Between August and October 2002,
several hundred UK radio amateurs applied for permits allowing experimentation on 5 MHz
(60 m). Most applicants were successful in
gaining the special authorisation (or 'Notice of Variation') required to carry out
experiments on the five 3kHz-wide channels made available under these arrangements.
I've had an interest in
building a simple VXO transmitter for many years, but was unsure that the limited tuning
range usually associated with VXOs would be useful on the busy HF bands. However,
when the channelised frequencies became available on 5 MHz, I thought that this would be
an ideal application for a VXO TX/Receive Converter project to be used alongside my
Elecraft K2 transceiver.
While much of the design is based on
136kHz QRP transmitter, described in the Autumn 2001 issue of SPRAT (SPRAT Nr. 108).
I am grateful to G-QRP Club member Dave Sergeant G3YMC (Bracknell, Berkshire) for
his ideas and advice during the design of the transmitter.
Those not having access to 5 MHz will find that all the circuit elements can
be adapted for similar projects on other HF bands.
As things turned out, my first QSO
using this TX design was with G3YMC - who was also using a very similar 5 watt VXO
TX. You don't need high power to work around the UK on 5 MHz! For details
of Dave's VXO TX, click here.
The transmitter uses: a bipolar VXO;
FET buffer; 2N2222 pre-driver + BC212 keying transistor; 2SC2166 driver; and a pair of
2SC2166 transistors in parallel in the 5 watt power amplifier (PA). The receive
converter uses a single-ended FET mixer, and a 24 MHz oscillator to up-convert the 5 MHz
band to 29 MHz. With portable operation in
mind, the Super-Sixty was
shoe-horned into a small diecast box measuring only 85
Crystal Oscillator (VXO) and Buffer
For the transmitter, channels FA (centre frequency of
5.260 MHz) and FB (5.280 MHz) were chosen.
The 50 pF tuning capacitor provides a tuning range of approximately 2 kHz
above the crystal frequency.
It is easier to tune a crystal to the high side of its intended
frequency of operation, so a 5.25900 MHz crystal is used for channel 'FA' (tuning 5.259 to
5.261 MHz); and a 5.27900 MHz crystal is used for channel 'FB' (tuning 5.279 to 5.281
MHz). The VXO drives a very effective FET buffer, which presents a chirp-free signal
to the driver stages.
The pre-driver stage uses an untuned 2N2222 common emitter
amplifier, keyed by a BC212 PNP transistor. In my experience, many simple TX designs
fail to provide adequate shaping of the keying waveform, resulting in very hard keying. To
reduce the likelihood of transmitting key clicks, the keying circuit in this transmitter
provides rise and fall times of 4 ms. The result is a very pleasant T9 note.
The keyed signal is coupled via a 0.01 uF capacitor to the 2SC2166 driver
The KEYING OUT line is for use with an external sidetone
generator. In my case, I wired the KEYING OUT line to a panel-mounted phono socket, and
then made up a connecting lead to the key jack of the K2. (When using the K2 in this
way, it is wise to inhibit CW transmit on the K2 by selecting the 'test' CW mode!)
Power Amplifier (PA) and Low Pass Filter (LPF)
L1 14 turns,
26 SWG enamelled copper wire on T50-2 ring core.
L2 19 turns, 26 SWG enamelled copper
wire on T50-2 ring core.
L3 20 turns, 26 SWG enamelled copper
wire on T50-2 ring core.
L4 19 turns, 26 SWG enamelled copper
wire on T50-2 ring core.
T1 14 bifilar turns, 22 SWG enamelled copper wire on T68-2 ring core. Twist two wires together at about one twist every
15 mm. Wind 14 turns on the ring core, and
label each of the two wires at the start of the winding with the identification numbers 1
and 3. Then label the other end of each wire
with 2 and 4, respectively. (In all cases,
one pass through the centre of the ring core counts as one turn; two passes as two turns,
The VXO stage was screened from the
other stages within the diecast box using 'walls' of copper-clad board soldered to a
'floor' of the same material. Earth (or
'ground') connections were made directly to the floor, keeping all earth connections as
short as possible especially around the driver and PA stages. If this is done consistently, Mr Murphy will be
more forgiving if you end up with longer connecting wires elsewhere. Each PA transistor draws 400 mA on key down. This can be checked by measuring the voltage
across one of the 1 ohm emitter resistors: 400 mV corresponds to a current of 400 mA. Increasing the value of the 10 ohm emitter
resistor in the driver stage will reduce the drive level to the PA.
The 'RX' antenna terminal should be
connected to the 5 MHz input terminal of the receive converter. Alternatively, the 'RX' terminal may be connected
directly to a separate 5 MHz receiver. The
neat thing about using a switched receive converter is that you dont have to keep
winding back every gain control on the main receiver each time you switch to transmit. Its well worth the trouble of incorporating
a receive converter for that reason alone! When
using a separate 5 MHz receiver with the Super-Sixty,
the +12 R line could be used to generate a mute signal for the
Although I made over 30 contacts using
the first prototype without fitting heatsinks to the PA transistors, these devices will
require a small heatsink to survive extended periods of key-down. Note that the tab of the 2SC2166 is internally
connected to the collector, so be sure to use a TO220 insulating kit! Although the driver transistor does not need a
heat sink, I found it convenient to mount all three 2SC2166s on the inside wall of the
diecast box. Initially, the VXO coil was
tuned for maximum signal, and then off-tuned very slightly (to a point where adjustment of
the core had less effect upon oscillator frequency).
The top of the adjustable core ended up being about 2 mm below the top of
the coil former.
This simple two transistor design is
similar to one that I have used for many years in a homemade 80 m transceiver, and
performs very well indeed. I have included
some modest protection at the output (the pair of 1N4004 diodes), in case the associated
29 MHz rig is keyed accidentally. To align
the converter, initially set the top of all adjustable cores about 2 mm below the top of
the coil former. Then, using a weak off-air
signal, peak all coils. As with the VXO, I
suggest off-tuning the 24 MHz oscillator coil very slightly.