Alta Bacchus Harstad Kiruna Lund Puffin Stamsund RPi 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15 Pixie Bergen Torvik Ystad Web server
WAN Network Timekeeping CPU/mem  EUMETCast Europe  DVB-S2 Ecast I/O Losses HVS losses RAMdisk T1/T2 Air °C  ADS-B GPS SNR  How-to

ADS-B: antenna and receiver comparisons

_________________________________________________________________________________________________________
Yagi antenna in the loft pointed at Edinburgh Airport ~11 km away.
2016-Mar-15, wk.11 RTL dongle, bronze cavity filter, no pre-amp
2016-Apr-14, wk.15 09:15 Replaced Bronze filter with JetVision one, not good!
2016-May-06, wk.18 08:45 FlightAware Pro stick receiver
2016-Aug-14, wk.32 13:15 Add 2-way splitter and fed Airspy/RasPi-10 from the other output
            Expect RasPi-10 to show increase, RasPi-7 to decrease
2017-Mar-13, wk.11 10:10 UTC Changed to FA Pro Plus stick without Bronze filter
RasPi-10/Airspy
Client - PC Ystad
Wi-Fi connected
Maximum range (km)
Message rate (msgs/sec)
FlightAware stats from RPi-10
RasPi-7/FA Pro Plus stick RX
Client - PC Stamsund
Wi-Fi connected
Maximum range (km)
Message rate (msgs/sec)
FlightAware stats from RPi-7
PlanePlotter share ZA
_________________________________________________________________________________________________________
Outside gain vertical (since 2016-Mar-15)
Cavity filter, HAB/Uputronics pre-amp. with SAW filter.
2016-Apr-14, wk.15 06:00 replaced PE1RKI filter with JetVision one until 08:30. Much worse. Back to PE1RKI.
2016-May-16, wk.20 11:00 updated Airspy to f/w 1.7, and airspy_adsb to v1.16
2016-Jul-17, wk.28 16:10 UTC Updated airspy_adsb to v1.17
2016-Aug-07, wk.31 17:05 UTC Replaced T-piece with Chinese 1-3 GHz splitter
2016-Aug-13, wk.32 11:20 UTC Changed scales from 500 to 600 maximum
RasPi-13/Airspy
Client - PC Kiruna
LAN connected
Maximum range (km)
Message rate (msgs/sec)
FlightAware stats from RPi-13
PlanePlotter share Xf
Beast receiver
Direct connect - PC Alta
Maximum range (km)
Message rate (msgs/sec)
FlightAware stats from PC Alta
_________________________________________________________________________________________________________
Indoor "3 dB gain" vertical antenna on a south-facing window ledge.
Pentland Hills block in this direction.
2016-May-16, wk.20 09:25 UTC, replaced Airspy R2 with Airspy mini.
2016-May-16, wk.20 10:00 UTC, updated airspy_adsb to v1.16
2016-May-22, wk.20 14:45 UTC, replaced RTL dongle with FlightAware stick
2016-Jul-17, wk.28 16:25 UTC, updated airspy_adsb to v1.17, added -n (highest native sample rate)
2016-Aug-04, wk.31 11:05 UTC, replaced T-piece with Chinese 1-3 GHz slpitter
2016-Aug-31, wk.35 14:40 UTC, replaced unreliable RasPi-8 with new model 3B
RasPi-9/Airspy mini
Client - PC Bergen
Wi-Fi connected
Maximum range (km)
Message rate (msgs/sec)
RasPi-8/FA Pro stick RX
Client - PC Torvik
Wi-Fi connected
Maximum range (km)
Message rate (msgs/sec)
_________________________________________________________________________________________________________

Other periods: day week month year

How these results were gathered - see here.

Discussion: Note that both message-rate and maximum range are only sampled at 5-minute intervals, the default for MRTG.  Thus there is an element of luck in the results in that if an approaching aircraft appears on the distant horizon just at the sample time on one system, but just after the sample time on another, the other system will only record the aircraft some five minutes later when it will be much closer.  This is why sometimes the Airspy-based systems may not show as having the maximum range.  Try instead to judge whether the maximum range values are more consistently higher on one system than on the other.
The higher message rate does not only imply that more planes are detected.  Whilst more planes would give a higher message rate, the better processing in some firmware or software may provide better detection in the presence of multiple signals.  In these circumstances, more messages can make for a smoother plot of the track in PlanePlotter.
The maximum range is limited to 500 km (270 nm) and the maximum message rate to 500/second in the plots above, to allow both to be plotted on one graph.

