Skip to main content

Full text of "RTTY V05N06"

See other formats

O W 

Ui ►< 


95 - 

r— < 

tz> 53 h 
E H , h 

f 3* 


S I c 

S3 f C 
w 5 

* to C 


- ' CL m 

© ^ o 

~ 2. 

» 5 .-. 





2 * 



K'j» 3 H 
« > ? — 

2 3 

0 s5 2 

» W H 

2> « 
cr E2 5? 
5.3? 2 

o o 
1- D. 
' * 




S I 










s *3 

- r 

60 >J3 

The W9TCJ Automatic Start RTTY Printer 




JUNE, 1957 
25 Cents 
Vol. 5, No. 6 


R T T Y 

R T T Y 


"Automatic Start" — Remote Control of 
Unattended Radio-Teleprinters 




One useful feature of the art of print- 
ing telegraphy — control of unattended 
teleprinters — has thus far not been 
employed much by RTTY amateurs. De- 
signs have been made by various individ- 
uals — notably W2BFD, W6AEE and 
W6IZJ — towards so-called “autostart 
systems,” but for various reasons they 
have not been generally adopted by the 
amateur RTTY fraternity. Systems in 
the past have suffered from complexity, 
nonselectivity, time-restrictions, frequen- 
cy drift, liability to running “wild” in 
absence of a RTTY signal and the like. 
Also most people prefer to have their 
printers turned on a directed control 
basis than to have them operate on any 
and all RTTY signals. This obviously 
conserves paper and machinery, and 
hence is of more utility than the mere 
desire of some individuals to “read other 
people’s mail.” 


The usual landline control of unattend- 
ed printers takes place in three steps: 
First, an initial “line-break” signal is 
transmitted — a space signal is sent for 
an instant and results in startups of 
printers on the line. Next, the message 
is sent. When this is finished, a “stop” 
signal is sent as the last keyboard man- 
ipulation (upper case “H”), resulting in 
shutdown of printer motors. This sys- 
tem is usable on landline circuits be- 
cause they are "solidly” reliable with 
negligible chances of false-signals com- 
ing upon the circuit to start up printers. 

Attempts to apply this landline pro- 
cedure to radio circuits are fraught with 
many problems. First of all, radio com- 
munications are notoriously subject to 
the vagaries of propagation, fading, in- 
terference and noise. This is especially 
true in amateur radio, and hence other 
control methods are needed. Two sys- 
tems have apparently been proposed and 
employed by a few individuals, as men- 

tioned in the introduction. The W2BFD 
system uses a time-clock programmer 
which checks the radio channel at period- 
ic intervals — and if there is a mark- 
signal present, a subsequent circuit turns 
on the printer motor, readying it to take 
the message. These "periodic intervals” 
are of the nature of a one-minute monit- 
oring on every exact hour, and so it 
imposes a time-restriction upon the 
transmitting operator. Furthermore a 
certain mark-signal has to be transmit- 
ted once every minute if the operator de- 
sires to keep the unattended printer go- 
ing during a long one-way transmission. 
This other special signal was designed in 
order to enable the printer to shut down 
by itself after one minute, to assure 
shutdown in case of disappearance of the 
received signal. 

The other system as employed on the 
west coast accepts any RTTY signal at 
any time and hence may operate the 
machine more or less all day every day 
— receiving uninteresting and tiresome 
tidbits of information originated by 
“ragchewer” RTTY stations. In other 
words this is a nonselective autostart 
system. . 

What the RTTY fraternity needs is 
some form of SELECTIVE calling or 
starting of the unattended printers. In 
this way directed calls can be made to 
specific stations, with as simple auxiliary 
equipment as possible, yet with reason- 
able chance for successful unattended 
reception. With such a set-up, one then 
can be away from home or otherwise 
engaged; and his printer will be on 
continuous monitoring duty, ready to 
take any messages directed to it. Later 
the fellow can check the printer to see 
what has been received, and he could 
answer the message in return via auto- 
start to the initiating operator. In this 
way busy people could keep in touch 
with each other, making use of machine- 
telegraphy’s great asset — the ability to 
print and store information for later 
perusal — as well as the ability to “call” 
the called party by machine by use of 
“BELL” signal. It then is not necessary 
to continuously monitor a sqawky and 

noisy radio receiver for a friend’s signal; 
let the machine take care of it. It is 
well to keep our precious machines in as 
constant use as possible — on monitoring 
duty or otherwise — so our friends will 
be able to contact us even if we are not 
as active on the air as we should be. 

3 — A Selective Control System for 
RTTY Autostart Printer: 

A good RTTY autostart system 
ought to have the following features: 

1. Selective calling and starting of spec- 
ific unattended printers, using a simple 
transmitted signal. 

2. Must be reasonably proof against 
QRM and QRN yet to be able to print 
100 percent on any RTTY signal under 
reasonable conditions, on any specified 

3. Must be on a continuous 24-hour 
monitoring basis so messages can be 
received at any time, with a minimum 
of clock timing problems. 

4. Simple circuitry, with a minimum of 
power-consuming components. 

5. A printer shutdown circuit, that is 
controllable by the transmitting operator 
and more importantly, that will shut 
down the printer automatically in case 
of signal disappearance or similar 

The first consideration — selective cal- 
ling and starting — means that only di- 
rected messages will be accepted at a 
given unattended printer — when the 
transmitting operator knows the exact 
procedure necessary. Ordinary RTTY or 
CW traffic, as well as noise, that happen 
to be on the autostart channel then do 
not operate the system and the printer 
is quiet. 

The second desriatum implies need for 
proper design of the overall receiving 
equipment. This takes in such factors 
as frequency stability, automatic gain 
control, and fairly selective mark and 
space filters. The design ought to be 
applicable for low' frequency amateur 
RTTY circuits as well as on VHF. After 
the radio-frequency control problem is 
solved, the autostart system can be used 
on any specific frequency on 80, 40, 20, 
15 or 2 meters, for instance. With such 
a design freedom, the system would be 
of general utility — for FSK or AFSK 

The third item is obvious. Taking due 
consideration of signal propagation con- 

ditions on the autostart channel (and 
also checking beforehand to make sure 
the channel is clear) the transmitting 
operator then can start up, “leave a 
message,” and shut down the remote un- 
attended printer at any convenient time. 

The fourth one is necessary inasmuch 
as 24-hour monitoring is involved. The 
equipment, including receiver, terminal 
unit, etc ought to be designed for low 
pow'er consumption, with as few tubes 
and parts as possible. Ideally, during 
standby (but on monitoring basis) the 
total power consumption by the entire 
system should not exceed say 30 or 40 
watts, even when using present-day avail- 
able tube equipment. Transistors could 
and will no doubt be eventually em- 
ployed and thus result in greatly re- 
duced power consumption, coupled with 
greater versatility. 

The final consideration is inherently 
simple. It should be that any pro- 
longed space-signal condition — such as 
due to operator shutting dowm printer 
or to QRM — will be certain to shut down 
the system. Of course this means that 
any CW signal that happens to be on 
space-signal frequency is apt to shut 
down the printer; or, likewise “hold the 
printer running” if on mark-signal fre- 
quency. This is tied up with item 2, 
above and it will be the transmitting 
operator’s responsibility to evaluate the 
chances of successful autostart by listen- 
ing on the channel beforehand. If neces- 
sary, the whole procedure could be re- 
peated a little while later to make sure 
the message is “delivered.” Finally 
there is the problem of wild printing in 
case of cessation of received signal; or 
in other words, on noise. The entire 
system needs to be properly designed 
with this possibility in mind. 

la- — The W9TCJ Autostart: Starting 
Signal Considerations: 

Following the general concepts pre- 
sented above, an autostart system W'as 
designed and built early in 1956 at W9 
TCJ. First, we must consider the quest- 
ion of a suitable start-signal. This could 
be of one of a number of forms, such 
as a special teleprinter character, or a 
series of mark-space reversals at a cer- 
tain low frequency. 

