Common and Precious 5

From HFUnderground

Jump to: navigation, search


Beacon "Common and Precious Nr. 5"

24. March 2023: Project off-air. Unexpected demolition works have started. Power has been turned off. A new site was found. Moving the beacon will happen in April.

Before anyone will ask the obvious: yeah, I went a bit over the top with this project. I kinda wanted to build it and then, when it was done, realised that maybeee... But oh well. It is what it is now. I'm not aiming for more Beacons in this power range. Except maybe one on 2200m. But only because the radition effiency of antennas here is around 0.1-1%.

All Common and Precious Beacons are under re-construction and some in for a re-location. The 2nd generation Beacons will transmit both on CW and RTTY. The new message will carry, besides the usual informative text, also live telemetry Data such as temperature and humidity. An AM2303 sensor is used.

This new Beacon, Nr. 5, is the first to run with the new software and features. If all works fine, the hardware of all other beacons will be upgraded and fitted with the new firmware.

History Backlog

- 22. January 2023: project went on air for the first time.

- 24. March 2023: Project off-air. Unexpected demolition works have started. Power has been turned off. A new site was found. Moving the beacon will happen in April.

Summary and details

  • Frequency: 4.095.400kHz (CW and Mark Frequency). RTTY: 170Hz shift, 50 Baud, 2 Stop-bits
  • RF Power: 50W
  • Operating since: 22. January 2023
  • Antenna: 5m vertical with two load coils. One at the base and the second approx 30cm below the capacity hat. It allows for a higher feed point impedance, higher bandwith and also a good radiation of both the inductive and capacitive elemements. The Antenna is installed approx 35m above ground. Matching to 50R+j0 is realised by a simple L/C network. The antenna is a converted lightning arrester. The basepoint was isolated from the mounting post by a sleeve made from PVC tube. A blotch of silicone works as a seal and keeps rainwater from seeping into it and causing corrosion together with the RF signal. The bottom is left open to allow moisture and possible condensation to evaporate. The C elemnt is a variable capacitor with additional fixed value styroflex capacitors. The inductor is an air-core type on a PVC tube. An iron power core suitable to handle the load would be too big to fit into the matching box. An additional 10k/2W resistor across the PL socket allows for any static buildup (caused wind / snow / rain or thunderstorm) to pass to ground. The existing grounding grid of the lightning arrester system was used as a counterpoise. (Note: re-tightening and greasing and existing joints may help to remove high contact resistance due to corrosion.)
  • Power Source: 24V/120W Switch Mode Power Supply. A special industrial type designed for dynamic loads was used.
  • Oscillator: pulled 4.096kHz X-tal based on a 74LS02 oscillator circuit. Frequency shifting is done by Varicap Diodes (Schottky Diodes were used since they're cheap and exhibit a high junction capacitance).
  • Keying: Atmel AVR ATTiny 461 (4kB Flash). An ATTiny26 was used first. First it all looked tight but good. Merging all functions eventually caused the MCU to run out of program memory. It was approx 100 bytes short (a bit more with debugging code still active). A pin compatible type with 4kb was used instead (the hardware had already been build). It also offers room for further features. Fun fact: software entirely done in assembler. Every bit of it. From reading the AM2302 to math and conversion and the CW/RTTY functions. The software is written in librarsies so the code exist as independent blocks that can be use for other/future projects as well. Keeping the code flexible was a key feature in this project.
  • Environmental Sensor: AM2303 (in instrument shelter), mounted outside (facing south may cause the temnperature reading during summer to be some degrees above the actual value. But since it's mounted high above the ground it can be assumed that present wind will keep it quite on-spot.)
  • Power Stage: Class E based on IRF640 FET @ 24V. Driver: MCP1407-E/P. Using a 7-pole (tripple PI) filter. 2nd harmonic supression greater than 33dB.
  • System effiency: approx 85%

Some additional notes

  • A 9.216.000MHz crystal allows to both derivate a 50Hz (20ms IRQ) Baud RTTY signal and also regular RS232 Baud rates from it (for debugging purposes).
  • Class E introduces quite some stress to Switch Mode Power Supplies when operating in CW mode. Experience from the recent years has shown that output capacitors will dry up rapidly or become short, oftean leading to a blown up power supply. Use a power supply that can at least provide double the amps drawn by the transmitter. Adding additional cap sto the output also helps. A classic transformer based power supply works also great. Be aware that any 100/120Hz ripple will be audible as AM modulation, thus leading to a higher signal bandwith.
  • Suitable (commonly available) FET drivers up to 8..10MHz: MCP1407-E/P, IR2121 or IR2125. Do NOT use the MCPxxE10/E14 series. Despite the datasheet stating good rise- and falltimes, those will become horrible above 1MHz. Discharging of the Gate to the low (off) state will not happen. Also, add some heatsink to the IC.



This site is a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for sites to earn advertising fees by advertising and linking to Some links may be affiliate links. We may get paid if you buy something or take an action after clicking one of these.