I made good progress with the portable S-band ground station this week.
I took the receiver to the OZ7SAT lab to measure its performance. Using the USRP+DBSRX and no LNA we could easily detect a -132 dBm CW signal with modest FFT integration (fraction of a second) in a GNU Radio spectrum scope. Using the LNA we could go down to about -138 dBm, i.e. an improvement in SNR of 6 dB. These figures were measured at an SNR ~5 dB. This is excellent, but please note that this is not real “sensitivity” in the traditional sense because we were not demodulating or decoding the signal. We were simply integrating the spectrum for a fraction of a second to detect the presence of the signal. The measurements were done by sampling a 250 kHz wide spectrum.
The gain (30dB) and noise figure (0.5dB) of the LNA was also confirmed in the lab for the 2.2 – 2.3 GHz range. In fact, we found out that the LNA has no built-in bandpass filter so it actually works very well outside this range. It keeps good gain (>28dB) and acceptable noise figure (~1dB) between 700 MHz and 2.45 GHz.
The antenna is also finished and ready!
Below you will find a video tour of the antenna showing the simple mount that was created using an old camera tripod. As @jetforme pointed out, a deluxe version could use motorized AltAz telescope mount such as this one and allow pointing the antenna using a computer. We’ll try that later!
All we need now is to finish the detector and recording software and we wait for the Moon to come further away from the Sun. With the Sun in the field of view the noise floor increases by ~3 dB and makes the detection unreliable.