The former scheme — i.e. a special tele- 
printer signal — would require a contin- 
uously running teleprinter receiving dis- 
tributor, which is undesirable in the light 
of its known mechanical characteristics. 
The other scheme was inspired by the 
use of “reed relays” in some remote 


R T T Y 

R T T Y 


control applications such as calling- of 
specific taxicabs by radio, without dis- 
turbing other cabs. Some such reed re- 
lays were obtained from Joe Juel, W9 
BGC, and work -was conducted with them, 
noting particularly the high “Q” of the 
resonant spring reed and need for “mag- 
netic bias.” As these reeds were tuned 
to specific odd-valued frequencies in the 
region 100 to 300 cps, efforts were ex- 
pended in devising “mixer” circuits, fed 
by local oscillators, so that the input 
frequency for “autostart buzzing” could 
be exactly a chosen frequency, say 60 
cps, the powerline frequency. However 
it was decided to try building a 60 cps 
reed relay; and this was done, with 
highly gratifying success, requiring only 
easily obtainable materials at W9TCJ. 

4b — Reed Relays — Construction and 

A reed-Relay consists of three main 
parts as shown in Figure 1; a clock- 
spring-type metallic reed, a magnetic ex- 
citing coil energized by an input signal, 
and a pair of contacts for operating sub- 
sequent circuits when the reed is caused 
to vibrate near maximum amplitude by 
the signal of proper frequency applied to 
the exciting coil. The reed may be com- 
pared to a high Q "L-C” circuit or a 
pendulum— the system responds sharply 
to a certain vibrational frequency. Reeds 
are employed in panel frequency meters 
for indicating, say, power line frequency, 
displaying one of a row of steel reeds 
vibrating at maximum amplitude, giving 
exact indicated frequency. A number of 
manufacturers make “reed-relays” but 
they seem to be only in the higher fre- 
quencies, above 100 cycles or so, and are 
relatively difficult to obtain. 

These reed-relays however can easily 
be constructed and tuned to any specific 
frequency using ordinary tools and mat- 
erials available to RTTY amateurs, and 
such was done at W9TCJ. I used a 
piece of clock spring and a 5000 ohm 
relay coil, both mounted on a heavy 
weighted metal base; the reed being 
solidly anchored at one end, and the coil 
mounted to that its pole-piece is close 
to the steel spring near its supported end. 
Application of an AC signal of proper 
frequency to the energizing coil then 
causes this reed to vibrate to maximum 
amplitude. Due to its high-Q (typically, 
several hundred) the vibrational frequen- 
cy ' s < l u *^ e critical, it has to be within 
0-2% °f the natural reed frequency, and 
this feature is all to the good, as it de- 
termines the autostart’s ability to reject 

undesired signals, QRM, RTTY, CW, etc. 
anything that is not supposed to turn on 
the printer. Furthermore, in order to 
enable the reed to follow the fundamen- 
tal frequency input of the signal, some 
amount of biasing DC is required in the 
energizing coil. More of this later. 

There is the problem of picking up a 
signal from the reed when it vibrates 
near maximum amplitude. The manu- 
factured reed-relays employ tiny con- 
tacts which are quite apt to get out of 
order and also impaires the mechanical 
“Q” of the reed to an extent. I could 
use a pair of contacts mounted on the 
reed-assembly like on a Vibroplex bug- 
key, but instead I hit upon the idea of 
using a photoelectric pickup device. As 
it turned out, it is a neat solution — all 
that is required is a radio dial lamp run 
on low voltage for a light source and a 
cadmium sulphide photo-cell. These two 
items then take place of the mechanical 
contacts and at the same time leaves 
the reed “Q” unaffected. The lamp and 
cell are placed at the free end of the 
reed so that when the latter vibrates to 
maximum amplitude, light is allowed up- 
on the cell and results in “contact.” This 
is shown in Figure 2, and an “exploded 
view” is shown in Figure 3. 

The frequency of the reed is adjusted 
by varying the length of free-vibrating 
portion of the spring. I used a piece 
of clock spring (such as from an old 
alarm clock of the Big Ben variety) 
about 4 inches long by 3/8 inch wide 
wide. The clamping device is designed 
so as to permit adjustment of the free- 
vibrating length. Applying 60 cps to 
the energizing coil (along with a little 
DC for magnetic bias) the length was 
adjusted until maximum amplitude was 
achieved. This is a very critical oper- 
ation much like aligning a high selectiv- 
!ty l-f amplifier, and in the final tuneup, 
the reed is exactly tuned by filing away 
at the free-end or adding solder upon it. 
The reed, thu^ tuned, has so much “Q” 
that if the frequency differs by 0.1 or 
0.2 cps from 60.0 cps there is a marked 
decrease in vibrational amplitude. And 
if the frequency is as much as half a 
cycle off, there is no response at all! 
This points up the importance of a 
stable 60.0 cps source; and so far as 
been determined the public utility source 
appears to hold very closely to 60.0 cps. 
(Recent monitoring measurements in- 
dicates deviations no greater than 0.01 
cps at W9TCJ’s QTH. It is believed that 
this kind of frequency stability can be 
expected of any large public utility 
power source at any place in the United 

Continuing with the discussion, one 
advantage of such a sharp reed is the 
possibility of setting up certain buzz- 
frequencies for specific stations to be 
called, using the same radio-frequency 
channel. Frequencies of 61.0, 62.0, 63.0, 
etc, could be chosen and transmitted, 
using simple reed-oscillators, to activitate 
certain printers only. The power line 
frequency, 60.0 cps, could perhaps be 
employed as an universal autostart sig- 
nal. Each autostart printer would have 
two reed-relays; one of the universal 
autostart frequency and the other for its 
assigned buzz-frequency. Utility of such 
features are obvious in communication 
work, wholesale bulletin reception as well 
as individual calling. 

This reed-relay is the heart of the 
whole W9TCJ autostart system and 
hence why we went into some detail on 
this device, pointing up its characteristics 
and advantages. Such relays, by the 
way, are coming into wide use for re- 
mote controls in communications and 

4c — Printer Motor Control and Message 

The photocell pickup on the reed-relay 
is fed into a single biased triode stage, 
fitted with a relay in its plate circuit. 
When light falls upon the cell, as dur- 
ing reed vibration, the triode conducts 
and closes the aforementioned relay; 
its contacts then cause “pick-up” of the 
main motor-start relay. An extra pair 
of contacts on the latter keeps it locked 
up so the motor runs. On the plate cir- 
cuit relay is another pair of contacts, 
this is employed to hold the printer 
magnet closed (on mark) during the 
starting up procedure so the machine 
does not run wild during the rest of the 
buzzing interval. 

Upon cessation of buzzing, the mes- 
sage may now be transmitted. The reed- 
relay decays down rapidly; and in a 
second or so, the triode stage cuts off, 
plate relay opens and releases the printer 
magnet to be keyed by the incoming 

4d — Printer Shutdown Control: 

As it is of course necessary to close 
down the printer after a message is 
sent, a "steady-space” signal is sent for 
a few seconds. This is accomplished by 
pressing the “Break” button on t*>e 
transmitting keyboard. Installed in the 
printer magnet-line is a RC charging 

network, with rectifier, and its output 
feeds into another relay that is kept 
closed if the incoming signal is on mark 
a great portion of the time — as it always 
is for a normal transmitting teleprinter 
signal. This RC network has its time 
constant so adjusted that if a space sig- 
nal is continued on the circuit longer 
than approximately one-fourth to one- 
half second, the network loses its charge 
and its relay opens. Its contacts then 
open the lockup circuit in the motor re- 
lay, allowing it to open, thus shutting 
down the printer motor. 

