AMATEUR RADIO
By GARRY CHATT, VK2YBX
A two-tone oscillator
for testing SSB transmitters
This simple two-tone test oscillator will allow
you to correctly adjust the output of an SSB
transmitter to prevent splatter. It uses just
three low cost transistors and can be quickly
lashed up on a piece of Veroboard.
As most amateurs are probably
aware, an incorrectly adjusted SSB
transmitter can be the cause of adjacent channel interference or
"splatter". This is a most undesirable situation. It not only in
conveniences other amateur operators but can also cause interference to television and radio
(a)
reception which should be avoided
at all costs.
By monitoring the output waveform of an SSB transmitter, this
problem can be detected and then
eliminated by correct adjustment of
the ALC (automatic level control)
and modulation circuits. This can
be done using a test signal from the
(b)
Fig.1: the output waveform (a) for a correctly adjusted transmitter and
the resultant spectrum analyser display (b), The IMD products are
suppressed by 30dB or so with respect to the output signal.
00
00
Fig.2: the effects of an overdriven output stage. The output waveform
is flattened and the IMD products now cause splatter.
46
SILICON CHIP
oscillator described here. It provides two test tones: one at 500Hz
and the other at 2.4kHz.
Spectrum analyser
Perhaps the best way of monitoring the output of any transmitter is
to use a spectrum analyser and
sample the output at a convenient
level.
Fig.1 shows the output waveform
of a correctly adjusted SSB
transmitter. As can be seen, the
two test tones are clearly visible
and the intermodulation distortion
(IMD) products are suppressed by
30dB or so with respect to the output signal. Hence there is virtually
no adjacent channel interference.
The corresponding modulation
display is sinusoidal, showing no
sign of an overdriven modulator.
While this is not a foolproof relationship, (ie, there can be other
causes of high IMD products in the
transmitter output), Fig.2 clearly
shows the effect of an overdriven
output stage. The modulation
waveform shows considerable flat
topping and the result on the output
can be seen on the corresponding
spectrum analyser display. The frequency spread of the transmitter
IMD products is sure to cause
"splatter" on adjacent frequencies.
Of course that is all very well for
those amateurs fortunate enough to
have access to a spectrum analyser. However, if you at least have or
can gain access to an oscilloscope,
monitoring the modulation output
waveform is a valuable and simple
method of predicting the purity of
the transmitter output signal.
.----------------------0+10-12v
CS
R6
PARTS LIST
+
4.7J:
R1
47k
R9
470k
1 piece of Veroboard
1 plastic case
3 BC549 NPN transistors
2400Hz
100k
C1
C4
.0082
.0018
....__ _ _ _ _ _ _ _ _ _ _ _ _ _ _....._.....__ _ _ _ _ _ _--nGND
Fig.3: the circuit for the two-tone test oscillator. Ql & Q2 form the 500Hz and
2400Hz oscillator stages. Their outputs are mixed in VR3 and fed to Q3 which
functions as an emitter follower. VR4 allows the output to be adjusted to a
suitable level for the transmitter.
CRO checks
A simple and practical method of
checking the linearity of an SSB
transmitter is to inject two harmonically unrelated frequencies into the microphone (or microphone
socket direct) and observe the output on an oscilloscope fed with an
RF probe and detector. The two
tone oscillator has constant output
which makes transmitter adjustment somewhat easier than
repeatedly whistling or screeching
into the microphone!
If an oscilloscope of sufficient
bandwidth (ie, higher than the
transmitter frequency) is available,
the RF output signal can be observed directly. If the modulator/
transmitter has good linearity, the
modulation pattern will be as close
to a pure sine wave as possible
(Fig. t).
As distortion increases, so do the
spurious transmitter products
(Fig.2). So it is prudent to operate
an SSB transmitter under the
cleanest modulation conditions
possible.
In practice, it is wise to operate
the transmitter at voice levels
slightly lower than the level achieved using a two tone input to ensure
that the modulator is not overdriven.
Circuit details
The two tone "oscillator" described here actually contains two
separate oscillators, one operating
at the nominal lower audio frequency limit likely to be encountered in
actual use, and the other operating
at the highest nominal audio frequency. Even though most amateurs
regard the "speech" limits of most
audio circuits as 300Hz to 3000Hz,
in practice 500Hz and 2400Hz are
the more likely limits.
Fig.3 shows the circuit diagram,
which is very simple. There are two
separate oscillators, involving Qt
and Q2, which are almost identical
Sallen-Key active filters, with a
high level of feedback to induce
oscillation. Some readers will recognise them as conventional transistor phase shift oscillators. Their
outputs typically have quite a low
level of distortion.
Transistors Qt and Q2 oscillate
at 2.4kHz and 500Hz respectively.
Both audio outputs are fed to VR3
which is a tMO trimpot. This serves
as a crude mixer while providing
sufficient isolation to ensure that
the two oscillators do not "pull"
each other. Trimpots VRt and VR2
are used to trim the frequency of
each oscillator.
The output from VR3 is fed to Q3,
an emitter follower, which provides
an output with an impedance of
around tkO via a tkO pot (VR4). The
---------o+10-12V
SQ
SPEAKER
Fig.4: this optional power amplifier
stage can be used to drive a
loudspeaker. It's based on a single IC.
Capacitors
1 1 OµF 16VW electrolytic
1 4. 7 µF 16VW electrolytic
3 .0082µF metallised polyester
3 .0018µF metallised polyester
Resistors (0.25W, 5%)
1 470k0
6 1 OOkO
2 47k0
1 1 kO
1 1 MO trimpot
2 1 kO trimpots
1 1 kO pot.
Optional power amplifier
1 LM386 audio amplifier IC
1 250µF 16VW electrolytic
1 1 OOµF 16VW electrolytic
1 1 OµF 16VW electrolytic
1 150 0.5W resistor
output is capacitively coupled and
could be fed directly into the
transmitter microphone socket. The
output level can be varied using
VR4 to a level suitable for most
transmitters.
A further addition can be made
to the circuit to enable the
oscillator to drive a speaker, which
can then be held close to the
transmitter microphone if direct
connection to the transmitter is not
desired. Of course, the frequency
response of the speaker must be
taken into account. A speaker having sufficiently wide bandwidth is
required so that both tones are
presented to the microphone at
much the same level.
In practice, the use of a communications type external speak-er
is acceptable, even though the efficiency of the actual speaker may be
less at 500Hz than it is at 2400Hz.
You will find that the speaker housing helps compensate the lower frequency level.
The power amplifier used in our
circuit is the LM386, which is connected using a minimum number of
components to provide a gain of 20.
The entire unit could be easily built
into a "jiffy" box on Veroboard and
could even be powered by an internally mounted 9 volt battery.
~
NOVEMBER 1989
47
