Sensitivity Comparison between A040RCV and FFTDSP for DSN reception
Objective
To investigate the relative minimum detectable signal of two commonly used audio waterfall programs, AO40RCV and FFTDSP.
Experiment
A low level signal from a signal generator was coupled into the IF output from the DSN downconverter, at approx 165MHz. The output from the signal generator was reduced until the signal was barely detectable on the waterfall and the power level noted.
Some experiments were done with FFTDSP to find the best settings. Two configurations were used - integration set to 1 and no smoothing, which refreshed the screen at a similar rate to AO40RCV, and an integration factor of 10 together with cosine smoothing to try to improve the sensitivity (at the expense of refresh rate). Refresh rate is important as too slow a response means that it is difficult to optimise the signal on screen, for example when peaking a dish.
Results
AO40RCV and FFTDSP with no smoothing and integration of 1 both detected a signal down to -91dBm input from the signal generator. Signals shown below. The signal is marginally easier to see with FFTDSP.
FFTDSP with -91dBm in AO40RCV with -91dBm in
If you are prepared to wait, increasing the integration factor (to 10 in these tests) and using smoothing, then FFTDSP can detect weaker signals, as noted above at the expense of screen refresh time. Here is FFTDSP with a signal 5dB weaker, which was not visible on AO40RCV, or FFTDSP with no smoothing and integration of 1.
FFTDSP with -96dBm in, integration =10, smoothing=cosine
Conclusion
For routine use, AO40RCV and FFTDSP with no smoothing and integration of 1 are able to detect weak signals at comparable S/N values. My preference is AO40RCV, as it has the advantage of having a useful spectrum analyser display in addition to the waterfall, making S/N measurements very easy.
back to Tech notes