This method of shutdown is simple 
and straighforward. It is noted that 
this same “break” system would be for 
starting up a landline printer, but here 
it shuts down the RTTY printer. This 
way assures that the autostart system 
just described will close down in case 
of cessation of signal or QRM on space- 

5— Reasons for Choice of 60.0 cps for 
Initial Autostart Signal: 

The choice of 60.0 cps as a buzzing 
frequency was decided upon early in the 
design period due to several factors to- 
wit: (1) This is the powerline frequency, 
available everywhere with some high 
degree of precision, as was indicated. 
(2) The 60 cps frequency can be easily 
injected into a RTTY circuit. (3) 60 cps 
is low enough in frequency to be passed 
with little attenuation through RTTY 
converters when superimposed upon a 
radio carrier. In other words, 60 cps 
mark-space reversals reproduce them- 
selves in the output from the TU’s dis- 
criminator. Higher frequencies may not 
pass through well due to certain “low- 
pass filter action” in printer circuits, 
while lower frequencies are apt to ontro- 
duce long build-up time in reed-relay, 
possibility of false starts due to various 
causes (mechanical shock to reed, high- 
speed CW) and other factors. 

(4) Also another factor enters the 
picture. In order to avoid possible false 
starts due to RTTY signal or high speed 
CW signal on the channel, the circuit 
response should be sharp and at a quite 
high frequency. This implies need for 
a “high-Q-system” — the tuned reed, and 
appreciable buildup time, this then auto- 
matically resulting in discriminatioin 
against undesired signals on the channel. 

All in all, 60.0 cps is a logical fre- 
quency to begin with, a reed tuned to 
that value has a sufficiently short build- 
up time (1.5 to 2 seconds) in response, 



R T T Y 

R T T Y 


and the frequency is available everywhere 
with close tolerance to operate the high- 
Q reed. And such a reed-relay can be 
easily constructed at low cost. 

6 — The W0BP Resonant Reed 

Here we will digress for a moment and 
mention an interesting development that 
Boyd Phelps, W0BP, has done along 
this line of autostart-autocall. During 
our discussions about systems using the 
resonant reed principle of selective re- 
sponse, Beep proposed and constructed 
a 6.1 cps hacksaw-blade reed, and vi- 
brated (like a Vibroplex key) by a re- 
lay coil placed near the reed. He ad- 
justed the length of this reed, weighted 
on its free end, so it responded to a 
frequency of 6.1 cps, the frequency of 
the teleprinter signal transmitting letter 
“0” repeatedly (key held down and 
transmitting cam allowed to revolve con- 
tinuously). Subsequent circuits were in- 
stalled to ring a bell, light a lamp, and 
start a clock upon receipt of such a 6.1 
cps signal after a few seconds buildup 
to maximum amplitude. Beep says that 
it is for “autocall” purposes only for the 
present. However I feel, that, with some 
further development, this system could 
be used for printer autostart; however 
the low frequency of 6.1 cps coupled 
with the necessarily high-Q character- 
istic of the reed make for a relatively 
long buildup time for the latter to get 
to full amplitude for energizing other 
circuits. This means at least 10 - 15 
seconds of letter “0” repeats. This ob- 
jection, however, may be minimized 
somewhat by transmitting the letter “I” 
which gives a frequency of 12.2 cps, 
feeding into a 12.2 cps reed with shorter 
buildup time. And with the 60 cps reed- 
relay in the W9TCJ autostart, about 2 
seconds of buzzing results in motor start. 

7 — Terminal Unit Design: 

Having described the manner of op- 
eration of the W9TCJ autostart system, 
it remains to investigate the signal cir- 
cuits needed for the system to function 
properly. This part will discuss only 
general principles and characteristics; 
the actual circuit will be in the second 
part of the paper. 

Inasmuch as the signal circuits will 
handle either the usual teleprinter signal 
or the automatic start “buzzing” signal, 
it follows that the circuit could be of 
almost any standard terminal unit de- 
sign. The buzzing signal is of not much 


higher frequency than that of the third 
harmonic involved in the teleprinter 
square-waves. Hence the TU ought to 
have reasonably broad mark and space 
band-pass filters — not only for good 
square-wave response but also to accom- 
modate slight frequency drifts or misal- 
ignment on either or both transmitter or 
receiver. A good value for such band- 
passes is 250-300 cps centered about 
mark and space frequencies. Any value 
wider than that only invites trouble due 
to QRM from adjacent channels. 

8 — Radio Frequency Control. 

Now we are faced with a major prob- 
lem, “radio frequency control,” before we 
can achieve successful FSK autostarts 
on the low frequency amateur bands. 
While the mark pnd space bandwidths in 
the autostart TU are of the order of 300 
cps, relatively wide by audio standards, 
it is more of a problem to set a FSK 
carrier exactly on the autostart channel 
so that the mark and space frequencies 
will be properly hetrodyned and caused 
to fall squarely through these audio fil- 
ters. This definitely calls for stable 
receivers and transmitters, plus the op- 
erator s ability to set up desired auto- 
start frequencies to the degree of pre- 
cision required. For an estimate, one 
should be able to set up to within 50 
cycles of the specified frequency. This 
is very easy to accomplish with a little 
auxiliary equipment and a proper pro- 
cedure. In fact, with the aid of a 100 
kc crystal standard plus a 10 kc multi- 
vibrator, one can adjust his transmitter 
frequency to within several cycles of 
7140.000 kc, for example, taking only a 
few seconds to accomplish the line-up. 
The surplus LM frequency meters also 
would serve excellently whr-n carefully 

The best way of achieving radio fre- 
quency control — i.e. calibration and 
stability - is to employ crystal control 
throughout in both transmitter and re- 
ceiver. Standard crystal FSK circuits 
exist, and inexpensive crystals are ob- 
tainable that have excellent calibration 
holding properties. An autostart chan- 
nel can be set up on a specified frequency 
and thereafter the operator only need 
transmit a buzz-signal for a few seconds 
in order to start up a distant printer. 
All in all, such a set-up is ideal for in- 
tercity RTTY circuits. Several separate 
autostart channels can be set up on eac h 
band, in addition to the several discrete 
buzzing-frequencies and thus realizing 
the equivalent of “picking up a telephone 
dialing a number, and talking to your 





friend at the other end." It has been 
many times demonstrated that properly 
aligned crystal-controls have more than 
the required degree of accuracy for gen- 
eral auto-start work with a minimum 
of frequency control worries. 

In conclusion the frequency control 
problem just discussed stems from the 
need for "dropping" the audio mark 
and space frequencies right through the 
filter “hatches” of the receiving TU. We 
wish to keep these hatches quite narrow 
to obtain all possible freedom from inter- 
ference by adjacent channel signals. We 
can zero-in on spot frequencies easily 
yet exactly, using auxiliary equipment 
or crystal control. There is no valid 
reason why this oannot be done by those 
really desiring reliable unattended print- 
er operation. The discipline and tech- 
niques needed will be of considerable 
benefit to the RTTY amateurs, generally 
and is definitely good for the progress 
of the radio art. Witness the need for 
oareful tuning on SSB phone signals as 
well as on RTTY signals, especially on 

It is obvious that this system of auto- 
start can be used to advantage on VHF, 
using AFSK (tone modulated radio- 
telephony) emissions. Assuming that 
the transmitting operator has the proper 
2125 and 2975 cps tones set up, there is 
a much-relaved attitude about radio fre- 
quency control here, and both transmit- 
ter and receiver need only be set that at 
least the AM or FM signal gets through. 

