SETI station - phase 3
last modified - 19082k
We use this station for a number of activities :
The following diagram shows all the components of the radiotelescope :
I acquired a 4m dish complete with mounting post. This came as a selection of petals and required assembly. The assembled dish is shown above.
I have positioned the dish to point due South, so that the vertical adjustment takes the dish from due South to overhead. I hope to eventually motorise the vertical movement.
I have two dish feeds
A home-made waveguide transition with 1/4wave probes at 90 degrees.
A helical dish feed my by the 'Olde Antenna Lab' of Denver
In sun transit tests, the helical dish feed out-performed the other dish feed by a couple of dBs.
This amplifier based on a design by DJ9BV in the DUBUS magazine. I am using a 4919 GaaSFET into an MMIC mounted just behind the dish feed. The GaaSFET gives a good noise figure ( measured 0.5dB at 1420MHz ) and the MMIC should overcome the cable loss in the 10m of RG213 from the dish to the Rack.
This is a communications receiver that covers up to 2GHz. The front-end technology is quite old, so we must use an external Low Noise Amplifier
This is an i386 clone running RedHat Linux 6.0. The processor is a Cyrix P166. The soundcard is soundblaster AWE32.
This is another i386 clone running RedHat 6.2. The motherboard is an ABIT BP6, with one Intel Celeron 566 and one Celeron 533, both running with a clock speed of 533MHz.
I am using a modified version of the specgram FFT program at present, however, I hope to replace this with the Setisearch program that I am currently developing. The program takes in audio from a receiver using a PC soundcard ( SoundBlaster 16 ). The audio bandwidth is approximately 20KHz, the soundcard sample rate is 44KHz giving a nyquist point of 22KHz. The software compensates for the input level using a sort or AGC algorithm with the soundcard's mixer. It then performs an FFT on the audio to convert it from the time domain to the frequency domain. The result is that each frequency from DC to 22KHz will have its own 10Hz bin. We look for interesting signals by scanning the bins to see if any of then have more than the average amount of signal in them The AGC level is used for Radio Astronomy work. It is written to a logfile every 2 minutes and this logfile can be used to plot graphs of sky noise as shown in the results section. If a signal is detected, the PC's parallel port is used to switch relays causing lights to flash and a siren to sound. A logfile is written and the tape recorder is started.
The graph shows the recovered audio power during a number of sun transit tests. The sun can be seen at 12:00 ( the clock is in GMT ) as the *huge* peak. In theory I should be able to work out my system noise temperature from this graph and the suns figures for that day. The other peaks are more interesting. There are a number of peaks at 02:00 ish. If these were Deep Sky Objects, then I would expect them to be both gaussian shaped and moving by 4 minutes per day. They are however saw-tooth and moving by 7 minutes per day. There are many other 'interesting' peaks. For a more complete set of data, including the source for this graph, look at the data below. It grows in size every day.
Update ( 05-07-2k ) I stopped the receiver frequency scanning and the 'sawtooth' responses have now become gaussian peaks as expected. This indicates that the peaks have a frequency response. The data file below the graph has been updated.
The graph shows a number of sun transits.
Here is the source data for the above graph
Here is my original page..