (send dit), the <lit signal line from the
matrix will be high. This will cause the dit
inverter to have a low output and one-half
For this article, nor gate logic was used of the "d it enable" gate will be enabled.
to implement the 11a11d functions; there- Since the space line is also at " low" level at
fore, the definition for ou r purpose of a this time , a trigger pulse will be sent
nand gate is a device used to give a high through capacitor C7 to the "<lit" one-
output when all of its input lines are shot. (A one-shot is a monostable device
low. Example: 3 input lines, OV on all used to generate a predetermined pu!se-
three input lines of a gate, produces width .) The dit time pulse determined by
+2 V output on the outpu t line of the the one-sho t is sent through the "dit or
gate. dah" gate and the "dit or dah/blank " gate
to e nable the "d it or dah send" gate. The
Symbol:
nand gate keys the oscillator circuit to
produce the dit signal.
Note that the zero placed before or At the same time the dit is being sent by
after the inverter, nor, and nand logic the one-shot to the oscillator , the "space"
gates defines the expec ted state of the one-shot logic is being reset via the "dit,
input or the output for the function to dah, or blank" gate, inverter, and "space
occur. enable" gate.
Flip-flop: A device used to store informa- Upon completion of the dit signal, the
tion a bit at a time . In the DIU "dit, dah, or blank" gate becomes high,
application , a string of flip-flops 1s used making the inverter output low. Since the
as a counter. The purpose of the coun- stop instruct ion has not been called for by
ter being to sequentially address the the matrix, the "space enable" gate pro-
required instructions for the DI U. duces a high output. The high output in
D
turn .sends a pulse through capacitor C3 to
Symbol: trigger the "space" one-shot. (The space
time period is used to separate the units of
a letter. Example: D = dah-space-dit-space-
dit.
Unit : Smallest bit of information sen t by The space period is the same as the
the DIU (dih, dah , or blank) . period for the dit. The space signal, besides
The DI U uses the MC7 00 series o f allowing for the time to distinguish the
Motorola integrated circuits due to their units of a letter, advances the counter
inexpensiveness and availability. through an inverter to the n ext unit and
resets the "dit" and "dah" one-shots by
Syslem Operation discharging capacitors CS and C7.
I\ 0\1 signa l through the start network If the diode matrix decodes the next
(sec Fig. I) from the transmitter keying sequence to be instruction 2 (send dah),
circui t resets all the flip -flops in the cou n- the dah signal line from the matrix will
139
s
T
A
~
N
E
T
w
~
XMIT
KEY CKT
All
...
11
PINS
All
PINS
I l
36V
Fig. I. Digital identification unit, logic and
schematic diagram.
become high and the dit signal li ne will space-dit-blank-dit-blank) Again th e loop
become low. When the space line becomes through the "space" one-shot is triggered
low , the "dah/blank enable" ga le will send and the counter is adva nced to the next
a pulse through capacitor CS, triggering the unit of in for mation.
"dah/blank" one-shot. The dah/blank p ulse The counter is advanced each time a
would then go through the "dit or dah/ unit of information is sent until it is
blank" gate while the dah signal from the advanced to the "stop" instruction. This
matrix would go through the " dit o r d ah" instruction causes a blank to be auto mati-
gate. T hese two gates would then enable cally sen t and stops the "space enable"
the "dit or dah send" gate to trigger t he gate from t riggering the space one-shot.
oscillator for the dash period . T he "space" The DIU remains in the stop state until a
one-shot is again triggered to advance t he reset pulse is sen t to the counter from the
counter to the next unit. transmitter keying circuit and the whole
If the diode matrix decodes the next process starts over again .
sequence to be instruction 3 (send a Of course the DI U works much faste r
blank), neither the dah nor dit line wiU than it can be described. Depending o n the
become high. The same will occur as above co mponent values selected for C2, C6, and
for the dah except that when the signal CB, the DIU can fu nction at any reasonable
reaches the coincidence gates the "d it or speed. The particular values used in the
dah" gate will not be enabled. T hus the prototype and listed for Fig. I (see Table I)
oscillator will not be keyed. This generates causes the unit to iden tify at the rate of 42
the blank period which is put between wpm (2 seconds fo r DE W6FNO). If a
letters. (Example: DE= dah-space-d it- faster or slower rate is desired, capacitor
140
values should be changed accord ingly. It feedb ack and bias networks - make up the
should be noted , however, tha t CB must be oscillator. The oscillator was d esigned to be
three times as large as C2 and C6 to give fed directly to the grid of the modulator in
the proper character formation. This is a the transmitter.
critical relationship and follows from the
fact that a space and dit are identical in The Diode Matrix
time length while a dah or b lank is three Up until now very little has been said
times the length of a dit. about the diod e matrix other than the fa ct
Transistor Q2 is used to lock on the that it determines what instruction to give
transmitter keying circuit while the digital the keying logic. T he actual constructio n
identificatio n unit is send ing its id entifica- of the matrix can be considerably sim pli-
tion code. If a timer is used in conjunction fied and consequently cheaper. Up to 70%
with the identifier, the transmitter will be of the dio des necessary for the diode
matrix can be eliminated by using math-
ematics. A much more sophisticated, eco-
Table I. Par ts list for DI U L ogic Board no mical, and space-saving layout can be
and Diode Matrix
DI U Logic Parts achieved using Boolean algebra. Thanks to
R28 10 K Y.W Mr. Karnaugh , it is not necessary to give a
R9,R10 3.3K Y.W complete d iscussion on Boolean algebra.
R1 , R11r33K Y.W
C1,C3,C5,C7 , C9 .05 td disc 25V
C10,C12 .008 Id disc 25V (1000 Hz)
C2*, C1, ca 10 fd/15V
C4, CG* 30 fd/15 V
*Most change in direct ratio
IC1 -3 Motorola 791P
IC4,6,7 Motorola 724P
ICS Motorola 789P
01-4 2M3415 or eq uivalent
01-3 1 N34 or equivalent
IC sockets wire-wrap type Vector R-714
22 pin PC socket Vector R-644
Matrix Parts
10-20 resistors 3.2K Y.W The IC logic board is shown here from the
60-100 diodes ge or si (all same type) compo nen t side. Note the use of IC receptacles,
22-pin PC socket Vector R6' which simplifies test, checkout, a nd replacemenr.
141
the top and along the side o f the chart It>".., 110 Il l 101 100
<"o 000 00> 011 010
correspond to the binary output of the
flip-flo ps - I for true or 0 for fal se. The 0 00
0 .19, l9, 7 tfJ. .
letters written diago nally in the top left
corner refe r to the six flip -fl ops. Example :
001
/\
td/
. /t ~,'\ r~.
.."
10 , .. I
"
Box 17 has flip-flop A tu.e, B false, C fa lse,
D fa lse, E true, and F false. Wri tte n in
0 11
fi'X. I
,,,--," 0
.. ~ J(
"\ ,.
Boolea n for m, 17 would be re presented b y 0 10 1t "
-;;~ . " "~.
AB'C'D'E F ', where the apost rophe after
the .lett er indicates that the fli p-flo p is false
110
.. ,w.,.." " "
,,
'
,.
" .. .. .. . ..
and, conversely, a le t te r witho ut a n apos- Ill p fli p
. ., ., ., .. ., ..
trophe is true.
To simpli fy the matrix, a Karnaugh ma p 101
.
41
is construc ted sepa ra tely ( F igs. 3 and 4) fo r
the di ts a nd da hs to be sent. From Fig. 2 ,
units 2, 3, 5, 8 , 13, 14 , 15, 16, 18, 19, 2 1,
100
.. ,,
"
,.
"
,
"
.
and 24 re prese nt the dits to be s ~ n t in the
message. In the dit Karnaugh ma p (Fig. 4) F ig . 4. K arnaugh map of d its to be gen erat ed in
a I is placed in each box corresponding to DE W6FNO.
the number. An X ( no t the X which
rep resents a blank ) is placed in all boxes
after the st9p cod e numqer. These a re Oo Oo Oo Ob Oc Oc 0d Oo Q, Ot ,
"d on't care" condition s because the coun-
36V
ter will not count to these codes.
4 , 2 1(
F rom the dit Karnaugh map ( Fig. 5) it
can be seen that t he t hird unit of informa-
tio n is a dit and that flip -flo p A is tr ue, B is
true, C is fa lse, D is fa lse, E is fa lse, a nd F F ig. S. Unit 3 infor m at ion - d ir.
is fa lse, or ABC'D'E' F'. T o put this in
matrix form, the Boolea n algebra tells us
tha t this d il would be represen ted by a
diode connected to Oa lead (the true lead
of fl ip-fl op A), a not her to Qb , ano ther to
di~
000 001 011 0 10 110 Il l 101 100
"'"
0 00
0
f11e
'
' ' 7
'
. M AfRI X
[,\ . f,d,,
r1J
0 01 '" <>---,rlHf-...--+"'+-+-"1f-.,.....,-l--+-+.C:W.~
#ti &(~
,.B." "
II t0 I
~ \~.
BCD
0 II
,. 2:) , 2 1
J(
,. I
,. dco
0 10
.. " .. " "\it, v1. 21
it\Cc
11 0
.. .. .. .. .. ,, "
.,
oll
,. ,, .. .. " ..NOT
,. (/ ., E t?
60 AeCo
101
.. .. ., ., .. .. .. .,
80<
000
.. "
., ,. ,. .. "
,.
ABC
i&co
l8CE
... o--'---;...~-l--1-':>.--+=-1-.;--1----'"'-'--I---<
142
Oc' (the false lead of nip-flop C), another map. In Figs. 3 and 4 the "not used"
to Qd ',ano ther to Oe '. Since there are only portion would be used. The upper portion
30 units of information, nip-flop F is not of Fig. 6 illustrates the wiring of the
used. A line may be used over any of the cou nter; note that it mates to the leads of
symbols to indicate the same thing as an the matrix.
apostrophe. It wou ld normally take six
diodes (seven when the F flip-flop is used) Construc tion of the DIU
to send this unit of information. See Fig. 6. Since the pub lication of the first article
Actua lly, it would take six diodes (seven on the DIU nume rous people have sough t
when the F flip-flop is used) for each unit printed circuit boards for the unit. In ord er
of information in the message or 29 x 6 = to obtain the DIU logic, matrix, or power
174 diodes. This includes the diodes supply boards, write to Keith Whitehurst,
Box 538, Claremont, Ca lif. 9 17 11.
143
Power Supply 3. To transmitter keying circuit - iden-
tifier off: I 0 MQ; identifier keyed: I O.Q.
Figure 7 illustrates the schematic.of the
4. To modulator circuit - high im-
DJU power supply. Table II lists parts
pedance DI U oscillator output.
required. Any power supply, however, may
Note that all de input lines to the DI U
be used if the power output is 3.6 V with
logic should be filtered. In some relay
less than 5% ripple (including voltage
spikes). circu its the ou tput of a bridge rec tifie r 1s
used to direc tly key the transmitter relay.
Installation So the pulsating de d oes not key the digital
The signa ls normally received and sent identification unit, a 60 F capacitor (or
to and from the DI U should meet the greate r) should be placed across the relay
following criteria: supply.
I. From power supply - 3.6V d e, well It is worthwhile to note that the original
filtered and regulated. prototype, after two and a half years, is
2. From transmitter keying cir- still operational atop Johnstone Peak in
cuit - OV, transmiller keyed; approxi- San Dimas, Ca lifornia , sending out for all
mately 6V transmiller unkeyed (fi ltered). to hear - DE W6FNO.
144
An IC Pulser for the
Amateur Experimenter
n designing an amateur pulse generator, speed , larger fanout, and a larger selec tion
I the first consideratio n has to be com-
pa tibility with ICs. From the standpoint of
of devices a rou nd which to design.
