Announcement

Collapse
No announcement yet.

Pi Zero feeding - bizarre statistics, any ideas?

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

  • Pi Zero feeding - bizarre statistics, any ideas?

    I've posted about this before but it has now become ridiculous!

    I fed with a 1st generation Pi B, then moved to Windows. Things stayed mostly the same with a range of 50NM or so. It's never more as I have mountains all around me.

    When Windows was ousted I switched to a Pi Zero and my range went down a fair bit. But it has now sunk to 0nm. As of this moment my statistics show:

    Aircraft seen: 108
    Positions reported: 61
    Maximum distance: 0nm
    Hits reported: 1,248

    Now I'm pretty sure 108 planes haven't flown directly overhead in the last 12 hours or so, and even if they had they'd be at 30,000ft plus which, in itself, is over 5 miles. I'm a couple of hundred miles from a major airport and not really on a major flight path unless there are odd weather systems being avoided. So 108 aircraft and 1,248 hits sounds as if it could be about right. But only 61 positions and 0nm doesn't.

    Should I suspect the dongle (RTL-SDR) or the Pi Zero? Power supply is checked out and fine, the antenna hasn't moved since I started feeding in about 2015. Any advice most welcome, thank you.

  • #2
    There is significant voltage drop in the cable to the pizero and in the pizero to the USB port and then another adapter.

    I'd recommend another pi with proper USB ports.

    Get a green radarbox dongle, they have a builtin filter/amplifier and are quite cheap.

    Hard to say what the problem is exactly.
    You can always try using the dongle via a powered USB hub if you have one.

    Comment


    • #3
      Originally posted by wiedehopf View Post
      There is significant voltage drop in the cable to the pizero and in the pizero to the USB port and then another adapter.

      I'd recommend another pi with proper USB ports.

      Get a green radarbox dongle, they have a builtin filter/amplifier and are quite cheap.

      Hard to say what the problem is exactly.
      You can always try using the dongle via a powered USB hub if you have one.
      I'm actually using a Y-lead to a single port on the Pi Zero - quite acceptable method according to the documentation! So the voltage at the dongle is identical to the voltage for the Pi and they are both 5.1V so I don't think that's the problem.

      But your reply has made me think I'm being somewhat stupid. I have an old Macbook so I could set it up on that and see if the dongle still performs as it did with a Windows laptop. That'd be the sensible way forward. Process of elimination. Thanks for switching my brain on!

      Comment


      • #4
        You'll need to run some Linux on the Macbook, but i don't suppose that's an issue?

        And the Y-lead is nice, but it might introduce extra noise on the USB line i suppose.

        You can try with a 6.7 cm paperclip in the SMA connector of the dongle instead of your current antenna.
        That should exclude a bad antenna.

        Comment


        • #5
          Originally posted by wiedehopf View Post
          You'll need to run some Linux on the Macbook, but i don't suppose that's an issue?

          And the Y-lead is nice, but it might introduce extra noise on the USB line i suppose.

          You can try with a 6.7 cm paperclip in the SMA connector of the dongle instead of your current antenna.
          That should exclude a bad antenna.
          Thanks for the ideas wiedehopf. The OSX software off the sharing page is actually a complete integrated package. Unzip it, click to start it, and off it goes. Nothing more needed so whoever put that together - five stars!

          Change of question now. How/where is range calculated in the feed software?

          I had a slight improvement using OSX rather than the Pi Zero but still nothing like it should be. 15nm instead of 0nm. So I stopped the feed software and just ran dump1090 using --interactive to get up the terminal display. And almost immediately it was showing an aircraft which was 53 miles away line of sight according to google maps. I simply noted the Hex and looked to see where it was on the FR24 main display.

          It was a full set of data - Hex, Flight, Altitude, Speed, Lat, Long, Track, Messages, Seen. With the "Seen" being a low number of seconds and messages in the hundreds. So the dongle is receiving everything correctly. It's getting passed to dump1090 correctly. But it can't work out range from the signal strength as the aerial would make too much difference. So I presume it's being computed from the aircraft's co-ordinates and my co-ordinates (which are right and show correctly on my FR24 statistics page).

          So where is that worked out? In the dongle? In the feeding software? Or at the FR24 servers? I can't believe it'd be the last as it would need pointless amounts of computing power for everyone feeding. And I can't believe it'd be my computer as I wouldn't have a Macbook and a Pi Zero which had both failed their basic Maths. So I guess it must be worked out in the dongle. But that doesn't really make sense either because when it's all set up and you put in the dongle's co-ordinates isn't that just sent to FR24 - and I don't think you even have to put them in if you don't choose to use MLAT? I'm not sure even dump1090 knows them when run by itself.

          I'm confused now.

          Comment


          • #6
            Well if you would run Linux on the laptop, you could check :P

            The range is calculated by the feed software i believe, it gets the location from the servers.
            Calculating a distance is not hard.

            You need much more consistent signal to get a range on the website though, it only counts aircraft you receive consistently for let's say 30 seconds or so.
            For the --interactive dump1090 needs a position configured somewhere.

            If you want to it right, you'll need a new dongle anyway and optimally a Raspberry Pi.

            Have you optimized your antenna? https://forum.flightradar24.com/thre...h-DVB-T-Dongle

            Comment


            • #7
              Originally posted by wiedehopf View Post
              The range is calculated by the feed software i believe, it gets the location from the servers.
              That would make sense - so it shouldn't differ from one machine to another.

              Originally posted by wiedehopf View Post
              You need much more consistent signal to get a range on the website though, it only counts aircraft you receive consistently for let's say 30 seconds or so.
              From the basic dump1090 - not the full feeding software - I am getting consistent signals. An aircraft that was tracked when it was probably 40 miles away sent 1,226 messages. According to the internet the interval between messages can vary, but according to my stopwatch that's two minutes or more of constant data.

