1
FET and 14 parts only!
INTRODUCTION
Some years ago a beautifully simple circuit for a 40m CW transceiver by W2UW
was published in GQRP SPRAT. This
inspired me to have a go at an even simpler version for 80m using just 14 parts
apart from the key and crystal earpiece. You would be hard pressed to get the parts
count lower and achieve solid and reliable local (and DX?) communications.
Basically the circuit shares a single
MPF102 FET between a very simple regenerative RX and a QRP crystal
oscillator. RX and TX switching consists
of a 4 pole 2 way switch which changes over the antenna and all 3 connections
to the shared FET. If you want a simpler TX-RX switch you could use two FETs
instead. Although a little PCB would be
neater, I just used ugly construction on a piece of single sided copper clad
board. The FET is mounted on the back of the switch. The small aluminium box
was obtained from Maplins.
RECEIVER
Feedback is provided by the tap on the main
toroid winding. You may need to optimise this but start with the tap point as
shown. A single turn antenna winding is over-wound on the main coil.
Regeneration is set by C3. Once regeneration is set to just oscillate it won't
need readjustment unless you change the power supply volts or antenna. The band
tuned is set by choosing values of C4 so that the variable capacitor
comfortably tunes across the 80m CW band. In my case the RX tunes 3.48 to
3.62MHz with a few hundred pFs of fixed capacitance at C4. The receiver is
quite remarkable for something this simple: mine measures around -100dBm
sensitivity for a usable signal level in the earpiece and picks up plenty of
CW and SSB stations.
NOTE: Please note that this simple
receiver will radiate a small carrier on the frequency to which
it is tuned (because it is set to be just in oscillation), so avoid
using this when operating in close physical proximity to other 80m
CW stations.
TRANSMITTER
On TX, I adjusted and fixed the value of C1
so that the oscillator started reliably with a 50 ohm load. In my case 68pF was
optimum. Observing on a scope, R3 was adjusted from zero until maximum RF
output was obtained, in my case about 18mW. Most people will use an ATU between
the output and antenna, but a low pass filter is recommended if not. By the
way, my antenna is just a 15m random wire with a central heating radiator as
ground.
RESULTS
It is early days yet but two solid local
two-way QSOs at 599 and one two-way QSO at 18kms with M1KTA suggest very many more miles are possible with a quiet band and a
clear frequency. Indeed G3XIZ has worked 3 countries (best DX 450kms) with his version and has been heard as far as Switzerland on his. See his neat layout in the photo of his version.
What amazes me is how
well this little circuit works: the RX is stable enough and not at all fiddly.
It is, all things considered, remarkably
sensitive and picks up plenty of European SSB and CW signals at night. On transmit, the signal reports have been good. Even buying all parts new it will not cost
you much more than a fish and chip supper. For me, this sort of circuit is the ultimate
QRP challenge. Making and using it has been real fun. Next I want to try
the same idea on 28MHz. Check my website
and blog pages for latest developments and news of further contacts. | 