This pulser was designed so that its
convenie nce , economy, o r versatility. it has output wo uld be compatible with DTL or
lm.:ome unattra ctive to build a11yrhi11g digi- TI"L and also with ECL The im plementa-
wl with o ut IC logic. Whal this means to tion of these two outputs is simplified by
pu lse r design is that th e pulser should be use of J('s that arc designed for buffer and
<:ompatible with the types of digital !Cs clock driver service. (The pulse genera tor
o n.: fin ds himself using most. section is shown in Fig. I , and the o utput
section in Fig. 2.)
Although RTL is the most widely used The free-running multivibrator whic h
IC logic fam ily in amateur circles, it is det ermines the basic "rep" rate , the delay
extremely do ubtful whether this situation one-shot , and the p11/se le11grh o ne-shot use
will last. The two types o f "current- HEP versions of MECL !Cs. The en tire
sinking" logic ( DT L and TTL) are far o ut basic pulse-forming system is similar to
in front of RT L in industry usage and that of my previous article (P11/.ie Ge11era-
gaining daily. The reasons for industry ror for rhe Amare11r, 73, Nov 1967). A
preference of DTL and TTL over RTL are number of improvements have been made,
several: bette r noise immunity. higher however.
-~ zv
145
different semiconducto r manufacturers.
01 L-f
BUF'F'ERS
fl These arc listed in Table I ; t hey arc 1w1
different families, but rather second-source
items.
Table I.
Srl('<1l1ed
Tv1>t- Manufa<.:1u1er Equivalent
"'"'\
=
146
exce pt th e th ree IJEP !Cs. Thc'e th ree IC\ Perform a 11ce
were rcv..:rsc-mo unttd to case layou t, al-
lowing sho rte r t race le ngth s. The p ulser will produ ce p ulses ( and
The 7-7 /8 x 7- 7 /8 in. p ane l is made to d elays) from a bo ut 50 ns to 30 ms.
fit a Bud CD-I 4 8 0 ca bine t which has DE LAY a nd P ULS E switches eac h h ave six
eno ugh panel s pace for all the controls and positions, and the variable con tro t a ssoci-
jacks. T he power transforme r is mo unted ated with e ach continuously varies each
( off th e board) inside t his cabine t. over at least ten-to -one . Pulse repe tition
.\ , to thL' DTL 1 I L output L".lp;1 h 1'11~ r;i tc i\ .1 <lJ t1 \ tahk fro m a bo ut 0 .5 Hz to I
o f l hL' pubc r. ii is tlcpc mknl P n 111 , '-'~" 1 Mll 1. with a six -p ositi on swi tch and variable
type of ou t pu t >lagc . L1L"11 01 1li- I C l co nt rol. The la be ling o f the three panel
ou t pu ts will <lrivc 24 l C L g.1IL"' I h;" ;in \Wlld1cs w;is minimal. wi th 20 0 k Hz to 2
1-:.C L fanoul o f 24 ). Bui the fan ou t ''' caLh 111. fin that order, so th at t he p eriod steps
o f th e t wo c urrenl-s in~ i n g Joi:il: n ut pu ts i' in the sa me d irec tio n as delay and pu lse
as fo llows: MC8 3::!P ~4 DTL lo;itl unit>. width ) fo r t he rep rate . These lab eled
SN7440N - 29 TTL (SN740 0i\ serio l freq uencies corresp on d to fre que ncies
load unit s, MC3025 P I lJ MTI L 111 lo a<l ll'i1/Ji11 each swit ch position , not the cente r
units. The load un its arc no t th e same fo r or eit he r e nd poin t. The DE LAY and
these three curre nt-sinking fu milies . so it i> P ULSE sw itc h positio ns were simi la rly
best to use the t y pe of o utput IC fo r the labe led f ro m 0.1 s lo I 0 ms fo r simplicit y .
sort of fam ily you u se most. It is q uit e all The regulated +SV and - 5.2 V are pro
right to use a ny of the types of curre n t vid ed b y a co m mo n power transformer and
sinkin g lo gic ICs with t he pulser (no ma tt er rectifier. Since the centertap of the trans-
which IC is used in the o ut put stage), bu t form er is grounded , the circuit may be
some reduc tio n o f fano ut may be expert co n sidered a s two full-wave rectifiers (one
enced with certain combinatio n s. positive and o n e negative) across the same
Fi g. 5. PC layo ur of IC pulser
147
it makes the best output stage of any of
the MECL II series. This is true because of
its exceptional ability to drive capacitive
loads.
Both the MCl460R and the HEP 232
(P NP power transistor) are diamond-shaped
and meant to be fastened to a heatsink.
They each have an aluminum bracket
attached to them to fulfill this requirement
for a d issi pa tor.
The entire circuit of the pulser is built
on a n etched circuit board. In fact, the
wafer switches arc assembled so that the
board is clamped in to the switch assembly.
The shafts of the switches are cut to 7 / 8 in.
to extend o ut the same length as the pot
shafts. T he enti re board is th en mo unt ed to
Circuit board, parts side. a pane l using 1-1 /8 in. spacers. Some care
must be exercised not to allow the spacers
to short any of the traces of th e circuit
transformer. Of course, the diode con- board to ground. This ca n be insured by
figu ration comes out the same as a fu ll-
using fiber washers bet ween board and
wave bridge: so an IC bridge l 11 l: P I 7 51 1
spacers. The one spacer in the power
used.
supply corner is intended to connect the
The positive rcgubtor is an i\IC I 460R.
board grou nd to 1hc panel, so no fiber
an inexpensive IC that provides cxcelknt
washer should be used at thal corner .
regulation with few external component s.
The negative regulator is an emitter fol- Construct ion
lower with a 5.6V zener and a german ium
1"11c entire pu lser is built on an etched
PNP transistor. Since the base-lo-emitter
circui1 hoard. including the power supply ,
drop of germanium transistors is about
whOSl' circui1 appears in Fig. 4. Figure 5 is
0.3V, 5.6V minus. 0.3V gives us clos..: to
a h.tlf-,calc eopy of the board.
the - 5.2V required for the ECL- ICs.
The ou t put section fo r lht ECL com- Since' !he clc:hcd board method has been
usl'd. some' specia lization in compo nents is
patible pulses is provided by an MC I 023 P.
Al though this IC is bi.lied as a duck dri11er, 11l'Cl"Ssary . T he C'omcll-Dubilicr BR I 000-15
148
Fig. 6. Component placement (etched sid e o f board shown).
149
IC/Photocell CompressorI AGC Unit
ne pro ble m with many compressors or Only the derivation o f t he control function
O audio age units is that th ey cannot be
convenie ntl y built in to an existing t ra ns-
is restricted to a n a udio frequ ency point in
a receiver or transmitter, since the IC
mitt er o r receiver, since lhe amplifier and ampli fie r used o perates at a udio fre-
co ntrol sections of the unit ca nnot be quencies. In an rf a mplifier, a n rf ac tuated
readily separated. Consequently , such units and rf control compressor could be
are usually placed in separate enclosures achieved. Although it has not yet bee n
and mounted in the micropho ne lead to a tried , it would seem that this la tte r
tra nsmitter or in the lo udspeaker o r head- approach might prod uce a highly effective
set audio ou tput leads of a receiver. SSB rf level compressor without the need
Most such conventio nal compressor/ for lwo side band filters as is required with
audio age circuits use transistor stages for SSB rf level dipping circ uits.
bo th amplificatio n a nd co ntrol func tions. Besides separat io n of the amplify ing and
It is difficult lo separate the stages physi- co ntrol stages, the use o f a photocell-lamp
cally unless additional coupling stages are mod ule also overcomes t he noise build-up
added , so t hat the a mplifying a nd control problem associated with convent ional au-
functi o ns ca n each be located where each dio comp resso rs. With such compressors,
ca n function best a nd where power and noise build-up occurs al the output of the
space in a transmitter o r receiver are most co mpressor during speech pauses because,
readily availahlc (sec Fig. I A) . witho ut speech input , the gain of the
The IC/ photocell unit to be described compressor rises to a high va lue and ampli-
overco mes most of these limitations. The fi es the self-noise of the first stage in the
amplifying and control functions can be com pressor lo a high value. Choosing a
sepa rated as desired ( Fig. I B) th rough the long lime constant in the gain co ntro l stage
isola tio n medium of a pho tocell-lamp mo - of the co mpressor will acl to suppress suc h
dule. Although described as a n audio com- noise bui ld-up, but o ne is quickly limit ed
pressor/age unit , the photocell-lam p mo- as Lo how far l his approach can be used. If
dule allows control to be achieved of a n rf the time constant is made too long, low
stage as well through its biasing netwo rk. level speech inputs to the compressor
I I
CO MPR( SSOR I TRANS MIT TCR I STllGES WITHIN
I REC[IV(R OR TRANSMIT TE R
I
I
AUDIO
IN PUT
' .. PHOTOC[lL,.
CO-.iPRCSSOlt
18 1
Fig. J. Amplifier stages of usual compressor ar e in ser ies with 1ransmit1er or audio chain (A).
Pho tocell compressor (B) works parallel 10 conlrolled stages and the phococell modu le provides
bot/J noise isolation from compressor amplifier and feedback isolalion belween conlrolled amplifier
stages.
150
following a higher intensity input will not > O
.....-~---~- 6 - 9\tOC
be amplified sufficiently. Since the ampli-
fier stages of the photocell compressor
0 01
need not be in series with the audio chain 0~1L
in a receiver or transmitter, its noise output ~e:!;. 1 ..v<>-l
is not renected in the controlled stages. 0
Also, the thermal-photoelectric interface
within the photocell-la mp module prevents Pl'IOTOC[Ll
01 LAMP MOOl.ILC
the coupling of noise or spu rious fre- CL AIRE X
quencies fro m t he audio amplifier of th e 0 o1 J; 0 ., CU .I JOO&
151
net work in a fai rly high impedance circuit.
In some units the microphone in put itself
or the interstage coupling point after the
first a udio ampl ifier might be used . Theo-
retically, the large resista nce range change
of the photocell resistive unit could pro-
duce voltage output changes in a high
impeda nce divider network of over 40 dB .
Another possible placement of the resistive
element which also prevents de from flow-
ing through it is shown in Fig. 38. As the
....: '" ' .. photocell module is driven harde r, the
Assembly of compressor/age units on Vector- resistive element increasingly shorts to
board. Photocell unit is shown next 10 output ground the audio bypass capacito r. Such a
transformer. Compressor level pot is ill other end
capacilor would be placed between a udio
of Vectorboard, and IC uni1 is in middle with
circuit components grouped around it. stages in a receiver o r transmitter to shunt
the audio signal to ground as compressor
action takes place . T here are various o ther
Construct ion placements possible for the resistive ele-
The photograph shows how a typical ment , such as in the bias line to the stages
photocell audio amplifier driver circuit can it is desired to conirol or even as a shunt
be assembled o n a piece of Vectorboard . element across a low level audio transfor-
The photocell module would, of cour~e, be mer. The placement that will achieve the
located remotely from the amplifie r cir- best control ca n be quickly determined by
cu it. A fin-type heat dissipatio n device experiment.
should be used on the IC, depend ing upon
the manufacturer's recommendation for AUOIO lEVCL
CONTROL IN
the unit used. A PC board type trim AECEIV(R OR
HUNSWIT TCR
pot entiometer, located to the right of the
/.:.
FIRST 001
STAG( >-i
IC, acts as the compression level control. AUO O
Placement
The objective in placing the resistive
elemen t of the photocell module in a
receiver o r tra nsmitter circuit is lo make
maximum use of its wide resistance change.
At the same time, the resistive element
can not be used such tha t it must dissipate '"'
over I / I 0 watt. One possible placement for Fig. 3. Possibilities for using the resistive section
the resistive element is shown in Fig. 3A, of the photocell for control of lhe signal ampli-
tude in an audio amplifier. Voltage divider
where it is used as part of a voltage divider method (A) and capacitance bypass mcihod (B).
