Announcement

Collapse
No announcement yet.

Flightradar24 decoder/feeder BETA testing (Win/RPi/Linux/OSX)

Collapse
This topic is closed.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    A new build is available now that makes small adjustment to the CPU usage, as well as the first Raspberry Pi build!

    I will check the altitude problem, but the "near aircraft" problem may be related to gain settings, we use adapted open source dump1090 code to demodulate data and therefore it's not been deeply investigated just yet.

    Comment


    • #17
      Ok, let me know if you need more data.

      Comment


      • #18
        For the Pi version - could you consider adding an auto-update feature?

        The program that is started would be a stub that establishes communication with the home server - this would then check for the presence and version of the main module, and if necessary update the main module before starting it (possibly enabled by the presence of an '--au=y' parameter)

        Comment


        • #19
          Yes it makes sense to have this as an option. It might be tricky to have a self update in terms of replacing the binary itself, but I can make the app exit with a specific code should there be an update, then all it takes is to download the new binary, swap it and replace which can be done with wget + tar.

          Comment


          • #20
            Feedback on RPi Beta 0.5

            (I cant spell receiver, misspell it and it doesn't say unrecognised parameter - it just craps out with invalid config message)

            I don't get the dump1090 map any more on http://rpi-ip:8080 (I'll miss it, can we have it back?)

            Can we link the feeder to adsbscope?

            Can't link to Radarview from chrome or Firefox (when given IP of Pi in settings, local data load failed 4 times) - this is nothing new though - it didn't work for me with the feeder arm-241/242.

            otherwise I get
            Code:
            [dvbt][i]Found 1 device(s):
            [dvbt][i]0: Realtek, RTL2838UHIDIR, SN: 00000001 (currently selected)
            Found Rafael Micro R820T tuner
            [dvbt][i]Using automatic gain control.
            Exact sample rate is: 2000000.052982 Hz
            [dvbt][i]Gain reported by device: 0.00
            
            root@raspberry-pi:~# cat /tmp/fr24feed.log
            [main][i]FR24 Decoder [0x02113000]
            [main][i]Built on 20140528-0558 (r:/Linux/armv6l)
            [main][i]Copyright 2008-2014 (c) Piotr Pawluczuk, Flightradar24 AB ( http://flightradar24.com )
            [main][i]Reader 0 started
            [main][i]Socket server started
            [main][i]Feed Network client started
            [feed][i]Downloading configuration
            [bs][i]Initializing server
            [bs][i]Starting server on *:30003
            [reader][i][0]Initializing reader
            [reader][i][0]Connecting to Generic receiver via (dvbt://local)
            [master][i]Starting processing thread
            [feed][c]Interval: 5s
            [feed][c]Latitude: 52.4762
            [feed][c]Longitude: -1.8986
            [feed][c]GND: YES
            [feed][c]NonADSB: YES
            [feed][i]defined 1 server
            [feed][n]EGBB19@83.140.21.66:8099/UDP
            [feed][n]connecting
            [i][stats]Stats thread started
            [reader][i][0]Connected to the receiver, authenticating
            [reader][i][0]Authenticated, processing messages
            [feed][n]connected
            [feed][n]switching to UDP
            [feed][n]working
            [feed][i]sent 42 planes in 2 packets
            [feed][i]sent 42 planes in 2 packets
            [feed][i]sent 40 planes in 1 packets
            Now running dump1090 (Malcolm Robb) and Rpi_242 feeder
            Last edited by peterhr; 2014-05-28, 19:04.

            Comment


            • #21
              Hi everyone,

              I'm new to ADS-B. Just got started a few days back as a feeder.
              So far i'm liking the Beta5 as is has an overall lower CPU utilization (~10%) compared to RTL1090 with fr24feed (~20%) on my low CPU spec that i run 24/7.
              I will try it on the RPi too in the next few days.

              Cheers!
              Last edited by who; 2014-05-29, 07:16.

              Comment


              • #22
                i know its early but are there plans for this in the future to take input from other devices, like GNS5890? I believe it gives its output as a raw serial stream.
                T-EGLF8

                Comment


                • #23
                  It will support pretty much all receivers except for airnav's radarbox.

                  Comment


                  • #24
                    @piopawlu: Gain setting seems little bit inconsistent. We need to use '--gain 340' in order to set 34.0

                    EDIT:

                    Could you also backport this fix? https://github.com/MalcolmRobb/dump1...95a138273fb15d
                    Last edited by kasper93; 2014-05-31, 14:54.

                    Comment


                    • #25
                      The decoder works good on my PI here, but i can still see a high CPU load. Dump1090 uses around 35%, the new FR24 Feeder (Version 20140528-0558) uses nearly 100%.
                      I see 2 threads of high load, one is round 94, the other is around 87%. Dump1090 was seen only as one thread, maybe there is some difference.

                      But thank you very much for this nice work - good to see that the "low level" T-Feeders are still welcome.

                      Comment


                      • #26
                        Originally posted by Yahoodi View Post
                        Dump1090 was seen only as one thread, maybe there is some difference.
                        It depends which version are you referring to. MalcolmRobb's fork have many changes to improve performance. Current FR24 Feeder performance should be similar to antirez's verions, because lib is based on the original dump1090. Unless feeder itself have big overhead.
                        Last edited by kasper93; 2014-05-31, 18:14.

                        Comment


                        • #27
                          Originally posted by kasper93 View Post
                          It depends which version are you referring to. MalcolmRobb's fork have many changes to improve performance. Current FR24 Feeder performance should be similar to antirez's verions, because lib is based on the original dump1090. Unless feeder itself have big overhead.
                          Ahh thanks for that hint ... i used the MalcolmRobb fork before. Now i tried it with antirez's version and i expect the same load.
                          It seems, that the option --aggressive makes the difference.

                          Comment


                          • #28
                            giving this a shot since my rtl1090/fr24 combo keeps stripping callsigns off a/c in my area.

                            Comment


                            • #29
                              what happens when it can't find the server? there was a brief outage on the fr24 side and now the feed is offline. i can't restart it remotely (feed is at work), does it keep sending data or self shutdown?

                              Comment


                              • #30
                                @kasper93: I'll have a look at this, but it may need to wait a while
                                @Filldo: It should keep reconnecting until it succeeds

                                Comment

                                Working...
                                X