You need the dark-blue (1090 Mhz only) filter
Announcing the new 1090 MHz only bandpass filter!
2 Likes
In my experience that dark blue 1090-only filter has benefits even in North America. Assuming you aren’t interested in UAT tracking, that is.
1 Like
Could I suggest a flexible coupling somewhere in that collection?
If half of it gets cantilevered off the table, you run the risk of snapping one of the SMA joints.
3 Likes
That’s not a bad idea! It’s secured pretty well and in an out-of-the-way location, but I’m certainly going to look into the flexible couplings. Do you have a link or pic to share what you had in mind?
EDIT: I ordered several of these: Data Alliance SMA Male Cables
That’s perfect. Doesn’t need to be long, you just want to eliminate the leverage if something goes wrong.
Can anyone tell me what the anomaly is that occurs at about 1930? It’s not the usual fluctuations that happen at night and I’ve never seen this happen at any time of day before. The light blue line and green area suddenly jump up for seemingly no reason. 
I did until I installed the Uputronics preamp/filter. After the addition of the preamp, the auto gain script seemed to throw my signal level out of whack. My understanding is you want the type of signal spread that I see from 1100 to 1900, and the auto gain script didn’t allow that.
Yep, I find 32 works the best the majority of the time.
I still can’t explain this sudden anomaly though. It started again last night around 1230. I have never seen this before until the last 24 hours.
Are your wired connections in order? I have seen similar things in the past and it turns out that the coaxial cable connectors were moist on the inside. Once I cleaned them and reconnected it was fine.
Good point since I recently updated some hardware. I added short SMA leads to allow for some flexibility in case anything is bumped (my entire setup is indoors, so I doubt moisture is an issue).
Before:
After:
However, even after all connections were checked and re-tightened, the signal anomaly is continuing:
are you sure the auto gain script is totally gone? the gain change looks to be approximately 3dB each time.
Restarting dump1090-fa will cause a break like that.
I ran the uninstall script here: https://github.com/wiedehopf/adsb-scripts/wiki/Automatic-gain-optimization-for-dump1090-fa and I’m also no longer able to manually run the script, so I assume it is uninstalled?
Also, my gain setting is not changing. It has been at 32 for several weeks now, so the gain setting itself is not responsible for the signal strength anomaly.
The break is because I powered the Pi down while tightening the connections.
Look through /var/log/syslog (for today), syslog.1 (yesterday), syslog.2.gz (day before), etc, for any clues at the times the gain changed.
Take a look at /var/log/piaware.log around those times to see if anything happened around then. Eg if I restart dump1090-fa I see things like:
Mar 14 20:35:59 piaware piaware[7038]: lost connection to dump1090-fa via faup1090
Mar 14 20:35:59 piaware piaware[7038]: faup1090 exited normally
Mar 14 20:35:59 piaware piaware[7038]: reconnecting to dump1090-fa
That should help narrow it down to a software issue or an external hardware issue.
1 Like
Thanks. I’ll look through the logs you mentioned tonight. Just for the record, the gain setting has not been changed. I’ve had it manually set to 32 this whole time.
Actually the last bit in the red box doesn’t show any abnormality …
I think this has happened to me once too in 3.7.2. I remember seeing a step up and, later on, down after a few hours and I hadn’t done anything nor had any gain related stuff installed. I put it down to “a glitch”. Perhaps it’s a bug in dump1090-fa that’s triggered by some edge case.
Forgot to mention, look at the other graphs at the same times. Anything interesting showing on those too? Note you can make all the graphs the same width which helps analysis.
Have you considered an external source of noise? Maybe something that’s only on when people are around.