152
T he sampli ng point fo r the compressor's level is achieved. Basically, the circuit
audio amplifier input is usually taken at should then be set, although one will have
some interstage point towards the high to initially adjust back and forth between
level end of the audio cha in in a unit. The the aud io level and compressor controls
sampling connection should have no effect until the optimum compression range is
upon the normal operation of the audio obtained.
amplifier in a unit. If possible, the sampling Summary
point shou ld be chosen such that the audio The photocell modu le is a very versatile
level control in a unit is located earlier in unit for compressor and age control cir-
the audio chain. The controlled point can cuits. Many more circuits are possible with
be either before or after the audio level it than the one illustrated here. For in-
control. stance, if one wanted to use a varying de
voltage as the control sour!:e for the
Opera tion module th is can be done by using a de
With the compressor control initia lly set amp lifier and direct output coupling in
for minimum gain, the audio level control place of the audio IC amplifier shown.
in a unit is set for the highest desired audio With some low power vacuum tube circuits
level. The compressor control is then ad- it is also possible to place the la mp of the
vanced until the audio level decreases photocell module in series with a cathode
significantly. The audio level control is resistor and achieve direc t control without
aga in advanced until the desired output any amplifier at all.
153
AC Switching with Self-powered IC'S
cure-all to ac switching problems? age; that is, when the power line voltage
A Nol so! But RCA's little CA3059
integrated circuit device goes a long way in
crosses zero. This happens twice each
cycle, or 120 times per second at 60 Hz. If
eliminating many of the bugs we inherited we switch at this precise moment, no
with Tesla 's genius. Formerly known as a current is flowing through the switch; thus,
monolith ic silicon zero-voltage switch, it is radio-frequency interference is eliminated,
now more handily identified by RCA's as is contact wear. Incandescent devices,
"CA" number. It is no bigger than a 2W such as tube heaters and pilot lights, which
resistor, yet it contains its own power have very low cold-resistance, are heated
supply and all the other functions shown in up "gradually" during a half-cycle if volt-
Fig. I. When used with a half-dozen o r so age is applied at one o f the "zero" crossing
other common components, it does many times. They are, therefore, not subjected to
wondrous things. Several are described in a d estructive high-current surge as before.
this article. In the same manner, we reduce the high
Basically, we would like to accomplish surge current and minimize the peak
switching when there is no incoming volt- inverse voltage imposed on our rectifiers
---------------,
I
I
I
I
I
60Hl
120~40
VIC
NTC'sE
Fig. 1. Functional block diagram of CA3059. Few ex1ernal components are required in this basic
circuit when used to trigger the gate of a triac. The sensor may be a temperaiure-sensitive thermistor, a
photoeleccric cell, or a simple off-on swicch.
154
when initially turning on a power supply. RCA-40526 se nsitive-gate triac with th e
The zero-crossing detector in the CA3059 CA3059 because of the low latching-
synchronizes the output pulses of its cir- current requirement of this triac.
cuit with the time of zero-vo ltage in the ac Figure 3 shows an application which is
cycle. Figure 2 shows th is relationship . useful in switching noninduct ive loads.
But. to be useful in switching normal Lamps and electric heaters can be con-
loads, an ex ternal power device is needed. trolled in this circu it by a simple switch,
Thyristors such as SCRs (half-wave) or photoelectric cell, or temperature-sensing
triacs (fu ll-wave) are ideal for this purpose. thermistor.
Either one is readily triggered by the I 00 s Inductive loads up to 200\V can be
pulses from the CA3059. The zero-voltage controlled by the configuration shown in
switch is designed primarily to trigger a Fig. 4. In this case, termina l 12 is con-
thyristor that switches a resistive load . nected to terminal 7, and the zero-crossing
Because the output pulse supplied by detector is inhibited. Whether a "high" or
the CA3059 is of short duration the " low" voltage is produced at terminal 4 is
latching current of the thyristor beco:Ues a then dependent only upon the state of the
significant fac tor in determining wheth er differe ntial comparator within the CA3059
other types of loa ds can be switched. (The integrated circuit , and not upon the zero
latching-curre nt value determines whether crossing of the incoming line voltage. Of
the thyristor will remain in con duction course, in this mode of operation, the
after the trigger pulse is removed.) Pro- CA3059 no longer operates as a zero-
visions are included in the CA3059 to also voltage switch. However, for many applica-
accommodate inductive loads. tions that involve the switching of low-
For example, for load currents that are current inductive loads, the amount of RF I
Jess than approximately 4A rms (or that generated can frequen tly be tolerated. Th is
are slightly inductive), it is possible to curcuitry is partic ularly useful as a differ-
re tard the output pu lse with respect to ential comparator. Such comparators have
the zero-voltage c rossing by insertion of found widespread use as limit detectors
ca pacitor Cx from termina l 5 to terminal 7 which compa re two analog input signals
as shown in Fig. I. The insertion of and provide a go/no-go output, depending
capacitor Cx permits switching of loads upon the relative magnitudes of these
that have a slight inductive component and signals. In many indust rial control applica-
that are greater power than approximately tions, a high-resolution, fast switching unit
200\V. For loads Jess than 200W it is is not essential. The CA3059 is ideally
recommended that the user empl~y the suited for use in such applications.
T ypical IC
PAR AMET E RS CA3059 Comparator
1. Sensitivity 50mV 2 mv
2 . Switching spoed >:ios 90 0$
3 . Output drive 4.SV at 3.2V at
GATE capability ~mA ~mA
PULSES U l'L 111 ..
155
..
.,. .
, ,.
~
"""
.... '""'
11'()\fA(
Fig. 3. Controller for res1s11ve loads. The RCA Fig. 4. Differential comparator circuit. The load
2N5444 triac can be used for load currents up to is switched on when the voltage difference
40A. The RCA 40668 triac wi/1 switch inter between Vs and VR becomes less than 50 V.
mediate loads and the 40526 will handle lighter Note the jumper between terminals 7 and 12,
loads and those which are somewhat inductive. which deactivates the antiRFI feature.
The circuit shown in Fig. 5 is connected This circuit can thus switch a load at
so that after switch S I is closed, electronic zero current regardless of whether it is
logic waits until the power-line voltage resistive or inductive. The maximum load
reaches zero before power is applied to th e current that can be switched depend s on
load. Conversely, when the switch is the rating .of the second triac. If it is an
opened, the load current continues until it RCA-2N5444, an rms current of 40A can
reaches zero. be switched.
INOUCTI V(
l<l>'O
zl
~
112 w
,.
l/2W 1,r
200
v oc
Fig. 5. Practical power switch. Heavy inductive loads can be handled by adding the 0.1 F capacitor
and by cascading the two triacs (RCA 40526 and2N5444). With the taller heavy-duty triac, Rl should
be the highest suitable value less than 10 kil
156
..
r-- - - -;:-- - -1 ~:;' T:
1
I ) t ouO'" +
o~._,,,.,.........-<;v--~--,~,,,..,,__~~ ~~~~~ :__- ~1
..... ..",
I I
..
:~~(.
I "
~ -- - f - --.._~~--~~~-~~+-~~+-_._~~+-~~--4---'
1
ltt C& ~'!I'
L- -t ~T~F!O~'!;_ _ - - - - -
Fig. 6. Schema!ic diagram of CA3059 zero-vol!age switch. This minia!ure device con!ains !he
equivalen! of 10 uansis!ors, 14 diodes, and 10 resis!ors.
157
2 Watt, 6 Meter Transmitter
Using the Heterodyne VFO
-y 12V IOOuF
I~
I rh ~
5-80pf I
I
120mw I
~Hz, -0-1--- I 1000_[/n _cm _cm FROM MOO.,
VFO 3-30pF
___ _J 620HM
LI I
!r80pF t:~r-~~~~-;:::::::~t:J~~
@RF
C3 J-, OUT
Notes: Baseboard Is coppercled,
8%" lg. X 4%''. H EP75 cose is
also th e colloctor, soldered to
L2. L 1 Is 9 turns No. 14 bare
copper, *"O.O. x 1lS" lg. L2
7X: turns coppor strap %" w ide
x *"0.0. x 3W' lg. Collec
SHIELD_/th 25
to r Is tapped et 2 turns, output
at 3 turns.
158
input power (not too likely), and where lnpul Matc hing and Base Circui t
you don't mind adjusting cable length. After trying out various combinations
However, if you're building up from onlv of circuits as shown by RCA, Motorola ;
120 mW of vfo output, you may want Fairchild, and o thers, the circu it o f Fig. !
maximum po wer transfer along with the was judged best. Inductor LI can also be
ability to match vario us cable le ngths, and an air-wound coil 5/ 8 of an inch O .D., 8
direct inputs (as in the case of packaging turns per in ch, with 5 1/2 turns ( not too
the entire rig in a box). critica I).
The circuit of Fig . 2 is reliable and good J ust a word of cautio n here about
for a "quickie" any time. But il does not overloading receiver inputs while testing
always furnish maximum drive or best transm itters. While check ing with the lab
input matc h unless the cable le ngth and receiver, with its ante nna only I 0 in. away
output tap on the vfo output inductor are from the I W rf output of the ampli fi er, the
adjusted. These latter are of course indica- receiver was completely blocked out and
tions of mismatch, but fo r short lengths of detuned by the rf. The add ition of only I
cable there is little loss, so you can operate in . of wire as a test antenn a bro ught things
that way if needed. back to norma l. A very pecu liar effect :
DC Base Lockup expect almost anything when you start to
This nasty little trouble has not been run power. In solid-state VHF, 'power "
mentioned in the rather large amount of refers to anything over I W.
literature persued here through recent Emitte r Circuit
years. I can just hear some know-it-all lads No trouble here. Two capacitors were
say ing, " Ho, Ho, he's just discovered the used for bypassing, wh ich are no t really
Poniatowsky effect." Well, maybe so , but I 00% in para llel. Thal is, with the ground
whatever its name is, it arrived here and I leads of the capacitors going to slightly
don' t like it! Here 's what happen s: With a n differe nt ground points and the two leads
rf choke from the base to ground and no cut ti r:g down the lead inductance by a
resistor, and a n rf input driving the collec- large fac tor, the emit ter is pretty well tied
tor to a bout 200 mA , cutting off the rf down to the ground plane rf-wise. A 2sn
excitation does not cu t off the collec tor pot controls rower output fro m abou t
curre nt. There is no rf invo lved and il 0. 25 to I .OW with 120 mW of rf input
seems to be a de type of hangup. J ust for power and 2W of de power.
fun I'll wail until someone tells me its Due to the base resistor requirements,
name. In the mean I ime, back on the no limiting resistor was needed in series
breadbo ard , the cure was easy. Just include with the 25.Q em itter pot which operates
a Ii Ille resista nce in the base circ uit. That's nicely as an rf power control (although 2W
why the 20.Q job is there. d e in put does not particularly stra in the
HEP-75) . A 100 F capacitor was later
shunted from the emitter to grou nd for
better modulat ion properties.
Collector Circuit
FROM Several requ irements must be met he re,
VFO c some of them not usually compa tible, but
things worked out quite well as you will
see in t he results section. A good high-Q
ind uc tor is desirable for maximum selec-
tivity when loaded by the an tenna, and at
the same time good heatsink ing is needed.