In the next part we will describe the 
W9TCJ autostart printer system, with 
circuits shown, including description of 
a stable receiver necessary for low-fre- 
quency amateur autostart work on a 24 
hour basis, for 80, or 40 spot frequency 
set-up And the final part will include 
some hints for frequency spotting and 
adjustment, along with a circuit for in- 
jecting the 60.0 cps buzzing signal into 
the RTTY transmitter. 


In the first part, some discussion was 
given on principles and design criteria 
underlying the automatic start system 
using the reed-relay method of selective 
response. The following material gives 
down to earth practical information on 
the circuits used in the W9TCJ system, 
together with appropriate discussions 
pertaining to particular features as they 
come up in the paper. We will now be- 
gin with the signal input to the receiver 

and work towards the final result — a 
message received on unattended basis. 

1 — The Radio Receiver: 

First of all, we have to keep in mind 
the need for “radio frequency control” 
and find out how we may achieve such 
a stable receiver that can be relied upon 
to monitor the given channel and accept 
any and all directed RTTY signals within 
a specified tolerance of plus or minus 
60 cps. One receiver that would fill the 
above bill is a Collins 75A; it being a 
double conversion superhet with a crystal 
controlled front end feeding into a first 
i-f amplifier. After this latter amplifier, 
there is a mixer fed by a permeability 
tuned oscillator covering a range in the 
2 me region and this latter oscillator is 
"ery stable primarily from the fact that 
its frequency is low and residual drifts 
are of minor importance. However such 
a receiver is very expensive and really 
should not be “assigned” to monitor a 
single frequency at all times. What one 
needs is a simpler and ideally spot-fre- 
quency receiver for that kind of monitor- 
ing duty. 

There is the possibility of using fixed- 
tune radio receivers, such as the Wilcox 
CW-3’s. This is la quite good idea, be- 
cause the front-end is crystal controlled 
and can be relied on to be highly stable 
for our purposes. However the i-f of 
such a receiver is quite broad and will 
take in interfering signals that happen 
to be adjacent to our autostart channel; 
and more seriously, the “audio-image” 
signals (from other side of zero beat). 
This suggests that such a receiver system 
will need Q-5er i-f selectivity and it could 
be installed using a low frequency i-f as 
an extra on the receiver itself. The 
final combination would give us about 
the right equipment needed for our auto- 
start set-up. Furthermore, as the ter- 
minal unit, here, has no limiter stage, 
we shall want AVC in our receiver de- 
sign in order to maintain the audio 
signal level fairly constant when re- 
ceiving RTTY. 

Speaking of the “Q-5er” why not make 
use of a BC-453, 190-500 kc Command 
receiver and fit it with a crystal con- 
trolled converter? All that is required 
is to build the converter and one then 
comes up with the “made to order” auto- 
start receiver! A most excellent descrip- 
tion of such a set-up appeared recently 
in the January, 1956 CQ magazine, by 
Don Stoner, W6TNS. 

A very similar set-up is used with the 
W9TCJ Autostart system and I have 
found it highly satisfactory and except- 

ionally stable; in fact the continuously 
running receiver can be set on channel 
and it will stay to within a few cycles 
for days and weeks on end without re- 
tuning. I heartly recommend this kind 
of receiver for spot-frequency working 
as it is so very stable and yet costs 
little to construct, all that is needed is 
a BC453 receiver in good condition, with 
a crystal and a few small parts for the 
converter and power supply. 

Of course my receiver has a few dif- 
ferences from the W6TNS model. In 
order to always have the crystal fre- 
quency lower than the signal frequency, 
to avoid ending up with “upside-down 
FSK signal output at receiver audio” I 
use separate crystals, namely 3200 and 
6700 kc, in the 6BE6 oscillator when 
working 80 or 40 meter bands. The rf 
ampjifier, using a 6BJ6, is entirely con- 
ventional; and coils for both that and 
the mixer are separate and switchable 
for either band. The Command receiver 
is tuned to near 400 kc when receiving 
RTTY signals on either band, and that 
i-f frequency is appreciably high enough 
to minimize image problems. I have 
not had any trouble with images on either 
band, and yet the ability to TUNE such 
a receiver anywhere in a band and still 
have a stable system makes the above 
combination a highly desireable one. 

The BC453 itself has been modified. 
The AVC line is returned to work right 
off the diode detector load (use 2 meg. 
resistor in series with AVC line to this 
point) instead of from the grid of the 
last 85 kc i-f stage. Bakelite pins on top 
of the three 85 kc transformers have 
pulled up and the latter’s alignment have 
ocen touched up in order to acheive an 
optimum i-f selectivity response for 
RTTY work. The audio stage has been 
modified to the extent of removal of the 
bypass capacitor that is across the out- 
put transformer as we want to have 
good output at 3 kc as well as at 2 kc 
from this stage. In regard to the BFO, 
it has been readjusted (Install a 40 mmf 
ceramic capacitor from BFO plate to 
ground. This shifts the BFO frequency 
low enough to set the mark and spoce 
signals symmetrically around the if 

peak). The series resistor that feeds B 
power to the BFO has been adjusted to 
obtain reduced BFO injection level; in 
order to (1) minimize the BFO effect 
upon the AVC and (2) to regulate the 
amount of heterodyning in the second 
detector. (Actually, the original 20,000 
ohm resistor will probably be adequate 
because the further offsetting of BFO 

frequency from i-f peak cuts down on the 
injection quite markedly and happens 
to be about right amount in my set-up). 
In fact, with such a limited BFO level, 
the audio output remains fairly constant 
even if the i-f signal level varies. This 
feature, coupled with AVC action, aids 
in providing ia constant level of audio out- 
put power over a wide range of rf input 
signal intensities, a necessity for work- 
ing into a limiterless terminal unit. 

Other receiver modifications include re- 
wiring the heater circuit and installation 
of a power supply of sufficient capacity 
for operating the entire receiver, conver- 
ter and terminal unit (except magnet 
keyer stage). Selenium rectifiers (this 
saves ten watts drain that otherwise re- 
sults with a rectifier tube filament) are 
employed with a transformer (Stancor 
PM-8419), and feeding into a single 20- 
30 henry choke backed up by a 40 mfd 
electrolytic capacitor. The voltage out- 
put from this supply is some 180 volts 
at about 50 mils overall drain, the volt- 
age being low enough so that there is 
no strain on receiver components, thus 
aiding in long-term reliability. No volt- 
age regulator tubes have been found 
necessary. The selenium rectifiers are 
120 volt 65 ma units and four of them 
are employed; two in each HV secondary 
leg with a 500 ohm series resistor to 
limit surge through the rectifiers. As 
in the original Command receiver setup, 
a 20,000 ohm potentiometer is installed 
to handle the currents from the rf and 
i-f amplifier cathodes; this is a supple- 
mentary gain control and is employed 
to adjust the autostart receiver sensitiv- 
ity threshhold. A switch is installed to 
short the AVC line to ground, in order 
to disable the AVC when il is necessary 
to do so as in during receiver tuneups." 

Recapitulating on the receiver portion: 

1. BC453 receiver with xtal controlled 

converter, run on selenium power 
supply, transformer operated. 

2. AVC line returned to diode detector, 

through a 2 meg resistor. AVC line 
has a switch so it can be shorted 
ground whenever necessa. y. 

3. BFO offset from i-f peak and injection 

level adjusted if necessary. 40 mmf. 
capacitor installed. 

4. Pull up i-f transformer bakelite pins 

and realign if necessary. 