Events (out of date)
2015-Sep-24 08:30 UTC Installed FlightRadar24 feeder on RasPi-7, but it didn't seem to be connecting to the dump1090 server on port 30005. Seems this needs to be configured as receiver="beast-tcp" rather than "avr-tcp" in the /etc/fr24feed.ini file. It then worked as expected.
2015-Jul-09 11:50 UTC Updated PiAware to 2.1-2
2015-Jul-03 20:15 UTC Updated PlaneFinder client to 3.0.2080
2015-Feb-13 08:15 UTC Added PlaneFinder client 3.0.450 to RasPi-7.  Noted about 7% CPU increase and about 30 MB memory increase.
2015-Feb-13 17:00 UTC Changed to RasPi-7 and PiAware 1.20
2015-Feb-08 12:00 UTC Added filter & pre-amp, better results
2014-Nov-19 18:15 UTC Updated PiAware to 1.18
2014-Oct-26 07:25 UTC Updated PiAware to 1.15
2014-Oct-04 07:30 UTC Updated PiAware to 1.13
2014-Sep-12 06:37 UTC Discovered that RasPi-3 hadn't been uploading to Plane Plotter since Sep-06 at 19:35 UTC. Restarted the 1090 processes.
2014-Sep-08 16:25 UTC Updated PiAware to 1.11
2014-Aug-31 15:15 UTC Updated PiAware to 1.9
2014-Aug-24 06:10 UTC Updated PiAware to 1.8
2014-Aug-16 11:00 UTC Installed PiAware 1.5 on RasPi-3
2014-Aug-10 10:55 UTC Updated dump1090 to V1.09.0608.14.
2014-May-27 15:30 UTC Updated dump1090 to V1.08.2705.14 which has more network buffering. Noticed a welcome CPU drop from ~35% to ~32%, which gave a slight CPU temperature drop too.
2014-Apr-02 06:05 UTC Connected back to the EDI-facing Yagi antenna.
2014-Mar-25 09:45 UTC Connected to main antenna instead of the EDI-facing yagi, and changed PC Alta (share tD) from The Mode-S Beast to the RPi for a test for John Locker.
2014-Mar-23 09:10 UTC Add --modeac to the parameters to add mode-A/C decoding. This increased the CPU load from averaging 35% to 45%.  Dropped in early April.
2014-Mar-11 07:20 UTC Updated to version 1.08.1003.14 which has a heartbeat to keep the connection with Plane Plotter alive.
2014-Feb-26 - Client PC Ystad is being fed from a microADSB receiver for a temporary test, so it will have no Wi-Fi I/O from the Raspberry Pi server. This also means that the Raspberry Pi is only feeding one PC with data rather than two, so its I/O rate would be expected to be halved.
Unexpectedly, this has rsulted in PC Stamsund's Plane Plotter restarting the TCP connection often during the night when there has been little traffic.
2014-Feb-24 11:20 UTC Updated dump1090 (and ppup1090) to version: 1.08.2302.14. Fixes: (1) missing blocks cause timestamp slips, and (2) incorrect beast binary input/output handling of 0x1a escape characters, but neither should affect my own system.
2013-Nov-10 11:40 UTC Edited /etc/init.d/dump1090.sh to include the ppup1090 program following suggestions from Steve Tickle. This allows the ppup1090 program to start automatically with the rest of the RPi. Tested by a warm reboot. Details are here.
2013-Nov-08 18:40 UTC RasPi-3 seemed to stop working at about the same time as I was setting up a new iPad with iOS-7. Co-incidence? Restarted OK, and did a manual start of ppup1090
2013-Nov-05 10:35 UTC Restarted PPupload function, as it had not been restarted at the Oct-20 reboot. dump1090 is at version 1.07.0710.13
2013-Oct-09 06:45 UTC Updated dump1090 to 1.07.06.10.13 to add the PPupload functionality.
2013-Sep-11 11:45 UTC RasPi-3 went off the air, not pingable. Power-off restart at 20:10, and it looks OK.
2013-Aug-21 14:00 UTC Updated dump1090 program to version 1.07.1908.13. Rebooted RasPi-3 to ensure it started OK, and to get a standardised starting condition.
2013-Aug-20 18:00 UTC Noted a zero message rate on PC Ystad from RasPi-3 dump1090 program. Tried pinging RasPi-3 and that was OK, tried stopping and restrting Plane Plotter but that didn't help. Tried closing and re-opening Plane Plotter but that didn't work. On RasPi-3, the dump1090 process was running (1 day 8 hours out of 8 days 36 hours up-time), but did a restart (sudo /etc/init.d/dump1090.sh start) and the TCP started flowing once again.
2013-Jul-24 18:20 UTC PC taking data from RasPi-3 showed no planes and zero message rate. RasPi-3 responding normally, and dump1090 process still running. Next time I looked, Plane Plotter had started seeing data again, so its automatic restart of the TCP/IP connection seems to have resolved the issue.
2013-May-30 06:20 UTC dump1090 appeared to have stopped sending data, so restarted it.
2013-May-18 11:05 UTC Updated dump1090 to 1.06.1405.13.
2013-May-16 19:40 UTC Added PC Ystad back in as receiver PC, RasPi now feeding two PCs.
2013-Apr-29 07:30 UTC Changed receiver PC from Ystad to Stamsund.
2013-Apr-29 04:50 UTC Moved the Raspberry Pi to the shack, connected to a yagi antenna pointing at the airport.
2013-Apr-28 07:10 UTC Moved the Raspberry Pi back to the spare room, may have been causing interference to 137 MHz APT reception. Mag mount antenna wasn't well located.
2013-Apr-19 16:40 UTC Updated dump1090 to 1.01.1924.13. Got start/stop script working.
2013-Apr-15 15:30 UTC Playing with script to start and stop dump1090, buffer size now 150. Later, buffer size: 500, number: 5.
2013-Apr-14 05:30 UTC Updated dump1090 to 1.00.1404.13 with buffer size 100. May reduce CPU overhead.
2013-Apr-09 14:00 UTC  Moved the Raspberry Pi into the loft, slightly better antenna.

Using the Raspberry Pi as a Wi-Fi Plane Plotter server
Installing and configuring the Raspberry Pi
Using the Raspberry Pi as a stratum-1 NTP server

 
Copyright © David Taylor  Edinburgh SatSignal home page Last modified: 2017 Apr 23 at 18:32:42