Fortunately , there is a design tha t will
accomplish both of these requ iremen ts at
F ig. 2. Simplified input circui!, 2 wan, 6 meters.
the same time. T he secre t; Plenty of
(Temporary) copper. Thus the edgewise-wound copper-
159
broadcasts reach you through a minimum put it on the air is to use an Amperex
of ten or a dozen transformers, so why TAA-300 integrated circuit, as shown in
worry about one more? Fig. 5. This little gem is actually a minia
So, for a modulation transformer, no t ture hifi se t all by itself, and it really tells
having found the ideal as yet, we use you what your own rig - with your own
another, back to back with the 4-TR voice mo dulating it - sounds like to others
winding of 8.Q, The nearest to 75.Q I on the band. To do this you need also a set
could find was the Lafayette AR-50 I, of earphones, with good padding to keep
which has an 8.Q winding on one side and a your voice from reaching your cars through
125.Q winding o n the other (centertapped). the air. The Lafayette Model 8X stereo
It worked out fine on the air, in the circuit headphones ($7 .95) are excellent for this,
shown in Fig. 4. As the power goes up on with 8.Q impedance per phone. Just con
this type of rig, with one or more HEP-75s, nect the two phones in parallel in a
possibly to I OW or so, the modulation three-circuit jack.
impedance will drop below the 48.Q region. Connect a tuned diode receiver in front
In this area there is the 1OW "universal" of the TAA-300 to pick up therrf from the
transformer with taps at 4, 8, and 16 o n transmitter, and use the Amperex external
one side and 8, 12, 16, 24, and 48 o n the circuit as shown in Fig. 5.
other side, at only $3.95. A major benefit The T AA-300 is an excellent example of
of low-impedance solid-state devices now a modern audio-type IC using eleven tran
becomes evident, as you don't have to pay sisters and five diodes, with a frequency
for much copper. response of 20 Hz to 25 kHz. Naturally ,
Jn my transmitter, only the final is. you don't need all that range for voice
being modulated. Tests were carried o ut communications work so the following
using only a 6V battery on the modulator modifications were used to cut down the
because the only 9V batteries I had on highs and reduce the lows. Replace Cl, 0 .6
hand were those little jobs for Jap radios F , with a 0.0 I, and install a 0.005 across
- and they will not furnish 500 mA ! the input jack. Regular treble and bass tone
Ex ce llent modulation reports were controls can be made up but they should
obtained on the air, however. be put in a minibox to minimize hum
A good idea for checking your own because there is a lot of gain at 60 Hz in
modulation on a dummy load before you the little tin can. I also put a pot in place
+9V
3 6 4
6 an
OUT
2
AMPEREX
TAA-300 9
(Use 1 F) AF AMPL.
8
>--1 C1
SET FOR
8 mA TOTAL
FROM
DIODE
DETECTOR
10
- 47
l.05)
F
162
of the 48.Q feedback shunt resistor, SS miles away. His report: "Modulation
because making the negative feedback very very good."
control variable has very interesting possi- ... KI CU .
bilities, as you will see when you operate
it. The TAA-300 also makes an ideal
amplifier for any amateur receiver using
transistors.
On The Air
163
Build Your Own 2m FM
164
don't. Then you have to lay out another $3
or $4 for a jeweler's chuck, which will take
a No. 80 drill ( l 3 Yz mils).
You do not have to drill the compo-
nent-lead holes exactly to size but the
closer you do the more rigid the parts will
be when mounted.
0
~ ..
series, which are just \12 in. long.
No. 48 or 49 bulbs are good fo r
check ing rf as you go along the multiplie r
chain from stage to stage. You sh ou ld JI l2
165
will show. This indicates the real frequency frequencies on a signal generator to a
of the main body of the rf present. Some homemade set of wavemeters.
transistor collector circuits not tapped
down on coils are especially notorious for The meter. This should be as sensitive as
this, and may exhibit two frequencies at possible. Lafayette has good ones down to
the same time. For example, there may be 50 A. Use a tap switch to put resistors in
energy al 72 and 96 MHz present. This is series to bring the voltage range up to I 0 V
an indication of mistuning, or overloading, or so for use with an active portion of the
or both. Tap the collector down on the rig such as the I W final circuit.
coil , don't load it so heavily to the next
stage, check it carefully with the diode The AF Amplifier. This item should not be
meter, and don't worry about a small neglected as it is at times a great aid to
remnant of off-frequency energy. After all , gett ing a trouble-free, noiseless ca rrier,
a multiplier is bound to have some of this which you can then modulate and be
present. Just get the main amount on proud of. The valuable RCA handbook,
frequency and be happy. And be sure the "Transistors, Thyristors, and Diode Man-
next stage also peaks on its desired fre- ual," has a lot to say about " discontinuous
jumps in amplitude or frequency as various
quency.
levels of d rive are encountered." These
A grid-dipper in the diode position can
little termites can be seen on the meter or
also be used for this work. A one-turn link
heard o n the af amplifier or can show up
around the low end of the grid-dipper coil
on both. Figure 3 shows a mounted version
and a cable will get you into small p laces in
of the af amplifier used here for this
a rig where you cannot insert the whole
purpose. It is a worthwhile and hand y little
dipper.
piece of equipment to have in a lot of
The diode detector. Figure 2 shows the situations, in both receiver tests and trans-
schematic of one of these useful pieces of mitter tuneup. Just plug ii into 13 of the
equipmen t which allow you to listen to diode detector in Fig. 2 and hear those
your transmitter multiplier stage as you unwanted clicks, whistles, rushing noises,
build it , and check the act!;lal frequency at squeals, etc., coming from what you may
the same time. I have a collection of them have wishfully thought was good clean rf in
lAfA't'ETlE your multiplier drive!
AUDIO AMPI.
Im portant notice! Overdrive is especial-
ly lo be avoided in multiplier chains with
non aOaaO transistors. Superregeneration is one of the
indications. Believe me, it can be a very
nasly bug!
Diode detector cable pro bes. llave a col-
lection of these on hand as in Fig. 4. You
GAIN TON(
can use them also to feed rf into a pilot
J2
light, connect up to your lab receiver, etc.
Fig . 3. Handy af assembly, top view Handy meter ja cks. Figure 5 shows an
elementary but nexible and useful meter-
here covering from 125 Kll z to 10 GHz. ing method for checking total or only one
With a good variable capacitor you can stage current.
generally run over three to one in frequen-
FOUNDATION RECEI VER
cy range - up to the UHF region at least.
From there on up things ge t a little more The basic design shown here is for a
difficult. low-cost single-conversion utilit y receiver
for 2 meter FM; particular attentio n is
These "receivers. because that's what given to easy-to-build i-f and discriminator
they really are, although of low sensitivity, modules for the I 0.7 MHz section. The rf
are especially helpful in transferring known is !unable from 144 to 148 MHz, with a
166
switch for AM use. This is a complete baseboard wi th small brackets made from
portable receiver, 1101 tied down to a large copper-clad . Small pointer knobs allow
ac.; com mu nica Lions receiver. peaking of these circuits. The rf stage has a
Discriminator action, with sample, is trimmer capacitor feeding the base which is
shown for easy understa nding and home- quite useful, resulting in a welcome balance
brewing. Double conversion with crystal between gain and self-oscillation. The mixer
control can be added la!er. also has a trimmer for its base input, which
The schematic of Fig. 6 shows how easy permits a selectivity adjustment for this
it can be. Remember, this is just a basic circuit .
receiver which, without double-conversion, The tunable oscillator was mounted on
is a rela lively broadban d, easy-to-tune job, the Miller slide-rule dial for mechanical
but it sure pulls in those interesting re- stability as shown in Fig. 7, and works
peaters! quite well - with the broadband i-f of
course. As I write, the repea ter band just
below 14 7 MHz is giving out with various
~"""""
IOR HIC.H tMPlDAh'C[
OVTP\.lf HSTS
JAC
"'
~
~O l80R10M'G
"''""'"'""'
COIPl.EO
'"" """""
f(lp.fl PNOll
(~
- _JL b +
TO . . .l.fl40till0
167
energy out on L6, such as to deliver 5V de
out of the diode, you've got a good front
end!
r 0~u"~~~
ule will eli;ninate the need for ave.
Trimmers are shown for C l and C2 , but
fixed capacitors of the proper value ma y be
"-::- _
- .. 0======c;
used to allow tu ning of the i-f coils at I 0.7
MHz by the variable tuning slug cores in L2
and L3 . Note that these Miller half-inch
,.L,,,
......
gems have very good electro magne tic as
we ll as electrostat ic shielding. due 10 the
cup-core type of construction , and are
available fo r use from 30 Mii z down to
I '.'S k l.ll .. They .:an also he easily ope ned Fig. 7. Dialmounted local oscillator
168
SHELD
+
I
I
I
I
I L3
+ OSC ll.LATOR
MOTOftOl.A
HEP' !le
SET TUNE
4.7K
r' IK
169
great time listening to those repeaters. F M TRANSM ITTER STRIP
llowcver. p lease bear in mind that the i-f The transmitter section measures I in.
sckctivity at 10.7 MHl is not sufl1c1cn1 for wide by 8 in. long, and it puts oul over a
1:ontinuecl use on 2 meters. except to tune watt on 146-147 MHz, with low-cost
\!1> lhe front end and to get acquainted with components. This miniaturized transmitter
what's going on in your neighborhood. is my logical step toward design and
even though it is fun' It will also start you ultimate const ruction of a ' 'shirt pocket"
off on discriminator work. which I find has portable transceiver. The parts for that one
><.llllC very fascinating aspects, such as the jump up a little in cost. because it takes a
extn:mc selectivity of the output curve for lot more tools to make subminiatures, such
as slc.!reo microscopes. special materials and
skills, 1cwelers tools, and so on.
50 pF .I
~AM OUT
47K m-
*ARCO 424
19-l!IO pF
47K
170
.....
- .....
....
cross over each other, and that the power
amplifier is well away from the oscillator.
Ultimate size is actually up to you, and
you can judge for yourself after laying out
the parts on hand . If you send for a
l~~[~l~~[~J.1.!~j~[~ll:t.1
the assembly, and after the wiring .YOU
could hardly tear them apart with your
fi ngers. I coun ted no less than 25 ground
.o~oo io no uooc
connect io ns to the copper on my own I 0
.
H::~~.:\~~.:1::.:1:r~1
in. strip.
Yo u ca n a lso make strip modules of any
F ig. 12. Dis.:riminator de output curve.
length you want such as modulator af,
receiver sections, etc., as shown in the
The B+ lead is red subminiaturc wire
receiver plans. This makes the task of
and goes from filter to filter along the
repairs or- improvement changes easier later
strip. The rf lead is green and goes from the
on. These shorter strips can be fastened
coil out put lap of each stage to the next
end on to each other and fastened down to
base coupling capacitor: tht: rest of the
the baseboard as shown.
connections practically fall in place for
soldering together. As you can see, there is Miniature F ilt ers.
still room left over!
Do not try to make up frequency
The detailed planning of the holes to be multipliers without rf filters in the de line
d rilled becomes a large portion of the ti" each stage, unless you ca re to experi-
work. Figure 15, component side and top ment with rf phasing in battery leads and
views, shows how to start this off. The that isn't good! (Every time I leave out the
next step is to make a li fe-size drilling filters 1 get into trouble!) You can make up
template using the components you have
&llll'lllt
o r intend to use. I mention this because ""'lOO
!t01f0"" VfllOI
INSULATED MOUNTING
STRIP
r
I in. ~
~ 0
jm11111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111
,,a,,./ /If;,
// -.// -/. V. /, . .///
;;7,~ ~;{ /j;/ ~f"v ../y
A , /,'/-' / ,(/.
__i1
f()'-, /7 ,
- ,112 In.
L ,
171
several filters between the high-po wer out-
put stage and the sensitive first stage.
However, if the filters are very good you
can bring the battery leads from each stage
to a commo n point , but this mu st be
checked carefully if you have to do it.