5. In the converter, take note that error 

in the Jan. ’56 article which has 
caused consternation, namely; in 


R T T Y 

R T T Y 


410j\. - + 


6 H 6 


(AGC * AVC) 




*. S.R : SELENIUM RECTIFIERS - HALFWAVE 65 ma. 115 v. *y„,. <JAN ' ® 57 ’ P * ' 





&4W- 71- - UiiUVi. lo 1TCJ 






R T T Y 


stead of the .005 mf capacitor called 
for in the 6BE6 oscillator circuit, 
use ia 50 mmf unit, and then the 
crystal will oscillate. 

6. Have a 20,000 ohm volume control on 

the receiver, to adjust the current 
through the cathodes of rf and i-f 
amplifier stages, as a means of sen- 
sitivity threshold. 

7. Audio bypass capacitor omitted. Out- 

put transformer impedance placed on 
high point. 

In short, following the above hints will 
give you a very fine little receiver for 
spot frequency operation. Overall power 
drain is about 30 to 40 watts, so leave 
it on continuously in order to have 24- 
hour monitoring and at same time have 
a well warmed up and stabilized receiver. 
One other item I would like to try out 
soon is a product detector to replace the 
diode second detector in the receiver to 
see if it might give an improvement. 
You might want to consider that, too. 

2a — The Terminal Unit with Autostart: 

Figure 4 show's the circuit of the ter- 
minal unit, employing a total of three 
tubes, for keying the teleprinter magnet 
as well as for autostart. In the interests 
of simplicity and low pow r er consumption 
as well as for other reasons, no limiter 
stage nor bandpass filter is used in this 
circuit. Audio selectivity is then de- 
pendant upon the toroidal filters used 
for mark and space discrimination. These 
88 mh telephone toroids were described 
in January, 1957 RTTY. together with 
a discriminator circuit and some oper- 
ating notes. The BC453 receiver audio 
stage feeds directly into the mark and 
space filters via an impedance matching 
transformer, Thordarson T-24S60. The 
outputs from the filters go into a pair 
of diodes, in the form of a 6H6 and this 
detector circuit is arranged so as to yield 
a total swing of 40 volts which next feeds 
into a triode DC amplifier. On mark 
the rectifier delivers negative 20 volts 
with respect to ground and on space it is 
positive 20 volts; assuming that things 
are properly adjusted at the receiver. 

The DC amplifier serves double duty, 
one for amplifying the DC swing enough 
to properly operate the maget keyer tube 
and the other, for driving the exciting 
coil of the reed -relay. This 5000 ohm 
coil is a part of the amplifier’s loud 
resistance and thus the flowing plate- 
current provides t h e needed ‘‘magnetic 
bias” to cause the reed to follow the 
fundamental frequency input. One triode 

thus suffices for both purposes and this 
is contained in one-half of a type 5963 
tube, a 12AU7 type tube with special 
processing to prevent ‘‘sleeping sickness" 
due to long periods of bias cut-off. The 
other section is employed for operating 
the autostart relay in its plate circuit; 
this relay serving two functions, to op- 
erate the motor-start circuit and to “shut 
off” the magnet keyer during the buzz- 
ing period. In order to operate this 
other triode circuit, a cadmium sulphide 
photocell, Clairex type CL-2, is mounted 
on the reed-relay assembly so that it 
“looks” 'at a radio dial lamp running 
at reduced voltage when the reed is 
vibrating at nearly full amplitude. This 
photocell functions as a light-sensitive 
variable resistance; it then dropping in 
value from many megohms in dark to 
some 500,000 ohms when exposed to light. 
One side of the photocell is connected 
to B plus, and the other side feeds into 
a load resistor. The resulting voltage 
drop then feeds into the grid of the re- 
lay triode, driving it positive and caus- 
ing the triode to conduct. During “dark” 
i.e. no buzzing, the triode is biased to 
cut-off by means of a NE-48 lamp placed 
in its cathode. As a partial provision 
for autostart sensitivity adjustment, a 
potentiometer is connected across the 
neon lamp, and the photocell load resistor 
is returned to the pot’s moveable con- 
tact: in this way, the amount of negative 
bias is controlled on the relay tube’s 

2b — The Magnet Keyer Circuit and 
Motor Shutdown System: 

Returning to the first section of the 
5963; i.e. the DC amplifier, the plate- 
swing output from the mark and space 
signal is fed to the grid of the 6W6GT 
magnet keyer tube via a couple of NE- 
2 “voltage steppers” and a pair of con- 
tacts on the autostart relay. The latter 
is needed in order to connect the grid 
of the keyer to a positive source of volt- 
age and thus keep the keyer conducting 
on steady mark during the buzzing period 
and resulting motor startup. When the 
reed-relay stops vibrating, light is cut 
off from the photocell; and this causes 
the aforementioned relay to open, thus 
releasing the grid of the keyer to be 
controlled by the incoming intelligence. 

The magnet keyer circuit uses a 
6W6GT arranged so that the tube func- 
tions as a switch, with the Model 26 tele- 
printer magnet in its pliate circuit. DC 
power is supplied by a small transformer 
operated selenium power supply, entirely 
separate from that of the receiver and 
running some 150 volts at 30 mils, as 

well as providing a small amount of 
current to the NE-2 “keep-alive” circuit 
at -170 volts. This too assures that the 
6W6GT grid will swing negative cut- 
off on space transitions. This is an ef- 
fective magnet keyer and is in use also 
in other W9TCJ terminal units. By ad- 
justing the series resistance shown and 
if necessary by using somewhat larger 
components, 60 mils can be obtained for 
operating a Model 15 if that is desired. 
As the W9TCJ system stands, it is used 
with a Model 26 that has not been re- 
wired; i.e. the signal output from the 
magnet keyer is merely hooked into the 
printer’s signal line as it was originally 
employed for landline work. This makes 
the keyboard circuit available for making 
tests on the printer and for locally shut- 
ting down the machine merely by press- 
ing the keyboard “break button.” 

Now r that mention has been made of 
the provision for “local printer shut- 
down,” it is well to discuss this in some 
detail and show how it is accomplished, 
not only locally but remotely. Pressing 
the “break” button opens the printer 
signal line and causes it to go on a steady 
space condition for the duration that the 
button is down. ’ Shown in Figure 4 is 
a simple R-C network, equipped with a 
rectifier, that feeds into a printer shut- 
down relay which contacts in turn con- 
trol the lockup circuits on the motor 
relay. The R-C network is placed in the 
signal line of the printer magnet keyer 
circuit and if a prolonged space (no 
current) condition is present, the charge 
on the capacitor is lost through the re- 
lay, and then it opens, thus releasing 
the motor relay. On mark, current flows 
thus charging the capacitor and keeping 
it supplied with energy so that the relay 
is held closed, thus keeping the motor 
relay lockup circuit closed. The selenium 
rectifier is used as a “one-way valve” 
to prevent the capacitor from emptying 
its charge back into the signal line in 
case of space transitions of a normal 
teleprinter signal. The time constant of 
the R-C network is adjusted so as to 
open the motor lockup circuit when a 

porlonged space signal is received, over 
one quarter or one half second. This is 
a simple yet effective method of printer 
shutdown, depending upon the fact that 
during a normal transmission, mark is 
always present a great portion of the 
time; and moreover the longest space 
signal in any teleprinter character is the 
“blank” consisting of 132 milliseconds. 

This R-C system is hooked across the 
series resistance that regulates the print- 
er line-current so the voltage drop that 
exists across that resistor energizes the 
circuit. It only takes about 5 mills 
from the line (adjust the resistance un- 
til at least 30 mills flows through the 
printer magnet itself) and tests have 
shown that there is no effect on the 
printer's range (20 to 110, centered on 
65). The oapacitor is a 16 mfd electro- 
lytic and it supplies energy through a 
resistor to the coil of the relay thus 
keeping it closed, so long as there are 
some mark energy pulses coming through 
the rectifier to keep the charge replen- 

Thus, the printer shutdown control is 
described and it is obvious that it is 
under the distant tranmitting operator’s 
direction; all that he need do is to trans- 
mit at least 5 seconds of steady-space, 
merely by holding down his keyboard 
“break” button. 

