Announcement

Collapse
No announcement yet.

Dynamic Range improvement.

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Dynamic Range improvement.

    For some time I’ve been using a discone antenna a few metres above the house roof with around 15m of good quality coax feeding a Pro Stick Plus with reasonable but not great results.
    I recently decided to upgrade, replacing the discone with a FlightAware aerial and a Rtl SDR-Radio Blog filtered LNA a few metres from the antenna. The original 15m coax still delivers the amplified signals to the receiver.

    First of all the Pro stick plus had to go as its internal LNA was now being swamped with too much signal, so my RTL V3 dongle took its place with some improvement.
    Despite running the dongle at 0db gain, aircraft taking off a few miles away were still dropping out so this was eventually fixed by inserting a 20db attenuator at the dongle end of the coax.

    The system is working fine and is much better than the original setup, but there is clearly a trade off between reliable decoding of strong local signals and getting higher message counts with longer range which is presumably down to the dynamic range the dongle is capable of handling.

    I wonder if anyone has progressed to alternatives such as the Airspy and managed better local and distant reception.

  • #2
    Omni directional antennas always perform better further away, due to a donut effect those overhead will always be effected.
    Either by overloading if the device is made too sensitive, or weak signal due to the lateral propagation
    Posts not to be taken as official support representation - Just a helpful uploader who tinkers

    Comment


    • #3
      Yes in my case the overhead signals are always rock solid.

      It’s the low level aircraft taking off a couple of miles away which would be liable to overload if I try to improve the sensitivity and range any further.

      Comment


      • #4
        My guess would be that somehow adjusting the gain didn't work, reducing the gain to 0 should be enough attenuation.
        Maybe there is something overloaded before the gain is applied, that's possible but i doubt it.

        The Airspy Mini works nicely, there is quite a long thread over at FA regarding its configuration and general stuff as well: https://discussions.flightaware.com/...guration/44343

        This installation is running an airspy: http://mtg.getmyip.com:747/tar1090/
        But don't expect that much, it's on a 10 m tower with flat land or ocean surrounding it.


        In regards to the gain adjustment which dump1090 version are you using?
        I'd recommend trying dump1090-fa and consider using the automatic nightly gain adjustment i added: https://github.com/wiedehopf/adsb-wi...for-dump1090fa

        Maybe you can also try reducing the voltage fed to the LNA, at 3 V my rtl-sdr LNA works just as well with a little less gain and runs quite a bit cooler.
        Last edited by wiedehopf; 2019-08-10, 10:41.

        Comment


        • #5
          I'm using Dump1090 1.15~dev and I did wonder why dropping the original gain setting of 49.6 down gradually to zero didn't really seem to be having the expected effect.

          The LNA allegedly provides around 27dB gain and at present I am using 3.3V from the Raspberry Pi to power it. I experimented with lower voltage levels to reduce the LNA gain but gave up and inserted the attenuator instead.

          I'm guessing the upgrade overall might have increased the signal levels by something around 30dB and I would have thought that after losing the internal LNA of the Pro Stick plus, then slightly reducing the --gain setting in the replacement dongle would have been about right.

          It does seem absurd to have to use an attenuator in the feed while the dongle is supposedly set to 0dB gain.

          For the record I set the gain in the Fr24 Feeder Settings webpage, "Process Arguments" with a value of --gain 0.0
          Just tried varying the gain a few times and it doesn't seem to be having any effect that I can see!

          I'll take a look at the Airspy thread thanks.
          Last edited by jackthom; 2019-08-10, 11:29.

          Comment


          • #6
            If you are using dump1090-mutability as a service and not started by fr24feed, you need to change those settings.

            You can check the first line of this command output:

            cat /etc/fr24feed.ini

            If it says beast-tcp, then dump1090-mutability is not started by fr24feed and the "process arguments" won't have any effect.
            Then you need to change the gain via these settings:

            sudo dpkg-reconfigure dump1090-mutability



            Alternatively you can switch to dump1090-fa:
            https://github.com/wiedehopf/adsb-sc...or-dump1090-fa
            Maybe even with an interface with improved load time and some other improvements:
            https://github.com/wiedehopf/tar1090


            Independently (also works with dump1090-mutability) you might want some graphs to check the message rate and other stats:
            https://github.com/wiedehopf/graphs1090#graphs1090
            Last edited by wiedehopf; 2019-08-10, 11:52.

            Comment


            • #7
              Thank you very much for that.

              I'd completely forgotten that FR24feed does not start dump-1090 in this case and a quick dpkg-reconfigure is now actually changing the gain.

              I'll remove the attenuator and play around with the gain for a while to push things as far as I can.

              Comment


              • #8
                I'm sure that'll improve things a bit for you

                Comment


                • #9
                  Alternatively you can switch to dump1090-fa:
                  https://github.com/wiedehopf/adsb-sc...or-dump1090-fa
                  Maybe even with an interface with improved load time and some other improvements:
                  https://github.com/wiedehopf/tar1090


                  Independently (also works with dump1090-mutability) you might want some graphs to check the message rate and other stats:
                  https://github.com/wiedehopf/graphs1090#graphs1090[/QUOTE]



                  Thank you for these scripts.

                  And wow, what a lot of planes do you see.

                  Comment


                  • #10
                    I’m still very happy at the moment with the package of Dump1090 1.15 which includes the performance graphs, as scripted by J Prochazka.
                    However I just got hold of another rpi 3b+ so I am now tempted to try out the goodies on offer from wiedehopf.

                    Comment


                    • #11
                      I am running my PiAware with his alternative dump1090 interface and the 1090graphs where i am totally happy with. Initially i thought about moving to Dump1090 Mutability, but i decided to keep it as it is.

                      Comment


                      • #12
                        You can still install separate web interfaces:

                        Webinterface from dump1090-mutability 1.15 using dump1090-fa data: https://github.com/wiedehopf/dump1090-retro-html


                        https://github.com/alkissack/Dump1090-OpenLayers3-html

                        (you either overwrite dump1090-fa html directory or create your own configuration file for it.
                        As easy as copying /etc/lighttpd/conf-enabled/88-dump1090-retro-html.conf and adapting it.
                        Or copy /etc/lighttpd/conf-enabled/89-dump1090-fa.conf it's basically the same.
                        Just be aware there is no need to recreate the stuff using port 8080.

                        Comment


                        • #13
                          I know, i am using it already since our exchange in a different forum

                          Comment

                          Working...
                          X