Chasing noise on a rooftop SDR setup
The noise floor on 137 MHz was 9 dB higher than it should have been. Finding out why took three weekends.
The method was dull but effective: record a five-minute spectrum, switch one thing off, record again, write the difference down.
| Change | Noise floor |
|---|---|
| Baseline | −92 dBm |
| Pi powered from a linear supply | −96 dBm |
| Ferrites on USB and power leads | −99 dBm |
| Neighbour's LED floodlight off | −101 dBm |
The last row is not something I can control, so the station now simply does better after midnight.