June, 195E PRACTICAL TELEVISION 522

the writer
WHENWHEN
sat down
with log tables
A involve at least two more
stages -so again they

TV kìttcrii
were not included in this
and book of valve data ' design.
to design this apparatus. Both horizontal and
it was with some sense vertical modulation bars
of misgiving. On the

Generator
can be used to test
one hand there was the linearity. The vertical
major expense involved bars arc the only good
in purchasing a reliable test f or horizontal
generator, and on the linearity. but again the
other the complexity of the horizontal bars are not strictly
problem for home construc- needed because each line of
tion. Nevertheless. the need the picture is a horizontal
IA as urgent for an instrument
bar and vertical linearity can
able to produce a workable be investigated by inspection.
imitation of the BBC Fo; the rest of the design
television waveform. it may be said briefly that the
It was realised from the sync pulses are a good
start that it might not work imitation of the British
properly. The British TV standard; the line fly -back
waveform is almost incredibly pulses are identical. the frame
complex when analysed in pulses closely similar. No
detail, and the means to front or back " porch " has
produce it synthetically in an been provided -again
accurate fashion must like- for
simplicity in circuitry -but
wise be complex. Complexity their loss has not been felt
in apparatus involves certain acutely.
undesirable matters such as To sum up then, it may he
cost. reliability, difficulty of This Useful Tesi Set Duplicates said that with the instrument
setting -up, portability, and so now to be described a steady
on. Besides these. the the BBC Waveform for Test
raster can be obtained con-
probability of unwanted stray Purposes By D. R. Bowman sisting of 200 lines repeating
couplings between the various
sections increases geometrically as the number six to ten lines suppressed frames
at per second. with
50
during
of circuits increases, and the cut -and -try develop- Thus a receiver will respond to it frame fly -back.
ment involved in experimenting with layouts BBC or ITV waveforms, without line or to just as the
might well be prohibitive in time and money. speed modification. In addition. the raster can frame
So to aim at simplicity seemed sensible for all modulated at will with any number be
possible reasons. between ten
and 300 vertical bars, which enables both linearity
In this case, simplicity in circuitry is necessarily and resolution to be investigated. Also,
tied up with simplification of the result to be oscillator incorporated can tune over thethewhole R.F.
achieved. And the first thing to do was to take of the I.F. range of receivers; and on harmonics
a critical look at the British TV waveform and the entire BBC television
to try to estimate what is essential and what is at once that this generator band. It may be said
is for the experienced
luxury. constructor only, because its simplicity is only
First, as to interlacing. It would be wise to relative!
have it, certainly; but to introduce it would have
meant at least four extra stages. and for this The Circuit
reason it was reluctantly decided to make do
without. After all, few receivers get á good The generator consists essentially of the
interlace, and on a good many de- focusing or following circuits:
" whites" is so bad that it is hardly useful when (a) A multivibrator running at approximately
it is attained. 10,000 pulses per second, as pulse train generator.
Next, both vertical and horizontal modulation (b) A multivibrator "counting down" to 700
was considered. Horizontal detail is a function pulses per second.
of the goodness of receiver design adjustment, (c) A multivibrator running at 50 pulses per
and therefore horizontal detail had to be second for frame pulse gating.
included in the form of vertical bars. These must (d) A gating stage. combining the outputs of
be adjustable at will from " coarse definition " multivibrators (a) and (e) above.
to " very fine detail." and therefore means has (e) A keying stage, to be explained later.
been provided to have between ten and 300 (f) A R.F. oscillator, generating sine waves at
vertical bars; this corresponds to 100 kc /s and frequencies between 100 kc /s and 3 Mc /s.
3 Mc /s definition, and any number of bars can (g) A long -tailed pair combining R.F. and snyc
be tuned in at will. pulses.
Vertical definition is, however, a function of (h) A R.F. oscillator,
sharpness of focus and steadiness of trace. and of stage (g), generating modulated R.F.
by the output
this can readily be investigated by other means quency suitable for application at the output fre-
than a pattern generator. Horizontal bars would I.F. amplifier of a TV receiver. to the aerial or