172
rf output jack which the usual FM receiver plifer from Lafayette, at $4.95 on hand,
cannot distinguish from true FM. Being that's what was installed, with a slight
crystal-controlled it is used by practically adjustment of the feedback resistor. This
all the FM mobile and base stations in the had nothing to do with the FM unit, it just
U.S. so it is 100% okay here. And of course happened that the Lafayette amplifier
with the crystals in there, yoti will be on sounded and acted awful funny at first.
the amateur FM channels, providing you And no wonder -, it was oscillating up in
buy them right. You have to pay around the I 00 kHz range! After trying to bypass
$7 for these but it seems well worth it. and decouple almost everything in the little
Certain designs of the af section of the brat my eyes bega n to focus on that
phase modulator, its tuneup, and the con- printed lead going from the sn output
nections to the phase modulator can be connection over to near the input, and sure
troublesome for the homebrewer, so con- enough that was it: too much feedback!
siderable time was spent to make it as An additional 50 kn resistor in series with
simplified and easy to adjust as possible. It the one already there did the trick and
also can be used in the receiver section as from then on nothing but good af came
the af amplifier because the frequency out. I mention this because it could happen
correction is done outside. The use of an 8 to you too. My 3W job, also Lafayette, is
or I 6n output connection into the phase suspect, possibly the same simple trouble.
modulator emitter circuit h"elps to stiffen The af output needed to drive the phase
the af drive and keep it clea n. modulator emitter is several hundred milli-
Phase modulation af sections in com- volts, and the low impedan ce allows the
mercial rigs are often qualified as "audio usual rf bypass capacitor of I 0,000 pF to
conditioning," or "processing" circuits, act simply as an additional af filter, which
which they are of course, but don't let that it does.
bother you. Excellent FM quality can be As a result, the entire tuneup is d one by
adjusting the value of the emitter resistor
and t he phase modulator tank tuning coil.
Neither are actually critical but should be
adjusted while listening to the 146 MHz
carrier on a good amateur narrowband FM
receiver. The emitter resistor will be heard
to kill the modulation when going much
below 2 kil and to bring in d istortion o n
large amounts of audio when going a lot
more than 2kr2. This latter condition also
causes a drop in the rf output. You may hit
Fig. 19. Phase modulator interconnect circuitry. it right the first ti me with the 2kr2 value;
I'm just pointing out that this resistor is
obtained by the use of an inductance of worth checking up on for a fin al value
large value, placed outside of the af ampli- when adjusting for best modulation.
fier, in the noncritical low-impedance out-
The actual phase modulallon resulting
put circuit. The inductance cuts down the
from VHrying the emitter voltage with
extra high audio modulating frequencies
caused by the phase modulator's tendency
to make the FM deviation directly propor-
tional to the modulating frequency, which XTAL
emphasizes the highs too much unless
corrected. Being outside of the af ampli- C, EMITTER
173
audio is adjusted by tuning, which is also - 44 dB , costs only $3.82 amateur net , and
smooth and noncritical. I used the tried- sounds good!
and-true method of listening to my own Almost any desired amount of highs and
voice with plenty of audio on the receiver lows can be obtained or surpressed by the
and a set of well-padd ed earphones (you manipula tion of the LC values in the
can get a very useful set for under $10 a~ modulator. If you use a Miller .9009
Lafayette) which keep your voice from wide-range adjustable inductor, 180-
reaching your ears directly through th e air. 750 mH you can hear the difference as you
It also cuts down aud io feedback. adjust the core in and out.
I started out with a large-scale layout
Tuning with af going into the phase fo r the parts, but you may wish to skip
modulator as per Fig. 19, you will notice that and go right to a life-size layout as in
good strong clean FM on either side of the Fig. 20. To make the life-size drilling
peak tuning. These points occur before the template, lay out the componen ts one after
146 MHz carrier output starts to drop from
the other, "standing up" o n a ~ in. strip of
detuning the phase modulator tank, so
good-grade white cardboard a nd mark the
don't worry about that part. In a ny case,
component lead holes, which should result
you are supposed to be following the phase in something similar to Fig. 20. A nice
modulator with enough saturated class C
fea ture of the cardboard method is the
multipliers and amplifiers to prevent any easy punching of the holes and the way it
variation in amplitude (otherwise known as holds the drill as you go through the strip.
AM!) I say "supposed to" because you Tape the template in place onto the insu-
don 't au tomatically get this condition. You lating vertical strip. Do not use anything
may have noticed an unduly large number
that melts under heat , though . Even if you
of tubes showing in ads for surplus com-
ruin part of the strip, or want to make a
mercial FM sets. This large number is due large change of one stage you can just saw
to the designer's wish to get all the benefits
that out and make up another section and
of FM into his package. In one box if
go ahead.
possible. You have to watch very carefully
when using !Cs for modulators, they tend
The 8 MHz Oscillator
to pick up rf and generate feedback with
their wideband audio circuits and some- Figure 2 1 shows the schematic of the
times as many as I I or even more transis- crystal oscillator stage. Note the apparent
tors in one little can. Just a word of what use of negative fee dback with the base
to look out for. It's hit me mo re than once. return through the crystal to a tap on the
Also, d on't put more than the specified ind uctance. It is only apparen t though, as
voltage on IC amps. You- can easily drop the crystal reverses the feedback phase,
d own with a resistor and a large bypass making it positive. It is a very powerful,
capacitor. sure-fire circuit.
t used my favorite mike on the input, The tap o n the coil also provides a good
the Asiatic 150, my favorite because it low-impedance matc h for the next base
only weiehs 3 oz, has the most ou tput, input. The coil itse lf is made from a Miller
9050 shield ed coil which has magne tic as
well as electrostatic shielding, and a good
adjustable core that works good mechanic-
LI ally (which is mo re than you can say for
-:i:r~~ :s ~ !SU TUT)
some of th ose types of cores).
TO AWf..111'11."""
"4ASI: llOOUl..ATOft
Remove the aluminum can by bending
back the fou r holding tabs and wind on
th ree turns of 30- o r 32-gage silk-covered
wire onto the existing winding of the coil.
Be sure and wind them in the same
Fig. 21. 8 MHz oscillaror schematic. direction as the turns that a re already
174
there. The oscillator coil will then look like "plop" and gradually build up on the other
Fig. 22, and is ready to mo unt o n the strip. side as you tune.
The wiring on the lead side of the strip I always start with a large calibrated
is shown in Fig. 23, where most of the variable capacity at C3 (some 500 or I 000
leads arc seen to fall in place quite well. pF, made from an old BC set three-ganger)
Insert the component leads through th e ahd then put in fixed values so that the
strip and bend them slightly in the direc- iron core tuning slug in the 905 0 coil tunes
tion they will go, such as the two base properly about Yi in. under the winding of
resistor leads which are bent towards the the coil.
base lead , as shown clearly in Fig. 23. Power can be adjusted by the emitter
When all the leads to be soldered in one resistor, and feedback by the number of
place are all touching each other, a fina l turns between ground and the osciJlator-
dressing can be done followed by soldering. coil tap. (These arc of course the number
In the examp le mentioned, the base lead of turns added to the Miller 9050.)
has th ree oth er wires soldered to it, a wire A 48 or 49 bulb, rated at 2V and 60mA,
from the crystal, the I kn resisto r, and the shoultl light up with about 50 to 100 mW
5 kn resistor. worth of rf with a 50-300 pF trimmer in
The ca n of th e 9050 coil has a tab series, as in Fig. 24. When the oscillator is
which should be soldered to ground. The properly tuned and under good power
ground lead o f some resistors (or all o f control via the test pot (in Fig. 24) and the
them) is not routed through the strip but is plus voltage is checked for the voltages you
soldered to the baseboard o n the compo- expect to see, the next stage can b e
nen t side of the strip. assembled. Of course if you wish, you can
When the oscillator is assembled and mark out the whole strip template, driJI all
wired, B+ can be brought in and the unit the holes, and mount and solder all
tested for rf. Some 5- 10 mA of curren t components except the coupling capacitor
and B+ to the next stage. This allows you
to test the oscillator by itself.
The 8 MHz Amplifier- Phase Modulator
-
HAllT !W This stage (Fig. 25) is not critical, other
OlllGINAL
than to keep the input base coupling
capacitor at a low value to avoid self-
oscillatio n. The only requirement is that
t~e tuning should be correct for phase
modulation.
Use the same methods of assemb ly,
wiring, and tuneup for power output as
"Fig . 22. Miller 9050 coil with added turns. with the oscillator stage. You do not nt:o.:ll
much gain, if any, in this stage.
should register and as soon as the oscillator The 8-24 MHz Tripler Stage
coil is resonated to the crystal frequency, A frequency multiplier has the advan-
the oscillator should show rf output to the tage that generally (though not always) it is
8 MHz tuned diode test set connected to free from self-oscillation, due mainly to the
the output tap lead . output and input circuits being on differ-
Check the oscillator carefully on a sharp ent frequ encies. The bias requirements are
receiver fo r its frequency-ho lding ability different in a tripler from those of a
while tuning the slug in and out o f reson- doubler o r class C straigh t-through ampli-
ance at 8 MHz. Actually this will be near fier, but this can be adjusted simply by
8. 130 MHz. (With a multiplication of X IS, varying the emitter resistor during tuneup
this should land on whatever 2 meter FM for maximum output o n the d esired fre-
chann el you're aimi ng for). It should come quency. Figure 26 shows the schema tic o f
into resonance on one side with a good this stage, where the base input coupling
175
Fig. 23. Oscillator wiring diagram (lead side of Fig. 24. Oscillator test setup.
strip).
..
Fig. 27. Wiring diagram, 8 MHz tripler. Fig. 28. 24 MHz tripler schematic.
176
Fig. 25. Schematic of phase modularor/amplifier. Fig. 26. 8 Mlb tripler schemarlc.
.........,.,.,
=.>-4.~-!-....J.r
l"'
Fig. 29. 24 MHz II'ipler wiring diagram. Fig. 30. 73 MHs doubler schematic
When the stage is assembled and wired for power control and noise , you are read y
and under test as done with the previous for the next stage, a doubler.
stages, once again, look out for those The Doubler to 146 MHz
undesired harmonics, especially the 64
This stage uses a Motorola HEP 75
MHz one in this case. It'll sneak up on you
(2N3866), always a lively powerful one for
i(_y9u 're _not real careful!
VHF. The schematic, shown in Fig. 30, is
quite similar to the others except for the
..... different transistor and another coil tap .
The base input capacito r worked out at 25
...... ..
>---?
......... pF maximum, with a 3951 emitter resistor
to keep the power up for maximum input
into the final stage. The collector lead is
cut off and the collector connection is
Fig. 3J. 146 MHz amplifier schematic.
made by solde ring a I /8-in.-wid e soft and
thin copper strap, which increases the heal-
The inductor may be fastened to the sinking as well as rf conduction, directly to
mounting strip with a nylon screw (Fig. the HEP 75 case.
29) for ruggedness. The variable capacitor
does all right standing up on end with the
fixed plates soldered to the baseboard , and
the movable plates brought out through
the strip with a piece of 16- or 18-gage
wire, where it is joined up with th e
collector, as can be plainly seen in Fig. 29.
Clip on your diode detector for power ''' '!'''' 4 1'''''''1 '' '''61'''1'''1''' ''61'''1'1
checks and freque ncy. You ca n't check this
latter too o ften , believe me. After testing Closeup of wiring side o f insulated srrip.
177
small iron, rub the iron gently on the case
two or three times for about one second
only to effect a good joint for the collector
strap over to the coil. The inductance is
not actua!Jy critical but should be correctly
tuned up and tapped for the collector as
well as the output tap. After all you do
need all the power you can get into that
final.
You could use a little larger emitter
resistor, for a little less current, but here
again power is a point to watch. I find
about 50 to 120 mW of rf at 146 MHz at
the output tap, depending on the plus
voltage also. Of course, you can play
around with up to 18 V if you want to push
out a bigger signal. Before buttoning up
this stage, check it once more for fre-
quency, please.
