dump1090 – Gain vs. Performance

I took the opportunity for some experiments with gain settings in dump1090-fa while testing a new built antenna for the ADSB receiver. I compared my previously built QFH antenna to a G7RGQ omni directional antenna that I built last week. Second object of test was a 10m active USB cable (from Amazon).

1090MHz QFH on the left. G7RGQ omni on the right.

In fact, the difference between these two setups are, the QFH is connected to the RTL-SDR recevier by a 6m long coax cable with roughly 6.5dB attenuation @ 1090MHz (bad solution, to be changed). While the G7RGQ was attached to a temporary bamboos pole, the RTL-SDR receiver directly connected to the antenna feed point and connected to the Raspberry Pi by an 10m active USB cable.

I started with 48dB gain setting that was used for the QFH before. Then stepping down to 25.4dB gain letting the system run for a while between changes.

Gain changes over time. Red marks on the x-axis are gain changes.

The following gain setting were applied at time, so you can compare with the above graphs:

Time Real Gain RTL-SDR [dB] Gain setting dump1090-fa
13:01 48.0 48
13:39 38.6 38
14:09 32.8 32
14:48 29.7 30
15:26 25.4 25
16:00 29.7 30

My conclusion

  • A 10m active USB cable works excellent to replace the 6m coax, hence the additional attenuation.
  • Highest gain is not giving best results. Here is why:
  • Higher the gain -> higher the noise -> lower the signal-to-noise ratio (SNR) -> lower the sensitivity (see signal level graph).
  • Lower the sensitivity -> lower the range.
  • Higher the gain -> higher the signal mean level -> lower the dynamic range of the RTL-SDR receiver -> lower the sensitivity.
  • Again, lower the sensitivity -> lower the range.
  • For maximum sensitivity, ergo range, the gain must set to a level where noise is lowest, mean level is not reaching maximum and peak level is above -3dBFS.

To archive the latter, I found that a good indication is the peak level in the signal graph. As long as the peak level is a flat line the gain is to high, also when mean level is reaching the -3dBFS limit. In this case lower the gain until peak level is changing dynamically but still being above the -3dBFS limit. Perform gain tests during a period with high aircraft traffic so you can expect a stable incoming signal. At night you will usually see the peak level dropping almost to noise floor when not aircraft is in range. But that´s normal.

This post is applicable for the ADSB receiver image from https://www.adsbreceiver.net/ where you have the web portal with performance graphs.

Leave a Reply