Hi everyone,
i am playing around with some older setup (RPi 3B+, blue FA-Feeder, Uputronics LNA)
Installed is readsb and i have intermittent performance drops which are self healing:
The same time the gain setting (autogain) drops as well:
And it shows a similar pattern to all other values (aircraft, message rate, signal level etc.) except range.
What are your thoughts on that?
Thanks
Do you have adaptive burst in dump1090-fa enabled ? Do the moments coincide with a close proximity aircraft at that moment? I get this behaviour when I get aircraft very close to my location with adaptive burst active.
See my original post:
For testing i’ve switched the gain to a fixed mode and the problems aren’t visible since then
What is “close”? I get aircraft over my location, but they are either at 30-40.000 feet or they are passing my location with a 10-15 km distance.
I don’t use readsb so I don’t know what the decoding mechanism underneath it is. I regard readsb at the same level as piaware so it might have a separate decoder like dump1090-fa 
When I say close mean within 2 NM of your location
readsb is the decoder which provides the same functionality as dump1090-fa with some additional features. Has nothing to do with piaware itself.
It is a fork of an earlier version of dump1090-fa.
Nah, they are not that close normally. But i will watch it. Sometimes a smaller aircraft or a heli is passing by.
It’s not a problem it’s the intended behaviour as tom correctly inferred.
Close by low flying traffic is generally much more interesting than traffic that’s far away.
You can set a minimum gain it won’t go below as stated in the docs.
Though of course that will mean you likely lose reception for aircraft that fly really close to your antenna.
Are the times random or similar each day?
Is it only during the day or through the night as well?
Try heating a cup of water in the microwave and see if you get a dropout.
I guess the rebound could be a bit quicker.
Can you try installing the current version, see if changes the duration of the low gain setting.
In general this is just a tradeoff. If you really care for the data from the aircraft that passes low / close by, you have to reduce gain.
Thanks for your replies.
I have it installed the day before yesterday, so it should be the most current version.
The dropouts are random. yesterday more, today only once at 12:40, later nothing.
Gain usually hovers around 20-23, during the given time it was low at round 4 then went up again to the normal values.
Some smaller dropouts occured as well, but it went never lower than 15
I’ll keep it monitored.
That would require a microwave which we do not have
Me writing:
I guess the rebound could be a bit quicker.
Meant: Let’s change that code a bit.
In other words no your version is not recent enough.
Sorry, was a misunderstanding.
installed now is
readsb version: 3.14.1678 wiedehopf git: ecdf45f (committed: Tue May 27 16:48:33 2025 0000)
correct one?
Interestingly the tar1090 shows V3.14.1736 (this didn’t change during upgrade)
Edit: After some minutes of keeping it running it seem to be that the gain is now more stable, no jumps every few minutes as before (updated before 21:30)
That’s the tar1090 version, not the readsb version.
Those version numbers are unrelated.
It shouldn’t change that behaviour … maybe i broke something.
Possibly just coincidence.
Real test will be some low flying aircraft close to your receiver.
You could use /?replay on one of the aggregation sites to determine what planes / helicopters are causing the gain to dip in the graphs.
Good morning.
gain is still changing, but only every few hours now, not every few minutes as before:
I am checking out once low flying aircraft are around.
Right now there is an A 320 flying low (6000ft) with a distance of 4.6NM
No impact to the gain or other values so far.
I am waiting for smaller aircraft or helis
I use piaware/skyaware to track planes and when one gets close and overloads the track appears as
Note the dotted trail when the RSSI exceeds about -1.6
That is more responsive than the Graphs1090 trace.
If the plane is close enough and loud enough then most aircraft in view will have their tracks interupted at that time.
I mainly use fixed gain and no Burst.
S.
That’s not how it works on the readsb/tar1090 combination. In readsb the line only cuts if out of range. At least in my installation. I do not see an “overload” in the line.
That’s what the automatic gain adjustment is for …
Put a manual gain that’s high enough and you’ll see the effect pretty easily.
Typically the frontend should recover quickly enough though not to affect other aircraft too much.
Don’t think i’ve seen the blackout @SweetPea11 described.
After all ADS-B is only very short bursts and each aircraft only sends at most 40 per second if the interrogration is crazy like in Europe and more typically 15 per second at most.
I had one drop at around 09:05 CEST, but if i am selecting the timeframe, I only get an HTTP 404 “no data for the given time”.