Power Amplifier
Refer to Fig. 31. Everything went along
nicely with this one also, as in the previous
178
Note that with sma ll emitter resistors of _collector circuit. Without the test b ulb or
under son the collector current can get any antenna loading you can expect self-
pretty high , so always keep a t least I on in oscillation as the HEP 7S gives plenty of
series, as you test for the best emi tter action on 146 MHz. The o utput tap can be
resistor value. Don't forget that I W is 100 led into the small but good and very useful
mA at 1OV. And for a watt out you will son ca ble (RG-174/U). Th is cable will
need more than I 00 mW even a t l SV. then go to your changeover switch or re-
" Big" transmitte rs (SO wa tters) use as low lay fo r the final assembly.
as 0 .1n at t imes in this place, and in some
of the new heat-su nk jobs (out o f our price As a final note o n the transmitter, each
range) the emitte r goes directly to gro und . tuned circuit of the multipliers and finals
I suggest currents of not over I 00 mA should also be adjusted while listening to
for this stage. The arra ngement shown uses the carrier modulation. I didn't find any of
about SO to 60 mA, depending on drive these at all critica l but " the books" say to
from the tripler, B+ voltage, and output do this to assure absence of phase-shift
loading. distortion in circuits afte r the phase modu-
Two bypass capacitors ar e used in the lator.
collector circuit. A test bulb (SV at 150 So, good luck with PM, it doesn ' t seem
mA), in series with a 1- 12 pF trimmer to too tough to build if you avoid t he fancy
ground , indicates rf outpu t a nd loads the stuff to start with.
179
The Phase-Locked Loop Comes of Age
t's been a little more than 13 years now about I 0 mA from an l 8V supply, and is
I since the idea of "synchronous detec-
tion" entered the ham radio world - and
hardly as large as a commemorative postage
stamp.
the odds are, unless you're a dedicated And to to p it off, the price of the device
VHF and DSB (yes, we said DSB) nut like I is surprisingly low, considering what you
am, you still haven't heard very much get. There are several different versions
abo ut it. with different features , but the least costly
Not that some of u s haven't done our of them is still a complete FM receiver
part. The initial publication of an article of which accepts a minimum 300 V signal
the subject back in 1957 (three years (across 3 ki1) at any frequency from I 00
before 73 's birth) was accomplished under Hz up to 60 MHz or so, with any deviation
Wayne Green's guiding hand in CQ Maga- up to +20% of center frequency , and
zine, and the next synchronous-detection produces 60 m V o f audio output across 8
bombshell (conservatively titled 50 dB ki1. This one costs $30 in single-lot quan-
Under the Noise - A Breakthrough) saw tities. according to the most recent price
the light of day in the short-lived pages of list, but that's all you need for a comple te
6-Up, 73's subsidiary VHF magazine of the receiver within its frequency range and
mid-60s. sensitivity.
But all the way through, the synchro- On top of this, remember that these arc
nous detection ideli has had a tough c urrent prices, and the unit is not yet in
obstacle to battle: While it does give a ll the wide use. If it ca tches on as it apparently
performance claimed for it, it requires should , the cost is bo und to come down.
several times as many components as does Remember when the CK-722 transistor
a conventional detection c ircuit. The 1957 sold for $7 .SO each? And now they' re
version used e ight double-sec tion tubes to down to 4 fo r $ 1 from the mail-order
give 16 stages; the 1964 edition (actually a houses?
completely separate implemen tation of the What Is Synchronous De tection?
same basic idea but giving o ther features) The words "synchron ous detection"
was solid-state and required 24 transistors. have been applied to many different detec-
Quite obviously, this is a larger stage tion schemes. but all of them share the idea
count than that of many complete receiver of using a stable loca l osc illator to mix
designs. So Jong as so many devices were with the incoming signal , and d eveloping
necessary, the synchronou s-detection idea an error signal should the local oscillator
just couldn't make it in the face of its get off frequency . You might call it a sort
much simpler competition. of superhe t receiver. with automatic fre-
But all that was c hanged in the last year que ncy contro l and a zero-frequency i-f, if
when a major manufacturer of integrated tha t brings any reasonable pic ture to mind.
circuits put together a single-chip circuit The most co mmon type of sy nch ronous
which does just about everything required reception c urren tly used (yes, it is bei ng
for the synchronous detector, in a single used - by radio astrono me rs, space com-
sta nda rd-sized 16-Jead dualinline IC pack- municatio ns systems . lo ng-range radar , and
age. Although it contains approximately 50 any place e lse that the ultimate in radio
transistors, the whole thing takes only communications is required, thus justifying
180
the comp lexit y and cost of the older audio and retain only the de componen t,
approaches) invo lves a "phase-locked which is a measure o f the drift in the loca l
loop. " oscillator, and then applied as an "error
The phase-locked loop includes a phase signal" to the local oscillato r to keep it
detector, a low-pass filt er, and a voltage- locked in phase with the incoming FM
con trolled oscillator. The phase detector is sign~l.
a circuit which accepts two different rf The filter keeps the voltage-cont ro lled
input signals, and produces a de-to-audio oscillator (VCO) from (ollowing the audio,
output signal which reflects the phase so we pick the audio off ahead of the filter,
differences between the two inputs. Thar and we have an FM detector (Fig. I ).
For AM reception , we do things a little
FM
IN differently. We lock t he VCO to the
incoming sign al, just as before, but we add
things.
LOW- PASS It works out to be something very much
FILTER
like SSB reception ; many y ears ago QST
pushed something they called "exalted
Fig. 1. Basic phase-locked loop or synchronous carrier receptio n ," which involved using
detection hookup for FM reception uses volrage
two i-f strips, one extremely sharp to p ick
controlled oscillator in a servo loop. Output of
VCO is cont i nually compared with input signal in out the carrier from between the sidebands
phase detector, which produces an o utput con for high -gain amplification, and the oth er
sisting of de voltage w hich varies as input to accept the sidebands.
frequency changes (which amounts to superim- Our phase-locked AM reception is j ust
posed audio). Low-pass filter wipes off audio and
de goes to VCO to keep it locked to input signal.
about the same, except that we generate a
Unfiltered error voltage from phase detecror is new carrier in the VCO which is phase-
the audio modulation of the signal. locked to the incoming carrier, and use
that to demodulate the sidebands. The
demodulation is accomplished in a " multi-
is, so long as both inputs arc in the sa me
plier circuit" which is more fa miliar to most
phase relation to each o ther ( usually 90
of us as a " produ ct detector" or "mixer."
degrees), o utput of the phase detector is
Another low-pass fil ter shaves off the
zero. If the phase of inpu t A begins to lag,
original input-frequency sign al, the VCO
output goes positive , and if inpu t A begins
signal , and th e sum frequency fro m th e
to lead the o ther, outpu t goes negative.
product detector 's output, leaving us the
If we cou ld kee p an oscillator abso lutely
difference frequency, which turns out to
stable on the cent er freq uency of an FM
be the audio we wanted to recover.
transmission , we could fee d its o utp ut to
Note that this arrangement (Fig. 2)
one input of the phase detector, and feed
cannot demo dula te an SSB sign al sin ce
the other input with the FM signal. The
there is no carrier presen t fo r the phase-
output would then faithfully reproduce the
lock loop to lo ck o nto. In such a case, it
phase differences, which would reproduce
would attempt to lock o n the strongest
the audio signal envelope.
sideband component present with notably
If we cou ld keep an oscillator absolu tely
less than sa tisfactory results.
stable on t he center frequency of an FM
transm ission, we could feed its output to T he 195"/ circuit , by Dr. John Costas
one input o f the phase detector, and feed W2C RR and G. K. Webb W~AHM /2 (both
the o ther input wi th the FM signal. The with General Electric in Syracuse, N. Y. at
output would then faith fully re prod uce the that time), overcame t he problem by
phase diffe rences, which would reproduce deriving the contro l voltage for phase-
the aud io signa l envelope. locking from the sidebands rather than
And that 's the way the phase-locked from the carrier. Unfo rtuna tely, their
loop (abbreviated PL L) works. The phase- circuit has no t yet been implemen ted on an
detector outpu t is fi ltered to remove any integrated-circuit chip so fa r as we have
181
I#
1 - FM i-f strip a nd demodulator fo r
OUT commercial FM receivers.
2 - Com mercial TV sound i-f and
demodulator.
3 - Tuned AM detector.
4 - Self-contained SCA (storecast
music) receiver.
5 - FM/ multiplex telemetry receiver.
6 - Signal conditioner and limiter .
Fig. 2. AM d etector using phase-locked loop is a
7 - Frequency-shift telegraph receiver.
bit more complex. Incoming AM is shifted 90 8 - Frequency selective multiplier and
degrees in phase and VCO is then l ocked to it. divider.
VCO output is used as bfo input to product
detector, to r ecover audio. Low-pass filter in 9 - High-linearity FM detector fo r
audio signal path removes intermediate, VCO, wide-deviation FM .
and sum frequencies, l eaving only the difference- T o his list , we can add several more
frequency outp ut of the product detector. directed specifi ca lly at amateur use:
I 0 - VHF FM mobile receiver fo r com-
been able to learn and so we must wait a pact cars.
bit longer for that happy time. 11 - VHF FM handheld receiver.
Right now, though, we can build an FM 12 - Frequency synthesis for high-
receiver with excellent performance which accuracy VHF FM transceivers (Boelke, 73,
will also be able to de tect conven tional Feb. 1970) .
AM, for a small fract ion of the cost of
conventional receiver circuits.
Some of Dr. Grebene's suggested appli-
How Ca n We Use Phase-Locked Loops? cations are outside our scope, such as the
No new component, no matter how te le mentry receiver, signal conditioner, and
interesting it may be, is of mu ch good to TV sound demod ulator. Several of the
many of us unless we can put it to use. The others te lescope into a single application
logical question at this point , the n , is wl)en signal frequencies are ignored - th at
" How can we use phase-locked loops?" of a versatile FM receiver. .So Jet's see how
In the initial report describing the IC to use the PLL ( phase-locked loop) IC as
phase-locked loop, _Dr. Alan B. Grebene an FM receiver, as an AM receiver, as a
listed nine electronic circuit functions for frequency multiplier, a nd as a FSK RTTY
which he fe lt it was well suited: converter.
16 15 14 13 12 II 10 9
NE5618
Fig. 3. Practical versio n s of PLL using
integrated circuit with external resistor s
1 2 3 4 5 6 and capacitors to select frequen cy of
NC NC ~c
operation. FM input may be applied to
R3 either pin 12 or 13 (or push-pull to
Cl
FINE TUNE (S u Teat ) both ).
s..
Tf
182
Let's emphasize that the circuits which in the left-hand graph of Fig. 4. This
follow are taken from the manufacturer's capacitor provides a "coarse" adjustment
published applications notes and literature ; of fre quency which ca n be trimmed by
we have not breadboarded any o f them and variation of capacitor value. For "fine"
so ca nnot guaran tee results (but let us frequency adjustment, current mu st be
know if you have troubles with any of injected into pin 6 through a series resistor
them). fro m the power supply ; the right-hand
FM Receiver graph of Fig. 4 shows the percentage of
FM reception is simple with the IC PLL ; frequ en cy cha nge achieved for various
that's the job it appears to have been values of current injectio n .
invented to do in the first place. When it's The other capacitors shown in Fig. 3 are
running " locked" to a signal, the average bypass capacitors (C2) , coupling capacitors
de level of the phase-detector output is (C3), and low-pass filter capacitors (C4) .
directly proportional to the freq uency of Their values are de pendent to some degree
the input signal. As the input fre q uency upon the center frequency chosen. T he
shifts with modulation , it's t his de output deem phasis capacitor (CS ) should be larger
that changes and causes the VCO to shift than 200 pF for commercial FM reception;
its frequency to remai n locked on the its value is subject to experiment for
input signal. communications purposes.
