I have recently changed my setup, moving my Nooelec Sawbird lna up below a new Paradar antenna with HDF400 coax coming down to an Airspy mini on a Pi5. The lna is housed in a waterproof junction box and all connections from the antenna down are bound in amalgamating tape so I’m pretty sure there’s been no ingress of rain water.I’m getting good results in the morning until about 10am when plane count, message rate and range all take a dive to about half of what they’ve been, this continues right through to the early hours when things pick up again. The weather here has been very settled for the period I’ve had this setup, high pressure dominating with no rain, sunny days and cold nights. I would say performance improves when the temperature drops into the thirties overnight and drops off when it goes up into the forties around mid morning. As the housed lna is attached to the antenna mast on the chimney it is not accessible to me, my days of scrambling up the roof are over even if it is not too high being a bungalow.
I’ve played around with the gain setting to no avail leaving it at 21 max for the Airspy. Any suggestions other than bringing it down and starting again.
Check for RF interference. It sounds like some equipment or machinery is starting at 10 am that is causing your issue.
Graphs1090 might help reveal more info.
rbrewis
What model Nooelec Sawbird and how are you powering it?
There seems to be something seriously wrong with your dynamic range on the signal level graph.
You only have about 12dB it should be nearer 40dB.
It may be your gain is much too high or possibly you have a large amount of noise (all the time).
Here are my graphs for comparison.
Airspy mini with RTL-SDR BLOG 1090 MHZ ADS-B LNA and Airspy decoder.
I have the Sawbird+ -ADSB module powered by the Airspy’s bias tee.
rbrewis
Sawbrid + ADSB datasheet shows 140mA current draw the Airspy Mini is only rated to supply 50mA.
Try an external bias-tee.
Also, that LNA is +34dB, so a gain setting of 21 is likely much too high as evidenced by your weakest signal being way up at -13db, I would shoot for something in the -36 range.
I believe the gain setting step is 3 dB approximately so I would give a setting of 10 a try.
In this case you may find it useful to edit /etc/default/graphs1090 and set all the graphs to wide. That way there will be more space for them and the vertical axes will also align, which is useful for correlating events, such as temperature.
all_large=yes
Then sudo systemctl restart collectd graphs1090 and reload the page in your web browser, it will take a few minutes to get back to normal operation.
I’ll give that a go thanks.
Thanks, I have a bias tee injector so I’ll connect that and play around with the gain again.