Flightradar24 Feeder Chat

[HighlyCool]

Could also depend on antenna polarization. ADS-B signals are vertically polarized. In turn your antenna for your feeder and the aircraft are vertically polarized antennas. The signal for the plane may be presenting itself to other antenna better than yours even though it is farther away as the best areas for reception are off at angles on the sides of the antenna(s) an not off of the tip(s). See below

[BrianG]

Seeing More With The Pi

This is VERY helpful. Part of what I'm doing here is creating a Raspberry Pi STEM Computer Room [Pi, Linux, Python] and being able to combine the Pi with my existing set-up is terrific. I will begin the set-up ASAP.

Many thanks,

see http://ins-knowledge.blogspot.co.uk/...ys-better.html

... maybe you could fill the hole using a pair of "raspberry pi" based receivers [ http://forum.flightradar24.com/threa...e-to-feed-FR24 ] to fill the hole with antennas pointed N-S (will fill overhead and E-W) and E-W (will fill overhead and N-S) if your F- antenna is pointed up-down (providing distance cover)

Put the working Pi's in a nice clear case like this http://www.adafruit.com/images/medium/ID859_MED.jpg9 and it could be part of the display (the Pi being designed as an educational computer)

It would be good to have a screen display what each receiver can see ... with the Pi you can point a browser at the Pi on port 8080 to see this [try the link in my sig under this post to see this] (I wonder if you could do this with the pi itself in graphical mode running it's own browser driving a small composite monitor from the video out)

[byteworks]

I would say that the definitive answer was given by FR24support and not explained by the antenna polarization. My radar station (a F-) is rarely seen on radar because the planes are picked way before by other station which will not "release" them until out of reception; the RadarView and other apps are clearly showing I am receiving all planes in my area from a max of 184 NM to local azimuth.
I know is quite frustrating not being able to boast yourself to the friends with your radar station, and maybe will feel better that the major contribution then is in increasing MLAT coverage.

Maybe still FR24 can do something to please these poor guys (including myself) ?

[peterhr]

I would say that the definitive answer was given by FR24support and not explained by the antenna polarization. My radar station (a F-) is rarely seen on radar because the planes are picked way before by other station which will not "release" them until out of reception; the RadarView and other apps are clearly showing I am receiving all planes in my area from a max of 184 NM to local azimuth.
I know is quite frustrating not being able to boast yourself to the friends with your radar station, and maybe will feel better that the major contribution then is in increasing MLAT coverage.

Maybe still FR24 can do something to please these poor guys (including myself) ?

I don't believe the person who first raised the issue was trying to 'boast' - he was just looking for a reason so he could explain it to others.

Same as people who have paid their own money for equipment to set up a 'T-' radar who then don't see it used on FR24 (due to nearby F- radars are not looking for 'bonus points' when they ask about it ... they are just wondering if their time and money expense spent was worth the bother.

[byteworks]

Of course I understood BrianG's reason to ask, I only showed yet another reason.

Yet is very interesting to analyze FR24 algorithm uses to select the follower station. In my case even the plane is near my location, under another station follow back on track, is usually picked by another station way at distance in front (all F-), which led to believe that followers are selected preferring big steps, from the ones next on track or maybe even the ones that are on the predefined known route - really don't know.

[Jarod]

My feeder is at a science museum in Wilson, NC. I'm still trying to figure out how an aircraft flying directly over me shows a different radar (I'm F-KGSB1) and yet one 70-80 miles away does show me as the radar. When I get this fully set-up at the museum (displayed on a large monitor) and an subsequent article printed in the local newspaper I need to be able to explain this.

How the radar works is really interesting. I mostly pick up aircraft on the runways and taxi-ing. Because i can literally look down on the runways at Schiphol.
But overhead? I am never being used my FR24. :-)

I would say that the definitive answer was given by FR24support and not explained by the antenna polarization. My radar station (a F-) is rarely seen on radar because the planes are picked way before by other station which will not "release" them until out of reception; the RadarView and other apps are clearly showing I am receiving all planes in my area from a max of 184 NM to local azimuth.
I know is quite frustrating not being able to boast yourself to the friends with your radar station, and maybe will feel better that the major contribution then is in increasing MLAT coverage.

Maybe still FR24 can do something to please these poor guys (including myself) ?

MLAT was the sole reason i applied to have a receiver installed. Instead of I i think we should think WE.


I been doing some tests with RTLSDR + ADS# at high locations.
http://x264.nl/schiphol_tower_2.jpg 90 meters off the ground, looking down on runway 18R/36L at Schiphol. As this is line of sight, the radar was chosen very often.
http://x264.nl/hello.png 110 meters off the ground, a bit futher away from schiphol, still looking down on the runways, the radar was almost never selected.
And for both goes, none are ever selected for overpassing aircraft, unless, as explained in this thread, the signals turn out to be directed to the antennas.

