QRP Labs
30m QRSS junkbox tx: construction PDF Print E-mail
Written by Hans Summers   
Monday, 15 June 2009 22:54

Here's a 30m QRSS beacon whose main objective is low cost and simplicity. I built this project entirely from junk box parts: the only component purchased specifically for this project was the 10,140kHz crystal I bought from Peter DL6NL for 10 Euros (Thanks OM!). Everything else is from my junk box, either leftovers from previous projects, or from dismantled consumer electronics equipment. For example the transistors used are 2N3904, these came from a broken DVD player. Instead of a real varicap (varactor) diode to provide some frequency shift, I use an ordinary 5mm red LED. See my Varicap Diodes page for measurements of LED's and other common diodes used as varicaps. This project follows from my previous 30m QRSS beacon project. The circuit is based closely on that used by Colin G6AVK and Peter DL6NL (thanks for both your advice etc.).

Click photos and diagrams for the large scale images!

Oscillator circuit: The oscillator is a single 2N2904 transistor. I had to put a 3.5uH inductor in series with the crystal frequency in order to lower it onto the desired 10,140kHz frequency. The frequency is tuned by a polyvaricon tuning capacitor from an old AM radio. The 5mm red LED provides frequency shift for FSK modes.
Oscillator: I built a polystyrene box to provide some protection against rapid temperature changes, for example due to drafts. It's just bits of old packaging polystyrene held together by rubber bands.
PA circuit: The power supply is an old laptop power supply, it provides 16V. Keying is via a TIP146 PNP Darlington pair in the collector lead of the 2N2219.
PA: Here's the main driver and power amplifier board. The "chassis" is a piece of unetched plain copper PCB. the PA circuit itself is built on a piece of matrix board, having copper islands around each hole. That board is fixed by four "legs" above to the chassis. Yes, construction is simple, it's ugly, but it works nicely. You can see the ferrite EMI shield on the end of the laptop cable, fixed to the bottom left of the chassis.
Heatsink: In this view from above, you get a good chance to see the heatsink properly. The PA transistor is a junkbox 2N2219 which is a TO5 style case. I didn't have any appropriate heatsinks for that. So I made one from a piece of thin brass sheet a couple of inches square, which I wrapped around the transistor and soldered together for a tight fit. Then I cut the sheet with scissors such that I created fins which could be bent outwards. Using this arrangement the 2N2219 and heatsink barely gets warm. Ugly, yes. But it works well and still costs nothing at all.
Output: Nice clean output on my HP1741A oscilloscope. When I tidied up the circuit the output power dropped a little. I am now getting 12V peak to peak output into the 50 ohm dummy load, which is abour 360mW of power. Frequency calibration was carried out using my panel mounting frequency counter in my Polyphase HF Receiver. I measured the known 10.000000MHz OCXO in my experimental DDS generator, in order to obtain a correction factor to apply to the inaccurate (uncalibrated) frequency counter reading.
50-ohm dummy load: I built a simple dummy load specially for this project. It consists of just 20 1/4-watt 1K resistors in parallel. These resistors got extremely hot in operation and there was a discomforting smell of gently frying resistor on the air. Later I realised that I had forgotten the 33nF coupling capacitor between the PA transistor and the low pass filter. With the oscilloscope in DC mode, it was clear that there was a DC offset of 16V in addition to the RF sinewave. This would have been producing an additional 5W of power dissipation in the resistors! No wonder they got more than a little warm.
Side view: From the side you get a good idea of how I the polystyrene oscillator box and how I strapped it under the main chassis using rubber bands. Yes, you guessed what I'm going to say next: it's ugly, but it works.

By the way, the knob on the left hand side sets the amount of frequency shift that is available to the modulation input for a voltage change of zero to 16V. The potentiometer allows the range to be set anywhere from zero to 30Hz.
Lowpass filter: The lowpass output filter is a 7-element filter design from the technical pages of the GQRP club. I built it on a small fragment of PCB, which I mounted at the back of the main PA board and at right angles. It uses T37-6 toroids which I had. The wire is unwound from the large toroid you find inside an old lighting dimmer. The capacitors are slightly off-value since I did not have the correct values, and I had to make them up by parallelling smaller values.
Pattern generator circuit: A standard astable multivibrator provides a simple pattern generator by applying the output to the shift/modulation input of the oscillator to produce some frequency shift keying. The astable spends a couple of seconds in each state. I added a 6K8/100uF integrator on the output to slow down the edges of the shift a little so that hopefully the QRSS screenshots will show the transitions.
Pattern generator: This is built on a small board which plugs into the main board via a header socket from the old DVD player.

Bill Of Materials


1x 78L09 : 9V, 100mA voltage regulator
1x 2N2904 Transistor
1x 470uF capacitor
1x 33nF capacitor
2x 12pF capacitor
2x 220pF capacitor
1x 39pF capacitor
1x 0-150pF "polyvaricon" AM radio variable capacitor
1x 3.5uH inductor (T50-2 toroid with 27 turns)
1x ordinary 5mm red LED
1x 10,140kHz quartz crystal
1x 220K resistor
1x 10K resistor
1x 3K3 resistor
1x 150 ohm resistor
1x 180 ohm resistor
1x scrap of matrix board
1x piece of packaging polystyrene to make an insulated box
3x rubber bands to hold the box together and to the PA


1x TIP146 Darlington Pair
1x 2N2904 Transistor
1x 2N2219 Transistor
1x 2-inch square brass shim to make a 2N2219 heatsink 2x RFC (I used T50-6 toroid with 30 turns)
2x T37-6 toroid with 19 turns
1x T37-6 toroid with 20 turns
2x 10nF capacitor
2x 33nF capacitor
1x 100nF capacitor
1x 4.7pF capacitor
2x 270pF capacitor (I used three 90pF in parallel)
2x 560pF capacitor (I used some caps in parallel, total 590pF)
1x 4K7 potentiometer
2x 4K7 resistor
2x 10K resistor
1x 220 ohm resistor
4x 348 ohm 1/4 watt resistor
1x 6K8 resistor
1x 1K resistor
1x 22K resistor
1x 6-way header socket
1x scrap of matrix board etc


2x 2N3904 transistors
2x 470uF capacitors
1x 100uF capacitors
2x 1K resistors
2x 22K resistors
1x 6K8 resistor
1x scrap of matrix board
1x 6-way header plug

Last Updated on Tuesday, 23 June 2009 17:55
© 2009-2017 Hans Summers
Web services by DataState Solutions