Announcement

Collapse
No announcement yet.

RSSI Values compressed

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

  • RSSI Values compressed

    I'm not sure this qualifies as a hardware problem TBH but wondered if others are seeing a very limited range of RSSI values on their /dump1090/gmap.html webpage.

    I'm using only the software provided in the last week or so by FlightRadar24 for the Raspberry Pi.

    Mine only ever covers the range from approx -30.5 to -34 dbFS. This happens with two independent raspberry pi setups and regardless of the aerial in use, the dongle in use (I have four to choose from, one of which is a Flightaware Pro Stick Plus) and is largely independent of the range of the aircraft. I've also played around with gain settings as well as agc on and off, but the only thing that varies is the range and number of aircraft messages decoded.

    For example an aircraft around 2 miles away on final approach and which is clearly visible to me as well as to the aerial, might give a reading of -30 or so while those a good bit further away, even up to 150nm, can be anything from say -32 to -34.

    I have experimented with aerials including a coaxial colinear and a discone using a good low loss coax for 1090MHz, both of these on the roof of the house and with a very good take-off in all directions.

    Next step will be to add a filtered LNA at the antenna end of the feeder although I'm seriously wondering whether even that will make my range of RSSI values appear more sensible.
    Last edited by jackthom; 2018-05-28, 11:22.

  • #2
    @jackthom

    First check what RF signals are existing in your area and picked by your aerial.

    Do I need a Filter? - Find RF Signals In Area


    Here is what I get in my area

    Last edited by abcd567; 2018-05-28, 14:55.

    Comment


    • #3
      @jackthom

      Here is a plot of Signal Strength and Noise Level in dBFS for last 24 hrs for my Receiver (FA Aerial + FA Filter + FA Pro Stick + Pi)
      dump1090-mutability v1.15~dev

      Last edited by abcd567; 2018-05-28, 15:12.

      Comment


      • #4
        Thanks abcd567. Here's a windows scan using my Jetvision dongle and the rooftop coco antenna. Looks like I definitely need some filtering before the LNA.




        It's the luck of the draw whether much 1090 activity is recorded but after a few passes it wasn't much better than this.

        Edit: I later tried the FlightAware Pro Stick Plus and its internal filter doesn't seem to make a great deal of difference to the 800MHz and 950MHz signals.
        Last edited by jackthom; 2018-05-30, 18:06. Reason: Image Insertion

        Comment


        • #5
          Yes, for locations with strong GSM850 and EGSM900 signals, the internal filter of ProStick+ is not sufficient and an external filter is required.

          Comment


          • #6
            Very useful info, thanks again.

            The FA Filter is certainly cheap enough so I'll have to give it a try.

            I'll report back when it arrives.

            Comment


            • #7
              What is your gain setting?

              If you have not changed it, then it will be default "max".

              Try few lower values of gain like 45, then 40, then 35, then 30. For each setting, for a while watch performance (Messages/sec, Number of planes, max distance aeroplanes are picked by your system). This way you will find the best gain setting for your system.

              Change gain setting by using your browser.

              (1) Go to settings page at address given below.
              For DVBT pligged into RPi:
              ip-of-pi:8754/settings.html

              For DVBT plugged into Windows computer (using windows feeder):
              localhost:8754/settings.html

              (2) In the field "Process Arguments" add argument --gain n, where n is the value of gain you want to set.

              (3) Click "Save" button, then "Restart" button. Both these buttons are at bottom-right of settings page.
              Last edited by abcd567; 2018-05-28, 20:56.

              Comment


              • #8
                Gain is set to max at present. I did try a few different lower settings by editing the entry in the Dump1090-mutability configuration file but I prefer the idea of using the extra arguments field in the web settings page.

                TBH I wasn’t entirely sure my gain entries had the correct syntax. Should the gain numbers in the config file be presented as a string within speech marks <gain=“number”>, as a following number ie <gain number> or in the form <gain=number>.

                I’ll certainly try out a few more alternative gain settings tomorrow.