I did however set the Filter on the website, following the radars. When your reception is good enough, your receiver will "fill in the blanks" for where other receivers aren't optimal.
I have a few fixed points in Belgium and Germany where my receiver is almost always selected. It would be awesome to see a world map with all receivers and their reception range.
Or some average and see where the holes are, but my guess is that FR24 are doing this internally.

Yup interesting stuff, ADS-B.

Edit: A coverage chart somethig like this: http://radio-tv-nederland.nl/dvbt/DV...hresholdGE.jpg
But then for ADS-B receivers...

[SoCalBrian]

FR24 already have two F- boxes in San Diego County.
F-KNKX1
F-KNKX2

Mike,
there is no F- feeders in San Diego, USA. Try to get a few there.

[HighlyCool]

I been doing some tests with RTLSDR + ADS# at high locations.
http://x264.nl/schiphol_tower_2.jpg 90 meters off the ground, looking down on runway 18R/36L at Schiphol. As this is line of sight, the radar was chosen very often.
http://x264.nl/hello.png 110 meters off the ground, a bit futher away from schiphol, still looking down on the runways, the radar was almost never selected.
And for both goes, none are ever selected for overpassing aircraft, unless, as explained in this thread, the signals turn out to be directed to the antennas.

The difference between the two sites and seeing taxiing and arriving/departing aircraft could very well be antenna related. A vertical antenna receives or radiates out of the sides. It is a ground-plane design and the lobes off off the sides of the antenna look like below


This means anything under the level of the bottom of the antenna is in a 'shadow' of the pattern and thusly will not be received very well if the signal is not reflected off of the ground or does not present itself to the side of the antenna.

The ADS-B antennas on the aircraft are mounted on the underside of the plane. So, this is again an issue where the vertical antenna is mounted in an upside down position. That means the signal will look like the above image, but upside down. If you put the images side by side with the 0 (zero) as the level plane, you will see a 'null' area where the reception is not good between the two antennas. In turn the site that is farther away and higher will have more of a shadow of the airport and would need a stronger signal to capture the transmission from the aircraft.

SO, when the closer receiver sees the aircraft on the ground, the feed is taken from that site. Once the aircraft departs that same site will hold the feed. Being a T-feeder (dongle), once an F-feeder (dedicated box) that sends more frequent data to the server than the T-feeder does the server sees the aircraft it will switch to using the feed from the F-feeder. The F-feeders also have a lot better sensitivity to the ADS-B signals as they are tuned specifically for that frequency and also have pre-amps associated with them to boost the signal, so they will usually have a greater receive distance than the dongles.

As for overhead, the same thing goes. If the aircraft is directly over the antenna, the tips of both antennas are pointing at each other and the aircraft and the feeder station in that 'null' for their antennas. There will still be some signal, but it may not be very strong and may be noisy or packets get corrupted by other aircraft that have a stronger signal to your system.

Bottom line, there are a lot of considerations on antenna placement, feed line lengths and signal loss, signal capture, processing speed, etc. that go into who gets listed. The F-feeders being dedicated receivers and processors for ADS-B are usually going to win out over the T-feeders given a proper installation and good location.

Hope my rambling makes sense.

[Oblivian]

They are GPS time synced too.. so the datapackets are often a few ms faster than any NTP tagged one

[vk2tv]

The difference between the two sites and seeing taxiing and arriving/departing aircraft could very well be antenna related. A vertical antenna receives or radiates out of the sides. It is a ground-plane design and the lobes off off the sides of the antenna look like below


This means anything under the level of the bottom of the antenna is in a 'shadow' of the pattern and thusly will not be received very well if the signal is not reflected off of the ground or does not present itself to the side of the antenna.

The ADS-B antennas on the aircraft are mounted on the underside of the plane. So, this is again an issue where the vertical antenna is mounted in an upside down position. That means the signal will look like the above image, but upside down. If you put the images side by side with the 0 (zero) as the level plane, you will see a 'null' area where the reception is not good between the two antennas. In turn the site that is farther away and higher will have more of a shadow of the airport and would need a stronger signal to capture the transmission from the aircraft.

SO, when the closer receiver sees the aircraft on the ground, the feed is taken from that site. Once the aircraft departs that same site will hold the feed. Being a T-feeder (dongle), once an F-feeder (dedicated box) that sends more frequent data to the server than the T-feeder does the server sees the aircraft it will switch to using the feed from the F-feeder. The F-feeders also have a lot better sensitivity to the ADS-B signals as they are tuned specifically for that frequency and also have pre-amps associated with them to boost the signal, so they will usually have a greater receive distance than the dongles.