The only problem to face in building an
Resistors R I together with capacito rs
FM receiver with the IC PL L is that of
C4 form the low-pass filter ; it's fed by an
setting it up for the proper in pu t signal
frequency. The VCO center frequency is internal impedance of 6 H2 at pin 14 or
set by an external timing capacitor, and by 15. Typical values for R 1 and C4 are son
varying the value o f this capacitor we can and 1 JOO pF.
work at any frequ ency from I 00 Hz up to The locking threshold of the circuit may
a guaranteed 15 MHz, with typical units be controlled by connecting resistor R2
operating to 30 MHz and operation as high across pins I 4 and 15. Resistor R2 is
as 60 MHz possible by a trick we'll pass on normally left out, but at high input signal
a little later. levels or high input signal frequency this
The connections necessary , as well as reduction in threshold may be necessary to
the external components required, are prevent instability . Approximately 6 kn is
shown in Fig. 3. Capacitor Cl is for timing ; a typical starting point ; the value of R2
its value in picofarads fo r frequ encies in should be as high as possible in any specific
the ra nge fro m I 00 Hz to 30 MHz is shown case, though .
eo
"
""" ..
""
10000
./
'\.
./v
u 1000
"r-... /
100
v
IO "" \
v
I 0
I02 t03 "'4 t05 I06 I07 108 0 4 8 10 12
Fig. 4. These two graphs show how operating frequency o f VCO in IC PLL is set. Graph a l left shows
approximate center frequency as value of Cl is changed th ro ugh the range from J F to 10 pF. This is
coarse frequency selling, determining lowest opera ting frequency. Graph at r ight shows how
frequency increases as current is f ed into pin 6; approxima tely 45% increase in frequ ency can be
atrained by this means. This provides "fine tuning" control.
183
Because of the comparatively high signal
level required , and the low input frequency
PLL 8 (when compared to the 2 meter band , for
2 3 instance) a VHF FM receiver for ham use
Cl of the PLL would require a converter
ahead of the PLL circuit . A block diagram
of su ch a hookup appears as Fig. 6. Note
IOK IOK that the converter could be crysta l-
controlled to produce a 28 MHz output ; by
adjustment of current into pin 6 the PLL
can easily be tuned over the full band from
a 30 MHz center frequency. This provides,
for less than $50, an FM receiver based on
the time-ho nored 75A4 principle of
'------ - -< 1-6V crystal-controlled fron t end and tu nable i-f.
Fig. 5. The NE560/NE561 IC PLL is ra ted for
operation only up to 15 MHz, although m any AM Receiver
units operate satisfactorily up to 30 MHz. Oper-
ating frequency for all units can be increased to The AM reception capability of the PLL
approximately 60 MHz by connecting pins 1, 2, is something o f an "extra." The PLL is
3, and 8 as shown here. This modification can be locked to input signal ca rrier frequency
applied to any of the accompanying circuits. The
5 kU pot serves as a fine- tuning contro l, replacing
and its outp ut is used as the local oscillator
the current injection into pin 6. If current. lo a built-in product detector. A 90-<legree
injectipn is to be used, the pot can be omitted phase shift is required to obtain proper
from this circuit. operation.
F igure 7 shows the hookup. Bypass and
coupling capacitors, all shown as 0.1 F ,
Tracking range may be contro Ued by should have low impedance at operating
current injection into pin 7 through resis- frequency (this circuit is intended to cover
to r R3. If R3 is omitted, tracking range the BC band rather than HF or VHF).
will be approx imately I 5% of center fre - Capacitor C 1 is selected to obtain the
quency. When R3 is set to a value which proper center frequency, and C2 sets the
permits 0.65 mA of current to now into bandwidth by filtering off high-frequ ency
the 600U impedance al pin 7, this figure is audio output.
reduced lo approximately 3%. The phase-shift network ( R l - 2, C3 - 4)
Other controls are possible, but should provides 90 degrees of phase shift for the rf
not be necessary when the PLL is used as
an FM receiver. Input signal level should be
at least 120 V to either pin 12 or 13 (the
unused input pin should be ac-grounded
through a bypass capacitor). Output at pin
9 across a 15 kU external resistor (which XTAL
PLL
must be in t he circuit; this is an open CONVERTER
emitter in the IC) should average 6 0 mV.
For VHF operatio n of the PLL, two IO
kU resistors and a 5 kU pot shou ld be
added externally as shown in Fig. 5 .
According to applications engineer Ralph Fig. 6. For ham use as a VHF FM receiver, a
Seymour, this modification extends the crystalcon trolled converter ahead of the PLL is
necessary. This converter need only provide a
frequency range up to 60 MHz. The 5 kU moderate output signal level, however, and can
pot provides convenient fine-tuning of be broadband with all receiver tuning being
center frequen cy, and may be omitted. accomplished at the PLL circuit.
184
input signal; th e sum. of RI a nd R2 should from pin 9 to ground . Opera tion will be
be less than 5 kn (2 kn each was the value improved by using a broadband untuned rf
used in the prototype) and CJ and C4 amplifier, but care is necessary to assure
should have reactance equal to the values that the PLL is not overdriven (maximum
of RI a nd R2. For BC operation , 82 pF input signal if IV rms). Maximum audio
was chosen. out put is 2 V p-p, and typical output is
In this hookup , the low-pass filter is not about 60 m V.
critical, since no audio is taken from the
loop itself; so a simple 0.0 I F capacitor Frequency Multiplication
from pin 14 to pin 15 suffi ces. The PLL IC can be used as a frequency
Tuning may be done in either of two multiplier in several ways. The simp lest is
ways. The first is more straigh tforward but me rely to set the center freq ue ncy to some
the second is elegant. T he fi rst way is to multiple of the actual input signal fre-
vary the value of timing capacitor C l. F or quency. However, as the higher (and thu s
BC operatio n , CI should tune fro m a weaker) harmonics are used for locking,
minimum of 220 pf (for 1600 k Hz) to a the lock range decreases. If input fre-
maxim um of 620 pF (for 50 0 kHz). The quency is fa irly stable and rapid tracking is
classic receiver align ment tech niq ue of no t required , this technique can be used to
padding capacitance is used at the high end multiply by 2, 3, 4, o r 5 times any input
of the range, but at the low e nd vary ing th e signal. Output of the VCO at pin 5 is a
current into pin 6 (fine-tuning control) square wave.
t akes the place of adjustment o f Action of the PLL as a multiplier can be
induc tance. improved by converting the input signal to
The second method of tuning the a square wave, which has much stronger
receiver uses a fix ed value fo r C I . This harmonics than does a sine. When this is
value is whatever is required Lo make the done, any output frequency up to 15 MHz
VCO operate at 940 k Hz (geome t ric mean can b e produced from its tenth subhar-
freque ncy) when the curre nt at pin 6 is monic (multiplication o f up to I 0 times).
zero. Pin 6 is then connected through a 1.2 Since th e PLL output is already a
kn resistor to the arm of a 5 kn pot (F ig. square wave, rich in ha rmonics, this offers
8) across the I 8V power supply. Varying an opportunity for use as a frequency
the pot tunes the rece iver across the BC ma rke r. Two PLLs in series cou ld produce a
band using o nly the fine-tuning feature. I MHz standard fro m a I 0 0 kHz crystal,
This receiver requires an a ntenna and a a nd in turn a I 0 MHz stan da rd fro m the I
good ground ; it must get at least JOO V MHz standard . One great advantage of the
I
Cl(S)F) "= ~(~Hr l
~~o-4 Cl
our
"' 8
NC NC
6
NC
5 4 3
,. l"
NE5168
(TOP VIEW I
.,
I" 9
NC
10
NC
II
NC
12 13 14 15 16
18V
Fig. 7. Th is is a BC-band AM receiver u sing the NE561 IC PLL (<he NE560 does not include AM
derec tion capability).
185
Extensions of these features make pos-
sible the construction of simple frequency
synthesizers; too many PLLs are required
to make them practical at present prices,
but when costs come down they may well
be worth investigating. Collins Radio's
book "Fundamentals of SSB" (out of print
Fig. 8. This simple arran gemen t for connection to
since 1962, unfortunately) gives the prin-
pin 6 of NES60/NES6 1 provides fine tuning over ciples involved if you're interested.
a 3:1 range. The 1. 2 kD.resistor limits maximum The hookup for use of the PLL as a
injection current. Timing capacitor should be set freq uency multiplier or divider is shown in
for lowest frequency with pot arm at ground, and
1.2 krl resistor then trimmed to set highest
Fig. 9.
frequency desired, with pot arm at hot end.
FSK Converter for RTTY
PLL multiplier as compared to a multi- Using the PLL as an RTTY FSK con-
vibrator or a conventional tuned-amplifier verter is almost identical to its use as an
multiplier is that the order of multipli- FM receiver, because FSK (or AFSK) is
cation can be changed merely by changing merely a means of carrying binary or
center frequency (such as with switched telegraph (mark-space) information by
timing capacitors). This would permit means of frequency modulation . The
generation of a 3.5 MHz square wave as the frequency shift involved is usually ra ther
seventh multiple of a 500 kHz standard , small - but so is that for FM, in com-
giving you strong markers at the bottom parison with the center frequen cy .
edge of every HF ham band. The FSK input, either at communications-
The PLL multiplier will divide fre- receiver i-f for FSK or at aud io frequency
quency just as easily as it multiplies. If for AFSK, is applied to the FM input of
center frequency is adjusted to be one- the PLL. The loop filter capacitor is made
third that of the input signal, the circuit smaller than usual to eliminate any possi-
still locks. This action occurs only for odd bility of "overshoot" in the output pulse,
submultiples, however ; for even divisions and a three-stage ladder filter removes the
(half, quarter, eighth) it doesn't work. You carrier component from the output (this is
can divide 3, 5 , 7 or 9 only. necessary only for AFSK). The center
RANGE FINE
CONTROL TUNE
Cl
lOK 101( U(
NC NC
e 4 1
NE5618
9 10 II 12 13
I NC '-----+--+-+-+--+--o~~~L~~:~ut
--jl---------' tll!V
PUSH-PULL
INPUTS .I
~1-~~~~~~~~
,+,
Fig. 9. PLL makes fine frequency multiplier or divider. For this application, audio outpu t connections
are ignored and the VCO output is used instead. If input is single-ended, one of the two push-pull
input leads should be bypassed to ground as shown by dotted iines. Circuit will multiply up to 10
times, and divide input frequency by 3, S, 7, or 9. Cl and fine-tuning adjustment must be set for
oper11tion near desired output frequency. When input is applied, VCO will lock to exact multiple or
odd submultiple of input if it is within lockin g range and o f adequate strength.
186
Cl
FINE
RANGE TUNE NC NC NC
8 7 6 5 4 3 2 1
NE560/NE561
+ 18 V
9 II 12 13 14 15
. ~L t 7*I
.I NC 1100 1100
'~o--ji---+---+-----...
IOK IOK
600
IOK
I ~
KEYING
PULSES
OUT
JUL
FSK AFSK
FREO . , , KHz 25:!0 Hz
Cl no pF OJ3 F
Fig . 10. RT T Y convener circu it is taken from compu ter data set applications note; da ta set is same as
AFSK convener, but gets i nput sign al from telephone line and so is not subject to su ch high levels of
int erference as is RTT Y. I nput may be either at i-f or a udio frequencies; table shows values of Cl for
bot h cases. Output consists of pulses which may drive a keying circuit for selector magne ts.
frequency is adjusted to produce abo ut different type of PLL chip together with
12V at the output when the input fre- an IC voltage1comparator at the output to
quency is at its lower figure. When input change the o utput levels to values which
frequ ency rises to its other figure, output are compatible with normal digital IC
will rise a maximum of 4V as the VCO chips. In this case, center frequency is
tracks the input-frequency change. This adjusted to produce a slightly positive
voltage change is coupled out through voltage at the output, when input is at the
output coupling capacitor to drive any low frequency.
external circuits desired, such as a magnet SCA Adapter
driver. The PLL's capability to receive FM at
The circuit is shown in Fig. 10. Figure almost any center fr equency comes in
11 shows an alternate circuit using a handy if you like background music from
+5V
5K .15
IOK
FSKo - - j 2 OUT
IN
6
3
NE5e5
..
