80 Metre CW Beacon

The local Homebrew Group has been talking about Peter Parker VK3YE's Carnegie Communicator project in Lo-Key. The general design of the circuit (based on the old "Oner" CW transceiver) lead in the usual round-about way to build this CW beacon. OK, so how does a top-band AM transceiver lead to a Morse beacon? Well, inspiration is a complex thing, and with the PSK beacon stuck in phase jitter hell, a CW beacon sounded like a good project to initially test the waters with.

Beacon Prototype Lash-up

The design is very brute-force and simplistic, closely following the old Oner design. A Pierce oscillator drives a FET output stage with a keyed drain and an output filter. The output is loaded directly at 50 Ohms with no impedance transformation, limiting the output power to ~1.5 Watts from a 13.8 Volt supply when you consider the saturation voltage of the BD140 keying BJT. I tried several output devices, including IRF510 and VN10KM, but settled on a paralleled pair of 2N7000s which provided the best performance from the limited drive. A single 2N7000 will work and produce slightly more output power as the drive sees only half the capacitive load, but the device will run stinking hot and is being driven outside its safe working area. Two devices get warm to the touch and output about 200 mW less, but will probably last forever. The IRF510 on the other hand is under-driven in this circuit because of its much larger gate capacitance. The output with an IRF510 is still OK, and if its all you have it will work and give a little under a watt out for about the same drain current (~210 mA). With a little more than 1.5 Watts out, the efficiency is horrific at only about 60%. Modification of the output network to run in class-E could improve this towards 95% and the output devices would run much cooler, one 2N7000 would probably be fine. The 60 volt Vds limit of the 2N7000 gives a limited safety factor in this mode of operation however.

Transmitter Circuit Diagram

The crystal is an NTSC colourburst one from the junkbox, the output frequency is around 3.57867 MHz, but the crystal isn't oven-controlled so it will probably drift around a bit. Please send me a reception report if you hear it!

The keying is performed by an ATtiny13V. A small board holds a socket for the microcontroller a LM78L05 regulator and a decoupling capacitor. The active-high output of the keyer is used to drive the keyline through an extra 2N7000 pull-down. The microcontroller software is available here.

The "backwave" (key-up output power) is suppressed only about -46 dBc, which may prove to be insufficient. The backwave output power is of a similar order of magnitude as the total output power of some of Michael Rainey's beacons! Of course it is still very much QRP, and currently I am loading up my 40 metre end-fed dipole using 18 turns on an FT120-43 and my end-fed coupler which is just a terrible antenna for 80 metres so the actual radiated power is probably quite small. I'll probably build a loaded vertical with top-hat for this beacon if I keep it running.



title type size
Transmitter Circuit Diagram Source application/postscript 15.456 kbytes


2008-08-03: Beacon Reports
Summary of the reports received for the 80 metre beacon experiment.
2008-07-13: New Beacon Antenna
I finally rig up a vertical antenna for my 80 metre CW beacon.
2008-06-29: More Beacon Work
I put the 80 metre beacon in a box and do some more work towards an antenna for it.