As for overhead, the same thing goes. If the aircraft is directly over the antenna, the tips of both antennas are pointing at each other and the aircraft and the feeder station in that 'null' for their antennas. There will still be some signal, but it may not be very strong and may be noisy or packets get corrupted by other aircraft that have a stronger signal to your system.

Bottom line, there are a lot of considerations on antenna placement, feed line lengths and signal loss, signal capture, processing speed, etc. that go into who gets listed. The F-feeders being dedicated receivers and processors for ADS-B are usually going to win out over the T-feeders given a proper installation and good location.

Hope my rambling makes sense.

All,

Without intending to offend anyone, polar diagrams are, for convenience, sometimes shown like the example shown here, with no pattern displayed for angles below zero elevation. In reality, for simple antennas such as the Ground Plane (GP) and Coaxial Dipole, the pattern below zero elevation is a mirror image of what's shown above zero. The image for a GP shows what I mean. GP_image.pngGP_Polar.png The important point is that the pattern does not vanish below zero elevation. Professional Polar Plots usually show the entire pattern because when one gets into gain antennas such as collinears the above and below zero patterns MAY be different.

Ray vk2tv
Retired communication technician

[HighlyCool]

All,

Without intending to offend anyone, polar diagrams are, for convenience, sometimes shown like the example shown here, with no pattern displayed for angles below zero elevation. In reality, for simple antennas such as the Ground Plane (GP) and Coaxial Dipole, the pattern below zero elevation is a mirror image of what's shown above zero. The image for a GP shows what I mean. [ATTACH=CONFIG]3238[/ATTACH][ATTACH=CONFIG]3239[/ATTACH] The important point is that the pattern does not vanish below zero elevation. Professional Polar Plots usually show the entire pattern because when one gets into gain antennas such as collinears the above and below zero patterns MAY be different.

Ray vk2tv
Retired communication technician

I agree it does not go to zero, but it does have nulls that exist and cause a difference in signal stregth. That was my whole point. In the end, the F-feeders are much faster and better a doing the ADS-B data than a DVB-T dongle and a computer. But antenna patterns and gains play into the factor significantly.

[vk2tv]

I agree it does not go to zero, but it does have nulls that exist and cause a difference in signal stregth. That was my whole point. In the end, the F-feeders are much faster and better a doing the ADS-B data than a DVB-T dongle and a computer. But antenna patterns and gains play into the factor significantly.

Hi Highlycool

I provided the explanation for those amongst us who might have taken the initial polar plot literally, so that they may better understand the real situation.

Ray

[peterhr]

They are GPS time synced too.. so the datapackets are often a few ms faster than any NTP tagged one

'Grand Prix Selected'?

I agree it does not go to zero, but it does have nulls that exist and cause a difference in signal stregth. That was my whole point. In the end, the F-feeders are much faster and better a doing the ADS-B data than a DVB-T dongle and a computer. But antenna patterns and gains play into the factor significantly.

The F- feeders do have a very sensitive tuner and a fast decoder, but the F- feeders also seem to send data packets in at a much higher rate than [the one packet per second or so used by] T- feeders, so F- feeder data is most often the most recent data for a particular plane received by the servers.

[Oblivian]

'Grand Prix Selected'?

Struggling to tell if you being funny or serious..?

PCs using NTP/internet time sync are already offset from UTC by up to 3-6ms before they even bundle a ready-to-upload data packet together with the time they received each ADSB packet. The F- units with direct GPS receivers are getting live time data and processing it internally simultaneously as its decoded and uploaded. In that sense I can see it's more than likely going to have more up to date packet upload time..

[peterhr]

Struggling to tell if you being funny or serious..?

PCs using NTP/internet time sync are already offset from UTC by up to 3-6ms before they even bundle a ready-to-upload data packet together with the time they received each ADSB packet. The F- units with direct GPS receivers are getting live time data and processing it internally simultaneously as its decoded and uploaded. In that sense I can see it's more than likely going to have more up to date packet upload time..

Part of NTP protocol includes a correction factor where by
client sends a request to the server, server responds with the correct time
Client takes that time from the server and adds to it half the round trip time that it took between sending request and receiving response (this does assume that the network delays are symmetric, the NTP request is timestamped by the client making the request - time stamp is returned as part of the NTP data packet), there is also a second correction factor that is applied which takes into account the time difference between the server receiving the request and the server responding.

With this correction it's possible that the client time could be slightly wrong and indeed the correction calculation could put it slightly in advance of the time server as well as behind.

See http://www.meinbergglobal.com/englis...ntp-packet.htm
and http://en.wikipedia.org/wiki/Network_Time_Protocol