From Webshed
Jump to: navigation, search

G7UVW QRSS 30 Meter Beacon 10.140 MHz

After seeing lots of information and designs and grabbed QRSS signals I've decided to have ago at getting a beacon on the air for some tests. The 10.140 MHz xtal was supplied by Chris G8OCV and the xtal oven was a rally find a couple of years ago.

The oscillator is pretty much a straight clone of Hans Summers' junk box QRSS beacon but with a smaller inductor in series with the xtal and a proper AM radio varicap (taken from a dead radio) instead of an LED for frequency shifting.

The whole thing is built on a small scrap of veroboard so that it will fit inside the aluminum box of my xtal oven. I decided to oven the whole oscillator instead of just the xtal for better frequency stability.

Oscillator - click for larger Oscillator - click for larger

The white stuff at the bottom of the oven block is white-tak just to electrically insulate the board from the metal block and hold the board fairly solidly.

The xtal oven I'm using is an old Pye ovened frequency standard. I think I paid about £3 for it at a radio rally. It wasn't a 10MHz standard as I'd hoped and a straight xtal swap didn't work with the existing oscillator, so it was replaced with a 10MHz rubidium standard from ebay. After sitting unused for some time on the shelf, it has been re-purposed for this project.

The oven itself is just a box fitted with two BD132 transistors as heaters. There is a simple 741 based circuit that compares a thermistor and a potentiometer and adjusts the transistor current to heat the box and thermistor. Temperature control is pretty good after about a 20min warm up.

First light RF

Fed with 8 volts from the internal supply of the oven, the oscillator starts up at 10.140240 MHz and drifts downwards to 10.140105.3 as the oven heats up. This final frequency holds stable to within 0.1 Hz for at least 3 hours.
Oscillator signal - click for larger

The signal from the oscillator isn't a pure sinewave, so there is some need for bandpass filtering to remove harmonics before this goes on air.

Update 10/03/2009

I removed the old buffer stage from the Pye frequency standard and replaced it with a single 2N2222 buffer. The Pye stage was probably better designed, but was also somewhat filtered thus reducing the power out at 10.140 MHz. A simple low pass filter has been fitted to the output and I now get something that looks closer to a sine wave.

Measured power out into a 50Ohm load is 800uW. Initial start up frequency is around 6.8 MHz !, rapidly rising at the oven heats up to around 10.3 MHz before settling down at a steady 10.140031 MHz. No idea why the oscillator starts so low, but you can watch it increase in frequency by ~100KHz per second until it jumps to 10.139xxx MHz and then settles.

Update 11/03/2009

The warm up and settling of the oscillator. The 6MHz start up is omitted to allow the scale to show deviations around 10.140 MHz Warmup.png

Update 11/03/2009

The low start up frequency problem seems to have cured itself, the oscillator now starts around 10.139 MHz and shortly settles at 10.1400035 MHz.

The oscillator feeds a small PA to provide about 2.8W into a 50ohm load (with PA supply at 12v). The output of the PA is bandpass filtered with a couple of re-tuned 10.7 MHz TOKO IF cans which also limit the power output to 13mW into 50ohm (PA supply at 8v). Photographs to follow. On air test to happen in next day or two.

Update 21/04/2009

The on-air test occurred, but I'm not 100% sure my signal was seen on the QRSS grabbers. A simple PA stage using a 74HC244 IC has been designed to increase the power output of the beacon.

Update 23/04/2009

Success! Had a play with the filters on the output last night with the help of Chris G8OCV. Now putting about 200mW into a known crap antenna. Spotted signal on I2NDT's grabber in Bergamo Italy.


The sawtooth pattern in the middle of the screen with some faint Morse either side is my signal.