2c — Motor Control Circuits: 

This portion of the system is largely 
self-explanatory, as references have been 
made to various relays used for motor 
control. There are four relays directly 
involved; (1) the reed-relay, which upon 
receipt of a buzz-signal of the proper 
frequency, causes the closure of (2) the 
autostart relay via phototube and triode 
relay stage. This relay closes the circuit 
to the (3) power contactor or motor relay, 
thus starting up the motor in the printer. 
During this time, and as long as the buzz 
is on, the autostart relay holds the 
magnet line on steady mark to preve. t 
wild printing. The motor relay has a 
pair of contacts that keep it locked up 
after cessation of buzzing, and the print- 

1G R 

er is now ready to take the message. And 
when the printer is to be shut down, the 
steady-space signal allows the (4) print- 
er shutdown relay to open the lockup cir- 
cuit, resulting in motor relay opening 
and and shutting down the motor. The 
model 26 teleprinter power cord is merely 
plugged into the motor control circuit, 
and no modification of any sort have been 
done to the machine or its table wiring, 
as it presently stands. 

3 — Conclusion : 

The entire W9TCJ autostart, employ- 
ing a method of selective response and 
useable on h-f amateur bands with FSK 
emission, is described and circuits have 
been shown. The equipment was designed 
with the factor of ease of control-signal- 
ling foremost in mind; and using a sim- 
plified terminal unit with autostart. The 
readily available 60.0 cps. “signal” from 
the power line serves excellently for the 
“buzzing signal” that is easily injected 
into the RTTY transmitter as mark-and- 
space reversals. The selective control is 
obtained by using a reed-relay that can 
be constructed at low cost, and, further- 
more, it could be extended to some other 
specific buzzing frequencies — resulting 
in a versatile individual-station calling 
system. Not only can different buzz- 
frequencies be used, but different radio- 
frequency channels as well, thus multi- 
plying the number of discrete stations 
that can be called. The 60 cps frequency 
could and should for the time being used 
as an “universal” buzz-frequency — and 
this brings up another idea. 

It is entirely possible that a system 
could be arranged, along these lines, so 
that the 60 cps buzzing signal results in 
motor starts on all printers within range. 
Now, as it is desired to operate only a 
certain printer on a directed control 
basis, one could transmit a special “tele- 
printer call-code” consisting of a com- 
bination of two letters, such as RWI1W. 

A pair of microswitches mounted so as to 
sense the code on the typing unit of the 
one “RW” printer called would then 
energize a lockup circuit to keep the 
machine running. All the rest of the 

T Y 

printers would close down, controlled by 
automatic shutdown delay circuits, oper- 
ating after say 15 seconds. In this way 
a message could be sent to a specific 
printer that is one of a multitude of 
machines using the same channel and 
buzz-frequency. Give this a thought. 

This W9TCJ autostart was first placed 
into operation in June, 1956, and tests 
were made to the home-printer from the 
Starved Rock Hamfest at Ottawa, Illinois 
just 100 miles south. The regular station 
was taken down there and set up for 
demonstration of ham RTTY. On 3622.5 
kc a number of calls were made to the 
home-printer — resulting in successful 
unattended printer reception there. My 
mother received these messages and was 
so enthused that she put in a long dis- 
tance telephone call to Ottawa to let me 
know that my one-way messages were 

Since then, the equipment has been on 
more or less continuous monitoring basis 
and has received some messages from 
W0BP and, more recently, from W8SDZ 
Keith Peterson of Swanton, Ohio, to 
whom my sincere thanks must be expres- 
sed for all the efforts that he went to 
to make a great many one-way trans- 
missions to aid me in evaluating and ad- 
justing the autostart system for best 
performance. Keith has proved his ability 
to set his transmitter close enough in 
frequency to cause hightly reiable auto- 
start reeeiption here at W9TCJ. 

The system is immune to false-starts 
due to CW or RTTY signals that happen 
to be on frequency, as well as to noise. 
One could ask now whether a 60 cps 
buzz of any kind that happens to get 
into the system could start it, and the 
answer is a “qualified no.” The reed- 
relay and autostart circuits have been 
adjusted so that nearly the full 40 volts 
of discriminator swing is required to 
bring the reed-relay to near maximum 
amplitude of vibration; and to cause 
that a FSK signal is required, containing 
60 cps equalized mark and space rever- 
sals, and with a 850 cycle shift, properly 

R T 

centered in the autostart channel. This 
is one design consideration aimed at 
avoiding false starts from such noise 
as “ITV” television receiver noise (horiz- 
ontal oscillator) that is rampant on the 
lower frequency amateur bands in locat- 
ions where there are nearby TV nabors. 
Actually as such a noise sweeps through 
the autostart channel, the printer will, 
very rarely, start up and run wild for 
just five or ten letters, then shut itself 
down. This happens on the average of 
once a week or less often, and there are 
three TV sets within 50-100 feet of the 
W9TCJ QHT. The tuning meter on the 
autostart receiver will respond and show 
such signals. I am not worried about 
this problem as it is inconsequential and 
the printer has never been found to be 
running for hours on no-signal; its print- 
er shutdown circuit has such a short 
time-constant that if one quarter or one 
half second of steady space is had, it 
operates. And the simplified terminal 
unit has a slight no-signal spacing bias 
tendency that serves to aid in shutdown. 
In fact, when a strong RTTY signal is 
taken off the air, the system shuts down 
by itself “at once” because the AVC in 
the receiver takes a certain time to re- 
adjust itself: and hence when carrier is 
cut, the audio disappears from the TU 
and the aforementioned spacing-bias 
tendency comes into play to shut down 
the printer. In regard to ITV and 
noise in general, the gain control on the 
BC453 receiver is adjusted so as to cause 
a barely perceptible reading on the tun- 
ing meter on no-signal. This way, one 
assures shutdown, and the threshhold is 
such that signals from stations nearby 
(of order of 300 miles or more) have 
sufficient signal strength to operate the 
W9TCJ system, on a relatively short in- 
door antenna wire. 

As for quality of printing. On good 
steady signals, the system prints 100 
percent perfect, as has been proved by 
reception from stations such as W0BP 
and W8SDZ, both stations some 300 to 
400 miles from here, as well as from 
stations closer by. On weak rapidly 
QSBing signals, often the system will 

T Y 17 

print tolerably well, yielding messages 
that have occasional hits, even though 
such a signal may print perfectly on 
another TU that has an efficient limiter 
on its input. So here in this present 
TU design, compromises have been taken 
in interests of simplicity, low power- 
drain, reliable shutdown. I tried a lim- 
iter on this autostart TU and discovered 
that the added circuit generates so much 
noise that the discriminator, DC amplif- 
ier loses its characteristic no-signal spac- 
ing bias tendency. The printer then tends 
to run continuously, printing wild on no 
signal. Additional circuits are being de- 
vised to assure this printer shutdown on 
noise alone, and this is part of some 
current development work 1 am doing, 
on a “Mark II” autostart TU. 