www.americanradiohistory.com
 PRACTICAL TELEVISION June, 1958
530
In addition, phase inverting circuits. differ- application for the following reasons:
entiating and pulse shaping circuits are included (a) It has fast rise and fall times, giving a good
as necessary.
rectangular output at V2 anode.
(b) Its output is positive -going at V2 anode.
(c) It is exceptionally stable, when used in this
Description circuit.
Stage (a) is an asymmetric multivibrator genera-
It seemed originally that a symmetrical multi-
ting positive pulses at its second anode, at a rate vibrator whose mark -to -space ratio was variable.
of nominally 10.000 per second, each pulse being prove most suitable, especially as for
about 8 ps in width. This output is fed into a would
reasons given later both
14th or 15th positive -going and negative-
cipsecs -.I k--47- /OO,/secs
going pulses are required.
/0000 pps These circuits were found,
Mu/tividvtrr however, to be susceptible to a
particular fault, which caused
j, the counting down to vary in
Lafferentiatg
circuit Y- Ì ratio, apparently at random.
The master oscillator stage (a),
_... 60 secs running at nominally 10.000
approx.
pulses per second; t h i s
700 pps
/vlu/tivibratorr synchronises a 700 pps multi -
5) vibrator which in turn triggers
a 50 pps multivibrator. The
Fig. 1. -" Counting down " with multivibrators.
three are locked
cathode- follower type phase -splitter giving both together. I f th e
positive and negative pulse outputs. count -down rat i o
Stage (a) is arranged also to trigger a second alters, either between
multivibrator running at a nominal speed of 700
stage (a) and stage (b)
pulses per second. The correct sync pulse is Une sync. pulses
or between (b) and
obtained by passing the positive -going pulses (c), the number of
(abc
through a differentiating stage. which produces
positive and negative " pips ": the negative-going trames per second Fig. 4.- Combining phase -splitter o
synchr
pip is used for triggering: and thus there is a varies and therefore,
delay of about 8 its in the
triggering. Fig. shows the wave-
1 Phase-soUtter! Stage tS)

form involved, together with the

pulse train eventually developed
at the second anode of the second
multivibrator.
There are some points worth
noting about these multivibrators,
and they are conveniently dis-
cussed at this stage. The particular
type of multivibrator used is the
direct - coupled cathode - coupled
circuit, shown in Fig. 2.
This circuit has been described
previously in PRACTICAL WIRELESS
(December, 1955), but it was there
intended essentially as a saw -tooth
generator. It was chosen for this
*TO L /VE
HEATER LEA!
(SEE TEXT)

LIST OF C. AND R. VALUES Sta

C1-500 pF C16-.1 pF R6-5.6 K R22-200 K R37-68 K
-
C2-.02 t.F C17-S pF R7- K R23-2.2 M R38-2.2 M
's
; C3-.01 pF C18-250 pF R8-10 R24-33 K R39-1.5 K
CI9-500 pF R9-2.2 M R25-27 K R40-2.2 K
C4-50pF R10-10 K R26-5.6 K R41-10 K
C5-.1 pF C20-25 pF R42-1.5 K
C6-500 pF C21-.1 iF R11-1 K R27-1.5 K
C22-15 pF R12-10 K R28-10 K R43-2.2 M
G7-.1 pF R13-I M var. R44-2.2 M
C8-.1 pF C23-.01 /F R29-2.2 M R45-see text
C9-.002 RF C24-50 pF var. R14- 100 K R30-2.2 M R46-470 K
C10-.002 pF C25-50 pF var. R15-4.7 KM R31- 1.8 K R47-2 K
C11-8 pF elec. R1-200 K var. R16-40
R17-27 K R32-15 K R48-27 K
C12-50 pF elec.
C13-.1 pF
C14-16 pF elec.
R2-200 K
R3-2 2 M
R4-40 K
-
R18-5.6 K
R19-1 K
R20-10 K
R33-33011
R34-2.2 M
R35-2.2 M
R49-5 K
R50-see text
VR1-50 K
C15-500 pF var. R21-200 K var. R36-680 K VR2-70 K
(low min. capacity) R5-25 K

www.americanradiohistory.com
 June, 1958 PRACTICAL TELEVISION 531

on the TV raster, the number of lines displayed. vibrator runs at 50 pps. This is, of course.
This gives rise to " jittering " which was found to in order to coincide with the mains frequency,
be quite unacceptable. and it is highly recommended that this
Fig. I shows how the synchronising is arranged feature is adhered to in operation. It may not
in this apparatus. It will be noted that the trailing matter to be a cycle per second out, or there-
edge of each pulse produces, on differentiations, a
negative -going sharp pip which affords a very H.Tt
accurate sync pulse suitable for synchronising the R4 R7
lower -speed multivibrator. Each pip is preceded
by a positive -going pip; this serves the purpose Output
PS
of " holding -off " the second multivibrator if it
was about to trigger before time, and very con- V/
sistent synchronisation is obtained. In 4 hours
operation, no change in division ratio has been C2
observed except that manifestly due to ignition
or mains interference.
The same device is used between the 700 pps Sync
and the 50 pps multivibrators. The latter needs
a variable mark -to -space ratio, but this can be
arranged without difficulty. IP2 CI
RB,npot
There is, of course, only a limited choice of R
multivibrator speeds obtainable by the controls
provided, but Table I (page 532) shows HT-
witch over points
that this need cause
no inconvenience. The
limit of choice is due
Fig. 2.- Asymmetric cathode -coupled multivibrator
general arrangement.
-
to the fact that count- abouts, but some commercial TV receivers do
ing down can occur not " lock " properly at any but the mains
in the whole- number frequency. The adjustment to the correct
Line ratios only. frequency will be dealt with later.
me sync. pu /se
/0 lines in duration) It will be noted that
uts to give both line and frame in every case the Phase-splitting Circuit
sation. "frame" multi- Besides synchronising stage (b) the output from
stage (a) is fed to the grid of a cathode -
Stage (e) Stage (f) v Stage (h) follower type phase -splitter. This valve
wrMNm HT+25ov has equal anode and cathode load
V9 ECC8/
=C/4 R50
} P49 resistors, and hence nearly equal but
rE
relatively " inverted " outputs are
EFSO obtained. Fig. 3 shows the two pulse
VIO trains.
25
C20 It will be seen that while the output
from the anode is the usual line -
frequency sync pulse train, that from
1 bL4 the cathode is sufficiently nearly the
/3
1 46
o
óö T2 C
frame -frequency sync pulse train. The
way in which the two are combined
C
/9 24
is explained in the next paragraph: here
L2
47
C2/ C23 it may be noted that if the two outputs
C25 L
5 HT are switched on and off at the required
PF OUTPUT
intervals a train of pulses similar to
that of Fig. 4 can be achieved.
Stase (q)
P32
'd)
931
4 SYNC.-
PF+
P39
R4/
HT+ 200
P42
V
The Gating Stage (Stage d)
In order to combine the pulse trains,
the gating stage of Fig. 5 is employed.
Two valves, VI and V2, have a com-
WO
9F C18
The line pulse
C+ E EFSO
mon anode load R.
SYN. trains from the phase -splitter arc passed
C/6

