Other aircraft were decoded during this period. It’s just not fast enough reducing the gain.
Mar 16 16:42:46 dump1090-fa[1556906]: rtlsdr: tuner gain set to 49.6 dB (gain step 28)
Mar 16 16:42:47 dump1090-fa[1556906]: rtlsdr: tuner gain set to 48.0 dB (gain step 27)
Mar 16 16:42:48 dump1090-fa[1556906]: rtlsdr: tuner gain set to 44.5 dB (gain step 26)
Mar 16 16:42:49 dump1090-fa[1556906]: rtlsdr: tuner gain set to 43.9 dB (gain step 25)
In this case you need another receiver (antenna+filter+dongle+RPi+dump1090-fa) for nearby helicopters. In this receiver, set gain to low enough value manually (by trial and error) to keep track of very close flying helicopters.
You can have 2 independent receivers on one Pi also, if you want to save cost of 2nd RPi
It’s not so important that it would justify two receivers.
Now I changed adaptive.c so that it makes two gain steps at once instead of only one gain step. That works reasonably well.
Isn’t the file adaptive.c part of source-code of dump1090-fa?
If I am right, then not many users here will be able to modify it and then recompile dump1090-fa.
After modification of source-code file adaptive.c and recompiling, the dump1090-fa is changing gain in steps of 2 (see below, from step 29 to step 27)
Mar 16 22:21:53 ubuntu-24 dump1090-fa[10372]: adaptive: changing gain from 58.6dB (step 29) to 48.0dB (step 27) because: high rate of loud undecoded messages
Is there a setting such as “adaptive_burst_loud_signal_value = -2.0” that sets the trigger point for the gain reduction? the dump1090 readme says " The “burst” adaptive gain mode listens for loud bursts of signal that were not successfully decoded as ADS-B messages".
I’d like to have the gain reduction triggered before the signal gets too loud. I’m guessing this would require 100% sampling, correct?
You’re into ‘crystal ball’ territory there.
Because you are trying to detect both weak and strong signals at the same time, it’s a real juggeling act where to set the gain.
The trouble is, you can’t predict when one transmitter will become ‘un-decodable’.
Reducing the gain unnecessarily will reduce the distant / weak signals without improving local / strong signals.
The common work-around is a ‘better’ receiver (meaning wider dynamic range).
The second option is to setup two receivers - one with high-gain, the other with low-gain.
What is needed is an SDR dongle that has better dynamic range than an RTL-SDR V3 dongle (say with a 12-bit A/D converter), but doesn’t cost $100 like the Airspy mini.
@geckoVN , I have used the two receivers (high+low gain) in the past but I’m willing to give up distant / weak signals so I can continue to track the very close aircraft (100’ - 200’ away) on the ground at the airport. @VirusPilot was getting close to a solution - was curious if there were any other single receiver options other than keeping the gain very low (~15).
My goal for using adaptive gain and adaptive burst was to prefer local traffic (pattern and ground) for our traffic display when we are on duty as controllers on the tower. With my settings in my fork this works quite nicely, I recently doubled the gain increase/decrease step size which also helps: https://github.com/flightaware/dump1090/compare/master…VirusPilot:dump1090:master
@VirusPilot , what’s the process to switch from dump1090-fa to your fork of dump1090? I’m not an experienced programmer so a list of steps would be helpful.