5
600
.0 5
- 5V
Fi g. 1 1. Allernate RTTY circuit uses N E5 65 IC, which is so new that i ts pr ice is not liste d ye t.
Maximum frequency of 565 is 500 kHz. This circu i t is designed to drive d igital I C devices, and type
5710 volrage comparator is included to adjusr o utput level to values suirable for digi tal !Cs. Po r is for
frequency adjustment.
187
~-------+-----+----+----+---O+I0-24V
1800
~K .018 .OIS
,001
~10 IK IK IK SCA
8
~~o---41-<>-ll-+-.----l2 ._..-~--~---1.IVV----OAUOIO
OUT
4700 NE!56!5
4700
--W.-i3
4700
4700
Fig. 12. Background music adapter for FM receivers uses type 565 IC PLL. 510 pF capacitors and 4. 7
kU resistors at input form high-pass filter to keep audio from FM set from overloading PLL; ladder
filter at output removes everything above about 10 kHz to keep from overloading audio amplifier s
following. Pot allows frequency to be adjusted to 6 7 kHz to pick off SCA subcarrier.
the " storecast" services sold by many required 21 pages merely to list prices of
commercial FM stations. These storecast current JC products).
signals are transmitted as FM of a 67 kHz Signetics is IQcated at 811 E. Arques
subcarrier which is itself a part of the Avenue, Sunnyvale CA 94086, and we
normal FM signal. To receive them with obtained our information from Ralph
the PLL, all you need do is set up the PLL Semour, linear applications supervisor.
for a center frequency of 67 kHz and tap They make the PLL in several d ifferent
off an output from your FM receiver ahead models, wit h model numbers to match.
of the deemphasis filter (which bypasses The one shown in most of these diagrams is
the 67 kHz signal to ground) to feed the the NE561 B, which is the most general in
PLL. PLL output will then be the back- application. It includes the AM-deteclion
ground music, which is free of all inter- capability, and quoted price in lots of 1-24
ruptions such as commercials or even is $37.50 each. The NE560B is virtually
station identification. identical but does not include the AM
One precaution may be necessary. The detection features, and lists at $30 each.
signal from the FM receiver to the PLL Models NE565 and SE565 (the N indicates
input should go through a high-pass fil ter commercial temperature range ; the S
to prevent any possibility of overload by means military range) are too new to be
the much stronger audio of the normal listed; they operate on ly up to 500 kHz
broadcast program or any accompanying and so presumably wou ld be lower in cost.
stereo information at 38 k Hz. A typical The 565 is used in the circuits of Figs. 11
hookup using the NE565 PLL is shown in and 12. Finally, the NE562 is inteRded for
Fig. 12. direct interface with digital logic systems
and, like the 565 it is too new to have an
The Source established price yet.
About the only thing we haven't yet Pin connectio ns for the 560, 56 I , and
told you about the PLL IC chip is who 565 are shown in Fig. 13.
makes it and where they can be obtained . The N57 l OT comparator shown in Fig.
The manufacturer is Signetics Corpor- 11 .se lls for $2.62 each in lo ts of 1- 24.
ation, a subsidiary of Coming Glass Works All of these (with the possible exception
and one of the leaders in the integrated- of the 562 and 565 , which were still
circuit industry for a number of years now. preliminary d esigns al this writing) may be
The PLL IC is only one of many chips in obtained on special order from the manu-
their line (their price list as of 4 May 197 0 facturer. In addition, the major mail-order
188
..., -.:AH(;("
oc
TUN[
vco
-!U.!-
vco
--?~ NC
,.
8 6
560
C'fOP VJEW)
3 2
..
...,,
9 10 II 12 13 14
..
16
' M ......
D<-
""'
M
Of"FS(T
-...
--;;-
"'"
f fl.TUt
vco
,,..,,
8
RANG[
oc
TIN
6
vco
OUT
"' ~"!.-
,.
AM
T
F M INPUTS
OAN'[ OHLY OH[ SIO( AHO
J.C GltCJll..I) TH[ OTHER ,.Oft
S#il.C- CNOE:O OPUlAfl()N ,
561 Fig. 13. Pin connections for types 560, 561, and
ITOP VIEW)
565 Plls are show n here. Only difference between
IO II 12 13 14 15
..
16
,..,,
560 and 561 is that 561 has Am detector
included where 560 has additional VCO output
lead.
189
Index
.
HOBBY & EXPERIMENT
TAB BAS IC T ECHNOLOGY
terest
542- TRANSISTOR PROJECTS FOR HOBBY 538 - COM PUTER CI RCU I T S & HOW T HE Y
ISTS & STU DENTS. Steckl er. 192 pages, 140 WOR K . Byron Wei s. 192 pages. 134 i llus. S4.95
illus. $4.95 530 - BASIC ELECTRONICS PROBLEMS
537 - 125 ONE-TRANSISTOR PROJECTS. SOLVED. D. A . Sm i th . 192 pages, 100 ill us. S4 .95
Rufus Turner. 192 pages, 125 il l us. SJ.95 528 - PULSE & SWITCHING CIRCU I TS.
524 - 104 EASY PROJECTS FOR THE Harvey Swear er . 256 p ages, 200 i l l us. S4.95
ELECTRONICS GADGETEER . Robert M . 510 - HOW TO READ E L ECTRONI C CI R CU IT
Brown. 160 pages, 104 illus. $3.95 DI AGRAMS. Brown & Lawrence. 192 pages. plus
487 - 64 HOBBY PROJECTS FOR HOME & 64-page schemat ic foldout sect ion . 140 i l lus. SJ.95
CAR. Brown & Olsen. 192 pages, 111 illus. SJ.95 112 - LEARN EL ECTRONICS B Y BUILD I NG.
486 - 104 SIMPLE ONE-TUBE PROJECTS. John Schroeder . 208 pages, 209 i l l us. S4.95
Robert M . Brown. 192 pages, 104 illu s. SJ.95 111 - BAS I C TRANS ISTOR COU RSE . Paul R.
464- ELECTRONIC HOBBYIST'S IC PROJECT K en ian. 224 pages. 176 il l us. S4 .95
HANDBOOK. Robert M . Brown. 160 pages $3.95 105 - BASIC TV COURSE . George Krav itz. 224
462-104 EASY TRANSISTOR PROJECTS YOU pages, 137 i l l us . S5.95
CAN BU I LO. Robert M . Brown. 224 pages, 104 104 - BASIC RADIO COURSE . John T. Fry e. 224
ill us. $3.95 p ages, 131 i l l us. S4.95
135 - RADIO CONTR OL MANUAL. Edward L. G l OO - BAS IC MATH COURSE F OR ELEC-
Safford. 192 pages, 147 illus. $3.95 T RONICS. Jacobwitz. 160 pages, 89 i ll us. S4 .95
129 - SKILL-BUILDING TRANSISTOR 99 - INDUSTR IAL E L ECTRON ICS MADE
PROJECTS & EXPERIME N TS. Lou Garner. 192 EASY . Tom Jaski. 288 pages. 239 i llus. S5.95
pages, 128 illus. SJ.95
122 - ADVANCED RADIO CONTROL. Edward SOLID-STATE TECHNOLOGY
L. Saf ford, Jr. 192 pages, 174 ill u s. S4.95 513 - UNDERST.ANDING SOLID-STAT E
93- RADIO CONTROL HANDBOOK : New 3rd Cl RCU ITS. Crowhurst. 192 pages, 150 ill us. S4.95
Edi tion . Howard McEntee. 320 pages, 240 i ll. 55.95 501 - WORKING WITH SEMICONDUCTORS.
89 - TRANSISTOR PROJECTS. Radio Elec- A. C. Saunders. 224 pages, 185 illus. S4.95
troni cs Staff. 160 pages, 123 ill us. 52.95 493 - SEMICONDUCTORS FROM A TO Z.
83 - FUN WITH ELECTRICITY. Tom Kennedy, Phi lli p Dahlen. 288 pages, 300 illus. S4.95
Jr. 128 pages, 95 illu s. $2.95 470 - TRANSISTOR CI RCUIT GUI DEBOOK.
74 - MODEL RADIO CONTROL. Edwar d L. Byron Weis. 224 pages, 104 illus. S4.95
Safford, Jr. 192 pages, 21 0 illus. $3.95 116 - GET TI N G STARTED WITH TRAN-
70 - ELECTRONIC PUZZLES & GAMES. SI STORS. Lou Garner. 160 pages, 89 il l us. SJ.95
Matthew Mandi. 128 pages, 72 illus. 52.50 94- TR A NS ISTO R S. Radio Electr onics Slaff. 96
69- ELECTRONIC HOBBYIST' S HANDBOOK. pages, 65 i llus. $2.50
Rufus Turn er. 160 pages, 118 i llus. $3.95 75- TRANSISTO RS- THEORY & PRACTICE.
Rufus Turner . 160 pages, 143 illus. SJ.95
AMATEUR RADIO 63 - TRANSISTOR CIRCUITS. Rufu s Turner.
I
160 pages, 146 illus. SJ.95
543 - AMATEUR RADIO EXTRA-CLASS
LICENSE STUDY GUIDE . 73 Magazine. 224 61 - TRANSISTOR TECHN IQUES. Radio
E lectronics Staff. 96 pages, 78 il lus. $2 .50
pages, 162 il lus. $4.95
527 - AMATEUR RADIO ADVANCED CLASS AUDIO & HI- Fl STEREO
LICENSE STUDY GUIDE . 73 Magazine. 192
pages, 73 illus. SJ.95 546 - ELECTRONIC MUSICAL INSTRU-
499- CB RADIO OPERATOR'S GU I DE. Brown MENTS. Crowhur st. 192 pages, 125 i llus. 54.95
& Lawrence. 224 pages, 138 i llus. 53.95 529 - HANDBOOK OF MAGNETI C RECOR-
469 - HAM RADIO INCENTIVE LICENSING DING. F inn Jorgensen. 224 pages, 90 illus. 54.95
GUIDE . Berl Simon. 160 pages, 314 Q & A, 35 534 - SERVIC I NG MODERN HI- Fl STEREO
ill us. SJ.95 SYSTEMS. Nor man Crowhursl . 224 pages, plus
468 - 104 HAM RADIO PROJECTS FOR schematic fol dout sect ion. 125 il lus. S4.95
NOVI CE & TECHNICIAN . Ber t Simon. 192 505 - INSTALLING & SERVICING HOME
pages, 104 ill us. SJ.95 AUDIO SYSTEMS. Jack Hobbs. 256 pages, 150
460 - VHF HAM RADIO HANDBOOK. E dward i l lus. $4.95
G. MacKinnon . 176 p ages, 100 illus. SJ.95 497 - TAPE RECORDI NG FOR FUN &
PROF IT. Wa lter Sa l m . 224 pages, 200 i ll us. S4.95
RAD IO SERVICING 494 - AUD I O SYSTEMS HANDBOOK. Norman
504 - HOW TO FIX TRANSISTOR RADIOS & Crowhurst. 192 pages, 125 illus. 54.95
PRINTED CIRCUITS. L eonard C. Lane. 256 120 - Hl-F I TROUBLES. Herman Burstein. 160
pages, 150 Illus. S4.95 pages, 130 i llus. SJ.95
429 - EASY WAY TO SERVICE RADIO 86 - I NSTALLING HI- Fl SYSTEMS. Mar kel l &
RECEIVERS.Sands. 176pages, 100 Illus. SJ.95 Stanton. 224 pages, 152 illu s. $5.95
78 - RAPID RADIO REPAIR . G. Warr en Heath. 79 - DESIGNING AND BUILDING HI-Fl
224 pages, 104 i llus. SJ.95 F URN I TURE. Markel l. 224 pages, 195 il lus. $4.95
76 - SERVICING TRANSISTOR RADIOS. 67 - ELEMENTS OF T APE RECORDER
Leonard D' Airo. 224 pages, 202 il l us. S4.95 CIR CU I T S. Burstein & Pollak. 224 pages, $4.95