In retrospect; it seems that machine 
telegraphy is at its best in two fields: 
(1) Wholesale message traffic handling, 
where the high machine printing speed 
is beneficial in transmitting large 
amounts of information over a circuit in 
the shortest possible time, and with the 
people who are relatively unskilled in 
“Morse Code,” they only need be good 
enough typists to operate these tele- 
printers. (2) Automatic unattended print- 
er operation, where messages can be 
transmitted to and left upon such 
machines without anybody being present 
during reception, thus a kind of “mes- 
sage recorder,” akin to a postman leav- 
ing a letter in one’s mailbox while ad- 
dressee is out or otherwise occupied. In 
this latter application, autostart printer 
are abviously of utility. And speaking 
of Amateur Radio, as one well knows, 
this game is a “catch as can catch” af- 
fair, when one wants to talk to a particu- 
lar station, he has to keep previously 
arranged “sked” or take a chance that 
the other fellow happens to be on the 
air at that particular time. I suppose 
that various ways have been devised by 
hams to enable instant call and answer 
possibilities to be realized on phone or 
CW but with varying degrees of success. 
And here, with our RTTY equipment, we 
have a golden opportunity to put our 
teleprinters on “instant call” at all times. 

18 R T 

Busy people, as has been remarked be- 
fore, can make good use of these auto- 
matic unattended message recording de- 
vices and also be able to call each other 
if necessary. In this way we can keep 
all our precious teleprinters on duty as 
much as possible, whether in actual com- 
munication or on standby monitoring on 
call. Let’s go autostart, we have the 
machines, and I have presented one 
method of autostart in this write-up. 
Can we devise a simpler and more re- 
liable autostart system? Such a system 
should have the desirable features of 
selective response, print well on directed 
signals at any reasonable time, and be 
immune to other signals and noise. 

The third and final part will now con- 
cern itself with some methods of radio 
frequency control and buzz-signal in- 
jection into a RTTY transmitter. 


1 — “Radio Frequency Spotting:” 

As pointed out in the preceding pages, 
the major problem of “radio frequency 
control” if we are to achieve reliable 
FSK autostarts on low frequency amateur 
bands. Frequency settings and stabilities 
assume equal importance in both trans- 
mitter and receiver ends of a proposed 
autostart circuit. It has indicated that 
one should be able to set up a frequency, 
on either transmitter or receiver, to with- 
in a tolerance of plus or minus 50 cps; 
and that is quite easy to accomplish this 
with the proper auxiliary equipment. 

In the second part, a description of a 
receiver that has satisfactory stability 
for this kind of work was given. It is 
indeed remarkable that the little BC453 
receiver, equipped with a crystal con- 
trolled converter, has adequate stability 
and can be relied upon to monitor an 
autostart RTTY channel on a continuous 
basis; and to do this for days and weeks 
without any need for retuning. Again I 

T Y 

repeat, I highly recommend this receiver 
set-up primarily because of its stability, 
selectivity, ease of tuning to different 
frequency, and low cost. 

Now we will discuss the transmitter 
frequency control problem. We want to 
be able to “aim” the think with the re- 
quired degree of precesion in order to 
score a direct hit upon the bullseye, so 
the autostart TU obtains its properly 
placed mark and space tones through 
the receiver as set up on the channel, 
one logical method is crystal control, 
and this is an ideal way when it is de- 
sired to have a complete autostart RTTY 
station that requires a minimum of at- 
tention of any sort once set up on a 
specified channel. In other words, only 
the teleprinter should be in plain view, 
equipped with the necessary controls for 
transmission, reception and the like — 
and the rest of the equipment placed out 
of the way in a locked cabinet. Life is 
getting complicated enough as it stands 
now-a-days and we desire to acheive the 
ultimate in simplicity and ease of oper- 
ating our radio teleprinter equipment. 
We want to get rid of frequency control 
worries and problems in the simplest 
manner we each are able to do, and we 
will now' discuss some various methods 
of frequency spotting. 

An ideal frequency spotter is a 100 
kc crystal standard, equipped with multi- 
vibrators for subdividing down to 10 kc, 
2.5 kc and 2 kc. This instrument gener- 
ates and places precise beats all the 
way through the radio spectrum, and 
w'hen the 100 kc crystal harmonic is 
synchronized to zero beat on a signal 
from WWV, then one obtains calibration 
accuracy of one part in a million or 
better. It is then practical to get a de- 
sired “reference point” accurate to several 
cycles on the 80 meter band, for instance. 
Such standards are available in kit form 
or otherwise, from various sources (One 
source: International Crystal Mfg. Co., 
Oklahoma City) and do not cost much, 
yet are of tremendous usefulness and 
necessity for the amateur interested in 
precise frequency spotting. 

R T 

Some of the higher priced receivers 
have built in crystal standards, and such 
are usable to correct the bandspread dial 
to obtain “direct frequency reading.” Pro- 
vided the dial is actually linear and does 
not deviate from indicated frequency 
points over the range between successive 
100 kc points, such a receiver should be 
capable of precise frequency spotting. 
It is important that the dial be calibrated 
in steps of integral kilocycles (or better) 
and that the tuning mechanism be free 
from backlash. The Collins 75A re- 
ceivers are supposed to be in this cate- 
gory and it is interesting to review some 
of the tests that Keith Petersen, W8SDZ 
and I conducted relative to frequency 
calibrations and spottings. In the Fall 
of 1956 Keith took interest in autostart 
and commenced tests and call to my 
equipment, depending on the dial cali- 
bration of this 75A receiver to align his 
transmitter VFO to the channel. Taking 
care to preset his receiver dial to a 100 
kc spot from his internal 100 kc standard, 
he set the receiver next to 3624.0 kc 
mark and afterwards tuned up his trans- 
mitter VFO to that frequency. My ob- 
servations of his resulting transmitter 
frequency, using my precise (used in 
ARRL FMT’s) equipment indicated that 
he was able to hit the channel consist- 
ently within 50 or 100 cycles and resulted 
in many successful autostarts here at 
W9TCJ. In fact, hardly any message 
w'as missed and this all the more re- 
markable, considering t.he many adverse 
factors that oan and do intrude to block 
the circuit between calling station and 
called station. 

Keith employed that method above for 
some time until trouble developed in 
the PTO part of his receiver, necessitat- 
ing replacement of that portion with a 
new unit obtained from the Collins 
factory. After the change was made, it 
turned out that the dial calibration be- 
came grossly nonlinear and the kilocycle 
divisions no longer indicated true fre- 
quency settings. Some dial settings 
were found to be as much as several 
hundred cycles off, especiaally when a 
ways from the 100 kc calibrating point. 

T Y 19 

So this indicates that commercially avail- 
able receiver equipment is not to be 
trusted too far as far as frequency cali- 
bration is concerned, even if equipped 
with 100 kc standards. For instance, 
most such dials only read to one kilo- 
cycle divisions and W'hen one considers 
the 50 cps tolerance required in the 
autostart system, this means one has 
to estimate his dial setting to wdthin 
l/20th of a division! And extreme care 
has to be taken to avoid dial-scale paral- 
lax, that phenomenon of changing dial 
reading due to change in view'ing angle 
upon dial from eye, with respect to the 
dial’s fidicuial mark. Considering all the 
above factors it is really amazing that 
Keith succeeded so well in leaving mes- 
sages on my unattended printer. 

Keith went to crystal control on his 
transmitter on the autostart channel and 
is now consistently successful in leaving 
messages here. He uses an International 
Crystal FA-9 unit, calibrated to a fre- 
quency 0.6 kc lower than the desired 
channel mark frequency, in a standard 
Collins 709-D-l type FSK circuit and 
obtains both the stability and full 850 
cycle shift on the 80-meter frequency 
of 3624.000 kc mark high. Yes, indeed, 
full shift is available on such a low fre- 
quency and for the purpose of informa- 
tion, it was found that a plated crystal 
such as International’s was capable of 
sufficient rubbery flexibility to go 850 
cycle shift. FT-243 type units do not 
permit such a degree of shift as they 
are differently manufactured. Further- 
more, the FSK oscillator circuit was al- 
tered to the extent of keeping the oscil- 
lator grid to ground capacity as low as 
possible to obtain sufficient shift at such 
a low frequency. Short leads ware found 
necessary as well as proper paits place- 
ment. No trimmer capacitor to grid 
from ground, or at least a very small 
unit as was found needed by Keith in 
his circuit. All in all, crystal control as 
he demonstrates results in satisfactory 
transmitter frequency calibration hold- 
ing and checks show' his frequency to 
be well within 50 cycles, often 10 or 20 
cycles from the channel frequency, plenty 



pood enough! I he plated crystals do 
not seem to drift or vary much with time, 
age, or even temperature changes so far 
as has been determined in the past few 
months of use in various applications. 
(I have about eight or ten such crystals 
in various places at W9TCJ; all by In- 

About using a transmitter VFO. Sure, 
it can he used, provided that it is cap- 
able of reasonable short term stability, 
enough to get a message across, and that 
some means of frequency reference point 
is available, such as a 100 kc standard 
with proper multivibrators. Or, as pres- 
ently employed in my station and at 
W9LDH’s, a newcomer to the field of 
autostart, the same International Crys- 
tal FA-9 units, calibrated to 3624.4 kc, 
are operated on grid dipper (Colpitts 
type) oscillators, yielding a reliable 
reference point for 3624 kc. The grid 
dipper is tuned or adjusted so as to 
obtain that desired frequency, depending 
upon other standards (or over the air 
signal from W9TCJ) to get that spot. 
Knowing the dial setting on the dipper, 
Spence and I, each, are able to generate 
the exact frequency at any time we de- 
sire to have it. Furthermore such crys- 
tals will eventually be used in transmit- 
ter FSK exciters, so they are a fine in- 
vestment at $3.00 each for the beginning 
RTTY autostarter. 

So much for frequency standards. By 
far, the 100 kc standard with 10 kc. 
multivibrator wi 1 1 deliver precise points 
10 kc apart all the way through the 
spectrum and therefore is a dependable 
piece of equipment, when properly em- 
ployed. It would be desirable to add a 
2.5 and 2 kc multivibrator; either value 
selectable by changing appropriate time 
constants in the MV stage, because us- 
ually a finer degree of frequency select- 
ion is desired, such as choosing 3624.000 
kc or 7137.500 kc spots for autostart 

2 — Transmitter Tuneup Methods: 

Various methods of checking for zero 
in on selected frequencies are available. 

My method is to operate a standard radio 
receiver (such as a BC348Q) in its sharp- 
est xtal selectivity position, i.e. Single 
signal with audio peaking and BFO 
circuits thrown in, and tune in a desired 
beat-signal from the standard. I ascer- 
tain I have the right signal by counting 
from the nearest 100 kc and/or 10 kc 
point and by reference to known receiver 
dial settings. Next, I adjust the VFO 
until its note beats against the standard 
signal. Several seconds of touch-up en- 
ables the VFO to be set to within a cycle 
or two of the standard beat. 1 do this 
daily in various measurement jobs and 
the experience gives me a clear under- 
standing of what I am doing. 

One method, as presently employed by 
VV9LDH, is merely to zero-beat the re- 
ceiver upon the standard (3624 kc signal 
from xtal on a grid-dipper), getting the 
null between the two audio beat notes. 
Then the VFO is turned on, and it is 
zeroed in on the previously-set receiver. 
Depending on the lowest frequency pass- 
able by the receiver audio stages, one 
could get to within perhaps 20 or 30 
cycles of the actual frequency. 

Still another method is to employ the 
standard RTTY tuning indicator system 
(scope, or magic eyes), having previously 
adjusted the receiver upon the standard 
signal to obtain maximum response on 
Mark and adjust the VFO so it gives 
maximum response to the same Mark 
frequency. My set-up includes a shift 
meter which, combined with the other 
indicators to show sense of shift, en- 
ables me to zero in to matter of several 
cycles in a short time. 

No doubt, there are other equally ac- 
ceptable methods of zero in, depending 
on the particular equipment and habits 
that a given experienced RTTYer has. 

It is only required to be able to get to 
within a few T cycles of a certain specified 
frequency; and furthermore, the specifi- 
cation is alw’ays on Mark-High basis, 
so make sure your VFO is on mark when 
tuning it in. 




R T T Y 

R T T Y 


Magnet Keyer and some other Relays are separate on another chassis, 
not shown, but all can be combined on one chassis 


3 — Receiver Tuneup Method: with 
Reference to W9TCJ System: 

Having a 3624 kc signal see up (as 
from a xtal dipper), the autostart re- 
ceiver is merely tuned for maximum 
reading on the tuning meter (200 micro- 
ampere meter shown in Figure 4). In 
order to resolve the ambiguity as to 
sense (Mark or Space), the receiver knob 
is always tuned in a counter-clockwise 
direction until the first peak is reached 
on the meter; this is the Mark setting; 
the second peak being the Space setting. 
A little experience will show what is 
necessary to do to be sure of having 
the receiver on channel. A pair of NE-2 
lamps could be readily wired into the 
TU so as to show Mark or Space sense 
during a tuneup and operation. 

4 — Buzzing Circuit: 

Figure 5 shows a suggested circuit for 
introducing the 60 cps buzzing signal 
into the FSK circuit of a RTTY trans- 
mitter. The signal is obtained by vibrat- 
ing a polar relay contact from off the 
AC powerline; the contacts involved then 
interrupt the keyboard circuit at a 60 
cps rate thus generating square waves. 
It will be necessary to check and adjust 
the contacts for best reproduction; this 
is easily done beforehand by using an 
oscilloscope: feeding a little voltage, DC 
through the signal circuit and observ- 
ing the waveform on the scope. This 
results in a satisfactory method of gen- 
erating 60 cps mark-space reversals, and 
the whole thing could be built into a 
small wood box. 

Alternately in some RTTY stransmitter 
circuits, employing direct electronic key- 
ing it is possible to introduce 60 cps 
voltage internally so as to transmit a 
buzz signal by means of a push button 
placed for that purpose. Such is done 
in the W9TCJ transmitter circuit mere- 
ly by feeding a little voltage from the 
6.3 volt line into the FSK Diode Driver 
grid. (Use series resistor to limit the 
current flow). 

Anyhow, the idea is to transmit 60 cps 
mark-space reversals over the RTTY 
circuit to activate the reed-relay in the 
autostart receiver system. As shown 
in Figure 6, the signal pattern is such 
that at least five seconds of 60 cps 
buzzing is required before transmitting 
a message. This assures that the reed- 
relay has a reasonable chance to come 
up to full amplitude. And at the con- 
clusion of the buzzing interval, the trans- 
mitted signal must rest on Mark; not 
on Space for even and instant or else 
the printer shutdown circuit may op- 
erate and thus close down the motor. 
The latter circuit has a purposely short 
time constant, of order of one half a 
second. This is the reason for the 
particular wiring of the start push- 
button in Figure 5, it is arranged so 
that when the button is up, it closes 
the keyboard loop. 

After the message is sent, the shut- 
down procedure is simple. Hold the key- 
board Break button down or the Stop 
button on the buzzing-box for at least 
five seconds; this transmits a steady 
Space signal for that duration. 

5 — Conclusion: 

This concludes the three-part paper 
on the W9TCJ Autostart System, to- 
gether with frequency setting and buzz- 
ing discussions. As it stands, the auto- 
start system is in excellent operating 
condition and monitors the channel on 
a 24-hour basis, ready to accept and take 
any directed message at any time from 
other stations. The standby power 
drain is about 30 or 40 watts. The 
equipment is adjusted so for all practical 
purposes it is immune to false-starts due 
to noise, etc. In short, I am well pleased 
with the set up as it stands and I be- 
lieve you would be as pleased with the 
system. How about going autostart, 

Vive Autostart! 

Vive Amateur Radioteletype! 

Vive Amateur Radio!