«. cu VALVE AND SWITCH DATA

R 1735 P43 t
V// FASO
34 R
38 \'I -6SN7 V7-EF50
V2 -6,15 V8-EF50
V9 -ECC81
V3 -6SN7
R33 -C12 937 P40
'. V4-6SN7 \V 10 -EF50
11 -EA50
HT-
V5 -X65 SI -2-pole, 2 -way
: V61X65. S2 -1 -pote, 2 -way
'heorelical. circuit of the Pattern Generator.

www.americanradiohistory.com
 532 PRACTICAL TELEVISION June, 1958
to the grids of VI and V2; and if this were all. inclusion of Rg. When VI is cut off and CI
the output developed across R would be zero. 1f is charging through R3, no grid current flows and
to a second grid of VI and V2 antiphase switch- therefore Rg has no effect. When VI cuts on
ing pulses are applied:
(a) When V1 is cut off by a negative switching
pulse only V2 is operating and positive -going
pulses appear across R. a)
(b) When V2 is cut off by a negative switching
pulse --simultaneously with the positive pulse on
V I -only negative -going pulses appear across R.-
By properly arranging the switching pulses. the
desired sync pulse train. consisting of both line (b)r
and frame sync pulses. is obtained.
It should be noted that in order to simulate the Fig. 3.- Outputs from phase -splitter : (a) cathode ;
(h) anode.
correct sync pulses for frame fly -back. several
" lines " have to be suppressed. and a frame sync and CI is discharging Rg is placed in the
pulse, consisting of several broad negative -going discharge path. altering the time -constant and
pulses. has to be substituted. This entails the achieving the desired result. It will now be seen
correct output from stage (c). the " frame " multi - that the frame pulse from the stage (c) multi-
vibrator. For adjustment. the mark -to -space ratio vibrator can not only be made to give the correct
number of the line sync pulses in
the output of the gating stage.
but also the correct number and
duration of frame sync pulses.
For the gating stage itself,
simple valves were not found
suitable. The best choice among
them was the 6F33 pentode. whose
suppressor characteristics were
tried in use. The ex- Government
VR116 also gave passable results.
In both cases. however. the
internal capacity between anode
and suppressor grid rather spoilt
i

the line pulse waveform. and in

I he completed unit.
addition a considerable difference
existed between " frame gating
is variable. Reference to Fig. 2 shows that the and " line sync " pulse voltages needed. The
fly-back time of the multivibrator depends on the simplest solution was found in the use of the
D.C. resistance of the grid- cathode path of VI triode -hexode X65. The oscillator section of this
and R6 in series. To alter the fly -back time this valve was only used indirectly. by virtue of the
resistance has to be changed. It cannot be done
by altering R6 because this materially affects the TABLE ONE
frequency of operation Stage (a) Count- Stage (b) Count- Stage (e)
of the stage. Instead.
therefore. a variable
resistor is introduced.
running
at -pps.
down
ratio at-
running
pps.
down
ratio at -
running
pps.
as shown in 9.600 12 800 16
Fig. 6. Here 9,800 14 ; 700 14
Ou t
9,750 13 750 15
part of Fig. V2 0,200 12 850 17 50 pps.
V/
2 is repro - 0,400 13 800 16
d u c e d.
showing the - --- 0,500
0.500
14
15
750
700
15
14
1,050 13 850 17
1,200 14 800 16
1,200 16 700 14

direct connection between the oscillator grid and

nno the third grid of the hexode section. To avoid
Line pu/ses
capacitance effects, and the introduction of hum,
the oscillator anode and grid are connected
nnr together.
puises On. The Vertical Bar Generator
The vertical bar generator, stage (f). has to
generate radio frequencies between 100 kc /s and
WFreme pulses bn
A Frame flyback time
R
3 Mc /s. For this purpose a rather unusual
oscillator is employed, shown in detail in Fig. 7.
Fig. 5.- Simplified circuit of gating stage. (To be continued)

www.americanradiohistory.com