I looked at some system at FR24 system. But never saw anything like a drop in the 24-hour statistics. As I said, I have a system working correctly on the North East side. Will post a scan from Both systems later on.
Well not too much you can do then except use another dongle without LNA.
I would probably switch to a external LNA anyway and prefer the rtl-sdr LNA over the uputronics one, but both might be sufficient to stop the interference.
I think about giving it a try. But first I want to identify the problem. I could image that my antenna, the receiver or some kind of electronical device is the source of the problem. Have a look an pi2 (left) and pi1 (right)
Pi1 Scan: https://1drv.ms/u/s!AhE767u2PS2lhQt-Cl1YLIWbsuY-
Pi2 Scan: https://1drv.ms/u/s!AhE767u2PS2lhQoaYyvJA8Il_79o
So itâs not a generally a problem of this specific area. Itâs my setup. Letâs swap the sticksâŚ
If you do a swap, swap the complete RPi with stick and power supply.
But something timed at pretty much 1400 every day is unlikely to be anything about the RPi, power supply or stick.
Probably just local interference coming from somewhere in the building.
Yes, right. But only one system has that much noise that the â1400 thingâ is an issue. I will change one component at a time. Changing multiple variables at one time doesnât sound like a good idee when looking for the reason.
Youâve asked why the light blue filter didnât work for me. Itâs easy. When I have installed the filter (tired on both setups) I could see significant less air craft. It wasnâT a matter of setting the gain value. When combine with old NooElec TV28T v2 it worked. But not with the FA sticks. So I uninstalled the filter.
Made a scan with swapped FA STicks. Still interferences at Pi1. Will change antenna back so small one and see what happens. Maybe itâs more comfortable to install some scan monitor software like this one here GitHub - xmikos/qspectrumanalyzer: Spectrum analyzer for multiple SDR platforms (PyQtGraph based GUI for soapy_power, hackrf_sweep, rtl_power, rx_power and other backends)
It looks from those images that both of your systems are receiving the interference but pi2 is receiving a lower level. That would appear to indicate it is external but closer to your pi1.
Swapping the 2 over should prove whether it really is external.
If that were the case, the noise âshouldâ be having the same effect on both the OP systems.
My guess was that something was causing the LNA to go into oscillation, but thatâs unlikely on two dongles.
What sort of PSU are you using? Are you able to unplug - rotate 180 degrees - re-plug?
(This makes a difference with some PSUs)
1 Like
The German Schuko is reversible (at least it was when I lived there!).
The silly french version is not, with its reverse sex earth pin.
Thx to all suggestions. Iâm on my way to Stuttgart so i cannot experiment right not. But for sure german current plugs are reversible. Will give it a try. For the moment please have a look at my 25 hours scan. It looks a bit weird to me:
Iâve got a rtl-sdr v3 Dongle and a âADS-B Triple Filtered LNA (Bias Tee Powered)â. I activate bias tee and it works. What I wonder about is set the right gain. Iâm using dump1090-mutability and only setting to âagcâ works. after that I see the same number of planes at the same distance as before. Before I can measure any improvements or waiting if the â1400 overloadingâ still happens, I have to figure out if my setting is right.
Edit: Attaching new scan. Gain set to âagcâ.
Try a gain of 23 as a start.
If you want to do a spectrum scan, try a gain of 20 for that.
That scan is really strange, there should be something at 1090 MHz.
Maybe itâs because of using AGC, donât know why you would use that for a scan.
You can use this command line for a scan:
sudo rtl_power -f 700M:1200M:100k -i 30 -c 50% -e 30m -g 20 -F 9 >scan.csv
Iâve played a bit with the gain and decided to use 23db:
sudo rtl_power -f 700M:1200M:100k -i 30 -c 50% -e 30m -g 23 -F 9 >scan.csv
Number of frequency hops: 358
Dongle bandwidth: 2793296Hz
Downsampling by: 1x
Cropping by: 50.00%
Total FFT bins: 11456
Logged FFT bins: 5728
FFT bin size: 87290.50Hz
Buffer size: 16384 bytes (2.93ms)
Reporting every 30 seconds
Found 1 device(s):
0: Realtek, RTL2838UHIDIR, SN: 00000001
Using device 0: Generic RTL2832U OEM
Found Rafael Micro R820T tuner
Tuner gain set to 22.90 dB.
Exact sample rate is: 2793296.086905 Hz
[R82XX] PLL not locked!
User cancel, exiting...
and
./heatmap.py scan.csv scan.png --ytick 60 --palette twente
generates this one:
Still looks weird, a bit distorted. 
When I postet an hour ago that my receiver doesnât show more aircraft than with the ne rtl-sdr dongle, that actually isnât right. Something happens after doing the scans. Iâve restartet the pi and now I have better distance and message rate. Every time I change the gain setting I restart the dump1090-mutability using
sudo systemctl restart dump1090-mutability
and I can see the immediate effect. But actually a restart is having an additional effect. 
Probably the bias-t wasnât active when you did the first scan.
That would explain why the 1090 MHz signals werenât visible.
Maybe using rtl_power switches off the bias-t or something.
After the reboot of the RPi check the plane count and then restart dump1090-mutability and check if it changes anything.
Maybe you need the bias-t activation on every start of dump1090-mutability.
(Not too complicated)
It looks like you have MASSIVE mobile phone signals interfering with your reception.
Canât even believe those scans are with the rtl-sdr LNA.
For me these are the scans looked like this:
1: ProStick Plus without external filter
2: ProStick Plus with added FA filter
3: RTL-SDR BLOG V.3. with RTL-SDR BLOG LNA
But maybe i used some other parameters.
What antenna does this system use?
What is clear though: the 1090 MHz area looks quite clean which should mean your reception is improved.
Notice that before with the Flightaware Pro Stick plus it was completely diffuse.
I think that 800 MHz is an LTE tower that must be on the adjacent building or otherwise very close.
Itâs probably on the building next to the receiver you were having problems with.
No clue what that tower does at 1400 hours but hey itâs very likely the problem was related to that tower.
Maybe try not using the twente color coding, just use the basic command for creating the picture (no need for a new scan, you can conver the existing csv)
./heatmap.py scan.csv scan.png
Not sure someone mentioned this, but this is probably no big surpise. The light blue filter is now sold as an 978 and 1090 filter, it was never good at filtering out European mobile phone frequencies, which is why FA now also sells a different version in a darker blue.
Iâm using Flightawareâs 26 inch antenna. Reception was improved two hours ago. After coming back and looking at the graphics I was disappointed. Something is going on here. The reception drops and drops again. This is ridiculous. I just donât know what happens here. Yes it went dark. Yes my wife was using a dect phone. Yes there are less aircraft in the air. Yes everybody is switching on the TV. But I still should see some high flying lonely planes. They should be there at every timeâŚ
My antenna:
My pi:
The âdropsâ:
After the two drops Iâve played again with the gain (which was at â26â when the drops happened).
It seems that the ideal gain value suddenly changes. What is it? 
Edit: PI2 on the North can see right now 46 aircrafts at a maximum distance of 315km.
1 Like
Thatâs a wideband LNA, it will only make your situation worse since it amplifies all frequencies, including cellphone frequencies that seem to be very strong in your area. Youâll need the ADSB-version of their LNA, it has filters that dampen the signals from cellphones and passes through 1090Mhz signals to the reciever.
I use the ADSB-version in my system and it provides a clean spectrum with only ADSB-signals (I have several strong cellphone transmitters in my area and the Uputronics preamp didnât filter enough so I got the RTL-SDR since it has better filtering.
Yes, rtl-sdr blog sell two types, this is the adsb one:
