FlightAware Discussions

HOWTO: Airspy mini and Airspy R2: Piaware / dump1090-fa configuration


Thanks for the link.

Wow, the SDR# sharp is also there.

I have been using SDR# on my windows desktop since 2012. Beautiful program. If I remember correctly, at that time it was not merged to air-spy.


Thanks! These are all my “hobby” projects :slight_smile:


hello wiedehopf, i have received the airspy mini dongle. i will install on my experimental RPI. i assume the “uninstall” command is sudo bash uninstall.sh


The commands should work as listed in the first post of this thread:
(HOWTO: Airspy mini and Airspy R2: Piaware / dump1090-fa configuration)

There is another post floating around in the middle of the thread but please use the Howto in the first post of the thread :slight_smile:

As stated the system needs to have piaware installed already.


just got back into town…am preparing to do install now…also, i see at the top of the posting the addition…thanks, i will advise over the next day or so…


hello wiedehopf, i just installed the airspy mini. followed directions. had to reboot after install to get it working…but now working. preliminary results.
my reference side is an RTL-SDR.com. my original experimental side was also an RTL-SDR.com with much lower gain. before i installed the airspy mini i set both RTL-SDR dongle gains to -10. (i have always had my reference side at -10 since i have about 9dB of loss from my LNA (ADSB 1090MHz RTLSDR ) to the input of the dongles. when the gains were the same the ref side showed slightly more total aircraft and about 11% more messages/sec.
when i installed the airspy mini on the experimental side (no changes to the ref side) the airspy side showed a few more total aircraft but 9% more messages/sec than the ref side so i am seeing nice improvements in the message rate.

i seem to remember comments you made re airspy and the json data. the only data i see on the performance graphs is the positions per hour data. there is no strong signal data etc, can you advise where you made the comments re airspy and .json data (assuming i remember correctly)?


You won’t get any reliable signal strength data out of airspy_adsb at the moment. Only thing you can do is watch the local map and check if aircraft that are low and close by disappear where the antenna should be able to see them. That would indicate too much gain.

You can get back message rates and stuff like that. I don’t know where i made those comments but i’ll check and get back to you :slight_smile:

Great that it’s working!
Not sure why it would require a restart though, i presume you changed dongles while the power was off?
I’ll include something that assures a clean dump1090-fa restart even in case it hangs that’s about the only thing i can think of.
Anyway not too important i hope.

If you feel like experimenting you can try running 20 MHz as a sample rate, for burgdorfer it worked just fine on the RPi :slight_smile:
Are the instructions in regards to changing settings clear enough or is there room for improvement?


regarding the instructions, yes, they were very clear. to change gain, sample rate etc i see the line NET="-1 30005:beast -c localhost :30104:beast". do i just add to this line to change gain and sample rate


No, just check below, you can just change the value in these lines:
GAIN=21 and


there is a line SAMPLE_RATE=12 but no gain line


oops, sorry, i see it now


You can get messages/s displayed in the graphs:

sed -iE -e 's/messages-local_acc/messages-remote_acc/g' /home/pi/adsb-receiver/build/portal/graphs/make-collectd-graphs.sh

If you need to switch back when you use a different dongle on that system again you can just get my version of the graph script:

cd /home/pi/adsb-receiver/build/portal/graphs
wget https://raw.githubusercontent.com/wiedehopf/adsb-receiver/freeze/build/portal/graphs/make-collectd-graphs.sh -O make-collectd-graphs.sh

The signal graph won’t work though.


If you are satisified with the setup i would suggest you get rid of that 9dB of loss to make full use of the airspy dynamic range. (At least the splitter, not sure what else you have in the signal path)

While a stronger signal might have overloaded the rtl-sdr dongle, it probably is better with the airspy.


i completely agree, i currently use the splitter for a “reference” dongle. once i get the airspy mini tuned i will remove the splitter (3.6dB loss per output (it’s a good minicircuits splitter)) and the drop the gain of the airspy to further tune. as it
is, i am amazed at how high the RSSI values are and no drop outs on the airspy side of the splitter. all other losses are due to long cable runs to roof (using lmr400 cable) and the bias t injector and lightening arrestor (the FA antenna is on the roof)


The RSSI values are not really full scale. It’s in reference to some other receiver for compatibility reasons.
So it’s fine if basically everything reads 0, doesn’t mean it’s overloaded.

Even without the splitter you might just be able run it on 21. Only real way to tell if your gain is too high is vanishing planes :wink:


If the lightning arrestor passes dc you could use the interal bias-t of the airspy and get rid of some more attenuation.
Not strictly necessary though i guess.

If you are done tweaking would you mind to run the setup as it is another day or so, and then switch the two dongles and run it for another day and a half?
(Guys over in the other thread are looking at dongle options and such a comparison is always nice)


hello wiedehopf. i will run 24 full hours with the splitter (V3 vs mini (with mini gain at 21 and sample rate =12)). then i will run 24 hours of V3 vs mini with mini sample rate =20 (i have an RPI 3B (but not plus version). i will provide data for this. then i will remove the splitter and run the mini only with 3.6dB higher signal input and adjust gain (if needed). i will provide all data when complete


Actually right now it looks like the v3 has a little bit better range and number of positions.
So it might be interesting to exchange the coax cables between the v3 and airspy mini as well. (Hard to tell if the splitter is symmetric)
(Also if you could just make a post here when you change the setup then i’ll take a screenshot of the position distribution by range on the FA stats page)

I’ll do some before and after comparisons with my mini when i get it but those are not as good as airtraffic changes to much. You really need a splitter for a good comparison :slight_smile:


kuox1 is the V3. as of 15:23 the data is:
kuox1 (V3) aircraft 1524 and positions 212057
kuox2 (mini) aircraft 1533 and positions 216022


Have you measured the loss through your mini circuits? On both of mine, the loss was 5+ db with max variation 0.2db between ports.

You may find your loss is more than thought.