              So, being a poor pensioner who finds everything expensive, and even more importantly for the sake of the planet, I really don't want to start replacing equipment which may not be faulty. I'm now also wondering if packets are being dropped somewhere because of my aged router or my ancient ADSL copper wire. Perhaps I should try FR24 support and see if they have any ideas.

              Comment


              • #8
                Well you didn't respond to the antenna thread i posted.

                It might just be that each time you are using a different position is what influences reception.
                It's basically impossible to tell what the problem is without either replacing parts or at least having reference parts to check.

                Try with a can below the antenna or different antenna position.
                Or just describe what kind of antenna you are using, most of the time there is room for optimization.

                Comment


                • #9
                  Originally posted by wiedehopf View Post
                  Well you didn't respond to the antenna thread i posted.

                  It might just be that each time you are using a different position is what influences reception.
                  It's basically impossible to tell what the problem is without either replacing parts or at least having reference parts to check.

                  Try with a can below the antenna or different antenna position.
                  Or just describe what kind of antenna you are using, most of the time there is room for optimization.
                  Little antenna! On a tin. Not the best I know but it's the same one I've been using for nearly five years, in the same spot, which gave me 50nm, then 0nm, and now 50nm again when using just dump1090 --interactive.

                  And likewise the dongle is performing as it should with --interactive. So logically it has to be somewhere further down the chain. Hence wondering if it's in the transmission to FR24 although my router isn't logging any errors. It's an old industrial strength BiPAC 7800DXL which probably gives more comprehensive information than almost any other ADSL router ever made! All just very strange!

                  Comment


                  • #10
                    The time column, is time since last decodable signal. If that is ever more than 0, then that's a big tell tale that you aren't getting enough for fr24 to care about in each upload.

                    2-4 per second. And not Evey one is a location packet. And if full off noise and not able to get the crc bit at the end. Can be non decodable

                    And can vary with gain.

                    Sent from my EML-L09 using Tapatalk
                    Posts not to be taken as official support representation - Just a helpful uploader who tinkers

                    Comment


                    • #11
                      Originally posted by Oblivian View Post
                      The time column, is time since last decodable signal. If that is ever more than 0, then that's a big tell tale that you aren't getting enough for fr24 to care about in each upload.

                      2-4 per second. And not Evey one is a location packet. And if full off noise and not able to get the crc bit at the end. Can be non decodable

                      And can vary with gain.

                      Sent from my EML-L09 using Tapatalk
                      Oh right! So does the dongle do any processing at all? Or does it just receive and amplify and shove it out to the USB port? And presumably ignore anything where the packet is incomplete or the checksum is no good? And then that is also bundled up and sent through to FR24 without any checks by the host computer?

                      If all that's right, was there a time when FR24 changed or improved its data checking? Maybe when Windows was depreciated? Because that would then start to make sense. And mean the dongle was actually out performing itself in my early days of feeding. Sorry for the layman's language but as I said to wiedehopf I'd like to understand the problem before replacing equipment. If my guess is near the mark and my data is and will forever be pretty much useless (and all the military have switched off their transponders anyway) I might ponder on whether to carry on.

                      Comment


                      • #12
                        Nothing. It is a radio receiver. Designed for tv.

                        S d r

                        Software, defined, radio.

                        What the software does is send a command to it go Change frequency and start receiving.

                        Dump1090 does the work. And as such is adjustable.

                        It gets the raw data off the frequency tuned. And decodes it. Hence the 'dump'

                        Everything else including fr24 uses the readable decoded data from there.

                        Sent from my EML-L09 using Tapatalk
                        Posts not to be taken as official support representation - Just a helpful uploader who tinkers

                        Comment


                        • #13
                          I should point out, I believe the fa prosticks have been tuned and locked (tv sticks drift and may not tune the same twice depending on heat etc)

                          And filters and amps added. So in that respect they do more optimising for the signal before outputting it.

                          Sent from my EML-L09 using Tapatalk
                          Posts not to be taken as official support representation - Just a helpful uploader who tinkers

                          Comment


                          • #14
                            Originally posted by Oblivian View Post
                            The time column, is time since last decodable signal. If that is ever more than 0, then that's a big tell tale that you aren't getting enough for fr24 to care about in each upload.

                            2-4 per second. And not Evey one is a location packet. And if full off noise and not able to get the crc bit at the end. Can be non decodable

                            And can vary with gain.

                            Sent from my EML-L09 using Tapatalk

                            Thanks for all the information.

                            So I've been monitoring it just on dump1090 --interactive and even with an aircraft close by it has problems holding the "Seen" at 0 secs. Little jumps to 1 sec or more at quite regular intervals. Dump1090 on startup says the gain is set to maximum (49.4 from memory), and the aerial is where it has always been, and the Macbook running it has 500mA at the USB port for the dongle which will easily be enough, and I've cleaned all contacts with quality switch cleaner.

                            So does all that suggest to you the dongle is weary after nearly five years of use? It's just a DVB-T stick - Macbook says the chip is RTL2838UHIDIR. Blue! And has always got quite warm. Are the flight aware dongles the way to go?

                            Comment


                            • #15
                              radarbox green dongle is better with interference and cheaper.

                              But FA dongles are also good, depending on interference an extra filter is required/improves reception.

                              Or best option in my opinion, this LNA together with the v3 or smartee dongle: https://www.rtl-sdr.com/new-product-...mhz-ads-b-lna/

                              Comment

                              Working...
                              X