                Comment


                • #9
                  Originally posted by jackthom View Post
                  TBH I wasn’t entirely sure my gain entries had the correct syntax. Should the gain numbers in the config file be presented as a string within speech marks <gain=“number”>, as a following number ie <gain number> or in the form <gain=number>.

                  I’ll certainly try out a few more alternative gain settings tomorrow.
                  In the file /etc/default/dump1090-mutability , l have used syntax:
                  Gain="max"
                  Gain=38

                  I did not try
                  Gain="38"

                  After setting a value of gain in the above file and Saving the file, you can check if it is implemented or not by following command:
                  Code:
                  #First restart dump1090-mutability to pick the new settings
                  sudo systemctl restart dump1090-mutability
                  
                  #Now check status
                  sudo systemctl status dump1090-mutability -l
                  The output of above status command will be several lines. One of the lines, which is longest line in the output, will have gain value in use.

                  Code:
                  /usr/bin/dump1090-mutability --measure-noise --net --device-index 0 --gain 38 ......
                  Last edited by abcd567; 2018-05-29, 01:23.

                  Comment


                  • #10
                    Originally posted by jackthom View Post
                    Gain is set to max at present. I did try a few different lower settings by editing the entry in the Dump1090-mutability configuration file but I prefer the idea of using the extra arguments field in the web settings page.
                    The "Process Arguments" in the settings page ip-of-pi:8754/settings.html are passed to dump1090-mutability (installed by fr4feed), if "Receiver = DVBT" is used.

                    If you are using a seperately installed dump1090-mutability, and using

                    "Receiver=AVR(TCP), Host/IP=127.0.0.1:30002"
                    or
                    "Receiver=ModeS Beast (TCP), Host/IP=127.0.0.1:30005",

                    then the "Process Arguments" will not be passed to your separetely installed dump1090-mutability.
                    Last edited by abcd567; 2018-05-29, 01:39.

                    Comment


                    • #11
                      I tried taking the gain down to 10 initially and the range and number of messages per second both took a nosedive, as might be expected.

                      Increasing the gain gradually up to max gave an improvement all the way as far as I could tell, the number of messages etc never seemed to go into reverse.

                      I'm regularly seeing aircraft around the 150nm mark and the system is working reasonably well, it's just that the RSSI values rarely get better than -31dbFS. They certainly don't seem to be following anything close to an inverse square law in relation to aircraft distance, so I guess I'll try to filter and amplify until the 1090mHz signal levels are less dwarfed by the local cellphone transmissions around 960MHz.

                      Comment


                      • #12
                        960 Mhz is too close to 1090 Mhz, and there are chances that Flightaware filter may not fully clear it. If this happens, then you will need a Cavity filter or Ceramic Filter.

                        Comment


                        • #13
                          I'm afraid that's right. If anything the FA filter caused the number of messages/sec to fall a little rather than improving things, presumably because of its ~2db insertion loss.

                          I already ordered a RTL-SDR.com 1090 LNA/filter last weekend (coming to UK from China), so I'll give that a try when it finally arrives. They claim its triple filtering achieves around -50db at 960MHz.

                          Hope that's not just sales talk.
                          Last edited by jackthom; 2018-05-31, 13:26.

                          Comment


                          • #14
                            Just an update on the original problem which wasn't fixed by the RTL-SDR 1090 LNA/filter.

                            I have set up an alternative Raspberry Pi 3B+ rig, this time using "dump1090-mutability_1.15~dev".
                            Testing this, while using the same aerials as before, the RSSI values are now looking much more sensible.
                            The previous odd compression of received signal strengths has gone and they now range between approx -1.5 and -33 dbFS.

                            I have a cavity filter on order from Osmocom to try and sort out the old setup (too late to cancel it) although I'm no longer so sure its entirely down to high signal levels from nearby cellular masts.
                            I guess in any case the cavity filter can't do any harm and it may well improve the whole system besides fixing this "problem", considering that no-one else seems to have have reported any similar issues with the previous version of dump1090-mutability.
                            Last edited by jackthom; 2018-09-03, 10:58.

                            Comment


                            • #15
                              jackthom: Would like to hear how you go with the Osmocom cavity filter

                              ylis

                              Comment

                              Working...
                              X